diff --git "a/checkpoint-19500/trainer_state.json" "b/checkpoint-19500/trainer_state.json" new file mode 100644--- /dev/null +++ "b/checkpoint-19500/trainer_state.json" @@ -0,0 +1,136533 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 15.575079872204473, + "eval_steps": 500, + "global_step": 19500, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0007987220447284345, + "grad_norm": 12.75, + "learning_rate": 0.0005, + "loss": 3.6897, + "step": 1 + }, + { + "epoch": 0.001597444089456869, + "grad_norm": 131.0, + "learning_rate": 0.0005, + "loss": 22.6333, + "step": 2 + }, + { + "epoch": 0.0023961661341853034, + "grad_norm": 15.0625, + "learning_rate": 0.0005, + "loss": 3.0277, + "step": 3 + }, + { + "epoch": 0.003194888178913738, + "grad_norm": 27.625, + "learning_rate": 0.0005, + "loss": 4.58, + "step": 4 + }, + { + "epoch": 0.003993610223642172, + "grad_norm": 1.8203125, + "learning_rate": 0.0005, + "loss": 1.9082, + "step": 5 + }, + { + "epoch": 0.004792332268370607, + "grad_norm": 5.96875, + "learning_rate": 0.0005, + "loss": 2.2409, + "step": 6 + }, + { + "epoch": 0.005591054313099041, + "grad_norm": 2.3125, + "learning_rate": 0.0005, + "loss": 1.9377, + "step": 7 + }, + { + "epoch": 0.006389776357827476, + "grad_norm": 4.71875, + "learning_rate": 0.0005, + "loss": 1.9207, + "step": 8 + }, + { + "epoch": 0.00718849840255591, + "grad_norm": 0.484375, + "learning_rate": 0.0005, + "loss": 1.7547, + "step": 9 + }, + { + "epoch": 0.007987220447284345, + "grad_norm": 2.15625, + "learning_rate": 0.0005, + "loss": 1.7161, + "step": 10 + }, + { + "epoch": 0.00878594249201278, + "grad_norm": 3.96875, + "learning_rate": 0.0005, + "loss": 1.7638, + "step": 11 + }, + { + "epoch": 0.009584664536741214, + "grad_norm": 2.109375, + "learning_rate": 0.0005, + "loss": 1.6822, + "step": 12 + }, + { + "epoch": 0.010383386581469648, + "grad_norm": 5.78125, + "learning_rate": 0.0005, + "loss": 1.7121, + "step": 13 + }, + { + "epoch": 0.011182108626198083, + "grad_norm": 5.71875, + "learning_rate": 0.0005, + "loss": 1.6905, + "step": 14 + }, + { + "epoch": 0.011980830670926517, + "grad_norm": 0.5234375, + "learning_rate": 0.0005, + "loss": 1.5379, + "step": 15 + }, + { + "epoch": 0.012779552715654952, + "grad_norm": 4.8125, + "learning_rate": 0.0005, + "loss": 1.6309, + "step": 16 + }, + { + "epoch": 0.013578274760383386, + "grad_norm": 4.75, + "learning_rate": 0.0005, + "loss": 1.6553, + "step": 17 + }, + { + "epoch": 0.01437699680511182, + "grad_norm": 1.390625, + "learning_rate": 0.0005, + "loss": 1.525, + "step": 18 + }, + { + "epoch": 0.015175718849840255, + "grad_norm": 4.28125, + "learning_rate": 0.0005, + "loss": 1.5676, + "step": 19 + }, + { + "epoch": 0.01597444089456869, + "grad_norm": 4.90625, + "learning_rate": 0.0005, + "loss": 1.6095, + "step": 20 + }, + { + "epoch": 0.016773162939297124, + "grad_norm": 2.78125, + "learning_rate": 0.0005, + "loss": 1.5417, + "step": 21 + }, + { + "epoch": 0.01757188498402556, + "grad_norm": 1.3515625, + "learning_rate": 0.0005, + "loss": 1.4745, + "step": 22 + }, + { + "epoch": 0.018370607028753993, + "grad_norm": 1.421875, + "learning_rate": 0.0005, + "loss": 1.4543, + "step": 23 + }, + { + "epoch": 0.019169329073482427, + "grad_norm": 0.29296875, + "learning_rate": 0.0005, + "loss": 1.4246, + "step": 24 + }, + { + "epoch": 0.019968051118210862, + "grad_norm": 0.78125, + "learning_rate": 0.0005, + "loss": 1.4123, + "step": 25 + }, + { + "epoch": 0.020766773162939296, + "grad_norm": 0.29296875, + "learning_rate": 0.0005, + "loss": 1.4111, + "step": 26 + }, + { + "epoch": 0.02156549520766773, + "grad_norm": 0.365234375, + "learning_rate": 0.0005, + "loss": 1.3884, + "step": 27 + }, + { + "epoch": 0.022364217252396165, + "grad_norm": 0.259765625, + "learning_rate": 0.0005, + "loss": 1.3844, + "step": 28 + }, + { + "epoch": 0.0231629392971246, + "grad_norm": 0.1708984375, + "learning_rate": 0.0005, + "loss": 1.376, + "step": 29 + }, + { + "epoch": 0.023961661341853034, + "grad_norm": 0.2138671875, + "learning_rate": 0.0005, + "loss": 1.3648, + "step": 30 + }, + { + "epoch": 0.02476038338658147, + "grad_norm": 0.1572265625, + "learning_rate": 0.0005, + "loss": 1.3598, + "step": 31 + }, + { + "epoch": 0.025559105431309903, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.3482, + "step": 32 + }, + { + "epoch": 0.026357827476038338, + "grad_norm": 0.181640625, + "learning_rate": 0.0005, + "loss": 1.3438, + "step": 33 + }, + { + "epoch": 0.027156549520766772, + "grad_norm": 0.1826171875, + "learning_rate": 0.0005, + "loss": 1.3365, + "step": 34 + }, + { + "epoch": 0.027955271565495207, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.3207, + "step": 35 + }, + { + "epoch": 0.02875399361022364, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.3332, + "step": 36 + }, + { + "epoch": 0.029552715654952075, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.3288, + "step": 37 + }, + { + "epoch": 0.03035143769968051, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.3114, + "step": 38 + }, + { + "epoch": 0.031150159744408944, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.295, + "step": 39 + }, + { + "epoch": 0.03194888178913738, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.3012, + "step": 40 + }, + { + "epoch": 0.03274760383386582, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.2937, + "step": 41 + }, + { + "epoch": 0.03354632587859425, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.2868, + "step": 42 + }, + { + "epoch": 0.034345047923322686, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.2911, + "step": 43 + }, + { + "epoch": 0.03514376996805112, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.2929, + "step": 44 + }, + { + "epoch": 0.035942492012779555, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.2667, + "step": 45 + }, + { + "epoch": 0.036741214057507986, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.2734, + "step": 46 + }, + { + "epoch": 0.037539936102236424, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.2674, + "step": 47 + }, + { + "epoch": 0.038338658146964855, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.2775, + "step": 48 + }, + { + "epoch": 0.03913738019169329, + "grad_norm": 0.1640625, + "learning_rate": 0.0005, + "loss": 1.2771, + "step": 49 + }, + { + "epoch": 0.039936102236421724, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.2572, + "step": 50 + }, + { + "epoch": 0.04073482428115016, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.2669, + "step": 51 + }, + { + "epoch": 0.04153354632587859, + "grad_norm": 0.16015625, + "learning_rate": 0.0005, + "loss": 1.2655, + "step": 52 + }, + { + "epoch": 0.04233226837060703, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.2605, + "step": 53 + }, + { + "epoch": 0.04313099041533546, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.2752, + "step": 54 + }, + { + "epoch": 0.0439297124600639, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.2713, + "step": 55 + }, + { + "epoch": 0.04472843450479233, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.2517, + "step": 56 + }, + { + "epoch": 0.04552715654952077, + "grad_norm": 0.1787109375, + "learning_rate": 0.0005, + "loss": 1.2529, + "step": 57 + }, + { + "epoch": 0.0463258785942492, + "grad_norm": 0.1796875, + "learning_rate": 0.0005, + "loss": 1.255, + "step": 58 + }, + { + "epoch": 0.04712460063897764, + "grad_norm": 0.17578125, + "learning_rate": 0.0005, + "loss": 1.2675, + "step": 59 + }, + { + "epoch": 0.04792332268370607, + "grad_norm": 0.251953125, + "learning_rate": 0.0005, + "loss": 1.2442, + "step": 60 + }, + { + "epoch": 0.048722044728434506, + "grad_norm": 0.24609375, + "learning_rate": 0.0005, + "loss": 1.2543, + "step": 61 + }, + { + "epoch": 0.04952076677316294, + "grad_norm": 0.162109375, + "learning_rate": 0.0005, + "loss": 1.248, + "step": 62 + }, + { + "epoch": 0.050319488817891375, + "grad_norm": 0.2021484375, + "learning_rate": 0.0005, + "loss": 1.2369, + "step": 63 + }, + { + "epoch": 0.051118210862619806, + "grad_norm": 0.2021484375, + "learning_rate": 0.0005, + "loss": 1.2378, + "step": 64 + }, + { + "epoch": 0.051916932907348244, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.2386, + "step": 65 + }, + { + "epoch": 0.052715654952076675, + "grad_norm": 0.1953125, + "learning_rate": 0.0005, + "loss": 1.2314, + "step": 66 + }, + { + "epoch": 0.05351437699680511, + "grad_norm": 0.1572265625, + "learning_rate": 0.0005, + "loss": 1.2345, + "step": 67 + }, + { + "epoch": 0.054313099041533544, + "grad_norm": 0.1708984375, + "learning_rate": 0.0005, + "loss": 1.2247, + "step": 68 + }, + { + "epoch": 0.05511182108626198, + "grad_norm": 0.205078125, + "learning_rate": 0.0005, + "loss": 1.2356, + "step": 69 + }, + { + "epoch": 0.05591054313099041, + "grad_norm": 0.1572265625, + "learning_rate": 0.0005, + "loss": 1.2471, + "step": 70 + }, + { + "epoch": 0.05670926517571885, + "grad_norm": 0.21484375, + "learning_rate": 0.0005, + "loss": 1.2299, + "step": 71 + }, + { + "epoch": 0.05750798722044728, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.2284, + "step": 72 + }, + { + "epoch": 0.05830670926517572, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.2306, + "step": 73 + }, + { + "epoch": 0.05910543130990415, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.2346, + "step": 74 + }, + { + "epoch": 0.05990415335463259, + "grad_norm": 0.1689453125, + "learning_rate": 0.0005, + "loss": 1.2334, + "step": 75 + }, + { + "epoch": 0.06070287539936102, + "grad_norm": 0.2109375, + "learning_rate": 0.0005, + "loss": 1.2322, + "step": 76 + }, + { + "epoch": 0.06150159744408946, + "grad_norm": 0.166015625, + "learning_rate": 0.0005, + "loss": 1.2159, + "step": 77 + }, + { + "epoch": 0.06230031948881789, + "grad_norm": 0.181640625, + "learning_rate": 0.0005, + "loss": 1.2154, + "step": 78 + }, + { + "epoch": 0.06309904153354633, + "grad_norm": 0.25390625, + "learning_rate": 0.0005, + "loss": 1.2258, + "step": 79 + }, + { + "epoch": 0.06389776357827476, + "grad_norm": 0.263671875, + "learning_rate": 0.0005, + "loss": 1.2462, + "step": 80 + }, + { + "epoch": 0.06469648562300319, + "grad_norm": 0.2001953125, + "learning_rate": 0.0005, + "loss": 1.2327, + "step": 81 + }, + { + "epoch": 0.06549520766773163, + "grad_norm": 0.30859375, + "learning_rate": 0.0005, + "loss": 1.2313, + "step": 82 + }, + { + "epoch": 0.06629392971246006, + "grad_norm": 0.390625, + "learning_rate": 0.0005, + "loss": 1.225, + "step": 83 + }, + { + "epoch": 0.0670926517571885, + "grad_norm": 0.208984375, + "learning_rate": 0.0005, + "loss": 1.2258, + "step": 84 + }, + { + "epoch": 0.06789137380191693, + "grad_norm": 0.208984375, + "learning_rate": 0.0005, + "loss": 1.2257, + "step": 85 + }, + { + "epoch": 0.06869009584664537, + "grad_norm": 0.173828125, + "learning_rate": 0.0005, + "loss": 1.2052, + "step": 86 + }, + { + "epoch": 0.0694888178913738, + "grad_norm": 0.2080078125, + "learning_rate": 0.0005, + "loss": 1.2235, + "step": 87 + }, + { + "epoch": 0.07028753993610223, + "grad_norm": 0.1982421875, + "learning_rate": 0.0005, + "loss": 1.2292, + "step": 88 + }, + { + "epoch": 0.07108626198083066, + "grad_norm": 0.1787109375, + "learning_rate": 0.0005, + "loss": 1.2237, + "step": 89 + }, + { + "epoch": 0.07188498402555911, + "grad_norm": 0.2294921875, + "learning_rate": 0.0005, + "loss": 1.218, + "step": 90 + }, + { + "epoch": 0.07268370607028754, + "grad_norm": 0.1953125, + "learning_rate": 0.0005, + "loss": 1.2208, + "step": 91 + }, + { + "epoch": 0.07348242811501597, + "grad_norm": 0.181640625, + "learning_rate": 0.0005, + "loss": 1.2202, + "step": 92 + }, + { + "epoch": 0.0742811501597444, + "grad_norm": 0.21875, + "learning_rate": 0.0005, + "loss": 1.2122, + "step": 93 + }, + { + "epoch": 0.07507987220447285, + "grad_norm": 0.1962890625, + "learning_rate": 0.0005, + "loss": 1.2253, + "step": 94 + }, + { + "epoch": 0.07587859424920128, + "grad_norm": 0.1875, + "learning_rate": 0.0005, + "loss": 1.2127, + "step": 95 + }, + { + "epoch": 0.07667731629392971, + "grad_norm": 0.1767578125, + "learning_rate": 0.0005, + "loss": 1.2082, + "step": 96 + }, + { + "epoch": 0.07747603833865814, + "grad_norm": 0.17578125, + "learning_rate": 0.0005, + "loss": 1.2104, + "step": 97 + }, + { + "epoch": 0.07827476038338659, + "grad_norm": 0.224609375, + "learning_rate": 0.0005, + "loss": 1.2082, + "step": 98 + }, + { + "epoch": 0.07907348242811502, + "grad_norm": 0.22265625, + "learning_rate": 0.0005, + "loss": 1.1995, + "step": 99 + }, + { + "epoch": 0.07987220447284345, + "grad_norm": 0.197265625, + "learning_rate": 0.0005, + "loss": 1.2066, + "step": 100 + }, + { + "epoch": 0.08067092651757188, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.2039, + "step": 101 + }, + { + "epoch": 0.08146964856230032, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.2018, + "step": 102 + }, + { + "epoch": 0.08226837060702875, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.1904, + "step": 103 + }, + { + "epoch": 0.08306709265175719, + "grad_norm": 0.1953125, + "learning_rate": 0.0005, + "loss": 1.2146, + "step": 104 + }, + { + "epoch": 0.08386581469648563, + "grad_norm": 0.1943359375, + "learning_rate": 0.0005, + "loss": 1.1988, + "step": 105 + }, + { + "epoch": 0.08466453674121406, + "grad_norm": 0.3515625, + "learning_rate": 0.0005, + "loss": 1.2121, + "step": 106 + }, + { + "epoch": 0.08546325878594249, + "grad_norm": 0.55078125, + "learning_rate": 0.0005, + "loss": 1.2155, + "step": 107 + }, + { + "epoch": 0.08626198083067092, + "grad_norm": 0.439453125, + "learning_rate": 0.0005, + "loss": 1.2159, + "step": 108 + }, + { + "epoch": 0.08706070287539937, + "grad_norm": 0.2392578125, + "learning_rate": 0.0005, + "loss": 1.2118, + "step": 109 + }, + { + "epoch": 0.0878594249201278, + "grad_norm": 0.310546875, + "learning_rate": 0.0005, + "loss": 1.2225, + "step": 110 + }, + { + "epoch": 0.08865814696485623, + "grad_norm": 0.2001953125, + "learning_rate": 0.0005, + "loss": 1.2074, + "step": 111 + }, + { + "epoch": 0.08945686900958466, + "grad_norm": 0.212890625, + "learning_rate": 0.0005, + "loss": 1.1986, + "step": 112 + }, + { + "epoch": 0.0902555910543131, + "grad_norm": 0.1748046875, + "learning_rate": 0.0005, + "loss": 1.2043, + "step": 113 + }, + { + "epoch": 0.09105431309904154, + "grad_norm": 0.18359375, + "learning_rate": 0.0005, + "loss": 1.1918, + "step": 114 + }, + { + "epoch": 0.09185303514376997, + "grad_norm": 0.169921875, + "learning_rate": 0.0005, + "loss": 1.2079, + "step": 115 + }, + { + "epoch": 0.0926517571884984, + "grad_norm": 0.16796875, + "learning_rate": 0.0005, + "loss": 1.1924, + "step": 116 + }, + { + "epoch": 0.09345047923322684, + "grad_norm": 0.16015625, + "learning_rate": 0.0005, + "loss": 1.1989, + "step": 117 + }, + { + "epoch": 0.09424920127795527, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.1899, + "step": 118 + }, + { + "epoch": 0.0950479233226837, + "grad_norm": 0.173828125, + "learning_rate": 0.0005, + "loss": 1.2007, + "step": 119 + }, + { + "epoch": 0.09584664536741214, + "grad_norm": 0.193359375, + "learning_rate": 0.0005, + "loss": 1.2056, + "step": 120 + }, + { + "epoch": 0.09664536741214058, + "grad_norm": 0.181640625, + "learning_rate": 0.0005, + "loss": 1.1956, + "step": 121 + }, + { + "epoch": 0.09744408945686901, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.1909, + "step": 122 + }, + { + "epoch": 0.09824281150159744, + "grad_norm": 0.17578125, + "learning_rate": 0.0005, + "loss": 1.191, + "step": 123 + }, + { + "epoch": 0.09904153354632587, + "grad_norm": 0.173828125, + "learning_rate": 0.0005, + "loss": 1.185, + "step": 124 + }, + { + "epoch": 0.09984025559105432, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.1787, + "step": 125 + }, + { + "epoch": 0.10063897763578275, + "grad_norm": 0.1943359375, + "learning_rate": 0.0005, + "loss": 1.1839, + "step": 126 + }, + { + "epoch": 0.10143769968051118, + "grad_norm": 0.1865234375, + "learning_rate": 0.0005, + "loss": 1.1906, + "step": 127 + }, + { + "epoch": 0.10223642172523961, + "grad_norm": 0.1669921875, + "learning_rate": 0.0005, + "loss": 1.1831, + "step": 128 + }, + { + "epoch": 0.10303514376996806, + "grad_norm": 0.1875, + "learning_rate": 0.0005, + "loss": 1.1883, + "step": 129 + }, + { + "epoch": 0.10383386581469649, + "grad_norm": 0.25, + "learning_rate": 0.0005, + "loss": 1.188, + "step": 130 + }, + { + "epoch": 0.10463258785942492, + "grad_norm": 0.46484375, + "learning_rate": 0.0005, + "loss": 1.1881, + "step": 131 + }, + { + "epoch": 0.10543130990415335, + "grad_norm": 1.1328125, + "learning_rate": 0.0005, + "loss": 1.2194, + "step": 132 + }, + { + "epoch": 0.1062300319488818, + "grad_norm": 0.78515625, + "learning_rate": 0.0005, + "loss": 1.2307, + "step": 133 + }, + { + "epoch": 0.10702875399361023, + "grad_norm": 0.328125, + "learning_rate": 0.0005, + "loss": 1.2056, + "step": 134 + }, + { + "epoch": 0.10782747603833866, + "grad_norm": 0.4765625, + "learning_rate": 0.0005, + "loss": 1.2139, + "step": 135 + }, + { + "epoch": 0.10862619808306709, + "grad_norm": 0.341796875, + "learning_rate": 0.0005, + "loss": 1.2103, + "step": 136 + }, + { + "epoch": 0.10942492012779553, + "grad_norm": 0.2197265625, + "learning_rate": 0.0005, + "loss": 1.21, + "step": 137 + }, + { + "epoch": 0.11022364217252396, + "grad_norm": 0.3203125, + "learning_rate": 0.0005, + "loss": 1.2066, + "step": 138 + }, + { + "epoch": 0.1110223642172524, + "grad_norm": 0.201171875, + "learning_rate": 0.0005, + "loss": 1.1956, + "step": 139 + }, + { + "epoch": 0.11182108626198083, + "grad_norm": 0.2392578125, + "learning_rate": 0.0005, + "loss": 1.2111, + "step": 140 + }, + { + "epoch": 0.11261980830670927, + "grad_norm": 0.1669921875, + "learning_rate": 0.0005, + "loss": 1.1994, + "step": 141 + }, + { + "epoch": 0.1134185303514377, + "grad_norm": 0.2080078125, + "learning_rate": 0.0005, + "loss": 1.2043, + "step": 142 + }, + { + "epoch": 0.11421725239616613, + "grad_norm": 0.1611328125, + "learning_rate": 0.0005, + "loss": 1.1971, + "step": 143 + }, + { + "epoch": 0.11501597444089456, + "grad_norm": 0.1806640625, + "learning_rate": 0.0005, + "loss": 1.198, + "step": 144 + }, + { + "epoch": 0.11581469648562301, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.1906, + "step": 145 + }, + { + "epoch": 0.11661341853035144, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.2011, + "step": 146 + }, + { + "epoch": 0.11741214057507987, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.1981, + "step": 147 + }, + { + "epoch": 0.1182108626198083, + "grad_norm": 0.1533203125, + "learning_rate": 0.0005, + "loss": 1.1947, + "step": 148 + }, + { + "epoch": 0.11900958466453675, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.1906, + "step": 149 + }, + { + "epoch": 0.11980830670926518, + "grad_norm": 0.16796875, + "learning_rate": 0.0005, + "loss": 1.1868, + "step": 150 + }, + { + "epoch": 0.12060702875399361, + "grad_norm": 0.177734375, + "learning_rate": 0.0005, + "loss": 1.1925, + "step": 151 + }, + { + "epoch": 0.12140575079872204, + "grad_norm": 0.162109375, + "learning_rate": 0.0005, + "loss": 1.1814, + "step": 152 + }, + { + "epoch": 0.12220447284345048, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.1895, + "step": 153 + }, + { + "epoch": 0.12300319488817892, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.1736, + "step": 154 + }, + { + "epoch": 0.12380191693290735, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.1839, + "step": 155 + }, + { + "epoch": 0.12460063897763578, + "grad_norm": 0.1728515625, + "learning_rate": 0.0005, + "loss": 1.1873, + "step": 156 + }, + { + "epoch": 0.1253993610223642, + "grad_norm": 0.166015625, + "learning_rate": 0.0005, + "loss": 1.1764, + "step": 157 + }, + { + "epoch": 0.12619808306709265, + "grad_norm": 0.1845703125, + "learning_rate": 0.0005, + "loss": 1.1907, + "step": 158 + }, + { + "epoch": 0.1269968051118211, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.1753, + "step": 159 + }, + { + "epoch": 0.12779552715654952, + "grad_norm": 0.169921875, + "learning_rate": 0.0005, + "loss": 1.1772, + "step": 160 + }, + { + "epoch": 0.12859424920127796, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.1759, + "step": 161 + }, + { + "epoch": 0.12939297124600638, + "grad_norm": 0.166015625, + "learning_rate": 0.0005, + "loss": 1.1729, + "step": 162 + }, + { + "epoch": 0.13019169329073482, + "grad_norm": 0.1611328125, + "learning_rate": 0.0005, + "loss": 1.1797, + "step": 163 + }, + { + "epoch": 0.13099041533546327, + "grad_norm": 0.1787109375, + "learning_rate": 0.0005, + "loss": 1.1736, + "step": 164 + }, + { + "epoch": 0.13178913738019168, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.1822, + "step": 165 + }, + { + "epoch": 0.13258785942492013, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.1573, + "step": 166 + }, + { + "epoch": 0.13338658146964857, + "grad_norm": 0.1806640625, + "learning_rate": 0.0005, + "loss": 1.1726, + "step": 167 + }, + { + "epoch": 0.134185303514377, + "grad_norm": 0.1689453125, + "learning_rate": 0.0005, + "loss": 1.1771, + "step": 168 + }, + { + "epoch": 0.13498402555910544, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.1711, + "step": 169 + }, + { + "epoch": 0.13578274760383385, + "grad_norm": 0.1533203125, + "learning_rate": 0.0005, + "loss": 1.1654, + "step": 170 + }, + { + "epoch": 0.1365814696485623, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.1681, + "step": 171 + }, + { + "epoch": 0.13738019169329074, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.1661, + "step": 172 + }, + { + "epoch": 0.13817891373801916, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.1708, + "step": 173 + }, + { + "epoch": 0.1389776357827476, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.1637, + "step": 174 + }, + { + "epoch": 0.13977635782747605, + "grad_norm": 0.169921875, + "learning_rate": 0.0005, + "loss": 1.1654, + "step": 175 + }, + { + "epoch": 0.14057507987220447, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.1703, + "step": 176 + }, + { + "epoch": 0.1413738019169329, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.1583, + "step": 177 + }, + { + "epoch": 0.14217252396166133, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.1642, + "step": 178 + }, + { + "epoch": 0.14297124600638977, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.1709, + "step": 179 + }, + { + "epoch": 0.14376996805111822, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.1582, + "step": 180 + }, + { + "epoch": 0.14456869009584664, + "grad_norm": 0.1689453125, + "learning_rate": 0.0005, + "loss": 1.1696, + "step": 181 + }, + { + "epoch": 0.14536741214057508, + "grad_norm": 0.1689453125, + "learning_rate": 0.0005, + "loss": 1.1603, + "step": 182 + }, + { + "epoch": 0.14616613418530353, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.1593, + "step": 183 + }, + { + "epoch": 0.14696485623003194, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.1728, + "step": 184 + }, + { + "epoch": 0.1477635782747604, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.1723, + "step": 185 + }, + { + "epoch": 0.1485623003194888, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.1687, + "step": 186 + }, + { + "epoch": 0.14936102236421725, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.1488, + "step": 187 + }, + { + "epoch": 0.1501597444089457, + "grad_norm": 0.197265625, + "learning_rate": 0.0005, + "loss": 1.161, + "step": 188 + }, + { + "epoch": 0.1509584664536741, + "grad_norm": 0.2197265625, + "learning_rate": 0.0005, + "loss": 1.162, + "step": 189 + }, + { + "epoch": 0.15175718849840256, + "grad_norm": 0.265625, + "learning_rate": 0.0005, + "loss": 1.1615, + "step": 190 + }, + { + "epoch": 0.152555910543131, + "grad_norm": 0.48828125, + "learning_rate": 0.0005, + "loss": 1.1742, + "step": 191 + }, + { + "epoch": 0.15335463258785942, + "grad_norm": 1.2734375, + "learning_rate": 0.0005, + "loss": 1.1867, + "step": 192 + }, + { + "epoch": 0.15415335463258786, + "grad_norm": 1.34375, + "learning_rate": 0.0005, + "loss": 1.2171, + "step": 193 + }, + { + "epoch": 0.15495207667731628, + "grad_norm": 0.2177734375, + "learning_rate": 0.0005, + "loss": 1.1803, + "step": 194 + }, + { + "epoch": 0.15575079872204473, + "grad_norm": 0.6015625, + "learning_rate": 0.0005, + "loss": 1.1918, + "step": 195 + }, + { + "epoch": 0.15654952076677317, + "grad_norm": 0.443359375, + "learning_rate": 0.0005, + "loss": 1.1942, + "step": 196 + }, + { + "epoch": 0.1573482428115016, + "grad_norm": 0.376953125, + "learning_rate": 0.0005, + "loss": 1.1905, + "step": 197 + }, + { + "epoch": 0.15814696485623003, + "grad_norm": 0.251953125, + "learning_rate": 0.0005, + "loss": 1.175, + "step": 198 + }, + { + "epoch": 0.15894568690095848, + "grad_norm": 0.244140625, + "learning_rate": 0.0005, + "loss": 1.1769, + "step": 199 + }, + { + "epoch": 0.1597444089456869, + "grad_norm": 0.21484375, + "learning_rate": 0.0005, + "loss": 1.1866, + "step": 200 + }, + { + "epoch": 0.16054313099041534, + "grad_norm": 0.1669921875, + "learning_rate": 0.0005, + "loss": 1.1811, + "step": 201 + }, + { + "epoch": 0.16134185303514376, + "grad_norm": 0.1787109375, + "learning_rate": 0.0005, + "loss": 1.1697, + "step": 202 + }, + { + "epoch": 0.1621405750798722, + "grad_norm": 0.17578125, + "learning_rate": 0.0005, + "loss": 1.1675, + "step": 203 + }, + { + "epoch": 0.16293929712460065, + "grad_norm": 0.193359375, + "learning_rate": 0.0005, + "loss": 1.1612, + "step": 204 + }, + { + "epoch": 0.16373801916932906, + "grad_norm": 0.1572265625, + "learning_rate": 0.0005, + "loss": 1.1648, + "step": 205 + }, + { + "epoch": 0.1645367412140575, + "grad_norm": 0.1708984375, + "learning_rate": 0.0005, + "loss": 1.1723, + "step": 206 + }, + { + "epoch": 0.16533546325878595, + "grad_norm": 0.1669921875, + "learning_rate": 0.0005, + "loss": 1.1696, + "step": 207 + }, + { + "epoch": 0.16613418530351437, + "grad_norm": 0.1611328125, + "learning_rate": 0.0005, + "loss": 1.1602, + "step": 208 + }, + { + "epoch": 0.16693290734824281, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.1654, + "step": 209 + }, + { + "epoch": 0.16773162939297126, + "grad_norm": 0.16796875, + "learning_rate": 0.0005, + "loss": 1.1691, + "step": 210 + }, + { + "epoch": 0.16853035143769968, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.1675, + "step": 211 + }, + { + "epoch": 0.16932907348242812, + "grad_norm": 0.189453125, + "learning_rate": 0.0005, + "loss": 1.1506, + "step": 212 + }, + { + "epoch": 0.17012779552715654, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.1614, + "step": 213 + }, + { + "epoch": 0.17092651757188498, + "grad_norm": 0.1943359375, + "learning_rate": 0.0005, + "loss": 1.1562, + "step": 214 + }, + { + "epoch": 0.17172523961661343, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.1528, + "step": 215 + }, + { + "epoch": 0.17252396166134185, + "grad_norm": 0.166015625, + "learning_rate": 0.0005, + "loss": 1.1544, + "step": 216 + }, + { + "epoch": 0.1733226837060703, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.1583, + "step": 217 + }, + { + "epoch": 0.17412140575079874, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.149, + "step": 218 + }, + { + "epoch": 0.17492012779552715, + "grad_norm": 0.173828125, + "learning_rate": 0.0005, + "loss": 1.151, + "step": 219 + }, + { + "epoch": 0.1757188498402556, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.156, + "step": 220 + }, + { + "epoch": 0.17651757188498401, + "grad_norm": 0.244140625, + "learning_rate": 0.0005, + "loss": 1.1533, + "step": 221 + }, + { + "epoch": 0.17731629392971246, + "grad_norm": 0.16796875, + "learning_rate": 0.0005, + "loss": 1.1595, + "step": 222 + }, + { + "epoch": 0.1781150159744409, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.1603, + "step": 223 + }, + { + "epoch": 0.17891373801916932, + "grad_norm": 0.1953125, + "learning_rate": 0.0005, + "loss": 1.1518, + "step": 224 + }, + { + "epoch": 0.17971246006389777, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.1643, + "step": 225 + }, + { + "epoch": 0.1805111821086262, + "grad_norm": 0.18359375, + "learning_rate": 0.0005, + "loss": 1.146, + "step": 226 + }, + { + "epoch": 0.18130990415335463, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.1517, + "step": 227 + }, + { + "epoch": 0.18210862619808307, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.1503, + "step": 228 + }, + { + "epoch": 0.1829073482428115, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.147, + "step": 229 + }, + { + "epoch": 0.18370607028753994, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.1567, + "step": 230 + }, + { + "epoch": 0.18450479233226838, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.1438, + "step": 231 + }, + { + "epoch": 0.1853035143769968, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.1536, + "step": 232 + }, + { + "epoch": 0.18610223642172524, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.1528, + "step": 233 + }, + { + "epoch": 0.1869009584664537, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.1626, + "step": 234 + }, + { + "epoch": 0.1876996805111821, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.1496, + "step": 235 + }, + { + "epoch": 0.18849840255591055, + "grad_norm": 0.1669921875, + "learning_rate": 0.0005, + "loss": 1.1523, + "step": 236 + }, + { + "epoch": 0.18929712460063897, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.1541, + "step": 237 + }, + { + "epoch": 0.1900958466453674, + "grad_norm": 0.1826171875, + "learning_rate": 0.0005, + "loss": 1.1467, + "step": 238 + }, + { + "epoch": 0.19089456869009586, + "grad_norm": 0.1748046875, + "learning_rate": 0.0005, + "loss": 1.162, + "step": 239 + }, + { + "epoch": 0.19169329073482427, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.1404, + "step": 240 + }, + { + "epoch": 0.19249201277955272, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.1466, + "step": 241 + }, + { + "epoch": 0.19329073482428116, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.1519, + "step": 242 + }, + { + "epoch": 0.19408945686900958, + "grad_norm": 0.169921875, + "learning_rate": 0.0005, + "loss": 1.16, + "step": 243 + }, + { + "epoch": 0.19488817891373802, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.1381, + "step": 244 + }, + { + "epoch": 0.19568690095846644, + "grad_norm": 0.1865234375, + "learning_rate": 0.0005, + "loss": 1.1541, + "step": 245 + }, + { + "epoch": 0.1964856230031949, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.1464, + "step": 246 + }, + { + "epoch": 0.19728434504792333, + "grad_norm": 0.1689453125, + "learning_rate": 0.0005, + "loss": 1.1512, + "step": 247 + }, + { + "epoch": 0.19808306709265175, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.1664, + "step": 248 + }, + { + "epoch": 0.1988817891373802, + "grad_norm": 0.1748046875, + "learning_rate": 0.0005, + "loss": 1.1438, + "step": 249 + }, + { + "epoch": 0.19968051118210864, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.1442, + "step": 250 + }, + { + "epoch": 0.20047923322683706, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.1574, + "step": 251 + }, + { + "epoch": 0.2012779552715655, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.1525, + "step": 252 + }, + { + "epoch": 0.20207667731629392, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.1535, + "step": 253 + }, + { + "epoch": 0.20287539936102236, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.1548, + "step": 254 + }, + { + "epoch": 0.2036741214057508, + "grad_norm": 0.1748046875, + "learning_rate": 0.0005, + "loss": 1.144, + "step": 255 + }, + { + "epoch": 0.20447284345047922, + "grad_norm": 0.25390625, + "learning_rate": 0.0005, + "loss": 1.1486, + "step": 256 + }, + { + "epoch": 0.20527156549520767, + "grad_norm": 0.4609375, + "learning_rate": 0.0005, + "loss": 1.1524, + "step": 257 + }, + { + "epoch": 0.20607028753993611, + "grad_norm": 1.1015625, + "learning_rate": 0.0005, + "loss": 1.1532, + "step": 258 + }, + { + "epoch": 0.20686900958466453, + "grad_norm": 1.7265625, + "learning_rate": 0.0005, + "loss": 1.1995, + "step": 259 + }, + { + "epoch": 0.20766773162939298, + "grad_norm": 0.57421875, + "learning_rate": 0.0005, + "loss": 1.1581, + "step": 260 + }, + { + "epoch": 0.2084664536741214, + "grad_norm": 1.609375, + "learning_rate": 0.0005, + "loss": 1.2112, + "step": 261 + }, + { + "epoch": 0.20926517571884984, + "grad_norm": 0.734375, + "learning_rate": 0.0005, + "loss": 1.1968, + "step": 262 + }, + { + "epoch": 0.21006389776357828, + "grad_norm": 0.486328125, + "learning_rate": 0.0005, + "loss": 1.1754, + "step": 263 + }, + { + "epoch": 0.2108626198083067, + "grad_norm": 0.5, + "learning_rate": 0.0005, + "loss": 1.1936, + "step": 264 + }, + { + "epoch": 0.21166134185303515, + "grad_norm": 0.36328125, + "learning_rate": 0.0005, + "loss": 1.1768, + "step": 265 + }, + { + "epoch": 0.2124600638977636, + "grad_norm": 0.2158203125, + "learning_rate": 0.0005, + "loss": 1.1779, + "step": 266 + }, + { + "epoch": 0.213258785942492, + "grad_norm": 0.2578125, + "learning_rate": 0.0005, + "loss": 1.176, + "step": 267 + }, + { + "epoch": 0.21405750798722045, + "grad_norm": 0.1982421875, + "learning_rate": 0.0005, + "loss": 1.1704, + "step": 268 + }, + { + "epoch": 0.21485623003194887, + "grad_norm": 0.1865234375, + "learning_rate": 0.0005, + "loss": 1.1641, + "step": 269 + }, + { + "epoch": 0.21565495207667731, + "grad_norm": 0.1962890625, + "learning_rate": 0.0005, + "loss": 1.1669, + "step": 270 + }, + { + "epoch": 0.21645367412140576, + "grad_norm": 0.1845703125, + "learning_rate": 0.0005, + "loss": 1.1597, + "step": 271 + }, + { + "epoch": 0.21725239616613418, + "grad_norm": 0.162109375, + "learning_rate": 0.0005, + "loss": 1.1687, + "step": 272 + }, + { + "epoch": 0.21805111821086262, + "grad_norm": 0.1708984375, + "learning_rate": 0.0005, + "loss": 1.1648, + "step": 273 + }, + { + "epoch": 0.21884984025559107, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.1612, + "step": 274 + }, + { + "epoch": 0.21964856230031948, + "grad_norm": 0.162109375, + "learning_rate": 0.0005, + "loss": 1.1593, + "step": 275 + }, + { + "epoch": 0.22044728434504793, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.1561, + "step": 276 + }, + { + "epoch": 0.22124600638977635, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.1633, + "step": 277 + }, + { + "epoch": 0.2220447284345048, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.1636, + "step": 278 + }, + { + "epoch": 0.22284345047923323, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.1603, + "step": 279 + }, + { + "epoch": 0.22364217252396165, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.1487, + "step": 280 + }, + { + "epoch": 0.2244408945686901, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.1538, + "step": 281 + }, + { + "epoch": 0.22523961661341854, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.1502, + "step": 282 + }, + { + "epoch": 0.22603833865814696, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.1464, + "step": 283 + }, + { + "epoch": 0.2268370607028754, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.1429, + "step": 284 + }, + { + "epoch": 0.22763578274760382, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.1509, + "step": 285 + }, + { + "epoch": 0.22843450479233227, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.1482, + "step": 286 + }, + { + "epoch": 0.2292332268370607, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.1456, + "step": 287 + }, + { + "epoch": 0.23003194888178913, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.1466, + "step": 288 + }, + { + "epoch": 0.23083067092651757, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.1476, + "step": 289 + }, + { + "epoch": 0.23162939297124602, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.1431, + "step": 290 + }, + { + "epoch": 0.23242811501597443, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.1368, + "step": 291 + }, + { + "epoch": 0.23322683706070288, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.1503, + "step": 292 + }, + { + "epoch": 0.2340255591054313, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.1473, + "step": 293 + }, + { + "epoch": 0.23482428115015974, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.1399, + "step": 294 + }, + { + "epoch": 0.2356230031948882, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.146, + "step": 295 + }, + { + "epoch": 0.2364217252396166, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.1397, + "step": 296 + }, + { + "epoch": 0.23722044728434505, + "grad_norm": 0.1875, + "learning_rate": 0.0005, + "loss": 1.1352, + "step": 297 + }, + { + "epoch": 0.2380191693290735, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.1542, + "step": 298 + }, + { + "epoch": 0.2388178913738019, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.1501, + "step": 299 + }, + { + "epoch": 0.23961661341853036, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.1447, + "step": 300 + }, + { + "epoch": 0.24041533546325877, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.1419, + "step": 301 + }, + { + "epoch": 0.24121405750798722, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.1453, + "step": 302 + }, + { + "epoch": 0.24201277955271566, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.1445, + "step": 303 + }, + { + "epoch": 0.24281150159744408, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.1336, + "step": 304 + }, + { + "epoch": 0.24361022364217252, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.1361, + "step": 305 + }, + { + "epoch": 0.24440894568690097, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.1425, + "step": 306 + }, + { + "epoch": 0.2452076677316294, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.1384, + "step": 307 + }, + { + "epoch": 0.24600638977635783, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.1411, + "step": 308 + }, + { + "epoch": 0.24680511182108625, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.1402, + "step": 309 + }, + { + "epoch": 0.2476038338658147, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.1458, + "step": 310 + }, + { + "epoch": 0.24840255591054314, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.1405, + "step": 311 + }, + { + "epoch": 0.24920127795527156, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.1299, + "step": 312 + }, + { + "epoch": 0.25, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.1471, + "step": 313 + }, + { + "epoch": 0.2507987220447284, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.143, + "step": 314 + }, + { + "epoch": 0.2515974440894569, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.1316, + "step": 315 + }, + { + "epoch": 0.2523961661341853, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.1278, + "step": 316 + }, + { + "epoch": 0.2531948881789137, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.1331, + "step": 317 + }, + { + "epoch": 0.2539936102236422, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.1323, + "step": 318 + }, + { + "epoch": 0.2547923322683706, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.1435, + "step": 319 + }, + { + "epoch": 0.25559105431309903, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.1457, + "step": 320 + }, + { + "epoch": 0.2563897763578275, + "grad_norm": 0.1611328125, + "learning_rate": 0.0005, + "loss": 1.1368, + "step": 321 + }, + { + "epoch": 0.2571884984025559, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.1356, + "step": 322 + }, + { + "epoch": 0.25798722044728434, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.1448, + "step": 323 + }, + { + "epoch": 0.25878594249201275, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.1317, + "step": 324 + }, + { + "epoch": 0.2595846645367412, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.1273, + "step": 325 + }, + { + "epoch": 0.26038338658146964, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.1294, + "step": 326 + }, + { + "epoch": 0.26118210862619806, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.1413, + "step": 327 + }, + { + "epoch": 0.26198083067092653, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.1443, + "step": 328 + }, + { + "epoch": 0.26277955271565495, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.1377, + "step": 329 + }, + { + "epoch": 0.26357827476038337, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.1413, + "step": 330 + }, + { + "epoch": 0.26437699680511184, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.1431, + "step": 331 + }, + { + "epoch": 0.26517571884984026, + "grad_norm": 0.16796875, + "learning_rate": 0.0005, + "loss": 1.139, + "step": 332 + }, + { + "epoch": 0.2659744408945687, + "grad_norm": 0.3125, + "learning_rate": 0.0005, + "loss": 1.1247, + "step": 333 + }, + { + "epoch": 0.26677316293929715, + "grad_norm": 0.328125, + "learning_rate": 0.0005, + "loss": 1.1276, + "step": 334 + }, + { + "epoch": 0.26757188498402557, + "grad_norm": 0.259765625, + "learning_rate": 0.0005, + "loss": 1.1361, + "step": 335 + }, + { + "epoch": 0.268370607028754, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.1391, + "step": 336 + }, + { + "epoch": 0.26916932907348246, + "grad_norm": 0.2451171875, + "learning_rate": 0.0005, + "loss": 1.1318, + "step": 337 + }, + { + "epoch": 0.26996805111821087, + "grad_norm": 0.2734375, + "learning_rate": 0.0005, + "loss": 1.1413, + "step": 338 + }, + { + "epoch": 0.2707667731629393, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.133, + "step": 339 + }, + { + "epoch": 0.2715654952076677, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.1295, + "step": 340 + }, + { + "epoch": 0.2723642172523962, + "grad_norm": 0.173828125, + "learning_rate": 0.0005, + "loss": 1.1469, + "step": 341 + }, + { + "epoch": 0.2731629392971246, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.1369, + "step": 342 + }, + { + "epoch": 0.273961661341853, + "grad_norm": 0.2314453125, + "learning_rate": 0.0005, + "loss": 1.1399, + "step": 343 + }, + { + "epoch": 0.2747603833865815, + "grad_norm": 0.1962890625, + "learning_rate": 0.0005, + "loss": 1.1384, + "step": 344 + }, + { + "epoch": 0.2755591054313099, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.138, + "step": 345 + }, + { + "epoch": 0.2763578274760383, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.1324, + "step": 346 + }, + { + "epoch": 0.2771565495207668, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.1349, + "step": 347 + }, + { + "epoch": 0.2779552715654952, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.1282, + "step": 348 + }, + { + "epoch": 0.2787539936102236, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.1255, + "step": 349 + }, + { + "epoch": 0.2795527156549521, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.1371, + "step": 350 + }, + { + "epoch": 0.2803514376996805, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.1312, + "step": 351 + }, + { + "epoch": 0.28115015974440893, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.1274, + "step": 352 + }, + { + "epoch": 0.2819488817891374, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.1203, + "step": 353 + }, + { + "epoch": 0.2827476038338658, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.129, + "step": 354 + }, + { + "epoch": 0.28354632587859424, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.1257, + "step": 355 + }, + { + "epoch": 0.28434504792332266, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.133, + "step": 356 + }, + { + "epoch": 0.28514376996805113, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.1362, + "step": 357 + }, + { + "epoch": 0.28594249201277955, + "grad_norm": 0.1767578125, + "learning_rate": 0.0005, + "loss": 1.1255, + "step": 358 + }, + { + "epoch": 0.28674121405750796, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.1298, + "step": 359 + }, + { + "epoch": 0.28753993610223644, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.1307, + "step": 360 + }, + { + "epoch": 0.28833865814696485, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.1364, + "step": 361 + }, + { + "epoch": 0.28913738019169327, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.1232, + "step": 362 + }, + { + "epoch": 0.28993610223642174, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.1229, + "step": 363 + }, + { + "epoch": 0.29073482428115016, + "grad_norm": 0.1572265625, + "learning_rate": 0.0005, + "loss": 1.1296, + "step": 364 + }, + { + "epoch": 0.2915335463258786, + "grad_norm": 0.240234375, + "learning_rate": 0.0005, + "loss": 1.1335, + "step": 365 + }, + { + "epoch": 0.29233226837060705, + "grad_norm": 0.203125, + "learning_rate": 0.0005, + "loss": 1.1294, + "step": 366 + }, + { + "epoch": 0.29313099041533547, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.1285, + "step": 367 + }, + { + "epoch": 0.2939297124600639, + "grad_norm": 0.1728515625, + "learning_rate": 0.0005, + "loss": 1.1336, + "step": 368 + }, + { + "epoch": 0.29472843450479236, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.1183, + "step": 369 + }, + { + "epoch": 0.2955271565495208, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.1297, + "step": 370 + }, + { + "epoch": 0.2963258785942492, + "grad_norm": 0.412109375, + "learning_rate": 0.0005, + "loss": 1.1281, + "step": 371 + }, + { + "epoch": 0.2971246006389776, + "grad_norm": 0.98828125, + "learning_rate": 0.0005, + "loss": 1.1473, + "step": 372 + }, + { + "epoch": 0.2979233226837061, + "grad_norm": 1.859375, + "learning_rate": 0.0005, + "loss": 1.1801, + "step": 373 + }, + { + "epoch": 0.2987220447284345, + "grad_norm": 0.54296875, + "learning_rate": 0.0005, + "loss": 1.1492, + "step": 374 + }, + { + "epoch": 0.2995207667731629, + "grad_norm": 0.93359375, + "learning_rate": 0.0005, + "loss": 1.1727, + "step": 375 + }, + { + "epoch": 0.3003194888178914, + "grad_norm": 0.6640625, + "learning_rate": 0.0005, + "loss": 1.1768, + "step": 376 + }, + { + "epoch": 0.3011182108626198, + "grad_norm": 0.61328125, + "learning_rate": 0.0005, + "loss": 1.1756, + "step": 377 + }, + { + "epoch": 0.3019169329073482, + "grad_norm": 0.34765625, + "learning_rate": 0.0005, + "loss": 1.159, + "step": 378 + }, + { + "epoch": 0.3027156549520767, + "grad_norm": 0.28515625, + "learning_rate": 0.0005, + "loss": 1.1505, + "step": 379 + }, + { + "epoch": 0.3035143769968051, + "grad_norm": 0.287109375, + "learning_rate": 0.0005, + "loss": 1.1577, + "step": 380 + }, + { + "epoch": 0.30431309904153353, + "grad_norm": 0.201171875, + "learning_rate": 0.0005, + "loss": 1.1527, + "step": 381 + }, + { + "epoch": 0.305111821086262, + "grad_norm": 0.1953125, + "learning_rate": 0.0005, + "loss": 1.1455, + "step": 382 + }, + { + "epoch": 0.3059105431309904, + "grad_norm": 0.21484375, + "learning_rate": 0.0005, + "loss": 1.1435, + "step": 383 + }, + { + "epoch": 0.30670926517571884, + "grad_norm": 0.1826171875, + "learning_rate": 0.0005, + "loss": 1.1382, + "step": 384 + }, + { + "epoch": 0.3075079872204473, + "grad_norm": 0.208984375, + "learning_rate": 0.0005, + "loss": 1.1401, + "step": 385 + }, + { + "epoch": 0.3083067092651757, + "grad_norm": 0.3046875, + "learning_rate": 0.0005, + "loss": 1.134, + "step": 386 + }, + { + "epoch": 0.30910543130990414, + "grad_norm": 0.2021484375, + "learning_rate": 0.0005, + "loss": 1.1486, + "step": 387 + }, + { + "epoch": 0.30990415335463256, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.142, + "step": 388 + }, + { + "epoch": 0.31070287539936103, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.1284, + "step": 389 + }, + { + "epoch": 0.31150159744408945, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.1367, + "step": 390 + }, + { + "epoch": 0.31230031948881787, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.13, + "step": 391 + }, + { + "epoch": 0.31309904153354634, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.1327, + "step": 392 + }, + { + "epoch": 0.31389776357827476, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.1333, + "step": 393 + }, + { + "epoch": 0.3146964856230032, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.1375, + "step": 394 + }, + { + "epoch": 0.31549520766773165, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.1328, + "step": 395 + }, + { + "epoch": 0.31629392971246006, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.1222, + "step": 396 + }, + { + "epoch": 0.3170926517571885, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.1357, + "step": 397 + }, + { + "epoch": 0.31789137380191695, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.1254, + "step": 398 + }, + { + "epoch": 0.31869009584664537, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.1337, + "step": 399 + }, + { + "epoch": 0.3194888178913738, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.1393, + "step": 400 + }, + { + "epoch": 0.32028753993610226, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.133, + "step": 401 + }, + { + "epoch": 0.3210862619808307, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.1283, + "step": 402 + }, + { + "epoch": 0.3218849840255591, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.1247, + "step": 403 + }, + { + "epoch": 0.3226837060702875, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.1338, + "step": 404 + }, + { + "epoch": 0.323482428115016, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.1305, + "step": 405 + }, + { + "epoch": 0.3242811501597444, + "grad_norm": 0.1875, + "learning_rate": 0.0005, + "loss": 1.1282, + "step": 406 + }, + { + "epoch": 0.3250798722044728, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.1281, + "step": 407 + }, + { + "epoch": 0.3258785942492013, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.1223, + "step": 408 + }, + { + "epoch": 0.3266773162939297, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.1345, + "step": 409 + }, + { + "epoch": 0.3274760383386581, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.1319, + "step": 410 + }, + { + "epoch": 0.3282747603833866, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.1251, + "step": 411 + }, + { + "epoch": 0.329073482428115, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.1324, + "step": 412 + }, + { + "epoch": 0.32987220447284343, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.1249, + "step": 413 + }, + { + "epoch": 0.3306709265175719, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.1269, + "step": 414 + }, + { + "epoch": 0.3314696485623003, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.1213, + "step": 415 + }, + { + "epoch": 0.33226837060702874, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.127, + "step": 416 + }, + { + "epoch": 0.3330670926517572, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.1242, + "step": 417 + }, + { + "epoch": 0.33386581469648563, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.1268, + "step": 418 + }, + { + "epoch": 0.33466453674121405, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.1178, + "step": 419 + }, + { + "epoch": 0.3354632587859425, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.1197, + "step": 420 + }, + { + "epoch": 0.33626198083067094, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.1202, + "step": 421 + }, + { + "epoch": 0.33706070287539935, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.1224, + "step": 422 + }, + { + "epoch": 0.33785942492012777, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.121, + "step": 423 + }, + { + "epoch": 0.33865814696485624, + "grad_norm": 0.2119140625, + "learning_rate": 0.0005, + "loss": 1.1228, + "step": 424 + }, + { + "epoch": 0.33945686900958466, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.1296, + "step": 425 + }, + { + "epoch": 0.3402555910543131, + "grad_norm": 0.291015625, + "learning_rate": 0.0005, + "loss": 1.1158, + "step": 426 + }, + { + "epoch": 0.34105431309904155, + "grad_norm": 0.2060546875, + "learning_rate": 0.0005, + "loss": 1.1176, + "step": 427 + }, + { + "epoch": 0.34185303514376997, + "grad_norm": 0.193359375, + "learning_rate": 0.0005, + "loss": 1.1271, + "step": 428 + }, + { + "epoch": 0.3426517571884984, + "grad_norm": 0.33984375, + "learning_rate": 0.0005, + "loss": 1.129, + "step": 429 + }, + { + "epoch": 0.34345047923322686, + "grad_norm": 0.177734375, + "learning_rate": 0.0005, + "loss": 1.1216, + "step": 430 + }, + { + "epoch": 0.3442492012779553, + "grad_norm": 0.2080078125, + "learning_rate": 0.0005, + "loss": 1.1222, + "step": 431 + }, + { + "epoch": 0.3450479233226837, + "grad_norm": 0.279296875, + "learning_rate": 0.0005, + "loss": 1.1313, + "step": 432 + }, + { + "epoch": 0.34584664536741216, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.1188, + "step": 433 + }, + { + "epoch": 0.3466453674121406, + "grad_norm": 0.26953125, + "learning_rate": 0.0005, + "loss": 1.1187, + "step": 434 + }, + { + "epoch": 0.347444089456869, + "grad_norm": 0.1962890625, + "learning_rate": 0.0005, + "loss": 1.1242, + "step": 435 + }, + { + "epoch": 0.34824281150159747, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.126, + "step": 436 + }, + { + "epoch": 0.3490415335463259, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.1118, + "step": 437 + }, + { + "epoch": 0.3498402555910543, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.1232, + "step": 438 + }, + { + "epoch": 0.3506389776357827, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.1179, + "step": 439 + }, + { + "epoch": 0.3514376996805112, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.1152, + "step": 440 + }, + { + "epoch": 0.3522364217252396, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.1251, + "step": 441 + }, + { + "epoch": 0.35303514376996803, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.1221, + "step": 442 + }, + { + "epoch": 0.3538338658146965, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.1254, + "step": 443 + }, + { + "epoch": 0.3546325878594249, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.1345, + "step": 444 + }, + { + "epoch": 0.35543130990415334, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.1168, + "step": 445 + }, + { + "epoch": 0.3562300319488818, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.1192, + "step": 446 + }, + { + "epoch": 0.3570287539936102, + "grad_norm": 0.162109375, + "learning_rate": 0.0005, + "loss": 1.1178, + "step": 447 + }, + { + "epoch": 0.35782747603833864, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.1207, + "step": 448 + }, + { + "epoch": 0.3586261980830671, + "grad_norm": 0.189453125, + "learning_rate": 0.0005, + "loss": 1.1265, + "step": 449 + }, + { + "epoch": 0.35942492012779553, + "grad_norm": 0.1953125, + "learning_rate": 0.0005, + "loss": 1.1157, + "step": 450 + }, + { + "epoch": 0.36022364217252395, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.1327, + "step": 451 + }, + { + "epoch": 0.3610223642172524, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.1269, + "step": 452 + }, + { + "epoch": 0.36182108626198084, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.1315, + "step": 453 + }, + { + "epoch": 0.36261980830670926, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.1191, + "step": 454 + }, + { + "epoch": 0.3634185303514377, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.1189, + "step": 455 + }, + { + "epoch": 0.36421725239616615, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.1242, + "step": 456 + }, + { + "epoch": 0.36501597444089456, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.1136, + "step": 457 + }, + { + "epoch": 0.365814696485623, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.1097, + "step": 458 + }, + { + "epoch": 0.36661341853035145, + "grad_norm": 0.1572265625, + "learning_rate": 0.0005, + "loss": 1.1236, + "step": 459 + }, + { + "epoch": 0.36741214057507987, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.1145, + "step": 460 + }, + { + "epoch": 0.3682108626198083, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.1147, + "step": 461 + }, + { + "epoch": 0.36900958466453676, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.1226, + "step": 462 + }, + { + "epoch": 0.3698083067092652, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.1234, + "step": 463 + }, + { + "epoch": 0.3706070287539936, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.1264, + "step": 464 + }, + { + "epoch": 0.37140575079872207, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.1194, + "step": 465 + }, + { + "epoch": 0.3722044728434505, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.1101, + "step": 466 + }, + { + "epoch": 0.3730031948881789, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.1133, + "step": 467 + }, + { + "epoch": 0.3738019169329074, + "grad_norm": 0.40234375, + "learning_rate": 0.0005, + "loss": 1.1222, + "step": 468 + }, + { + "epoch": 0.3746006389776358, + "grad_norm": 0.68359375, + "learning_rate": 0.0005, + "loss": 1.121, + "step": 469 + }, + { + "epoch": 0.3753993610223642, + "grad_norm": 1.234375, + "learning_rate": 0.0005, + "loss": 1.137, + "step": 470 + }, + { + "epoch": 0.3761980830670926, + "grad_norm": 1.09375, + "learning_rate": 0.0005, + "loss": 1.1463, + "step": 471 + }, + { + "epoch": 0.3769968051118211, + "grad_norm": 0.431640625, + "learning_rate": 0.0005, + "loss": 1.1271, + "step": 472 + }, + { + "epoch": 0.3777955271565495, + "grad_norm": 0.42578125, + "learning_rate": 0.0005, + "loss": 1.1262, + "step": 473 + }, + { + "epoch": 0.37859424920127793, + "grad_norm": 0.5625, + "learning_rate": 0.0005, + "loss": 1.1252, + "step": 474 + }, + { + "epoch": 0.3793929712460064, + "grad_norm": 0.208984375, + "learning_rate": 0.0005, + "loss": 1.1165, + "step": 475 + }, + { + "epoch": 0.3801916932907348, + "grad_norm": 0.42578125, + "learning_rate": 0.0005, + "loss": 1.1201, + "step": 476 + }, + { + "epoch": 0.38099041533546324, + "grad_norm": 0.2392578125, + "learning_rate": 0.0005, + "loss": 1.1309, + "step": 477 + }, + { + "epoch": 0.3817891373801917, + "grad_norm": 0.30859375, + "learning_rate": 0.0005, + "loss": 1.1169, + "step": 478 + }, + { + "epoch": 0.38258785942492013, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.1224, + "step": 479 + }, + { + "epoch": 0.38338658146964855, + "grad_norm": 0.328125, + "learning_rate": 0.0005, + "loss": 1.1142, + "step": 480 + }, + { + "epoch": 0.384185303514377, + "grad_norm": 0.173828125, + "learning_rate": 0.0005, + "loss": 1.1156, + "step": 481 + }, + { + "epoch": 0.38498402555910544, + "grad_norm": 0.2333984375, + "learning_rate": 0.0005, + "loss": 1.1236, + "step": 482 + }, + { + "epoch": 0.38578274760383385, + "grad_norm": 0.1953125, + "learning_rate": 0.0005, + "loss": 1.1237, + "step": 483 + }, + { + "epoch": 0.3865814696485623, + "grad_norm": 0.171875, + "learning_rate": 0.0005, + "loss": 1.1188, + "step": 484 + }, + { + "epoch": 0.38738019169329074, + "grad_norm": 0.1748046875, + "learning_rate": 0.0005, + "loss": 1.1125, + "step": 485 + }, + { + "epoch": 0.38817891373801916, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.1224, + "step": 486 + }, + { + "epoch": 0.3889776357827476, + "grad_norm": 0.197265625, + "learning_rate": 0.0005, + "loss": 1.1102, + "step": 487 + }, + { + "epoch": 0.38977635782747605, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.1192, + "step": 488 + }, + { + "epoch": 0.39057507987220447, + "grad_norm": 0.1943359375, + "learning_rate": 0.0005, + "loss": 1.1155, + "step": 489 + }, + { + "epoch": 0.3913738019169329, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.1109, + "step": 490 + }, + { + "epoch": 0.39217252396166136, + "grad_norm": 0.220703125, + "learning_rate": 0.0005, + "loss": 1.1218, + "step": 491 + }, + { + "epoch": 0.3929712460063898, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.1197, + "step": 492 + }, + { + "epoch": 0.3937699680511182, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.1296, + "step": 493 + }, + { + "epoch": 0.39456869009584666, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.1173, + "step": 494 + }, + { + "epoch": 0.3953674121405751, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.1196, + "step": 495 + }, + { + "epoch": 0.3961661341853035, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.1126, + "step": 496 + }, + { + "epoch": 0.39696485623003197, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.1059, + "step": 497 + }, + { + "epoch": 0.3977635782747604, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.1153, + "step": 498 + }, + { + "epoch": 0.3985623003194888, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.1224, + "step": 499 + }, + { + "epoch": 0.3993610223642173, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.1206, + "step": 500 + }, + { + "epoch": 0.4001597444089457, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.109, + "step": 501 + }, + { + "epoch": 0.4009584664536741, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.1218, + "step": 502 + }, + { + "epoch": 0.40175718849840253, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.1223, + "step": 503 + }, + { + "epoch": 0.402555910543131, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.1182, + "step": 504 + }, + { + "epoch": 0.4033546325878594, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.123, + "step": 505 + }, + { + "epoch": 0.40415335463258784, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.1184, + "step": 506 + }, + { + "epoch": 0.4049520766773163, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.1057, + "step": 507 + }, + { + "epoch": 0.4057507987220447, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.0982, + "step": 508 + }, + { + "epoch": 0.40654952076677314, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.1176, + "step": 509 + }, + { + "epoch": 0.4073482428115016, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.1152, + "step": 510 + }, + { + "epoch": 0.40814696485623003, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.1221, + "step": 511 + }, + { + "epoch": 0.40894568690095845, + "grad_norm": 0.1640625, + "learning_rate": 0.0005, + "loss": 1.1183, + "step": 512 + }, + { + "epoch": 0.4097444089456869, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.1093, + "step": 513 + }, + { + "epoch": 0.41054313099041534, + "grad_norm": 0.1845703125, + "learning_rate": 0.0005, + "loss": 1.1134, + "step": 514 + }, + { + "epoch": 0.41134185303514376, + "grad_norm": 0.259765625, + "learning_rate": 0.0005, + "loss": 1.1209, + "step": 515 + }, + { + "epoch": 0.41214057507987223, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.1104, + "step": 516 + }, + { + "epoch": 0.41293929712460065, + "grad_norm": 0.22265625, + "learning_rate": 0.0005, + "loss": 1.1189, + "step": 517 + }, + { + "epoch": 0.41373801916932906, + "grad_norm": 0.33984375, + "learning_rate": 0.0005, + "loss": 1.1134, + "step": 518 + }, + { + "epoch": 0.4145367412140575, + "grad_norm": 0.2099609375, + "learning_rate": 0.0005, + "loss": 1.1115, + "step": 519 + }, + { + "epoch": 0.41533546325878595, + "grad_norm": 0.1669921875, + "learning_rate": 0.0005, + "loss": 1.1102, + "step": 520 + }, + { + "epoch": 0.41613418530351437, + "grad_norm": 0.2275390625, + "learning_rate": 0.0005, + "loss": 1.1153, + "step": 521 + }, + { + "epoch": 0.4169329073482428, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.1097, + "step": 522 + }, + { + "epoch": 0.41773162939297126, + "grad_norm": 0.1962890625, + "learning_rate": 0.0005, + "loss": 1.1127, + "step": 523 + }, + { + "epoch": 0.4185303514376997, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.1126, + "step": 524 + }, + { + "epoch": 0.4193290734824281, + "grad_norm": 0.2001953125, + "learning_rate": 0.0005, + "loss": 1.1151, + "step": 525 + }, + { + "epoch": 0.42012779552715657, + "grad_norm": 0.1923828125, + "learning_rate": 0.0005, + "loss": 1.1149, + "step": 526 + }, + { + "epoch": 0.420926517571885, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.1092, + "step": 527 + }, + { + "epoch": 0.4217252396166134, + "grad_norm": 0.173828125, + "learning_rate": 0.0005, + "loss": 1.1165, + "step": 528 + }, + { + "epoch": 0.4225239616613419, + "grad_norm": 0.2333984375, + "learning_rate": 0.0005, + "loss": 1.1071, + "step": 529 + }, + { + "epoch": 0.4233226837060703, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.1106, + "step": 530 + }, + { + "epoch": 0.4241214057507987, + "grad_norm": 0.2099609375, + "learning_rate": 0.0005, + "loss": 1.1117, + "step": 531 + }, + { + "epoch": 0.4249201277955272, + "grad_norm": 0.2421875, + "learning_rate": 0.0005, + "loss": 1.112, + "step": 532 + }, + { + "epoch": 0.4257188498402556, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.1068, + "step": 533 + }, + { + "epoch": 0.426517571884984, + "grad_norm": 0.34765625, + "learning_rate": 0.0005, + "loss": 1.1119, + "step": 534 + }, + { + "epoch": 0.4273162939297125, + "grad_norm": 0.40234375, + "learning_rate": 0.0005, + "loss": 1.105, + "step": 535 + }, + { + "epoch": 0.4281150159744409, + "grad_norm": 0.251953125, + "learning_rate": 0.0005, + "loss": 1.1161, + "step": 536 + }, + { + "epoch": 0.4289137380191693, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.1158, + "step": 537 + }, + { + "epoch": 0.42971246006389774, + "grad_norm": 0.240234375, + "learning_rate": 0.0005, + "loss": 1.1235, + "step": 538 + }, + { + "epoch": 0.4305111821086262, + "grad_norm": 0.232421875, + "learning_rate": 0.0005, + "loss": 1.1158, + "step": 539 + }, + { + "epoch": 0.43130990415335463, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.1099, + "step": 540 + }, + { + "epoch": 0.43210862619808305, + "grad_norm": 0.2021484375, + "learning_rate": 0.0005, + "loss": 1.1045, + "step": 541 + }, + { + "epoch": 0.4329073482428115, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.1139, + "step": 542 + }, + { + "epoch": 0.43370607028753994, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.1139, + "step": 543 + }, + { + "epoch": 0.43450479233226835, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.126, + "step": 544 + }, + { + "epoch": 0.4353035143769968, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.1072, + "step": 545 + }, + { + "epoch": 0.43610223642172524, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.1103, + "step": 546 + }, + { + "epoch": 0.43690095846645366, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.111, + "step": 547 + }, + { + "epoch": 0.43769968051118213, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.1141, + "step": 548 + }, + { + "epoch": 0.43849840255591055, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.1091, + "step": 549 + }, + { + "epoch": 0.43929712460063897, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.1031, + "step": 550 + }, + { + "epoch": 0.44009584664536744, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.1076, + "step": 551 + }, + { + "epoch": 0.44089456869009586, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.1142, + "step": 552 + }, + { + "epoch": 0.4416932907348243, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.1074, + "step": 553 + }, + { + "epoch": 0.4424920127795527, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.1137, + "step": 554 + }, + { + "epoch": 0.44329073482428116, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.1066, + "step": 555 + }, + { + "epoch": 0.4440894568690096, + "grad_norm": 0.291015625, + "learning_rate": 0.0005, + "loss": 1.1009, + "step": 556 + }, + { + "epoch": 0.444888178913738, + "grad_norm": 0.39453125, + "learning_rate": 0.0005, + "loss": 1.1078, + "step": 557 + }, + { + "epoch": 0.44568690095846647, + "grad_norm": 0.298828125, + "learning_rate": 0.0005, + "loss": 1.1088, + "step": 558 + }, + { + "epoch": 0.4464856230031949, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.112, + "step": 559 + }, + { + "epoch": 0.4472843450479233, + "grad_norm": 0.251953125, + "learning_rate": 0.0005, + "loss": 1.108, + "step": 560 + }, + { + "epoch": 0.4480830670926518, + "grad_norm": 0.19921875, + "learning_rate": 0.0005, + "loss": 1.108, + "step": 561 + }, + { + "epoch": 0.4488817891373802, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.1101, + "step": 562 + }, + { + "epoch": 0.4496805111821086, + "grad_norm": 0.1669921875, + "learning_rate": 0.0005, + "loss": 1.1174, + "step": 563 + }, + { + "epoch": 0.4504792332268371, + "grad_norm": 0.1533203125, + "learning_rate": 0.0005, + "loss": 1.1184, + "step": 564 + }, + { + "epoch": 0.4512779552715655, + "grad_norm": 0.1640625, + "learning_rate": 0.0005, + "loss": 1.1089, + "step": 565 + }, + { + "epoch": 0.4520766773162939, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.1091, + "step": 566 + }, + { + "epoch": 0.4528753993610224, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.1129, + "step": 567 + }, + { + "epoch": 0.4536741214057508, + "grad_norm": 0.1845703125, + "learning_rate": 0.0005, + "loss": 1.1067, + "step": 568 + }, + { + "epoch": 0.4544728434504792, + "grad_norm": 0.201171875, + "learning_rate": 0.0005, + "loss": 1.104, + "step": 569 + }, + { + "epoch": 0.45527156549520764, + "grad_norm": 0.185546875, + "learning_rate": 0.0005, + "loss": 1.1136, + "step": 570 + }, + { + "epoch": 0.4560702875399361, + "grad_norm": 0.16796875, + "learning_rate": 0.0005, + "loss": 1.1129, + "step": 571 + }, + { + "epoch": 0.45686900958466453, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.0965, + "step": 572 + }, + { + "epoch": 0.45766773162939295, + "grad_norm": 0.251953125, + "learning_rate": 0.0005, + "loss": 1.1054, + "step": 573 + }, + { + "epoch": 0.4584664536741214, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.1134, + "step": 574 + }, + { + "epoch": 0.45926517571884984, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.099, + "step": 575 + }, + { + "epoch": 0.46006389776357826, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.1155, + "step": 576 + }, + { + "epoch": 0.46086261980830673, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.1075, + "step": 577 + }, + { + "epoch": 0.46166134185303515, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.1058, + "step": 578 + }, + { + "epoch": 0.46246006389776356, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.1054, + "step": 579 + }, + { + "epoch": 0.46325878594249204, + "grad_norm": 0.1728515625, + "learning_rate": 0.0005, + "loss": 1.109, + "step": 580 + }, + { + "epoch": 0.46405750798722045, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.1043, + "step": 581 + }, + { + "epoch": 0.46485623003194887, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.1141, + "step": 582 + }, + { + "epoch": 0.46565495207667734, + "grad_norm": 0.2001953125, + "learning_rate": 0.0005, + "loss": 1.11, + "step": 583 + }, + { + "epoch": 0.46645367412140576, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.1029, + "step": 584 + }, + { + "epoch": 0.4672523961661342, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.1132, + "step": 585 + }, + { + "epoch": 0.4680511182108626, + "grad_norm": 0.19140625, + "learning_rate": 0.0005, + "loss": 1.1099, + "step": 586 + }, + { + "epoch": 0.46884984025559107, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.1056, + "step": 587 + }, + { + "epoch": 0.4696485623003195, + "grad_norm": 0.2021484375, + "learning_rate": 0.0005, + "loss": 1.1117, + "step": 588 + }, + { + "epoch": 0.4704472843450479, + "grad_norm": 0.369140625, + "learning_rate": 0.0005, + "loss": 1.1087, + "step": 589 + }, + { + "epoch": 0.4712460063897764, + "grad_norm": 0.384765625, + "learning_rate": 0.0005, + "loss": 1.1152, + "step": 590 + }, + { + "epoch": 0.4720447284345048, + "grad_norm": 0.349609375, + "learning_rate": 0.0005, + "loss": 1.1014, + "step": 591 + }, + { + "epoch": 0.4728434504792332, + "grad_norm": 0.212890625, + "learning_rate": 0.0005, + "loss": 1.1081, + "step": 592 + }, + { + "epoch": 0.4736421725239617, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.1115, + "step": 593 + }, + { + "epoch": 0.4744408945686901, + "grad_norm": 0.33984375, + "learning_rate": 0.0005, + "loss": 1.1067, + "step": 594 + }, + { + "epoch": 0.4752396166134185, + "grad_norm": 0.3359375, + "learning_rate": 0.0005, + "loss": 1.1093, + "step": 595 + }, + { + "epoch": 0.476038338658147, + "grad_norm": 0.189453125, + "learning_rate": 0.0005, + "loss": 1.1143, + "step": 596 + }, + { + "epoch": 0.4768370607028754, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.1083, + "step": 597 + }, + { + "epoch": 0.4776357827476038, + "grad_norm": 0.2138671875, + "learning_rate": 0.0005, + "loss": 1.1031, + "step": 598 + }, + { + "epoch": 0.4784345047923323, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.1094, + "step": 599 + }, + { + "epoch": 0.4792332268370607, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.1034, + "step": 600 + }, + { + "epoch": 0.48003194888178913, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.1134, + "step": 601 + }, + { + "epoch": 0.48083067092651754, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.1126, + "step": 602 + }, + { + "epoch": 0.481629392971246, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.1012, + "step": 603 + }, + { + "epoch": 0.48242811501597443, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.1026, + "step": 604 + }, + { + "epoch": 0.48322683706070285, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.105, + "step": 605 + }, + { + "epoch": 0.4840255591054313, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0959, + "step": 606 + }, + { + "epoch": 0.48482428115015974, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.1045, + "step": 607 + }, + { + "epoch": 0.48562300319488816, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.1116, + "step": 608 + }, + { + "epoch": 0.48642172523961663, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.1238, + "step": 609 + }, + { + "epoch": 0.48722044728434505, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.1133, + "step": 610 + }, + { + "epoch": 0.48801916932907347, + "grad_norm": 0.2060546875, + "learning_rate": 0.0005, + "loss": 1.1091, + "step": 611 + }, + { + "epoch": 0.48881789137380194, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.1126, + "step": 612 + }, + { + "epoch": 0.48961661341853036, + "grad_norm": 0.2353515625, + "learning_rate": 0.0005, + "loss": 1.1088, + "step": 613 + }, + { + "epoch": 0.4904153354632588, + "grad_norm": 0.1796875, + "learning_rate": 0.0005, + "loss": 1.1027, + "step": 614 + }, + { + "epoch": 0.49121405750798725, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.1112, + "step": 615 + }, + { + "epoch": 0.49201277955271566, + "grad_norm": 0.166015625, + "learning_rate": 0.0005, + "loss": 1.1058, + "step": 616 + }, + { + "epoch": 0.4928115015974441, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.1106, + "step": 617 + }, + { + "epoch": 0.4936102236421725, + "grad_norm": 0.16015625, + "learning_rate": 0.0005, + "loss": 1.0974, + "step": 618 + }, + { + "epoch": 0.49440894568690097, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.1098, + "step": 619 + }, + { + "epoch": 0.4952076677316294, + "grad_norm": 0.263671875, + "learning_rate": 0.0005, + "loss": 1.1025, + "step": 620 + }, + { + "epoch": 0.4960063897763578, + "grad_norm": 0.234375, + "learning_rate": 0.0005, + "loss": 1.1047, + "step": 621 + }, + { + "epoch": 0.4968051118210863, + "grad_norm": 0.2119140625, + "learning_rate": 0.0005, + "loss": 1.1037, + "step": 622 + }, + { + "epoch": 0.4976038338658147, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.1063, + "step": 623 + }, + { + "epoch": 0.4984025559105431, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.1039, + "step": 624 + }, + { + "epoch": 0.4992012779552716, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.1091, + "step": 625 + }, + { + "epoch": 0.5, + "grad_norm": 0.16796875, + "learning_rate": 0.0005, + "loss": 1.1117, + "step": 626 + }, + { + "epoch": 0.5007987220447284, + "grad_norm": 0.1533203125, + "learning_rate": 0.0005, + "loss": 1.1133, + "step": 627 + }, + { + "epoch": 0.5015974440894568, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.104, + "step": 628 + }, + { + "epoch": 0.5023961661341853, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.1, + "step": 629 + }, + { + "epoch": 0.5031948881789138, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.1037, + "step": 630 + }, + { + "epoch": 0.5039936102236422, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.1105, + "step": 631 + }, + { + "epoch": 0.5047923322683706, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.1141, + "step": 632 + }, + { + "epoch": 0.505591054313099, + "grad_norm": 0.162109375, + "learning_rate": 0.0005, + "loss": 1.1048, + "step": 633 + }, + { + "epoch": 0.5063897763578274, + "grad_norm": 0.177734375, + "learning_rate": 0.0005, + "loss": 1.1123, + "step": 634 + }, + { + "epoch": 0.5071884984025559, + "grad_norm": 0.171875, + "learning_rate": 0.0005, + "loss": 1.1118, + "step": 635 + }, + { + "epoch": 0.5079872204472844, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.105, + "step": 636 + }, + { + "epoch": 0.5087859424920128, + "grad_norm": 0.216796875, + "learning_rate": 0.0005, + "loss": 1.1117, + "step": 637 + }, + { + "epoch": 0.5095846645367412, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0943, + "step": 638 + }, + { + "epoch": 0.5103833865814696, + "grad_norm": 0.201171875, + "learning_rate": 0.0005, + "loss": 1.1123, + "step": 639 + }, + { + "epoch": 0.5111821086261981, + "grad_norm": 0.181640625, + "learning_rate": 0.0005, + "loss": 1.1034, + "step": 640 + }, + { + "epoch": 0.5119808306709265, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.1145, + "step": 641 + }, + { + "epoch": 0.512779552715655, + "grad_norm": 0.2119140625, + "learning_rate": 0.0005, + "loss": 1.106, + "step": 642 + }, + { + "epoch": 0.5135782747603834, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.1089, + "step": 643 + }, + { + "epoch": 0.5143769968051118, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0985, + "step": 644 + }, + { + "epoch": 0.5151757188498403, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.1046, + "step": 645 + }, + { + "epoch": 0.5159744408945687, + "grad_norm": 0.251953125, + "learning_rate": 0.0005, + "loss": 1.1167, + "step": 646 + }, + { + "epoch": 0.5167731629392971, + "grad_norm": 0.255859375, + "learning_rate": 0.0005, + "loss": 1.1155, + "step": 647 + }, + { + "epoch": 0.5175718849840255, + "grad_norm": 0.1962890625, + "learning_rate": 0.0005, + "loss": 1.1156, + "step": 648 + }, + { + "epoch": 0.518370607028754, + "grad_norm": 0.1533203125, + "learning_rate": 0.0005, + "loss": 1.1109, + "step": 649 + }, + { + "epoch": 0.5191693290734825, + "grad_norm": 0.16796875, + "learning_rate": 0.0005, + "loss": 1.114, + "step": 650 + }, + { + "epoch": 0.5199680511182109, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.1104, + "step": 651 + }, + { + "epoch": 0.5207667731629393, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0981, + "step": 652 + }, + { + "epoch": 0.5215654952076677, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0948, + "step": 653 + }, + { + "epoch": 0.5223642172523961, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.1, + "step": 654 + }, + { + "epoch": 0.5231629392971247, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.0995, + "step": 655 + }, + { + "epoch": 0.5239616613418531, + "grad_norm": 0.212890625, + "learning_rate": 0.0005, + "loss": 1.0972, + "step": 656 + }, + { + "epoch": 0.5247603833865815, + "grad_norm": 0.1953125, + "learning_rate": 0.0005, + "loss": 1.0985, + "step": 657 + }, + { + "epoch": 0.5255591054313099, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.1098, + "step": 658 + }, + { + "epoch": 0.5263578274760383, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.1023, + "step": 659 + }, + { + "epoch": 0.5271565495207667, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.1137, + "step": 660 + }, + { + "epoch": 0.5279552715654952, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.1056, + "step": 661 + }, + { + "epoch": 0.5287539936102237, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.1164, + "step": 662 + }, + { + "epoch": 0.5295527156549521, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0962, + "step": 663 + }, + { + "epoch": 0.5303514376996805, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.101, + "step": 664 + }, + { + "epoch": 0.5311501597444089, + "grad_norm": 0.1845703125, + "learning_rate": 0.0005, + "loss": 1.1065, + "step": 665 + }, + { + "epoch": 0.5319488817891374, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0955, + "step": 666 + }, + { + "epoch": 0.5327476038338658, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.1011, + "step": 667 + }, + { + "epoch": 0.5335463258785943, + "grad_norm": 0.2294921875, + "learning_rate": 0.0005, + "loss": 1.1019, + "step": 668 + }, + { + "epoch": 0.5343450479233227, + "grad_norm": 0.2333984375, + "learning_rate": 0.0005, + "loss": 1.0947, + "step": 669 + }, + { + "epoch": 0.5351437699680511, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.1021, + "step": 670 + }, + { + "epoch": 0.5359424920127795, + "grad_norm": 0.2470703125, + "learning_rate": 0.0005, + "loss": 1.0986, + "step": 671 + }, + { + "epoch": 0.536741214057508, + "grad_norm": 0.306640625, + "learning_rate": 0.0005, + "loss": 1.098, + "step": 672 + }, + { + "epoch": 0.5375399361022364, + "grad_norm": 0.318359375, + "learning_rate": 0.0005, + "loss": 1.099, + "step": 673 + }, + { + "epoch": 0.5383386581469649, + "grad_norm": 0.349609375, + "learning_rate": 0.0005, + "loss": 1.1018, + "step": 674 + }, + { + "epoch": 0.5391373801916933, + "grad_norm": 0.44921875, + "learning_rate": 0.0005, + "loss": 1.105, + "step": 675 + }, + { + "epoch": 0.5399361022364217, + "grad_norm": 0.73828125, + "learning_rate": 0.0005, + "loss": 1.1084, + "step": 676 + }, + { + "epoch": 0.5407348242811502, + "grad_norm": 1.375, + "learning_rate": 0.0005, + "loss": 1.1432, + "step": 677 + }, + { + "epoch": 0.5415335463258786, + "grad_norm": 0.60546875, + "learning_rate": 0.0005, + "loss": 1.1199, + "step": 678 + }, + { + "epoch": 0.542332268370607, + "grad_norm": 0.36328125, + "learning_rate": 0.0005, + "loss": 1.1087, + "step": 679 + }, + { + "epoch": 0.5431309904153354, + "grad_norm": 0.65625, + "learning_rate": 0.0005, + "loss": 1.1163, + "step": 680 + }, + { + "epoch": 0.5439297124600639, + "grad_norm": 0.427734375, + "learning_rate": 0.0005, + "loss": 1.1134, + "step": 681 + }, + { + "epoch": 0.5447284345047924, + "grad_norm": 0.421875, + "learning_rate": 0.0005, + "loss": 1.1157, + "step": 682 + }, + { + "epoch": 0.5455271565495208, + "grad_norm": 0.6171875, + "learning_rate": 0.0005, + "loss": 1.1113, + "step": 683 + }, + { + "epoch": 0.5463258785942492, + "grad_norm": 0.44921875, + "learning_rate": 0.0005, + "loss": 1.1141, + "step": 684 + }, + { + "epoch": 0.5471246006389776, + "grad_norm": 0.2158203125, + "learning_rate": 0.0005, + "loss": 1.1062, + "step": 685 + }, + { + "epoch": 0.547923322683706, + "grad_norm": 0.408203125, + "learning_rate": 0.0005, + "loss": 1.1039, + "step": 686 + }, + { + "epoch": 0.5487220447284346, + "grad_norm": 0.1640625, + "learning_rate": 0.0005, + "loss": 1.1026, + "step": 687 + }, + { + "epoch": 0.549520766773163, + "grad_norm": 0.310546875, + "learning_rate": 0.0005, + "loss": 1.1019, + "step": 688 + }, + { + "epoch": 0.5503194888178914, + "grad_norm": 0.2412109375, + "learning_rate": 0.0005, + "loss": 1.1028, + "step": 689 + }, + { + "epoch": 0.5511182108626198, + "grad_norm": 0.1767578125, + "learning_rate": 0.0005, + "loss": 1.1059, + "step": 690 + }, + { + "epoch": 0.5519169329073482, + "grad_norm": 0.2421875, + "learning_rate": 0.0005, + "loss": 1.1102, + "step": 691 + }, + { + "epoch": 0.5527156549520766, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.0941, + "step": 692 + }, + { + "epoch": 0.5535143769968051, + "grad_norm": 0.310546875, + "learning_rate": 0.0005, + "loss": 1.1009, + "step": 693 + }, + { + "epoch": 0.5543130990415336, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.115, + "step": 694 + }, + { + "epoch": 0.555111821086262, + "grad_norm": 0.263671875, + "learning_rate": 0.0005, + "loss": 1.1034, + "step": 695 + }, + { + "epoch": 0.5559105431309904, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.1011, + "step": 696 + }, + { + "epoch": 0.5567092651757188, + "grad_norm": 0.2197265625, + "learning_rate": 0.0005, + "loss": 1.1155, + "step": 697 + }, + { + "epoch": 0.5575079872204473, + "grad_norm": 0.177734375, + "learning_rate": 0.0005, + "loss": 1.1012, + "step": 698 + }, + { + "epoch": 0.5583067092651757, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.1088, + "step": 699 + }, + { + "epoch": 0.5591054313099042, + "grad_norm": 0.216796875, + "learning_rate": 0.0005, + "loss": 1.0977, + "step": 700 + }, + { + "epoch": 0.5599041533546326, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.097, + "step": 701 + }, + { + "epoch": 0.560702875399361, + "grad_norm": 0.341796875, + "learning_rate": 0.0005, + "loss": 1.0966, + "step": 702 + }, + { + "epoch": 0.5615015974440895, + "grad_norm": 0.2314453125, + "learning_rate": 0.0005, + "loss": 1.0994, + "step": 703 + }, + { + "epoch": 0.5623003194888179, + "grad_norm": 0.177734375, + "learning_rate": 0.0005, + "loss": 1.1057, + "step": 704 + }, + { + "epoch": 0.5630990415335463, + "grad_norm": 0.333984375, + "learning_rate": 0.0005, + "loss": 1.101, + "step": 705 + }, + { + "epoch": 0.5638977635782748, + "grad_norm": 0.208984375, + "learning_rate": 0.0005, + "loss": 1.1105, + "step": 706 + }, + { + "epoch": 0.5646964856230032, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0914, + "step": 707 + }, + { + "epoch": 0.5654952076677316, + "grad_norm": 0.22265625, + "learning_rate": 0.0005, + "loss": 1.099, + "step": 708 + }, + { + "epoch": 0.5662939297124601, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0949, + "step": 709 + }, + { + "epoch": 0.5670926517571885, + "grad_norm": 0.2255859375, + "learning_rate": 0.0005, + "loss": 1.1047, + "step": 710 + }, + { + "epoch": 0.5678913738019169, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.1047, + "step": 711 + }, + { + "epoch": 0.5686900958466453, + "grad_norm": 0.1953125, + "learning_rate": 0.0005, + "loss": 1.0924, + "step": 712 + }, + { + "epoch": 0.5694888178913738, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.1027, + "step": 713 + }, + { + "epoch": 0.5702875399361023, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0892, + "step": 714 + }, + { + "epoch": 0.5710862619808307, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.1007, + "step": 715 + }, + { + "epoch": 0.5718849840255591, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0926, + "step": 716 + }, + { + "epoch": 0.5726837060702875, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.0934, + "step": 717 + }, + { + "epoch": 0.5734824281150159, + "grad_norm": 0.2021484375, + "learning_rate": 0.0005, + "loss": 1.1134, + "step": 718 + }, + { + "epoch": 0.5742811501597445, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.093, + "step": 719 + }, + { + "epoch": 0.5750798722044729, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0982, + "step": 720 + }, + { + "epoch": 0.5758785942492013, + "grad_norm": 0.1708984375, + "learning_rate": 0.0005, + "loss": 1.0993, + "step": 721 + }, + { + "epoch": 0.5766773162939297, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.1032, + "step": 722 + }, + { + "epoch": 0.5774760383386581, + "grad_norm": 0.1630859375, + "learning_rate": 0.0005, + "loss": 1.0936, + "step": 723 + }, + { + "epoch": 0.5782747603833865, + "grad_norm": 0.2021484375, + "learning_rate": 0.0005, + "loss": 1.1067, + "step": 724 + }, + { + "epoch": 0.579073482428115, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0952, + "step": 725 + }, + { + "epoch": 0.5798722044728435, + "grad_norm": 0.3125, + "learning_rate": 0.0005, + "loss": 1.1092, + "step": 726 + }, + { + "epoch": 0.5806709265175719, + "grad_norm": 0.1669921875, + "learning_rate": 0.0005, + "loss": 1.1094, + "step": 727 + }, + { + "epoch": 0.5814696485623003, + "grad_norm": 0.2138671875, + "learning_rate": 0.0005, + "loss": 1.0918, + "step": 728 + }, + { + "epoch": 0.5822683706070287, + "grad_norm": 0.244140625, + "learning_rate": 0.0005, + "loss": 1.0901, + "step": 729 + }, + { + "epoch": 0.5830670926517572, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.093, + "step": 730 + }, + { + "epoch": 0.5838658146964856, + "grad_norm": 0.36328125, + "learning_rate": 0.0005, + "loss": 1.1086, + "step": 731 + }, + { + "epoch": 0.5846645367412141, + "grad_norm": 0.392578125, + "learning_rate": 0.0005, + "loss": 1.0994, + "step": 732 + }, + { + "epoch": 0.5854632587859425, + "grad_norm": 0.265625, + "learning_rate": 0.0005, + "loss": 1.0974, + "step": 733 + }, + { + "epoch": 0.5862619808306709, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.1002, + "step": 734 + }, + { + "epoch": 0.5870607028753994, + "grad_norm": 0.2333984375, + "learning_rate": 0.0005, + "loss": 1.0885, + "step": 735 + }, + { + "epoch": 0.5878594249201278, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.1018, + "step": 736 + }, + { + "epoch": 0.5886581469648562, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.1028, + "step": 737 + }, + { + "epoch": 0.5894568690095847, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0968, + "step": 738 + }, + { + "epoch": 0.5902555910543131, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.0953, + "step": 739 + }, + { + "epoch": 0.5910543130990416, + "grad_norm": 0.232421875, + "learning_rate": 0.0005, + "loss": 1.0898, + "step": 740 + }, + { + "epoch": 0.59185303514377, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.1016, + "step": 741 + }, + { + "epoch": 0.5926517571884984, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0917, + "step": 742 + }, + { + "epoch": 0.5934504792332268, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0913, + "step": 743 + }, + { + "epoch": 0.5942492012779552, + "grad_norm": 0.2451171875, + "learning_rate": 0.0005, + "loss": 1.1103, + "step": 744 + }, + { + "epoch": 0.5950479233226837, + "grad_norm": 0.18359375, + "learning_rate": 0.0005, + "loss": 1.0922, + "step": 745 + }, + { + "epoch": 0.5958466453674122, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0992, + "step": 746 + }, + { + "epoch": 0.5966453674121406, + "grad_norm": 0.17578125, + "learning_rate": 0.0005, + "loss": 1.0986, + "step": 747 + }, + { + "epoch": 0.597444089456869, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0949, + "step": 748 + }, + { + "epoch": 0.5982428115015974, + "grad_norm": 0.1640625, + "learning_rate": 0.0005, + "loss": 1.0967, + "step": 749 + }, + { + "epoch": 0.5990415335463258, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0959, + "step": 750 + }, + { + "epoch": 0.5998402555910544, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.1002, + "step": 751 + }, + { + "epoch": 0.6006389776357828, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0989, + "step": 752 + }, + { + "epoch": 0.6014376996805112, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.101, + "step": 753 + }, + { + "epoch": 0.6022364217252396, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.1073, + "step": 754 + }, + { + "epoch": 0.603035143769968, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.1072, + "step": 755 + }, + { + "epoch": 0.6038338658146964, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0928, + "step": 756 + }, + { + "epoch": 0.604632587859425, + "grad_norm": 0.2041015625, + "learning_rate": 0.0005, + "loss": 1.101, + "step": 757 + }, + { + "epoch": 0.6054313099041534, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0982, + "step": 758 + }, + { + "epoch": 0.6062300319488818, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0974, + "step": 759 + }, + { + "epoch": 0.6070287539936102, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.108, + "step": 760 + }, + { + "epoch": 0.6078274760383386, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0865, + "step": 761 + }, + { + "epoch": 0.6086261980830671, + "grad_norm": 0.181640625, + "learning_rate": 0.0005, + "loss": 1.1098, + "step": 762 + }, + { + "epoch": 0.6094249201277955, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.1042, + "step": 763 + }, + { + "epoch": 0.610223642172524, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.1076, + "step": 764 + }, + { + "epoch": 0.6110223642172524, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.1062, + "step": 765 + }, + { + "epoch": 0.6118210862619808, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0958, + "step": 766 + }, + { + "epoch": 0.6126198083067093, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.1007, + "step": 767 + }, + { + "epoch": 0.6134185303514377, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0934, + "step": 768 + }, + { + "epoch": 0.6142172523961661, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.1004, + "step": 769 + }, + { + "epoch": 0.6150159744408946, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.1008, + "step": 770 + }, + { + "epoch": 0.615814696485623, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.101, + "step": 771 + }, + { + "epoch": 0.6166134185303515, + "grad_norm": 0.1689453125, + "learning_rate": 0.0005, + "loss": 1.0958, + "step": 772 + }, + { + "epoch": 0.6174121405750799, + "grad_norm": 0.228515625, + "learning_rate": 0.0005, + "loss": 1.0959, + "step": 773 + }, + { + "epoch": 0.6182108626198083, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0989, + "step": 774 + }, + { + "epoch": 0.6190095846645367, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.1091, + "step": 775 + }, + { + "epoch": 0.6198083067092651, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0957, + "step": 776 + }, + { + "epoch": 0.6206070287539937, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0976, + "step": 777 + }, + { + "epoch": 0.6214057507987221, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.1017, + "step": 778 + }, + { + "epoch": 0.6222044728434505, + "grad_norm": 0.169921875, + "learning_rate": 0.0005, + "loss": 1.0988, + "step": 779 + }, + { + "epoch": 0.6230031948881789, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.1009, + "step": 780 + }, + { + "epoch": 0.6238019169329073, + "grad_norm": 0.16015625, + "learning_rate": 0.0005, + "loss": 1.0916, + "step": 781 + }, + { + "epoch": 0.6246006389776357, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.1063, + "step": 782 + }, + { + "epoch": 0.6253993610223643, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.1076, + "step": 783 + }, + { + "epoch": 0.6261980830670927, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.1029, + "step": 784 + }, + { + "epoch": 0.6269968051118211, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.1023, + "step": 785 + }, + { + "epoch": 0.6277955271565495, + "grad_norm": 0.1767578125, + "learning_rate": 0.0005, + "loss": 1.105, + "step": 786 + }, + { + "epoch": 0.6285942492012779, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0951, + "step": 787 + }, + { + "epoch": 0.6293929712460063, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.1068, + "step": 788 + }, + { + "epoch": 0.6301916932907349, + "grad_norm": 0.1640625, + "learning_rate": 0.0005, + "loss": 1.0971, + "step": 789 + }, + { + "epoch": 0.6309904153354633, + "grad_norm": 0.1748046875, + "learning_rate": 0.0005, + "loss": 1.0895, + "step": 790 + }, + { + "epoch": 0.6317891373801917, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0959, + "step": 791 + }, + { + "epoch": 0.6325878594249201, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0991, + "step": 792 + }, + { + "epoch": 0.6333865814696485, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0954, + "step": 793 + }, + { + "epoch": 0.634185303514377, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.094, + "step": 794 + }, + { + "epoch": 0.6349840255591054, + "grad_norm": 0.1748046875, + "learning_rate": 0.0005, + "loss": 1.0914, + "step": 795 + }, + { + "epoch": 0.6357827476038339, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0954, + "step": 796 + }, + { + "epoch": 0.6365814696485623, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.09, + "step": 797 + }, + { + "epoch": 0.6373801916932907, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0878, + "step": 798 + }, + { + "epoch": 0.6381789137380192, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.1031, + "step": 799 + }, + { + "epoch": 0.6389776357827476, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0994, + "step": 800 + }, + { + "epoch": 0.639776357827476, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0951, + "step": 801 + }, + { + "epoch": 0.6405750798722045, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.0976, + "step": 802 + }, + { + "epoch": 0.6413738019169329, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.091, + "step": 803 + }, + { + "epoch": 0.6421725239616614, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.1027, + "step": 804 + }, + { + "epoch": 0.6429712460063898, + "grad_norm": 0.1962890625, + "learning_rate": 0.0005, + "loss": 1.0933, + "step": 805 + }, + { + "epoch": 0.6437699680511182, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.0876, + "step": 806 + }, + { + "epoch": 0.6445686900958466, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.1034, + "step": 807 + }, + { + "epoch": 0.645367412140575, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.0915, + "step": 808 + }, + { + "epoch": 0.6461661341853036, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0921, + "step": 809 + }, + { + "epoch": 0.646964856230032, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.0951, + "step": 810 + }, + { + "epoch": 0.6477635782747604, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0959, + "step": 811 + }, + { + "epoch": 0.6485623003194888, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0996, + "step": 812 + }, + { + "epoch": 0.6493610223642172, + "grad_norm": 0.216796875, + "learning_rate": 0.0005, + "loss": 1.1001, + "step": 813 + }, + { + "epoch": 0.6501597444089456, + "grad_norm": 0.236328125, + "learning_rate": 0.0005, + "loss": 1.0944, + "step": 814 + }, + { + "epoch": 0.6509584664536742, + "grad_norm": 0.1533203125, + "learning_rate": 0.0005, + "loss": 1.0936, + "step": 815 + }, + { + "epoch": 0.6517571884984026, + "grad_norm": 0.1845703125, + "learning_rate": 0.0005, + "loss": 1.0999, + "step": 816 + }, + { + "epoch": 0.652555910543131, + "grad_norm": 0.34375, + "learning_rate": 0.0005, + "loss": 1.1, + "step": 817 + }, + { + "epoch": 0.6533546325878594, + "grad_norm": 0.40625, + "learning_rate": 0.0005, + "loss": 1.0889, + "step": 818 + }, + { + "epoch": 0.6541533546325878, + "grad_norm": 0.4921875, + "learning_rate": 0.0005, + "loss": 1.0971, + "step": 819 + }, + { + "epoch": 0.6549520766773163, + "grad_norm": 0.625, + "learning_rate": 0.0005, + "loss": 1.0997, + "step": 820 + }, + { + "epoch": 0.6557507987220448, + "grad_norm": 0.89453125, + "learning_rate": 0.0005, + "loss": 1.1009, + "step": 821 + }, + { + "epoch": 0.6565495207667732, + "grad_norm": 1.0546875, + "learning_rate": 0.0005, + "loss": 1.1061, + "step": 822 + }, + { + "epoch": 0.6573482428115016, + "grad_norm": 0.66796875, + "learning_rate": 0.0005, + "loss": 1.1118, + "step": 823 + }, + { + "epoch": 0.65814696485623, + "grad_norm": 0.36328125, + "learning_rate": 0.0005, + "loss": 1.1035, + "step": 824 + }, + { + "epoch": 0.6589456869009584, + "grad_norm": 0.58203125, + "learning_rate": 0.0005, + "loss": 1.1017, + "step": 825 + }, + { + "epoch": 0.6597444089456869, + "grad_norm": 0.484375, + "learning_rate": 0.0005, + "loss": 1.11, + "step": 826 + }, + { + "epoch": 0.6605431309904153, + "grad_norm": 0.326171875, + "learning_rate": 0.0005, + "loss": 1.098, + "step": 827 + }, + { + "epoch": 0.6613418530351438, + "grad_norm": 0.44921875, + "learning_rate": 0.0005, + "loss": 1.1007, + "step": 828 + }, + { + "epoch": 0.6621405750798722, + "grad_norm": 0.37109375, + "learning_rate": 0.0005, + "loss": 1.0971, + "step": 829 + }, + { + "epoch": 0.6629392971246006, + "grad_norm": 0.359375, + "learning_rate": 0.0005, + "loss": 1.1047, + "step": 830 + }, + { + "epoch": 0.6637380191693291, + "grad_norm": 0.390625, + "learning_rate": 0.0005, + "loss": 1.1009, + "step": 831 + }, + { + "epoch": 0.6645367412140575, + "grad_norm": 0.1572265625, + "learning_rate": 0.0005, + "loss": 1.1, + "step": 832 + }, + { + "epoch": 0.6653354632587859, + "grad_norm": 0.30078125, + "learning_rate": 0.0005, + "loss": 1.1117, + "step": 833 + }, + { + "epoch": 0.6661341853035144, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0995, + "step": 834 + }, + { + "epoch": 0.6669329073482428, + "grad_norm": 0.251953125, + "learning_rate": 0.0005, + "loss": 1.1038, + "step": 835 + }, + { + "epoch": 0.6677316293929713, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0961, + "step": 836 + }, + { + "epoch": 0.6685303514376997, + "grad_norm": 0.232421875, + "learning_rate": 0.0005, + "loss": 1.097, + "step": 837 + }, + { + "epoch": 0.6693290734824281, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.1009, + "step": 838 + }, + { + "epoch": 0.6701277955271565, + "grad_norm": 0.1962890625, + "learning_rate": 0.0005, + "loss": 1.0916, + "step": 839 + }, + { + "epoch": 0.670926517571885, + "grad_norm": 0.173828125, + "learning_rate": 0.0005, + "loss": 1.0939, + "step": 840 + }, + { + "epoch": 0.6717252396166135, + "grad_norm": 0.17578125, + "learning_rate": 0.0005, + "loss": 1.0962, + "step": 841 + }, + { + "epoch": 0.6725239616613419, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0938, + "step": 842 + }, + { + "epoch": 0.6733226837060703, + "grad_norm": 0.18359375, + "learning_rate": 0.0005, + "loss": 1.0986, + "step": 843 + }, + { + "epoch": 0.6741214057507987, + "grad_norm": 0.19921875, + "learning_rate": 0.0005, + "loss": 1.1042, + "step": 844 + }, + { + "epoch": 0.6749201277955271, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0863, + "step": 845 + }, + { + "epoch": 0.6757188498402555, + "grad_norm": 0.1787109375, + "learning_rate": 0.0005, + "loss": 1.0977, + "step": 846 + }, + { + "epoch": 0.6765175718849841, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0901, + "step": 847 + }, + { + "epoch": 0.6773162939297125, + "grad_norm": 0.1611328125, + "learning_rate": 0.0005, + "loss": 1.095, + "step": 848 + }, + { + "epoch": 0.6781150159744409, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0845, + "step": 849 + }, + { + "epoch": 0.6789137380191693, + "grad_norm": 0.16796875, + "learning_rate": 0.0005, + "loss": 1.09, + "step": 850 + }, + { + "epoch": 0.6797124600638977, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0969, + "step": 851 + }, + { + "epoch": 0.6805111821086262, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0982, + "step": 852 + }, + { + "epoch": 0.6813099041533547, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0919, + "step": 853 + }, + { + "epoch": 0.6821086261980831, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0875, + "step": 854 + }, + { + "epoch": 0.6829073482428115, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.097, + "step": 855 + }, + { + "epoch": 0.6837060702875399, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0961, + "step": 856 + }, + { + "epoch": 0.6845047923322684, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0938, + "step": 857 + }, + { + "epoch": 0.6853035143769968, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0877, + "step": 858 + }, + { + "epoch": 0.6861022364217252, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.095, + "step": 859 + }, + { + "epoch": 0.6869009584664537, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.099, + "step": 860 + }, + { + "epoch": 0.6876996805111821, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.094, + "step": 861 + }, + { + "epoch": 0.6884984025559105, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.1022, + "step": 862 + }, + { + "epoch": 0.689297124600639, + "grad_norm": 0.2041015625, + "learning_rate": 0.0005, + "loss": 1.0878, + "step": 863 + }, + { + "epoch": 0.6900958466453674, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0871, + "step": 864 + }, + { + "epoch": 0.6908945686900958, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.1039, + "step": 865 + }, + { + "epoch": 0.6916932907348243, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0948, + "step": 866 + }, + { + "epoch": 0.6924920127795527, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.0928, + "step": 867 + }, + { + "epoch": 0.6932907348242812, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0889, + "step": 868 + }, + { + "epoch": 0.6940894568690096, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0918, + "step": 869 + }, + { + "epoch": 0.694888178913738, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0829, + "step": 870 + }, + { + "epoch": 0.6956869009584664, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0913, + "step": 871 + }, + { + "epoch": 0.6964856230031949, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0961, + "step": 872 + }, + { + "epoch": 0.6972843450479234, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.088, + "step": 873 + }, + { + "epoch": 0.6980830670926518, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.1016, + "step": 874 + }, + { + "epoch": 0.6988817891373802, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0894, + "step": 875 + }, + { + "epoch": 0.6996805111821086, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0877, + "step": 876 + }, + { + "epoch": 0.700479233226837, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.1, + "step": 877 + }, + { + "epoch": 0.7012779552715654, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0927, + "step": 878 + }, + { + "epoch": 0.702076677316294, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0885, + "step": 879 + }, + { + "epoch": 0.7028753993610224, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0924, + "step": 880 + }, + { + "epoch": 0.7036741214057508, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0911, + "step": 881 + }, + { + "epoch": 0.7044728434504792, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.1065, + "step": 882 + }, + { + "epoch": 0.7052715654952076, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.099, + "step": 883 + }, + { + "epoch": 0.7060702875399361, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.1007, + "step": 884 + }, + { + "epoch": 0.7068690095846646, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0913, + "step": 885 + }, + { + "epoch": 0.707667731629393, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.1015, + "step": 886 + }, + { + "epoch": 0.7084664536741214, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0969, + "step": 887 + }, + { + "epoch": 0.7092651757188498, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.093, + "step": 888 + }, + { + "epoch": 0.7100638977635783, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0864, + "step": 889 + }, + { + "epoch": 0.7108626198083067, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0909, + "step": 890 + }, + { + "epoch": 0.7116613418530351, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0877, + "step": 891 + }, + { + "epoch": 0.7124600638977636, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.092, + "step": 892 + }, + { + "epoch": 0.713258785942492, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0959, + "step": 893 + }, + { + "epoch": 0.7140575079872205, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0912, + "step": 894 + }, + { + "epoch": 0.7148562300319489, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0952, + "step": 895 + }, + { + "epoch": 0.7156549520766773, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0899, + "step": 896 + }, + { + "epoch": 0.7164536741214057, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0902, + "step": 897 + }, + { + "epoch": 0.7172523961661342, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0869, + "step": 898 + }, + { + "epoch": 0.7180511182108626, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0954, + "step": 899 + }, + { + "epoch": 0.7188498402555911, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0889, + "step": 900 + }, + { + "epoch": 0.7196485623003195, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0895, + "step": 901 + }, + { + "epoch": 0.7204472843450479, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0921, + "step": 902 + }, + { + "epoch": 0.7212460063897763, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0939, + "step": 903 + }, + { + "epoch": 0.7220447284345048, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.098, + "step": 904 + }, + { + "epoch": 0.7228434504792333, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0981, + "step": 905 + }, + { + "epoch": 0.7236421725239617, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0942, + "step": 906 + }, + { + "epoch": 0.7244408945686901, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0806, + "step": 907 + }, + { + "epoch": 0.7252396166134185, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.1022, + "step": 908 + }, + { + "epoch": 0.7260383386581469, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0903, + "step": 909 + }, + { + "epoch": 0.7268370607028753, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0938, + "step": 910 + }, + { + "epoch": 0.7276357827476039, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0948, + "step": 911 + }, + { + "epoch": 0.7284345047923323, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0935, + "step": 912 + }, + { + "epoch": 0.7292332268370607, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0789, + "step": 913 + }, + { + "epoch": 0.7300319488817891, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0889, + "step": 914 + }, + { + "epoch": 0.7308306709265175, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.098, + "step": 915 + }, + { + "epoch": 0.731629392971246, + "grad_norm": 0.181640625, + "learning_rate": 0.0005, + "loss": 1.1049, + "step": 916 + }, + { + "epoch": 0.7324281150159745, + "grad_norm": 0.1640625, + "learning_rate": 0.0005, + "loss": 1.0978, + "step": 917 + }, + { + "epoch": 0.7332268370607029, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0838, + "step": 918 + }, + { + "epoch": 0.7340255591054313, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0888, + "step": 919 + }, + { + "epoch": 0.7348242811501597, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.098, + "step": 920 + }, + { + "epoch": 0.7356230031948882, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0932, + "step": 921 + }, + { + "epoch": 0.7364217252396166, + "grad_norm": 0.1962890625, + "learning_rate": 0.0005, + "loss": 1.0907, + "step": 922 + }, + { + "epoch": 0.737220447284345, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.094, + "step": 923 + }, + { + "epoch": 0.7380191693290735, + "grad_norm": 0.28125, + "learning_rate": 0.0005, + "loss": 1.0898, + "step": 924 + }, + { + "epoch": 0.7388178913738019, + "grad_norm": 0.318359375, + "learning_rate": 0.0005, + "loss": 1.0854, + "step": 925 + }, + { + "epoch": 0.7396166134185304, + "grad_norm": 0.2412109375, + "learning_rate": 0.0005, + "loss": 1.0957, + "step": 926 + }, + { + "epoch": 0.7404153354632588, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0906, + "step": 927 + }, + { + "epoch": 0.7412140575079872, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0964, + "step": 928 + }, + { + "epoch": 0.7420127795527156, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.1057, + "step": 929 + }, + { + "epoch": 0.7428115015974441, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0921, + "step": 930 + }, + { + "epoch": 0.7436102236421726, + "grad_norm": 0.1630859375, + "learning_rate": 0.0005, + "loss": 1.0984, + "step": 931 + }, + { + "epoch": 0.744408945686901, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0943, + "step": 932 + }, + { + "epoch": 0.7452076677316294, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0942, + "step": 933 + }, + { + "epoch": 0.7460063897763578, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0992, + "step": 934 + }, + { + "epoch": 0.7468051118210862, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0884, + "step": 935 + }, + { + "epoch": 0.7476038338658147, + "grad_norm": 0.1845703125, + "learning_rate": 0.0005, + "loss": 1.0868, + "step": 936 + }, + { + "epoch": 0.7484025559105432, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0928, + "step": 937 + }, + { + "epoch": 0.7492012779552716, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0955, + "step": 938 + }, + { + "epoch": 0.75, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0853, + "step": 939 + }, + { + "epoch": 0.7507987220447284, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0936, + "step": 940 + }, + { + "epoch": 0.7515974440894568, + "grad_norm": 0.1826171875, + "learning_rate": 0.0005, + "loss": 1.091, + "step": 941 + }, + { + "epoch": 0.7523961661341853, + "grad_norm": 0.234375, + "learning_rate": 0.0005, + "loss": 1.0941, + "step": 942 + }, + { + "epoch": 0.7531948881789138, + "grad_norm": 0.169921875, + "learning_rate": 0.0005, + "loss": 1.0953, + "step": 943 + }, + { + "epoch": 0.7539936102236422, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0925, + "step": 944 + }, + { + "epoch": 0.7547923322683706, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0988, + "step": 945 + }, + { + "epoch": 0.755591054313099, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.094, + "step": 946 + }, + { + "epoch": 0.7563897763578274, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0916, + "step": 947 + }, + { + "epoch": 0.7571884984025559, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0953, + "step": 948 + }, + { + "epoch": 0.7579872204472844, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0829, + "step": 949 + }, + { + "epoch": 0.7587859424920128, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0906, + "step": 950 + }, + { + "epoch": 0.7595846645367412, + "grad_norm": 0.16015625, + "learning_rate": 0.0005, + "loss": 1.0949, + "step": 951 + }, + { + "epoch": 0.7603833865814696, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0883, + "step": 952 + }, + { + "epoch": 0.7611821086261981, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0839, + "step": 953 + }, + { + "epoch": 0.7619808306709265, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0911, + "step": 954 + }, + { + "epoch": 0.762779552715655, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0874, + "step": 955 + }, + { + "epoch": 0.7635782747603834, + "grad_norm": 0.166015625, + "learning_rate": 0.0005, + "loss": 1.0909, + "step": 956 + }, + { + "epoch": 0.7643769968051118, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0969, + "step": 957 + }, + { + "epoch": 0.7651757188498403, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0984, + "step": 958 + }, + { + "epoch": 0.7659744408945687, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0908, + "step": 959 + }, + { + "epoch": 0.7667731629392971, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0902, + "step": 960 + }, + { + "epoch": 0.7675718849840255, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0877, + "step": 961 + }, + { + "epoch": 0.768370607028754, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0933, + "step": 962 + }, + { + "epoch": 0.7691693290734825, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0861, + "step": 963 + }, + { + "epoch": 0.7699680511182109, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.094, + "step": 964 + }, + { + "epoch": 0.7707667731629393, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0882, + "step": 965 + }, + { + "epoch": 0.7715654952076677, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0914, + "step": 966 + }, + { + "epoch": 0.7723642172523961, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0927, + "step": 967 + }, + { + "epoch": 0.7731629392971247, + "grad_norm": 0.1611328125, + "learning_rate": 0.0005, + "loss": 1.0893, + "step": 968 + }, + { + "epoch": 0.7739616613418531, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0957, + "step": 969 + }, + { + "epoch": 0.7747603833865815, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0791, + "step": 970 + }, + { + "epoch": 0.7755591054313099, + "grad_norm": 0.169921875, + "learning_rate": 0.0005, + "loss": 1.0836, + "step": 971 + }, + { + "epoch": 0.7763578274760383, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0806, + "step": 972 + }, + { + "epoch": 0.7771565495207667, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0851, + "step": 973 + }, + { + "epoch": 0.7779552715654952, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0818, + "step": 974 + }, + { + "epoch": 0.7787539936102237, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0891, + "step": 975 + }, + { + "epoch": 0.7795527156549521, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.0804, + "step": 976 + }, + { + "epoch": 0.7803514376996805, + "grad_norm": 0.1728515625, + "learning_rate": 0.0005, + "loss": 1.0846, + "step": 977 + }, + { + "epoch": 0.7811501597444089, + "grad_norm": 0.2421875, + "learning_rate": 0.0005, + "loss": 1.0896, + "step": 978 + }, + { + "epoch": 0.7819488817891374, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0827, + "step": 979 + }, + { + "epoch": 0.7827476038338658, + "grad_norm": 0.4296875, + "learning_rate": 0.0005, + "loss": 1.0867, + "step": 980 + }, + { + "epoch": 0.7835463258785943, + "grad_norm": 0.59375, + "learning_rate": 0.0005, + "loss": 1.0926, + "step": 981 + }, + { + "epoch": 0.7843450479233227, + "grad_norm": 0.98828125, + "learning_rate": 0.0005, + "loss": 1.1099, + "step": 982 + }, + { + "epoch": 0.7851437699680511, + "grad_norm": 1.65625, + "learning_rate": 0.0005, + "loss": 1.1436, + "step": 983 + }, + { + "epoch": 0.7859424920127795, + "grad_norm": 0.271484375, + "learning_rate": 0.0005, + "loss": 1.093, + "step": 984 + }, + { + "epoch": 0.786741214057508, + "grad_norm": 0.76171875, + "learning_rate": 0.0005, + "loss": 1.1047, + "step": 985 + }, + { + "epoch": 0.7875399361022364, + "grad_norm": 0.3046875, + "learning_rate": 0.0005, + "loss": 1.1084, + "step": 986 + }, + { + "epoch": 0.7883386581469649, + "grad_norm": 0.36328125, + "learning_rate": 0.0005, + "loss": 1.1068, + "step": 987 + }, + { + "epoch": 0.7891373801916933, + "grad_norm": 0.3203125, + "learning_rate": 0.0005, + "loss": 1.0918, + "step": 988 + }, + { + "epoch": 0.7899361022364217, + "grad_norm": 0.3203125, + "learning_rate": 0.0005, + "loss": 1.1008, + "step": 989 + }, + { + "epoch": 0.7907348242811502, + "grad_norm": 0.2421875, + "learning_rate": 0.0005, + "loss": 1.0996, + "step": 990 + }, + { + "epoch": 0.7915335463258786, + "grad_norm": 0.3046875, + "learning_rate": 0.0005, + "loss": 1.0992, + "step": 991 + }, + { + "epoch": 0.792332268370607, + "grad_norm": 0.44921875, + "learning_rate": 0.0005, + "loss": 1.0967, + "step": 992 + }, + { + "epoch": 0.7931309904153354, + "grad_norm": 0.55859375, + "learning_rate": 0.0005, + "loss": 1.1046, + "step": 993 + }, + { + "epoch": 0.7939297124600639, + "grad_norm": 0.2080078125, + "learning_rate": 0.0005, + "loss": 1.1084, + "step": 994 + }, + { + "epoch": 0.7947284345047924, + "grad_norm": 0.185546875, + "learning_rate": 0.0005, + "loss": 1.094, + "step": 995 + }, + { + "epoch": 0.7955271565495208, + "grad_norm": 0.2412109375, + "learning_rate": 0.0005, + "loss": 1.1029, + "step": 996 + }, + { + "epoch": 0.7963258785942492, + "grad_norm": 0.2353515625, + "learning_rate": 0.0005, + "loss": 1.0991, + "step": 997 + }, + { + "epoch": 0.7971246006389776, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.0948, + "step": 998 + }, + { + "epoch": 0.797923322683706, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.0943, + "step": 999 + }, + { + "epoch": 0.7987220447284346, + "grad_norm": 0.185546875, + "learning_rate": 0.0005, + "loss": 1.0951, + "step": 1000 + }, + { + "epoch": 0.799520766773163, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0894, + "step": 1001 + }, + { + "epoch": 0.8003194888178914, + "grad_norm": 0.1796875, + "learning_rate": 0.0005, + "loss": 1.0891, + "step": 1002 + }, + { + "epoch": 0.8011182108626198, + "grad_norm": 0.2431640625, + "learning_rate": 0.0005, + "loss": 1.1023, + "step": 1003 + }, + { + "epoch": 0.8019169329073482, + "grad_norm": 0.1943359375, + "learning_rate": 0.0005, + "loss": 1.0822, + "step": 1004 + }, + { + "epoch": 0.8027156549520766, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0964, + "step": 1005 + }, + { + "epoch": 0.8035143769968051, + "grad_norm": 0.203125, + "learning_rate": 0.0005, + "loss": 1.0973, + "step": 1006 + }, + { + "epoch": 0.8043130990415336, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0878, + "step": 1007 + }, + { + "epoch": 0.805111821086262, + "grad_norm": 0.2451171875, + "learning_rate": 0.0005, + "loss": 1.1023, + "step": 1008 + }, + { + "epoch": 0.8059105431309904, + "grad_norm": 0.419921875, + "learning_rate": 0.0005, + "loss": 1.0951, + "step": 1009 + }, + { + "epoch": 0.8067092651757188, + "grad_norm": 0.4453125, + "learning_rate": 0.0005, + "loss": 1.0868, + "step": 1010 + }, + { + "epoch": 0.8075079872204473, + "grad_norm": 0.31640625, + "learning_rate": 0.0005, + "loss": 1.0928, + "step": 1011 + }, + { + "epoch": 0.8083067092651757, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0903, + "step": 1012 + }, + { + "epoch": 0.8091054313099042, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.0958, + "step": 1013 + }, + { + "epoch": 0.8099041533546326, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.091, + "step": 1014 + }, + { + "epoch": 0.810702875399361, + "grad_norm": 0.267578125, + "learning_rate": 0.0005, + "loss": 1.0844, + "step": 1015 + }, + { + "epoch": 0.8115015974440895, + "grad_norm": 0.2177734375, + "learning_rate": 0.0005, + "loss": 1.0916, + "step": 1016 + }, + { + "epoch": 0.8123003194888179, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.097, + "step": 1017 + }, + { + "epoch": 0.8130990415335463, + "grad_norm": 0.1669921875, + "learning_rate": 0.0005, + "loss": 1.0832, + "step": 1018 + }, + { + "epoch": 0.8138977635782748, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0877, + "step": 1019 + }, + { + "epoch": 0.8146964856230032, + "grad_norm": 0.173828125, + "learning_rate": 0.0005, + "loss": 1.0896, + "step": 1020 + }, + { + "epoch": 0.8154952076677316, + "grad_norm": 0.177734375, + "learning_rate": 0.0005, + "loss": 1.0924, + "step": 1021 + }, + { + "epoch": 0.8162939297124601, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 1022 + }, + { + "epoch": 0.8170926517571885, + "grad_norm": 0.3046875, + "learning_rate": 0.0005, + "loss": 1.0936, + "step": 1023 + }, + { + "epoch": 0.8178913738019169, + "grad_norm": 0.275390625, + "learning_rate": 0.0005, + "loss": 1.0883, + "step": 1024 + }, + { + "epoch": 0.8186900958466453, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.1013, + "step": 1025 + }, + { + "epoch": 0.8194888178913738, + "grad_norm": 0.27734375, + "learning_rate": 0.0005, + "loss": 1.0877, + "step": 1026 + }, + { + "epoch": 0.8202875399361023, + "grad_norm": 0.193359375, + "learning_rate": 0.0005, + "loss": 1.0892, + "step": 1027 + }, + { + "epoch": 0.8210862619808307, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.096, + "step": 1028 + }, + { + "epoch": 0.8218849840255591, + "grad_norm": 0.30078125, + "learning_rate": 0.0005, + "loss": 1.0931, + "step": 1029 + }, + { + "epoch": 0.8226837060702875, + "grad_norm": 0.2099609375, + "learning_rate": 0.0005, + "loss": 1.0819, + "step": 1030 + }, + { + "epoch": 0.8234824281150159, + "grad_norm": 0.1611328125, + "learning_rate": 0.0005, + "loss": 1.0793, + "step": 1031 + }, + { + "epoch": 0.8242811501597445, + "grad_norm": 0.30859375, + "learning_rate": 0.0005, + "loss": 1.0945, + "step": 1032 + }, + { + "epoch": 0.8250798722044729, + "grad_norm": 0.201171875, + "learning_rate": 0.0005, + "loss": 1.0897, + "step": 1033 + }, + { + "epoch": 0.8258785942492013, + "grad_norm": 0.1640625, + "learning_rate": 0.0005, + "loss": 1.0832, + "step": 1034 + }, + { + "epoch": 0.8266773162939297, + "grad_norm": 0.29296875, + "learning_rate": 0.0005, + "loss": 1.0993, + "step": 1035 + }, + { + "epoch": 0.8274760383386581, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.085, + "step": 1036 + }, + { + "epoch": 0.8282747603833865, + "grad_norm": 0.25390625, + "learning_rate": 0.0005, + "loss": 1.0988, + "step": 1037 + }, + { + "epoch": 0.829073482428115, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0944, + "step": 1038 + }, + { + "epoch": 0.8298722044728435, + "grad_norm": 0.216796875, + "learning_rate": 0.0005, + "loss": 1.0874, + "step": 1039 + }, + { + "epoch": 0.8306709265175719, + "grad_norm": 0.1884765625, + "learning_rate": 0.0005, + "loss": 1.0905, + "step": 1040 + }, + { + "epoch": 0.8314696485623003, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0875, + "step": 1041 + }, + { + "epoch": 0.8322683706070287, + "grad_norm": 0.19921875, + "learning_rate": 0.0005, + "loss": 1.0913, + "step": 1042 + }, + { + "epoch": 0.8330670926517572, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.088, + "step": 1043 + }, + { + "epoch": 0.8338658146964856, + "grad_norm": 0.28125, + "learning_rate": 0.0005, + "loss": 1.0856, + "step": 1044 + }, + { + "epoch": 0.8346645367412141, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0985, + "step": 1045 + }, + { + "epoch": 0.8354632587859425, + "grad_norm": 0.1689453125, + "learning_rate": 0.0005, + "loss": 1.089, + "step": 1046 + }, + { + "epoch": 0.8362619808306709, + "grad_norm": 0.17578125, + "learning_rate": 0.0005, + "loss": 1.0911, + "step": 1047 + }, + { + "epoch": 0.8370607028753994, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0805, + "step": 1048 + }, + { + "epoch": 0.8378594249201278, + "grad_norm": 0.283203125, + "learning_rate": 0.0005, + "loss": 1.0875, + "step": 1049 + }, + { + "epoch": 0.8386581469648562, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0914, + "step": 1050 + }, + { + "epoch": 0.8394568690095847, + "grad_norm": 0.1953125, + "learning_rate": 0.0005, + "loss": 1.0965, + "step": 1051 + }, + { + "epoch": 0.8402555910543131, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0824, + "step": 1052 + }, + { + "epoch": 0.8410543130990416, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0852, + "step": 1053 + }, + { + "epoch": 0.84185303514377, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0919, + "step": 1054 + }, + { + "epoch": 0.8426517571884984, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.079, + "step": 1055 + }, + { + "epoch": 0.8434504792332268, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0865, + "step": 1056 + }, + { + "epoch": 0.8442492012779552, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0869, + "step": 1057 + }, + { + "epoch": 0.8450479233226837, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0846, + "step": 1058 + }, + { + "epoch": 0.8458466453674122, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.091, + "step": 1059 + }, + { + "epoch": 0.8466453674121406, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0842, + "step": 1060 + }, + { + "epoch": 0.847444089456869, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0926, + "step": 1061 + }, + { + "epoch": 0.8482428115015974, + "grad_norm": 0.166015625, + "learning_rate": 0.0005, + "loss": 1.0897, + "step": 1062 + }, + { + "epoch": 0.8490415335463258, + "grad_norm": 0.234375, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 1063 + }, + { + "epoch": 0.8498402555910544, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0861, + "step": 1064 + }, + { + "epoch": 0.8506389776357828, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0836, + "step": 1065 + }, + { + "epoch": 0.8514376996805112, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0834, + "step": 1066 + }, + { + "epoch": 0.8522364217252396, + "grad_norm": 0.201171875, + "learning_rate": 0.0005, + "loss": 1.0924, + "step": 1067 + }, + { + "epoch": 0.853035143769968, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0851, + "step": 1068 + }, + { + "epoch": 0.8538338658146964, + "grad_norm": 0.234375, + "learning_rate": 0.0005, + "loss": 1.0814, + "step": 1069 + }, + { + "epoch": 0.854632587859425, + "grad_norm": 0.1572265625, + "learning_rate": 0.0005, + "loss": 1.0881, + "step": 1070 + }, + { + "epoch": 0.8554313099041534, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0923, + "step": 1071 + }, + { + "epoch": 0.8562300319488818, + "grad_norm": 0.2099609375, + "learning_rate": 0.0005, + "loss": 1.0908, + "step": 1072 + }, + { + "epoch": 0.8570287539936102, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.093, + "step": 1073 + }, + { + "epoch": 0.8578274760383386, + "grad_norm": 0.2041015625, + "learning_rate": 0.0005, + "loss": 1.0885, + "step": 1074 + }, + { + "epoch": 0.8586261980830671, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0924, + "step": 1075 + }, + { + "epoch": 0.8594249201277955, + "grad_norm": 0.232421875, + "learning_rate": 0.0005, + "loss": 1.0808, + "step": 1076 + }, + { + "epoch": 0.860223642172524, + "grad_norm": 0.17578125, + "learning_rate": 0.0005, + "loss": 1.1008, + "step": 1077 + }, + { + "epoch": 0.8610223642172524, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0886, + "step": 1078 + }, + { + "epoch": 0.8618210862619808, + "grad_norm": 0.291015625, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 1079 + }, + { + "epoch": 0.8626198083067093, + "grad_norm": 0.2431640625, + "learning_rate": 0.0005, + "loss": 1.0846, + "step": 1080 + }, + { + "epoch": 0.8634185303514377, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 1081 + }, + { + "epoch": 0.8642172523961661, + "grad_norm": 0.2734375, + "learning_rate": 0.0005, + "loss": 1.0852, + "step": 1082 + }, + { + "epoch": 0.8650159744408946, + "grad_norm": 0.2265625, + "learning_rate": 0.0005, + "loss": 1.094, + "step": 1083 + }, + { + "epoch": 0.865814696485623, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.0917, + "step": 1084 + }, + { + "epoch": 0.8666134185303515, + "grad_norm": 0.31640625, + "learning_rate": 0.0005, + "loss": 1.083, + "step": 1085 + }, + { + "epoch": 0.8674121405750799, + "grad_norm": 0.2275390625, + "learning_rate": 0.0005, + "loss": 1.0811, + "step": 1086 + }, + { + "epoch": 0.8682108626198083, + "grad_norm": 0.1533203125, + "learning_rate": 0.0005, + "loss": 1.0851, + "step": 1087 + }, + { + "epoch": 0.8690095846645367, + "grad_norm": 0.3359375, + "learning_rate": 0.0005, + "loss": 1.0864, + "step": 1088 + }, + { + "epoch": 0.8698083067092651, + "grad_norm": 0.248046875, + "learning_rate": 0.0005, + "loss": 1.0827, + "step": 1089 + }, + { + "epoch": 0.8706070287539937, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0798, + "step": 1090 + }, + { + "epoch": 0.8714057507987221, + "grad_norm": 0.33984375, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 1091 + }, + { + "epoch": 0.8722044728434505, + "grad_norm": 0.2138671875, + "learning_rate": 0.0005, + "loss": 1.0875, + "step": 1092 + }, + { + "epoch": 0.8730031948881789, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.083, + "step": 1093 + }, + { + "epoch": 0.8738019169329073, + "grad_norm": 0.1826171875, + "learning_rate": 0.0005, + "loss": 1.084, + "step": 1094 + }, + { + "epoch": 0.8746006389776357, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0858, + "step": 1095 + }, + { + "epoch": 0.8753993610223643, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0874, + "step": 1096 + }, + { + "epoch": 0.8761980830670927, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0766, + "step": 1097 + }, + { + "epoch": 0.8769968051118211, + "grad_norm": 0.2021484375, + "learning_rate": 0.0005, + "loss": 1.0877, + "step": 1098 + }, + { + "epoch": 0.8777955271565495, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0892, + "step": 1099 + }, + { + "epoch": 0.8785942492012779, + "grad_norm": 0.1904296875, + "learning_rate": 0.0005, + "loss": 1.0986, + "step": 1100 + }, + { + "epoch": 0.8793929712460063, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0998, + "step": 1101 + }, + { + "epoch": 0.8801916932907349, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0854, + "step": 1102 + }, + { + "epoch": 0.8809904153354633, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0893, + "step": 1103 + }, + { + "epoch": 0.8817891373801917, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0885, + "step": 1104 + }, + { + "epoch": 0.8825878594249201, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0817, + "step": 1105 + }, + { + "epoch": 0.8833865814696485, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0922, + "step": 1106 + }, + { + "epoch": 0.884185303514377, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0759, + "step": 1107 + }, + { + "epoch": 0.8849840255591054, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.091, + "step": 1108 + }, + { + "epoch": 0.8857827476038339, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0789, + "step": 1109 + }, + { + "epoch": 0.8865814696485623, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0887, + "step": 1110 + }, + { + "epoch": 0.8873801916932907, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0874, + "step": 1111 + }, + { + "epoch": 0.8881789137380192, + "grad_norm": 0.1796875, + "learning_rate": 0.0005, + "loss": 1.0835, + "step": 1112 + }, + { + "epoch": 0.8889776357827476, + "grad_norm": 0.19921875, + "learning_rate": 0.0005, + "loss": 1.0868, + "step": 1113 + }, + { + "epoch": 0.889776357827476, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.086, + "step": 1114 + }, + { + "epoch": 0.8905750798722045, + "grad_norm": 0.2109375, + "learning_rate": 0.0005, + "loss": 1.0867, + "step": 1115 + }, + { + "epoch": 0.8913738019169329, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0872, + "step": 1116 + }, + { + "epoch": 0.8921725239616614, + "grad_norm": 0.2265625, + "learning_rate": 0.0005, + "loss": 1.0827, + "step": 1117 + }, + { + "epoch": 0.8929712460063898, + "grad_norm": 0.2490234375, + "learning_rate": 0.0005, + "loss": 1.0891, + "step": 1118 + }, + { + "epoch": 0.8937699680511182, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0796, + "step": 1119 + }, + { + "epoch": 0.8945686900958466, + "grad_norm": 0.19921875, + "learning_rate": 0.0005, + "loss": 1.0846, + "step": 1120 + }, + { + "epoch": 0.895367412140575, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0921, + "step": 1121 + }, + { + "epoch": 0.8961661341853036, + "grad_norm": 0.25, + "learning_rate": 0.0005, + "loss": 1.0829, + "step": 1122 + }, + { + "epoch": 0.896964856230032, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0829, + "step": 1123 + }, + { + "epoch": 0.8977635782747604, + "grad_norm": 0.1689453125, + "learning_rate": 0.0005, + "loss": 1.0834, + "step": 1124 + }, + { + "epoch": 0.8985623003194888, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.084, + "step": 1125 + }, + { + "epoch": 0.8993610223642172, + "grad_norm": 0.173828125, + "learning_rate": 0.0005, + "loss": 1.0891, + "step": 1126 + }, + { + "epoch": 0.9001597444089456, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0832, + "step": 1127 + }, + { + "epoch": 0.9009584664536742, + "grad_norm": 0.1826171875, + "learning_rate": 0.0005, + "loss": 1.0878, + "step": 1128 + }, + { + "epoch": 0.9017571884984026, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.0905, + "step": 1129 + }, + { + "epoch": 0.902555910543131, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0819, + "step": 1130 + }, + { + "epoch": 0.9033546325878594, + "grad_norm": 0.24609375, + "learning_rate": 0.0005, + "loss": 1.075, + "step": 1131 + }, + { + "epoch": 0.9041533546325878, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0869, + "step": 1132 + }, + { + "epoch": 0.9049520766773163, + "grad_norm": 0.1572265625, + "learning_rate": 0.0005, + "loss": 1.0782, + "step": 1133 + }, + { + "epoch": 0.9057507987220448, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0847, + "step": 1134 + }, + { + "epoch": 0.9065495207667732, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0902, + "step": 1135 + }, + { + "epoch": 0.9073482428115016, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0802, + "step": 1136 + }, + { + "epoch": 0.90814696485623, + "grad_norm": 0.16796875, + "learning_rate": 0.0005, + "loss": 1.0892, + "step": 1137 + }, + { + "epoch": 0.9089456869009584, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0893, + "step": 1138 + }, + { + "epoch": 0.9097444089456869, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0871, + "step": 1139 + }, + { + "epoch": 0.9105431309904153, + "grad_norm": 0.16796875, + "learning_rate": 0.0005, + "loss": 1.0892, + "step": 1140 + }, + { + "epoch": 0.9113418530351438, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0836, + "step": 1141 + }, + { + "epoch": 0.9121405750798722, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.0851, + "step": 1142 + }, + { + "epoch": 0.9129392971246006, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0927, + "step": 1143 + }, + { + "epoch": 0.9137380191693291, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0807, + "step": 1144 + }, + { + "epoch": 0.9145367412140575, + "grad_norm": 0.197265625, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 1145 + }, + { + "epoch": 0.9153354632587859, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0758, + "step": 1146 + }, + { + "epoch": 0.9161341853035144, + "grad_norm": 0.294921875, + "learning_rate": 0.0005, + "loss": 1.0803, + "step": 1147 + }, + { + "epoch": 0.9169329073482428, + "grad_norm": 0.287109375, + "learning_rate": 0.0005, + "loss": 1.084, + "step": 1148 + }, + { + "epoch": 0.9177316293929713, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0849, + "step": 1149 + }, + { + "epoch": 0.9185303514376997, + "grad_norm": 0.37890625, + "learning_rate": 0.0005, + "loss": 1.0848, + "step": 1150 + }, + { + "epoch": 0.9193290734824281, + "grad_norm": 0.609375, + "learning_rate": 0.0005, + "loss": 1.0988, + "step": 1151 + }, + { + "epoch": 0.9201277955271565, + "grad_norm": 0.7734375, + "learning_rate": 0.0005, + "loss": 1.0967, + "step": 1152 + }, + { + "epoch": 0.920926517571885, + "grad_norm": 0.7421875, + "learning_rate": 0.0005, + "loss": 1.0897, + "step": 1153 + }, + { + "epoch": 0.9217252396166135, + "grad_norm": 0.40625, + "learning_rate": 0.0005, + "loss": 1.0835, + "step": 1154 + }, + { + "epoch": 0.9225239616613419, + "grad_norm": 0.208984375, + "learning_rate": 0.0005, + "loss": 1.0824, + "step": 1155 + }, + { + "epoch": 0.9233226837060703, + "grad_norm": 0.458984375, + "learning_rate": 0.0005, + "loss": 1.0861, + "step": 1156 + }, + { + "epoch": 0.9241214057507987, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0935, + "step": 1157 + }, + { + "epoch": 0.9249201277955271, + "grad_norm": 0.4609375, + "learning_rate": 0.0005, + "loss": 1.0897, + "step": 1158 + }, + { + "epoch": 0.9257188498402555, + "grad_norm": 0.30078125, + "learning_rate": 0.0005, + "loss": 1.0795, + "step": 1159 + }, + { + "epoch": 0.9265175718849841, + "grad_norm": 0.2314453125, + "learning_rate": 0.0005, + "loss": 1.0891, + "step": 1160 + }, + { + "epoch": 0.9273162939297125, + "grad_norm": 0.51953125, + "learning_rate": 0.0005, + "loss": 1.0865, + "step": 1161 + }, + { + "epoch": 0.9281150159744409, + "grad_norm": 0.322265625, + "learning_rate": 0.0005, + "loss": 1.089, + "step": 1162 + }, + { + "epoch": 0.9289137380191693, + "grad_norm": 0.259765625, + "learning_rate": 0.0005, + "loss": 1.0821, + "step": 1163 + }, + { + "epoch": 0.9297124600638977, + "grad_norm": 0.6953125, + "learning_rate": 0.0005, + "loss": 1.0908, + "step": 1164 + }, + { + "epoch": 0.9305111821086262, + "grad_norm": 0.68359375, + "learning_rate": 0.0005, + "loss": 1.0902, + "step": 1165 + }, + { + "epoch": 0.9313099041533547, + "grad_norm": 0.26171875, + "learning_rate": 0.0005, + "loss": 1.0892, + "step": 1166 + }, + { + "epoch": 0.9321086261980831, + "grad_norm": 0.333984375, + "learning_rate": 0.0005, + "loss": 1.0887, + "step": 1167 + }, + { + "epoch": 0.9329073482428115, + "grad_norm": 0.474609375, + "learning_rate": 0.0005, + "loss": 1.0911, + "step": 1168 + }, + { + "epoch": 0.9337060702875399, + "grad_norm": 0.1689453125, + "learning_rate": 0.0005, + "loss": 1.085, + "step": 1169 + }, + { + "epoch": 0.9345047923322684, + "grad_norm": 0.30859375, + "learning_rate": 0.0005, + "loss": 1.0868, + "step": 1170 + }, + { + "epoch": 0.9353035143769968, + "grad_norm": 0.21875, + "learning_rate": 0.0005, + "loss": 1.0827, + "step": 1171 + }, + { + "epoch": 0.9361022364217252, + "grad_norm": 0.1572265625, + "learning_rate": 0.0005, + "loss": 1.0918, + "step": 1172 + }, + { + "epoch": 0.9369009584664537, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0822, + "step": 1173 + }, + { + "epoch": 0.9376996805111821, + "grad_norm": 0.189453125, + "learning_rate": 0.0005, + "loss": 1.0846, + "step": 1174 + }, + { + "epoch": 0.9384984025559105, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0841, + "step": 1175 + }, + { + "epoch": 0.939297124600639, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0876, + "step": 1176 + }, + { + "epoch": 0.9400958466453674, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0866, + "step": 1177 + }, + { + "epoch": 0.9408945686900958, + "grad_norm": 0.1962890625, + "learning_rate": 0.0005, + "loss": 1.0796, + "step": 1178 + }, + { + "epoch": 0.9416932907348243, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0844, + "step": 1179 + }, + { + "epoch": 0.9424920127795527, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0888, + "step": 1180 + }, + { + "epoch": 0.9432907348242812, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0833, + "step": 1181 + }, + { + "epoch": 0.9440894568690096, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0821, + "step": 1182 + }, + { + "epoch": 0.944888178913738, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0958, + "step": 1183 + }, + { + "epoch": 0.9456869009584664, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0766, + "step": 1184 + }, + { + "epoch": 0.9464856230031949, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0933, + "step": 1185 + }, + { + "epoch": 0.9472843450479234, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0816, + "step": 1186 + }, + { + "epoch": 0.9480830670926518, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0824, + "step": 1187 + }, + { + "epoch": 0.9488817891373802, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0811, + "step": 1188 + }, + { + "epoch": 0.9496805111821086, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0796, + "step": 1189 + }, + { + "epoch": 0.950479233226837, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.078, + "step": 1190 + }, + { + "epoch": 0.9512779552715654, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.0792, + "step": 1191 + }, + { + "epoch": 0.952076677316294, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0821, + "step": 1192 + }, + { + "epoch": 0.9528753993610224, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.073, + "step": 1193 + }, + { + "epoch": 0.9536741214057508, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 1194 + }, + { + "epoch": 0.9544728434504792, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 1195 + }, + { + "epoch": 0.9552715654952076, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.085, + "step": 1196 + }, + { + "epoch": 0.9560702875399361, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.0856, + "step": 1197 + }, + { + "epoch": 0.9568690095846646, + "grad_norm": 0.185546875, + "learning_rate": 0.0005, + "loss": 1.08, + "step": 1198 + }, + { + "epoch": 0.957667731629393, + "grad_norm": 0.2333984375, + "learning_rate": 0.0005, + "loss": 1.0859, + "step": 1199 + }, + { + "epoch": 0.9584664536741214, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0902, + "step": 1200 + }, + { + "epoch": 0.9592651757188498, + "grad_norm": 0.34375, + "learning_rate": 0.0005, + "loss": 1.0886, + "step": 1201 + }, + { + "epoch": 0.9600638977635783, + "grad_norm": 0.1884765625, + "learning_rate": 0.0005, + "loss": 1.0847, + "step": 1202 + }, + { + "epoch": 0.9608626198083067, + "grad_norm": 0.1953125, + "learning_rate": 0.0005, + "loss": 1.0882, + "step": 1203 + }, + { + "epoch": 0.9616613418530351, + "grad_norm": 0.24609375, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 1204 + }, + { + "epoch": 0.9624600638977636, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0816, + "step": 1205 + }, + { + "epoch": 0.963258785942492, + "grad_norm": 0.251953125, + "learning_rate": 0.0005, + "loss": 1.0798, + "step": 1206 + }, + { + "epoch": 0.9640575079872205, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0735, + "step": 1207 + }, + { + "epoch": 0.9648562300319489, + "grad_norm": 0.1943359375, + "learning_rate": 0.0005, + "loss": 1.0796, + "step": 1208 + }, + { + "epoch": 0.9656549520766773, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0827, + "step": 1209 + }, + { + "epoch": 0.9664536741214057, + "grad_norm": 0.1689453125, + "learning_rate": 0.0005, + "loss": 1.0843, + "step": 1210 + }, + { + "epoch": 0.9672523961661342, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0894, + "step": 1211 + }, + { + "epoch": 0.9680511182108626, + "grad_norm": 0.1689453125, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 1212 + }, + { + "epoch": 0.9688498402555911, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0844, + "step": 1213 + }, + { + "epoch": 0.9696485623003195, + "grad_norm": 0.21484375, + "learning_rate": 0.0005, + "loss": 1.0878, + "step": 1214 + }, + { + "epoch": 0.9704472843450479, + "grad_norm": 0.251953125, + "learning_rate": 0.0005, + "loss": 1.0859, + "step": 1215 + }, + { + "epoch": 0.9712460063897763, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0787, + "step": 1216 + }, + { + "epoch": 0.9720447284345048, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0799, + "step": 1217 + }, + { + "epoch": 0.9728434504792333, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0901, + "step": 1218 + }, + { + "epoch": 0.9736421725239617, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 1219 + }, + { + "epoch": 0.9744408945686901, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.0844, + "step": 1220 + }, + { + "epoch": 0.9752396166134185, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0789, + "step": 1221 + }, + { + "epoch": 0.9760383386581469, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.085, + "step": 1222 + }, + { + "epoch": 0.9768370607028753, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0848, + "step": 1223 + }, + { + "epoch": 0.9776357827476039, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0836, + "step": 1224 + }, + { + "epoch": 0.9784345047923323, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0819, + "step": 1225 + }, + { + "epoch": 0.9792332268370607, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 1226 + }, + { + "epoch": 0.9800319488817891, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0808, + "step": 1227 + }, + { + "epoch": 0.9808306709265175, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0764, + "step": 1228 + }, + { + "epoch": 0.981629392971246, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0716, + "step": 1229 + }, + { + "epoch": 0.9824281150159745, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0824, + "step": 1230 + }, + { + "epoch": 0.9832268370607029, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0834, + "step": 1231 + }, + { + "epoch": 0.9840255591054313, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0845, + "step": 1232 + }, + { + "epoch": 0.9848242811501597, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0845, + "step": 1233 + }, + { + "epoch": 0.9856230031948882, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.084, + "step": 1234 + }, + { + "epoch": 0.9864217252396166, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.075, + "step": 1235 + }, + { + "epoch": 0.987220447284345, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.073, + "step": 1236 + }, + { + "epoch": 0.9880191693290735, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0834, + "step": 1237 + }, + { + "epoch": 0.9888178913738019, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0758, + "step": 1238 + }, + { + "epoch": 0.9896166134185304, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.076, + "step": 1239 + }, + { + "epoch": 0.9904153354632588, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0785, + "step": 1240 + }, + { + "epoch": 0.9912140575079872, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0899, + "step": 1241 + }, + { + "epoch": 0.9920127795527156, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0775, + "step": 1242 + }, + { + "epoch": 0.9928115015974441, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.0783, + "step": 1243 + }, + { + "epoch": 0.9936102236421726, + "grad_norm": 0.2392578125, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 1244 + }, + { + "epoch": 0.994408945686901, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0795, + "step": 1245 + }, + { + "epoch": 0.9952076677316294, + "grad_norm": 0.1982421875, + "learning_rate": 0.0005, + "loss": 1.0939, + "step": 1246 + }, + { + "epoch": 0.9960063897763578, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0856, + "step": 1247 + }, + { + "epoch": 0.9968051118210862, + "grad_norm": 0.16796875, + "learning_rate": 0.0005, + "loss": 1.0844, + "step": 1248 + }, + { + "epoch": 0.9976038338658147, + "grad_norm": 0.1572265625, + "learning_rate": 0.0005, + "loss": 1.0835, + "step": 1249 + }, + { + "epoch": 0.9984025559105432, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.0761, + "step": 1250 + }, + { + "epoch": 0.9992012779552716, + "grad_norm": 0.2177734375, + "learning_rate": 0.0005, + "loss": 1.0761, + "step": 1251 + }, + { + "epoch": 1.0, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0793, + "step": 1252 + }, + { + "epoch": 1.0007987220447285, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0799, + "step": 1253 + }, + { + "epoch": 1.0015974440894568, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0818, + "step": 1254 + }, + { + "epoch": 1.0023961661341854, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.084, + "step": 1255 + }, + { + "epoch": 1.0031948881789137, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0899, + "step": 1256 + }, + { + "epoch": 1.0039936102236422, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0779, + "step": 1257 + }, + { + "epoch": 1.0047923322683705, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0794, + "step": 1258 + }, + { + "epoch": 1.005591054313099, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.082, + "step": 1259 + }, + { + "epoch": 1.0063897763578276, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0912, + "step": 1260 + }, + { + "epoch": 1.0071884984025559, + "grad_norm": 0.2001953125, + "learning_rate": 0.0005, + "loss": 1.0751, + "step": 1261 + }, + { + "epoch": 1.0079872204472844, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0827, + "step": 1262 + }, + { + "epoch": 1.0087859424920127, + "grad_norm": 0.19140625, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 1263 + }, + { + "epoch": 1.0095846645367412, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 1264 + }, + { + "epoch": 1.0103833865814698, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0811, + "step": 1265 + }, + { + "epoch": 1.011182108626198, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0774, + "step": 1266 + }, + { + "epoch": 1.0119808306709266, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0795, + "step": 1267 + }, + { + "epoch": 1.012779552715655, + "grad_norm": 0.1630859375, + "learning_rate": 0.0005, + "loss": 1.0866, + "step": 1268 + }, + { + "epoch": 1.0135782747603834, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0744, + "step": 1269 + }, + { + "epoch": 1.0143769968051117, + "grad_norm": 0.228515625, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 1270 + }, + { + "epoch": 1.0151757188498403, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0755, + "step": 1271 + }, + { + "epoch": 1.0159744408945688, + "grad_norm": 0.16796875, + "learning_rate": 0.0005, + "loss": 1.0849, + "step": 1272 + }, + { + "epoch": 1.016773162939297, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0815, + "step": 1273 + }, + { + "epoch": 1.0175718849840256, + "grad_norm": 0.169921875, + "learning_rate": 0.0005, + "loss": 1.0729, + "step": 1274 + }, + { + "epoch": 1.018370607028754, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0805, + "step": 1275 + }, + { + "epoch": 1.0191693290734825, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 1276 + }, + { + "epoch": 1.0199680511182108, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.074, + "step": 1277 + }, + { + "epoch": 1.0207667731629393, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0781, + "step": 1278 + }, + { + "epoch": 1.0215654952076678, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 1279 + }, + { + "epoch": 1.0223642172523961, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0847, + "step": 1280 + }, + { + "epoch": 1.0231629392971247, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0866, + "step": 1281 + }, + { + "epoch": 1.023961661341853, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 1282 + }, + { + "epoch": 1.0247603833865815, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 1283 + }, + { + "epoch": 1.0255591054313098, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0804, + "step": 1284 + }, + { + "epoch": 1.0263578274760383, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0793, + "step": 1285 + }, + { + "epoch": 1.0271565495207668, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.078, + "step": 1286 + }, + { + "epoch": 1.0279552715654952, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0866, + "step": 1287 + }, + { + "epoch": 1.0287539936102237, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0778, + "step": 1288 + }, + { + "epoch": 1.029552715654952, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0821, + "step": 1289 + }, + { + "epoch": 1.0303514376996805, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 1290 + }, + { + "epoch": 1.031150159744409, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.077, + "step": 1291 + }, + { + "epoch": 1.0319488817891374, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0788, + "step": 1292 + }, + { + "epoch": 1.0327476038338659, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0751, + "step": 1293 + }, + { + "epoch": 1.0335463258785942, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0761, + "step": 1294 + }, + { + "epoch": 1.0343450479233227, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0765, + "step": 1295 + }, + { + "epoch": 1.035143769968051, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0852, + "step": 1296 + }, + { + "epoch": 1.0359424920127795, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 1297 + }, + { + "epoch": 1.036741214057508, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0763, + "step": 1298 + }, + { + "epoch": 1.0375399361022364, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 1299 + }, + { + "epoch": 1.038338658146965, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 1300 + }, + { + "epoch": 1.0391373801916932, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 1301 + }, + { + "epoch": 1.0399361022364217, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0763, + "step": 1302 + }, + { + "epoch": 1.04073482428115, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 1303 + }, + { + "epoch": 1.0415335463258786, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.074, + "step": 1304 + }, + { + "epoch": 1.042332268370607, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0794, + "step": 1305 + }, + { + "epoch": 1.0431309904153354, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0781, + "step": 1306 + }, + { + "epoch": 1.043929712460064, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0775, + "step": 1307 + }, + { + "epoch": 1.0447284345047922, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0764, + "step": 1308 + }, + { + "epoch": 1.0455271565495208, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0831, + "step": 1309 + }, + { + "epoch": 1.0463258785942493, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0747, + "step": 1310 + }, + { + "epoch": 1.0471246006389776, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 1311 + }, + { + "epoch": 1.0479233226837061, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 1312 + }, + { + "epoch": 1.0487220447284344, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0783, + "step": 1313 + }, + { + "epoch": 1.049520766773163, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.073, + "step": 1314 + }, + { + "epoch": 1.0503194888178913, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 1315 + }, + { + "epoch": 1.0511182108626198, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0831, + "step": 1316 + }, + { + "epoch": 1.0519169329073483, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0862, + "step": 1317 + }, + { + "epoch": 1.0527156549520766, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0861, + "step": 1318 + }, + { + "epoch": 1.0535143769968052, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0835, + "step": 1319 + }, + { + "epoch": 1.0543130990415335, + "grad_norm": 0.1572265625, + "learning_rate": 0.0005, + "loss": 1.0818, + "step": 1320 + }, + { + "epoch": 1.055111821086262, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0782, + "step": 1321 + }, + { + "epoch": 1.0559105431309903, + "grad_norm": 0.203125, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 1322 + }, + { + "epoch": 1.0567092651757188, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0763, + "step": 1323 + }, + { + "epoch": 1.0575079872204474, + "grad_norm": 0.1943359375, + "learning_rate": 0.0005, + "loss": 1.0787, + "step": 1324 + }, + { + "epoch": 1.0583067092651757, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0768, + "step": 1325 + }, + { + "epoch": 1.0591054313099042, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0754, + "step": 1326 + }, + { + "epoch": 1.0599041533546325, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0761, + "step": 1327 + }, + { + "epoch": 1.060702875399361, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0858, + "step": 1328 + }, + { + "epoch": 1.0615015974440896, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.0829, + "step": 1329 + }, + { + "epoch": 1.0623003194888179, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0815, + "step": 1330 + }, + { + "epoch": 1.0630990415335464, + "grad_norm": 0.1728515625, + "learning_rate": 0.0005, + "loss": 1.0766, + "step": 1331 + }, + { + "epoch": 1.0638977635782747, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0801, + "step": 1332 + }, + { + "epoch": 1.0646964856230032, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0762, + "step": 1333 + }, + { + "epoch": 1.0654952076677315, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0847, + "step": 1334 + }, + { + "epoch": 1.06629392971246, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 1335 + }, + { + "epoch": 1.0670926517571886, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0786, + "step": 1336 + }, + { + "epoch": 1.067891373801917, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0795, + "step": 1337 + }, + { + "epoch": 1.0686900958466454, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0762, + "step": 1338 + }, + { + "epoch": 1.0694888178913737, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0764, + "step": 1339 + }, + { + "epoch": 1.0702875399361023, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0736, + "step": 1340 + }, + { + "epoch": 1.0710862619808306, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0804, + "step": 1341 + }, + { + "epoch": 1.071884984025559, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.0821, + "step": 1342 + }, + { + "epoch": 1.0726837060702876, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 1343 + }, + { + "epoch": 1.073482428115016, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0816, + "step": 1344 + }, + { + "epoch": 1.0742811501597445, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 1345 + }, + { + "epoch": 1.0750798722044728, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0779, + "step": 1346 + }, + { + "epoch": 1.0758785942492013, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 1347 + }, + { + "epoch": 1.0766773162939298, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 1348 + }, + { + "epoch": 1.0774760383386581, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0848, + "step": 1349 + }, + { + "epoch": 1.0782747603833867, + "grad_norm": 0.236328125, + "learning_rate": 0.0005, + "loss": 1.0768, + "step": 1350 + }, + { + "epoch": 1.079073482428115, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0792, + "step": 1351 + }, + { + "epoch": 1.0798722044728435, + "grad_norm": 0.1826171875, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 1352 + }, + { + "epoch": 1.0806709265175718, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0838, + "step": 1353 + }, + { + "epoch": 1.0814696485623003, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.0786, + "step": 1354 + }, + { + "epoch": 1.0822683706070289, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.0822, + "step": 1355 + }, + { + "epoch": 1.0830670926517572, + "grad_norm": 0.2890625, + "learning_rate": 0.0005, + "loss": 1.074, + "step": 1356 + }, + { + "epoch": 1.0838658146964857, + "grad_norm": 0.267578125, + "learning_rate": 0.0005, + "loss": 1.0815, + "step": 1357 + }, + { + "epoch": 1.084664536741214, + "grad_norm": 0.1884765625, + "learning_rate": 0.0005, + "loss": 1.0796, + "step": 1358 + }, + { + "epoch": 1.0854632587859425, + "grad_norm": 0.2021484375, + "learning_rate": 0.0005, + "loss": 1.0759, + "step": 1359 + }, + { + "epoch": 1.0862619808306708, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0737, + "step": 1360 + }, + { + "epoch": 1.0870607028753994, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0834, + "step": 1361 + }, + { + "epoch": 1.0878594249201279, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 1362 + }, + { + "epoch": 1.0886581469648562, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0779, + "step": 1363 + }, + { + "epoch": 1.0894568690095847, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0816, + "step": 1364 + }, + { + "epoch": 1.090255591054313, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 1365 + }, + { + "epoch": 1.0910543130990416, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0798, + "step": 1366 + }, + { + "epoch": 1.09185303514377, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 1367 + }, + { + "epoch": 1.0926517571884984, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 1368 + }, + { + "epoch": 1.093450479233227, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0836, + "step": 1369 + }, + { + "epoch": 1.0942492012779552, + "grad_norm": 0.21484375, + "learning_rate": 0.0005, + "loss": 1.0813, + "step": 1370 + }, + { + "epoch": 1.0950479233226837, + "grad_norm": 0.2412109375, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 1371 + }, + { + "epoch": 1.095846645367412, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.0859, + "step": 1372 + }, + { + "epoch": 1.0966453674121406, + "grad_norm": 0.369140625, + "learning_rate": 0.0005, + "loss": 1.0804, + "step": 1373 + }, + { + "epoch": 1.097444089456869, + "grad_norm": 0.25, + "learning_rate": 0.0005, + "loss": 1.075, + "step": 1374 + }, + { + "epoch": 1.0982428115015974, + "grad_norm": 0.17578125, + "learning_rate": 0.0005, + "loss": 1.0799, + "step": 1375 + }, + { + "epoch": 1.099041533546326, + "grad_norm": 0.1884765625, + "learning_rate": 0.0005, + "loss": 1.0785, + "step": 1376 + }, + { + "epoch": 1.0998402555910542, + "grad_norm": 0.1572265625, + "learning_rate": 0.0005, + "loss": 1.0802, + "step": 1377 + }, + { + "epoch": 1.1006389776357828, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0735, + "step": 1378 + }, + { + "epoch": 1.101437699680511, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0792, + "step": 1379 + }, + { + "epoch": 1.1022364217252396, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0861, + "step": 1380 + }, + { + "epoch": 1.1030351437699681, + "grad_norm": 0.1806640625, + "learning_rate": 0.0005, + "loss": 1.0808, + "step": 1381 + }, + { + "epoch": 1.1038338658146964, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0764, + "step": 1382 + }, + { + "epoch": 1.104632587859425, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 1383 + }, + { + "epoch": 1.1054313099041533, + "grad_norm": 0.29296875, + "learning_rate": 0.0005, + "loss": 1.0798, + "step": 1384 + }, + { + "epoch": 1.1062300319488818, + "grad_norm": 0.275390625, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 1385 + }, + { + "epoch": 1.1070287539936103, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0753, + "step": 1386 + }, + { + "epoch": 1.1078274760383386, + "grad_norm": 0.30078125, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 1387 + }, + { + "epoch": 1.1086261980830672, + "grad_norm": 0.27734375, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 1388 + }, + { + "epoch": 1.1094249201277955, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0764, + "step": 1389 + }, + { + "epoch": 1.110223642172524, + "grad_norm": 0.1689453125, + "learning_rate": 0.0005, + "loss": 1.0778, + "step": 1390 + }, + { + "epoch": 1.1110223642172523, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 1391 + }, + { + "epoch": 1.1118210862619808, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0744, + "step": 1392 + }, + { + "epoch": 1.1126198083067094, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0752, + "step": 1393 + }, + { + "epoch": 1.1134185303514377, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0828, + "step": 1394 + }, + { + "epoch": 1.1142172523961662, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0807, + "step": 1395 + }, + { + "epoch": 1.1150159744408945, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0792, + "step": 1396 + }, + { + "epoch": 1.115814696485623, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0797, + "step": 1397 + }, + { + "epoch": 1.1166134185303513, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0837, + "step": 1398 + }, + { + "epoch": 1.1174121405750799, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 1399 + }, + { + "epoch": 1.1182108626198084, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 1400 + }, + { + "epoch": 1.1190095846645367, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0798, + "step": 1401 + }, + { + "epoch": 1.1198083067092652, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0774, + "step": 1402 + }, + { + "epoch": 1.1206070287539935, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0873, + "step": 1403 + }, + { + "epoch": 1.121405750798722, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0834, + "step": 1404 + }, + { + "epoch": 1.1222044728434506, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0744, + "step": 1405 + }, + { + "epoch": 1.123003194888179, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0783, + "step": 1406 + }, + { + "epoch": 1.1238019169329074, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0796, + "step": 1407 + }, + { + "epoch": 1.1246006389776357, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.074, + "step": 1408 + }, + { + "epoch": 1.1253993610223643, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.084, + "step": 1409 + }, + { + "epoch": 1.1261980830670926, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 1410 + }, + { + "epoch": 1.126996805111821, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0873, + "step": 1411 + }, + { + "epoch": 1.1277955271565494, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0802, + "step": 1412 + }, + { + "epoch": 1.128594249201278, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0736, + "step": 1413 + }, + { + "epoch": 1.1293929712460065, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 1414 + }, + { + "epoch": 1.1301916932907348, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 1415 + }, + { + "epoch": 1.1309904153354633, + "grad_norm": 0.2041015625, + "learning_rate": 0.0005, + "loss": 1.0868, + "step": 1416 + }, + { + "epoch": 1.1317891373801916, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.0768, + "step": 1417 + }, + { + "epoch": 1.1325878594249201, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 1418 + }, + { + "epoch": 1.1333865814696487, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 1419 + }, + { + "epoch": 1.134185303514377, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0749, + "step": 1420 + }, + { + "epoch": 1.1349840255591055, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 1421 + }, + { + "epoch": 1.1357827476038338, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.0772, + "step": 1422 + }, + { + "epoch": 1.1365814696485623, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.0794, + "step": 1423 + }, + { + "epoch": 1.1373801916932909, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 1424 + }, + { + "epoch": 1.1381789137380192, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0764, + "step": 1425 + }, + { + "epoch": 1.1389776357827477, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 1426 + }, + { + "epoch": 1.139776357827476, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.079, + "step": 1427 + }, + { + "epoch": 1.1405750798722045, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 1428 + }, + { + "epoch": 1.1413738019169328, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.081, + "step": 1429 + }, + { + "epoch": 1.1421725239616614, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0827, + "step": 1430 + }, + { + "epoch": 1.1429712460063897, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0785, + "step": 1431 + }, + { + "epoch": 1.1437699680511182, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0791, + "step": 1432 + }, + { + "epoch": 1.1445686900958467, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 1433 + }, + { + "epoch": 1.145367412140575, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 1434 + }, + { + "epoch": 1.1461661341853036, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.075, + "step": 1435 + }, + { + "epoch": 1.1469648562300319, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0801, + "step": 1436 + }, + { + "epoch": 1.1477635782747604, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0759, + "step": 1437 + }, + { + "epoch": 1.148562300319489, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0843, + "step": 1438 + }, + { + "epoch": 1.1493610223642172, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0801, + "step": 1439 + }, + { + "epoch": 1.1501597444089458, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0855, + "step": 1440 + }, + { + "epoch": 1.150958466453674, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 1441 + }, + { + "epoch": 1.1517571884984026, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 1442 + }, + { + "epoch": 1.1525559105431311, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0855, + "step": 1443 + }, + { + "epoch": 1.1533546325878594, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0822, + "step": 1444 + }, + { + "epoch": 1.154153354632588, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0726, + "step": 1445 + }, + { + "epoch": 1.1549520766773163, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 1446 + }, + { + "epoch": 1.1557507987220448, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 1447 + }, + { + "epoch": 1.156549520766773, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0798, + "step": 1448 + }, + { + "epoch": 1.1573482428115016, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 1449 + }, + { + "epoch": 1.15814696485623, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 1450 + }, + { + "epoch": 1.1589456869009584, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0719, + "step": 1451 + }, + { + "epoch": 1.159744408945687, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 1452 + }, + { + "epoch": 1.1605431309904153, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0747, + "step": 1453 + }, + { + "epoch": 1.1613418530351438, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 1454 + }, + { + "epoch": 1.1621405750798721, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0831, + "step": 1455 + }, + { + "epoch": 1.1629392971246006, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 1456 + }, + { + "epoch": 1.1637380191693292, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0775, + "step": 1457 + }, + { + "epoch": 1.1645367412140575, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.079, + "step": 1458 + }, + { + "epoch": 1.165335463258786, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0755, + "step": 1459 + }, + { + "epoch": 1.1661341853035143, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0845, + "step": 1460 + }, + { + "epoch": 1.1669329073482428, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0889, + "step": 1461 + }, + { + "epoch": 1.1677316293929714, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0736, + "step": 1462 + }, + { + "epoch": 1.1685303514376997, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.079, + "step": 1463 + }, + { + "epoch": 1.1693290734824282, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.073, + "step": 1464 + }, + { + "epoch": 1.1701277955271565, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.075, + "step": 1465 + }, + { + "epoch": 1.170926517571885, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0761, + "step": 1466 + }, + { + "epoch": 1.1717252396166133, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 1467 + }, + { + "epoch": 1.1725239616613419, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 1468 + }, + { + "epoch": 1.1733226837060702, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0761, + "step": 1469 + }, + { + "epoch": 1.1741214057507987, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0802, + "step": 1470 + }, + { + "epoch": 1.1749201277955272, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0744, + "step": 1471 + }, + { + "epoch": 1.1757188498402555, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0736, + "step": 1472 + }, + { + "epoch": 1.176517571884984, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 1473 + }, + { + "epoch": 1.1773162939297124, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 1474 + }, + { + "epoch": 1.178115015974441, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0783, + "step": 1475 + }, + { + "epoch": 1.1789137380191694, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0832, + "step": 1476 + }, + { + "epoch": 1.1797124600638977, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 1477 + }, + { + "epoch": 1.1805111821086263, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0761, + "step": 1478 + }, + { + "epoch": 1.1813099041533546, + "grad_norm": 0.1640625, + "learning_rate": 0.0005, + "loss": 1.0761, + "step": 1479 + }, + { + "epoch": 1.182108626198083, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0847, + "step": 1480 + }, + { + "epoch": 1.1829073482428114, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 1481 + }, + { + "epoch": 1.18370607028754, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0729, + "step": 1482 + }, + { + "epoch": 1.1845047923322685, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 1483 + }, + { + "epoch": 1.1853035143769968, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0785, + "step": 1484 + }, + { + "epoch": 1.1861022364217253, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0813, + "step": 1485 + }, + { + "epoch": 1.1869009584664536, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 1486 + }, + { + "epoch": 1.1876996805111821, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 1487 + }, + { + "epoch": 1.1884984025559104, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 1488 + }, + { + "epoch": 1.189297124600639, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0832, + "step": 1489 + }, + { + "epoch": 1.1900958466453675, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 1490 + }, + { + "epoch": 1.1908945686900958, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0774, + "step": 1491 + }, + { + "epoch": 1.1916932907348243, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 1492 + }, + { + "epoch": 1.1924920127795526, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0801, + "step": 1493 + }, + { + "epoch": 1.1932907348242812, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 1494 + }, + { + "epoch": 1.1940894568690097, + "grad_norm": 0.2392578125, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 1495 + }, + { + "epoch": 1.194888178913738, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 1496 + }, + { + "epoch": 1.1956869009584665, + "grad_norm": 0.1923828125, + "learning_rate": 0.0005, + "loss": 1.0769, + "step": 1497 + }, + { + "epoch": 1.1964856230031948, + "grad_norm": 0.263671875, + "learning_rate": 0.0005, + "loss": 1.076, + "step": 1498 + }, + { + "epoch": 1.1972843450479234, + "grad_norm": 0.197265625, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 1499 + }, + { + "epoch": 1.1980830670926517, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0819, + "step": 1500 + }, + { + "epoch": 1.1988817891373802, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0803, + "step": 1501 + }, + { + "epoch": 1.1996805111821087, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0753, + "step": 1502 + }, + { + "epoch": 1.200479233226837, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0729, + "step": 1503 + }, + { + "epoch": 1.2012779552715656, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 1504 + }, + { + "epoch": 1.2020766773162939, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0769, + "step": 1505 + }, + { + "epoch": 1.2028753993610224, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.0852, + "step": 1506 + }, + { + "epoch": 1.2036741214057507, + "grad_norm": 0.2197265625, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 1507 + }, + { + "epoch": 1.2044728434504792, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.078, + "step": 1508 + }, + { + "epoch": 1.2052715654952078, + "grad_norm": 0.349609375, + "learning_rate": 0.0005, + "loss": 1.0766, + "step": 1509 + }, + { + "epoch": 1.206070287539936, + "grad_norm": 0.373046875, + "learning_rate": 0.0005, + "loss": 1.0754, + "step": 1510 + }, + { + "epoch": 1.2068690095846646, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 1511 + }, + { + "epoch": 1.207667731629393, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 1512 + }, + { + "epoch": 1.2084664536741214, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0762, + "step": 1513 + }, + { + "epoch": 1.20926517571885, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 1514 + }, + { + "epoch": 1.2100638977635783, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 1515 + }, + { + "epoch": 1.2108626198083068, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0735, + "step": 1516 + }, + { + "epoch": 1.211661341853035, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0792, + "step": 1517 + }, + { + "epoch": 1.2124600638977636, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 1518 + }, + { + "epoch": 1.213258785942492, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 1519 + }, + { + "epoch": 1.2140575079872205, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 1520 + }, + { + "epoch": 1.2148562300319488, + "grad_norm": 0.1806640625, + "learning_rate": 0.0005, + "loss": 1.0814, + "step": 1521 + }, + { + "epoch": 1.2156549520766773, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 1522 + }, + { + "epoch": 1.2164536741214058, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0783, + "step": 1523 + }, + { + "epoch": 1.2172523961661341, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0788, + "step": 1524 + }, + { + "epoch": 1.2180511182108626, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 1525 + }, + { + "epoch": 1.218849840255591, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0832, + "step": 1526 + }, + { + "epoch": 1.2196485623003195, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 1527 + }, + { + "epoch": 1.220447284345048, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0825, + "step": 1528 + }, + { + "epoch": 1.2212460063897763, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0795, + "step": 1529 + }, + { + "epoch": 1.2220447284345048, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 1530 + }, + { + "epoch": 1.2228434504792332, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 1531 + }, + { + "epoch": 1.2236421725239617, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.081, + "step": 1532 + }, + { + "epoch": 1.2244408945686902, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0775, + "step": 1533 + }, + { + "epoch": 1.2252396166134185, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 1534 + }, + { + "epoch": 1.226038338658147, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 1535 + }, + { + "epoch": 1.2268370607028753, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0849, + "step": 1536 + }, + { + "epoch": 1.2276357827476039, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0884, + "step": 1537 + }, + { + "epoch": 1.2284345047923322, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 1538 + }, + { + "epoch": 1.2292332268370607, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0782, + "step": 1539 + }, + { + "epoch": 1.230031948881789, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 1540 + }, + { + "epoch": 1.2308306709265175, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 1541 + }, + { + "epoch": 1.231629392971246, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 1542 + }, + { + "epoch": 1.2324281150159744, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 1543 + }, + { + "epoch": 1.233226837060703, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 1544 + }, + { + "epoch": 1.2340255591054312, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0737, + "step": 1545 + }, + { + "epoch": 1.2348242811501597, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0752, + "step": 1546 + }, + { + "epoch": 1.2356230031948883, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 1547 + }, + { + "epoch": 1.2364217252396166, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0807, + "step": 1548 + }, + { + "epoch": 1.237220447284345, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0748, + "step": 1549 + }, + { + "epoch": 1.2380191693290734, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 1550 + }, + { + "epoch": 1.238817891373802, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0788, + "step": 1551 + }, + { + "epoch": 1.2396166134185305, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 1552 + }, + { + "epoch": 1.2404153354632588, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0781, + "step": 1553 + }, + { + "epoch": 1.2412140575079873, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 1554 + }, + { + "epoch": 1.2420127795527156, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 1555 + }, + { + "epoch": 1.2428115015974441, + "grad_norm": 0.169921875, + "learning_rate": 0.0005, + "loss": 1.0735, + "step": 1556 + }, + { + "epoch": 1.2436102236421724, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 1557 + }, + { + "epoch": 1.244408945686901, + "grad_norm": 0.205078125, + "learning_rate": 0.0005, + "loss": 1.0795, + "step": 1558 + }, + { + "epoch": 1.2452076677316293, + "grad_norm": 0.283203125, + "learning_rate": 0.0005, + "loss": 1.0863, + "step": 1559 + }, + { + "epoch": 1.2460063897763578, + "grad_norm": 0.251953125, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 1560 + }, + { + "epoch": 1.2468051118210863, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0783, + "step": 1561 + }, + { + "epoch": 1.2476038338658146, + "grad_norm": 0.2177734375, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 1562 + }, + { + "epoch": 1.2484025559105432, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0805, + "step": 1563 + }, + { + "epoch": 1.2492012779552715, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0805, + "step": 1564 + }, + { + "epoch": 1.25, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 1565 + }, + { + "epoch": 1.2507987220447285, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0809, + "step": 1566 + }, + { + "epoch": 1.2515974440894568, + "grad_norm": 0.265625, + "learning_rate": 0.0005, + "loss": 1.0877, + "step": 1567 + }, + { + "epoch": 1.2523961661341854, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 1568 + }, + { + "epoch": 1.2531948881789137, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 1569 + }, + { + "epoch": 1.2539936102236422, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.076, + "step": 1570 + }, + { + "epoch": 1.2547923322683707, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0759, + "step": 1571 + }, + { + "epoch": 1.255591054313099, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 1572 + }, + { + "epoch": 1.2563897763578276, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0759, + "step": 1573 + }, + { + "epoch": 1.2571884984025559, + "grad_norm": 0.23046875, + "learning_rate": 0.0005, + "loss": 1.0789, + "step": 1574 + }, + { + "epoch": 1.2579872204472844, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0735, + "step": 1575 + }, + { + "epoch": 1.2587859424920127, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 1576 + }, + { + "epoch": 1.2595846645367412, + "grad_norm": 0.1611328125, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 1577 + }, + { + "epoch": 1.2603833865814695, + "grad_norm": 0.162109375, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 1578 + }, + { + "epoch": 1.261182108626198, + "grad_norm": 0.421875, + "learning_rate": 0.0005, + "loss": 1.078, + "step": 1579 + }, + { + "epoch": 1.2619808306709266, + "grad_norm": 0.625, + "learning_rate": 0.0005, + "loss": 1.0768, + "step": 1580 + }, + { + "epoch": 1.262779552715655, + "grad_norm": 0.85546875, + "learning_rate": 0.0005, + "loss": 1.0764, + "step": 1581 + }, + { + "epoch": 1.2635782747603834, + "grad_norm": 0.91796875, + "learning_rate": 0.0005, + "loss": 1.0864, + "step": 1582 + }, + { + "epoch": 1.2643769968051117, + "grad_norm": 0.5859375, + "learning_rate": 0.0005, + "loss": 1.0748, + "step": 1583 + }, + { + "epoch": 1.2651757188498403, + "grad_norm": 0.23828125, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 1584 + }, + { + "epoch": 1.2659744408945688, + "grad_norm": 0.7109375, + "learning_rate": 0.0005, + "loss": 1.0848, + "step": 1585 + }, + { + "epoch": 1.266773162939297, + "grad_norm": 0.310546875, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 1586 + }, + { + "epoch": 1.2675718849840256, + "grad_norm": 0.423828125, + "learning_rate": 0.0005, + "loss": 1.0794, + "step": 1587 + }, + { + "epoch": 1.268370607028754, + "grad_norm": 0.4921875, + "learning_rate": 0.0005, + "loss": 1.0761, + "step": 1588 + }, + { + "epoch": 1.2691693290734825, + "grad_norm": 0.1943359375, + "learning_rate": 0.0005, + "loss": 1.0759, + "step": 1589 + }, + { + "epoch": 1.269968051118211, + "grad_norm": 0.6171875, + "learning_rate": 0.0005, + "loss": 1.0764, + "step": 1590 + }, + { + "epoch": 1.2707667731629393, + "grad_norm": 0.4140625, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 1591 + }, + { + "epoch": 1.2715654952076676, + "grad_norm": 0.296875, + "learning_rate": 0.0005, + "loss": 1.0783, + "step": 1592 + }, + { + "epoch": 1.2723642172523961, + "grad_norm": 0.75, + "learning_rate": 0.0005, + "loss": 1.092, + "step": 1593 + }, + { + "epoch": 1.2731629392971247, + "grad_norm": 0.52734375, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 1594 + }, + { + "epoch": 1.273961661341853, + "grad_norm": 0.3203125, + "learning_rate": 0.0005, + "loss": 1.0781, + "step": 1595 + }, + { + "epoch": 1.2747603833865815, + "grad_norm": 0.73828125, + "learning_rate": 0.0005, + "loss": 1.0852, + "step": 1596 + }, + { + "epoch": 1.2755591054313098, + "grad_norm": 0.396484375, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 1597 + }, + { + "epoch": 1.2763578274760383, + "grad_norm": 0.48046875, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 1598 + }, + { + "epoch": 1.2771565495207668, + "grad_norm": 0.6640625, + "learning_rate": 0.0005, + "loss": 1.0815, + "step": 1599 + }, + { + "epoch": 1.2779552715654952, + "grad_norm": 0.177734375, + "learning_rate": 0.0005, + "loss": 1.0803, + "step": 1600 + }, + { + "epoch": 1.2787539936102237, + "grad_norm": 0.50390625, + "learning_rate": 0.0005, + "loss": 1.0751, + "step": 1601 + }, + { + "epoch": 1.279552715654952, + "grad_norm": 0.2734375, + "learning_rate": 0.0005, + "loss": 1.0759, + "step": 1602 + }, + { + "epoch": 1.2803514376996805, + "grad_norm": 0.25390625, + "learning_rate": 0.0005, + "loss": 1.0788, + "step": 1603 + }, + { + "epoch": 1.281150159744409, + "grad_norm": 0.2314453125, + "learning_rate": 0.0005, + "loss": 1.0736, + "step": 1604 + }, + { + "epoch": 1.2819488817891374, + "grad_norm": 0.169921875, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 1605 + }, + { + "epoch": 1.2827476038338659, + "grad_norm": 0.2099609375, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 1606 + }, + { + "epoch": 1.2835463258785942, + "grad_norm": 0.2099609375, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 1607 + }, + { + "epoch": 1.2843450479233227, + "grad_norm": 0.265625, + "learning_rate": 0.0005, + "loss": 1.075, + "step": 1608 + }, + { + "epoch": 1.2851437699680512, + "grad_norm": 0.1943359375, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 1609 + }, + { + "epoch": 1.2859424920127795, + "grad_norm": 0.380859375, + "learning_rate": 0.0005, + "loss": 1.0822, + "step": 1610 + }, + { + "epoch": 1.2867412140575079, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0826, + "step": 1611 + }, + { + "epoch": 1.2875399361022364, + "grad_norm": 0.23046875, + "learning_rate": 0.0005, + "loss": 1.0689, + "step": 1612 + }, + { + "epoch": 1.288338658146965, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0796, + "step": 1613 + }, + { + "epoch": 1.2891373801916932, + "grad_norm": 0.18359375, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 1614 + }, + { + "epoch": 1.2899361022364217, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 1615 + }, + { + "epoch": 1.29073482428115, + "grad_norm": 0.17578125, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 1616 + }, + { + "epoch": 1.2915335463258786, + "grad_norm": 0.1728515625, + "learning_rate": 0.0005, + "loss": 1.077, + "step": 1617 + }, + { + "epoch": 1.292332268370607, + "grad_norm": 0.28515625, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 1618 + }, + { + "epoch": 1.2931309904153354, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 1619 + }, + { + "epoch": 1.293929712460064, + "grad_norm": 0.205078125, + "learning_rate": 0.0005, + "loss": 1.0822, + "step": 1620 + }, + { + "epoch": 1.2947284345047922, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0752, + "step": 1621 + }, + { + "epoch": 1.2955271565495208, + "grad_norm": 0.1845703125, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 1622 + }, + { + "epoch": 1.2963258785942493, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0781, + "step": 1623 + }, + { + "epoch": 1.2971246006389776, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.0786, + "step": 1624 + }, + { + "epoch": 1.2979233226837061, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 1625 + }, + { + "epoch": 1.2987220447284344, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 1626 + }, + { + "epoch": 1.299520766773163, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 1627 + }, + { + "epoch": 1.3003194888178915, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 1628 + }, + { + "epoch": 1.3011182108626198, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 1629 + }, + { + "epoch": 1.3019169329073481, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0747, + "step": 1630 + }, + { + "epoch": 1.3027156549520766, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 1631 + }, + { + "epoch": 1.3035143769968052, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 1632 + }, + { + "epoch": 1.3043130990415335, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.0737, + "step": 1633 + }, + { + "epoch": 1.305111821086262, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 1634 + }, + { + "epoch": 1.3059105431309903, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0735, + "step": 1635 + }, + { + "epoch": 1.3067092651757188, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 1636 + }, + { + "epoch": 1.3075079872204474, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.08, + "step": 1637 + }, + { + "epoch": 1.3083067092651757, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0729, + "step": 1638 + }, + { + "epoch": 1.3091054313099042, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.077, + "step": 1639 + }, + { + "epoch": 1.3099041533546325, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.078, + "step": 1640 + }, + { + "epoch": 1.310702875399361, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0747, + "step": 1641 + }, + { + "epoch": 1.3115015974440896, + "grad_norm": 0.1533203125, + "learning_rate": 0.0005, + "loss": 1.0775, + "step": 1642 + }, + { + "epoch": 1.3123003194888179, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0754, + "step": 1643 + }, + { + "epoch": 1.3130990415335464, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 1644 + }, + { + "epoch": 1.3138977635782747, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 1645 + }, + { + "epoch": 1.3146964856230032, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 1646 + }, + { + "epoch": 1.3154952076677318, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0833, + "step": 1647 + }, + { + "epoch": 1.31629392971246, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 1648 + }, + { + "epoch": 1.3170926517571884, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 1649 + }, + { + "epoch": 1.317891373801917, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 1650 + }, + { + "epoch": 1.3186900958466454, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 1651 + }, + { + "epoch": 1.3194888178913737, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0754, + "step": 1652 + }, + { + "epoch": 1.3202875399361023, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0785, + "step": 1653 + }, + { + "epoch": 1.3210862619808306, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0755, + "step": 1654 + }, + { + "epoch": 1.321884984025559, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0798, + "step": 1655 + }, + { + "epoch": 1.3226837060702876, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0749, + "step": 1656 + }, + { + "epoch": 1.323482428115016, + "grad_norm": 0.1962890625, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 1657 + }, + { + "epoch": 1.3242811501597445, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 1658 + }, + { + "epoch": 1.3250798722044728, + "grad_norm": 0.1708984375, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 1659 + }, + { + "epoch": 1.3258785942492013, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 1660 + }, + { + "epoch": 1.3266773162939298, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 1661 + }, + { + "epoch": 1.3274760383386581, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 1662 + }, + { + "epoch": 1.3282747603833867, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0742, + "step": 1663 + }, + { + "epoch": 1.329073482428115, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 1664 + }, + { + "epoch": 1.3298722044728435, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0771, + "step": 1665 + }, + { + "epoch": 1.330670926517572, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 1666 + }, + { + "epoch": 1.3314696485623003, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 1667 + }, + { + "epoch": 1.3322683706070286, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0767, + "step": 1668 + }, + { + "epoch": 1.3330670926517572, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0723, + "step": 1669 + }, + { + "epoch": 1.3338658146964857, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 1670 + }, + { + "epoch": 1.334664536741214, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 1671 + }, + { + "epoch": 1.3354632587859425, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.078, + "step": 1672 + }, + { + "epoch": 1.3362619808306708, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 1673 + }, + { + "epoch": 1.3370607028753994, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 1674 + }, + { + "epoch": 1.3378594249201279, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 1675 + }, + { + "epoch": 1.3386581469648562, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 1676 + }, + { + "epoch": 1.3394568690095847, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 1677 + }, + { + "epoch": 1.340255591054313, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.074, + "step": 1678 + }, + { + "epoch": 1.3410543130990416, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 1679 + }, + { + "epoch": 1.34185303514377, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0754, + "step": 1680 + }, + { + "epoch": 1.3426517571884984, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 1681 + }, + { + "epoch": 1.343450479233227, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0744, + "step": 1682 + }, + { + "epoch": 1.3442492012779552, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 1683 + }, + { + "epoch": 1.3450479233226837, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 1684 + }, + { + "epoch": 1.3458466453674123, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0788, + "step": 1685 + }, + { + "epoch": 1.3466453674121406, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 1686 + }, + { + "epoch": 1.3474440894568689, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 1687 + }, + { + "epoch": 1.3482428115015974, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0781, + "step": 1688 + }, + { + "epoch": 1.349041533546326, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 1689 + }, + { + "epoch": 1.3498402555910542, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0686, + "step": 1690 + }, + { + "epoch": 1.3506389776357828, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 1691 + }, + { + "epoch": 1.351437699680511, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 1692 + }, + { + "epoch": 1.3522364217252396, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 1693 + }, + { + "epoch": 1.3530351437699681, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 1694 + }, + { + "epoch": 1.3538338658146964, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 1695 + }, + { + "epoch": 1.354632587859425, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0726, + "step": 1696 + }, + { + "epoch": 1.3554313099041533, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 1697 + }, + { + "epoch": 1.3562300319488818, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 1698 + }, + { + "epoch": 1.3570287539936103, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 1699 + }, + { + "epoch": 1.3578274760383386, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 1700 + }, + { + "epoch": 1.3586261980830672, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 1701 + }, + { + "epoch": 1.3594249201277955, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0754, + "step": 1702 + }, + { + "epoch": 1.360223642172524, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0758, + "step": 1703 + }, + { + "epoch": 1.3610223642172525, + "grad_norm": 0.162109375, + "learning_rate": 0.0005, + "loss": 1.079, + "step": 1704 + }, + { + "epoch": 1.3618210862619808, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0752, + "step": 1705 + }, + { + "epoch": 1.3626198083067091, + "grad_norm": 0.1669921875, + "learning_rate": 0.0005, + "loss": 1.0819, + "step": 1706 + }, + { + "epoch": 1.3634185303514377, + "grad_norm": 0.19140625, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 1707 + }, + { + "epoch": 1.3642172523961662, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 1708 + }, + { + "epoch": 1.3650159744408945, + "grad_norm": 0.1962890625, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 1709 + }, + { + "epoch": 1.365814696485623, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0764, + "step": 1710 + }, + { + "epoch": 1.3666134185303513, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0815, + "step": 1711 + }, + { + "epoch": 1.3674121405750799, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 1712 + }, + { + "epoch": 1.3682108626198084, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 1713 + }, + { + "epoch": 1.3690095846645367, + "grad_norm": 0.1689453125, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 1714 + }, + { + "epoch": 1.3698083067092652, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 1715 + }, + { + "epoch": 1.3706070287539935, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.073, + "step": 1716 + }, + { + "epoch": 1.371405750798722, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 1717 + }, + { + "epoch": 1.3722044728434506, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 1718 + }, + { + "epoch": 1.373003194888179, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 1719 + }, + { + "epoch": 1.3738019169329074, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.08, + "step": 1720 + }, + { + "epoch": 1.3746006389776357, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0802, + "step": 1721 + }, + { + "epoch": 1.3753993610223643, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0768, + "step": 1722 + }, + { + "epoch": 1.3761980830670926, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 1723 + }, + { + "epoch": 1.376996805111821, + "grad_norm": 0.1533203125, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 1724 + }, + { + "epoch": 1.3777955271565494, + "grad_norm": 0.1845703125, + "learning_rate": 0.0005, + "loss": 1.073, + "step": 1725 + }, + { + "epoch": 1.378594249201278, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 1726 + }, + { + "epoch": 1.3793929712460065, + "grad_norm": 0.234375, + "learning_rate": 0.0005, + "loss": 1.0818, + "step": 1727 + }, + { + "epoch": 1.3801916932907348, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0779, + "step": 1728 + }, + { + "epoch": 1.3809904153354633, + "grad_norm": 0.177734375, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 1729 + }, + { + "epoch": 1.3817891373801916, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 1730 + }, + { + "epoch": 1.3825878594249201, + "grad_norm": 0.279296875, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 1731 + }, + { + "epoch": 1.3833865814696487, + "grad_norm": 0.20703125, + "learning_rate": 0.0005, + "loss": 1.0752, + "step": 1732 + }, + { + "epoch": 1.384185303514377, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.082, + "step": 1733 + }, + { + "epoch": 1.3849840255591055, + "grad_norm": 0.240234375, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 1734 + }, + { + "epoch": 1.3857827476038338, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 1735 + }, + { + "epoch": 1.3865814696485623, + "grad_norm": 0.30078125, + "learning_rate": 0.0005, + "loss": 1.077, + "step": 1736 + }, + { + "epoch": 1.3873801916932909, + "grad_norm": 0.283203125, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 1737 + }, + { + "epoch": 1.3881789137380192, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 1738 + }, + { + "epoch": 1.3889776357827475, + "grad_norm": 0.318359375, + "learning_rate": 0.0005, + "loss": 1.075, + "step": 1739 + }, + { + "epoch": 1.389776357827476, + "grad_norm": 0.1533203125, + "learning_rate": 0.0005, + "loss": 1.0744, + "step": 1740 + }, + { + "epoch": 1.3905750798722045, + "grad_norm": 0.1728515625, + "learning_rate": 0.0005, + "loss": 1.0744, + "step": 1741 + }, + { + "epoch": 1.3913738019169328, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 1742 + }, + { + "epoch": 1.3921725239616614, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 1743 + }, + { + "epoch": 1.3929712460063897, + "grad_norm": 0.2080078125, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 1744 + }, + { + "epoch": 1.3937699680511182, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0761, + "step": 1745 + }, + { + "epoch": 1.3945686900958467, + "grad_norm": 0.1669921875, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 1746 + }, + { + "epoch": 1.395367412140575, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 1747 + }, + { + "epoch": 1.3961661341853036, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.078, + "step": 1748 + }, + { + "epoch": 1.3969648562300319, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 1749 + }, + { + "epoch": 1.3977635782747604, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 1750 + }, + { + "epoch": 1.398562300319489, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0744, + "step": 1751 + }, + { + "epoch": 1.3993610223642172, + "grad_norm": 0.1923828125, + "learning_rate": 0.0005, + "loss": 1.0749, + "step": 1752 + }, + { + "epoch": 1.4001597444089458, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0726, + "step": 1753 + }, + { + "epoch": 1.400958466453674, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.074, + "step": 1754 + }, + { + "epoch": 1.4017571884984026, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 1755 + }, + { + "epoch": 1.4025559105431311, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 1756 + }, + { + "epoch": 1.4033546325878594, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 1757 + }, + { + "epoch": 1.4041533546325877, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 1758 + }, + { + "epoch": 1.4049520766773163, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0765, + "step": 1759 + }, + { + "epoch": 1.4057507987220448, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 1760 + }, + { + "epoch": 1.406549520766773, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 1761 + }, + { + "epoch": 1.4073482428115016, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 1762 + }, + { + "epoch": 1.40814696485623, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 1763 + }, + { + "epoch": 1.4089456869009584, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0743, + "step": 1764 + }, + { + "epoch": 1.409744408945687, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0795, + "step": 1765 + }, + { + "epoch": 1.4105431309904153, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 1766 + }, + { + "epoch": 1.4113418530351438, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 1767 + }, + { + "epoch": 1.4121405750798721, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 1768 + }, + { + "epoch": 1.4129392971246006, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 1769 + }, + { + "epoch": 1.4137380191693292, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0752, + "step": 1770 + }, + { + "epoch": 1.4145367412140575, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 1771 + }, + { + "epoch": 1.415335463258786, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.0726, + "step": 1772 + }, + { + "epoch": 1.4161341853035143, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0758, + "step": 1773 + }, + { + "epoch": 1.4169329073482428, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0764, + "step": 1774 + }, + { + "epoch": 1.4177316293929714, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 1775 + }, + { + "epoch": 1.4185303514376997, + "grad_norm": 0.1572265625, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 1776 + }, + { + "epoch": 1.419329073482428, + "grad_norm": 0.1533203125, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 1777 + }, + { + "epoch": 1.4201277955271565, + "grad_norm": 0.216796875, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 1778 + }, + { + "epoch": 1.420926517571885, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 1779 + }, + { + "epoch": 1.4217252396166133, + "grad_norm": 0.185546875, + "learning_rate": 0.0005, + "loss": 1.0757, + "step": 1780 + }, + { + "epoch": 1.4225239616613419, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 1781 + }, + { + "epoch": 1.4233226837060702, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 1782 + }, + { + "epoch": 1.4241214057507987, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 1783 + }, + { + "epoch": 1.4249201277955272, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 1784 + }, + { + "epoch": 1.4257188498402555, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 1785 + }, + { + "epoch": 1.426517571884984, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 1786 + }, + { + "epoch": 1.4273162939297124, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 1787 + }, + { + "epoch": 1.428115015974441, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 1788 + }, + { + "epoch": 1.4289137380191694, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 1789 + }, + { + "epoch": 1.4297124600638977, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 1790 + }, + { + "epoch": 1.4305111821086263, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0788, + "step": 1791 + }, + { + "epoch": 1.4313099041533546, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 1792 + }, + { + "epoch": 1.432108626198083, + "grad_norm": 0.169921875, + "learning_rate": 0.0005, + "loss": 1.0686, + "step": 1793 + }, + { + "epoch": 1.4329073482428116, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 1794 + }, + { + "epoch": 1.43370607028754, + "grad_norm": 0.19921875, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 1795 + }, + { + "epoch": 1.4345047923322682, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.076, + "step": 1796 + }, + { + "epoch": 1.4353035143769968, + "grad_norm": 0.185546875, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 1797 + }, + { + "epoch": 1.4361022364217253, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 1798 + }, + { + "epoch": 1.4369009584664536, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 1799 + }, + { + "epoch": 1.4376996805111821, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0822, + "step": 1800 + }, + { + "epoch": 1.4384984025559104, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 1801 + }, + { + "epoch": 1.439297124600639, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 1802 + }, + { + "epoch": 1.4400958466453675, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 1803 + }, + { + "epoch": 1.4408945686900958, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 1804 + }, + { + "epoch": 1.4416932907348243, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 1805 + }, + { + "epoch": 1.4424920127795526, + "grad_norm": 0.1796875, + "learning_rate": 0.0005, + "loss": 1.0786, + "step": 1806 + }, + { + "epoch": 1.4432907348242812, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 1807 + }, + { + "epoch": 1.4440894568690097, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0779, + "step": 1808 + }, + { + "epoch": 1.444888178913738, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 1809 + }, + { + "epoch": 1.4456869009584665, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 1810 + }, + { + "epoch": 1.4464856230031948, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0728, + "step": 1811 + }, + { + "epoch": 1.4472843450479234, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 1812 + }, + { + "epoch": 1.4480830670926519, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0748, + "step": 1813 + }, + { + "epoch": 1.4488817891373802, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 1814 + }, + { + "epoch": 1.4496805111821085, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.074, + "step": 1815 + }, + { + "epoch": 1.450479233226837, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 1816 + }, + { + "epoch": 1.4512779552715656, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 1817 + }, + { + "epoch": 1.4520766773162939, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0806, + "step": 1818 + }, + { + "epoch": 1.4528753993610224, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 1819 + }, + { + "epoch": 1.4536741214057507, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 1820 + }, + { + "epoch": 1.4544728434504792, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0756, + "step": 1821 + }, + { + "epoch": 1.4552715654952078, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 1822 + }, + { + "epoch": 1.456070287539936, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 1823 + }, + { + "epoch": 1.4568690095846646, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 1824 + }, + { + "epoch": 1.457667731629393, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 1825 + }, + { + "epoch": 1.4584664536741214, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 1826 + }, + { + "epoch": 1.45926517571885, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 1827 + }, + { + "epoch": 1.4600638977635783, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 1828 + }, + { + "epoch": 1.4608626198083068, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 1829 + }, + { + "epoch": 1.461661341853035, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 1830 + }, + { + "epoch": 1.4624600638977636, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 1831 + }, + { + "epoch": 1.4632587859424921, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 1832 + }, + { + "epoch": 1.4640575079872205, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0793, + "step": 1833 + }, + { + "epoch": 1.4648562300319488, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0738, + "step": 1834 + }, + { + "epoch": 1.4656549520766773, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 1835 + }, + { + "epoch": 1.4664536741214058, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 1836 + }, + { + "epoch": 1.4672523961661341, + "grad_norm": 0.2001953125, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 1837 + }, + { + "epoch": 1.4680511182108626, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 1838 + }, + { + "epoch": 1.468849840255591, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0766, + "step": 1839 + }, + { + "epoch": 1.4696485623003195, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0758, + "step": 1840 + }, + { + "epoch": 1.470447284345048, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 1841 + }, + { + "epoch": 1.4712460063897763, + "grad_norm": 0.1875, + "learning_rate": 0.0005, + "loss": 1.0837, + "step": 1842 + }, + { + "epoch": 1.4720447284345048, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 1843 + }, + { + "epoch": 1.4728434504792332, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 1844 + }, + { + "epoch": 1.4736421725239617, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 1845 + }, + { + "epoch": 1.4744408945686902, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 1846 + }, + { + "epoch": 1.4752396166134185, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.0726, + "step": 1847 + }, + { + "epoch": 1.476038338658147, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 1848 + }, + { + "epoch": 1.4768370607028753, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0761, + "step": 1849 + }, + { + "epoch": 1.4776357827476039, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 1850 + }, + { + "epoch": 1.4784345047923324, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 1851 + }, + { + "epoch": 1.4792332268370607, + "grad_norm": 0.16015625, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 1852 + }, + { + "epoch": 1.480031948881789, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 1853 + }, + { + "epoch": 1.4808306709265175, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0813, + "step": 1854 + }, + { + "epoch": 1.481629392971246, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0744, + "step": 1855 + }, + { + "epoch": 1.4824281150159744, + "grad_norm": 0.1669921875, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 1856 + }, + { + "epoch": 1.483226837060703, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 1857 + }, + { + "epoch": 1.4840255591054312, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 1858 + }, + { + "epoch": 1.4848242811501597, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 1859 + }, + { + "epoch": 1.4856230031948883, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 1860 + }, + { + "epoch": 1.4864217252396166, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 1861 + }, + { + "epoch": 1.487220447284345, + "grad_norm": 0.1630859375, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 1862 + }, + { + "epoch": 1.4880191693290734, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 1863 + }, + { + "epoch": 1.488817891373802, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 1864 + }, + { + "epoch": 1.4896166134185305, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 1865 + }, + { + "epoch": 1.4904153354632588, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 1866 + }, + { + "epoch": 1.4912140575079873, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 1867 + }, + { + "epoch": 1.4920127795527156, + "grad_norm": 0.1826171875, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 1868 + }, + { + "epoch": 1.4928115015974441, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 1869 + }, + { + "epoch": 1.4936102236421724, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 1870 + }, + { + "epoch": 1.494408945686901, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 1871 + }, + { + "epoch": 1.4952076677316293, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 1872 + }, + { + "epoch": 1.4960063897763578, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0721, + "step": 1873 + }, + { + "epoch": 1.4968051118210863, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0755, + "step": 1874 + }, + { + "epoch": 1.4976038338658146, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 1875 + }, + { + "epoch": 1.4984025559105432, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 1876 + }, + { + "epoch": 1.4992012779552715, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 1877 + }, + { + "epoch": 1.5, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0717, + "step": 1878 + }, + { + "epoch": 1.5007987220447285, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 1879 + }, + { + "epoch": 1.5015974440894568, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 1880 + }, + { + "epoch": 1.5023961661341851, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 1881 + }, + { + "epoch": 1.5031948881789137, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 1882 + }, + { + "epoch": 1.5039936102236422, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 1883 + }, + { + "epoch": 1.5047923322683707, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 1884 + }, + { + "epoch": 1.505591054313099, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 1885 + }, + { + "epoch": 1.5063897763578273, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 1886 + }, + { + "epoch": 1.5071884984025559, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 1887 + }, + { + "epoch": 1.5079872204472844, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 1888 + }, + { + "epoch": 1.508785942492013, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 1889 + }, + { + "epoch": 1.5095846645367412, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 1890 + }, + { + "epoch": 1.5103833865814695, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.0744, + "step": 1891 + }, + { + "epoch": 1.511182108626198, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 1892 + }, + { + "epoch": 1.5119808306709266, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 1893 + }, + { + "epoch": 1.5127795527156551, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 1894 + }, + { + "epoch": 1.5135782747603834, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0766, + "step": 1895 + }, + { + "epoch": 1.5143769968051117, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 1896 + }, + { + "epoch": 1.5151757188498403, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 1897 + }, + { + "epoch": 1.5159744408945688, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 1898 + }, + { + "epoch": 1.516773162939297, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 1899 + }, + { + "epoch": 1.5175718849840254, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0693, + "step": 1900 + }, + { + "epoch": 1.518370607028754, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 1901 + }, + { + "epoch": 1.5191693290734825, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0782, + "step": 1902 + }, + { + "epoch": 1.519968051118211, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 1903 + }, + { + "epoch": 1.5207667731629393, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 1904 + }, + { + "epoch": 1.5215654952076676, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 1905 + }, + { + "epoch": 1.5223642172523961, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 1906 + }, + { + "epoch": 1.5231629392971247, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 1907 + }, + { + "epoch": 1.5239616613418532, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 1908 + }, + { + "epoch": 1.5247603833865815, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 1909 + }, + { + "epoch": 1.5255591054313098, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 1910 + }, + { + "epoch": 1.5263578274760383, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 1911 + }, + { + "epoch": 1.5271565495207668, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0789, + "step": 1912 + }, + { + "epoch": 1.5279552715654952, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 1913 + }, + { + "epoch": 1.5287539936102237, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 1914 + }, + { + "epoch": 1.529552715654952, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 1915 + }, + { + "epoch": 1.5303514376996805, + "grad_norm": 0.24609375, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 1916 + }, + { + "epoch": 1.531150159744409, + "grad_norm": 0.1669921875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 1917 + }, + { + "epoch": 1.5319488817891374, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 1918 + }, + { + "epoch": 1.5327476038338657, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 1919 + }, + { + "epoch": 1.5335463258785942, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 1920 + }, + { + "epoch": 1.5343450479233227, + "grad_norm": 0.2080078125, + "learning_rate": 0.0005, + "loss": 1.079, + "step": 1921 + }, + { + "epoch": 1.5351437699680512, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 1922 + }, + { + "epoch": 1.5359424920127795, + "grad_norm": 0.1669921875, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 1923 + }, + { + "epoch": 1.5367412140575079, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 1924 + }, + { + "epoch": 1.5375399361022364, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 1925 + }, + { + "epoch": 1.538338658146965, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.074, + "step": 1926 + }, + { + "epoch": 1.5391373801916934, + "grad_norm": 0.236328125, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 1927 + }, + { + "epoch": 1.5399361022364217, + "grad_norm": 0.234375, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 1928 + }, + { + "epoch": 1.54073482428115, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 1929 + }, + { + "epoch": 1.5415335463258786, + "grad_norm": 0.1884765625, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 1930 + }, + { + "epoch": 1.542332268370607, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 1931 + }, + { + "epoch": 1.5431309904153354, + "grad_norm": 0.189453125, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 1932 + }, + { + "epoch": 1.543929712460064, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 1933 + }, + { + "epoch": 1.5447284345047922, + "grad_norm": 0.2216796875, + "learning_rate": 0.0005, + "loss": 1.0741, + "step": 1934 + }, + { + "epoch": 1.5455271565495208, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 1935 + }, + { + "epoch": 1.5463258785942493, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 1936 + }, + { + "epoch": 1.5471246006389776, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0802, + "step": 1937 + }, + { + "epoch": 1.547923322683706, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 1938 + }, + { + "epoch": 1.5487220447284344, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 1939 + }, + { + "epoch": 1.549520766773163, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 1940 + }, + { + "epoch": 1.5503194888178915, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0793, + "step": 1941 + }, + { + "epoch": 1.5511182108626198, + "grad_norm": 0.1796875, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 1942 + }, + { + "epoch": 1.5519169329073481, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 1943 + }, + { + "epoch": 1.5527156549520766, + "grad_norm": 0.21484375, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 1944 + }, + { + "epoch": 1.5535143769968052, + "grad_norm": 0.21484375, + "learning_rate": 0.0005, + "loss": 1.0733, + "step": 1945 + }, + { + "epoch": 1.5543130990415337, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 1946 + }, + { + "epoch": 1.555111821086262, + "grad_norm": 0.365234375, + "learning_rate": 0.0005, + "loss": 1.0759, + "step": 1947 + }, + { + "epoch": 1.5559105431309903, + "grad_norm": 0.38671875, + "learning_rate": 0.0005, + "loss": 1.0751, + "step": 1948 + }, + { + "epoch": 1.5567092651757188, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 1949 + }, + { + "epoch": 1.5575079872204474, + "grad_norm": 0.1982421875, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 1950 + }, + { + "epoch": 1.5583067092651757, + "grad_norm": 0.18359375, + "learning_rate": 0.0005, + "loss": 1.0755, + "step": 1951 + }, + { + "epoch": 1.5591054313099042, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 1952 + }, + { + "epoch": 1.5599041533546325, + "grad_norm": 0.22265625, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 1953 + }, + { + "epoch": 1.560702875399361, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 1954 + }, + { + "epoch": 1.5615015974440896, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 1955 + }, + { + "epoch": 1.5623003194888179, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0749, + "step": 1956 + }, + { + "epoch": 1.5630990415335462, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 1957 + }, + { + "epoch": 1.5638977635782747, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0727, + "step": 1958 + }, + { + "epoch": 1.5646964856230032, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 1959 + }, + { + "epoch": 1.5654952076677318, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 1960 + }, + { + "epoch": 1.56629392971246, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 1961 + }, + { + "epoch": 1.5670926517571884, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 1962 + }, + { + "epoch": 1.567891373801917, + "grad_norm": 0.193359375, + "learning_rate": 0.0005, + "loss": 1.0709, + "step": 1963 + }, + { + "epoch": 1.5686900958466454, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 1964 + }, + { + "epoch": 1.569488817891374, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 1965 + }, + { + "epoch": 1.5702875399361023, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 1966 + }, + { + "epoch": 1.5710862619808306, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 1967 + }, + { + "epoch": 1.571884984025559, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 1968 + }, + { + "epoch": 1.5726837060702876, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 1969 + }, + { + "epoch": 1.573482428115016, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 1970 + }, + { + "epoch": 1.5742811501597445, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0782, + "step": 1971 + }, + { + "epoch": 1.5750798722044728, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0706, + "step": 1972 + }, + { + "epoch": 1.5758785942492013, + "grad_norm": 0.162109375, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 1973 + }, + { + "epoch": 1.5766773162939298, + "grad_norm": 0.2490234375, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 1974 + }, + { + "epoch": 1.5774760383386581, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 1975 + }, + { + "epoch": 1.5782747603833864, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 1976 + }, + { + "epoch": 1.579073482428115, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 1977 + }, + { + "epoch": 1.5798722044728435, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 1978 + }, + { + "epoch": 1.580670926517572, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 1979 + }, + { + "epoch": 1.5814696485623003, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 1980 + }, + { + "epoch": 1.5822683706070286, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 1981 + }, + { + "epoch": 1.5830670926517572, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 1982 + }, + { + "epoch": 1.5838658146964857, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 1983 + }, + { + "epoch": 1.5846645367412142, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 1984 + }, + { + "epoch": 1.5854632587859425, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 1985 + }, + { + "epoch": 1.5862619808306708, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 1986 + }, + { + "epoch": 1.5870607028753994, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 1987 + }, + { + "epoch": 1.5878594249201279, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0736, + "step": 1988 + }, + { + "epoch": 1.5886581469648562, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 1989 + }, + { + "epoch": 1.5894568690095847, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 1990 + }, + { + "epoch": 1.590255591054313, + "grad_norm": 0.171875, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 1991 + }, + { + "epoch": 1.5910543130990416, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 1992 + }, + { + "epoch": 1.59185303514377, + "grad_norm": 0.1806640625, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 1993 + }, + { + "epoch": 1.5926517571884984, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 1994 + }, + { + "epoch": 1.5934504792332267, + "grad_norm": 0.1826171875, + "learning_rate": 0.0005, + "loss": 1.0772, + "step": 1995 + }, + { + "epoch": 1.5942492012779552, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 1996 + }, + { + "epoch": 1.5950479233226837, + "grad_norm": 0.2119140625, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 1997 + }, + { + "epoch": 1.5958466453674123, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.078, + "step": 1998 + }, + { + "epoch": 1.5966453674121406, + "grad_norm": 0.1943359375, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 1999 + }, + { + "epoch": 1.5974440894568689, + "grad_norm": 0.1689453125, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 2000 + }, + { + "epoch": 1.5982428115015974, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 2001 + }, + { + "epoch": 1.599041533546326, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 2002 + }, + { + "epoch": 1.5998402555910545, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 2003 + }, + { + "epoch": 1.6006389776357828, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 2004 + }, + { + "epoch": 1.601437699680511, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 2005 + }, + { + "epoch": 1.6022364217252396, + "grad_norm": 0.1826171875, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 2006 + }, + { + "epoch": 1.6030351437699681, + "grad_norm": 0.349609375, + "learning_rate": 0.0005, + "loss": 1.0807, + "step": 2007 + }, + { + "epoch": 1.6038338658146964, + "grad_norm": 0.2216796875, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 2008 + }, + { + "epoch": 1.604632587859425, + "grad_norm": 0.2578125, + "learning_rate": 0.0005, + "loss": 1.0729, + "step": 2009 + }, + { + "epoch": 1.6054313099041533, + "grad_norm": 0.39453125, + "learning_rate": 0.0005, + "loss": 1.0737, + "step": 2010 + }, + { + "epoch": 1.6062300319488818, + "grad_norm": 0.35546875, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 2011 + }, + { + "epoch": 1.6070287539936103, + "grad_norm": 0.43359375, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 2012 + }, + { + "epoch": 1.6078274760383386, + "grad_norm": 0.33984375, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 2013 + }, + { + "epoch": 1.608626198083067, + "grad_norm": 0.21875, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 2014 + }, + { + "epoch": 1.6094249201277955, + "grad_norm": 0.388671875, + "learning_rate": 0.0005, + "loss": 1.0693, + "step": 2015 + }, + { + "epoch": 1.610223642172524, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 2016 + }, + { + "epoch": 1.6110223642172525, + "grad_norm": 0.314453125, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 2017 + }, + { + "epoch": 1.6118210862619808, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0732, + "step": 2018 + }, + { + "epoch": 1.6126198083067091, + "grad_norm": 0.2099609375, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 2019 + }, + { + "epoch": 1.6134185303514377, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0766, + "step": 2020 + }, + { + "epoch": 1.6142172523961662, + "grad_norm": 0.2080078125, + "learning_rate": 0.0005, + "loss": 1.0768, + "step": 2021 + }, + { + "epoch": 1.6150159744408947, + "grad_norm": 0.1533203125, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 2022 + }, + { + "epoch": 1.615814696485623, + "grad_norm": 0.2060546875, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 2023 + }, + { + "epoch": 1.6166134185303513, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 2024 + }, + { + "epoch": 1.6174121405750799, + "grad_norm": 0.1806640625, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 2025 + }, + { + "epoch": 1.6182108626198084, + "grad_norm": 0.220703125, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 2026 + }, + { + "epoch": 1.6190095846645367, + "grad_norm": 0.19140625, + "learning_rate": 0.0005, + "loss": 1.0739, + "step": 2027 + }, + { + "epoch": 1.619808306709265, + "grad_norm": 0.2080078125, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 2028 + }, + { + "epoch": 1.6206070287539935, + "grad_norm": 0.1767578125, + "learning_rate": 0.0005, + "loss": 1.0697, + "step": 2029 + }, + { + "epoch": 1.621405750798722, + "grad_norm": 0.2275390625, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 2030 + }, + { + "epoch": 1.6222044728434506, + "grad_norm": 0.2236328125, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 2031 + }, + { + "epoch": 1.623003194888179, + "grad_norm": 0.2158203125, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 2032 + }, + { + "epoch": 1.6238019169329072, + "grad_norm": 0.462890625, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 2033 + }, + { + "epoch": 1.6246006389776357, + "grad_norm": 0.2119140625, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 2034 + }, + { + "epoch": 1.6253993610223643, + "grad_norm": 0.1943359375, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 2035 + }, + { + "epoch": 1.6261980830670928, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 2036 + }, + { + "epoch": 1.626996805111821, + "grad_norm": 0.216796875, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 2037 + }, + { + "epoch": 1.6277955271565494, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.0731, + "step": 2038 + }, + { + "epoch": 1.628594249201278, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 2039 + }, + { + "epoch": 1.6293929712460065, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 2040 + }, + { + "epoch": 1.630191693290735, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 2041 + }, + { + "epoch": 1.6309904153354633, + "grad_norm": 0.18359375, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 2042 + }, + { + "epoch": 1.6317891373801916, + "grad_norm": 0.1845703125, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 2043 + }, + { + "epoch": 1.6325878594249201, + "grad_norm": 0.294921875, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 2044 + }, + { + "epoch": 1.6333865814696487, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 2045 + }, + { + "epoch": 1.634185303514377, + "grad_norm": 0.259765625, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 2046 + }, + { + "epoch": 1.6349840255591053, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 2047 + }, + { + "epoch": 1.6357827476038338, + "grad_norm": 0.1962890625, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 2048 + }, + { + "epoch": 1.6365814696485623, + "grad_norm": 0.1904296875, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 2049 + }, + { + "epoch": 1.6373801916932909, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 2050 + }, + { + "epoch": 1.6381789137380192, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 2051 + }, + { + "epoch": 1.6389776357827475, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 2052 + }, + { + "epoch": 1.639776357827476, + "grad_norm": 0.17578125, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 2053 + }, + { + "epoch": 1.6405750798722045, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 2054 + }, + { + "epoch": 1.641373801916933, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 2055 + }, + { + "epoch": 1.6421725239616614, + "grad_norm": 0.16015625, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 2056 + }, + { + "epoch": 1.6429712460063897, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 2057 + }, + { + "epoch": 1.6437699680511182, + "grad_norm": 0.212890625, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 2058 + }, + { + "epoch": 1.6445686900958467, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 2059 + }, + { + "epoch": 1.645367412140575, + "grad_norm": 0.30859375, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 2060 + }, + { + "epoch": 1.6461661341853036, + "grad_norm": 0.19140625, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 2061 + }, + { + "epoch": 1.6469648562300319, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0686, + "step": 2062 + }, + { + "epoch": 1.6477635782747604, + "grad_norm": 0.353515625, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 2063 + }, + { + "epoch": 1.648562300319489, + "grad_norm": 0.30859375, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 2064 + }, + { + "epoch": 1.6493610223642172, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 2065 + }, + { + "epoch": 1.6501597444089455, + "grad_norm": 0.359375, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 2066 + }, + { + "epoch": 1.650958466453674, + "grad_norm": 0.322265625, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 2067 + }, + { + "epoch": 1.6517571884984026, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 2068 + }, + { + "epoch": 1.6525559105431311, + "grad_norm": 0.1689453125, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 2069 + }, + { + "epoch": 1.6533546325878594, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 2070 + }, + { + "epoch": 1.6541533546325877, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 2071 + }, + { + "epoch": 1.6549520766773163, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 2072 + }, + { + "epoch": 1.6557507987220448, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 2073 + }, + { + "epoch": 1.6565495207667733, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 2074 + }, + { + "epoch": 1.6573482428115016, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 2075 + }, + { + "epoch": 1.65814696485623, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 2076 + }, + { + "epoch": 1.6589456869009584, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0689, + "step": 2077 + }, + { + "epoch": 1.659744408945687, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 2078 + }, + { + "epoch": 1.6605431309904153, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 2079 + }, + { + "epoch": 1.6613418530351438, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 2080 + }, + { + "epoch": 1.6621405750798721, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 2081 + }, + { + "epoch": 1.6629392971246006, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 2082 + }, + { + "epoch": 1.6637380191693292, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 2083 + }, + { + "epoch": 1.6645367412140575, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 2084 + }, + { + "epoch": 1.6653354632587858, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 2085 + }, + { + "epoch": 1.6661341853035143, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 2086 + }, + { + "epoch": 1.6669329073482428, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 2087 + }, + { + "epoch": 1.6677316293929714, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 2088 + }, + { + "epoch": 1.6685303514376997, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 2089 + }, + { + "epoch": 1.669329073482428, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 2090 + }, + { + "epoch": 1.6701277955271565, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 2091 + }, + { + "epoch": 1.670926517571885, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 2092 + }, + { + "epoch": 1.6717252396166136, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 2093 + }, + { + "epoch": 1.6725239616613419, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 2094 + }, + { + "epoch": 1.6733226837060702, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 2095 + }, + { + "epoch": 1.6741214057507987, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 2096 + }, + { + "epoch": 1.6749201277955272, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 2097 + }, + { + "epoch": 1.6757188498402555, + "grad_norm": 0.1982421875, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 2098 + }, + { + "epoch": 1.676517571884984, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 2099 + }, + { + "epoch": 1.6773162939297124, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 2100 + }, + { + "epoch": 1.678115015974441, + "grad_norm": 0.1748046875, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 2101 + }, + { + "epoch": 1.6789137380191694, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 2102 + }, + { + "epoch": 1.6797124600638977, + "grad_norm": 0.2216796875, + "learning_rate": 0.0005, + "loss": 1.0761, + "step": 2103 + }, + { + "epoch": 1.680511182108626, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 2104 + }, + { + "epoch": 1.6813099041533546, + "grad_norm": 0.1669921875, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 2105 + }, + { + "epoch": 1.682108626198083, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 2106 + }, + { + "epoch": 1.6829073482428116, + "grad_norm": 0.1923828125, + "learning_rate": 0.0005, + "loss": 1.0773, + "step": 2107 + }, + { + "epoch": 1.68370607028754, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0714, + "step": 2108 + }, + { + "epoch": 1.6845047923322682, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 2109 + }, + { + "epoch": 1.6853035143769968, + "grad_norm": 0.169921875, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 2110 + }, + { + "epoch": 1.6861022364217253, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 2111 + }, + { + "epoch": 1.6869009584664538, + "grad_norm": 0.2060546875, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 2112 + }, + { + "epoch": 1.6876996805111821, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 2113 + }, + { + "epoch": 1.6884984025559104, + "grad_norm": 0.19921875, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 2114 + }, + { + "epoch": 1.689297124600639, + "grad_norm": 0.27734375, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 2115 + }, + { + "epoch": 1.6900958466453675, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 2116 + }, + { + "epoch": 1.6908945686900958, + "grad_norm": 0.349609375, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 2117 + }, + { + "epoch": 1.6916932907348243, + "grad_norm": 0.2099609375, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 2118 + }, + { + "epoch": 1.6924920127795526, + "grad_norm": 0.208984375, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 2119 + }, + { + "epoch": 1.6932907348242812, + "grad_norm": 0.4140625, + "learning_rate": 0.0005, + "loss": 1.0745, + "step": 2120 + }, + { + "epoch": 1.6940894568690097, + "grad_norm": 0.25390625, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 2121 + }, + { + "epoch": 1.694888178913738, + "grad_norm": 0.1708984375, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 2122 + }, + { + "epoch": 1.6956869009584663, + "grad_norm": 0.345703125, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 2123 + }, + { + "epoch": 1.6964856230031948, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 2124 + }, + { + "epoch": 1.6972843450479234, + "grad_norm": 0.2119140625, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 2125 + }, + { + "epoch": 1.6980830670926519, + "grad_norm": 0.2294921875, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 2126 + }, + { + "epoch": 1.6988817891373802, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 2127 + }, + { + "epoch": 1.6996805111821085, + "grad_norm": 0.28515625, + "learning_rate": 0.0005, + "loss": 1.0767, + "step": 2128 + }, + { + "epoch": 1.700479233226837, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0724, + "step": 2129 + }, + { + "epoch": 1.7012779552715656, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 2130 + }, + { + "epoch": 1.702076677316294, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 2131 + }, + { + "epoch": 1.7028753993610224, + "grad_norm": 0.267578125, + "learning_rate": 0.0005, + "loss": 1.074, + "step": 2132 + }, + { + "epoch": 1.7036741214057507, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 2133 + }, + { + "epoch": 1.7044728434504792, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 2134 + }, + { + "epoch": 1.7052715654952078, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 2135 + }, + { + "epoch": 1.706070287539936, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 2136 + }, + { + "epoch": 1.7068690095846646, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 2137 + }, + { + "epoch": 1.707667731629393, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 2138 + }, + { + "epoch": 1.7084664536741214, + "grad_norm": 0.181640625, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 2139 + }, + { + "epoch": 1.70926517571885, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 2140 + }, + { + "epoch": 1.7100638977635783, + "grad_norm": 0.1865234375, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 2141 + }, + { + "epoch": 1.7108626198083066, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 2142 + }, + { + "epoch": 1.711661341853035, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 2143 + }, + { + "epoch": 1.7124600638977636, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 2144 + }, + { + "epoch": 1.7132587859424921, + "grad_norm": 0.181640625, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 2145 + }, + { + "epoch": 1.7140575079872205, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 2146 + }, + { + "epoch": 1.7148562300319488, + "grad_norm": 0.27734375, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 2147 + }, + { + "epoch": 1.7156549520766773, + "grad_norm": 0.2060546875, + "learning_rate": 0.0005, + "loss": 1.0702, + "step": 2148 + }, + { + "epoch": 1.7164536741214058, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 2149 + }, + { + "epoch": 1.7172523961661343, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 2150 + }, + { + "epoch": 1.7180511182108626, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 2151 + }, + { + "epoch": 1.718849840255591, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 2152 + }, + { + "epoch": 1.7196485623003195, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 2153 + }, + { + "epoch": 1.720447284345048, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 2154 + }, + { + "epoch": 1.7212460063897763, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 2155 + }, + { + "epoch": 1.7220447284345048, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 2156 + }, + { + "epoch": 1.7228434504792332, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 2157 + }, + { + "epoch": 1.7236421725239617, + "grad_norm": 0.1669921875, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 2158 + }, + { + "epoch": 1.7244408945686902, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 2159 + }, + { + "epoch": 1.7252396166134185, + "grad_norm": 0.169921875, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 2160 + }, + { + "epoch": 1.7260383386581468, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0715, + "step": 2161 + }, + { + "epoch": 1.7268370607028753, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 2162 + }, + { + "epoch": 1.7276357827476039, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 2163 + }, + { + "epoch": 1.7284345047923324, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 2164 + }, + { + "epoch": 1.7292332268370607, + "grad_norm": 0.173828125, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 2165 + }, + { + "epoch": 1.730031948881789, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 2166 + }, + { + "epoch": 1.7308306709265175, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 2167 + }, + { + "epoch": 1.731629392971246, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 2168 + }, + { + "epoch": 1.7324281150159746, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 2169 + }, + { + "epoch": 1.733226837060703, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 2170 + }, + { + "epoch": 1.7340255591054312, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 2171 + }, + { + "epoch": 1.7348242811501597, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 2172 + }, + { + "epoch": 1.7356230031948883, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 2173 + }, + { + "epoch": 1.7364217252396166, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 2174 + }, + { + "epoch": 1.7372204472843449, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 2175 + }, + { + "epoch": 1.7380191693290734, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 2176 + }, + { + "epoch": 1.738817891373802, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 2177 + }, + { + "epoch": 1.7396166134185305, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 2178 + }, + { + "epoch": 1.7404153354632588, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0734, + "step": 2179 + }, + { + "epoch": 1.741214057507987, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 2180 + }, + { + "epoch": 1.7420127795527156, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 2181 + }, + { + "epoch": 1.7428115015974441, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 2182 + }, + { + "epoch": 1.7436102236421727, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0701, + "step": 2183 + }, + { + "epoch": 1.744408945686901, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 2184 + }, + { + "epoch": 1.7452076677316293, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 2185 + }, + { + "epoch": 1.7460063897763578, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 2186 + }, + { + "epoch": 1.7468051118210863, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 2187 + }, + { + "epoch": 1.7476038338658149, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 2188 + }, + { + "epoch": 1.7484025559105432, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 2189 + }, + { + "epoch": 1.7492012779552715, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 2190 + }, + { + "epoch": 1.75, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0693, + "step": 2191 + }, + { + "epoch": 1.7507987220447285, + "grad_norm": 0.162109375, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 2192 + }, + { + "epoch": 1.7515974440894568, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 2193 + }, + { + "epoch": 1.7523961661341851, + "grad_norm": 0.2333984375, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 2194 + }, + { + "epoch": 1.7531948881789137, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 2195 + }, + { + "epoch": 1.7539936102236422, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 2196 + }, + { + "epoch": 1.7547923322683707, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 2197 + }, + { + "epoch": 1.755591054313099, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 2198 + }, + { + "epoch": 1.7563897763578273, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 2199 + }, + { + "epoch": 1.7571884984025559, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 2200 + }, + { + "epoch": 1.7579872204472844, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 2201 + }, + { + "epoch": 1.758785942492013, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 2202 + }, + { + "epoch": 1.7595846645367412, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 2203 + }, + { + "epoch": 1.7603833865814695, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 2204 + }, + { + "epoch": 1.761182108626198, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 2205 + }, + { + "epoch": 1.7619808306709266, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 2206 + }, + { + "epoch": 1.7627795527156551, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 2207 + }, + { + "epoch": 1.7635782747603834, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 2208 + }, + { + "epoch": 1.7643769968051117, + "grad_norm": 0.173828125, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 2209 + }, + { + "epoch": 1.7651757188498403, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 2210 + }, + { + "epoch": 1.7659744408945688, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 2211 + }, + { + "epoch": 1.766773162939297, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 2212 + }, + { + "epoch": 1.7675718849840254, + "grad_norm": 0.2294921875, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 2213 + }, + { + "epoch": 1.768370607028754, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 2214 + }, + { + "epoch": 1.7691693290734825, + "grad_norm": 0.205078125, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 2215 + }, + { + "epoch": 1.769968051118211, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 2216 + }, + { + "epoch": 1.7707667731629393, + "grad_norm": 0.1982421875, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 2217 + }, + { + "epoch": 1.7715654952076676, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 2218 + }, + { + "epoch": 1.7723642172523961, + "grad_norm": 0.162109375, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 2219 + }, + { + "epoch": 1.7731629392971247, + "grad_norm": 0.1669921875, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 2220 + }, + { + "epoch": 1.7739616613418532, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 2221 + }, + { + "epoch": 1.7747603833865815, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 2222 + }, + { + "epoch": 1.7755591054313098, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 2223 + }, + { + "epoch": 1.7763578274760383, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 2224 + }, + { + "epoch": 1.7771565495207668, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 2225 + }, + { + "epoch": 1.7779552715654952, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 2226 + }, + { + "epoch": 1.7787539936102237, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 2227 + }, + { + "epoch": 1.779552715654952, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0781, + "step": 2228 + }, + { + "epoch": 1.7803514376996805, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 2229 + }, + { + "epoch": 1.781150159744409, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 2230 + }, + { + "epoch": 1.7819488817891374, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 2231 + }, + { + "epoch": 1.7827476038338657, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 2232 + }, + { + "epoch": 1.7835463258785942, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0689, + "step": 2233 + }, + { + "epoch": 1.7843450479233227, + "grad_norm": 0.1630859375, + "learning_rate": 0.0005, + "loss": 1.071, + "step": 2234 + }, + { + "epoch": 1.7851437699680512, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 2235 + }, + { + "epoch": 1.7859424920127795, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 2236 + }, + { + "epoch": 1.7867412140575079, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 2237 + }, + { + "epoch": 1.7875399361022364, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 2238 + }, + { + "epoch": 1.788338658146965, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 2239 + }, + { + "epoch": 1.7891373801916934, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 2240 + }, + { + "epoch": 1.7899361022364217, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0806, + "step": 2241 + }, + { + "epoch": 1.79073482428115, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0693, + "step": 2242 + }, + { + "epoch": 1.7915335463258786, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 2243 + }, + { + "epoch": 1.792332268370607, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 2244 + }, + { + "epoch": 1.7931309904153354, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 2245 + }, + { + "epoch": 1.793929712460064, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 2246 + }, + { + "epoch": 1.7947284345047922, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 2247 + }, + { + "epoch": 1.7955271565495208, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 2248 + }, + { + "epoch": 1.7963258785942493, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 2249 + }, + { + "epoch": 1.7971246006389776, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 2250 + }, + { + "epoch": 1.797923322683706, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 2251 + }, + { + "epoch": 1.7987220447284344, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 2252 + }, + { + "epoch": 1.799520766773163, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 2253 + }, + { + "epoch": 1.8003194888178915, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 2254 + }, + { + "epoch": 1.8011182108626198, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 2255 + }, + { + "epoch": 1.8019169329073481, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 2256 + }, + { + "epoch": 1.8027156549520766, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 2257 + }, + { + "epoch": 1.8035143769968052, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 2258 + }, + { + "epoch": 1.8043130990415337, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0752, + "step": 2259 + }, + { + "epoch": 1.805111821086262, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 2260 + }, + { + "epoch": 1.8059105431309903, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 2261 + }, + { + "epoch": 1.8067092651757188, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 2262 + }, + { + "epoch": 1.8075079872204474, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 2263 + }, + { + "epoch": 1.8083067092651757, + "grad_norm": 0.1630859375, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 2264 + }, + { + "epoch": 1.8091054313099042, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 2265 + }, + { + "epoch": 1.8099041533546325, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 2266 + }, + { + "epoch": 1.810702875399361, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 2267 + }, + { + "epoch": 1.8115015974440896, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 2268 + }, + { + "epoch": 1.8123003194888179, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0692, + "step": 2269 + }, + { + "epoch": 1.8130990415335462, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 2270 + }, + { + "epoch": 1.8138977635782747, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 2271 + }, + { + "epoch": 1.8146964856230032, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 2272 + }, + { + "epoch": 1.8154952076677318, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 2273 + }, + { + "epoch": 1.81629392971246, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 2274 + }, + { + "epoch": 1.8170926517571884, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 2275 + }, + { + "epoch": 1.817891373801917, + "grad_norm": 0.1904296875, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 2276 + }, + { + "epoch": 1.8186900958466454, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0673, + "step": 2277 + }, + { + "epoch": 1.819488817891374, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 2278 + }, + { + "epoch": 1.8202875399361023, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 2279 + }, + { + "epoch": 1.8210862619808306, + "grad_norm": 0.1533203125, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 2280 + }, + { + "epoch": 1.821884984025559, + "grad_norm": 0.232421875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 2281 + }, + { + "epoch": 1.8226837060702876, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 2282 + }, + { + "epoch": 1.823482428115016, + "grad_norm": 0.359375, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 2283 + }, + { + "epoch": 1.8242811501597445, + "grad_norm": 0.38671875, + "learning_rate": 0.0005, + "loss": 1.07, + "step": 2284 + }, + { + "epoch": 1.8250798722044728, + "grad_norm": 0.2265625, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 2285 + }, + { + "epoch": 1.8258785942492013, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 2286 + }, + { + "epoch": 1.8266773162939298, + "grad_norm": 0.18359375, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 2287 + }, + { + "epoch": 1.8274760383386581, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 2288 + }, + { + "epoch": 1.8282747603833864, + "grad_norm": 0.1708984375, + "learning_rate": 0.0005, + "loss": 1.0705, + "step": 2289 + }, + { + "epoch": 1.829073482428115, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 2290 + }, + { + "epoch": 1.8298722044728435, + "grad_norm": 0.1806640625, + "learning_rate": 0.0005, + "loss": 1.0694, + "step": 2291 + }, + { + "epoch": 1.830670926517572, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 2292 + }, + { + "epoch": 1.8314696485623003, + "grad_norm": 0.162109375, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 2293 + }, + { + "epoch": 1.8322683706070286, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 2294 + }, + { + "epoch": 1.8330670926517572, + "grad_norm": 0.1533203125, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 2295 + }, + { + "epoch": 1.8338658146964857, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 2296 + }, + { + "epoch": 1.8346645367412142, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 2297 + }, + { + "epoch": 1.8354632587859425, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 2298 + }, + { + "epoch": 1.8362619808306708, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 2299 + }, + { + "epoch": 1.8370607028753994, + "grad_norm": 0.2353515625, + "learning_rate": 0.0005, + "loss": 1.0689, + "step": 2300 + }, + { + "epoch": 1.8378594249201279, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 2301 + }, + { + "epoch": 1.8386581469648562, + "grad_norm": 0.177734375, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 2302 + }, + { + "epoch": 1.8394568690095847, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 2303 + }, + { + "epoch": 1.840255591054313, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 2304 + }, + { + "epoch": 1.8410543130990416, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 2305 + }, + { + "epoch": 1.84185303514377, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 2306 + }, + { + "epoch": 1.8426517571884984, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 2307 + }, + { + "epoch": 1.8434504792332267, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 2308 + }, + { + "epoch": 1.8442492012779552, + "grad_norm": 0.19140625, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 2309 + }, + { + "epoch": 1.8450479233226837, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 2310 + }, + { + "epoch": 1.8458466453674123, + "grad_norm": 0.23046875, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 2311 + }, + { + "epoch": 1.8466453674121406, + "grad_norm": 0.162109375, + "learning_rate": 0.0005, + "loss": 1.0696, + "step": 2312 + }, + { + "epoch": 1.8474440894568689, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 2313 + }, + { + "epoch": 1.8482428115015974, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 2314 + }, + { + "epoch": 1.849041533546326, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 2315 + }, + { + "epoch": 1.8498402555910545, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 2316 + }, + { + "epoch": 1.8506389776357828, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 2317 + }, + { + "epoch": 1.851437699680511, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0682, + "step": 2318 + }, + { + "epoch": 1.8522364217252396, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.044, + "step": 2319 + }, + { + "epoch": 1.8530351437699681, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 2320 + }, + { + "epoch": 1.8538338658146964, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 2321 + }, + { + "epoch": 1.854632587859425, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0776, + "step": 2322 + }, + { + "epoch": 1.8554313099041533, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 2323 + }, + { + "epoch": 1.8562300319488818, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 2324 + }, + { + "epoch": 1.8570287539936103, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 2325 + }, + { + "epoch": 1.8578274760383386, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 2326 + }, + { + "epoch": 1.858626198083067, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 2327 + }, + { + "epoch": 1.8594249201277955, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 2328 + }, + { + "epoch": 1.860223642172524, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 2329 + }, + { + "epoch": 1.8610223642172525, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 2330 + }, + { + "epoch": 1.8618210862619808, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 2331 + }, + { + "epoch": 1.8626198083067091, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 2332 + }, + { + "epoch": 1.8634185303514377, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 2333 + }, + { + "epoch": 1.8642172523961662, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 2334 + }, + { + "epoch": 1.8650159744408947, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 2335 + }, + { + "epoch": 1.865814696485623, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 2336 + }, + { + "epoch": 1.8666134185303513, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 2337 + }, + { + "epoch": 1.8674121405750799, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 2338 + }, + { + "epoch": 1.8682108626198084, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 2339 + }, + { + "epoch": 1.8690095846645367, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.068, + "step": 2340 + }, + { + "epoch": 1.869808306709265, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 2341 + }, + { + "epoch": 1.8706070287539935, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 2342 + }, + { + "epoch": 1.871405750798722, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 2343 + }, + { + "epoch": 1.8722044728434506, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 2344 + }, + { + "epoch": 1.873003194888179, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 2345 + }, + { + "epoch": 1.8738019169329072, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0693, + "step": 2346 + }, + { + "epoch": 1.8746006389776357, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 2347 + }, + { + "epoch": 1.8753993610223643, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 2348 + }, + { + "epoch": 1.8761980830670928, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 2349 + }, + { + "epoch": 1.876996805111821, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 2350 + }, + { + "epoch": 1.8777955271565494, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 2351 + }, + { + "epoch": 1.878594249201278, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 2352 + }, + { + "epoch": 1.8793929712460065, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 2353 + }, + { + "epoch": 1.880191693290735, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 2354 + }, + { + "epoch": 1.8809904153354633, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.042, + "step": 2355 + }, + { + "epoch": 1.8817891373801916, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 2356 + }, + { + "epoch": 1.8825878594249201, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0736, + "step": 2357 + }, + { + "epoch": 1.8833865814696487, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 2358 + }, + { + "epoch": 1.884185303514377, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0758, + "step": 2359 + }, + { + "epoch": 1.8849840255591053, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 2360 + }, + { + "epoch": 1.8857827476038338, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 2361 + }, + { + "epoch": 1.8865814696485623, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 2362 + }, + { + "epoch": 1.8873801916932909, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 2363 + }, + { + "epoch": 1.8881789137380192, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0686, + "step": 2364 + }, + { + "epoch": 1.8889776357827475, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 2365 + }, + { + "epoch": 1.889776357827476, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 2366 + }, + { + "epoch": 1.8905750798722045, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 2367 + }, + { + "epoch": 1.891373801916933, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 2368 + }, + { + "epoch": 1.8921725239616614, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 2369 + }, + { + "epoch": 1.8929712460063897, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 2370 + }, + { + "epoch": 1.8937699680511182, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 2371 + }, + { + "epoch": 1.8945686900958467, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 2372 + }, + { + "epoch": 1.895367412140575, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 2373 + }, + { + "epoch": 1.8961661341853036, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 2374 + }, + { + "epoch": 1.8969648562300319, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0686, + "step": 2375 + }, + { + "epoch": 1.8977635782747604, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 2376 + }, + { + "epoch": 1.898562300319489, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 2377 + }, + { + "epoch": 1.8993610223642172, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 2378 + }, + { + "epoch": 1.9001597444089455, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 2379 + }, + { + "epoch": 1.900958466453674, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 2380 + }, + { + "epoch": 1.9017571884984026, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 2381 + }, + { + "epoch": 1.9025559105431311, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 2382 + }, + { + "epoch": 1.9033546325878594, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 2383 + }, + { + "epoch": 1.9041533546325877, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 2384 + }, + { + "epoch": 1.9049520766773163, + "grad_norm": 0.203125, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 2385 + }, + { + "epoch": 1.9057507987220448, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 2386 + }, + { + "epoch": 1.9065495207667733, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 2387 + }, + { + "epoch": 1.9073482428115016, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 2388 + }, + { + "epoch": 1.90814696485623, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 2389 + }, + { + "epoch": 1.9089456869009584, + "grad_norm": 0.220703125, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 2390 + }, + { + "epoch": 1.909744408945687, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 2391 + }, + { + "epoch": 1.9105431309904153, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 2392 + }, + { + "epoch": 1.9113418530351438, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0659, + "step": 2393 + }, + { + "epoch": 1.9121405750798721, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0712, + "step": 2394 + }, + { + "epoch": 1.9129392971246006, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 2395 + }, + { + "epoch": 1.9137380191693292, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0441, + "step": 2396 + }, + { + "epoch": 1.9145367412140575, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 2397 + }, + { + "epoch": 1.9153354632587858, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 2398 + }, + { + "epoch": 1.9161341853035143, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 2399 + }, + { + "epoch": 1.9169329073482428, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 2400 + }, + { + "epoch": 1.9177316293929714, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 2401 + }, + { + "epoch": 1.9185303514376997, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 2402 + }, + { + "epoch": 1.919329073482428, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 2403 + }, + { + "epoch": 1.9201277955271565, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 2404 + }, + { + "epoch": 1.920926517571885, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 2405 + }, + { + "epoch": 1.9217252396166136, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 2406 + }, + { + "epoch": 1.9225239616613419, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0707, + "step": 2407 + }, + { + "epoch": 1.9233226837060702, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0725, + "step": 2408 + }, + { + "epoch": 1.9241214057507987, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 2409 + }, + { + "epoch": 1.9249201277955272, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 2410 + }, + { + "epoch": 1.9257188498402555, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 2411 + }, + { + "epoch": 1.926517571884984, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0729, + "step": 2412 + }, + { + "epoch": 1.9273162939297124, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 2413 + }, + { + "epoch": 1.928115015974441, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 2414 + }, + { + "epoch": 1.9289137380191694, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 2415 + }, + { + "epoch": 1.9297124600638977, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 2416 + }, + { + "epoch": 1.930511182108626, + "grad_norm": 0.18359375, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 2417 + }, + { + "epoch": 1.9313099041533546, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 2418 + }, + { + "epoch": 1.932108626198083, + "grad_norm": 0.1611328125, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 2419 + }, + { + "epoch": 1.9329073482428116, + "grad_norm": 0.220703125, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 2420 + }, + { + "epoch": 1.93370607028754, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 2421 + }, + { + "epoch": 1.9345047923322682, + "grad_norm": 0.251953125, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 2422 + }, + { + "epoch": 1.9353035143769968, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 2423 + }, + { + "epoch": 1.9361022364217253, + "grad_norm": 0.1787109375, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 2424 + }, + { + "epoch": 1.9369009584664538, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 2425 + }, + { + "epoch": 1.9376996805111821, + "grad_norm": 0.193359375, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 2426 + }, + { + "epoch": 1.9384984025559104, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 2427 + }, + { + "epoch": 1.939297124600639, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 2428 + }, + { + "epoch": 1.9400958466453675, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 2429 + }, + { + "epoch": 1.9408945686900958, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 2430 + }, + { + "epoch": 1.9416932907348243, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 2431 + }, + { + "epoch": 1.9424920127795526, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 2432 + }, + { + "epoch": 1.9432907348242812, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 2433 + }, + { + "epoch": 1.9440894568690097, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 2434 + }, + { + "epoch": 1.944888178913738, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 2435 + }, + { + "epoch": 1.9456869009584663, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 2436 + }, + { + "epoch": 1.9464856230031948, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 2437 + }, + { + "epoch": 1.9472843450479234, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 2438 + }, + { + "epoch": 1.9480830670926519, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 2439 + }, + { + "epoch": 1.9488817891373802, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 2440 + }, + { + "epoch": 1.9496805111821085, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 2441 + }, + { + "epoch": 1.950479233226837, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 2442 + }, + { + "epoch": 1.9512779552715656, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 2443 + }, + { + "epoch": 1.952076677316294, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 2444 + }, + { + "epoch": 1.9528753993610224, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 2445 + }, + { + "epoch": 1.9536741214057507, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 2446 + }, + { + "epoch": 1.9544728434504792, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 2447 + }, + { + "epoch": 1.9552715654952078, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 2448 + }, + { + "epoch": 1.956070287539936, + "grad_norm": 0.1826171875, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 2449 + }, + { + "epoch": 1.9568690095846646, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 2450 + }, + { + "epoch": 1.957667731629393, + "grad_norm": 0.3671875, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 2451 + }, + { + "epoch": 1.9584664536741214, + "grad_norm": 0.267578125, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 2452 + }, + { + "epoch": 1.95926517571885, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 2453 + }, + { + "epoch": 1.9600638977635783, + "grad_norm": 0.220703125, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 2454 + }, + { + "epoch": 1.9608626198083066, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 2455 + }, + { + "epoch": 1.961661341853035, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 2456 + }, + { + "epoch": 1.9624600638977636, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 2457 + }, + { + "epoch": 1.9632587859424921, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 2458 + }, + { + "epoch": 1.9640575079872205, + "grad_norm": 0.1572265625, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 2459 + }, + { + "epoch": 1.9648562300319488, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 2460 + }, + { + "epoch": 1.9656549520766773, + "grad_norm": 0.2255859375, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 2461 + }, + { + "epoch": 1.9664536741214058, + "grad_norm": 0.23828125, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 2462 + }, + { + "epoch": 1.9672523961661343, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 2463 + }, + { + "epoch": 1.9680511182108626, + "grad_norm": 0.265625, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 2464 + }, + { + "epoch": 1.968849840255591, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 2465 + }, + { + "epoch": 1.9696485623003195, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 2466 + }, + { + "epoch": 1.970447284345048, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 2467 + }, + { + "epoch": 1.9712460063897763, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 2468 + }, + { + "epoch": 1.9720447284345048, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 2469 + }, + { + "epoch": 1.9728434504792332, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 2470 + }, + { + "epoch": 1.9736421725239617, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.069, + "step": 2471 + }, + { + "epoch": 1.9744408945686902, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 2472 + }, + { + "epoch": 1.9752396166134185, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 2473 + }, + { + "epoch": 1.9760383386581468, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 2474 + }, + { + "epoch": 1.9768370607028753, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 2475 + }, + { + "epoch": 1.9776357827476039, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 2476 + }, + { + "epoch": 1.9784345047923324, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 2477 + }, + { + "epoch": 1.9792332268370607, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 2478 + }, + { + "epoch": 1.980031948881789, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 2479 + }, + { + "epoch": 1.9808306709265175, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 2480 + }, + { + "epoch": 1.981629392971246, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 2481 + }, + { + "epoch": 1.9824281150159746, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 2482 + }, + { + "epoch": 1.983226837060703, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 2483 + }, + { + "epoch": 1.9840255591054312, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 2484 + }, + { + "epoch": 1.9848242811501597, + "grad_norm": 0.171875, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 2485 + }, + { + "epoch": 1.9856230031948883, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0699, + "step": 2486 + }, + { + "epoch": 1.9864217252396166, + "grad_norm": 0.1826171875, + "learning_rate": 0.0005, + "loss": 1.0708, + "step": 2487 + }, + { + "epoch": 1.9872204472843449, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 2488 + }, + { + "epoch": 1.9880191693290734, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 2489 + }, + { + "epoch": 1.988817891373802, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.0686, + "step": 2490 + }, + { + "epoch": 1.9896166134185305, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 2491 + }, + { + "epoch": 1.9904153354632588, + "grad_norm": 0.2001953125, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 2492 + }, + { + "epoch": 1.991214057507987, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 2493 + }, + { + "epoch": 1.9920127795527156, + "grad_norm": 0.1533203125, + "learning_rate": 0.0005, + "loss": 1.0633, + "step": 2494 + }, + { + "epoch": 1.9928115015974441, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 2495 + }, + { + "epoch": 1.9936102236421727, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 2496 + }, + { + "epoch": 1.994408945686901, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 2497 + }, + { + "epoch": 1.9952076677316293, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 2498 + }, + { + "epoch": 1.9960063897763578, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 2499 + }, + { + "epoch": 1.9968051118210863, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 2500 + }, + { + "epoch": 1.9976038338658149, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0688, + "step": 2501 + }, + { + "epoch": 1.9984025559105432, + "grad_norm": 0.169921875, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 2502 + }, + { + "epoch": 1.9992012779552715, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 2503 + }, + { + "epoch": 2.0, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 2504 + }, + { + "epoch": 2.0007987220447285, + "grad_norm": 0.193359375, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 2505 + }, + { + "epoch": 2.001597444089457, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 2506 + }, + { + "epoch": 2.002396166134185, + "grad_norm": 0.205078125, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 2507 + }, + { + "epoch": 2.0031948881789137, + "grad_norm": 0.1630859375, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 2508 + }, + { + "epoch": 2.003993610223642, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 2509 + }, + { + "epoch": 2.0047923322683707, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 2510 + }, + { + "epoch": 2.0055910543130993, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 2511 + }, + { + "epoch": 2.0063897763578273, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 2512 + }, + { + "epoch": 2.007188498402556, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 2513 + }, + { + "epoch": 2.0079872204472844, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 2514 + }, + { + "epoch": 2.008785942492013, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 2515 + }, + { + "epoch": 2.009584664536741, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 2516 + }, + { + "epoch": 2.0103833865814695, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 2517 + }, + { + "epoch": 2.011182108626198, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 2518 + }, + { + "epoch": 2.0119808306709266, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 2519 + }, + { + "epoch": 2.012779552715655, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 2520 + }, + { + "epoch": 2.013578274760383, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 2521 + }, + { + "epoch": 2.0143769968051117, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0726, + "step": 2522 + }, + { + "epoch": 2.0151757188498403, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 2523 + }, + { + "epoch": 2.015974440894569, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 2524 + }, + { + "epoch": 2.0167731629392973, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 2525 + }, + { + "epoch": 2.0175718849840254, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 2526 + }, + { + "epoch": 2.018370607028754, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 2527 + }, + { + "epoch": 2.0191693290734825, + "grad_norm": 0.2294921875, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 2528 + }, + { + "epoch": 2.019968051118211, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 2529 + }, + { + "epoch": 2.0207667731629395, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 2530 + }, + { + "epoch": 2.0215654952076676, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 2531 + }, + { + "epoch": 2.022364217252396, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 2532 + }, + { + "epoch": 2.0231629392971247, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 2533 + }, + { + "epoch": 2.023961661341853, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0722, + "step": 2534 + }, + { + "epoch": 2.0247603833865813, + "grad_norm": 0.1533203125, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 2535 + }, + { + "epoch": 2.02555910543131, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 2536 + }, + { + "epoch": 2.0263578274760383, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 2537 + }, + { + "epoch": 2.027156549520767, + "grad_norm": 0.2734375, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 2538 + }, + { + "epoch": 2.0279552715654954, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 2539 + }, + { + "epoch": 2.0287539936102235, + "grad_norm": 0.2119140625, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 2540 + }, + { + "epoch": 2.029552715654952, + "grad_norm": 0.220703125, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 2541 + }, + { + "epoch": 2.0303514376996805, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 2542 + }, + { + "epoch": 2.031150159744409, + "grad_norm": 0.345703125, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 2543 + }, + { + "epoch": 2.0319488817891376, + "grad_norm": 0.349609375, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 2544 + }, + { + "epoch": 2.0327476038338657, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 2545 + }, + { + "epoch": 2.033546325878594, + "grad_norm": 0.326171875, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 2546 + }, + { + "epoch": 2.0343450479233227, + "grad_norm": 0.2890625, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 2547 + }, + { + "epoch": 2.0351437699680512, + "grad_norm": 0.1533203125, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 2548 + }, + { + "epoch": 2.0359424920127798, + "grad_norm": 0.3984375, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 2549 + }, + { + "epoch": 2.036741214057508, + "grad_norm": 0.205078125, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 2550 + }, + { + "epoch": 2.0375399361022364, + "grad_norm": 0.271484375, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 2551 + }, + { + "epoch": 2.038338658146965, + "grad_norm": 0.5, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 2552 + }, + { + "epoch": 2.0391373801916934, + "grad_norm": 0.455078125, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 2553 + }, + { + "epoch": 2.0399361022364215, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 2554 + }, + { + "epoch": 2.04073482428115, + "grad_norm": 0.31640625, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 2555 + }, + { + "epoch": 2.0415335463258786, + "grad_norm": 0.330078125, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 2556 + }, + { + "epoch": 2.042332268370607, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 2557 + }, + { + "epoch": 2.0431309904153356, + "grad_norm": 0.208984375, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 2558 + }, + { + "epoch": 2.0439297124600637, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0695, + "step": 2559 + }, + { + "epoch": 2.0447284345047922, + "grad_norm": 0.251953125, + "learning_rate": 0.0005, + "loss": 1.0674, + "step": 2560 + }, + { + "epoch": 2.0455271565495208, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 2561 + }, + { + "epoch": 2.0463258785942493, + "grad_norm": 0.275390625, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 2562 + }, + { + "epoch": 2.047124600638978, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 2563 + }, + { + "epoch": 2.047923322683706, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 2564 + }, + { + "epoch": 2.0487220447284344, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 2565 + }, + { + "epoch": 2.049520766773163, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 2566 + }, + { + "epoch": 2.0503194888178915, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 2567 + }, + { + "epoch": 2.0511182108626196, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 2568 + }, + { + "epoch": 2.051916932907348, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 2569 + }, + { + "epoch": 2.0527156549520766, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0693, + "step": 2570 + }, + { + "epoch": 2.053514376996805, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 2571 + }, + { + "epoch": 2.0543130990415337, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 2572 + }, + { + "epoch": 2.055111821086262, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 2573 + }, + { + "epoch": 2.0559105431309903, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 2574 + }, + { + "epoch": 2.056709265175719, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 2575 + }, + { + "epoch": 2.0575079872204474, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 2576 + }, + { + "epoch": 2.058306709265176, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 2577 + }, + { + "epoch": 2.059105431309904, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 2578 + }, + { + "epoch": 2.0599041533546325, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 2579 + }, + { + "epoch": 2.060702875399361, + "grad_norm": 0.2041015625, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 2580 + }, + { + "epoch": 2.0615015974440896, + "grad_norm": 0.19921875, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 2581 + }, + { + "epoch": 2.062300319488818, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 2582 + }, + { + "epoch": 2.063099041533546, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 2583 + }, + { + "epoch": 2.0638977635782747, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 2584 + }, + { + "epoch": 2.0646964856230032, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 2585 + }, + { + "epoch": 2.0654952076677318, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 2586 + }, + { + "epoch": 2.06629392971246, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 2587 + }, + { + "epoch": 2.0670926517571884, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 2588 + }, + { + "epoch": 2.067891373801917, + "grad_norm": 0.1669921875, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 2589 + }, + { + "epoch": 2.0686900958466454, + "grad_norm": 0.291015625, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 2590 + }, + { + "epoch": 2.069488817891374, + "grad_norm": 0.1689453125, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 2591 + }, + { + "epoch": 2.070287539936102, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 2592 + }, + { + "epoch": 2.0710862619808306, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 2593 + }, + { + "epoch": 2.071884984025559, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 2594 + }, + { + "epoch": 2.0726837060702876, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 2595 + }, + { + "epoch": 2.073482428115016, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0677, + "step": 2596 + }, + { + "epoch": 2.0742811501597442, + "grad_norm": 0.171875, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 2597 + }, + { + "epoch": 2.0750798722044728, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 2598 + }, + { + "epoch": 2.0758785942492013, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 2599 + }, + { + "epoch": 2.07667731629393, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 2600 + }, + { + "epoch": 2.0774760383386583, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 2601 + }, + { + "epoch": 2.0782747603833864, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 2602 + }, + { + "epoch": 2.079073482428115, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 2603 + }, + { + "epoch": 2.0798722044728435, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0735, + "step": 2604 + }, + { + "epoch": 2.080670926517572, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 2605 + }, + { + "epoch": 2.0814696485623, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 2606 + }, + { + "epoch": 2.0822683706070286, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 2607 + }, + { + "epoch": 2.083067092651757, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 2608 + }, + { + "epoch": 2.0838658146964857, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 2609 + }, + { + "epoch": 2.084664536741214, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 2610 + }, + { + "epoch": 2.0854632587859423, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 2611 + }, + { + "epoch": 2.086261980830671, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 2612 + }, + { + "epoch": 2.0870607028753994, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 2613 + }, + { + "epoch": 2.087859424920128, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 2614 + }, + { + "epoch": 2.0886581469648564, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0666, + "step": 2615 + }, + { + "epoch": 2.0894568690095845, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 2616 + }, + { + "epoch": 2.090255591054313, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 2617 + }, + { + "epoch": 2.0910543130990416, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 2618 + }, + { + "epoch": 2.09185303514377, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 2619 + }, + { + "epoch": 2.0926517571884986, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 2620 + }, + { + "epoch": 2.0934504792332267, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 2621 + }, + { + "epoch": 2.094249201277955, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 2622 + }, + { + "epoch": 2.0950479233226837, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 2623 + }, + { + "epoch": 2.0958466453674123, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 2624 + }, + { + "epoch": 2.0966453674121404, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 2625 + }, + { + "epoch": 2.097444089456869, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 2626 + }, + { + "epoch": 2.0982428115015974, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 2627 + }, + { + "epoch": 2.099041533546326, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 2628 + }, + { + "epoch": 2.0998402555910545, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 2629 + }, + { + "epoch": 2.1006389776357826, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 2630 + }, + { + "epoch": 2.101437699680511, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 2631 + }, + { + "epoch": 2.1022364217252396, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0636, + "step": 2632 + }, + { + "epoch": 2.103035143769968, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 2633 + }, + { + "epoch": 2.1038338658146967, + "grad_norm": 0.24609375, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 2634 + }, + { + "epoch": 2.1046325878594248, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 2635 + }, + { + "epoch": 2.1054313099041533, + "grad_norm": 0.166015625, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 2636 + }, + { + "epoch": 2.106230031948882, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 2637 + }, + { + "epoch": 2.1070287539936103, + "grad_norm": 0.271484375, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 2638 + }, + { + "epoch": 2.107827476038339, + "grad_norm": 0.1611328125, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 2639 + }, + { + "epoch": 2.108626198083067, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 2640 + }, + { + "epoch": 2.1094249201277955, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 2641 + }, + { + "epoch": 2.110223642172524, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 2642 + }, + { + "epoch": 2.1110223642172525, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 2643 + }, + { + "epoch": 2.1118210862619806, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 2644 + }, + { + "epoch": 2.112619808306709, + "grad_norm": 0.181640625, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 2645 + }, + { + "epoch": 2.1134185303514377, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 2646 + }, + { + "epoch": 2.114217252396166, + "grad_norm": 0.2060546875, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 2647 + }, + { + "epoch": 2.1150159744408947, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 2648 + }, + { + "epoch": 2.115814696485623, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 2649 + }, + { + "epoch": 2.1166134185303513, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 2650 + }, + { + "epoch": 2.11741214057508, + "grad_norm": 0.169921875, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 2651 + }, + { + "epoch": 2.1182108626198084, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 2652 + }, + { + "epoch": 2.119009584664537, + "grad_norm": 0.20703125, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 2653 + }, + { + "epoch": 2.119808306709265, + "grad_norm": 0.236328125, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 2654 + }, + { + "epoch": 2.1206070287539935, + "grad_norm": 0.216796875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 2655 + }, + { + "epoch": 2.121405750798722, + "grad_norm": 0.1845703125, + "learning_rate": 0.0005, + "loss": 1.0735, + "step": 2656 + }, + { + "epoch": 2.1222044728434506, + "grad_norm": 0.345703125, + "learning_rate": 0.0005, + "loss": 1.0669, + "step": 2657 + }, + { + "epoch": 2.123003194888179, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 2658 + }, + { + "epoch": 2.123801916932907, + "grad_norm": 0.193359375, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 2659 + }, + { + "epoch": 2.1246006389776357, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 2660 + }, + { + "epoch": 2.1253993610223643, + "grad_norm": 0.1884765625, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 2661 + }, + { + "epoch": 2.126198083067093, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 2662 + }, + { + "epoch": 2.126996805111821, + "grad_norm": 0.23046875, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 2663 + }, + { + "epoch": 2.1277955271565494, + "grad_norm": 0.1708984375, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 2664 + }, + { + "epoch": 2.128594249201278, + "grad_norm": 0.30078125, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 2665 + }, + { + "epoch": 2.1293929712460065, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 2666 + }, + { + "epoch": 2.130191693290735, + "grad_norm": 0.322265625, + "learning_rate": 0.0005, + "loss": 1.0684, + "step": 2667 + }, + { + "epoch": 2.130990415335463, + "grad_norm": 0.1982421875, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 2668 + }, + { + "epoch": 2.1317891373801916, + "grad_norm": 0.232421875, + "learning_rate": 0.0005, + "loss": 1.0671, + "step": 2669 + }, + { + "epoch": 2.13258785942492, + "grad_norm": 0.1572265625, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 2670 + }, + { + "epoch": 2.1333865814696487, + "grad_norm": 0.2099609375, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 2671 + }, + { + "epoch": 2.134185303514377, + "grad_norm": 0.25, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 2672 + }, + { + "epoch": 2.1349840255591053, + "grad_norm": 0.1630859375, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 2673 + }, + { + "epoch": 2.135782747603834, + "grad_norm": 0.263671875, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 2674 + }, + { + "epoch": 2.1365814696485623, + "grad_norm": 0.193359375, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 2675 + }, + { + "epoch": 2.137380191693291, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 2676 + }, + { + "epoch": 2.1381789137380194, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 2677 + }, + { + "epoch": 2.1389776357827475, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 2678 + }, + { + "epoch": 2.139776357827476, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 2679 + }, + { + "epoch": 2.1405750798722045, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 2680 + }, + { + "epoch": 2.141373801916933, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 2681 + }, + { + "epoch": 2.142172523961661, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 2682 + }, + { + "epoch": 2.1429712460063897, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 2683 + }, + { + "epoch": 2.143769968051118, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.0713, + "step": 2684 + }, + { + "epoch": 2.1445686900958467, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 2685 + }, + { + "epoch": 2.1453674121405752, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 2686 + }, + { + "epoch": 2.1461661341853033, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 2687 + }, + { + "epoch": 2.146964856230032, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 2688 + }, + { + "epoch": 2.1477635782747604, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 2689 + }, + { + "epoch": 2.148562300319489, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 2690 + }, + { + "epoch": 2.1493610223642174, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 2691 + }, + { + "epoch": 2.1501597444089455, + "grad_norm": 0.193359375, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 2692 + }, + { + "epoch": 2.150958466453674, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 2693 + }, + { + "epoch": 2.1517571884984026, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 2694 + }, + { + "epoch": 2.152555910543131, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 2695 + }, + { + "epoch": 2.1533546325878596, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 2696 + }, + { + "epoch": 2.1541533546325877, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 2697 + }, + { + "epoch": 2.1549520766773163, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 2698 + }, + { + "epoch": 2.155750798722045, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 2699 + }, + { + "epoch": 2.1565495207667733, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 2700 + }, + { + "epoch": 2.1573482428115014, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 2701 + }, + { + "epoch": 2.15814696485623, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 2702 + }, + { + "epoch": 2.1589456869009584, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 2703 + }, + { + "epoch": 2.159744408945687, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 2704 + }, + { + "epoch": 2.1605431309904155, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 2705 + }, + { + "epoch": 2.1613418530351436, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 2706 + }, + { + "epoch": 2.162140575079872, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 2707 + }, + { + "epoch": 2.1629392971246006, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 2708 + }, + { + "epoch": 2.163738019169329, + "grad_norm": 0.1845703125, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 2709 + }, + { + "epoch": 2.1645367412140577, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 2710 + }, + { + "epoch": 2.165335463258786, + "grad_norm": 0.306640625, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 2711 + }, + { + "epoch": 2.1661341853035143, + "grad_norm": 0.2119140625, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 2712 + }, + { + "epoch": 2.166932907348243, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0472, + "step": 2713 + }, + { + "epoch": 2.1677316293929714, + "grad_norm": 0.322265625, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 2714 + }, + { + "epoch": 2.1685303514377, + "grad_norm": 0.2333984375, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 2715 + }, + { + "epoch": 2.169329073482428, + "grad_norm": 0.2080078125, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 2716 + }, + { + "epoch": 2.1701277955271565, + "grad_norm": 0.49609375, + "learning_rate": 0.0005, + "loss": 1.0686, + "step": 2717 + }, + { + "epoch": 2.170926517571885, + "grad_norm": 0.41015625, + "learning_rate": 0.0005, + "loss": 1.0751, + "step": 2718 + }, + { + "epoch": 2.1717252396166136, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 2719 + }, + { + "epoch": 2.1725239616613417, + "grad_norm": 0.283203125, + "learning_rate": 0.0005, + "loss": 1.0644, + "step": 2720 + }, + { + "epoch": 2.17332268370607, + "grad_norm": 0.1796875, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 2721 + }, + { + "epoch": 2.1741214057507987, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 2722 + }, + { + "epoch": 2.1749201277955272, + "grad_norm": 0.2578125, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 2723 + }, + { + "epoch": 2.1757188498402558, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 2724 + }, + { + "epoch": 2.176517571884984, + "grad_norm": 0.384765625, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 2725 + }, + { + "epoch": 2.1773162939297124, + "grad_norm": 0.357421875, + "learning_rate": 0.0005, + "loss": 1.0663, + "step": 2726 + }, + { + "epoch": 2.178115015974441, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 2727 + }, + { + "epoch": 2.1789137380191694, + "grad_norm": 0.439453125, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 2728 + }, + { + "epoch": 2.179712460063898, + "grad_norm": 0.2490234375, + "learning_rate": 0.0005, + "loss": 1.0678, + "step": 2729 + }, + { + "epoch": 2.180511182108626, + "grad_norm": 0.21875, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 2730 + }, + { + "epoch": 2.1813099041533546, + "grad_norm": 0.36328125, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 2731 + }, + { + "epoch": 2.182108626198083, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 2732 + }, + { + "epoch": 2.1829073482428116, + "grad_norm": 0.216796875, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 2733 + }, + { + "epoch": 2.18370607028754, + "grad_norm": 0.23828125, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 2734 + }, + { + "epoch": 2.1845047923322682, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 2735 + }, + { + "epoch": 2.1853035143769968, + "grad_norm": 0.1630859375, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 2736 + }, + { + "epoch": 2.1861022364217253, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 2737 + }, + { + "epoch": 2.186900958466454, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 2738 + }, + { + "epoch": 2.187699680511182, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 2739 + }, + { + "epoch": 2.1884984025559104, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 2740 + }, + { + "epoch": 2.189297124600639, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 2741 + }, + { + "epoch": 2.1900958466453675, + "grad_norm": 0.205078125, + "learning_rate": 0.0005, + "loss": 1.0651, + "step": 2742 + }, + { + "epoch": 2.190894568690096, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 2743 + }, + { + "epoch": 2.191693290734824, + "grad_norm": 0.3203125, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 2744 + }, + { + "epoch": 2.1924920127795526, + "grad_norm": 0.197265625, + "learning_rate": 0.0005, + "loss": 1.0482, + "step": 2745 + }, + { + "epoch": 2.193290734824281, + "grad_norm": 0.1982421875, + "learning_rate": 0.0005, + "loss": 1.0675, + "step": 2746 + }, + { + "epoch": 2.1940894568690097, + "grad_norm": 0.326171875, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 2747 + }, + { + "epoch": 2.194888178913738, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 2748 + }, + { + "epoch": 2.1956869009584663, + "grad_norm": 0.287109375, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 2749 + }, + { + "epoch": 2.196485623003195, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 2750 + }, + { + "epoch": 2.1972843450479234, + "grad_norm": 0.193359375, + "learning_rate": 0.0005, + "loss": 1.065, + "step": 2751 + }, + { + "epoch": 2.198083067092652, + "grad_norm": 0.234375, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 2752 + }, + { + "epoch": 2.1988817891373804, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 2753 + }, + { + "epoch": 2.1996805111821085, + "grad_norm": 0.2470703125, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 2754 + }, + { + "epoch": 2.200479233226837, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 2755 + }, + { + "epoch": 2.2012779552715656, + "grad_norm": 0.31640625, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 2756 + }, + { + "epoch": 2.202076677316294, + "grad_norm": 0.19140625, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 2757 + }, + { + "epoch": 2.202875399361022, + "grad_norm": 0.1943359375, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 2758 + }, + { + "epoch": 2.2036741214057507, + "grad_norm": 0.275390625, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 2759 + }, + { + "epoch": 2.2044728434504792, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 2760 + }, + { + "epoch": 2.2052715654952078, + "grad_norm": 0.337890625, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 2761 + }, + { + "epoch": 2.2060702875399363, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 2762 + }, + { + "epoch": 2.2068690095846644, + "grad_norm": 0.32421875, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 2763 + }, + { + "epoch": 2.207667731629393, + "grad_norm": 0.193359375, + "learning_rate": 0.0005, + "loss": 1.0661, + "step": 2764 + }, + { + "epoch": 2.2084664536741214, + "grad_norm": 0.1748046875, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 2765 + }, + { + "epoch": 2.20926517571885, + "grad_norm": 0.185546875, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 2766 + }, + { + "epoch": 2.2100638977635785, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 2767 + }, + { + "epoch": 2.2108626198083066, + "grad_norm": 0.234375, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 2768 + }, + { + "epoch": 2.211661341853035, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 2769 + }, + { + "epoch": 2.2124600638977636, + "grad_norm": 0.2255859375, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 2770 + }, + { + "epoch": 2.213258785942492, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 2771 + }, + { + "epoch": 2.2140575079872207, + "grad_norm": 0.2490234375, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 2772 + }, + { + "epoch": 2.2148562300319488, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 2773 + }, + { + "epoch": 2.2156549520766773, + "grad_norm": 0.1875, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 2774 + }, + { + "epoch": 2.216453674121406, + "grad_norm": 0.234375, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 2775 + }, + { + "epoch": 2.2172523961661343, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 2776 + }, + { + "epoch": 2.2180511182108624, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0691, + "step": 2777 + }, + { + "epoch": 2.218849840255591, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0704, + "step": 2778 + }, + { + "epoch": 2.2196485623003195, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 2779 + }, + { + "epoch": 2.220447284345048, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.064, + "step": 2780 + }, + { + "epoch": 2.2212460063897765, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 2781 + }, + { + "epoch": 2.2220447284345046, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 2782 + }, + { + "epoch": 2.222843450479233, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 2783 + }, + { + "epoch": 2.2236421725239617, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 2784 + }, + { + "epoch": 2.22444089456869, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 2785 + }, + { + "epoch": 2.2252396166134187, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 2786 + }, + { + "epoch": 2.226038338658147, + "grad_norm": 0.18359375, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 2787 + }, + { + "epoch": 2.2268370607028753, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 2788 + }, + { + "epoch": 2.227635782747604, + "grad_norm": 0.1611328125, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 2789 + }, + { + "epoch": 2.2284345047923324, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 2790 + }, + { + "epoch": 2.229233226837061, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 2791 + }, + { + "epoch": 2.230031948881789, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 2792 + }, + { + "epoch": 2.2308306709265175, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 2793 + }, + { + "epoch": 2.231629392971246, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 2794 + }, + { + "epoch": 2.2324281150159746, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 2795 + }, + { + "epoch": 2.2332268370607027, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 2796 + }, + { + "epoch": 2.234025559105431, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 2797 + }, + { + "epoch": 2.2348242811501597, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0624, + "step": 2798 + }, + { + "epoch": 2.2356230031948883, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 2799 + }, + { + "epoch": 2.236421725239617, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.0442, + "step": 2800 + }, + { + "epoch": 2.237220447284345, + "grad_norm": 0.236328125, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 2801 + }, + { + "epoch": 2.2380191693290734, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 2802 + }, + { + "epoch": 2.238817891373802, + "grad_norm": 0.1845703125, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 2803 + }, + { + "epoch": 2.2396166134185305, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 2804 + }, + { + "epoch": 2.2404153354632586, + "grad_norm": 0.26171875, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 2805 + }, + { + "epoch": 2.241214057507987, + "grad_norm": 0.2451171875, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 2806 + }, + { + "epoch": 2.2420127795527156, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 2807 + }, + { + "epoch": 2.242811501597444, + "grad_norm": 0.2109375, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 2808 + }, + { + "epoch": 2.2436102236421727, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 2809 + }, + { + "epoch": 2.244408945686901, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.0648, + "step": 2810 + }, + { + "epoch": 2.2452076677316293, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 2811 + }, + { + "epoch": 2.246006389776358, + "grad_norm": 0.189453125, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 2812 + }, + { + "epoch": 2.2468051118210863, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 2813 + }, + { + "epoch": 2.247603833865815, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 2814 + }, + { + "epoch": 2.248402555910543, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 2815 + }, + { + "epoch": 2.2492012779552715, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 2816 + }, + { + "epoch": 2.25, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 2817 + }, + { + "epoch": 2.2507987220447285, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 2818 + }, + { + "epoch": 2.251597444089457, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 2819 + }, + { + "epoch": 2.252396166134185, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 2820 + }, + { + "epoch": 2.2531948881789137, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 2821 + }, + { + "epoch": 2.253993610223642, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 2822 + }, + { + "epoch": 2.2547923322683707, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.0635, + "step": 2823 + }, + { + "epoch": 2.255591054313099, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 2824 + }, + { + "epoch": 2.2563897763578273, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 2825 + }, + { + "epoch": 2.257188498402556, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 2826 + }, + { + "epoch": 2.2579872204472844, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0657, + "step": 2827 + }, + { + "epoch": 2.258785942492013, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 2828 + }, + { + "epoch": 2.2595846645367414, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 2829 + }, + { + "epoch": 2.2603833865814695, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 2830 + }, + { + "epoch": 2.261182108626198, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 2831 + }, + { + "epoch": 2.2619808306709266, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 2832 + }, + { + "epoch": 2.262779552715655, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 2833 + }, + { + "epoch": 2.263578274760383, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 2834 + }, + { + "epoch": 2.2643769968051117, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 2835 + }, + { + "epoch": 2.2651757188498403, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 2836 + }, + { + "epoch": 2.265974440894569, + "grad_norm": 0.1748046875, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 2837 + }, + { + "epoch": 2.2667731629392973, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 2838 + }, + { + "epoch": 2.2675718849840254, + "grad_norm": 0.21875, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 2839 + }, + { + "epoch": 2.268370607028754, + "grad_norm": 0.251953125, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 2840 + }, + { + "epoch": 2.2691693290734825, + "grad_norm": 0.1708984375, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 2841 + }, + { + "epoch": 2.269968051118211, + "grad_norm": 0.5390625, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 2842 + }, + { + "epoch": 2.270766773162939, + "grad_norm": 0.6484375, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 2843 + }, + { + "epoch": 2.2715654952076676, + "grad_norm": 0.6015625, + "learning_rate": 0.0005, + "loss": 1.0718, + "step": 2844 + }, + { + "epoch": 2.272364217252396, + "grad_norm": 0.3203125, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 2845 + }, + { + "epoch": 2.2731629392971247, + "grad_norm": 0.2451171875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 2846 + }, + { + "epoch": 2.273961661341853, + "grad_norm": 0.5625, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 2847 + }, + { + "epoch": 2.2747603833865817, + "grad_norm": 0.361328125, + "learning_rate": 0.0005, + "loss": 1.0621, + "step": 2848 + }, + { + "epoch": 2.27555910543131, + "grad_norm": 0.2490234375, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 2849 + }, + { + "epoch": 2.2763578274760383, + "grad_norm": 0.65625, + "learning_rate": 0.0005, + "loss": 1.0653, + "step": 2850 + }, + { + "epoch": 2.277156549520767, + "grad_norm": 0.53125, + "learning_rate": 0.0005, + "loss": 1.0746, + "step": 2851 + }, + { + "epoch": 2.2779552715654954, + "grad_norm": 0.1875, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 2852 + }, + { + "epoch": 2.2787539936102235, + "grad_norm": 0.78125, + "learning_rate": 0.0005, + "loss": 1.0747, + "step": 2853 + }, + { + "epoch": 2.279552715654952, + "grad_norm": 0.72265625, + "learning_rate": 0.0005, + "loss": 1.0711, + "step": 2854 + }, + { + "epoch": 2.2803514376996805, + "grad_norm": 0.205078125, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 2855 + }, + { + "epoch": 2.281150159744409, + "grad_norm": 0.62109375, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 2856 + }, + { + "epoch": 2.2819488817891376, + "grad_norm": 0.53515625, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 2857 + }, + { + "epoch": 2.2827476038338657, + "grad_norm": 0.2041015625, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 2858 + }, + { + "epoch": 2.283546325878594, + "grad_norm": 0.470703125, + "learning_rate": 0.0005, + "loss": 1.072, + "step": 2859 + }, + { + "epoch": 2.2843450479233227, + "grad_norm": 0.1767578125, + "learning_rate": 0.0005, + "loss": 1.0612, + "step": 2860 + }, + { + "epoch": 2.2851437699680512, + "grad_norm": 0.734375, + "learning_rate": 0.0005, + "loss": 1.0649, + "step": 2861 + }, + { + "epoch": 2.2859424920127793, + "grad_norm": 0.45703125, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 2862 + }, + { + "epoch": 2.286741214057508, + "grad_norm": 0.287109375, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 2863 + }, + { + "epoch": 2.2875399361022364, + "grad_norm": 0.439453125, + "learning_rate": 0.0005, + "loss": 1.0683, + "step": 2864 + }, + { + "epoch": 2.288338658146965, + "grad_norm": 0.2158203125, + "learning_rate": 0.0005, + "loss": 1.0703, + "step": 2865 + }, + { + "epoch": 2.2891373801916934, + "grad_norm": 0.29296875, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 2866 + }, + { + "epoch": 2.289936102236422, + "grad_norm": 0.2373046875, + "learning_rate": 0.0005, + "loss": 1.0681, + "step": 2867 + }, + { + "epoch": 2.29073482428115, + "grad_norm": 0.2099609375, + "learning_rate": 0.0005, + "loss": 1.0632, + "step": 2868 + }, + { + "epoch": 2.2915335463258786, + "grad_norm": 0.283203125, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 2869 + }, + { + "epoch": 2.292332268370607, + "grad_norm": 0.203125, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 2870 + }, + { + "epoch": 2.2931309904153356, + "grad_norm": 0.30078125, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 2871 + }, + { + "epoch": 2.2939297124600637, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 2872 + }, + { + "epoch": 2.2947284345047922, + "grad_norm": 0.291015625, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 2873 + }, + { + "epoch": 2.2955271565495208, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 2874 + }, + { + "epoch": 2.2963258785942493, + "grad_norm": 0.193359375, + "learning_rate": 0.0005, + "loss": 1.0652, + "step": 2875 + }, + { + "epoch": 2.297124600638978, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 2876 + }, + { + "epoch": 2.297923322683706, + "grad_norm": 0.2109375, + "learning_rate": 0.0005, + "loss": 1.0668, + "step": 2877 + }, + { + "epoch": 2.2987220447284344, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 2878 + }, + { + "epoch": 2.299520766773163, + "grad_norm": 0.2080078125, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 2879 + }, + { + "epoch": 2.3003194888178915, + "grad_norm": 0.1796875, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 2880 + }, + { + "epoch": 2.3011182108626196, + "grad_norm": 0.25390625, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 2881 + }, + { + "epoch": 2.301916932907348, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.073, + "step": 2882 + }, + { + "epoch": 2.3027156549520766, + "grad_norm": 0.255859375, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 2883 + }, + { + "epoch": 2.303514376996805, + "grad_norm": 0.1640625, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 2884 + }, + { + "epoch": 2.3043130990415337, + "grad_norm": 0.357421875, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 2885 + }, + { + "epoch": 2.3051118210862622, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 2886 + }, + { + "epoch": 2.3059105431309903, + "grad_norm": 0.265625, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 2887 + }, + { + "epoch": 2.306709265175719, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0687, + "step": 2888 + }, + { + "epoch": 2.3075079872204474, + "grad_norm": 0.2099609375, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 2889 + }, + { + "epoch": 2.308306709265176, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 2890 + }, + { + "epoch": 2.309105431309904, + "grad_norm": 0.271484375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 2891 + }, + { + "epoch": 2.3099041533546325, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 2892 + }, + { + "epoch": 2.310702875399361, + "grad_norm": 0.255859375, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 2893 + }, + { + "epoch": 2.3115015974440896, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0662, + "step": 2894 + }, + { + "epoch": 2.312300319488818, + "grad_norm": 0.1640625, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 2895 + }, + { + "epoch": 2.313099041533546, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 2896 + }, + { + "epoch": 2.3138977635782747, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 2897 + }, + { + "epoch": 2.3146964856230032, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 2898 + }, + { + "epoch": 2.3154952076677318, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 2899 + }, + { + "epoch": 2.31629392971246, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 2900 + }, + { + "epoch": 2.3170926517571884, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 2901 + }, + { + "epoch": 2.317891373801917, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 2902 + }, + { + "epoch": 2.3186900958466454, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 2903 + }, + { + "epoch": 2.319488817891374, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 2904 + }, + { + "epoch": 2.3202875399361025, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0629, + "step": 2905 + }, + { + "epoch": 2.3210862619808306, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 2906 + }, + { + "epoch": 2.321884984025559, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 2907 + }, + { + "epoch": 2.3226837060702876, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 2908 + }, + { + "epoch": 2.323482428115016, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0698, + "step": 2909 + }, + { + "epoch": 2.3242811501597442, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 2910 + }, + { + "epoch": 2.3250798722044728, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 2911 + }, + { + "epoch": 2.3258785942492013, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 2912 + }, + { + "epoch": 2.32667731629393, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0639, + "step": 2913 + }, + { + "epoch": 2.3274760383386583, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 2914 + }, + { + "epoch": 2.3282747603833864, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 2915 + }, + { + "epoch": 2.329073482428115, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 2916 + }, + { + "epoch": 2.3298722044728435, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 2917 + }, + { + "epoch": 2.330670926517572, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 2918 + }, + { + "epoch": 2.3314696485623, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 2919 + }, + { + "epoch": 2.3322683706070286, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 2920 + }, + { + "epoch": 2.333067092651757, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 2921 + }, + { + "epoch": 2.3338658146964857, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 2922 + }, + { + "epoch": 2.334664536741214, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0459, + "step": 2923 + }, + { + "epoch": 2.3354632587859427, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 2924 + }, + { + "epoch": 2.336261980830671, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 2925 + }, + { + "epoch": 2.3370607028753994, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 2926 + }, + { + "epoch": 2.337859424920128, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 2927 + }, + { + "epoch": 2.3386581469648564, + "grad_norm": 0.1630859375, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 2928 + }, + { + "epoch": 2.3394568690095845, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 2929 + }, + { + "epoch": 2.340255591054313, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 2930 + }, + { + "epoch": 2.3410543130990416, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 2931 + }, + { + "epoch": 2.34185303514377, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 2932 + }, + { + "epoch": 2.3426517571884986, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0464, + "step": 2933 + }, + { + "epoch": 2.3434504792332267, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 2934 + }, + { + "epoch": 2.344249201277955, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 2935 + }, + { + "epoch": 2.3450479233226837, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 2936 + }, + { + "epoch": 2.3458466453674123, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 2937 + }, + { + "epoch": 2.3466453674121404, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 2938 + }, + { + "epoch": 2.347444089456869, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 2939 + }, + { + "epoch": 2.3482428115015974, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0457, + "step": 2940 + }, + { + "epoch": 2.349041533546326, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 2941 + }, + { + "epoch": 2.3498402555910545, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 2942 + }, + { + "epoch": 2.3506389776357826, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 2943 + }, + { + "epoch": 2.351437699680511, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 2944 + }, + { + "epoch": 2.3522364217252396, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 2945 + }, + { + "epoch": 2.353035143769968, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 2946 + }, + { + "epoch": 2.3538338658146967, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 2947 + }, + { + "epoch": 2.3546325878594248, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 2948 + }, + { + "epoch": 2.3554313099041533, + "grad_norm": 0.20703125, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 2949 + }, + { + "epoch": 2.356230031948882, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 2950 + }, + { + "epoch": 2.3570287539936103, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 2951 + }, + { + "epoch": 2.357827476038339, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 2952 + }, + { + "epoch": 2.358626198083067, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 2953 + }, + { + "epoch": 2.3594249201277955, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 2954 + }, + { + "epoch": 2.360223642172524, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 2955 + }, + { + "epoch": 2.3610223642172525, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 2956 + }, + { + "epoch": 2.3618210862619806, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 2957 + }, + { + "epoch": 2.362619808306709, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 2958 + }, + { + "epoch": 2.3634185303514377, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.0448, + "step": 2959 + }, + { + "epoch": 2.364217252396166, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 2960 + }, + { + "epoch": 2.3650159744408947, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 2961 + }, + { + "epoch": 2.365814696485623, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 2962 + }, + { + "epoch": 2.3666134185303513, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 2963 + }, + { + "epoch": 2.36741214057508, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 2964 + }, + { + "epoch": 2.3682108626198084, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0642, + "step": 2965 + }, + { + "epoch": 2.369009584664537, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 2966 + }, + { + "epoch": 2.369808306709265, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 2967 + }, + { + "epoch": 2.3706070287539935, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 2968 + }, + { + "epoch": 2.371405750798722, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 2969 + }, + { + "epoch": 2.3722044728434506, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 2970 + }, + { + "epoch": 2.373003194888179, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 2971 + }, + { + "epoch": 2.373801916932907, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 2972 + }, + { + "epoch": 2.3746006389776357, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0647, + "step": 2973 + }, + { + "epoch": 2.3753993610223643, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 2974 + }, + { + "epoch": 2.376198083067093, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 2975 + }, + { + "epoch": 2.376996805111821, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 2976 + }, + { + "epoch": 2.3777955271565494, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 2977 + }, + { + "epoch": 2.378594249201278, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 2978 + }, + { + "epoch": 2.3793929712460065, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 2979 + }, + { + "epoch": 2.380191693290735, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 2980 + }, + { + "epoch": 2.380990415335463, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 2981 + }, + { + "epoch": 2.3817891373801916, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 2982 + }, + { + "epoch": 2.38258785942492, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 2983 + }, + { + "epoch": 2.3833865814696487, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 2984 + }, + { + "epoch": 2.384185303514377, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 2985 + }, + { + "epoch": 2.3849840255591053, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 2986 + }, + { + "epoch": 2.385782747603834, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 2987 + }, + { + "epoch": 2.3865814696485623, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 2988 + }, + { + "epoch": 2.387380191693291, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 2989 + }, + { + "epoch": 2.3881789137380194, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 2990 + }, + { + "epoch": 2.3889776357827475, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 2991 + }, + { + "epoch": 2.389776357827476, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 2992 + }, + { + "epoch": 2.3905750798722045, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 2993 + }, + { + "epoch": 2.391373801916933, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0458, + "step": 2994 + }, + { + "epoch": 2.392172523961661, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0464, + "step": 2995 + }, + { + "epoch": 2.3929712460063897, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 2996 + }, + { + "epoch": 2.393769968051118, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 2997 + }, + { + "epoch": 2.3945686900958467, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 2998 + }, + { + "epoch": 2.3953674121405752, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.0679, + "step": 2999 + }, + { + "epoch": 2.3961661341853033, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 3000 + }, + { + "epoch": 2.396964856230032, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 3001 + }, + { + "epoch": 2.3977635782747604, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 3002 + }, + { + "epoch": 2.398562300319489, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 3003 + }, + { + "epoch": 2.3993610223642174, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 3004 + }, + { + "epoch": 2.4001597444089455, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 3005 + }, + { + "epoch": 2.400958466453674, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 3006 + }, + { + "epoch": 2.4017571884984026, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 3007 + }, + { + "epoch": 2.402555910543131, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 3008 + }, + { + "epoch": 2.4033546325878596, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 3009 + }, + { + "epoch": 2.4041533546325877, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 3010 + }, + { + "epoch": 2.4049520766773163, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 3011 + }, + { + "epoch": 2.405750798722045, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 3012 + }, + { + "epoch": 2.4065495207667733, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 3013 + }, + { + "epoch": 2.4073482428115014, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 3014 + }, + { + "epoch": 2.40814696485623, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 3015 + }, + { + "epoch": 2.4089456869009584, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 3016 + }, + { + "epoch": 2.409744408945687, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 3017 + }, + { + "epoch": 2.4105431309904155, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 3018 + }, + { + "epoch": 2.4113418530351436, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 3019 + }, + { + "epoch": 2.412140575079872, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 3020 + }, + { + "epoch": 2.4129392971246006, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0658, + "step": 3021 + }, + { + "epoch": 2.413738019169329, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 3022 + }, + { + "epoch": 2.4145367412140573, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 3023 + }, + { + "epoch": 2.415335463258786, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 3024 + }, + { + "epoch": 2.4161341853035143, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 3025 + }, + { + "epoch": 2.416932907348243, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 3026 + }, + { + "epoch": 2.4177316293929714, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0452, + "step": 3027 + }, + { + "epoch": 2.4185303514377, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 3028 + }, + { + "epoch": 2.419329073482428, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 3029 + }, + { + "epoch": 2.4201277955271565, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 3030 + }, + { + "epoch": 2.420926517571885, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 3031 + }, + { + "epoch": 2.4217252396166136, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 3032 + }, + { + "epoch": 2.4225239616613417, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 3033 + }, + { + "epoch": 2.42332268370607, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 3034 + }, + { + "epoch": 2.4241214057507987, + "grad_norm": 0.1630859375, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 3035 + }, + { + "epoch": 2.4249201277955272, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 3036 + }, + { + "epoch": 2.4257188498402558, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 3037 + }, + { + "epoch": 2.426517571884984, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 3038 + }, + { + "epoch": 2.4273162939297124, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 3039 + }, + { + "epoch": 2.428115015974441, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 3040 + }, + { + "epoch": 2.4289137380191694, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 3041 + }, + { + "epoch": 2.4297124600638975, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 3042 + }, + { + "epoch": 2.430511182108626, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0456, + "step": 3043 + }, + { + "epoch": 2.4313099041533546, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 3044 + }, + { + "epoch": 2.432108626198083, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 3045 + }, + { + "epoch": 2.4329073482428116, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 3046 + }, + { + "epoch": 2.43370607028754, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 3047 + }, + { + "epoch": 2.4345047923322682, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 3048 + }, + { + "epoch": 2.4353035143769968, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0667, + "step": 3049 + }, + { + "epoch": 2.4361022364217253, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 3050 + }, + { + "epoch": 2.436900958466454, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 3051 + }, + { + "epoch": 2.437699680511182, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 3052 + }, + { + "epoch": 2.4384984025559104, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 3053 + }, + { + "epoch": 2.439297124600639, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 3054 + }, + { + "epoch": 2.4400958466453675, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 3055 + }, + { + "epoch": 2.440894568690096, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 3056 + }, + { + "epoch": 2.441693290734824, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 3057 + }, + { + "epoch": 2.4424920127795526, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 3058 + }, + { + "epoch": 2.443290734824281, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 3059 + }, + { + "epoch": 2.4440894568690097, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0655, + "step": 3060 + }, + { + "epoch": 2.4448881789137378, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 3061 + }, + { + "epoch": 2.4456869009584663, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 3062 + }, + { + "epoch": 2.446485623003195, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 3063 + }, + { + "epoch": 2.4472843450479234, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 3064 + }, + { + "epoch": 2.448083067092652, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 3065 + }, + { + "epoch": 2.4488817891373804, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 3066 + }, + { + "epoch": 2.4496805111821085, + "grad_norm": 0.1923828125, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 3067 + }, + { + "epoch": 2.450479233226837, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 3068 + }, + { + "epoch": 2.4512779552715656, + "grad_norm": 0.212890625, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 3069 + }, + { + "epoch": 2.452076677316294, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 3070 + }, + { + "epoch": 2.452875399361022, + "grad_norm": 0.296875, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 3071 + }, + { + "epoch": 2.4536741214057507, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 3072 + }, + { + "epoch": 2.4544728434504792, + "grad_norm": 0.244140625, + "learning_rate": 0.0005, + "loss": 1.0646, + "step": 3073 + }, + { + "epoch": 2.4552715654952078, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 3074 + }, + { + "epoch": 2.4560702875399363, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 3075 + }, + { + "epoch": 2.4568690095846644, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 3076 + }, + { + "epoch": 2.457667731629393, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 3077 + }, + { + "epoch": 2.4584664536741214, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 3078 + }, + { + "epoch": 2.45926517571885, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 3079 + }, + { + "epoch": 2.460063897763578, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 3080 + }, + { + "epoch": 2.4608626198083066, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 3081 + }, + { + "epoch": 2.461661341853035, + "grad_norm": 0.1923828125, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 3082 + }, + { + "epoch": 2.4624600638977636, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 3083 + }, + { + "epoch": 2.463258785942492, + "grad_norm": 0.2041015625, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 3084 + }, + { + "epoch": 2.4640575079872207, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 3085 + }, + { + "epoch": 2.4648562300319488, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 3086 + }, + { + "epoch": 2.4656549520766773, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 3087 + }, + { + "epoch": 2.466453674121406, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 3088 + }, + { + "epoch": 2.4672523961661343, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 3089 + }, + { + "epoch": 2.4680511182108624, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 3090 + }, + { + "epoch": 2.468849840255591, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 3091 + }, + { + "epoch": 2.4696485623003195, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 3092 + }, + { + "epoch": 2.470447284345048, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 3093 + }, + { + "epoch": 2.4712460063897765, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 3094 + }, + { + "epoch": 2.4720447284345046, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 3095 + }, + { + "epoch": 2.472843450479233, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 3096 + }, + { + "epoch": 2.4736421725239617, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 3097 + }, + { + "epoch": 2.47444089456869, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 3098 + }, + { + "epoch": 2.4752396166134183, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 3099 + }, + { + "epoch": 2.476038338658147, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 3100 + }, + { + "epoch": 2.4768370607028753, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 3101 + }, + { + "epoch": 2.477635782747604, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 3102 + }, + { + "epoch": 2.4784345047923324, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 3103 + }, + { + "epoch": 2.479233226837061, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 3104 + }, + { + "epoch": 2.480031948881789, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 3105 + }, + { + "epoch": 2.4808306709265175, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 3106 + }, + { + "epoch": 2.481629392971246, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 3107 + }, + { + "epoch": 2.4824281150159746, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 3108 + }, + { + "epoch": 2.4832268370607027, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 3109 + }, + { + "epoch": 2.484025559105431, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 3110 + }, + { + "epoch": 2.4848242811501597, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 3111 + }, + { + "epoch": 2.4856230031948883, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 3112 + }, + { + "epoch": 2.486421725239617, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 3113 + }, + { + "epoch": 2.487220447284345, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 3114 + }, + { + "epoch": 2.4880191693290734, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 3115 + }, + { + "epoch": 2.488817891373802, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 3116 + }, + { + "epoch": 2.4896166134185305, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 3117 + }, + { + "epoch": 2.4904153354632586, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 3118 + }, + { + "epoch": 2.491214057507987, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0444, + "step": 3119 + }, + { + "epoch": 2.4920127795527156, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 3120 + }, + { + "epoch": 2.492811501597444, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 3121 + }, + { + "epoch": 2.4936102236421727, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 3122 + }, + { + "epoch": 2.494408945686901, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 3123 + }, + { + "epoch": 2.4952076677316293, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 3124 + }, + { + "epoch": 2.496006389776358, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 3125 + }, + { + "epoch": 2.4968051118210863, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 3126 + }, + { + "epoch": 2.497603833865815, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0488, + "step": 3127 + }, + { + "epoch": 2.498402555910543, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 3128 + }, + { + "epoch": 2.4992012779552715, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0455, + "step": 3129 + }, + { + "epoch": 2.5, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 3130 + }, + { + "epoch": 2.5007987220447285, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 3131 + }, + { + "epoch": 2.501597444089457, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 3132 + }, + { + "epoch": 2.502396166134185, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0582, + "step": 3133 + }, + { + "epoch": 2.5031948881789137, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 3134 + }, + { + "epoch": 2.503993610223642, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 3135 + }, + { + "epoch": 2.5047923322683707, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 3136 + }, + { + "epoch": 2.505591054313099, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 3137 + }, + { + "epoch": 2.5063897763578273, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 3138 + }, + { + "epoch": 2.507188498402556, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 3139 + }, + { + "epoch": 2.5079872204472844, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 3140 + }, + { + "epoch": 2.508785942492013, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 3141 + }, + { + "epoch": 2.5095846645367414, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 3142 + }, + { + "epoch": 2.5103833865814695, + "grad_norm": 0.169921875, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 3143 + }, + { + "epoch": 2.511182108626198, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 3144 + }, + { + "epoch": 2.5119808306709266, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 3145 + }, + { + "epoch": 2.512779552715655, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 3146 + }, + { + "epoch": 2.513578274760383, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 3147 + }, + { + "epoch": 2.5143769968051117, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0441, + "step": 3148 + }, + { + "epoch": 2.5151757188498403, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0685, + "step": 3149 + }, + { + "epoch": 2.515974440894569, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0605, + "step": 3150 + }, + { + "epoch": 2.5167731629392973, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 3151 + }, + { + "epoch": 2.5175718849840254, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 3152 + }, + { + "epoch": 2.518370607028754, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 3153 + }, + { + "epoch": 2.5191693290734825, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0468, + "step": 3154 + }, + { + "epoch": 2.519968051118211, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 3155 + }, + { + "epoch": 2.520766773162939, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 3156 + }, + { + "epoch": 2.5215654952076676, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0591, + "step": 3157 + }, + { + "epoch": 2.522364217252396, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 3158 + }, + { + "epoch": 2.5231629392971247, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 3159 + }, + { + "epoch": 2.523961661341853, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 3160 + }, + { + "epoch": 2.5247603833865817, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 3161 + }, + { + "epoch": 2.52555910543131, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 3162 + }, + { + "epoch": 2.5263578274760383, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 3163 + }, + { + "epoch": 2.527156549520767, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0446, + "step": 3164 + }, + { + "epoch": 2.527955271565495, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 3165 + }, + { + "epoch": 2.5287539936102235, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 3166 + }, + { + "epoch": 2.529552715654952, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0434, + "step": 3167 + }, + { + "epoch": 2.5303514376996805, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 3168 + }, + { + "epoch": 2.531150159744409, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 3169 + }, + { + "epoch": 2.5319488817891376, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 3170 + }, + { + "epoch": 2.5327476038338657, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 3171 + }, + { + "epoch": 2.533546325878594, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 3172 + }, + { + "epoch": 2.5343450479233227, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 3173 + }, + { + "epoch": 2.5351437699680512, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 3174 + }, + { + "epoch": 2.5359424920127793, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 3175 + }, + { + "epoch": 2.536741214057508, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0482, + "step": 3176 + }, + { + "epoch": 2.5375399361022364, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 3177 + }, + { + "epoch": 2.538338658146965, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 3178 + }, + { + "epoch": 2.5391373801916934, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0638, + "step": 3179 + }, + { + "epoch": 2.539936102236422, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 3180 + }, + { + "epoch": 2.54073482428115, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 3181 + }, + { + "epoch": 2.5415335463258786, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0664, + "step": 3182 + }, + { + "epoch": 2.542332268370607, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 3183 + }, + { + "epoch": 2.543130990415335, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 3184 + }, + { + "epoch": 2.5439297124600637, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 3185 + }, + { + "epoch": 2.5447284345047922, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 3186 + }, + { + "epoch": 2.5455271565495208, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 3187 + }, + { + "epoch": 2.5463258785942493, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 3188 + }, + { + "epoch": 2.547124600638978, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 3189 + }, + { + "epoch": 2.547923322683706, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 3190 + }, + { + "epoch": 2.5487220447284344, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 3191 + }, + { + "epoch": 2.549520766773163, + "grad_norm": 0.1572265625, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 3192 + }, + { + "epoch": 2.5503194888178915, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 3193 + }, + { + "epoch": 2.5511182108626196, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 3194 + }, + { + "epoch": 2.551916932907348, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 3195 + }, + { + "epoch": 2.5527156549520766, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 3196 + }, + { + "epoch": 2.553514376996805, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0672, + "step": 3197 + }, + { + "epoch": 2.5543130990415337, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 3198 + }, + { + "epoch": 2.5551118210862622, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0376, + "step": 3199 + }, + { + "epoch": 2.5559105431309903, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 3200 + }, + { + "epoch": 2.556709265175719, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 3201 + }, + { + "epoch": 2.5575079872204474, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 3202 + }, + { + "epoch": 2.5583067092651754, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 3203 + }, + { + "epoch": 2.559105431309904, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 3204 + }, + { + "epoch": 2.5599041533546325, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 3205 + }, + { + "epoch": 2.560702875399361, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 3206 + }, + { + "epoch": 2.5615015974440896, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 3207 + }, + { + "epoch": 2.562300319488818, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 3208 + }, + { + "epoch": 2.563099041533546, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 3209 + }, + { + "epoch": 2.5638977635782747, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0656, + "step": 3210 + }, + { + "epoch": 2.5646964856230032, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 3211 + }, + { + "epoch": 2.5654952076677318, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 3212 + }, + { + "epoch": 2.56629392971246, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 3213 + }, + { + "epoch": 2.5670926517571884, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 3214 + }, + { + "epoch": 2.567891373801917, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 3215 + }, + { + "epoch": 2.5686900958466454, + "grad_norm": 0.166015625, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 3216 + }, + { + "epoch": 2.569488817891374, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 3217 + }, + { + "epoch": 2.5702875399361025, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 3218 + }, + { + "epoch": 2.5710862619808306, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 3219 + }, + { + "epoch": 2.571884984025559, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 3220 + }, + { + "epoch": 2.5726837060702876, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 3221 + }, + { + "epoch": 2.5734824281150157, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 3222 + }, + { + "epoch": 2.5742811501597442, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 3223 + }, + { + "epoch": 2.5750798722044728, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0449, + "step": 3224 + }, + { + "epoch": 2.5758785942492013, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 3225 + }, + { + "epoch": 2.57667731629393, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 3226 + }, + { + "epoch": 2.5774760383386583, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 3227 + }, + { + "epoch": 2.5782747603833864, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 3228 + }, + { + "epoch": 2.579073482428115, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 3229 + }, + { + "epoch": 2.5798722044728435, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 3230 + }, + { + "epoch": 2.580670926517572, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 3231 + }, + { + "epoch": 2.5814696485623, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 3232 + }, + { + "epoch": 2.5822683706070286, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 3233 + }, + { + "epoch": 2.583067092651757, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0447, + "step": 3234 + }, + { + "epoch": 2.5838658146964857, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 3235 + }, + { + "epoch": 2.584664536741214, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 3236 + }, + { + "epoch": 2.5854632587859427, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 3237 + }, + { + "epoch": 2.586261980830671, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 3238 + }, + { + "epoch": 2.5870607028753994, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 3239 + }, + { + "epoch": 2.587859424920128, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 3240 + }, + { + "epoch": 2.588658146964856, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 3241 + }, + { + "epoch": 2.5894568690095845, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 3242 + }, + { + "epoch": 2.590255591054313, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 3243 + }, + { + "epoch": 2.5910543130990416, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.063, + "step": 3244 + }, + { + "epoch": 2.59185303514377, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 3245 + }, + { + "epoch": 2.5926517571884986, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 3246 + }, + { + "epoch": 2.5934504792332267, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 3247 + }, + { + "epoch": 2.594249201277955, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 3248 + }, + { + "epoch": 2.5950479233226837, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 3249 + }, + { + "epoch": 2.5958466453674123, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 3250 + }, + { + "epoch": 2.5966453674121404, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0609, + "step": 3251 + }, + { + "epoch": 2.597444089456869, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 3252 + }, + { + "epoch": 2.5982428115015974, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 3253 + }, + { + "epoch": 2.599041533546326, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 3254 + }, + { + "epoch": 2.5998402555910545, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 3255 + }, + { + "epoch": 2.600638977635783, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 3256 + }, + { + "epoch": 2.601437699680511, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 3257 + }, + { + "epoch": 2.6022364217252396, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 3258 + }, + { + "epoch": 2.603035143769968, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 3259 + }, + { + "epoch": 2.6038338658146962, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 3260 + }, + { + "epoch": 2.6046325878594248, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 3261 + }, + { + "epoch": 2.6054313099041533, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 3262 + }, + { + "epoch": 2.606230031948882, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0627, + "step": 3263 + }, + { + "epoch": 2.6070287539936103, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 3264 + }, + { + "epoch": 2.607827476038339, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0604, + "step": 3265 + }, + { + "epoch": 2.608626198083067, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 3266 + }, + { + "epoch": 2.6094249201277955, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 3267 + }, + { + "epoch": 2.610223642172524, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0448, + "step": 3268 + }, + { + "epoch": 2.6110223642172525, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 3269 + }, + { + "epoch": 2.6118210862619806, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 3270 + }, + { + "epoch": 2.612619808306709, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0421, + "step": 3271 + }, + { + "epoch": 2.6134185303514377, + "grad_norm": 0.2236328125, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 3272 + }, + { + "epoch": 2.614217252396166, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 3273 + }, + { + "epoch": 2.6150159744408947, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 3274 + }, + { + "epoch": 2.6158146964856233, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 3275 + }, + { + "epoch": 2.6166134185303513, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 3276 + }, + { + "epoch": 2.61741214057508, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 3277 + }, + { + "epoch": 2.6182108626198084, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0442, + "step": 3278 + }, + { + "epoch": 2.6190095846645365, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 3279 + }, + { + "epoch": 2.619808306709265, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 3280 + }, + { + "epoch": 2.6206070287539935, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 3281 + }, + { + "epoch": 2.621405750798722, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 3282 + }, + { + "epoch": 2.6222044728434506, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0429, + "step": 3283 + }, + { + "epoch": 2.623003194888179, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 3284 + }, + { + "epoch": 2.623801916932907, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0457, + "step": 3285 + }, + { + "epoch": 2.6246006389776357, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 3286 + }, + { + "epoch": 2.6253993610223643, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 3287 + }, + { + "epoch": 2.626198083067093, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 3288 + }, + { + "epoch": 2.626996805111821, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 3289 + }, + { + "epoch": 2.6277955271565494, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 3290 + }, + { + "epoch": 2.628594249201278, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 3291 + }, + { + "epoch": 2.6293929712460065, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 3292 + }, + { + "epoch": 2.630191693290735, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 3293 + }, + { + "epoch": 2.6309904153354635, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 3294 + }, + { + "epoch": 2.6317891373801916, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 3295 + }, + { + "epoch": 2.63258785942492, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 3296 + }, + { + "epoch": 2.6333865814696487, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 3297 + }, + { + "epoch": 2.6341853035143767, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 3298 + }, + { + "epoch": 2.6349840255591053, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 3299 + }, + { + "epoch": 2.635782747603834, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0472, + "step": 3300 + }, + { + "epoch": 2.6365814696485623, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 3301 + }, + { + "epoch": 2.637380191693291, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 3302 + }, + { + "epoch": 2.6381789137380194, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 3303 + }, + { + "epoch": 2.6389776357827475, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 3304 + }, + { + "epoch": 2.639776357827476, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 3305 + }, + { + "epoch": 2.6405750798722045, + "grad_norm": 0.1708984375, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 3306 + }, + { + "epoch": 2.641373801916933, + "grad_norm": 0.349609375, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 3307 + }, + { + "epoch": 2.642172523961661, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 3308 + }, + { + "epoch": 2.6429712460063897, + "grad_norm": 0.2392578125, + "learning_rate": 0.0005, + "loss": 1.044, + "step": 3309 + }, + { + "epoch": 2.643769968051118, + "grad_norm": 0.2333984375, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 3310 + }, + { + "epoch": 2.6445686900958467, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 3311 + }, + { + "epoch": 2.6453674121405752, + "grad_norm": 0.359375, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 3312 + }, + { + "epoch": 2.6461661341853038, + "grad_norm": 0.216796875, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 3313 + }, + { + "epoch": 2.646964856230032, + "grad_norm": 0.2119140625, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 3314 + }, + { + "epoch": 2.6477635782747604, + "grad_norm": 0.3671875, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 3315 + }, + { + "epoch": 2.648562300319489, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 3316 + }, + { + "epoch": 2.649361022364217, + "grad_norm": 0.169921875, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 3317 + }, + { + "epoch": 2.6501597444089455, + "grad_norm": 0.173828125, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 3318 + }, + { + "epoch": 2.650958466453674, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 3319 + }, + { + "epoch": 2.6517571884984026, + "grad_norm": 0.2578125, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 3320 + }, + { + "epoch": 2.652555910543131, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0448, + "step": 3321 + }, + { + "epoch": 2.6533546325878596, + "grad_norm": 0.326171875, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 3322 + }, + { + "epoch": 2.6541533546325877, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 3323 + }, + { + "epoch": 2.6549520766773163, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 3324 + }, + { + "epoch": 2.655750798722045, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 3325 + }, + { + "epoch": 2.6565495207667733, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 3326 + }, + { + "epoch": 2.6573482428115014, + "grad_norm": 0.208984375, + "learning_rate": 0.0005, + "loss": 1.0626, + "step": 3327 + }, + { + "epoch": 2.65814696485623, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 3328 + }, + { + "epoch": 2.6589456869009584, + "grad_norm": 0.1884765625, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 3329 + }, + { + "epoch": 2.659744408945687, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 3330 + }, + { + "epoch": 2.6605431309904155, + "grad_norm": 0.19140625, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 3331 + }, + { + "epoch": 2.661341853035144, + "grad_norm": 0.1787109375, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 3332 + }, + { + "epoch": 2.662140575079872, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 3333 + }, + { + "epoch": 2.6629392971246006, + "grad_norm": 0.271484375, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 3334 + }, + { + "epoch": 2.663738019169329, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 3335 + }, + { + "epoch": 2.6645367412140573, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 3336 + }, + { + "epoch": 2.665335463258786, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 3337 + }, + { + "epoch": 2.6661341853035143, + "grad_norm": 0.197265625, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 3338 + }, + { + "epoch": 2.666932907348243, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 3339 + }, + { + "epoch": 2.6677316293929714, + "grad_norm": 0.1845703125, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 3340 + }, + { + "epoch": 2.6685303514377, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.0488, + "step": 3341 + }, + { + "epoch": 2.669329073482428, + "grad_norm": 0.267578125, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 3342 + }, + { + "epoch": 2.6701277955271565, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 3343 + }, + { + "epoch": 2.670926517571885, + "grad_norm": 0.34765625, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 3344 + }, + { + "epoch": 2.6717252396166136, + "grad_norm": 0.2470703125, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 3345 + }, + { + "epoch": 2.6725239616613417, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.0466, + "step": 3346 + }, + { + "epoch": 2.67332268370607, + "grad_norm": 0.2890625, + "learning_rate": 0.0005, + "loss": 1.0611, + "step": 3347 + }, + { + "epoch": 2.6741214057507987, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 3348 + }, + { + "epoch": 2.6749201277955272, + "grad_norm": 0.25390625, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 3349 + }, + { + "epoch": 2.6757188498402558, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 3350 + }, + { + "epoch": 2.6765175718849843, + "grad_norm": 0.2412109375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 3351 + }, + { + "epoch": 2.6773162939297124, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 3352 + }, + { + "epoch": 2.678115015974441, + "grad_norm": 0.17578125, + "learning_rate": 0.0005, + "loss": 1.042, + "step": 3353 + }, + { + "epoch": 2.6789137380191694, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 3354 + }, + { + "epoch": 2.6797124600638975, + "grad_norm": 0.1943359375, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 3355 + }, + { + "epoch": 2.680511182108626, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0618, + "step": 3356 + }, + { + "epoch": 2.6813099041533546, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 3357 + }, + { + "epoch": 2.682108626198083, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 3358 + }, + { + "epoch": 2.6829073482428116, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 3359 + }, + { + "epoch": 2.68370607028754, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 3360 + }, + { + "epoch": 2.6845047923322682, + "grad_norm": 0.1826171875, + "learning_rate": 0.0005, + "loss": 1.0689, + "step": 3361 + }, + { + "epoch": 2.6853035143769968, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 3362 + }, + { + "epoch": 2.6861022364217253, + "grad_norm": 0.26953125, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 3363 + }, + { + "epoch": 2.686900958466454, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 3364 + }, + { + "epoch": 2.687699680511182, + "grad_norm": 0.19921875, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 3365 + }, + { + "epoch": 2.6884984025559104, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 3366 + }, + { + "epoch": 2.689297124600639, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 3367 + }, + { + "epoch": 2.6900958466453675, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 3368 + }, + { + "epoch": 2.690894568690096, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 3369 + }, + { + "epoch": 2.6916932907348246, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0599, + "step": 3370 + }, + { + "epoch": 2.6924920127795526, + "grad_norm": 0.1640625, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 3371 + }, + { + "epoch": 2.693290734824281, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.066, + "step": 3372 + }, + { + "epoch": 2.6940894568690097, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 3373 + }, + { + "epoch": 2.6948881789137378, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 3374 + }, + { + "epoch": 2.6956869009584663, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 3375 + }, + { + "epoch": 2.696485623003195, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0455, + "step": 3376 + }, + { + "epoch": 2.6972843450479234, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 3377 + }, + { + "epoch": 2.698083067092652, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0476, + "step": 3378 + }, + { + "epoch": 2.6988817891373804, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 3379 + }, + { + "epoch": 2.6996805111821085, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 3380 + }, + { + "epoch": 2.700479233226837, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0665, + "step": 3381 + }, + { + "epoch": 2.7012779552715656, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 3382 + }, + { + "epoch": 2.702076677316294, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0408, + "step": 3383 + }, + { + "epoch": 2.702875399361022, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 3384 + }, + { + "epoch": 2.7036741214057507, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 3385 + }, + { + "epoch": 2.7044728434504792, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 3386 + }, + { + "epoch": 2.7052715654952078, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0608, + "step": 3387 + }, + { + "epoch": 2.7060702875399363, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0396, + "step": 3388 + }, + { + "epoch": 2.706869009584665, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 3389 + }, + { + "epoch": 2.707667731629393, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 3390 + }, + { + "epoch": 2.7084664536741214, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.046, + "step": 3391 + }, + { + "epoch": 2.70926517571885, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 3392 + }, + { + "epoch": 2.710063897763578, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 3393 + }, + { + "epoch": 2.7108626198083066, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 3394 + }, + { + "epoch": 2.711661341853035, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 3395 + }, + { + "epoch": 2.7124600638977636, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 3396 + }, + { + "epoch": 2.713258785942492, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0459, + "step": 3397 + }, + { + "epoch": 2.7140575079872207, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 3398 + }, + { + "epoch": 2.7148562300319488, + "grad_norm": 0.1767578125, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 3399 + }, + { + "epoch": 2.7156549520766773, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 3400 + }, + { + "epoch": 2.716453674121406, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 3401 + }, + { + "epoch": 2.7172523961661343, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 3402 + }, + { + "epoch": 2.7180511182108624, + "grad_norm": 0.2138671875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 3403 + }, + { + "epoch": 2.718849840255591, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 3404 + }, + { + "epoch": 2.7196485623003195, + "grad_norm": 0.298828125, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 3405 + }, + { + "epoch": 2.720447284345048, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 3406 + }, + { + "epoch": 2.7212460063897765, + "grad_norm": 0.1806640625, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 3407 + }, + { + "epoch": 2.722044728434505, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 3408 + }, + { + "epoch": 2.722843450479233, + "grad_norm": 0.236328125, + "learning_rate": 0.0005, + "loss": 1.0467, + "step": 3409 + }, + { + "epoch": 2.7236421725239617, + "grad_norm": 0.162109375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 3410 + }, + { + "epoch": 2.72444089456869, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 3411 + }, + { + "epoch": 2.7252396166134183, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 3412 + }, + { + "epoch": 2.726038338658147, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 3413 + }, + { + "epoch": 2.7268370607028753, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 3414 + }, + { + "epoch": 2.727635782747604, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 3415 + }, + { + "epoch": 2.7284345047923324, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 3416 + }, + { + "epoch": 2.729233226837061, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 3417 + }, + { + "epoch": 2.730031948881789, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 3418 + }, + { + "epoch": 2.7308306709265175, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 3419 + }, + { + "epoch": 2.731629392971246, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 3420 + }, + { + "epoch": 2.7324281150159746, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 3421 + }, + { + "epoch": 2.7332268370607027, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 3422 + }, + { + "epoch": 2.734025559105431, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 3423 + }, + { + "epoch": 2.7348242811501597, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 3424 + }, + { + "epoch": 2.7356230031948883, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 3425 + }, + { + "epoch": 2.736421725239617, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 3426 + }, + { + "epoch": 2.737220447284345, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 3427 + }, + { + "epoch": 2.7380191693290734, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 3428 + }, + { + "epoch": 2.738817891373802, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 3429 + }, + { + "epoch": 2.7396166134185305, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 3430 + }, + { + "epoch": 2.7404153354632586, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0453, + "step": 3431 + }, + { + "epoch": 2.741214057507987, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 3432 + }, + { + "epoch": 2.7420127795527156, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 3433 + }, + { + "epoch": 2.742811501597444, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 3434 + }, + { + "epoch": 2.7436102236421727, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 3435 + }, + { + "epoch": 2.744408945686901, + "grad_norm": 0.185546875, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 3436 + }, + { + "epoch": 2.7452076677316293, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.041, + "step": 3437 + }, + { + "epoch": 2.746006389776358, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 3438 + }, + { + "epoch": 2.7468051118210863, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.048, + "step": 3439 + }, + { + "epoch": 2.747603833865815, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 3440 + }, + { + "epoch": 2.748402555910543, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0592, + "step": 3441 + }, + { + "epoch": 2.7492012779552715, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 3442 + }, + { + "epoch": 2.75, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 3443 + }, + { + "epoch": 2.7507987220447285, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 3444 + }, + { + "epoch": 2.751597444089457, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 3445 + }, + { + "epoch": 2.752396166134185, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 3446 + }, + { + "epoch": 2.7531948881789137, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 3447 + }, + { + "epoch": 2.753993610223642, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0444, + "step": 3448 + }, + { + "epoch": 2.7547923322683707, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 3449 + }, + { + "epoch": 2.755591054313099, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0453, + "step": 3450 + }, + { + "epoch": 2.7563897763578273, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 3451 + }, + { + "epoch": 2.757188498402556, + "grad_norm": 0.197265625, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 3452 + }, + { + "epoch": 2.7579872204472844, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 3453 + }, + { + "epoch": 2.758785942492013, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 3454 + }, + { + "epoch": 2.7595846645367414, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 3455 + }, + { + "epoch": 2.7603833865814695, + "grad_norm": 0.162109375, + "learning_rate": 0.0005, + "loss": 1.0641, + "step": 3456 + }, + { + "epoch": 2.761182108626198, + "grad_norm": 0.193359375, + "learning_rate": 0.0005, + "loss": 1.0431, + "step": 3457 + }, + { + "epoch": 2.7619808306709266, + "grad_norm": 0.330078125, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 3458 + }, + { + "epoch": 2.762779552715655, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 3459 + }, + { + "epoch": 2.763578274760383, + "grad_norm": 0.275390625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 3460 + }, + { + "epoch": 2.7643769968051117, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 3461 + }, + { + "epoch": 2.7651757188498403, + "grad_norm": 0.1884765625, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 3462 + }, + { + "epoch": 2.765974440894569, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 3463 + }, + { + "epoch": 2.7667731629392973, + "grad_norm": 0.2119140625, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 3464 + }, + { + "epoch": 2.7675718849840254, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 3465 + }, + { + "epoch": 2.768370607028754, + "grad_norm": 0.40234375, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 3466 + }, + { + "epoch": 2.7691693290734825, + "grad_norm": 0.31640625, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 3467 + }, + { + "epoch": 2.769968051118211, + "grad_norm": 0.181640625, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 3468 + }, + { + "epoch": 2.770766773162939, + "grad_norm": 0.4765625, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 3469 + }, + { + "epoch": 2.7715654952076676, + "grad_norm": 0.23046875, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 3470 + }, + { + "epoch": 2.772364217252396, + "grad_norm": 0.1953125, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 3471 + }, + { + "epoch": 2.7731629392971247, + "grad_norm": 0.208984375, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 3472 + }, + { + "epoch": 2.773961661341853, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 3473 + }, + { + "epoch": 2.7747603833865817, + "grad_norm": 0.21484375, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 3474 + }, + { + "epoch": 2.77555910543131, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0464, + "step": 3475 + }, + { + "epoch": 2.7763578274760383, + "grad_norm": 0.296875, + "learning_rate": 0.0005, + "loss": 1.0441, + "step": 3476 + }, + { + "epoch": 2.777156549520767, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 3477 + }, + { + "epoch": 2.777955271565495, + "grad_norm": 0.234375, + "learning_rate": 0.0005, + "loss": 1.0454, + "step": 3478 + }, + { + "epoch": 2.7787539936102235, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 3479 + }, + { + "epoch": 2.779552715654952, + "grad_norm": 0.2021484375, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 3480 + }, + { + "epoch": 2.7803514376996805, + "grad_norm": 0.19140625, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 3481 + }, + { + "epoch": 2.781150159744409, + "grad_norm": 0.1787109375, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 3482 + }, + { + "epoch": 2.7819488817891376, + "grad_norm": 0.275390625, + "learning_rate": 0.0005, + "loss": 1.0623, + "step": 3483 + }, + { + "epoch": 2.7827476038338657, + "grad_norm": 0.1767578125, + "learning_rate": 0.0005, + "loss": 1.0437, + "step": 3484 + }, + { + "epoch": 2.783546325878594, + "grad_norm": 0.478515625, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 3485 + }, + { + "epoch": 2.7843450479233227, + "grad_norm": 0.23046875, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 3486 + }, + { + "epoch": 2.7851437699680512, + "grad_norm": 0.2294921875, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 3487 + }, + { + "epoch": 2.7859424920127793, + "grad_norm": 0.27734375, + "learning_rate": 0.0005, + "loss": 1.0411, + "step": 3488 + }, + { + "epoch": 2.786741214057508, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 3489 + }, + { + "epoch": 2.7875399361022364, + "grad_norm": 0.392578125, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 3490 + }, + { + "epoch": 2.788338658146965, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 3491 + }, + { + "epoch": 2.7891373801916934, + "grad_norm": 0.30078125, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 3492 + }, + { + "epoch": 2.789936102236422, + "grad_norm": 0.248046875, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 3493 + }, + { + "epoch": 2.79073482428115, + "grad_norm": 0.1611328125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 3494 + }, + { + "epoch": 2.7915335463258786, + "grad_norm": 0.201171875, + "learning_rate": 0.0005, + "loss": 1.0455, + "step": 3495 + }, + { + "epoch": 2.792332268370607, + "grad_norm": 0.19921875, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 3496 + }, + { + "epoch": 2.793130990415335, + "grad_norm": 0.31640625, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 3497 + }, + { + "epoch": 2.7939297124600637, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 3498 + }, + { + "epoch": 2.7947284345047922, + "grad_norm": 0.349609375, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 3499 + }, + { + "epoch": 2.7955271565495208, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 3500 + }, + { + "epoch": 2.7963258785942493, + "grad_norm": 0.408203125, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 3501 + }, + { + "epoch": 2.797124600638978, + "grad_norm": 0.1904296875, + "learning_rate": 0.0005, + "loss": 1.0436, + "step": 3502 + }, + { + "epoch": 2.797923322683706, + "grad_norm": 0.1826171875, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 3503 + }, + { + "epoch": 2.7987220447284344, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 3504 + }, + { + "epoch": 2.799520766773163, + "grad_norm": 0.1748046875, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 3505 + }, + { + "epoch": 2.8003194888178915, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 3506 + }, + { + "epoch": 2.8011182108626196, + "grad_norm": 0.2431640625, + "learning_rate": 0.0005, + "loss": 1.0405, + "step": 3507 + }, + { + "epoch": 2.801916932907348, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 3508 + }, + { + "epoch": 2.8027156549520766, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0414, + "step": 3509 + }, + { + "epoch": 2.803514376996805, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 3510 + }, + { + "epoch": 2.8043130990415337, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 3511 + }, + { + "epoch": 2.8051118210862622, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 3512 + }, + { + "epoch": 2.8059105431309903, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 3513 + }, + { + "epoch": 2.806709265175719, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 3514 + }, + { + "epoch": 2.8075079872204474, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0476, + "step": 3515 + }, + { + "epoch": 2.8083067092651754, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0393, + "step": 3516 + }, + { + "epoch": 2.809105431309904, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 3517 + }, + { + "epoch": 2.8099041533546325, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 3518 + }, + { + "epoch": 2.810702875399361, + "grad_norm": 0.1611328125, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 3519 + }, + { + "epoch": 2.8115015974440896, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.067, + "step": 3520 + }, + { + "epoch": 2.812300319488818, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 3521 + }, + { + "epoch": 2.813099041533546, + "grad_norm": 0.166015625, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 3522 + }, + { + "epoch": 2.8138977635782747, + "grad_norm": 0.2158203125, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 3523 + }, + { + "epoch": 2.8146964856230032, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 3524 + }, + { + "epoch": 2.8154952076677318, + "grad_norm": 0.2421875, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 3525 + }, + { + "epoch": 2.81629392971246, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 3526 + }, + { + "epoch": 2.8170926517571884, + "grad_norm": 0.306640625, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 3527 + }, + { + "epoch": 2.817891373801917, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 3528 + }, + { + "epoch": 2.8186900958466454, + "grad_norm": 0.1826171875, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 3529 + }, + { + "epoch": 2.819488817891374, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.039, + "step": 3530 + }, + { + "epoch": 2.8202875399361025, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0456, + "step": 3531 + }, + { + "epoch": 2.8210862619808306, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 3532 + }, + { + "epoch": 2.821884984025559, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 3533 + }, + { + "epoch": 2.8226837060702876, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 3534 + }, + { + "epoch": 2.8234824281150157, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 3535 + }, + { + "epoch": 2.8242811501597442, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 3536 + }, + { + "epoch": 2.8250798722044728, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 3537 + }, + { + "epoch": 2.8258785942492013, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 3538 + }, + { + "epoch": 2.82667731629393, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 3539 + }, + { + "epoch": 2.8274760383386583, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 3540 + }, + { + "epoch": 2.8282747603833864, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0427, + "step": 3541 + }, + { + "epoch": 2.829073482428115, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 3542 + }, + { + "epoch": 2.8298722044728435, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0654, + "step": 3543 + }, + { + "epoch": 2.830670926517572, + "grad_norm": 0.177734375, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 3544 + }, + { + "epoch": 2.8314696485623, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 3545 + }, + { + "epoch": 2.8322683706070286, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0451, + "step": 3546 + }, + { + "epoch": 2.833067092651757, + "grad_norm": 0.1640625, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 3547 + }, + { + "epoch": 2.8338658146964857, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 3548 + }, + { + "epoch": 2.834664536741214, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 3549 + }, + { + "epoch": 2.8354632587859427, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 3550 + }, + { + "epoch": 2.836261980830671, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 3551 + }, + { + "epoch": 2.8370607028753994, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0417, + "step": 3552 + }, + { + "epoch": 2.837859424920128, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 3553 + }, + { + "epoch": 2.838658146964856, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0446, + "step": 3554 + }, + { + "epoch": 2.8394568690095845, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0614, + "step": 3555 + }, + { + "epoch": 2.840255591054313, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 3556 + }, + { + "epoch": 2.8410543130990416, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 3557 + }, + { + "epoch": 2.84185303514377, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 3558 + }, + { + "epoch": 2.8426517571884986, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0619, + "step": 3559 + }, + { + "epoch": 2.8434504792332267, + "grad_norm": 0.1669921875, + "learning_rate": 0.0005, + "loss": 1.0424, + "step": 3560 + }, + { + "epoch": 2.844249201277955, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 3561 + }, + { + "epoch": 2.8450479233226837, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 3562 + }, + { + "epoch": 2.8458466453674123, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 3563 + }, + { + "epoch": 2.8466453674121404, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 3564 + }, + { + "epoch": 2.847444089456869, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 3565 + }, + { + "epoch": 2.8482428115015974, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.0407, + "step": 3566 + }, + { + "epoch": 2.849041533546326, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 3567 + }, + { + "epoch": 2.8498402555910545, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 3568 + }, + { + "epoch": 2.850638977635783, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 3569 + }, + { + "epoch": 2.851437699680511, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 3570 + }, + { + "epoch": 2.8522364217252396, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 3571 + }, + { + "epoch": 2.853035143769968, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0468, + "step": 3572 + }, + { + "epoch": 2.8538338658146962, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.0459, + "step": 3573 + }, + { + "epoch": 2.8546325878594248, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0457, + "step": 3574 + }, + { + "epoch": 2.8554313099041533, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 3575 + }, + { + "epoch": 2.856230031948882, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 3576 + }, + { + "epoch": 2.8570287539936103, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 3577 + }, + { + "epoch": 2.857827476038339, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 3578 + }, + { + "epoch": 2.858626198083067, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 3579 + }, + { + "epoch": 2.8594249201277955, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 3580 + }, + { + "epoch": 2.860223642172524, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 3581 + }, + { + "epoch": 2.8610223642172525, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 3582 + }, + { + "epoch": 2.8618210862619806, + "grad_norm": 0.177734375, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 3583 + }, + { + "epoch": 2.862619808306709, + "grad_norm": 0.3359375, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 3584 + }, + { + "epoch": 2.8634185303514377, + "grad_norm": 0.2021484375, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 3585 + }, + { + "epoch": 2.864217252396166, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0458, + "step": 3586 + }, + { + "epoch": 2.8650159744408947, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0676, + "step": 3587 + }, + { + "epoch": 2.8658146964856233, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 3588 + }, + { + "epoch": 2.8666134185303513, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 3589 + }, + { + "epoch": 2.86741214057508, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0602, + "step": 3590 + }, + { + "epoch": 2.8682108626198084, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 3591 + }, + { + "epoch": 2.8690095846645365, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 3592 + }, + { + "epoch": 2.869808306709265, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 3593 + }, + { + "epoch": 2.8706070287539935, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 3594 + }, + { + "epoch": 2.871405750798722, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0631, + "step": 3595 + }, + { + "epoch": 2.8722044728434506, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 3596 + }, + { + "epoch": 2.873003194888179, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 3597 + }, + { + "epoch": 2.873801916932907, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 3598 + }, + { + "epoch": 2.8746006389776357, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 3599 + }, + { + "epoch": 2.8753993610223643, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.046, + "step": 3600 + }, + { + "epoch": 2.876198083067093, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 3601 + }, + { + "epoch": 2.876996805111821, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 3602 + }, + { + "epoch": 2.8777955271565494, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 3603 + }, + { + "epoch": 2.878594249201278, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 3604 + }, + { + "epoch": 2.8793929712460065, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 3605 + }, + { + "epoch": 2.880191693290735, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 3606 + }, + { + "epoch": 2.8809904153354635, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 3607 + }, + { + "epoch": 2.8817891373801916, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 3608 + }, + { + "epoch": 2.88258785942492, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 3609 + }, + { + "epoch": 2.8833865814696487, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0442, + "step": 3610 + }, + { + "epoch": 2.8841853035143767, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 3611 + }, + { + "epoch": 2.8849840255591053, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 3612 + }, + { + "epoch": 2.885782747603834, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0402, + "step": 3613 + }, + { + "epoch": 2.8865814696485623, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 3614 + }, + { + "epoch": 2.887380191693291, + "grad_norm": 0.169921875, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 3615 + }, + { + "epoch": 2.8881789137380194, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 3616 + }, + { + "epoch": 2.8889776357827475, + "grad_norm": 0.21484375, + "learning_rate": 0.0005, + "loss": 1.0423, + "step": 3617 + }, + { + "epoch": 2.889776357827476, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 3618 + }, + { + "epoch": 2.8905750798722045, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 3619 + }, + { + "epoch": 2.891373801916933, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 3620 + }, + { + "epoch": 2.892172523961661, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 3621 + }, + { + "epoch": 2.8929712460063897, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 3622 + }, + { + "epoch": 2.893769968051118, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0451, + "step": 3623 + }, + { + "epoch": 2.8945686900958467, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 3624 + }, + { + "epoch": 2.8953674121405752, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 3625 + }, + { + "epoch": 2.8961661341853038, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 3626 + }, + { + "epoch": 2.896964856230032, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0454, + "step": 3627 + }, + { + "epoch": 2.8977635782747604, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 3628 + }, + { + "epoch": 2.898562300319489, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0447, + "step": 3629 + }, + { + "epoch": 2.899361022364217, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 3630 + }, + { + "epoch": 2.9001597444089455, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 3631 + }, + { + "epoch": 2.900958466453674, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 3632 + }, + { + "epoch": 2.9017571884984026, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 3633 + }, + { + "epoch": 2.902555910543131, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 3634 + }, + { + "epoch": 2.9033546325878596, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 3635 + }, + { + "epoch": 2.9041533546325877, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 3636 + }, + { + "epoch": 2.9049520766773163, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0468, + "step": 3637 + }, + { + "epoch": 2.905750798722045, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 3638 + }, + { + "epoch": 2.9065495207667733, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0446, + "step": 3639 + }, + { + "epoch": 2.9073482428115014, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0422, + "step": 3640 + }, + { + "epoch": 2.90814696485623, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0446, + "step": 3641 + }, + { + "epoch": 2.9089456869009584, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 3642 + }, + { + "epoch": 2.909744408945687, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 3643 + }, + { + "epoch": 2.9105431309904155, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 3644 + }, + { + "epoch": 2.911341853035144, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 3645 + }, + { + "epoch": 2.912140575079872, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 3646 + }, + { + "epoch": 2.9129392971246006, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 3647 + }, + { + "epoch": 2.913738019169329, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 3648 + }, + { + "epoch": 2.9145367412140573, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0446, + "step": 3649 + }, + { + "epoch": 2.915335463258786, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 3650 + }, + { + "epoch": 2.9161341853035143, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 3651 + }, + { + "epoch": 2.916932907348243, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 3652 + }, + { + "epoch": 2.9177316293929714, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 3653 + }, + { + "epoch": 2.9185303514377, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 3654 + }, + { + "epoch": 2.919329073482428, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 3655 + }, + { + "epoch": 2.9201277955271565, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0637, + "step": 3656 + }, + { + "epoch": 2.920926517571885, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 3657 + }, + { + "epoch": 2.9217252396166136, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 3658 + }, + { + "epoch": 2.9225239616613417, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 3659 + }, + { + "epoch": 2.92332268370607, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0451, + "step": 3660 + }, + { + "epoch": 2.9241214057507987, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 3661 + }, + { + "epoch": 2.9249201277955272, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.045, + "step": 3662 + }, + { + "epoch": 2.9257188498402558, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 3663 + }, + { + "epoch": 2.9265175718849843, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.06, + "step": 3664 + }, + { + "epoch": 2.9273162939297124, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 3665 + }, + { + "epoch": 2.928115015974441, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 3666 + }, + { + "epoch": 2.9289137380191694, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0464, + "step": 3667 + }, + { + "epoch": 2.9297124600638975, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 3668 + }, + { + "epoch": 2.930511182108626, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 3669 + }, + { + "epoch": 2.9313099041533546, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 3670 + }, + { + "epoch": 2.932108626198083, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0467, + "step": 3671 + }, + { + "epoch": 2.9329073482428116, + "grad_norm": 0.1533203125, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 3672 + }, + { + "epoch": 2.93370607028754, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 3673 + }, + { + "epoch": 2.9345047923322682, + "grad_norm": 0.1767578125, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 3674 + }, + { + "epoch": 2.9353035143769968, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 3675 + }, + { + "epoch": 2.9361022364217253, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 3676 + }, + { + "epoch": 2.936900958466454, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 3677 + }, + { + "epoch": 2.937699680511182, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0466, + "step": 3678 + }, + { + "epoch": 2.9384984025559104, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 3679 + }, + { + "epoch": 2.939297124600639, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 3680 + }, + { + "epoch": 2.9400958466453675, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 3681 + }, + { + "epoch": 2.940894568690096, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 3682 + }, + { + "epoch": 2.9416932907348246, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 3683 + }, + { + "epoch": 2.9424920127795526, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 3684 + }, + { + "epoch": 2.943290734824281, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 3685 + }, + { + "epoch": 2.9440894568690097, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 3686 + }, + { + "epoch": 2.9448881789137378, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 3687 + }, + { + "epoch": 2.9456869009584663, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 3688 + }, + { + "epoch": 2.946485623003195, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 3689 + }, + { + "epoch": 2.9472843450479234, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 3690 + }, + { + "epoch": 2.948083067092652, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 3691 + }, + { + "epoch": 2.9488817891373804, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0468, + "step": 3692 + }, + { + "epoch": 2.9496805111821085, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0466, + "step": 3693 + }, + { + "epoch": 2.950479233226837, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 3694 + }, + { + "epoch": 2.9512779552715656, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 3695 + }, + { + "epoch": 2.952076677316294, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 3696 + }, + { + "epoch": 2.952875399361022, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 3697 + }, + { + "epoch": 2.9536741214057507, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0464, + "step": 3698 + }, + { + "epoch": 2.9544728434504792, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 3699 + }, + { + "epoch": 2.9552715654952078, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 3700 + }, + { + "epoch": 2.9560702875399363, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 3701 + }, + { + "epoch": 2.956869009584665, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 3702 + }, + { + "epoch": 2.957667731629393, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 3703 + }, + { + "epoch": 2.9584664536741214, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 3704 + }, + { + "epoch": 2.95926517571885, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 3705 + }, + { + "epoch": 2.960063897763578, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0488, + "step": 3706 + }, + { + "epoch": 2.9608626198083066, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0634, + "step": 3707 + }, + { + "epoch": 2.961661341853035, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0468, + "step": 3708 + }, + { + "epoch": 2.9624600638977636, + "grad_norm": 0.1572265625, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 3709 + }, + { + "epoch": 2.963258785942492, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 3710 + }, + { + "epoch": 2.9640575079872207, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 3711 + }, + { + "epoch": 2.9648562300319488, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0482, + "step": 3712 + }, + { + "epoch": 2.9656549520766773, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 3713 + }, + { + "epoch": 2.966453674121406, + "grad_norm": 0.185546875, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 3714 + }, + { + "epoch": 2.9672523961661343, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 3715 + }, + { + "epoch": 2.9680511182108624, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 3716 + }, + { + "epoch": 2.968849840255591, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 3717 + }, + { + "epoch": 2.9696485623003195, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 3718 + }, + { + "epoch": 2.970447284345048, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 3719 + }, + { + "epoch": 2.9712460063897765, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0472, + "step": 3720 + }, + { + "epoch": 2.972044728434505, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 3721 + }, + { + "epoch": 2.972843450479233, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 3722 + }, + { + "epoch": 2.9736421725239617, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 3723 + }, + { + "epoch": 2.97444089456869, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 3724 + }, + { + "epoch": 2.9752396166134183, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 3725 + }, + { + "epoch": 2.976038338658147, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 3726 + }, + { + "epoch": 2.9768370607028753, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 3727 + }, + { + "epoch": 2.977635782747604, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 3728 + }, + { + "epoch": 2.9784345047923324, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 3729 + }, + { + "epoch": 2.979233226837061, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 3730 + }, + { + "epoch": 2.980031948881789, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 3731 + }, + { + "epoch": 2.9808306709265175, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 3732 + }, + { + "epoch": 2.981629392971246, + "grad_norm": 0.1572265625, + "learning_rate": 0.0005, + "loss": 1.0465, + "step": 3733 + }, + { + "epoch": 2.9824281150159746, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 3734 + }, + { + "epoch": 2.9832268370607027, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 3735 + }, + { + "epoch": 2.984025559105431, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 3736 + }, + { + "epoch": 2.9848242811501597, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 3737 + }, + { + "epoch": 2.9856230031948883, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 3738 + }, + { + "epoch": 2.986421725239617, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0433, + "step": 3739 + }, + { + "epoch": 2.987220447284345, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 3740 + }, + { + "epoch": 2.9880191693290734, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 3741 + }, + { + "epoch": 2.988817891373802, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 3742 + }, + { + "epoch": 2.9896166134185305, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 3743 + }, + { + "epoch": 2.9904153354632586, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0464, + "step": 3744 + }, + { + "epoch": 2.991214057507987, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 3745 + }, + { + "epoch": 2.9920127795527156, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 3746 + }, + { + "epoch": 2.992811501597444, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 3747 + }, + { + "epoch": 2.9936102236421727, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 3748 + }, + { + "epoch": 2.994408945686901, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 3749 + }, + { + "epoch": 2.9952076677316293, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 3750 + }, + { + "epoch": 2.996006389776358, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 3751 + }, + { + "epoch": 2.9968051118210863, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 3752 + }, + { + "epoch": 2.997603833865815, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 3753 + }, + { + "epoch": 2.998402555910543, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 3754 + }, + { + "epoch": 2.9992012779552715, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0573, + "step": 3755 + }, + { + "epoch": 3.0, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 3756 + }, + { + "epoch": 3.0007987220447285, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0446, + "step": 3757 + }, + { + "epoch": 3.001597444089457, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 3758 + }, + { + "epoch": 3.002396166134185, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0452, + "step": 3759 + }, + { + "epoch": 3.0031948881789137, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 3760 + }, + { + "epoch": 3.003993610223642, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 3761 + }, + { + "epoch": 3.0047923322683707, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 3762 + }, + { + "epoch": 3.0055910543130993, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 3763 + }, + { + "epoch": 3.0063897763578273, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 3764 + }, + { + "epoch": 3.007188498402556, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 3765 + }, + { + "epoch": 3.0079872204472844, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 3766 + }, + { + "epoch": 3.008785942492013, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 3767 + }, + { + "epoch": 3.009584664536741, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 3768 + }, + { + "epoch": 3.0103833865814695, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0456, + "step": 3769 + }, + { + "epoch": 3.011182108626198, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 3770 + }, + { + "epoch": 3.0119808306709266, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0456, + "step": 3771 + }, + { + "epoch": 3.012779552715655, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 3772 + }, + { + "epoch": 3.013578274760383, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 3773 + }, + { + "epoch": 3.0143769968051117, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0433, + "step": 3774 + }, + { + "epoch": 3.0151757188498403, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 3775 + }, + { + "epoch": 3.015974440894569, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 3776 + }, + { + "epoch": 3.0167731629392973, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.048, + "step": 3777 + }, + { + "epoch": 3.0175718849840254, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.0431, + "step": 3778 + }, + { + "epoch": 3.018370607028754, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 3779 + }, + { + "epoch": 3.0191693290734825, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0436, + "step": 3780 + }, + { + "epoch": 3.019968051118211, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 3781 + }, + { + "epoch": 3.0207667731629395, + "grad_norm": 0.189453125, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 3782 + }, + { + "epoch": 3.0215654952076676, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0425, + "step": 3783 + }, + { + "epoch": 3.022364217252396, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 3784 + }, + { + "epoch": 3.0231629392971247, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 3785 + }, + { + "epoch": 3.023961661341853, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 3786 + }, + { + "epoch": 3.0247603833865813, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 3787 + }, + { + "epoch": 3.02555910543131, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 3788 + }, + { + "epoch": 3.0263578274760383, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 3789 + }, + { + "epoch": 3.027156549520767, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0457, + "step": 3790 + }, + { + "epoch": 3.0279552715654954, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 3791 + }, + { + "epoch": 3.0287539936102235, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 3792 + }, + { + "epoch": 3.029552715654952, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 3793 + }, + { + "epoch": 3.0303514376996805, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0434, + "step": 3794 + }, + { + "epoch": 3.031150159744409, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0434, + "step": 3795 + }, + { + "epoch": 3.0319488817891376, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 3796 + }, + { + "epoch": 3.0327476038338657, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 3797 + }, + { + "epoch": 3.033546325878594, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0426, + "step": 3798 + }, + { + "epoch": 3.0343450479233227, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 3799 + }, + { + "epoch": 3.0351437699680512, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.058, + "step": 3800 + }, + { + "epoch": 3.0359424920127798, + "grad_norm": 0.1923828125, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 3801 + }, + { + "epoch": 3.036741214057508, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0467, + "step": 3802 + }, + { + "epoch": 3.0375399361022364, + "grad_norm": 0.1796875, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 3803 + }, + { + "epoch": 3.038338658146965, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 3804 + }, + { + "epoch": 3.0391373801916934, + "grad_norm": 0.2119140625, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 3805 + }, + { + "epoch": 3.0399361022364215, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 3806 + }, + { + "epoch": 3.04073482428115, + "grad_norm": 0.185546875, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 3807 + }, + { + "epoch": 3.0415335463258786, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 3808 + }, + { + "epoch": 3.042332268370607, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 3809 + }, + { + "epoch": 3.0431309904153356, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 3810 + }, + { + "epoch": 3.0439297124600637, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 3811 + }, + { + "epoch": 3.0447284345047922, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 3812 + }, + { + "epoch": 3.0455271565495208, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 3813 + }, + { + "epoch": 3.0463258785942493, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 3814 + }, + { + "epoch": 3.047124600638978, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0424, + "step": 3815 + }, + { + "epoch": 3.047923322683706, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 3816 + }, + { + "epoch": 3.0487220447284344, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0409, + "step": 3817 + }, + { + "epoch": 3.049520766773163, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0468, + "step": 3818 + }, + { + "epoch": 3.0503194888178915, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 3819 + }, + { + "epoch": 3.0511182108626196, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0399, + "step": 3820 + }, + { + "epoch": 3.051916932907348, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.0482, + "step": 3821 + }, + { + "epoch": 3.0527156549520766, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0607, + "step": 3822 + }, + { + "epoch": 3.053514376996805, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 3823 + }, + { + "epoch": 3.0543130990415337, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 3824 + }, + { + "epoch": 3.055111821086262, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0457, + "step": 3825 + }, + { + "epoch": 3.0559105431309903, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 3826 + }, + { + "epoch": 3.056709265175719, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0622, + "step": 3827 + }, + { + "epoch": 3.0575079872204474, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 3828 + }, + { + "epoch": 3.058306709265176, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 3829 + }, + { + "epoch": 3.059105431309904, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0425, + "step": 3830 + }, + { + "epoch": 3.0599041533546325, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0451, + "step": 3831 + }, + { + "epoch": 3.060702875399361, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 3832 + }, + { + "epoch": 3.0615015974440896, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 3833 + }, + { + "epoch": 3.062300319488818, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 3834 + }, + { + "epoch": 3.063099041533546, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 3835 + }, + { + "epoch": 3.0638977635782747, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 3836 + }, + { + "epoch": 3.0646964856230032, + "grad_norm": 0.1630859375, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 3837 + }, + { + "epoch": 3.0654952076677318, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 3838 + }, + { + "epoch": 3.06629392971246, + "grad_norm": 0.193359375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 3839 + }, + { + "epoch": 3.0670926517571884, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 3840 + }, + { + "epoch": 3.067891373801917, + "grad_norm": 0.2197265625, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 3841 + }, + { + "epoch": 3.0686900958466454, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 3842 + }, + { + "epoch": 3.069488817891374, + "grad_norm": 0.203125, + "learning_rate": 0.0005, + "loss": 1.0375, + "step": 3843 + }, + { + "epoch": 3.070287539936102, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 3844 + }, + { + "epoch": 3.0710862619808306, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 3845 + }, + { + "epoch": 3.071884984025559, + "grad_norm": 0.181640625, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 3846 + }, + { + "epoch": 3.0726837060702876, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 3847 + }, + { + "epoch": 3.073482428115016, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 3848 + }, + { + "epoch": 3.0742811501597442, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 3849 + }, + { + "epoch": 3.0750798722044728, + "grad_norm": 0.1611328125, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 3850 + }, + { + "epoch": 3.0758785942492013, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 3851 + }, + { + "epoch": 3.07667731629393, + "grad_norm": 0.1923828125, + "learning_rate": 0.0005, + "loss": 1.0446, + "step": 3852 + }, + { + "epoch": 3.0774760383386583, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.046, + "step": 3853 + }, + { + "epoch": 3.0782747603833864, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0472, + "step": 3854 + }, + { + "epoch": 3.079073482428115, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 3855 + }, + { + "epoch": 3.0798722044728435, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 3856 + }, + { + "epoch": 3.080670926517572, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 3857 + }, + { + "epoch": 3.0814696485623, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 3858 + }, + { + "epoch": 3.0822683706070286, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 3859 + }, + { + "epoch": 3.083067092651757, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 3860 + }, + { + "epoch": 3.0838658146964857, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 3861 + }, + { + "epoch": 3.084664536741214, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 3862 + }, + { + "epoch": 3.0854632587859423, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 3863 + }, + { + "epoch": 3.086261980830671, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 3864 + }, + { + "epoch": 3.0870607028753994, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 3865 + }, + { + "epoch": 3.087859424920128, + "grad_norm": 0.166015625, + "learning_rate": 0.0005, + "loss": 1.0464, + "step": 3866 + }, + { + "epoch": 3.0886581469648564, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 3867 + }, + { + "epoch": 3.0894568690095845, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 3868 + }, + { + "epoch": 3.090255591054313, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 3869 + }, + { + "epoch": 3.0910543130990416, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 3870 + }, + { + "epoch": 3.09185303514377, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0396, + "step": 3871 + }, + { + "epoch": 3.0926517571884986, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 3872 + }, + { + "epoch": 3.0934504792332267, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 3873 + }, + { + "epoch": 3.094249201277955, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 3874 + }, + { + "epoch": 3.0950479233226837, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 3875 + }, + { + "epoch": 3.0958466453674123, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 3876 + }, + { + "epoch": 3.0966453674121404, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 3877 + }, + { + "epoch": 3.097444089456869, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 3878 + }, + { + "epoch": 3.0982428115015974, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 3879 + }, + { + "epoch": 3.099041533546326, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 3880 + }, + { + "epoch": 3.0998402555910545, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 3881 + }, + { + "epoch": 3.1006389776357826, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.045, + "step": 3882 + }, + { + "epoch": 3.101437699680511, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 3883 + }, + { + "epoch": 3.1022364217252396, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.048, + "step": 3884 + }, + { + "epoch": 3.103035143769968, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0443, + "step": 3885 + }, + { + "epoch": 3.1038338658146967, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 3886 + }, + { + "epoch": 3.1046325878594248, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 3887 + }, + { + "epoch": 3.1054313099041533, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.0454, + "step": 3888 + }, + { + "epoch": 3.106230031948882, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 3889 + }, + { + "epoch": 3.1070287539936103, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 3890 + }, + { + "epoch": 3.107827476038339, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 3891 + }, + { + "epoch": 3.108626198083067, + "grad_norm": 0.171875, + "learning_rate": 0.0005, + "loss": 1.0466, + "step": 3892 + }, + { + "epoch": 3.1094249201277955, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0645, + "step": 3893 + }, + { + "epoch": 3.110223642172524, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 3894 + }, + { + "epoch": 3.1110223642172525, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 3895 + }, + { + "epoch": 3.1118210862619806, + "grad_norm": 0.2333984375, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 3896 + }, + { + "epoch": 3.112619808306709, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 3897 + }, + { + "epoch": 3.1134185303514377, + "grad_norm": 0.267578125, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 3898 + }, + { + "epoch": 3.114217252396166, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 3899 + }, + { + "epoch": 3.1150159744408947, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 3900 + }, + { + "epoch": 3.115814696485623, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 3901 + }, + { + "epoch": 3.1166134185303513, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 3902 + }, + { + "epoch": 3.11741214057508, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.0454, + "step": 3903 + }, + { + "epoch": 3.1182108626198084, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 3904 + }, + { + "epoch": 3.119009584664537, + "grad_norm": 0.197265625, + "learning_rate": 0.0005, + "loss": 1.044, + "step": 3905 + }, + { + "epoch": 3.119808306709265, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 3906 + }, + { + "epoch": 3.1206070287539935, + "grad_norm": 0.2236328125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 3907 + }, + { + "epoch": 3.121405750798722, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0472, + "step": 3908 + }, + { + "epoch": 3.1222044728434506, + "grad_norm": 0.33203125, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 3909 + }, + { + "epoch": 3.123003194888179, + "grad_norm": 0.25, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 3910 + }, + { + "epoch": 3.123801916932907, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 3911 + }, + { + "epoch": 3.1246006389776357, + "grad_norm": 0.259765625, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 3912 + }, + { + "epoch": 3.1253993610223643, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 3913 + }, + { + "epoch": 3.126198083067093, + "grad_norm": 0.26953125, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 3914 + }, + { + "epoch": 3.126996805111821, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0467, + "step": 3915 + }, + { + "epoch": 3.1277955271565494, + "grad_norm": 0.1884765625, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 3916 + }, + { + "epoch": 3.128594249201278, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0447, + "step": 3917 + }, + { + "epoch": 3.1293929712460065, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0441, + "step": 3918 + }, + { + "epoch": 3.130191693290735, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 3919 + }, + { + "epoch": 3.130990415335463, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 3920 + }, + { + "epoch": 3.1317891373801916, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0465, + "step": 3921 + }, + { + "epoch": 3.13258785942492, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 3922 + }, + { + "epoch": 3.1333865814696487, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 3923 + }, + { + "epoch": 3.134185303514377, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 3924 + }, + { + "epoch": 3.1349840255591053, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 3925 + }, + { + "epoch": 3.135782747603834, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 3926 + }, + { + "epoch": 3.1365814696485623, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 3927 + }, + { + "epoch": 3.137380191693291, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 3928 + }, + { + "epoch": 3.1381789137380194, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 3929 + }, + { + "epoch": 3.1389776357827475, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 3930 + }, + { + "epoch": 3.139776357827476, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0476, + "step": 3931 + }, + { + "epoch": 3.1405750798722045, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 3932 + }, + { + "epoch": 3.141373801916933, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.045, + "step": 3933 + }, + { + "epoch": 3.142172523961661, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 3934 + }, + { + "epoch": 3.1429712460063897, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 3935 + }, + { + "epoch": 3.143769968051118, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 3936 + }, + { + "epoch": 3.1445686900958467, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 3937 + }, + { + "epoch": 3.1453674121405752, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.046, + "step": 3938 + }, + { + "epoch": 3.1461661341853033, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 3939 + }, + { + "epoch": 3.146964856230032, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 3940 + }, + { + "epoch": 3.1477635782747604, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 3941 + }, + { + "epoch": 3.148562300319489, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 3942 + }, + { + "epoch": 3.1493610223642174, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 3943 + }, + { + "epoch": 3.1501597444089455, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 3944 + }, + { + "epoch": 3.150958466453674, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 3945 + }, + { + "epoch": 3.1517571884984026, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 3946 + }, + { + "epoch": 3.152555910543131, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0459, + "step": 3947 + }, + { + "epoch": 3.1533546325878596, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0468, + "step": 3948 + }, + { + "epoch": 3.1541533546325877, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 3949 + }, + { + "epoch": 3.1549520766773163, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.0413, + "step": 3950 + }, + { + "epoch": 3.155750798722045, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 3951 + }, + { + "epoch": 3.1565495207667733, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 3952 + }, + { + "epoch": 3.1573482428115014, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 3953 + }, + { + "epoch": 3.15814696485623, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 3954 + }, + { + "epoch": 3.1589456869009584, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 3955 + }, + { + "epoch": 3.159744408945687, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0598, + "step": 3956 + }, + { + "epoch": 3.1605431309904155, + "grad_norm": 0.1904296875, + "learning_rate": 0.0005, + "loss": 1.0415, + "step": 3957 + }, + { + "epoch": 3.1613418530351436, + "grad_norm": 0.2109375, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 3958 + }, + { + "epoch": 3.162140575079872, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0458, + "step": 3959 + }, + { + "epoch": 3.1629392971246006, + "grad_norm": 0.20703125, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 3960 + }, + { + "epoch": 3.163738019169329, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 3961 + }, + { + "epoch": 3.1645367412140577, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 3962 + }, + { + "epoch": 3.165335463258786, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 3963 + }, + { + "epoch": 3.1661341853035143, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0399, + "step": 3964 + }, + { + "epoch": 3.166932907348243, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0456, + "step": 3965 + }, + { + "epoch": 3.1677316293929714, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0422, + "step": 3966 + }, + { + "epoch": 3.1685303514377, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0437, + "step": 3967 + }, + { + "epoch": 3.169329073482428, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0454, + "step": 3968 + }, + { + "epoch": 3.1701277955271565, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.061, + "step": 3969 + }, + { + "epoch": 3.170926517571885, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 3970 + }, + { + "epoch": 3.1717252396166136, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 3971 + }, + { + "epoch": 3.1725239616613417, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0472, + "step": 3972 + }, + { + "epoch": 3.17332268370607, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 3973 + }, + { + "epoch": 3.1741214057507987, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 3974 + }, + { + "epoch": 3.1749201277955272, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 3975 + }, + { + "epoch": 3.1757188498402558, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 3976 + }, + { + "epoch": 3.176517571884984, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0596, + "step": 3977 + }, + { + "epoch": 3.1773162939297124, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 3978 + }, + { + "epoch": 3.178115015974441, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0404, + "step": 3979 + }, + { + "epoch": 3.1789137380191694, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 3980 + }, + { + "epoch": 3.179712460063898, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 3981 + }, + { + "epoch": 3.180511182108626, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 3982 + }, + { + "epoch": 3.1813099041533546, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.048, + "step": 3983 + }, + { + "epoch": 3.182108626198083, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 3984 + }, + { + "epoch": 3.1829073482428116, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 3985 + }, + { + "epoch": 3.18370607028754, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 3986 + }, + { + "epoch": 3.1845047923322682, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 3987 + }, + { + "epoch": 3.1853035143769968, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 3988 + }, + { + "epoch": 3.1861022364217253, + "grad_norm": 0.267578125, + "learning_rate": 0.0005, + "loss": 1.0442, + "step": 3989 + }, + { + "epoch": 3.186900958466454, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 3990 + }, + { + "epoch": 3.187699680511182, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 3991 + }, + { + "epoch": 3.1884984025559104, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 3992 + }, + { + "epoch": 3.189297124600639, + "grad_norm": 0.1669921875, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 3993 + }, + { + "epoch": 3.1900958466453675, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0455, + "step": 3994 + }, + { + "epoch": 3.190894568690096, + "grad_norm": 0.248046875, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 3995 + }, + { + "epoch": 3.191693290734824, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 3996 + }, + { + "epoch": 3.1924920127795526, + "grad_norm": 0.220703125, + "learning_rate": 0.0005, + "loss": 1.0452, + "step": 3997 + }, + { + "epoch": 3.193290734824281, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0429, + "step": 3998 + }, + { + "epoch": 3.1940894568690097, + "grad_norm": 0.203125, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 3999 + }, + { + "epoch": 3.194888178913738, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 4000 + }, + { + "epoch": 3.1956869009584663, + "grad_norm": 0.1796875, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 4001 + }, + { + "epoch": 3.196485623003195, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 4002 + }, + { + "epoch": 3.1972843450479234, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0433, + "step": 4003 + }, + { + "epoch": 3.198083067092652, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0578, + "step": 4004 + }, + { + "epoch": 3.1988817891373804, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0414, + "step": 4005 + }, + { + "epoch": 3.1996805111821085, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 4006 + }, + { + "epoch": 3.200479233226837, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 4007 + }, + { + "epoch": 3.2012779552715656, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 4008 + }, + { + "epoch": 3.202076677316294, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 4009 + }, + { + "epoch": 3.202875399361022, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 4010 + }, + { + "epoch": 3.2036741214057507, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 4011 + }, + { + "epoch": 3.2044728434504792, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.045, + "step": 4012 + }, + { + "epoch": 3.2052715654952078, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 4013 + }, + { + "epoch": 3.2060702875399363, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 4014 + }, + { + "epoch": 3.2068690095846644, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 4015 + }, + { + "epoch": 3.207667731629393, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0422, + "step": 4016 + }, + { + "epoch": 3.2084664536741214, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 4017 + }, + { + "epoch": 3.20926517571885, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 4018 + }, + { + "epoch": 3.2100638977635785, + "grad_norm": 0.19921875, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 4019 + }, + { + "epoch": 3.2108626198083066, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 4020 + }, + { + "epoch": 3.211661341853035, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 4021 + }, + { + "epoch": 3.2124600638977636, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 4022 + }, + { + "epoch": 3.213258785942492, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 4023 + }, + { + "epoch": 3.2140575079872207, + "grad_norm": 0.1533203125, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 4024 + }, + { + "epoch": 3.2148562300319488, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0425, + "step": 4025 + }, + { + "epoch": 3.2156549520766773, + "grad_norm": 0.1923828125, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 4026 + }, + { + "epoch": 3.216453674121406, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 4027 + }, + { + "epoch": 3.2172523961661343, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0445, + "step": 4028 + }, + { + "epoch": 3.2180511182108624, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.04, + "step": 4029 + }, + { + "epoch": 3.218849840255591, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 4030 + }, + { + "epoch": 3.2196485623003195, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 4031 + }, + { + "epoch": 3.220447284345048, + "grad_norm": 0.1640625, + "learning_rate": 0.0005, + "loss": 1.0625, + "step": 4032 + }, + { + "epoch": 3.2212460063897765, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0603, + "step": 4033 + }, + { + "epoch": 3.2220447284345046, + "grad_norm": 0.275390625, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 4034 + }, + { + "epoch": 3.222843450479233, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0397, + "step": 4035 + }, + { + "epoch": 3.2236421725239617, + "grad_norm": 0.234375, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 4036 + }, + { + "epoch": 3.22444089456869, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0455, + "step": 4037 + }, + { + "epoch": 3.2252396166134187, + "grad_norm": 0.1630859375, + "learning_rate": 0.0005, + "loss": 1.0585, + "step": 4038 + }, + { + "epoch": 3.226038338658147, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 4039 + }, + { + "epoch": 3.2268370607028753, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 4040 + }, + { + "epoch": 3.227635782747604, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 4041 + }, + { + "epoch": 3.2284345047923324, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 4042 + }, + { + "epoch": 3.229233226837061, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 4043 + }, + { + "epoch": 3.230031948881789, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0445, + "step": 4044 + }, + { + "epoch": 3.2308306709265175, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.0379, + "step": 4045 + }, + { + "epoch": 3.231629392971246, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.043, + "step": 4046 + }, + { + "epoch": 3.2324281150159746, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0452, + "step": 4047 + }, + { + "epoch": 3.2332268370607027, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 4048 + }, + { + "epoch": 3.234025559105431, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0584, + "step": 4049 + }, + { + "epoch": 3.2348242811501597, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0435, + "step": 4050 + }, + { + "epoch": 3.2356230031948883, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 4051 + }, + { + "epoch": 3.236421725239617, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 4052 + }, + { + "epoch": 3.237220447284345, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 4053 + }, + { + "epoch": 3.2380191693290734, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 4054 + }, + { + "epoch": 3.238817891373802, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0558, + "step": 4055 + }, + { + "epoch": 3.2396166134185305, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0402, + "step": 4056 + }, + { + "epoch": 3.2404153354632586, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 4057 + }, + { + "epoch": 3.241214057507987, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 4058 + }, + { + "epoch": 3.2420127795527156, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0447, + "step": 4059 + }, + { + "epoch": 3.242811501597444, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 4060 + }, + { + "epoch": 3.2436102236421727, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 4061 + }, + { + "epoch": 3.244408945686901, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0423, + "step": 4062 + }, + { + "epoch": 3.2452076677316293, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 4063 + }, + { + "epoch": 3.246006389776358, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.044, + "step": 4064 + }, + { + "epoch": 3.2468051118210863, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 4065 + }, + { + "epoch": 3.247603833865815, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 4066 + }, + { + "epoch": 3.248402555910543, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 4067 + }, + { + "epoch": 3.2492012779552715, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0472, + "step": 4068 + }, + { + "epoch": 3.25, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.044, + "step": 4069 + }, + { + "epoch": 3.2507987220447285, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 4070 + }, + { + "epoch": 3.251597444089457, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0402, + "step": 4071 + }, + { + "epoch": 3.252396166134185, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 4072 + }, + { + "epoch": 3.2531948881789137, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 4073 + }, + { + "epoch": 3.253993610223642, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0468, + "step": 4074 + }, + { + "epoch": 3.2547923322683707, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 4075 + }, + { + "epoch": 3.255591054313099, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 4076 + }, + { + "epoch": 3.2563897763578273, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 4077 + }, + { + "epoch": 3.257188498402556, + "grad_norm": 0.2265625, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 4078 + }, + { + "epoch": 3.2579872204472844, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 4079 + }, + { + "epoch": 3.258785942492013, + "grad_norm": 0.259765625, + "learning_rate": 0.0005, + "loss": 1.0476, + "step": 4080 + }, + { + "epoch": 3.2595846645367414, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 4081 + }, + { + "epoch": 3.2603833865814695, + "grad_norm": 0.18359375, + "learning_rate": 0.0005, + "loss": 1.0446, + "step": 4082 + }, + { + "epoch": 3.261182108626198, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0426, + "step": 4083 + }, + { + "epoch": 3.2619808306709266, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 4084 + }, + { + "epoch": 3.262779552715655, + "grad_norm": 0.2021484375, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 4085 + }, + { + "epoch": 3.263578274760383, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 4086 + }, + { + "epoch": 3.2643769968051117, + "grad_norm": 0.21875, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 4087 + }, + { + "epoch": 3.2651757188498403, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 4088 + }, + { + "epoch": 3.265974440894569, + "grad_norm": 0.2060546875, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 4089 + }, + { + "epoch": 3.2667731629392973, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 4090 + }, + { + "epoch": 3.2675718849840254, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 4091 + }, + { + "epoch": 3.268370607028754, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0458, + "step": 4092 + }, + { + "epoch": 3.2691693290734825, + "grad_norm": 0.255859375, + "learning_rate": 0.0005, + "loss": 1.0458, + "step": 4093 + }, + { + "epoch": 3.269968051118211, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 4094 + }, + { + "epoch": 3.270766773162939, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 4095 + }, + { + "epoch": 3.2715654952076676, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.046, + "step": 4096 + }, + { + "epoch": 3.272364217252396, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 4097 + }, + { + "epoch": 3.2731629392971247, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 4098 + }, + { + "epoch": 3.273961661341853, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 4099 + }, + { + "epoch": 3.2747603833865817, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 4100 + }, + { + "epoch": 3.27555910543131, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 4101 + }, + { + "epoch": 3.2763578274760383, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0435, + "step": 4102 + }, + { + "epoch": 3.277156549520767, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 4103 + }, + { + "epoch": 3.2779552715654954, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 4104 + }, + { + "epoch": 3.2787539936102235, + "grad_norm": 0.2265625, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 4105 + }, + { + "epoch": 3.279552715654952, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.044, + "step": 4106 + }, + { + "epoch": 3.2803514376996805, + "grad_norm": 0.16796875, + "learning_rate": 0.0005, + "loss": 1.0406, + "step": 4107 + }, + { + "epoch": 3.281150159744409, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0439, + "step": 4108 + }, + { + "epoch": 3.2819488817891376, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 4109 + }, + { + "epoch": 3.2827476038338657, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 4110 + }, + { + "epoch": 3.283546325878594, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 4111 + }, + { + "epoch": 3.2843450479233227, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0476, + "step": 4112 + }, + { + "epoch": 3.2851437699680512, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 4113 + }, + { + "epoch": 3.2859424920127793, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 4114 + }, + { + "epoch": 3.286741214057508, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 4115 + }, + { + "epoch": 3.2875399361022364, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.044, + "step": 4116 + }, + { + "epoch": 3.288338658146965, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0429, + "step": 4117 + }, + { + "epoch": 3.2891373801916934, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0476, + "step": 4118 + }, + { + "epoch": 3.289936102236422, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 4119 + }, + { + "epoch": 3.29073482428115, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 4120 + }, + { + "epoch": 3.2915335463258786, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0405, + "step": 4121 + }, + { + "epoch": 3.292332268370607, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 4122 + }, + { + "epoch": 3.2931309904153356, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 4123 + }, + { + "epoch": 3.2939297124600637, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 4124 + }, + { + "epoch": 3.2947284345047922, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0419, + "step": 4125 + }, + { + "epoch": 3.2955271565495208, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 4126 + }, + { + "epoch": 3.2963258785942493, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 4127 + }, + { + "epoch": 3.297124600638978, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 4128 + }, + { + "epoch": 3.297923322683706, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 4129 + }, + { + "epoch": 3.2987220447284344, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 4130 + }, + { + "epoch": 3.299520766773163, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0437, + "step": 4131 + }, + { + "epoch": 3.3003194888178915, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 4132 + }, + { + "epoch": 3.3011182108626196, + "grad_norm": 0.177734375, + "learning_rate": 0.0005, + "loss": 1.0466, + "step": 4133 + }, + { + "epoch": 3.301916932907348, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 4134 + }, + { + "epoch": 3.3027156549520766, + "grad_norm": 0.1748046875, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 4135 + }, + { + "epoch": 3.303514376996805, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 4136 + }, + { + "epoch": 3.3043130990415337, + "grad_norm": 0.2080078125, + "learning_rate": 0.0005, + "loss": 1.0587, + "step": 4137 + }, + { + "epoch": 3.3051118210862622, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 4138 + }, + { + "epoch": 3.3059105431309903, + "grad_norm": 0.1640625, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 4139 + }, + { + "epoch": 3.306709265175719, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 4140 + }, + { + "epoch": 3.3075079872204474, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 4141 + }, + { + "epoch": 3.308306709265176, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0469, + "step": 4142 + }, + { + "epoch": 3.309105431309904, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 4143 + }, + { + "epoch": 3.3099041533546325, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0469, + "step": 4144 + }, + { + "epoch": 3.310702875399361, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0441, + "step": 4145 + }, + { + "epoch": 3.3115015974440896, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 4146 + }, + { + "epoch": 3.312300319488818, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0468, + "step": 4147 + }, + { + "epoch": 3.313099041533546, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0447, + "step": 4148 + }, + { + "epoch": 3.3138977635782747, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 4149 + }, + { + "epoch": 3.3146964856230032, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 4150 + }, + { + "epoch": 3.3154952076677318, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 4151 + }, + { + "epoch": 3.31629392971246, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0553, + "step": 4152 + }, + { + "epoch": 3.3170926517571884, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0589, + "step": 4153 + }, + { + "epoch": 3.317891373801917, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 4154 + }, + { + "epoch": 3.3186900958466454, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.048, + "step": 4155 + }, + { + "epoch": 3.319488817891374, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 4156 + }, + { + "epoch": 3.3202875399361025, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 4157 + }, + { + "epoch": 3.3210862619808306, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 4158 + }, + { + "epoch": 3.321884984025559, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 4159 + }, + { + "epoch": 3.3226837060702876, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0446, + "step": 4160 + }, + { + "epoch": 3.323482428115016, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 4161 + }, + { + "epoch": 3.3242811501597442, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 4162 + }, + { + "epoch": 3.3250798722044728, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.0446, + "step": 4163 + }, + { + "epoch": 3.3258785942492013, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 4164 + }, + { + "epoch": 3.32667731629393, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 4165 + }, + { + "epoch": 3.3274760383386583, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 4166 + }, + { + "epoch": 3.3282747603833864, + "grad_norm": 0.2001953125, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 4167 + }, + { + "epoch": 3.329073482428115, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 4168 + }, + { + "epoch": 3.3298722044728435, + "grad_norm": 0.2578125, + "learning_rate": 0.0005, + "loss": 1.0428, + "step": 4169 + }, + { + "epoch": 3.330670926517572, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 4170 + }, + { + "epoch": 3.3314696485623, + "grad_norm": 0.1611328125, + "learning_rate": 0.0005, + "loss": 1.048, + "step": 4171 + }, + { + "epoch": 3.3322683706070286, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0435, + "step": 4172 + }, + { + "epoch": 3.333067092651757, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 4173 + }, + { + "epoch": 3.3338658146964857, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0464, + "step": 4174 + }, + { + "epoch": 3.334664536741214, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 4175 + }, + { + "epoch": 3.3354632587859427, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0447, + "step": 4176 + }, + { + "epoch": 3.336261980830671, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0643, + "step": 4177 + }, + { + "epoch": 3.3370607028753994, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 4178 + }, + { + "epoch": 3.337859424920128, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 4179 + }, + { + "epoch": 3.3386581469648564, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0447, + "step": 4180 + }, + { + "epoch": 3.3394568690095845, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 4181 + }, + { + "epoch": 3.340255591054313, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 4182 + }, + { + "epoch": 3.3410543130990416, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0427, + "step": 4183 + }, + { + "epoch": 3.34185303514377, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0405, + "step": 4184 + }, + { + "epoch": 3.3426517571884986, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 4185 + }, + { + "epoch": 3.3434504792332267, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 4186 + }, + { + "epoch": 3.344249201277955, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 4187 + }, + { + "epoch": 3.3450479233226837, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 4188 + }, + { + "epoch": 3.3458466453674123, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 4189 + }, + { + "epoch": 3.3466453674121404, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0431, + "step": 4190 + }, + { + "epoch": 3.347444089456869, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0488, + "step": 4191 + }, + { + "epoch": 3.3482428115015974, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0421, + "step": 4192 + }, + { + "epoch": 3.349041533546326, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 4193 + }, + { + "epoch": 3.3498402555910545, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 4194 + }, + { + "epoch": 3.3506389776357826, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 4195 + }, + { + "epoch": 3.351437699680511, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 4196 + }, + { + "epoch": 3.3522364217252396, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 4197 + }, + { + "epoch": 3.353035143769968, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.054, + "step": 4198 + }, + { + "epoch": 3.3538338658146967, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.043, + "step": 4199 + }, + { + "epoch": 3.3546325878594248, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0455, + "step": 4200 + }, + { + "epoch": 3.3554313099041533, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 4201 + }, + { + "epoch": 3.356230031948882, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0458, + "step": 4202 + }, + { + "epoch": 3.3570287539936103, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0432, + "step": 4203 + }, + { + "epoch": 3.357827476038339, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0482, + "step": 4204 + }, + { + "epoch": 3.358626198083067, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 4205 + }, + { + "epoch": 3.3594249201277955, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 4206 + }, + { + "epoch": 3.360223642172524, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 4207 + }, + { + "epoch": 3.3610223642172525, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.034, + "step": 4208 + }, + { + "epoch": 3.3618210862619806, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 4209 + }, + { + "epoch": 3.362619808306709, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 4210 + }, + { + "epoch": 3.3634185303514377, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 4211 + }, + { + "epoch": 3.364217252396166, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 4212 + }, + { + "epoch": 3.3650159744408947, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 4213 + }, + { + "epoch": 3.365814696485623, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 4214 + }, + { + "epoch": 3.3666134185303513, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 4215 + }, + { + "epoch": 3.36741214057508, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0448, + "step": 4216 + }, + { + "epoch": 3.3682108626198084, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 4217 + }, + { + "epoch": 3.369009584664537, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 4218 + }, + { + "epoch": 3.369808306709265, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 4219 + }, + { + "epoch": 3.3706070287539935, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 4220 + }, + { + "epoch": 3.371405750798722, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0409, + "step": 4221 + }, + { + "epoch": 3.3722044728434506, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0488, + "step": 4222 + }, + { + "epoch": 3.373003194888179, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0465, + "step": 4223 + }, + { + "epoch": 3.373801916932907, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 4224 + }, + { + "epoch": 3.3746006389776357, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0436, + "step": 4225 + }, + { + "epoch": 3.3753993610223643, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 4226 + }, + { + "epoch": 3.376198083067093, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 4227 + }, + { + "epoch": 3.376996805111821, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0444, + "step": 4228 + }, + { + "epoch": 3.3777955271565494, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 4229 + }, + { + "epoch": 3.378594249201278, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 4230 + }, + { + "epoch": 3.3793929712460065, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0539, + "step": 4231 + }, + { + "epoch": 3.380191693290735, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 4232 + }, + { + "epoch": 3.380990415335463, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.056, + "step": 4233 + }, + { + "epoch": 3.3817891373801916, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 4234 + }, + { + "epoch": 3.38258785942492, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 4235 + }, + { + "epoch": 3.3833865814696487, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0458, + "step": 4236 + }, + { + "epoch": 3.384185303514377, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0405, + "step": 4237 + }, + { + "epoch": 3.3849840255591053, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 4238 + }, + { + "epoch": 3.385782747603834, + "grad_norm": 0.189453125, + "learning_rate": 0.0005, + "loss": 1.0424, + "step": 4239 + }, + { + "epoch": 3.3865814696485623, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0416, + "step": 4240 + }, + { + "epoch": 3.387380191693291, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 4241 + }, + { + "epoch": 3.3881789137380194, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0434, + "step": 4242 + }, + { + "epoch": 3.3889776357827475, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 4243 + }, + { + "epoch": 3.389776357827476, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 4244 + }, + { + "epoch": 3.3905750798722045, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 4245 + }, + { + "epoch": 3.391373801916933, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 4246 + }, + { + "epoch": 3.392172523961661, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 4247 + }, + { + "epoch": 3.3929712460063897, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 4248 + }, + { + "epoch": 3.393769968051118, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 4249 + }, + { + "epoch": 3.3945686900958467, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 4250 + }, + { + "epoch": 3.3953674121405752, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 4251 + }, + { + "epoch": 3.3961661341853033, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0467, + "step": 4252 + }, + { + "epoch": 3.396964856230032, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0429, + "step": 4253 + }, + { + "epoch": 3.3977635782747604, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0416, + "step": 4254 + }, + { + "epoch": 3.398562300319489, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 4255 + }, + { + "epoch": 3.3993610223642174, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0457, + "step": 4256 + }, + { + "epoch": 3.4001597444089455, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0434, + "step": 4257 + }, + { + "epoch": 3.400958466453674, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 4258 + }, + { + "epoch": 3.4017571884984026, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 4259 + }, + { + "epoch": 3.402555910543131, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 4260 + }, + { + "epoch": 3.4033546325878596, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 4261 + }, + { + "epoch": 3.4041533546325877, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 4262 + }, + { + "epoch": 3.4049520766773163, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0456, + "step": 4263 + }, + { + "epoch": 3.405750798722045, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0456, + "step": 4264 + }, + { + "epoch": 3.4065495207667733, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0452, + "step": 4265 + }, + { + "epoch": 3.4073482428115014, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 4266 + }, + { + "epoch": 3.40814696485623, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0453, + "step": 4267 + }, + { + "epoch": 3.4089456869009584, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0429, + "step": 4268 + }, + { + "epoch": 3.409744408945687, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 4269 + }, + { + "epoch": 3.4105431309904155, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0409, + "step": 4270 + }, + { + "epoch": 3.4113418530351436, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 4271 + }, + { + "epoch": 3.412140575079872, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 4272 + }, + { + "epoch": 3.4129392971246006, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0416, + "step": 4273 + }, + { + "epoch": 3.413738019169329, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 4274 + }, + { + "epoch": 3.4145367412140573, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 4275 + }, + { + "epoch": 3.415335463258786, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0452, + "step": 4276 + }, + { + "epoch": 3.4161341853035143, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0452, + "step": 4277 + }, + { + "epoch": 3.416932907348243, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 4278 + }, + { + "epoch": 3.4177316293929714, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 4279 + }, + { + "epoch": 3.4185303514377, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0565, + "step": 4280 + }, + { + "epoch": 3.419329073482428, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 4281 + }, + { + "epoch": 3.4201277955271565, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0452, + "step": 4282 + }, + { + "epoch": 3.420926517571885, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 4283 + }, + { + "epoch": 3.4217252396166136, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 4284 + }, + { + "epoch": 3.4225239616613417, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0469, + "step": 4285 + }, + { + "epoch": 3.42332268370607, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0384, + "step": 4286 + }, + { + "epoch": 3.4241214057507987, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0446, + "step": 4287 + }, + { + "epoch": 3.4249201277955272, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 4288 + }, + { + "epoch": 3.4257188498402558, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.043, + "step": 4289 + }, + { + "epoch": 3.426517571884984, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0445, + "step": 4290 + }, + { + "epoch": 3.4273162939297124, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 4291 + }, + { + "epoch": 3.428115015974441, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0417, + "step": 4292 + }, + { + "epoch": 3.4289137380191694, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0389, + "step": 4293 + }, + { + "epoch": 3.4297124600638975, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 4294 + }, + { + "epoch": 3.430511182108626, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0407, + "step": 4295 + }, + { + "epoch": 3.4313099041533546, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0416, + "step": 4296 + }, + { + "epoch": 3.432108626198083, + "grad_norm": 0.212890625, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 4297 + }, + { + "epoch": 3.4329073482428116, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 4298 + }, + { + "epoch": 3.43370607028754, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0429, + "step": 4299 + }, + { + "epoch": 3.4345047923322682, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0416, + "step": 4300 + }, + { + "epoch": 3.4353035143769968, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 4301 + }, + { + "epoch": 3.4361022364217253, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 4302 + }, + { + "epoch": 3.436900958466454, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0464, + "step": 4303 + }, + { + "epoch": 3.437699680511182, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0576, + "step": 4304 + }, + { + "epoch": 3.4384984025559104, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0385, + "step": 4305 + }, + { + "epoch": 3.439297124600639, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 4306 + }, + { + "epoch": 3.4400958466453675, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 4307 + }, + { + "epoch": 3.440894568690096, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0456, + "step": 4308 + }, + { + "epoch": 3.441693290734824, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0453, + "step": 4309 + }, + { + "epoch": 3.4424920127795526, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 4310 + }, + { + "epoch": 3.443290734824281, + "grad_norm": 0.18359375, + "learning_rate": 0.0005, + "loss": 1.0446, + "step": 4311 + }, + { + "epoch": 3.4440894568690097, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0459, + "step": 4312 + }, + { + "epoch": 3.4448881789137378, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 4313 + }, + { + "epoch": 3.4456869009584663, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 4314 + }, + { + "epoch": 3.446485623003195, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 4315 + }, + { + "epoch": 3.4472843450479234, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0442, + "step": 4316 + }, + { + "epoch": 3.448083067092652, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 4317 + }, + { + "epoch": 3.4488817891373804, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 4318 + }, + { + "epoch": 3.4496805111821085, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 4319 + }, + { + "epoch": 3.450479233226837, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0428, + "step": 4320 + }, + { + "epoch": 3.4512779552715656, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.045, + "step": 4321 + }, + { + "epoch": 3.452076677316294, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 4322 + }, + { + "epoch": 3.452875399361022, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0465, + "step": 4323 + }, + { + "epoch": 3.4536741214057507, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.046, + "step": 4324 + }, + { + "epoch": 3.4544728434504792, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0417, + "step": 4325 + }, + { + "epoch": 3.4552715654952078, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 4326 + }, + { + "epoch": 3.4560702875399363, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0597, + "step": 4327 + }, + { + "epoch": 3.4568690095846644, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0404, + "step": 4328 + }, + { + "epoch": 3.457667731629393, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 4329 + }, + { + "epoch": 3.4584664536741214, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 4330 + }, + { + "epoch": 3.45926517571885, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0432, + "step": 4331 + }, + { + "epoch": 3.460063897763578, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0554, + "step": 4332 + }, + { + "epoch": 3.4608626198083066, + "grad_norm": 0.16015625, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 4333 + }, + { + "epoch": 3.461661341853035, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.0457, + "step": 4334 + }, + { + "epoch": 3.4624600638977636, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0424, + "step": 4335 + }, + { + "epoch": 3.463258785942492, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 4336 + }, + { + "epoch": 3.4640575079872207, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 4337 + }, + { + "epoch": 3.4648562300319488, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0409, + "step": 4338 + }, + { + "epoch": 3.4656549520766773, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0433, + "step": 4339 + }, + { + "epoch": 3.466453674121406, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.042, + "step": 4340 + }, + { + "epoch": 3.4672523961661343, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0436, + "step": 4341 + }, + { + "epoch": 3.4680511182108624, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 4342 + }, + { + "epoch": 3.468849840255591, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0423, + "step": 4343 + }, + { + "epoch": 3.4696485623003195, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0411, + "step": 4344 + }, + { + "epoch": 3.470447284345048, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0417, + "step": 4345 + }, + { + "epoch": 3.4712460063897765, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 4346 + }, + { + "epoch": 3.4720447284345046, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0426, + "step": 4347 + }, + { + "epoch": 3.472843450479233, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 4348 + }, + { + "epoch": 3.4736421725239617, + "grad_norm": 0.1572265625, + "learning_rate": 0.0005, + "loss": 1.0447, + "step": 4349 + }, + { + "epoch": 3.47444089456869, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0482, + "step": 4350 + }, + { + "epoch": 3.4752396166134183, + "grad_norm": 0.18359375, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 4351 + }, + { + "epoch": 3.476038338658147, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0442, + "step": 4352 + }, + { + "epoch": 3.4768370607028753, + "grad_norm": 0.267578125, + "learning_rate": 0.0005, + "loss": 1.0465, + "step": 4353 + }, + { + "epoch": 3.477635782747604, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 4354 + }, + { + "epoch": 3.4784345047923324, + "grad_norm": 0.1806640625, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 4355 + }, + { + "epoch": 3.479233226837061, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 4356 + }, + { + "epoch": 3.480031948881789, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 4357 + }, + { + "epoch": 3.4808306709265175, + "grad_norm": 0.16015625, + "learning_rate": 0.0005, + "loss": 1.0575, + "step": 4358 + }, + { + "epoch": 3.481629392971246, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0453, + "step": 4359 + }, + { + "epoch": 3.4824281150159746, + "grad_norm": 0.189453125, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 4360 + }, + { + "epoch": 3.4832268370607027, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0421, + "step": 4361 + }, + { + "epoch": 3.484025559105431, + "grad_norm": 0.1806640625, + "learning_rate": 0.0005, + "loss": 1.0488, + "step": 4362 + }, + { + "epoch": 3.4848242811501597, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 4363 + }, + { + "epoch": 3.4856230031948883, + "grad_norm": 0.1708984375, + "learning_rate": 0.0005, + "loss": 1.037, + "step": 4364 + }, + { + "epoch": 3.486421725239617, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0593, + "step": 4365 + }, + { + "epoch": 3.487220447284345, + "grad_norm": 0.1630859375, + "learning_rate": 0.0005, + "loss": 1.0456, + "step": 4366 + }, + { + "epoch": 3.4880191693290734, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.0414, + "step": 4367 + }, + { + "epoch": 3.488817891373802, + "grad_norm": 0.162109375, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 4368 + }, + { + "epoch": 3.4896166134185305, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 4369 + }, + { + "epoch": 3.4904153354632586, + "grad_norm": 0.1767578125, + "learning_rate": 0.0005, + "loss": 1.038, + "step": 4370 + }, + { + "epoch": 3.491214057507987, + "grad_norm": 0.2314453125, + "learning_rate": 0.0005, + "loss": 1.0459, + "step": 4371 + }, + { + "epoch": 3.4920127795527156, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 4372 + }, + { + "epoch": 3.492811501597444, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0454, + "step": 4373 + }, + { + "epoch": 3.4936102236421727, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 4374 + }, + { + "epoch": 3.494408945686901, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0442, + "step": 4375 + }, + { + "epoch": 3.4952076677316293, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 4376 + }, + { + "epoch": 3.496006389776358, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 4377 + }, + { + "epoch": 3.4968051118210863, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 4378 + }, + { + "epoch": 3.497603833865815, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0428, + "step": 4379 + }, + { + "epoch": 3.498402555910543, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 4380 + }, + { + "epoch": 3.4992012779552715, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 4381 + }, + { + "epoch": 3.5, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.048, + "step": 4382 + }, + { + "epoch": 3.5007987220447285, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0432, + "step": 4383 + }, + { + "epoch": 3.501597444089457, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0628, + "step": 4384 + }, + { + "epoch": 3.502396166134185, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0613, + "step": 4385 + }, + { + "epoch": 3.5031948881789137, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 4386 + }, + { + "epoch": 3.503993610223642, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 4387 + }, + { + "epoch": 3.5047923322683707, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0451, + "step": 4388 + }, + { + "epoch": 3.505591054313099, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 4389 + }, + { + "epoch": 3.5063897763578273, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 4390 + }, + { + "epoch": 3.507188498402556, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0427, + "step": 4391 + }, + { + "epoch": 3.5079872204472844, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 4392 + }, + { + "epoch": 3.508785942492013, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0407, + "step": 4393 + }, + { + "epoch": 3.5095846645367414, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0467, + "step": 4394 + }, + { + "epoch": 3.5103833865814695, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 4395 + }, + { + "epoch": 3.511182108626198, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 4396 + }, + { + "epoch": 3.5119808306709266, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 4397 + }, + { + "epoch": 3.512779552715655, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0391, + "step": 4398 + }, + { + "epoch": 3.513578274760383, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 4399 + }, + { + "epoch": 3.5143769968051117, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0452, + "step": 4400 + }, + { + "epoch": 3.5151757188498403, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0452, + "step": 4401 + }, + { + "epoch": 3.515974440894569, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 4402 + }, + { + "epoch": 3.5167731629392973, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 4403 + }, + { + "epoch": 3.5175718849840254, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 4404 + }, + { + "epoch": 3.518370607028754, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0399, + "step": 4405 + }, + { + "epoch": 3.5191693290734825, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 4406 + }, + { + "epoch": 3.519968051118211, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0468, + "step": 4407 + }, + { + "epoch": 3.520766773162939, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 4408 + }, + { + "epoch": 3.5215654952076676, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 4409 + }, + { + "epoch": 3.522364217252396, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0458, + "step": 4410 + }, + { + "epoch": 3.5231629392971247, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 4411 + }, + { + "epoch": 3.523961661341853, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0445, + "step": 4412 + }, + { + "epoch": 3.5247603833865817, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 4413 + }, + { + "epoch": 3.52555910543131, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0616, + "step": 4414 + }, + { + "epoch": 3.5263578274760383, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 4415 + }, + { + "epoch": 3.527156549520767, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 4416 + }, + { + "epoch": 3.527955271565495, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 4417 + }, + { + "epoch": 3.5287539936102235, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0435, + "step": 4418 + }, + { + "epoch": 3.529552715654952, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 4419 + }, + { + "epoch": 3.5303514376996805, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0455, + "step": 4420 + }, + { + "epoch": 3.531150159744409, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0454, + "step": 4421 + }, + { + "epoch": 3.5319488817891376, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0446, + "step": 4422 + }, + { + "epoch": 3.5327476038338657, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 4423 + }, + { + "epoch": 3.533546325878594, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 4424 + }, + { + "epoch": 3.5343450479233227, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0594, + "step": 4425 + }, + { + "epoch": 3.5351437699680512, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0456, + "step": 4426 + }, + { + "epoch": 3.5359424920127793, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0456, + "step": 4427 + }, + { + "epoch": 3.536741214057508, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.045, + "step": 4428 + }, + { + "epoch": 3.5375399361022364, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 4429 + }, + { + "epoch": 3.538338658146965, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 4430 + }, + { + "epoch": 3.5391373801916934, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0407, + "step": 4431 + }, + { + "epoch": 3.539936102236422, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0464, + "step": 4432 + }, + { + "epoch": 3.54073482428115, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 4433 + }, + { + "epoch": 3.5415335463258786, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 4434 + }, + { + "epoch": 3.542332268370607, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 4435 + }, + { + "epoch": 3.543130990415335, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 4436 + }, + { + "epoch": 3.5439297124600637, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 4437 + }, + { + "epoch": 3.5447284345047922, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 4438 + }, + { + "epoch": 3.5455271565495208, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0449, + "step": 4439 + }, + { + "epoch": 3.5463258785942493, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0464, + "step": 4440 + }, + { + "epoch": 3.547124600638978, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 4441 + }, + { + "epoch": 3.547923322683706, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0407, + "step": 4442 + }, + { + "epoch": 3.5487220447284344, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0548, + "step": 4443 + }, + { + "epoch": 3.549520766773163, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0391, + "step": 4444 + }, + { + "epoch": 3.5503194888178915, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0472, + "step": 4445 + }, + { + "epoch": 3.5511182108626196, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 4446 + }, + { + "epoch": 3.551916932907348, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0455, + "step": 4447 + }, + { + "epoch": 3.5527156549520766, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 4448 + }, + { + "epoch": 3.553514376996805, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0453, + "step": 4449 + }, + { + "epoch": 3.5543130990415337, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 4450 + }, + { + "epoch": 3.5551118210862622, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 4451 + }, + { + "epoch": 3.5559105431309903, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 4452 + }, + { + "epoch": 3.556709265175719, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0617, + "step": 4453 + }, + { + "epoch": 3.5575079872204474, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 4454 + }, + { + "epoch": 3.5583067092651754, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0399, + "step": 4455 + }, + { + "epoch": 3.559105431309904, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0429, + "step": 4456 + }, + { + "epoch": 3.5599041533546325, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 4457 + }, + { + "epoch": 3.560702875399361, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0454, + "step": 4458 + }, + { + "epoch": 3.5615015974440896, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.0456, + "step": 4459 + }, + { + "epoch": 3.562300319488818, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 4460 + }, + { + "epoch": 3.563099041533546, + "grad_norm": 0.16015625, + "learning_rate": 0.0005, + "loss": 1.0468, + "step": 4461 + }, + { + "epoch": 3.5638977635782747, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0436, + "step": 4462 + }, + { + "epoch": 3.5646964856230032, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 4463 + }, + { + "epoch": 3.5654952076677318, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 4464 + }, + { + "epoch": 3.56629392971246, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 4465 + }, + { + "epoch": 3.5670926517571884, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0417, + "step": 4466 + }, + { + "epoch": 3.567891373801917, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 4467 + }, + { + "epoch": 3.5686900958466454, + "grad_norm": 0.16796875, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 4468 + }, + { + "epoch": 3.569488817891374, + "grad_norm": 0.2001953125, + "learning_rate": 0.0005, + "loss": 1.0431, + "step": 4469 + }, + { + "epoch": 3.5702875399361025, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 4470 + }, + { + "epoch": 3.5710862619808306, + "grad_norm": 0.19140625, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 4471 + }, + { + "epoch": 3.571884984025559, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 4472 + }, + { + "epoch": 3.5726837060702876, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0458, + "step": 4473 + }, + { + "epoch": 3.5734824281150157, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0428, + "step": 4474 + }, + { + "epoch": 3.5742811501597442, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 4475 + }, + { + "epoch": 3.5750798722044728, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0403, + "step": 4476 + }, + { + "epoch": 3.5758785942492013, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0444, + "step": 4477 + }, + { + "epoch": 3.57667731629393, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 4478 + }, + { + "epoch": 3.5774760383386583, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 4479 + }, + { + "epoch": 3.5782747603833864, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 4480 + }, + { + "epoch": 3.579073482428115, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0453, + "step": 4481 + }, + { + "epoch": 3.5798722044728435, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0476, + "step": 4482 + }, + { + "epoch": 3.580670926517572, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 4483 + }, + { + "epoch": 3.5814696485623, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 4484 + }, + { + "epoch": 3.5822683706070286, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 4485 + }, + { + "epoch": 3.583067092651757, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0466, + "step": 4486 + }, + { + "epoch": 3.5838658146964857, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 4487 + }, + { + "epoch": 3.584664536741214, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 4488 + }, + { + "epoch": 3.5854632587859427, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 4489 + }, + { + "epoch": 3.586261980830671, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 4490 + }, + { + "epoch": 3.5870607028753994, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0411, + "step": 4491 + }, + { + "epoch": 3.587859424920128, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0372, + "step": 4492 + }, + { + "epoch": 3.588658146964856, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 4493 + }, + { + "epoch": 3.5894568690095845, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0441, + "step": 4494 + }, + { + "epoch": 3.590255591054313, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 4495 + }, + { + "epoch": 3.5910543130990416, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 4496 + }, + { + "epoch": 3.59185303514377, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 4497 + }, + { + "epoch": 3.5926517571884986, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0411, + "step": 4498 + }, + { + "epoch": 3.5934504792332267, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0469, + "step": 4499 + }, + { + "epoch": 3.594249201277955, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 4500 + }, + { + "epoch": 3.5950479233226837, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 4501 + }, + { + "epoch": 3.5958466453674123, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 4502 + }, + { + "epoch": 3.5966453674121404, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0393, + "step": 4503 + }, + { + "epoch": 3.597444089456869, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 4504 + }, + { + "epoch": 3.5982428115015974, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.046, + "step": 4505 + }, + { + "epoch": 3.599041533546326, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.045, + "step": 4506 + }, + { + "epoch": 3.5998402555910545, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 4507 + }, + { + "epoch": 3.600638977635783, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 4508 + }, + { + "epoch": 3.601437699680511, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 4509 + }, + { + "epoch": 3.6022364217252396, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.0447, + "step": 4510 + }, + { + "epoch": 3.603035143769968, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 4511 + }, + { + "epoch": 3.6038338658146962, + "grad_norm": 0.1640625, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 4512 + }, + { + "epoch": 3.6046325878594248, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 4513 + }, + { + "epoch": 3.6054313099041533, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0412, + "step": 4514 + }, + { + "epoch": 3.606230031948882, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 4515 + }, + { + "epoch": 3.6070287539936103, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.0469, + "step": 4516 + }, + { + "epoch": 3.607827476038339, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 4517 + }, + { + "epoch": 3.608626198083067, + "grad_norm": 0.2119140625, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 4518 + }, + { + "epoch": 3.6094249201277955, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0427, + "step": 4519 + }, + { + "epoch": 3.610223642172524, + "grad_norm": 0.267578125, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 4520 + }, + { + "epoch": 3.6110223642172525, + "grad_norm": 0.17578125, + "learning_rate": 0.0005, + "loss": 1.0513, + "step": 4521 + }, + { + "epoch": 3.6118210862619806, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 4522 + }, + { + "epoch": 3.612619808306709, + "grad_norm": 0.31640625, + "learning_rate": 0.0005, + "loss": 1.0449, + "step": 4523 + }, + { + "epoch": 3.6134185303514377, + "grad_norm": 0.162109375, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 4524 + }, + { + "epoch": 3.614217252396166, + "grad_norm": 0.1826171875, + "learning_rate": 0.0005, + "loss": 1.0549, + "step": 4525 + }, + { + "epoch": 3.6150159744408947, + "grad_norm": 0.1904296875, + "learning_rate": 0.0005, + "loss": 1.0458, + "step": 4526 + }, + { + "epoch": 3.6158146964856233, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.046, + "step": 4527 + }, + { + "epoch": 3.6166134185303513, + "grad_norm": 0.19921875, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 4528 + }, + { + "epoch": 3.61741214057508, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 4529 + }, + { + "epoch": 3.6182108626198084, + "grad_norm": 0.1796875, + "learning_rate": 0.0005, + "loss": 1.0455, + "step": 4530 + }, + { + "epoch": 3.6190095846645365, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0459, + "step": 4531 + }, + { + "epoch": 3.619808306709265, + "grad_norm": 0.265625, + "learning_rate": 0.0005, + "loss": 1.0555, + "step": 4532 + }, + { + "epoch": 3.6206070287539935, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0564, + "step": 4533 + }, + { + "epoch": 3.621405750798722, + "grad_norm": 0.37890625, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 4534 + }, + { + "epoch": 3.6222044728434506, + "grad_norm": 0.23828125, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 4535 + }, + { + "epoch": 3.623003194888179, + "grad_norm": 0.23828125, + "learning_rate": 0.0005, + "loss": 1.0375, + "step": 4536 + }, + { + "epoch": 3.623801916932907, + "grad_norm": 0.427734375, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 4537 + }, + { + "epoch": 3.6246006389776357, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 4538 + }, + { + "epoch": 3.6253993610223643, + "grad_norm": 0.298828125, + "learning_rate": 0.0005, + "loss": 1.0426, + "step": 4539 + }, + { + "epoch": 3.626198083067093, + "grad_norm": 0.1953125, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 4540 + }, + { + "epoch": 3.626996805111821, + "grad_norm": 0.16015625, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 4541 + }, + { + "epoch": 3.6277955271565494, + "grad_norm": 0.189453125, + "learning_rate": 0.0005, + "loss": 1.0422, + "step": 4542 + }, + { + "epoch": 3.628594249201278, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 4543 + }, + { + "epoch": 3.6293929712460065, + "grad_norm": 0.275390625, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 4544 + }, + { + "epoch": 3.630191693290735, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 4545 + }, + { + "epoch": 3.6309904153354635, + "grad_norm": 0.26171875, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 4546 + }, + { + "epoch": 3.6317891373801916, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0465, + "step": 4547 + }, + { + "epoch": 3.63258785942492, + "grad_norm": 0.240234375, + "learning_rate": 0.0005, + "loss": 1.0538, + "step": 4548 + }, + { + "epoch": 3.6333865814696487, + "grad_norm": 0.2216796875, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 4549 + }, + { + "epoch": 3.6341853035143767, + "grad_norm": 0.1806640625, + "learning_rate": 0.0005, + "loss": 1.0399, + "step": 4550 + }, + { + "epoch": 3.6349840255591053, + "grad_norm": 0.232421875, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 4551 + }, + { + "epoch": 3.635782747603834, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0469, + "step": 4552 + }, + { + "epoch": 3.6365814696485623, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0453, + "step": 4553 + }, + { + "epoch": 3.637380191693291, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 4554 + }, + { + "epoch": 3.6381789137380194, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 4555 + }, + { + "epoch": 3.6389776357827475, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0472, + "step": 4556 + }, + { + "epoch": 3.639776357827476, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0404, + "step": 4557 + }, + { + "epoch": 3.6405750798722045, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 4558 + }, + { + "epoch": 3.641373801916933, + "grad_norm": 0.1708984375, + "learning_rate": 0.0005, + "loss": 1.048, + "step": 4559 + }, + { + "epoch": 3.642172523961661, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 4560 + }, + { + "epoch": 3.6429712460063897, + "grad_norm": 0.2177734375, + "learning_rate": 0.0005, + "loss": 1.0423, + "step": 4561 + }, + { + "epoch": 3.643769968051118, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0431, + "step": 4562 + }, + { + "epoch": 3.6445686900958467, + "grad_norm": 0.28515625, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 4563 + }, + { + "epoch": 3.6453674121405752, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 4564 + }, + { + "epoch": 3.6461661341853038, + "grad_norm": 0.26953125, + "learning_rate": 0.0005, + "loss": 1.0458, + "step": 4565 + }, + { + "epoch": 3.646964856230032, + "grad_norm": 0.189453125, + "learning_rate": 0.0005, + "loss": 1.0476, + "step": 4566 + }, + { + "epoch": 3.6477635782747604, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0422, + "step": 4567 + }, + { + "epoch": 3.648562300319489, + "grad_norm": 0.2353515625, + "learning_rate": 0.0005, + "loss": 1.038, + "step": 4568 + }, + { + "epoch": 3.649361022364217, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 4569 + }, + { + "epoch": 3.6501597444089455, + "grad_norm": 0.380859375, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 4570 + }, + { + "epoch": 3.650958466453674, + "grad_norm": 0.2021484375, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 4571 + }, + { + "epoch": 3.6517571884984026, + "grad_norm": 0.2275390625, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 4572 + }, + { + "epoch": 3.652555910543131, + "grad_norm": 0.26953125, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 4573 + }, + { + "epoch": 3.6533546325878596, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 4574 + }, + { + "epoch": 3.6541533546325877, + "grad_norm": 0.2490234375, + "learning_rate": 0.0005, + "loss": 1.0393, + "step": 4575 + }, + { + "epoch": 3.6549520766773163, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 4576 + }, + { + "epoch": 3.655750798722045, + "grad_norm": 0.349609375, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 4577 + }, + { + "epoch": 3.6565495207667733, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.0454, + "step": 4578 + }, + { + "epoch": 3.6573482428115014, + "grad_norm": 0.203125, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 4579 + }, + { + "epoch": 3.65814696485623, + "grad_norm": 0.212890625, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 4580 + }, + { + "epoch": 3.6589456869009584, + "grad_norm": 0.173828125, + "learning_rate": 0.0005, + "loss": 1.0457, + "step": 4581 + }, + { + "epoch": 3.659744408945687, + "grad_norm": 0.353515625, + "learning_rate": 0.0005, + "loss": 1.0457, + "step": 4582 + }, + { + "epoch": 3.6605431309904155, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 4583 + }, + { + "epoch": 3.661341853035144, + "grad_norm": 0.30078125, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 4584 + }, + { + "epoch": 3.662140575079872, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.0429, + "step": 4585 + }, + { + "epoch": 3.6629392971246006, + "grad_norm": 0.1640625, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 4586 + }, + { + "epoch": 3.663738019169329, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 4587 + }, + { + "epoch": 3.6645367412140573, + "grad_norm": 0.1826171875, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 4588 + }, + { + "epoch": 3.665335463258786, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.0439, + "step": 4589 + }, + { + "epoch": 3.6661341853035143, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 4590 + }, + { + "epoch": 3.666932907348243, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0417, + "step": 4591 + }, + { + "epoch": 3.6677316293929714, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 4592 + }, + { + "epoch": 3.6685303514377, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 4593 + }, + { + "epoch": 3.669329073482428, + "grad_norm": 0.208984375, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 4594 + }, + { + "epoch": 3.6701277955271565, + "grad_norm": 0.1806640625, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 4595 + }, + { + "epoch": 3.670926517571885, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 4596 + }, + { + "epoch": 3.6717252396166136, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 4597 + }, + { + "epoch": 3.6725239616613417, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 4598 + }, + { + "epoch": 3.67332268370607, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 4599 + }, + { + "epoch": 3.6741214057507987, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 4600 + }, + { + "epoch": 3.6749201277955272, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0465, + "step": 4601 + }, + { + "epoch": 3.6757188498402558, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 4602 + }, + { + "epoch": 3.6765175718849843, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 4603 + }, + { + "epoch": 3.6773162939297124, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0404, + "step": 4604 + }, + { + "epoch": 3.678115015974441, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.0467, + "step": 4605 + }, + { + "epoch": 3.6789137380191694, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 4606 + }, + { + "epoch": 3.6797124600638975, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.045, + "step": 4607 + }, + { + "epoch": 3.680511182108626, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 4608 + }, + { + "epoch": 3.6813099041533546, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0449, + "step": 4609 + }, + { + "epoch": 3.682108626198083, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0431, + "step": 4610 + }, + { + "epoch": 3.6829073482428116, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 4611 + }, + { + "epoch": 3.68370607028754, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 4612 + }, + { + "epoch": 3.6845047923322682, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0453, + "step": 4613 + }, + { + "epoch": 3.6853035143769968, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 4614 + }, + { + "epoch": 3.6861022364217253, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0457, + "step": 4615 + }, + { + "epoch": 3.686900958466454, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0448, + "step": 4616 + }, + { + "epoch": 3.687699680511182, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 4617 + }, + { + "epoch": 3.6884984025559104, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 4618 + }, + { + "epoch": 3.689297124600639, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0426, + "step": 4619 + }, + { + "epoch": 3.6900958466453675, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 4620 + }, + { + "epoch": 3.690894568690096, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 4621 + }, + { + "epoch": 3.6916932907348246, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.044, + "step": 4622 + }, + { + "epoch": 3.6924920127795526, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0407, + "step": 4623 + }, + { + "epoch": 3.693290734824281, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 4624 + }, + { + "epoch": 3.6940894568690097, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 4625 + }, + { + "epoch": 3.6948881789137378, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0414, + "step": 4626 + }, + { + "epoch": 3.6956869009584663, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 4627 + }, + { + "epoch": 3.696485623003195, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 4628 + }, + { + "epoch": 3.6972843450479234, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0448, + "step": 4629 + }, + { + "epoch": 3.698083067092652, + "grad_norm": 0.173828125, + "learning_rate": 0.0005, + "loss": 1.0595, + "step": 4630 + }, + { + "epoch": 3.6988817891373804, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0456, + "step": 4631 + }, + { + "epoch": 3.6996805111821085, + "grad_norm": 0.27734375, + "learning_rate": 0.0005, + "loss": 1.0437, + "step": 4632 + }, + { + "epoch": 3.700479233226837, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.0442, + "step": 4633 + }, + { + "epoch": 3.7012779552715656, + "grad_norm": 0.17578125, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 4634 + }, + { + "epoch": 3.702076677316294, + "grad_norm": 0.1689453125, + "learning_rate": 0.0005, + "loss": 1.046, + "step": 4635 + }, + { + "epoch": 3.702875399361022, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0488, + "step": 4636 + }, + { + "epoch": 3.7036741214057507, + "grad_norm": 0.2041015625, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 4637 + }, + { + "epoch": 3.7044728434504792, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0426, + "step": 4638 + }, + { + "epoch": 3.7052715654952078, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0428, + "step": 4639 + }, + { + "epoch": 3.7060702875399363, + "grad_norm": 0.173828125, + "learning_rate": 0.0005, + "loss": 1.0458, + "step": 4640 + }, + { + "epoch": 3.706869009584665, + "grad_norm": 0.23828125, + "learning_rate": 0.0005, + "loss": 1.0465, + "step": 4641 + }, + { + "epoch": 3.707667731629393, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0413, + "step": 4642 + }, + { + "epoch": 3.7084664536741214, + "grad_norm": 0.16015625, + "learning_rate": 0.0005, + "loss": 1.0416, + "step": 4643 + }, + { + "epoch": 3.70926517571885, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.04, + "step": 4644 + }, + { + "epoch": 3.710063897763578, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 4645 + }, + { + "epoch": 3.7108626198083066, + "grad_norm": 0.1611328125, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 4646 + }, + { + "epoch": 3.711661341853035, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 4647 + }, + { + "epoch": 3.7124600638977636, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 4648 + }, + { + "epoch": 3.713258785942492, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0416, + "step": 4649 + }, + { + "epoch": 3.7140575079872207, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 4650 + }, + { + "epoch": 3.7148562300319488, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 4651 + }, + { + "epoch": 3.7156549520766773, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 4652 + }, + { + "epoch": 3.716453674121406, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 4653 + }, + { + "epoch": 3.7172523961661343, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 4654 + }, + { + "epoch": 3.7180511182108624, + "grad_norm": 0.205078125, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 4655 + }, + { + "epoch": 3.718849840255591, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 4656 + }, + { + "epoch": 3.7196485623003195, + "grad_norm": 0.25390625, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 4657 + }, + { + "epoch": 3.720447284345048, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 4658 + }, + { + "epoch": 3.7212460063897765, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0425, + "step": 4659 + }, + { + "epoch": 3.722044728434505, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.041, + "step": 4660 + }, + { + "epoch": 3.722843450479233, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0469, + "step": 4661 + }, + { + "epoch": 3.7236421725239617, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0435, + "step": 4662 + }, + { + "epoch": 3.72444089456869, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0546, + "step": 4663 + }, + { + "epoch": 3.7252396166134183, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0422, + "step": 4664 + }, + { + "epoch": 3.726038338658147, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.042, + "step": 4665 + }, + { + "epoch": 3.7268370607028753, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 4666 + }, + { + "epoch": 3.727635782747604, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 4667 + }, + { + "epoch": 3.7284345047923324, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 4668 + }, + { + "epoch": 3.729233226837061, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0422, + "step": 4669 + }, + { + "epoch": 3.730031948881789, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 4670 + }, + { + "epoch": 3.7308306709265175, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0455, + "step": 4671 + }, + { + "epoch": 3.731629392971246, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0384, + "step": 4672 + }, + { + "epoch": 3.7324281150159746, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 4673 + }, + { + "epoch": 3.7332268370607027, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0447, + "step": 4674 + }, + { + "epoch": 3.734025559105431, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.046, + "step": 4675 + }, + { + "epoch": 3.7348242811501597, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.041, + "step": 4676 + }, + { + "epoch": 3.7356230031948883, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 4677 + }, + { + "epoch": 3.736421725239617, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0421, + "step": 4678 + }, + { + "epoch": 3.737220447284345, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 4679 + }, + { + "epoch": 3.7380191693290734, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0396, + "step": 4680 + }, + { + "epoch": 3.738817891373802, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0397, + "step": 4681 + }, + { + "epoch": 3.7396166134185305, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 4682 + }, + { + "epoch": 3.7404153354632586, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 4683 + }, + { + "epoch": 3.741214057507987, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0437, + "step": 4684 + }, + { + "epoch": 3.7420127795527156, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0449, + "step": 4685 + }, + { + "epoch": 3.742811501597444, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0393, + "step": 4686 + }, + { + "epoch": 3.7436102236421727, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 4687 + }, + { + "epoch": 3.744408945686901, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 4688 + }, + { + "epoch": 3.7452076677316293, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0377, + "step": 4689 + }, + { + "epoch": 3.746006389776358, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0404, + "step": 4690 + }, + { + "epoch": 3.7468051118210863, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 4691 + }, + { + "epoch": 3.747603833865815, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0472, + "step": 4692 + }, + { + "epoch": 3.748402555910543, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 4693 + }, + { + "epoch": 3.7492012779552715, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 4694 + }, + { + "epoch": 3.75, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0446, + "step": 4695 + }, + { + "epoch": 3.7507987220447285, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 4696 + }, + { + "epoch": 3.751597444089457, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0431, + "step": 4697 + }, + { + "epoch": 3.752396166134185, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0451, + "step": 4698 + }, + { + "epoch": 3.7531948881789137, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 4699 + }, + { + "epoch": 3.753993610223642, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 4700 + }, + { + "epoch": 3.7547923322683707, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 4701 + }, + { + "epoch": 3.755591054313099, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 4702 + }, + { + "epoch": 3.7563897763578273, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0417, + "step": 4703 + }, + { + "epoch": 3.757188498402556, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 4704 + }, + { + "epoch": 3.7579872204472844, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0588, + "step": 4705 + }, + { + "epoch": 3.758785942492013, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.041, + "step": 4706 + }, + { + "epoch": 3.7595846645367414, + "grad_norm": 0.1630859375, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 4707 + }, + { + "epoch": 3.7603833865814695, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 4708 + }, + { + "epoch": 3.761182108626198, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0434, + "step": 4709 + }, + { + "epoch": 3.7619808306709266, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0567, + "step": 4710 + }, + { + "epoch": 3.762779552715655, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0458, + "step": 4711 + }, + { + "epoch": 3.763578274760383, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 4712 + }, + { + "epoch": 3.7643769968051117, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0464, + "step": 4713 + }, + { + "epoch": 3.7651757188498403, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0403, + "step": 4714 + }, + { + "epoch": 3.765974440894569, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 4715 + }, + { + "epoch": 3.7667731629392973, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 4716 + }, + { + "epoch": 3.7675718849840254, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 4717 + }, + { + "epoch": 3.768370607028754, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 4718 + }, + { + "epoch": 3.7691693290734825, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0441, + "step": 4719 + }, + { + "epoch": 3.769968051118211, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 4720 + }, + { + "epoch": 3.770766773162939, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 4721 + }, + { + "epoch": 3.7715654952076676, + "grad_norm": 0.18359375, + "learning_rate": 0.0005, + "loss": 1.0445, + "step": 4722 + }, + { + "epoch": 3.772364217252396, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 4723 + }, + { + "epoch": 3.7731629392971247, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0606, + "step": 4724 + }, + { + "epoch": 3.773961661341853, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 4725 + }, + { + "epoch": 3.7747603833865817, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.043, + "step": 4726 + }, + { + "epoch": 3.77555910543131, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.038, + "step": 4727 + }, + { + "epoch": 3.7763578274760383, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0466, + "step": 4728 + }, + { + "epoch": 3.777156549520767, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 4729 + }, + { + "epoch": 3.777955271565495, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 4730 + }, + { + "epoch": 3.7787539936102235, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0432, + "step": 4731 + }, + { + "epoch": 3.779552715654952, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 4732 + }, + { + "epoch": 3.7803514376996805, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.046, + "step": 4733 + }, + { + "epoch": 3.781150159744409, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.043, + "step": 4734 + }, + { + "epoch": 3.7819488817891376, + "grad_norm": 0.1611328125, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 4735 + }, + { + "epoch": 3.7827476038338657, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 4736 + }, + { + "epoch": 3.783546325878594, + "grad_norm": 0.1826171875, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 4737 + }, + { + "epoch": 3.7843450479233227, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 4738 + }, + { + "epoch": 3.7851437699680512, + "grad_norm": 0.1953125, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 4739 + }, + { + "epoch": 3.7859424920127793, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 4740 + }, + { + "epoch": 3.786741214057508, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.0451, + "step": 4741 + }, + { + "epoch": 3.7875399361022364, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0467, + "step": 4742 + }, + { + "epoch": 3.788338658146965, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 4743 + }, + { + "epoch": 3.7891373801916934, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0457, + "step": 4744 + }, + { + "epoch": 3.789936102236422, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 4745 + }, + { + "epoch": 3.79073482428115, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 4746 + }, + { + "epoch": 3.7915335463258786, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0446, + "step": 4747 + }, + { + "epoch": 3.792332268370607, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0413, + "step": 4748 + }, + { + "epoch": 3.793130990415335, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 4749 + }, + { + "epoch": 3.7939297124600637, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0482, + "step": 4750 + }, + { + "epoch": 3.7947284345047922, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 4751 + }, + { + "epoch": 3.7955271565495208, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 4752 + }, + { + "epoch": 3.7963258785942493, + "grad_norm": 0.2197265625, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 4753 + }, + { + "epoch": 3.797124600638978, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 4754 + }, + { + "epoch": 3.797923322683706, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 4755 + }, + { + "epoch": 3.7987220447284344, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0429, + "step": 4756 + }, + { + "epoch": 3.799520766773163, + "grad_norm": 0.1787109375, + "learning_rate": 0.0005, + "loss": 1.0416, + "step": 4757 + }, + { + "epoch": 3.8003194888178915, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 4758 + }, + { + "epoch": 3.8011182108626196, + "grad_norm": 0.189453125, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 4759 + }, + { + "epoch": 3.801916932907348, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0385, + "step": 4760 + }, + { + "epoch": 3.8027156549520766, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 4761 + }, + { + "epoch": 3.803514376996805, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.0434, + "step": 4762 + }, + { + "epoch": 3.8043130990415337, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 4763 + }, + { + "epoch": 3.8051118210862622, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 4764 + }, + { + "epoch": 3.8059105431309903, + "grad_norm": 0.162109375, + "learning_rate": 0.0005, + "loss": 1.0422, + "step": 4765 + }, + { + "epoch": 3.806709265175719, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0412, + "step": 4766 + }, + { + "epoch": 3.8075079872204474, + "grad_norm": 0.185546875, + "learning_rate": 0.0005, + "loss": 1.0401, + "step": 4767 + }, + { + "epoch": 3.8083067092651754, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.039, + "step": 4768 + }, + { + "epoch": 3.809105431309904, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 4769 + }, + { + "epoch": 3.8099041533546325, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0431, + "step": 4770 + }, + { + "epoch": 3.810702875399361, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 4771 + }, + { + "epoch": 3.8115015974440896, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 4772 + }, + { + "epoch": 3.812300319488818, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0457, + "step": 4773 + }, + { + "epoch": 3.813099041533546, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 4774 + }, + { + "epoch": 3.8138977635782747, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0437, + "step": 4775 + }, + { + "epoch": 3.8146964856230032, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 4776 + }, + { + "epoch": 3.8154952076677318, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0413, + "step": 4777 + }, + { + "epoch": 3.81629392971246, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0414, + "step": 4778 + }, + { + "epoch": 3.8170926517571884, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 4779 + }, + { + "epoch": 3.817891373801917, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 4780 + }, + { + "epoch": 3.8186900958466454, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 4781 + }, + { + "epoch": 3.819488817891374, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0439, + "step": 4782 + }, + { + "epoch": 3.8202875399361025, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 4783 + }, + { + "epoch": 3.8210862619808306, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 4784 + }, + { + "epoch": 3.821884984025559, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0409, + "step": 4785 + }, + { + "epoch": 3.8226837060702876, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 4786 + }, + { + "epoch": 3.8234824281150157, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 4787 + }, + { + "epoch": 3.8242811501597442, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 4788 + }, + { + "epoch": 3.8250798722044728, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.053, + "step": 4789 + }, + { + "epoch": 3.8258785942492013, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0428, + "step": 4790 + }, + { + "epoch": 3.82667731629393, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 4791 + }, + { + "epoch": 3.8274760383386583, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 4792 + }, + { + "epoch": 3.8282747603833864, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0443, + "step": 4793 + }, + { + "epoch": 3.829073482428115, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 4794 + }, + { + "epoch": 3.8298722044728435, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0467, + "step": 4795 + }, + { + "epoch": 3.830670926517572, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 4796 + }, + { + "epoch": 3.8314696485623, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0431, + "step": 4797 + }, + { + "epoch": 3.8322683706070286, + "grad_norm": 0.224609375, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 4798 + }, + { + "epoch": 3.833067092651757, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 4799 + }, + { + "epoch": 3.8338658146964857, + "grad_norm": 0.169921875, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 4800 + }, + { + "epoch": 3.834664536741214, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 4801 + }, + { + "epoch": 3.8354632587859427, + "grad_norm": 0.1640625, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 4802 + }, + { + "epoch": 3.836261980830671, + "grad_norm": 0.1728515625, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 4803 + }, + { + "epoch": 3.8370607028753994, + "grad_norm": 0.25, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 4804 + }, + { + "epoch": 3.837859424920128, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 4805 + }, + { + "epoch": 3.838658146964856, + "grad_norm": 0.208984375, + "learning_rate": 0.0005, + "loss": 1.0439, + "step": 4806 + }, + { + "epoch": 3.8394568690095845, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0472, + "step": 4807 + }, + { + "epoch": 3.840255591054313, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.045, + "step": 4808 + }, + { + "epoch": 3.8410543130990416, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0451, + "step": 4809 + }, + { + "epoch": 3.84185303514377, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 4810 + }, + { + "epoch": 3.8426517571884986, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 4811 + }, + { + "epoch": 3.8434504792332267, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 4812 + }, + { + "epoch": 3.844249201277955, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 4813 + }, + { + "epoch": 3.8450479233226837, + "grad_norm": 0.1904296875, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 4814 + }, + { + "epoch": 3.8458466453674123, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0434, + "step": 4815 + }, + { + "epoch": 3.8466453674121404, + "grad_norm": 0.1572265625, + "learning_rate": 0.0005, + "loss": 1.0435, + "step": 4816 + }, + { + "epoch": 3.847444089456869, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 4817 + }, + { + "epoch": 3.8482428115015974, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.0472, + "step": 4818 + }, + { + "epoch": 3.849041533546326, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0435, + "step": 4819 + }, + { + "epoch": 3.8498402555910545, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 4820 + }, + { + "epoch": 3.850638977635783, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 4821 + }, + { + "epoch": 3.851437699680511, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 4822 + }, + { + "epoch": 3.8522364217252396, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0435, + "step": 4823 + }, + { + "epoch": 3.853035143769968, + "grad_norm": 0.2060546875, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 4824 + }, + { + "epoch": 3.8538338658146962, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 4825 + }, + { + "epoch": 3.8546325878594248, + "grad_norm": 0.1689453125, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 4826 + }, + { + "epoch": 3.8554313099041533, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 4827 + }, + { + "epoch": 3.856230031948882, + "grad_norm": 0.169921875, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 4828 + }, + { + "epoch": 3.8570287539936103, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0451, + "step": 4829 + }, + { + "epoch": 3.857827476038339, + "grad_norm": 0.166015625, + "learning_rate": 0.0005, + "loss": 1.0419, + "step": 4830 + }, + { + "epoch": 3.858626198083067, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 4831 + }, + { + "epoch": 3.8594249201277955, + "grad_norm": 0.232421875, + "learning_rate": 0.0005, + "loss": 1.0444, + "step": 4832 + }, + { + "epoch": 3.860223642172524, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0424, + "step": 4833 + }, + { + "epoch": 3.8610223642172525, + "grad_norm": 0.19140625, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 4834 + }, + { + "epoch": 3.8618210862619806, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 4835 + }, + { + "epoch": 3.862619808306709, + "grad_norm": 0.212890625, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 4836 + }, + { + "epoch": 3.8634185303514377, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0426, + "step": 4837 + }, + { + "epoch": 3.864217252396166, + "grad_norm": 0.2333984375, + "learning_rate": 0.0005, + "loss": 1.0437, + "step": 4838 + }, + { + "epoch": 3.8650159744408947, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0442, + "step": 4839 + }, + { + "epoch": 3.8658146964856233, + "grad_norm": 0.1533203125, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 4840 + }, + { + "epoch": 3.8666134185303513, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 4841 + }, + { + "epoch": 3.86741214057508, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.0544, + "step": 4842 + }, + { + "epoch": 3.8682108626198084, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0416, + "step": 4843 + }, + { + "epoch": 3.8690095846645365, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 4844 + }, + { + "epoch": 3.869808306709265, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 4845 + }, + { + "epoch": 3.8706070287539935, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.0472, + "step": 4846 + }, + { + "epoch": 3.871405750798722, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0451, + "step": 4847 + }, + { + "epoch": 3.8722044728434506, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0437, + "step": 4848 + }, + { + "epoch": 3.873003194888179, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0415, + "step": 4849 + }, + { + "epoch": 3.873801916932907, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0563, + "step": 4850 + }, + { + "epoch": 3.8746006389776357, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.043, + "step": 4851 + }, + { + "epoch": 3.8753993610223643, + "grad_norm": 0.1845703125, + "learning_rate": 0.0005, + "loss": 1.0414, + "step": 4852 + }, + { + "epoch": 3.876198083067093, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 4853 + }, + { + "epoch": 3.876996805111821, + "grad_norm": 0.201171875, + "learning_rate": 0.0005, + "loss": 1.0406, + "step": 4854 + }, + { + "epoch": 3.8777955271565494, + "grad_norm": 0.2216796875, + "learning_rate": 0.0005, + "loss": 1.0482, + "step": 4855 + }, + { + "epoch": 3.878594249201278, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.0433, + "step": 4856 + }, + { + "epoch": 3.8793929712460065, + "grad_norm": 0.29296875, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 4857 + }, + { + "epoch": 3.880191693290735, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 4858 + }, + { + "epoch": 3.8809904153354635, + "grad_norm": 0.2138671875, + "learning_rate": 0.0005, + "loss": 1.0448, + "step": 4859 + }, + { + "epoch": 3.8817891373801916, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.038, + "step": 4860 + }, + { + "epoch": 3.88258785942492, + "grad_norm": 0.1533203125, + "learning_rate": 0.0005, + "loss": 1.0452, + "step": 4861 + }, + { + "epoch": 3.8833865814696487, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0568, + "step": 4862 + }, + { + "epoch": 3.8841853035143767, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0458, + "step": 4863 + }, + { + "epoch": 3.8849840255591053, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0407, + "step": 4864 + }, + { + "epoch": 3.885782747603834, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 4865 + }, + { + "epoch": 3.8865814696485623, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 4866 + }, + { + "epoch": 3.887380191693291, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 4867 + }, + { + "epoch": 3.8881789137380194, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0439, + "step": 4868 + }, + { + "epoch": 3.8889776357827475, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0433, + "step": 4869 + }, + { + "epoch": 3.889776357827476, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 4870 + }, + { + "epoch": 3.8905750798722045, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0431, + "step": 4871 + }, + { + "epoch": 3.891373801916933, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 4872 + }, + { + "epoch": 3.892172523961661, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0465, + "step": 4873 + }, + { + "epoch": 3.8929712460063897, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 4874 + }, + { + "epoch": 3.893769968051118, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.045, + "step": 4875 + }, + { + "epoch": 3.8945686900958467, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 4876 + }, + { + "epoch": 3.8953674121405752, + "grad_norm": 0.28515625, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 4877 + }, + { + "epoch": 3.8961661341853038, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0459, + "step": 4878 + }, + { + "epoch": 3.896964856230032, + "grad_norm": 0.1689453125, + "learning_rate": 0.0005, + "loss": 1.0409, + "step": 4879 + }, + { + "epoch": 3.8977635782747604, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0402, + "step": 4880 + }, + { + "epoch": 3.898562300319489, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 4881 + }, + { + "epoch": 3.899361022364217, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 4882 + }, + { + "epoch": 3.9001597444089455, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0426, + "step": 4883 + }, + { + "epoch": 3.900958466453674, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0435, + "step": 4884 + }, + { + "epoch": 3.9017571884984026, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0433, + "step": 4885 + }, + { + "epoch": 3.902555910543131, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.039, + "step": 4886 + }, + { + "epoch": 3.9033546325878596, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0448, + "step": 4887 + }, + { + "epoch": 3.9041533546325877, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0393, + "step": 4888 + }, + { + "epoch": 3.9049520766773163, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0403, + "step": 4889 + }, + { + "epoch": 3.905750798722045, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 4890 + }, + { + "epoch": 3.9065495207667733, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0413, + "step": 4891 + }, + { + "epoch": 3.9073482428115014, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0425, + "step": 4892 + }, + { + "epoch": 3.90814696485623, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 4893 + }, + { + "epoch": 3.9089456869009584, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.043, + "step": 4894 + }, + { + "epoch": 3.909744408945687, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0378, + "step": 4895 + }, + { + "epoch": 3.9105431309904155, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0455, + "step": 4896 + }, + { + "epoch": 3.911341853035144, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0428, + "step": 4897 + }, + { + "epoch": 3.912140575079872, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 4898 + }, + { + "epoch": 3.9129392971246006, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0408, + "step": 4899 + }, + { + "epoch": 3.913738019169329, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0394, + "step": 4900 + }, + { + "epoch": 3.9145367412140573, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0583, + "step": 4901 + }, + { + "epoch": 3.915335463258786, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 4902 + }, + { + "epoch": 3.9161341853035143, + "grad_norm": 0.17578125, + "learning_rate": 0.0005, + "loss": 1.0432, + "step": 4903 + }, + { + "epoch": 3.916932907348243, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 4904 + }, + { + "epoch": 3.9177316293929714, + "grad_norm": 0.26953125, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 4905 + }, + { + "epoch": 3.9185303514377, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0615, + "step": 4906 + }, + { + "epoch": 3.919329073482428, + "grad_norm": 0.271484375, + "learning_rate": 0.0005, + "loss": 1.0444, + "step": 4907 + }, + { + "epoch": 3.9201277955271565, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0453, + "step": 4908 + }, + { + "epoch": 3.920926517571885, + "grad_norm": 0.17578125, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 4909 + }, + { + "epoch": 3.9217252396166136, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0408, + "step": 4910 + }, + { + "epoch": 3.9225239616613417, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 4911 + }, + { + "epoch": 3.92332268370607, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 4912 + }, + { + "epoch": 3.9241214057507987, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 4913 + }, + { + "epoch": 3.9249201277955272, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0415, + "step": 4914 + }, + { + "epoch": 3.9257188498402558, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 4915 + }, + { + "epoch": 3.9265175718849843, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 4916 + }, + { + "epoch": 3.9273162939297124, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.04, + "step": 4917 + }, + { + "epoch": 3.928115015974441, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0437, + "step": 4918 + }, + { + "epoch": 3.9289137380191694, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0442, + "step": 4919 + }, + { + "epoch": 3.9297124600638975, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0464, + "step": 4920 + }, + { + "epoch": 3.930511182108626, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0428, + "step": 4921 + }, + { + "epoch": 3.9313099041533546, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0416, + "step": 4922 + }, + { + "epoch": 3.932108626198083, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0452, + "step": 4923 + }, + { + "epoch": 3.9329073482428116, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 4924 + }, + { + "epoch": 3.93370607028754, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.0449, + "step": 4925 + }, + { + "epoch": 3.9345047923322682, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0403, + "step": 4926 + }, + { + "epoch": 3.9353035143769968, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 4927 + }, + { + "epoch": 3.9361022364217253, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0434, + "step": 4928 + }, + { + "epoch": 3.936900958466454, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0406, + "step": 4929 + }, + { + "epoch": 3.937699680511182, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 4930 + }, + { + "epoch": 3.9384984025559104, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 4931 + }, + { + "epoch": 3.939297124600639, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 4932 + }, + { + "epoch": 3.9400958466453675, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0402, + "step": 4933 + }, + { + "epoch": 3.940894568690096, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0408, + "step": 4934 + }, + { + "epoch": 3.9416932907348246, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0437, + "step": 4935 + }, + { + "epoch": 3.9424920127795526, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 4936 + }, + { + "epoch": 3.943290734824281, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 4937 + }, + { + "epoch": 3.9440894568690097, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0408, + "step": 4938 + }, + { + "epoch": 3.9448881789137378, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.0404, + "step": 4939 + }, + { + "epoch": 3.9456869009584663, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 4940 + }, + { + "epoch": 3.946485623003195, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0401, + "step": 4941 + }, + { + "epoch": 3.9472843450479234, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 4942 + }, + { + "epoch": 3.948083067092652, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0453, + "step": 4943 + }, + { + "epoch": 3.9488817891373804, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 4944 + }, + { + "epoch": 3.9496805111821085, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0395, + "step": 4945 + }, + { + "epoch": 3.950479233226837, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.043, + "step": 4946 + }, + { + "epoch": 3.9512779552715656, + "grad_norm": 0.162109375, + "learning_rate": 0.0005, + "loss": 1.0391, + "step": 4947 + }, + { + "epoch": 3.952076677316294, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0389, + "step": 4948 + }, + { + "epoch": 3.952875399361022, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.039, + "step": 4949 + }, + { + "epoch": 3.9536741214057507, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 4950 + }, + { + "epoch": 3.9544728434504792, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0466, + "step": 4951 + }, + { + "epoch": 3.9552715654952078, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 4952 + }, + { + "epoch": 3.9560702875399363, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 4953 + }, + { + "epoch": 3.956869009584665, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 4954 + }, + { + "epoch": 3.957667731629393, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 4955 + }, + { + "epoch": 3.9584664536741214, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 4956 + }, + { + "epoch": 3.95926517571885, + "grad_norm": 0.2353515625, + "learning_rate": 0.0005, + "loss": 1.0405, + "step": 4957 + }, + { + "epoch": 3.960063897763578, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 4958 + }, + { + "epoch": 3.9608626198083066, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0435, + "step": 4959 + }, + { + "epoch": 3.961661341853035, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0444, + "step": 4960 + }, + { + "epoch": 3.9624600638977636, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0405, + "step": 4961 + }, + { + "epoch": 3.963258785942492, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0425, + "step": 4962 + }, + { + "epoch": 3.9640575079872207, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 4963 + }, + { + "epoch": 3.9648562300319488, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0409, + "step": 4964 + }, + { + "epoch": 3.9656549520766773, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0385, + "step": 4965 + }, + { + "epoch": 3.966453674121406, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0446, + "step": 4966 + }, + { + "epoch": 3.9672523961661343, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 4967 + }, + { + "epoch": 3.9680511182108624, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0436, + "step": 4968 + }, + { + "epoch": 3.968849840255591, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0601, + "step": 4969 + }, + { + "epoch": 3.9696485623003195, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0388, + "step": 4970 + }, + { + "epoch": 3.970447284345048, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 4971 + }, + { + "epoch": 3.9712460063897765, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 4972 + }, + { + "epoch": 3.972044728434505, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0407, + "step": 4973 + }, + { + "epoch": 3.972843450479233, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 4974 + }, + { + "epoch": 3.9736421725239617, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.037, + "step": 4975 + }, + { + "epoch": 3.97444089456869, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 4976 + }, + { + "epoch": 3.9752396166134183, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0525, + "step": 4977 + }, + { + "epoch": 3.976038338658147, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0444, + "step": 4978 + }, + { + "epoch": 3.9768370607028753, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0429, + "step": 4979 + }, + { + "epoch": 3.977635782747604, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 4980 + }, + { + "epoch": 3.9784345047923324, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 4981 + }, + { + "epoch": 3.979233226837061, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0444, + "step": 4982 + }, + { + "epoch": 3.980031948881789, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0401, + "step": 4983 + }, + { + "epoch": 3.9808306709265175, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0395, + "step": 4984 + }, + { + "epoch": 3.981629392971246, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0419, + "step": 4985 + }, + { + "epoch": 3.9824281150159746, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 4986 + }, + { + "epoch": 3.9832268370607027, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 4987 + }, + { + "epoch": 3.984025559105431, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0429, + "step": 4988 + }, + { + "epoch": 3.9848242811501597, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 4989 + }, + { + "epoch": 3.9856230031948883, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0452, + "step": 4990 + }, + { + "epoch": 3.986421725239617, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 4991 + }, + { + "epoch": 3.987220447284345, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 4992 + }, + { + "epoch": 3.9880191693290734, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 4993 + }, + { + "epoch": 3.988817891373802, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0451, + "step": 4994 + }, + { + "epoch": 3.9896166134185305, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0434, + "step": 4995 + }, + { + "epoch": 3.9904153354632586, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 4996 + }, + { + "epoch": 3.991214057507987, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0561, + "step": 4997 + }, + { + "epoch": 3.9920127795527156, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0417, + "step": 4998 + }, + { + "epoch": 3.992811501597444, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 4999 + }, + { + "epoch": 3.9936102236421727, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 5000 + }, + { + "epoch": 3.994408945686901, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0424, + "step": 5001 + }, + { + "epoch": 3.9952076677316293, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 5002 + }, + { + "epoch": 3.996006389776358, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0472, + "step": 5003 + }, + { + "epoch": 3.9968051118210863, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 5004 + }, + { + "epoch": 3.997603833865815, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.046, + "step": 5005 + }, + { + "epoch": 3.998402555910543, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0448, + "step": 5006 + }, + { + "epoch": 3.9992012779552715, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0453, + "step": 5007 + }, + { + "epoch": 4.0, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0407, + "step": 5008 + }, + { + "epoch": 4.0007987220447285, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0422, + "step": 5009 + }, + { + "epoch": 4.001597444089457, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0464, + "step": 5010 + }, + { + "epoch": 4.002396166134186, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0425, + "step": 5011 + }, + { + "epoch": 4.003194888178914, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 5012 + }, + { + "epoch": 4.003993610223642, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 5013 + }, + { + "epoch": 4.00479233226837, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 5014 + }, + { + "epoch": 4.005591054313099, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 5015 + }, + { + "epoch": 4.006389776357827, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.042, + "step": 5016 + }, + { + "epoch": 4.007188498402556, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 5017 + }, + { + "epoch": 4.007987220447284, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0412, + "step": 5018 + }, + { + "epoch": 4.008785942492013, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0396, + "step": 5019 + }, + { + "epoch": 4.0095846645367414, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.041, + "step": 5020 + }, + { + "epoch": 4.01038338658147, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0472, + "step": 5021 + }, + { + "epoch": 4.0111821086261985, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0447, + "step": 5022 + }, + { + "epoch": 4.011980830670926, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0404, + "step": 5023 + }, + { + "epoch": 4.012779552715655, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 5024 + }, + { + "epoch": 4.013578274760383, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0515, + "step": 5025 + }, + { + "epoch": 4.014376996805112, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0456, + "step": 5026 + }, + { + "epoch": 4.01517571884984, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 5027 + }, + { + "epoch": 4.015974440894569, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0444, + "step": 5028 + }, + { + "epoch": 4.016773162939297, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0488, + "step": 5029 + }, + { + "epoch": 4.017571884984026, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 5030 + }, + { + "epoch": 4.018370607028754, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0457, + "step": 5031 + }, + { + "epoch": 4.019169329073482, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 5032 + }, + { + "epoch": 4.0199680511182105, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0425, + "step": 5033 + }, + { + "epoch": 4.020766773162939, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0405, + "step": 5034 + }, + { + "epoch": 4.021565495207668, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 5035 + }, + { + "epoch": 4.022364217252396, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.038, + "step": 5036 + }, + { + "epoch": 4.023162939297125, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 5037 + }, + { + "epoch": 4.023961661341853, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 5038 + }, + { + "epoch": 4.024760383386582, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.045, + "step": 5039 + }, + { + "epoch": 4.02555910543131, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 5040 + }, + { + "epoch": 4.026357827476039, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0451, + "step": 5041 + }, + { + "epoch": 4.027156549520766, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 5042 + }, + { + "epoch": 4.027955271565495, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 5043 + }, + { + "epoch": 4.0287539936102235, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0429, + "step": 5044 + }, + { + "epoch": 4.029552715654952, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0458, + "step": 5045 + }, + { + "epoch": 4.0303514376996805, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0397, + "step": 5046 + }, + { + "epoch": 4.031150159744409, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0394, + "step": 5047 + }, + { + "epoch": 4.031948881789138, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0431, + "step": 5048 + }, + { + "epoch": 4.032747603833866, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 5049 + }, + { + "epoch": 4.033546325878595, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.0412, + "step": 5050 + }, + { + "epoch": 4.034345047923322, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 5051 + }, + { + "epoch": 4.035143769968051, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 5052 + }, + { + "epoch": 4.035942492012779, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0416, + "step": 5053 + }, + { + "epoch": 4.036741214057508, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 5054 + }, + { + "epoch": 4.037539936102236, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0436, + "step": 5055 + }, + { + "epoch": 4.038338658146965, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0415, + "step": 5056 + }, + { + "epoch": 4.039137380191693, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0441, + "step": 5057 + }, + { + "epoch": 4.039936102236422, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0399, + "step": 5058 + }, + { + "epoch": 4.0407348242811505, + "grad_norm": 0.189453125, + "learning_rate": 0.0005, + "loss": 1.0384, + "step": 5059 + }, + { + "epoch": 4.041533546325879, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0407, + "step": 5060 + }, + { + "epoch": 4.042332268370607, + "grad_norm": 0.205078125, + "learning_rate": 0.0005, + "loss": 1.0464, + "step": 5061 + }, + { + "epoch": 4.043130990415335, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.038, + "step": 5062 + }, + { + "epoch": 4.043929712460064, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.0384, + "step": 5063 + }, + { + "epoch": 4.044728434504792, + "grad_norm": 0.197265625, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 5064 + }, + { + "epoch": 4.045527156549521, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0404, + "step": 5065 + }, + { + "epoch": 4.046325878594249, + "grad_norm": 0.2138671875, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 5066 + }, + { + "epoch": 4.047124600638978, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 5067 + }, + { + "epoch": 4.047923322683706, + "grad_norm": 0.2333984375, + "learning_rate": 0.0005, + "loss": 1.0456, + "step": 5068 + }, + { + "epoch": 4.048722044728435, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0472, + "step": 5069 + }, + { + "epoch": 4.0495207667731625, + "grad_norm": 0.30859375, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 5070 + }, + { + "epoch": 4.050319488817891, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 5071 + }, + { + "epoch": 4.05111821086262, + "grad_norm": 0.1630859375, + "learning_rate": 0.0005, + "loss": 1.0579, + "step": 5072 + }, + { + "epoch": 4.051916932907348, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 5073 + }, + { + "epoch": 4.052715654952077, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 5074 + }, + { + "epoch": 4.053514376996805, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 5075 + }, + { + "epoch": 4.054313099041534, + "grad_norm": 0.181640625, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 5076 + }, + { + "epoch": 4.055111821086262, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0445, + "step": 5077 + }, + { + "epoch": 4.055910543130991, + "grad_norm": 0.1943359375, + "learning_rate": 0.0005, + "loss": 1.0424, + "step": 5078 + }, + { + "epoch": 4.056709265175719, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0378, + "step": 5079 + }, + { + "epoch": 4.057507987220447, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 5080 + }, + { + "epoch": 4.0583067092651754, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0403, + "step": 5081 + }, + { + "epoch": 4.059105431309904, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0379, + "step": 5082 + }, + { + "epoch": 4.0599041533546325, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 5083 + }, + { + "epoch": 4.060702875399361, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0426, + "step": 5084 + }, + { + "epoch": 4.06150159744409, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0399, + "step": 5085 + }, + { + "epoch": 4.062300319488818, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0452, + "step": 5086 + }, + { + "epoch": 4.063099041533547, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0395, + "step": 5087 + }, + { + "epoch": 4.063897763578275, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0533, + "step": 5088 + }, + { + "epoch": 4.064696485623003, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0441, + "step": 5089 + }, + { + "epoch": 4.065495207667731, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 5090 + }, + { + "epoch": 4.06629392971246, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0421, + "step": 5091 + }, + { + "epoch": 4.067092651757188, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.046, + "step": 5092 + }, + { + "epoch": 4.067891373801917, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0369, + "step": 5093 + }, + { + "epoch": 4.068690095846645, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0456, + "step": 5094 + }, + { + "epoch": 4.069488817891374, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0417, + "step": 5095 + }, + { + "epoch": 4.0702875399361025, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0415, + "step": 5096 + }, + { + "epoch": 4.071086261980831, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0458, + "step": 5097 + }, + { + "epoch": 4.0718849840255595, + "grad_norm": 0.166015625, + "learning_rate": 0.0005, + "loss": 1.0572, + "step": 5098 + }, + { + "epoch": 4.072683706070287, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0467, + "step": 5099 + }, + { + "epoch": 4.073482428115016, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0449, + "step": 5100 + }, + { + "epoch": 4.074281150159744, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 5101 + }, + { + "epoch": 4.075079872204473, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0414, + "step": 5102 + }, + { + "epoch": 4.075878594249201, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 5103 + }, + { + "epoch": 4.07667731629393, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 5104 + }, + { + "epoch": 4.077476038338658, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 5105 + }, + { + "epoch": 4.078274760383387, + "grad_norm": 0.2021484375, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 5106 + }, + { + "epoch": 4.079073482428115, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0388, + "step": 5107 + }, + { + "epoch": 4.079872204472843, + "grad_norm": 0.1640625, + "learning_rate": 0.0005, + "loss": 1.0369, + "step": 5108 + }, + { + "epoch": 4.080670926517572, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0447, + "step": 5109 + }, + { + "epoch": 4.0814696485623, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0428, + "step": 5110 + }, + { + "epoch": 4.082268370607029, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 5111 + }, + { + "epoch": 4.083067092651757, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0447, + "step": 5112 + }, + { + "epoch": 4.083865814696486, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0397, + "step": 5113 + }, + { + "epoch": 4.084664536741214, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 5114 + }, + { + "epoch": 4.085463258785943, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 5115 + }, + { + "epoch": 4.086261980830671, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 5116 + }, + { + "epoch": 4.0870607028754, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0454, + "step": 5117 + }, + { + "epoch": 4.087859424920127, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 5118 + }, + { + "epoch": 4.088658146964856, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0465, + "step": 5119 + }, + { + "epoch": 4.0894568690095845, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 5120 + }, + { + "epoch": 4.090255591054313, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0437, + "step": 5121 + }, + { + "epoch": 4.0910543130990416, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0373, + "step": 5122 + }, + { + "epoch": 4.09185303514377, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0379, + "step": 5123 + }, + { + "epoch": 4.092651757188499, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.048, + "step": 5124 + }, + { + "epoch": 4.093450479233227, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 5125 + }, + { + "epoch": 4.094249201277956, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 5126 + }, + { + "epoch": 4.095047923322683, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0427, + "step": 5127 + }, + { + "epoch": 4.095846645367412, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0429, + "step": 5128 + }, + { + "epoch": 4.09664536741214, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 5129 + }, + { + "epoch": 4.097444089456869, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.038, + "step": 5130 + }, + { + "epoch": 4.098242811501597, + "grad_norm": 0.1611328125, + "learning_rate": 0.0005, + "loss": 1.0436, + "step": 5131 + }, + { + "epoch": 4.099041533546326, + "grad_norm": 0.1689453125, + "learning_rate": 0.0005, + "loss": 1.0476, + "step": 5132 + }, + { + "epoch": 4.0998402555910545, + "grad_norm": 0.1787109375, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 5133 + }, + { + "epoch": 4.100638977635783, + "grad_norm": 0.244140625, + "learning_rate": 0.0005, + "loss": 1.042, + "step": 5134 + }, + { + "epoch": 4.1014376996805115, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0444, + "step": 5135 + }, + { + "epoch": 4.102236421725239, + "grad_norm": 0.314453125, + "learning_rate": 0.0005, + "loss": 1.0397, + "step": 5136 + }, + { + "epoch": 4.103035143769968, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 5137 + }, + { + "epoch": 4.103833865814696, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.0454, + "step": 5138 + }, + { + "epoch": 4.104632587859425, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.0427, + "step": 5139 + }, + { + "epoch": 4.105431309904153, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0409, + "step": 5140 + }, + { + "epoch": 4.106230031948882, + "grad_norm": 0.1865234375, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 5141 + }, + { + "epoch": 4.10702875399361, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 5142 + }, + { + "epoch": 4.107827476038339, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0445, + "step": 5143 + }, + { + "epoch": 4.108626198083067, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 5144 + }, + { + "epoch": 4.109424920127796, + "grad_norm": 0.171875, + "learning_rate": 0.0005, + "loss": 1.0422, + "step": 5145 + }, + { + "epoch": 4.110223642172524, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 5146 + }, + { + "epoch": 4.111022364217252, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0431, + "step": 5147 + }, + { + "epoch": 4.111821086261981, + "grad_norm": 0.181640625, + "learning_rate": 0.0005, + "loss": 1.0396, + "step": 5148 + }, + { + "epoch": 4.112619808306709, + "grad_norm": 0.208984375, + "learning_rate": 0.0005, + "loss": 1.0399, + "step": 5149 + }, + { + "epoch": 4.113418530351438, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0376, + "step": 5150 + }, + { + "epoch": 4.114217252396166, + "grad_norm": 0.267578125, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 5151 + }, + { + "epoch": 4.115015974440895, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 5152 + }, + { + "epoch": 4.115814696485623, + "grad_norm": 0.423828125, + "learning_rate": 0.0005, + "loss": 1.0388, + "step": 5153 + }, + { + "epoch": 4.116613418530352, + "grad_norm": 0.1787109375, + "learning_rate": 0.0005, + "loss": 1.0384, + "step": 5154 + }, + { + "epoch": 4.11741214057508, + "grad_norm": 0.2373046875, + "learning_rate": 0.0005, + "loss": 1.0403, + "step": 5155 + }, + { + "epoch": 4.118210862619808, + "grad_norm": 0.201171875, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 5156 + }, + { + "epoch": 4.1190095846645365, + "grad_norm": 0.17578125, + "learning_rate": 0.0005, + "loss": 1.042, + "step": 5157 + }, + { + "epoch": 4.119808306709265, + "grad_norm": 0.30078125, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 5158 + }, + { + "epoch": 4.1206070287539935, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0439, + "step": 5159 + }, + { + "epoch": 4.121405750798722, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0375, + "step": 5160 + }, + { + "epoch": 4.122204472843451, + "grad_norm": 0.1572265625, + "learning_rate": 0.0005, + "loss": 1.0414, + "step": 5161 + }, + { + "epoch": 4.123003194888179, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0444, + "step": 5162 + }, + { + "epoch": 4.123801916932908, + "grad_norm": 0.1904296875, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 5163 + }, + { + "epoch": 4.124600638977636, + "grad_norm": 0.33984375, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 5164 + }, + { + "epoch": 4.125399361022364, + "grad_norm": 0.16015625, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 5165 + }, + { + "epoch": 4.126198083067092, + "grad_norm": 0.2353515625, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 5166 + }, + { + "epoch": 4.126996805111821, + "grad_norm": 0.30078125, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 5167 + }, + { + "epoch": 4.127795527156549, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 5168 + }, + { + "epoch": 4.128594249201278, + "grad_norm": 0.296875, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 5169 + }, + { + "epoch": 4.1293929712460065, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 5170 + }, + { + "epoch": 4.130191693290735, + "grad_norm": 0.1787109375, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 5171 + }, + { + "epoch": 4.1309904153354635, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 5172 + }, + { + "epoch": 4.131789137380192, + "grad_norm": 0.27734375, + "learning_rate": 0.0005, + "loss": 1.0422, + "step": 5173 + }, + { + "epoch": 4.13258785942492, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 5174 + }, + { + "epoch": 4.133386581469648, + "grad_norm": 0.23828125, + "learning_rate": 0.0005, + "loss": 1.0427, + "step": 5175 + }, + { + "epoch": 4.134185303514377, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 5176 + }, + { + "epoch": 4.134984025559105, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 5177 + }, + { + "epoch": 4.135782747603834, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0586, + "step": 5178 + }, + { + "epoch": 4.136581469648562, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0425, + "step": 5179 + }, + { + "epoch": 4.137380191693291, + "grad_norm": 0.259765625, + "learning_rate": 0.0005, + "loss": 1.0559, + "step": 5180 + }, + { + "epoch": 4.138178913738019, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.038, + "step": 5181 + }, + { + "epoch": 4.138977635782748, + "grad_norm": 0.314453125, + "learning_rate": 0.0005, + "loss": 1.0427, + "step": 5182 + }, + { + "epoch": 4.139776357827476, + "grad_norm": 0.1806640625, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 5183 + }, + { + "epoch": 4.140575079872204, + "grad_norm": 0.1796875, + "learning_rate": 0.0005, + "loss": 1.0434, + "step": 5184 + }, + { + "epoch": 4.141373801916933, + "grad_norm": 0.2275390625, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 5185 + }, + { + "epoch": 4.142172523961661, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0459, + "step": 5186 + }, + { + "epoch": 4.14297124600639, + "grad_norm": 0.3984375, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 5187 + }, + { + "epoch": 4.143769968051118, + "grad_norm": 0.16796875, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 5188 + }, + { + "epoch": 4.144568690095847, + "grad_norm": 0.265625, + "learning_rate": 0.0005, + "loss": 1.0427, + "step": 5189 + }, + { + "epoch": 4.145367412140575, + "grad_norm": 0.236328125, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 5190 + }, + { + "epoch": 4.146166134185304, + "grad_norm": 0.244140625, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 5191 + }, + { + "epoch": 4.146964856230032, + "grad_norm": 0.48046875, + "learning_rate": 0.0005, + "loss": 1.0391, + "step": 5192 + }, + { + "epoch": 4.147763578274761, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.0453, + "step": 5193 + }, + { + "epoch": 4.1485623003194885, + "grad_norm": 0.3359375, + "learning_rate": 0.0005, + "loss": 1.0372, + "step": 5194 + }, + { + "epoch": 4.149361022364217, + "grad_norm": 0.3203125, + "learning_rate": 0.0005, + "loss": 1.0409, + "step": 5195 + }, + { + "epoch": 4.1501597444089455, + "grad_norm": 0.18359375, + "learning_rate": 0.0005, + "loss": 1.0423, + "step": 5196 + }, + { + "epoch": 4.150958466453674, + "grad_norm": 0.439453125, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 5197 + }, + { + "epoch": 4.151757188498403, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0435, + "step": 5198 + }, + { + "epoch": 4.152555910543131, + "grad_norm": 0.3671875, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 5199 + }, + { + "epoch": 4.15335463258786, + "grad_norm": 0.185546875, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 5200 + }, + { + "epoch": 4.154153354632588, + "grad_norm": 0.224609375, + "learning_rate": 0.0005, + "loss": 1.0413, + "step": 5201 + }, + { + "epoch": 4.154952076677317, + "grad_norm": 0.248046875, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 5202 + }, + { + "epoch": 4.155750798722044, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.0453, + "step": 5203 + }, + { + "epoch": 4.156549520766773, + "grad_norm": 0.25390625, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 5204 + }, + { + "epoch": 4.157348242811501, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0442, + "step": 5205 + }, + { + "epoch": 4.15814696485623, + "grad_norm": 0.220703125, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 5206 + }, + { + "epoch": 4.1589456869009584, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 5207 + }, + { + "epoch": 4.159744408945687, + "grad_norm": 0.25, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 5208 + }, + { + "epoch": 4.1605431309904155, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 5209 + }, + { + "epoch": 4.161341853035144, + "grad_norm": 0.275390625, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 5210 + }, + { + "epoch": 4.162140575079873, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 5211 + }, + { + "epoch": 4.1629392971246, + "grad_norm": 0.328125, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 5212 + }, + { + "epoch": 4.163738019169329, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0417, + "step": 5213 + }, + { + "epoch": 4.164536741214057, + "grad_norm": 0.171875, + "learning_rate": 0.0005, + "loss": 1.0393, + "step": 5214 + }, + { + "epoch": 4.165335463258786, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0421, + "step": 5215 + }, + { + "epoch": 4.166134185303514, + "grad_norm": 0.287109375, + "learning_rate": 0.0005, + "loss": 1.0412, + "step": 5216 + }, + { + "epoch": 4.166932907348243, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0466, + "step": 5217 + }, + { + "epoch": 4.167731629392971, + "grad_norm": 0.2080078125, + "learning_rate": 0.0005, + "loss": 1.0469, + "step": 5218 + }, + { + "epoch": 4.1685303514377, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 5219 + }, + { + "epoch": 4.169329073482428, + "grad_norm": 0.203125, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 5220 + }, + { + "epoch": 4.170127795527157, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0395, + "step": 5221 + }, + { + "epoch": 4.170926517571885, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 5222 + }, + { + "epoch": 4.171725239616613, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 5223 + }, + { + "epoch": 4.172523961661342, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 5224 + }, + { + "epoch": 4.17332268370607, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0444, + "step": 5225 + }, + { + "epoch": 4.174121405750799, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 5226 + }, + { + "epoch": 4.174920127795527, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0384, + "step": 5227 + }, + { + "epoch": 4.175718849840256, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0407, + "step": 5228 + }, + { + "epoch": 4.176517571884984, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 5229 + }, + { + "epoch": 4.177316293929713, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0458, + "step": 5230 + }, + { + "epoch": 4.178115015974441, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0453, + "step": 5231 + }, + { + "epoch": 4.178913738019169, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 5232 + }, + { + "epoch": 4.1797124600638975, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 5233 + }, + { + "epoch": 4.180511182108626, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 5234 + }, + { + "epoch": 4.181309904153355, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 5235 + }, + { + "epoch": 4.182108626198083, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.046, + "step": 5236 + }, + { + "epoch": 4.182907348242812, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0452, + "step": 5237 + }, + { + "epoch": 4.18370607028754, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.043, + "step": 5238 + }, + { + "epoch": 4.184504792332269, + "grad_norm": 0.18359375, + "learning_rate": 0.0005, + "loss": 1.0413, + "step": 5239 + }, + { + "epoch": 4.185303514376997, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 5240 + }, + { + "epoch": 4.186102236421725, + "grad_norm": 0.2265625, + "learning_rate": 0.0005, + "loss": 1.0423, + "step": 5241 + }, + { + "epoch": 4.186900958466453, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0395, + "step": 5242 + }, + { + "epoch": 4.187699680511182, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 5243 + }, + { + "epoch": 4.18849840255591, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0396, + "step": 5244 + }, + { + "epoch": 4.189297124600639, + "grad_norm": 0.162109375, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 5245 + }, + { + "epoch": 4.1900958466453675, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 5246 + }, + { + "epoch": 4.190894568690096, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0429, + "step": 5247 + }, + { + "epoch": 4.1916932907348246, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0409, + "step": 5248 + }, + { + "epoch": 4.192492012779553, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 5249 + }, + { + "epoch": 4.193290734824281, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 5250 + }, + { + "epoch": 4.194089456869009, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0491, + "step": 5251 + }, + { + "epoch": 4.194888178913738, + "grad_norm": 0.181640625, + "learning_rate": 0.0005, + "loss": 1.0369, + "step": 5252 + }, + { + "epoch": 4.195686900958466, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0439, + "step": 5253 + }, + { + "epoch": 4.196485623003195, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0426, + "step": 5254 + }, + { + "epoch": 4.197284345047923, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0466, + "step": 5255 + }, + { + "epoch": 4.198083067092652, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 5256 + }, + { + "epoch": 4.19888178913738, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.0429, + "step": 5257 + }, + { + "epoch": 4.199680511182109, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0411, + "step": 5258 + }, + { + "epoch": 4.2004792332268375, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 5259 + }, + { + "epoch": 4.201277955271565, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0411, + "step": 5260 + }, + { + "epoch": 4.202076677316294, + "grad_norm": 0.19921875, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 5261 + }, + { + "epoch": 4.202875399361022, + "grad_norm": 0.1533203125, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 5262 + }, + { + "epoch": 4.203674121405751, + "grad_norm": 0.181640625, + "learning_rate": 0.0005, + "loss": 1.0421, + "step": 5263 + }, + { + "epoch": 4.204472843450479, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.0482, + "step": 5264 + }, + { + "epoch": 4.205271565495208, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 5265 + }, + { + "epoch": 4.206070287539936, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 5266 + }, + { + "epoch": 4.206869009584665, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0436, + "step": 5267 + }, + { + "epoch": 4.207667731629393, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0433, + "step": 5268 + }, + { + "epoch": 4.208466453674121, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0453, + "step": 5269 + }, + { + "epoch": 4.2092651757188495, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 5270 + }, + { + "epoch": 4.210063897763578, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0413, + "step": 5271 + }, + { + "epoch": 4.210862619808307, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0468, + "step": 5272 + }, + { + "epoch": 4.211661341853035, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0476, + "step": 5273 + }, + { + "epoch": 4.212460063897764, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0411, + "step": 5274 + }, + { + "epoch": 4.213258785942492, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0499, + "step": 5275 + }, + { + "epoch": 4.214057507987221, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 5276 + }, + { + "epoch": 4.214856230031949, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0452, + "step": 5277 + }, + { + "epoch": 4.215654952076678, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.0458, + "step": 5278 + }, + { + "epoch": 4.216453674121405, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0431, + "step": 5279 + }, + { + "epoch": 4.217252396166134, + "grad_norm": 0.263671875, + "learning_rate": 0.0005, + "loss": 1.043, + "step": 5280 + }, + { + "epoch": 4.218051118210862, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 5281 + }, + { + "epoch": 4.218849840255591, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 5282 + }, + { + "epoch": 4.2196485623003195, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0373, + "step": 5283 + }, + { + "epoch": 4.220447284345048, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 5284 + }, + { + "epoch": 4.2212460063897765, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0436, + "step": 5285 + }, + { + "epoch": 4.222044728434505, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0411, + "step": 5286 + }, + { + "epoch": 4.222843450479234, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0433, + "step": 5287 + }, + { + "epoch": 4.223642172523961, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 5288 + }, + { + "epoch": 4.22444089456869, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.043, + "step": 5289 + }, + { + "epoch": 4.225239616613418, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0393, + "step": 5290 + }, + { + "epoch": 4.226038338658147, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 5291 + }, + { + "epoch": 4.226837060702875, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 5292 + }, + { + "epoch": 4.227635782747604, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 5293 + }, + { + "epoch": 4.228434504792332, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0407, + "step": 5294 + }, + { + "epoch": 4.229233226837061, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0385, + "step": 5295 + }, + { + "epoch": 4.2300319488817895, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0445, + "step": 5296 + }, + { + "epoch": 4.230830670926518, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0444, + "step": 5297 + }, + { + "epoch": 4.231629392971246, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 5298 + }, + { + "epoch": 4.232428115015974, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0439, + "step": 5299 + }, + { + "epoch": 4.233226837060703, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 5300 + }, + { + "epoch": 4.234025559105431, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0413, + "step": 5301 + }, + { + "epoch": 4.23482428115016, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 5302 + }, + { + "epoch": 4.235623003194888, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 5303 + }, + { + "epoch": 4.236421725239617, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0406, + "step": 5304 + }, + { + "epoch": 4.237220447284345, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 5305 + }, + { + "epoch": 4.238019169329074, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0428, + "step": 5306 + }, + { + "epoch": 4.2388178913738015, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 5307 + }, + { + "epoch": 4.23961661341853, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 5308 + }, + { + "epoch": 4.2404153354632586, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.039, + "step": 5309 + }, + { + "epoch": 4.241214057507987, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 5310 + }, + { + "epoch": 4.242012779552716, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 5311 + }, + { + "epoch": 4.242811501597444, + "grad_norm": 0.208984375, + "learning_rate": 0.0005, + "loss": 1.046, + "step": 5312 + }, + { + "epoch": 4.243610223642173, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0464, + "step": 5313 + }, + { + "epoch": 4.244408945686901, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0459, + "step": 5314 + }, + { + "epoch": 4.24520766773163, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0361, + "step": 5315 + }, + { + "epoch": 4.246006389776358, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 5316 + }, + { + "epoch": 4.246805111821086, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.046, + "step": 5317 + }, + { + "epoch": 4.247603833865814, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.039, + "step": 5318 + }, + { + "epoch": 4.248402555910543, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 5319 + }, + { + "epoch": 4.2492012779552715, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0437, + "step": 5320 + }, + { + "epoch": 4.25, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0437, + "step": 5321 + }, + { + "epoch": 4.2507987220447285, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 5322 + }, + { + "epoch": 4.251597444089457, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0403, + "step": 5323 + }, + { + "epoch": 4.252396166134186, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0442, + "step": 5324 + }, + { + "epoch": 4.253194888178914, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 5325 + }, + { + "epoch": 4.253993610223642, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 5326 + }, + { + "epoch": 4.25479233226837, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 5327 + }, + { + "epoch": 4.255591054313099, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0433, + "step": 5328 + }, + { + "epoch": 4.256389776357827, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 5329 + }, + { + "epoch": 4.257188498402556, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0369, + "step": 5330 + }, + { + "epoch": 4.257987220447284, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 5331 + }, + { + "epoch": 4.258785942492013, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0391, + "step": 5332 + }, + { + "epoch": 4.2595846645367414, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 5333 + }, + { + "epoch": 4.26038338658147, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0404, + "step": 5334 + }, + { + "epoch": 4.261182108626198, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 5335 + }, + { + "epoch": 4.261980830670926, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0412, + "step": 5336 + }, + { + "epoch": 4.262779552715655, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 5337 + }, + { + "epoch": 4.263578274760383, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 5338 + }, + { + "epoch": 4.264376996805112, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 5339 + }, + { + "epoch": 4.26517571884984, + "grad_norm": 0.1953125, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 5340 + }, + { + "epoch": 4.265974440894569, + "grad_norm": 0.2099609375, + "learning_rate": 0.0005, + "loss": 1.0391, + "step": 5341 + }, + { + "epoch": 4.266773162939297, + "grad_norm": 0.1826171875, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 5342 + }, + { + "epoch": 4.267571884984026, + "grad_norm": 0.40625, + "learning_rate": 0.0005, + "loss": 1.0397, + "step": 5343 + }, + { + "epoch": 4.268370607028754, + "grad_norm": 0.2431640625, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 5344 + }, + { + "epoch": 4.269169329073483, + "grad_norm": 0.255859375, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 5345 + }, + { + "epoch": 4.2699680511182105, + "grad_norm": 0.58984375, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 5346 + }, + { + "epoch": 4.270766773162939, + "grad_norm": 0.53125, + "learning_rate": 0.0005, + "loss": 1.062, + "step": 5347 + }, + { + "epoch": 4.271565495207668, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0451, + "step": 5348 + }, + { + "epoch": 4.272364217252396, + "grad_norm": 0.7890625, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 5349 + }, + { + "epoch": 4.273162939297125, + "grad_norm": 0.71875, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 5350 + }, + { + "epoch": 4.273961661341853, + "grad_norm": 0.1572265625, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 5351 + }, + { + "epoch": 4.274760383386582, + "grad_norm": 0.86328125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 5352 + }, + { + "epoch": 4.27555910543131, + "grad_norm": 0.6328125, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 5353 + }, + { + "epoch": 4.276357827476039, + "grad_norm": 0.353515625, + "learning_rate": 0.0005, + "loss": 1.042, + "step": 5354 + }, + { + "epoch": 4.277156549520766, + "grad_norm": 0.8359375, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 5355 + }, + { + "epoch": 4.277955271565495, + "grad_norm": 0.19140625, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 5356 + }, + { + "epoch": 4.2787539936102235, + "grad_norm": 0.466796875, + "learning_rate": 0.0005, + "loss": 1.0433, + "step": 5357 + }, + { + "epoch": 4.279552715654952, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0464, + "step": 5358 + }, + { + "epoch": 4.2803514376996805, + "grad_norm": 0.451171875, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 5359 + }, + { + "epoch": 4.281150159744409, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 5360 + }, + { + "epoch": 4.281948881789138, + "grad_norm": 0.322265625, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 5361 + }, + { + "epoch": 4.282747603833866, + "grad_norm": 0.208984375, + "learning_rate": 0.0005, + "loss": 1.0393, + "step": 5362 + }, + { + "epoch": 4.283546325878595, + "grad_norm": 0.4921875, + "learning_rate": 0.0005, + "loss": 1.037, + "step": 5363 + }, + { + "epoch": 4.284345047923322, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0402, + "step": 5364 + }, + { + "epoch": 4.285143769968051, + "grad_norm": 0.55078125, + "learning_rate": 0.0005, + "loss": 1.057, + "step": 5365 + }, + { + "epoch": 4.285942492012779, + "grad_norm": 0.224609375, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 5366 + }, + { + "epoch": 4.286741214057508, + "grad_norm": 0.25, + "learning_rate": 0.0005, + "loss": 1.0452, + "step": 5367 + }, + { + "epoch": 4.287539936102236, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0432, + "step": 5368 + }, + { + "epoch": 4.288338658146965, + "grad_norm": 0.1708984375, + "learning_rate": 0.0005, + "loss": 1.0396, + "step": 5369 + }, + { + "epoch": 4.289137380191693, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 5370 + }, + { + "epoch": 4.289936102236422, + "grad_norm": 0.25, + "learning_rate": 0.0005, + "loss": 1.048, + "step": 5371 + }, + { + "epoch": 4.2907348242811505, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0433, + "step": 5372 + }, + { + "epoch": 4.291533546325878, + "grad_norm": 0.267578125, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 5373 + }, + { + "epoch": 4.292332268370607, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 5374 + }, + { + "epoch": 4.293130990415335, + "grad_norm": 0.236328125, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 5375 + }, + { + "epoch": 4.293929712460064, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0395, + "step": 5376 + }, + { + "epoch": 4.294728434504792, + "grad_norm": 0.25390625, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 5377 + }, + { + "epoch": 4.295527156549521, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 5378 + }, + { + "epoch": 4.296325878594249, + "grad_norm": 0.26171875, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 5379 + }, + { + "epoch": 4.297124600638978, + "grad_norm": 0.189453125, + "learning_rate": 0.0005, + "loss": 1.0441, + "step": 5380 + }, + { + "epoch": 4.297923322683706, + "grad_norm": 0.283203125, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 5381 + }, + { + "epoch": 4.298722044728435, + "grad_norm": 0.203125, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 5382 + }, + { + "epoch": 4.2995207667731625, + "grad_norm": 0.2734375, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 5383 + }, + { + "epoch": 4.300319488817891, + "grad_norm": 0.30078125, + "learning_rate": 0.0005, + "loss": 1.041, + "step": 5384 + }, + { + "epoch": 4.30111821086262, + "grad_norm": 0.5546875, + "learning_rate": 0.0005, + "loss": 1.0458, + "step": 5385 + }, + { + "epoch": 4.301916932907348, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 5386 + }, + { + "epoch": 4.302715654952077, + "grad_norm": 0.515625, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 5387 + }, + { + "epoch": 4.303514376996805, + "grad_norm": 0.201171875, + "learning_rate": 0.0005, + "loss": 1.038, + "step": 5388 + }, + { + "epoch": 4.304313099041534, + "grad_norm": 0.287109375, + "learning_rate": 0.0005, + "loss": 1.0431, + "step": 5389 + }, + { + "epoch": 4.305111821086262, + "grad_norm": 0.166015625, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 5390 + }, + { + "epoch": 4.305910543130991, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.04, + "step": 5391 + }, + { + "epoch": 4.306709265175719, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0399, + "step": 5392 + }, + { + "epoch": 4.307507987220447, + "grad_norm": 0.1962890625, + "learning_rate": 0.0005, + "loss": 1.0406, + "step": 5393 + }, + { + "epoch": 4.3083067092651754, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 5394 + }, + { + "epoch": 4.309105431309904, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 5395 + }, + { + "epoch": 4.3099041533546325, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.0395, + "step": 5396 + }, + { + "epoch": 4.310702875399361, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0532, + "step": 5397 + }, + { + "epoch": 4.31150159744409, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0379, + "step": 5398 + }, + { + "epoch": 4.312300319488818, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0464, + "step": 5399 + }, + { + "epoch": 4.313099041533547, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 5400 + }, + { + "epoch": 4.313897763578275, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.044, + "step": 5401 + }, + { + "epoch": 4.314696485623003, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 5402 + }, + { + "epoch": 4.315495207667731, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 5403 + }, + { + "epoch": 4.31629392971246, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 5404 + }, + { + "epoch": 4.317092651757188, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0409, + "step": 5405 + }, + { + "epoch": 4.317891373801917, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0404, + "step": 5406 + }, + { + "epoch": 4.318690095846645, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0442, + "step": 5407 + }, + { + "epoch": 4.319488817891374, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0407, + "step": 5408 + }, + { + "epoch": 4.3202875399361025, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 5409 + }, + { + "epoch": 4.321086261980831, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0468, + "step": 5410 + }, + { + "epoch": 4.321884984025559, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 5411 + }, + { + "epoch": 4.322683706070287, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0413, + "step": 5412 + }, + { + "epoch": 4.323482428115016, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.038, + "step": 5413 + }, + { + "epoch": 4.324281150159744, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0396, + "step": 5414 + }, + { + "epoch": 4.325079872204473, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0369, + "step": 5415 + }, + { + "epoch": 4.325878594249201, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0453, + "step": 5416 + }, + { + "epoch": 4.32667731629393, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0402, + "step": 5417 + }, + { + "epoch": 4.327476038338658, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0432, + "step": 5418 + }, + { + "epoch": 4.328274760383387, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 5419 + }, + { + "epoch": 4.329073482428115, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0378, + "step": 5420 + }, + { + "epoch": 4.329872204472843, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0488, + "step": 5421 + }, + { + "epoch": 4.330670926517572, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.044, + "step": 5422 + }, + { + "epoch": 4.3314696485623, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0401, + "step": 5423 + }, + { + "epoch": 4.332268370607029, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 5424 + }, + { + "epoch": 4.333067092651757, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0448, + "step": 5425 + }, + { + "epoch": 4.333865814696486, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0433, + "step": 5426 + }, + { + "epoch": 4.334664536741214, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0571, + "step": 5427 + }, + { + "epoch": 4.335463258785943, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0446, + "step": 5428 + }, + { + "epoch": 4.336261980830671, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0476, + "step": 5429 + }, + { + "epoch": 4.3370607028754, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0439, + "step": 5430 + }, + { + "epoch": 4.337859424920127, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 5431 + }, + { + "epoch": 4.338658146964856, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 5432 + }, + { + "epoch": 4.3394568690095845, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0434, + "step": 5433 + }, + { + "epoch": 4.340255591054313, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0456, + "step": 5434 + }, + { + "epoch": 4.3410543130990416, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 5435 + }, + { + "epoch": 4.34185303514377, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 5436 + }, + { + "epoch": 4.342651757188499, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0429, + "step": 5437 + }, + { + "epoch": 4.343450479233227, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 5438 + }, + { + "epoch": 4.344249201277956, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 5439 + }, + { + "epoch": 4.345047923322683, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0467, + "step": 5440 + }, + { + "epoch": 4.345846645367412, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.041, + "step": 5441 + }, + { + "epoch": 4.34664536741214, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0436, + "step": 5442 + }, + { + "epoch": 4.347444089456869, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0422, + "step": 5443 + }, + { + "epoch": 4.348242811501597, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0415, + "step": 5444 + }, + { + "epoch": 4.349041533546326, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0412, + "step": 5445 + }, + { + "epoch": 4.3498402555910545, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0377, + "step": 5446 + }, + { + "epoch": 4.350638977635783, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0394, + "step": 5447 + }, + { + "epoch": 4.3514376996805115, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 5448 + }, + { + "epoch": 4.352236421725239, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 5449 + }, + { + "epoch": 4.353035143769968, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 5450 + }, + { + "epoch": 4.353833865814696, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0431, + "step": 5451 + }, + { + "epoch": 4.354632587859425, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0439, + "step": 5452 + }, + { + "epoch": 4.355431309904153, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 5453 + }, + { + "epoch": 4.356230031948882, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 5454 + }, + { + "epoch": 4.35702875399361, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0419, + "step": 5455 + }, + { + "epoch": 4.357827476038339, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 5456 + }, + { + "epoch": 4.358626198083067, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0537, + "step": 5457 + }, + { + "epoch": 4.359424920127796, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 5458 + }, + { + "epoch": 4.360223642172524, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0406, + "step": 5459 + }, + { + "epoch": 4.361022364217252, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0581, + "step": 5460 + }, + { + "epoch": 4.361821086261981, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 5461 + }, + { + "epoch": 4.362619808306709, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 5462 + }, + { + "epoch": 4.363418530351438, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 5463 + }, + { + "epoch": 4.364217252396166, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0385, + "step": 5464 + }, + { + "epoch": 4.365015974440895, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 5465 + }, + { + "epoch": 4.365814696485623, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 5466 + }, + { + "epoch": 4.366613418530352, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.048, + "step": 5467 + }, + { + "epoch": 4.36741214057508, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0465, + "step": 5468 + }, + { + "epoch": 4.368210862619808, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0452, + "step": 5469 + }, + { + "epoch": 4.3690095846645365, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 5470 + }, + { + "epoch": 4.369808306709265, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0431, + "step": 5471 + }, + { + "epoch": 4.3706070287539935, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 5472 + }, + { + "epoch": 4.371405750798722, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0436, + "step": 5473 + }, + { + "epoch": 4.372204472843451, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0436, + "step": 5474 + }, + { + "epoch": 4.373003194888179, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 5475 + }, + { + "epoch": 4.373801916932908, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 5476 + }, + { + "epoch": 4.374600638977636, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 5477 + }, + { + "epoch": 4.375399361022364, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 5478 + }, + { + "epoch": 4.376198083067092, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 5479 + }, + { + "epoch": 4.376996805111821, + "grad_norm": 0.1640625, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 5480 + }, + { + "epoch": 4.377795527156549, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 5481 + }, + { + "epoch": 4.378594249201278, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0464, + "step": 5482 + }, + { + "epoch": 4.3793929712460065, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 5483 + }, + { + "epoch": 4.380191693290735, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0469, + "step": 5484 + }, + { + "epoch": 4.3809904153354635, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0388, + "step": 5485 + }, + { + "epoch": 4.381789137380192, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.0457, + "step": 5486 + }, + { + "epoch": 4.38258785942492, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 5487 + }, + { + "epoch": 4.383386581469648, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 5488 + }, + { + "epoch": 4.384185303514377, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 5489 + }, + { + "epoch": 4.384984025559105, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 5490 + }, + { + "epoch": 4.385782747603834, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 5491 + }, + { + "epoch": 4.386581469648562, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0439, + "step": 5492 + }, + { + "epoch": 4.387380191693291, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0446, + "step": 5493 + }, + { + "epoch": 4.388178913738019, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.044, + "step": 5494 + }, + { + "epoch": 4.388977635782748, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.036, + "step": 5495 + }, + { + "epoch": 4.389776357827476, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0354, + "step": 5496 + }, + { + "epoch": 4.390575079872204, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.048, + "step": 5497 + }, + { + "epoch": 4.391373801916933, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.042, + "step": 5498 + }, + { + "epoch": 4.392172523961661, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 5499 + }, + { + "epoch": 4.39297124600639, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0399, + "step": 5500 + }, + { + "epoch": 4.393769968051118, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0441, + "step": 5501 + }, + { + "epoch": 4.394568690095847, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 5502 + }, + { + "epoch": 4.395367412140575, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0535, + "step": 5503 + }, + { + "epoch": 4.396166134185304, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0403, + "step": 5504 + }, + { + "epoch": 4.396964856230032, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0377, + "step": 5505 + }, + { + "epoch": 4.397763578274761, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 5506 + }, + { + "epoch": 4.3985623003194885, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.04, + "step": 5507 + }, + { + "epoch": 4.399361022364217, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0414, + "step": 5508 + }, + { + "epoch": 4.4001597444089455, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0425, + "step": 5509 + }, + { + "epoch": 4.400958466453674, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 5510 + }, + { + "epoch": 4.401757188498403, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 5511 + }, + { + "epoch": 4.402555910543131, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 5512 + }, + { + "epoch": 4.40335463258786, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0406, + "step": 5513 + }, + { + "epoch": 4.404153354632588, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0389, + "step": 5514 + }, + { + "epoch": 4.404952076677317, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0409, + "step": 5515 + }, + { + "epoch": 4.405750798722044, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0412, + "step": 5516 + }, + { + "epoch": 4.406549520766773, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0452, + "step": 5517 + }, + { + "epoch": 4.407348242811501, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 5518 + }, + { + "epoch": 4.40814696485623, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0391, + "step": 5519 + }, + { + "epoch": 4.4089456869009584, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0459, + "step": 5520 + }, + { + "epoch": 4.409744408945687, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0455, + "step": 5521 + }, + { + "epoch": 4.4105431309904155, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0468, + "step": 5522 + }, + { + "epoch": 4.411341853035144, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0422, + "step": 5523 + }, + { + "epoch": 4.412140575079873, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0443, + "step": 5524 + }, + { + "epoch": 4.4129392971246, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 5525 + }, + { + "epoch": 4.413738019169329, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 5526 + }, + { + "epoch": 4.414536741214057, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0409, + "step": 5527 + }, + { + "epoch": 4.415335463258786, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0423, + "step": 5528 + }, + { + "epoch": 4.416134185303514, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0338, + "step": 5529 + }, + { + "epoch": 4.416932907348243, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 5530 + }, + { + "epoch": 4.417731629392971, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 5531 + }, + { + "epoch": 4.4185303514377, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.0394, + "step": 5532 + }, + { + "epoch": 4.419329073482428, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0429, + "step": 5533 + }, + { + "epoch": 4.420127795527157, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 5534 + }, + { + "epoch": 4.420926517571885, + "grad_norm": 0.1572265625, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 5535 + }, + { + "epoch": 4.421725239616613, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0413, + "step": 5536 + }, + { + "epoch": 4.422523961661342, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0444, + "step": 5537 + }, + { + "epoch": 4.42332268370607, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.045, + "step": 5538 + }, + { + "epoch": 4.424121405750799, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0435, + "step": 5539 + }, + { + "epoch": 4.424920127795527, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 5540 + }, + { + "epoch": 4.425718849840256, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0422, + "step": 5541 + }, + { + "epoch": 4.426517571884984, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0377, + "step": 5542 + }, + { + "epoch": 4.427316293929713, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 5543 + }, + { + "epoch": 4.428115015974441, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0432, + "step": 5544 + }, + { + "epoch": 4.428913738019169, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0439, + "step": 5545 + }, + { + "epoch": 4.4297124600638975, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 5546 + }, + { + "epoch": 4.430511182108626, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 5547 + }, + { + "epoch": 4.431309904153355, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0396, + "step": 5548 + }, + { + "epoch": 4.432108626198083, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0439, + "step": 5549 + }, + { + "epoch": 4.432907348242812, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 5550 + }, + { + "epoch": 4.43370607028754, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 5551 + }, + { + "epoch": 4.434504792332269, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 5552 + }, + { + "epoch": 4.435303514376997, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0354, + "step": 5553 + }, + { + "epoch": 4.436102236421725, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 5554 + }, + { + "epoch": 4.436900958466453, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.037, + "step": 5555 + }, + { + "epoch": 4.437699680511182, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0395, + "step": 5556 + }, + { + "epoch": 4.43849840255591, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0451, + "step": 5557 + }, + { + "epoch": 4.439297124600639, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0417, + "step": 5558 + }, + { + "epoch": 4.4400958466453675, + "grad_norm": 0.16796875, + "learning_rate": 0.0005, + "loss": 1.0421, + "step": 5559 + }, + { + "epoch": 4.440894568690096, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0432, + "step": 5560 + }, + { + "epoch": 4.4416932907348246, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0434, + "step": 5561 + }, + { + "epoch": 4.442492012779553, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0422, + "step": 5562 + }, + { + "epoch": 4.443290734824281, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 5563 + }, + { + "epoch": 4.444089456869009, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0406, + "step": 5564 + }, + { + "epoch": 4.444888178913738, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0385, + "step": 5565 + }, + { + "epoch": 4.445686900958466, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0411, + "step": 5566 + }, + { + "epoch": 4.446485623003195, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.039, + "step": 5567 + }, + { + "epoch": 4.447284345047923, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0454, + "step": 5568 + }, + { + "epoch": 4.448083067092652, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 5569 + }, + { + "epoch": 4.44888178913738, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.044, + "step": 5570 + }, + { + "epoch": 4.449680511182109, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 5571 + }, + { + "epoch": 4.4504792332268375, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 5572 + }, + { + "epoch": 4.451277955271565, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 5573 + }, + { + "epoch": 4.452076677316294, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.046, + "step": 5574 + }, + { + "epoch": 4.452875399361022, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 5575 + }, + { + "epoch": 4.453674121405751, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.041, + "step": 5576 + }, + { + "epoch": 4.454472843450479, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0411, + "step": 5577 + }, + { + "epoch": 4.455271565495208, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0439, + "step": 5578 + }, + { + "epoch": 4.456070287539936, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 5579 + }, + { + "epoch": 4.456869009584665, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 5580 + }, + { + "epoch": 4.457667731629393, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.048, + "step": 5581 + }, + { + "epoch": 4.458466453674122, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0413, + "step": 5582 + }, + { + "epoch": 4.4592651757188495, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0379, + "step": 5583 + }, + { + "epoch": 4.460063897763578, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0445, + "step": 5584 + }, + { + "epoch": 4.460862619808307, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0453, + "step": 5585 + }, + { + "epoch": 4.461661341853035, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 5586 + }, + { + "epoch": 4.462460063897764, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0373, + "step": 5587 + }, + { + "epoch": 4.463258785942492, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 5588 + }, + { + "epoch": 4.464057507987221, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0373, + "step": 5589 + }, + { + "epoch": 4.464856230031949, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0407, + "step": 5590 + }, + { + "epoch": 4.465654952076678, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.038, + "step": 5591 + }, + { + "epoch": 4.466453674121405, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 5592 + }, + { + "epoch": 4.467252396166134, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 5593 + }, + { + "epoch": 4.468051118210862, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 5594 + }, + { + "epoch": 4.468849840255591, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 5595 + }, + { + "epoch": 4.4696485623003195, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0416, + "step": 5596 + }, + { + "epoch": 4.470447284345048, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 5597 + }, + { + "epoch": 4.4712460063897765, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0379, + "step": 5598 + }, + { + "epoch": 4.472044728434505, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0379, + "step": 5599 + }, + { + "epoch": 4.472843450479234, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0434, + "step": 5600 + }, + { + "epoch": 4.473642172523961, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 5601 + }, + { + "epoch": 4.47444089456869, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0416, + "step": 5602 + }, + { + "epoch": 4.475239616613418, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0379, + "step": 5603 + }, + { + "epoch": 4.476038338658147, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0432, + "step": 5604 + }, + { + "epoch": 4.476837060702875, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 5605 + }, + { + "epoch": 4.477635782747604, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 5606 + }, + { + "epoch": 4.478434504792332, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0388, + "step": 5607 + }, + { + "epoch": 4.479233226837061, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0449, + "step": 5608 + }, + { + "epoch": 4.4800319488817895, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0399, + "step": 5609 + }, + { + "epoch": 4.480830670926517, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0574, + "step": 5610 + }, + { + "epoch": 4.481629392971246, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 5611 + }, + { + "epoch": 4.482428115015974, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 5612 + }, + { + "epoch": 4.483226837060703, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.04, + "step": 5613 + }, + { + "epoch": 4.484025559105431, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 5614 + }, + { + "epoch": 4.48482428115016, + "grad_norm": 0.1640625, + "learning_rate": 0.0005, + "loss": 1.037, + "step": 5615 + }, + { + "epoch": 4.485623003194888, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0385, + "step": 5616 + }, + { + "epoch": 4.486421725239617, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0414, + "step": 5617 + }, + { + "epoch": 4.487220447284345, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0376, + "step": 5618 + }, + { + "epoch": 4.488019169329074, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0401, + "step": 5619 + }, + { + "epoch": 4.488817891373802, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0419, + "step": 5620 + }, + { + "epoch": 4.48961661341853, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0439, + "step": 5621 + }, + { + "epoch": 4.4904153354632586, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 5622 + }, + { + "epoch": 4.491214057507987, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 5623 + }, + { + "epoch": 4.492012779552716, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 5624 + }, + { + "epoch": 4.492811501597444, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0402, + "step": 5625 + }, + { + "epoch": 4.493610223642173, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 5626 + }, + { + "epoch": 4.494408945686901, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0469, + "step": 5627 + }, + { + "epoch": 4.49520766773163, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 5628 + }, + { + "epoch": 4.496006389776358, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 5629 + }, + { + "epoch": 4.496805111821086, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0541, + "step": 5630 + }, + { + "epoch": 4.497603833865814, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0434, + "step": 5631 + }, + { + "epoch": 4.498402555910543, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 5632 + }, + { + "epoch": 4.4992012779552715, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0449, + "step": 5633 + }, + { + "epoch": 4.5, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0457, + "step": 5634 + }, + { + "epoch": 4.5007987220447285, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 5635 + }, + { + "epoch": 4.501597444089457, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0435, + "step": 5636 + }, + { + "epoch": 4.502396166134186, + "grad_norm": 0.1669921875, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 5637 + }, + { + "epoch": 4.503194888178914, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0452, + "step": 5638 + }, + { + "epoch": 4.503993610223642, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 5639 + }, + { + "epoch": 4.50479233226837, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0425, + "step": 5640 + }, + { + "epoch": 4.505591054313099, + "grad_norm": 0.248046875, + "learning_rate": 0.0005, + "loss": 1.037, + "step": 5641 + }, + { + "epoch": 4.506389776357827, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0375, + "step": 5642 + }, + { + "epoch": 4.507188498402556, + "grad_norm": 0.26171875, + "learning_rate": 0.0005, + "loss": 1.0432, + "step": 5643 + }, + { + "epoch": 4.507987220447284, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0394, + "step": 5644 + }, + { + "epoch": 4.508785942492013, + "grad_norm": 0.3203125, + "learning_rate": 0.0005, + "loss": 1.0448, + "step": 5645 + }, + { + "epoch": 4.5095846645367414, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.044, + "step": 5646 + }, + { + "epoch": 4.51038338658147, + "grad_norm": 0.32421875, + "learning_rate": 0.0005, + "loss": 1.0429, + "step": 5647 + }, + { + "epoch": 4.511182108626198, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0405, + "step": 5648 + }, + { + "epoch": 4.511980830670926, + "grad_norm": 0.26171875, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 5649 + }, + { + "epoch": 4.512779552715655, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0439, + "step": 5650 + }, + { + "epoch": 4.513578274760383, + "grad_norm": 0.2255859375, + "learning_rate": 0.0005, + "loss": 1.0401, + "step": 5651 + }, + { + "epoch": 4.514376996805112, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0547, + "step": 5652 + }, + { + "epoch": 4.51517571884984, + "grad_norm": 0.2373046875, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 5653 + }, + { + "epoch": 4.515974440894569, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0353, + "step": 5654 + }, + { + "epoch": 4.516773162939297, + "grad_norm": 0.3203125, + "learning_rate": 0.0005, + "loss": 1.0378, + "step": 5655 + }, + { + "epoch": 4.517571884984026, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0407, + "step": 5656 + }, + { + "epoch": 4.518370607028754, + "grad_norm": 0.1748046875, + "learning_rate": 0.0005, + "loss": 1.0426, + "step": 5657 + }, + { + "epoch": 4.519169329073483, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0375, + "step": 5658 + }, + { + "epoch": 4.5199680511182105, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.0375, + "step": 5659 + }, + { + "epoch": 4.520766773162939, + "grad_norm": 0.1953125, + "learning_rate": 0.0005, + "loss": 1.0454, + "step": 5660 + }, + { + "epoch": 4.521565495207668, + "grad_norm": 0.25, + "learning_rate": 0.0005, + "loss": 1.037, + "step": 5661 + }, + { + "epoch": 4.522364217252396, + "grad_norm": 0.212890625, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 5662 + }, + { + "epoch": 4.523162939297125, + "grad_norm": 0.349609375, + "learning_rate": 0.0005, + "loss": 1.0447, + "step": 5663 + }, + { + "epoch": 4.523961661341853, + "grad_norm": 0.2421875, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 5664 + }, + { + "epoch": 4.524760383386582, + "grad_norm": 0.63671875, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 5665 + }, + { + "epoch": 4.52555910543131, + "grad_norm": 0.34765625, + "learning_rate": 0.0005, + "loss": 1.0441, + "step": 5666 + }, + { + "epoch": 4.526357827476039, + "grad_norm": 0.25390625, + "learning_rate": 0.0005, + "loss": 1.042, + "step": 5667 + }, + { + "epoch": 4.527156549520766, + "grad_norm": 0.375, + "learning_rate": 0.0005, + "loss": 1.0408, + "step": 5668 + }, + { + "epoch": 4.527955271565495, + "grad_norm": 0.18359375, + "learning_rate": 0.0005, + "loss": 1.0397, + "step": 5669 + }, + { + "epoch": 4.5287539936102235, + "grad_norm": 0.5234375, + "learning_rate": 0.0005, + "loss": 1.0482, + "step": 5670 + }, + { + "epoch": 4.529552715654952, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0468, + "step": 5671 + }, + { + "epoch": 4.5303514376996805, + "grad_norm": 0.427734375, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 5672 + }, + { + "epoch": 4.531150159744409, + "grad_norm": 0.17578125, + "learning_rate": 0.0005, + "loss": 1.0412, + "step": 5673 + }, + { + "epoch": 4.531948881789138, + "grad_norm": 0.2578125, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 5674 + }, + { + "epoch": 4.532747603833866, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0557, + "step": 5675 + }, + { + "epoch": 4.533546325878595, + "grad_norm": 0.2109375, + "learning_rate": 0.0005, + "loss": 1.0451, + "step": 5676 + }, + { + "epoch": 4.534345047923322, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 5677 + }, + { + "epoch": 4.535143769968051, + "grad_norm": 0.2109375, + "learning_rate": 0.0005, + "loss": 1.042, + "step": 5678 + }, + { + "epoch": 4.535942492012779, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.041, + "step": 5679 + }, + { + "epoch": 4.536741214057508, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0402, + "step": 5680 + }, + { + "epoch": 4.537539936102236, + "grad_norm": 0.1611328125, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 5681 + }, + { + "epoch": 4.538338658146965, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 5682 + }, + { + "epoch": 4.539137380191693, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0433, + "step": 5683 + }, + { + "epoch": 4.539936102236422, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 5684 + }, + { + "epoch": 4.5407348242811505, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0399, + "step": 5685 + }, + { + "epoch": 4.541533546325878, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0348, + "step": 5686 + }, + { + "epoch": 4.542332268370607, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0397, + "step": 5687 + }, + { + "epoch": 4.543130990415335, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 5688 + }, + { + "epoch": 4.543929712460064, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0397, + "step": 5689 + }, + { + "epoch": 4.544728434504792, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 5690 + }, + { + "epoch": 4.545527156549521, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0527, + "step": 5691 + }, + { + "epoch": 4.546325878594249, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.043, + "step": 5692 + }, + { + "epoch": 4.547124600638978, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.045, + "step": 5693 + }, + { + "epoch": 4.547923322683706, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 5694 + }, + { + "epoch": 4.548722044728435, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 5695 + }, + { + "epoch": 4.549520766773163, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 5696 + }, + { + "epoch": 4.550319488817891, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0449, + "step": 5697 + }, + { + "epoch": 4.55111821086262, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0425, + "step": 5698 + }, + { + "epoch": 4.551916932907348, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0389, + "step": 5699 + }, + { + "epoch": 4.552715654952077, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0361, + "step": 5700 + }, + { + "epoch": 4.553514376996805, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0397, + "step": 5701 + }, + { + "epoch": 4.554313099041534, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 5702 + }, + { + "epoch": 4.555111821086262, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0414, + "step": 5703 + }, + { + "epoch": 4.555910543130991, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 5704 + }, + { + "epoch": 4.556709265175719, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0373, + "step": 5705 + }, + { + "epoch": 4.557507987220447, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0393, + "step": 5706 + }, + { + "epoch": 4.5583067092651754, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0531, + "step": 5707 + }, + { + "epoch": 4.559105431309904, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 5708 + }, + { + "epoch": 4.5599041533546325, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0413, + "step": 5709 + }, + { + "epoch": 4.560702875399361, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0472, + "step": 5710 + }, + { + "epoch": 4.56150159744409, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0423, + "step": 5711 + }, + { + "epoch": 4.562300319488818, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 5712 + }, + { + "epoch": 4.563099041533547, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 5713 + }, + { + "epoch": 4.563897763578275, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0369, + "step": 5714 + }, + { + "epoch": 4.564696485623003, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0422, + "step": 5715 + }, + { + "epoch": 4.565495207667731, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0497, + "step": 5716 + }, + { + "epoch": 4.56629392971246, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 5717 + }, + { + "epoch": 4.567092651757188, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 5718 + }, + { + "epoch": 4.567891373801917, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 5719 + }, + { + "epoch": 4.568690095846645, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0448, + "step": 5720 + }, + { + "epoch": 4.569488817891374, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0386, + "step": 5721 + }, + { + "epoch": 4.5702875399361025, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0416, + "step": 5722 + }, + { + "epoch": 4.571086261980831, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0449, + "step": 5723 + }, + { + "epoch": 4.571884984025559, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0379, + "step": 5724 + }, + { + "epoch": 4.572683706070287, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0456, + "step": 5725 + }, + { + "epoch": 4.573482428115016, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0404, + "step": 5726 + }, + { + "epoch": 4.574281150159744, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.044, + "step": 5727 + }, + { + "epoch": 4.575079872204473, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 5728 + }, + { + "epoch": 4.575878594249201, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0433, + "step": 5729 + }, + { + "epoch": 4.57667731629393, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.039, + "step": 5730 + }, + { + "epoch": 4.577476038338658, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 5731 + }, + { + "epoch": 4.578274760383387, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0428, + "step": 5732 + }, + { + "epoch": 4.5790734824281145, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 5733 + }, + { + "epoch": 4.579872204472844, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0369, + "step": 5734 + }, + { + "epoch": 4.580670926517572, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0441, + "step": 5735 + }, + { + "epoch": 4.5814696485623, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 5736 + }, + { + "epoch": 4.582268370607029, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 5737 + }, + { + "epoch": 4.583067092651757, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0416, + "step": 5738 + }, + { + "epoch": 4.583865814696486, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 5739 + }, + { + "epoch": 4.584664536741214, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0415, + "step": 5740 + }, + { + "epoch": 4.585463258785943, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0457, + "step": 5741 + }, + { + "epoch": 4.586261980830671, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0443, + "step": 5742 + }, + { + "epoch": 4.5870607028754, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 5743 + }, + { + "epoch": 4.587859424920127, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0466, + "step": 5744 + }, + { + "epoch": 4.588658146964856, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0526, + "step": 5745 + }, + { + "epoch": 4.5894568690095845, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.043, + "step": 5746 + }, + { + "epoch": 4.590255591054313, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0354, + "step": 5747 + }, + { + "epoch": 4.5910543130990416, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 5748 + }, + { + "epoch": 4.59185303514377, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 5749 + }, + { + "epoch": 4.592651757188499, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 5750 + }, + { + "epoch": 4.593450479233227, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 5751 + }, + { + "epoch": 4.594249201277956, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0439, + "step": 5752 + }, + { + "epoch": 4.595047923322683, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 5753 + }, + { + "epoch": 4.595846645367412, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 5754 + }, + { + "epoch": 4.59664536741214, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 5755 + }, + { + "epoch": 4.597444089456869, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 5756 + }, + { + "epoch": 4.598242811501597, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 5757 + }, + { + "epoch": 4.599041533546326, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0447, + "step": 5758 + }, + { + "epoch": 4.5998402555910545, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0424, + "step": 5759 + }, + { + "epoch": 4.600638977635783, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0464, + "step": 5760 + }, + { + "epoch": 4.6014376996805115, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0411, + "step": 5761 + }, + { + "epoch": 4.602236421725239, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0421, + "step": 5762 + }, + { + "epoch": 4.603035143769968, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0413, + "step": 5763 + }, + { + "epoch": 4.603833865814696, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 5764 + }, + { + "epoch": 4.604632587859425, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0434, + "step": 5765 + }, + { + "epoch": 4.605431309904153, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0405, + "step": 5766 + }, + { + "epoch": 4.606230031948882, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0361, + "step": 5767 + }, + { + "epoch": 4.60702875399361, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 5768 + }, + { + "epoch": 4.607827476038339, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 5769 + }, + { + "epoch": 4.608626198083067, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0412, + "step": 5770 + }, + { + "epoch": 4.609424920127795, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0427, + "step": 5771 + }, + { + "epoch": 4.6102236421725244, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0415, + "step": 5772 + }, + { + "epoch": 4.611022364217252, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0377, + "step": 5773 + }, + { + "epoch": 4.611821086261981, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0424, + "step": 5774 + }, + { + "epoch": 4.612619808306709, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0441, + "step": 5775 + }, + { + "epoch": 4.613418530351438, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0468, + "step": 5776 + }, + { + "epoch": 4.614217252396166, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0562, + "step": 5777 + }, + { + "epoch": 4.615015974440895, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0406, + "step": 5778 + }, + { + "epoch": 4.615814696485623, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0405, + "step": 5779 + }, + { + "epoch": 4.616613418530352, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 5780 + }, + { + "epoch": 4.61741214057508, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 5781 + }, + { + "epoch": 4.618210862619808, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.037, + "step": 5782 + }, + { + "epoch": 4.6190095846645365, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0403, + "step": 5783 + }, + { + "epoch": 4.619808306709265, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0423, + "step": 5784 + }, + { + "epoch": 4.6206070287539935, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0425, + "step": 5785 + }, + { + "epoch": 4.621405750798722, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0386, + "step": 5786 + }, + { + "epoch": 4.622204472843451, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0407, + "step": 5787 + }, + { + "epoch": 4.623003194888179, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 5788 + }, + { + "epoch": 4.623801916932908, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0464, + "step": 5789 + }, + { + "epoch": 4.624600638977636, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0441, + "step": 5790 + }, + { + "epoch": 4.625399361022364, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0433, + "step": 5791 + }, + { + "epoch": 4.626198083067092, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 5792 + }, + { + "epoch": 4.626996805111821, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0433, + "step": 5793 + }, + { + "epoch": 4.627795527156549, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.059, + "step": 5794 + }, + { + "epoch": 4.628594249201278, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.0458, + "step": 5795 + }, + { + "epoch": 4.6293929712460065, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 5796 + }, + { + "epoch": 4.630191693290735, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.0435, + "step": 5797 + }, + { + "epoch": 4.6309904153354635, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 5798 + }, + { + "epoch": 4.631789137380192, + "grad_norm": 0.1875, + "learning_rate": 0.0005, + "loss": 1.0443, + "step": 5799 + }, + { + "epoch": 4.63258785942492, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 5800 + }, + { + "epoch": 4.633386581469648, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.04, + "step": 5801 + }, + { + "epoch": 4.634185303514377, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0444, + "step": 5802 + }, + { + "epoch": 4.634984025559105, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0399, + "step": 5803 + }, + { + "epoch": 4.635782747603834, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 5804 + }, + { + "epoch": 4.636581469648562, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0405, + "step": 5805 + }, + { + "epoch": 4.637380191693291, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.0434, + "step": 5806 + }, + { + "epoch": 4.638178913738019, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 5807 + }, + { + "epoch": 4.638977635782748, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0434, + "step": 5808 + }, + { + "epoch": 4.6397763578274756, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0432, + "step": 5809 + }, + { + "epoch": 4.640575079872205, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0426, + "step": 5810 + }, + { + "epoch": 4.641373801916933, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0389, + "step": 5811 + }, + { + "epoch": 4.642172523961661, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 5812 + }, + { + "epoch": 4.64297124600639, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.043, + "step": 5813 + }, + { + "epoch": 4.643769968051118, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 5814 + }, + { + "epoch": 4.644568690095847, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 5815 + }, + { + "epoch": 4.645367412140575, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 5816 + }, + { + "epoch": 4.646166134185304, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0417, + "step": 5817 + }, + { + "epoch": 4.646964856230032, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0439, + "step": 5818 + }, + { + "epoch": 4.647763578274761, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0437, + "step": 5819 + }, + { + "epoch": 4.6485623003194885, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0375, + "step": 5820 + }, + { + "epoch": 4.649361022364217, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0379, + "step": 5821 + }, + { + "epoch": 4.6501597444089455, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 5822 + }, + { + "epoch": 4.650958466453674, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 5823 + }, + { + "epoch": 4.651757188498403, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0384, + "step": 5824 + }, + { + "epoch": 4.652555910543131, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 5825 + }, + { + "epoch": 4.65335463258786, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0403, + "step": 5826 + }, + { + "epoch": 4.654153354632588, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0425, + "step": 5827 + }, + { + "epoch": 4.654952076677317, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 5828 + }, + { + "epoch": 4.655750798722044, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.042, + "step": 5829 + }, + { + "epoch": 4.656549520766773, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.048, + "step": 5830 + }, + { + "epoch": 4.657348242811501, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0472, + "step": 5831 + }, + { + "epoch": 4.65814696485623, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0454, + "step": 5832 + }, + { + "epoch": 4.6589456869009584, + "grad_norm": 0.189453125, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 5833 + }, + { + "epoch": 4.659744408945687, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0435, + "step": 5834 + }, + { + "epoch": 4.6605431309904155, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.036, + "step": 5835 + }, + { + "epoch": 4.661341853035144, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0516, + "step": 5836 + }, + { + "epoch": 4.662140575079873, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 5837 + }, + { + "epoch": 4.6629392971246, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.055, + "step": 5838 + }, + { + "epoch": 4.663738019169329, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0402, + "step": 5839 + }, + { + "epoch": 4.664536741214057, + "grad_norm": 0.1943359375, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 5840 + }, + { + "epoch": 4.665335463258786, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.037, + "step": 5841 + }, + { + "epoch": 4.666134185303514, + "grad_norm": 0.23046875, + "learning_rate": 0.0005, + "loss": 1.0449, + "step": 5842 + }, + { + "epoch": 4.666932907348243, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 5843 + }, + { + "epoch": 4.667731629392971, + "grad_norm": 0.19921875, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 5844 + }, + { + "epoch": 4.6685303514377, + "grad_norm": 0.193359375, + "learning_rate": 0.0005, + "loss": 1.0384, + "step": 5845 + }, + { + "epoch": 4.669329073482428, + "grad_norm": 0.265625, + "learning_rate": 0.0005, + "loss": 1.0467, + "step": 5846 + }, + { + "epoch": 4.670127795527156, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0401, + "step": 5847 + }, + { + "epoch": 4.6709265175718855, + "grad_norm": 0.24609375, + "learning_rate": 0.0005, + "loss": 1.0394, + "step": 5848 + }, + { + "epoch": 4.671725239616613, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0396, + "step": 5849 + }, + { + "epoch": 4.672523961661342, + "grad_norm": 0.2001953125, + "learning_rate": 0.0005, + "loss": 1.0455, + "step": 5850 + }, + { + "epoch": 4.67332268370607, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.046, + "step": 5851 + }, + { + "epoch": 4.674121405750799, + "grad_norm": 0.1923828125, + "learning_rate": 0.0005, + "loss": 1.044, + "step": 5852 + }, + { + "epoch": 4.674920127795527, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0409, + "step": 5853 + }, + { + "epoch": 4.675718849840256, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 5854 + }, + { + "epoch": 4.676517571884984, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0522, + "step": 5855 + }, + { + "epoch": 4.677316293929713, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0426, + "step": 5856 + }, + { + "epoch": 4.678115015974441, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 5857 + }, + { + "epoch": 4.678913738019169, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0428, + "step": 5858 + }, + { + "epoch": 4.6797124600638975, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0467, + "step": 5859 + }, + { + "epoch": 4.680511182108626, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 5860 + }, + { + "epoch": 4.681309904153355, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0415, + "step": 5861 + }, + { + "epoch": 4.682108626198083, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0424, + "step": 5862 + }, + { + "epoch": 4.682907348242812, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.0448, + "step": 5863 + }, + { + "epoch": 4.68370607028754, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0456, + "step": 5864 + }, + { + "epoch": 4.684504792332269, + "grad_norm": 0.1630859375, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 5865 + }, + { + "epoch": 4.685303514376997, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0431, + "step": 5866 + }, + { + "epoch": 4.686102236421725, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 5867 + }, + { + "epoch": 4.686900958466453, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0439, + "step": 5868 + }, + { + "epoch": 4.687699680511182, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0443, + "step": 5869 + }, + { + "epoch": 4.68849840255591, + "grad_norm": 0.1669921875, + "learning_rate": 0.0005, + "loss": 1.0468, + "step": 5870 + }, + { + "epoch": 4.689297124600639, + "grad_norm": 0.26953125, + "learning_rate": 0.0005, + "loss": 1.04, + "step": 5871 + }, + { + "epoch": 4.6900958466453675, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 5872 + }, + { + "epoch": 4.690894568690096, + "grad_norm": 0.267578125, + "learning_rate": 0.0005, + "loss": 1.0373, + "step": 5873 + }, + { + "epoch": 4.6916932907348246, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.0384, + "step": 5874 + }, + { + "epoch": 4.692492012779553, + "grad_norm": 0.2314453125, + "learning_rate": 0.0005, + "loss": 1.0459, + "step": 5875 + }, + { + "epoch": 4.693290734824281, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.041, + "step": 5876 + }, + { + "epoch": 4.694089456869009, + "grad_norm": 0.24609375, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 5877 + }, + { + "epoch": 4.694888178913738, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0409, + "step": 5878 + }, + { + "epoch": 4.695686900958466, + "grad_norm": 0.216796875, + "learning_rate": 0.0005, + "loss": 1.0447, + "step": 5879 + }, + { + "epoch": 4.696485623003195, + "grad_norm": 0.173828125, + "learning_rate": 0.0005, + "loss": 1.0396, + "step": 5880 + }, + { + "epoch": 4.697284345047923, + "grad_norm": 0.28125, + "learning_rate": 0.0005, + "loss": 1.041, + "step": 5881 + }, + { + "epoch": 4.698083067092652, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 5882 + }, + { + "epoch": 4.69888178913738, + "grad_norm": 0.1904296875, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 5883 + }, + { + "epoch": 4.699680511182109, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 5884 + }, + { + "epoch": 4.700479233226837, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 5885 + }, + { + "epoch": 4.701277955271565, + "grad_norm": 0.171875, + "learning_rate": 0.0005, + "loss": 1.0498, + "step": 5886 + }, + { + "epoch": 4.702076677316294, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.049, + "step": 5887 + }, + { + "epoch": 4.702875399361022, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0408, + "step": 5888 + }, + { + "epoch": 4.703674121405751, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 5889 + }, + { + "epoch": 4.704472843450479, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 5890 + }, + { + "epoch": 4.705271565495208, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 5891 + }, + { + "epoch": 4.706070287539936, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0451, + "step": 5892 + }, + { + "epoch": 4.706869009584665, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0439, + "step": 5893 + }, + { + "epoch": 4.707667731629393, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0469, + "step": 5894 + }, + { + "epoch": 4.708466453674122, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 5895 + }, + { + "epoch": 4.7092651757188495, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 5896 + }, + { + "epoch": 4.710063897763578, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 5897 + }, + { + "epoch": 4.710862619808307, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0431, + "step": 5898 + }, + { + "epoch": 4.711661341853035, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0505, + "step": 5899 + }, + { + "epoch": 4.712460063897764, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 5900 + }, + { + "epoch": 4.713258785942492, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0466, + "step": 5901 + }, + { + "epoch": 4.714057507987221, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0409, + "step": 5902 + }, + { + "epoch": 4.714856230031949, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0436, + "step": 5903 + }, + { + "epoch": 4.715654952076678, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 5904 + }, + { + "epoch": 4.716453674121405, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0385, + "step": 5905 + }, + { + "epoch": 4.717252396166134, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0425, + "step": 5906 + }, + { + "epoch": 4.718051118210862, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.042, + "step": 5907 + }, + { + "epoch": 4.718849840255591, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0397, + "step": 5908 + }, + { + "epoch": 4.7196485623003195, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 5909 + }, + { + "epoch": 4.720447284345048, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 5910 + }, + { + "epoch": 4.7212460063897765, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 5911 + }, + { + "epoch": 4.722044728434505, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0391, + "step": 5912 + }, + { + "epoch": 4.722843450479234, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 5913 + }, + { + "epoch": 4.723642172523961, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.04, + "step": 5914 + }, + { + "epoch": 4.72444089456869, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0399, + "step": 5915 + }, + { + "epoch": 4.725239616613418, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 5916 + }, + { + "epoch": 4.726038338658147, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0391, + "step": 5917 + }, + { + "epoch": 4.726837060702875, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0454, + "step": 5918 + }, + { + "epoch": 4.727635782747604, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0445, + "step": 5919 + }, + { + "epoch": 4.728434504792332, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0454, + "step": 5920 + }, + { + "epoch": 4.729233226837061, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 5921 + }, + { + "epoch": 4.7300319488817895, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0447, + "step": 5922 + }, + { + "epoch": 4.730830670926517, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0427, + "step": 5923 + }, + { + "epoch": 4.731629392971246, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 5924 + }, + { + "epoch": 4.732428115015974, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0413, + "step": 5925 + }, + { + "epoch": 4.733226837060703, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 5926 + }, + { + "epoch": 4.734025559105431, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0457, + "step": 5927 + }, + { + "epoch": 4.73482428115016, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 5928 + }, + { + "epoch": 4.735623003194888, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0449, + "step": 5929 + }, + { + "epoch": 4.736421725239617, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0423, + "step": 5930 + }, + { + "epoch": 4.737220447284345, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 5931 + }, + { + "epoch": 4.738019169329074, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0396, + "step": 5932 + }, + { + "epoch": 4.738817891373802, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0448, + "step": 5933 + }, + { + "epoch": 4.73961661341853, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0424, + "step": 5934 + }, + { + "epoch": 4.7404153354632586, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0361, + "step": 5935 + }, + { + "epoch": 4.741214057507987, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0411, + "step": 5936 + }, + { + "epoch": 4.742012779552716, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0445, + "step": 5937 + }, + { + "epoch": 4.742811501597444, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0455, + "step": 5938 + }, + { + "epoch": 4.743610223642173, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0449, + "step": 5939 + }, + { + "epoch": 4.744408945686901, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 5940 + }, + { + "epoch": 4.74520766773163, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 5941 + }, + { + "epoch": 4.746006389776358, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 5942 + }, + { + "epoch": 4.746805111821086, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.0432, + "step": 5943 + }, + { + "epoch": 4.747603833865814, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0508, + "step": 5944 + }, + { + "epoch": 4.748402555910543, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0507, + "step": 5945 + }, + { + "epoch": 4.7492012779552715, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0403, + "step": 5946 + }, + { + "epoch": 4.75, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.038, + "step": 5947 + }, + { + "epoch": 4.7507987220447285, + "grad_norm": 0.19140625, + "learning_rate": 0.0005, + "loss": 1.0452, + "step": 5948 + }, + { + "epoch": 4.751597444089457, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0388, + "step": 5949 + }, + { + "epoch": 4.752396166134186, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0455, + "step": 5950 + }, + { + "epoch": 4.753194888178914, + "grad_norm": 0.16796875, + "learning_rate": 0.0005, + "loss": 1.0446, + "step": 5951 + }, + { + "epoch": 4.753993610223642, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 5952 + }, + { + "epoch": 4.75479233226837, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0394, + "step": 5953 + }, + { + "epoch": 4.755591054313099, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0431, + "step": 5954 + }, + { + "epoch": 4.756389776357827, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0413, + "step": 5955 + }, + { + "epoch": 4.757188498402556, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 5956 + }, + { + "epoch": 4.757987220447284, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 5957 + }, + { + "epoch": 4.758785942492013, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 5958 + }, + { + "epoch": 4.7595846645367414, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 5959 + }, + { + "epoch": 4.76038338658147, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 5960 + }, + { + "epoch": 4.761182108626198, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0552, + "step": 5961 + }, + { + "epoch": 4.761980830670926, + "grad_norm": 0.162109375, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 5962 + }, + { + "epoch": 4.762779552715655, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 5963 + }, + { + "epoch": 4.763578274760383, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0426, + "step": 5964 + }, + { + "epoch": 4.764376996805112, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.045, + "step": 5965 + }, + { + "epoch": 4.76517571884984, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0389, + "step": 5966 + }, + { + "epoch": 4.765974440894569, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0452, + "step": 5967 + }, + { + "epoch": 4.766773162939297, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0439, + "step": 5968 + }, + { + "epoch": 4.767571884984026, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0404, + "step": 5969 + }, + { + "epoch": 4.768370607028754, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0468, + "step": 5970 + }, + { + "epoch": 4.769169329073483, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 5971 + }, + { + "epoch": 4.7699680511182105, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0452, + "step": 5972 + }, + { + "epoch": 4.770766773162939, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 5973 + }, + { + "epoch": 4.771565495207668, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0511, + "step": 5974 + }, + { + "epoch": 4.772364217252396, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 5975 + }, + { + "epoch": 4.773162939297125, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 5976 + }, + { + "epoch": 4.773961661341853, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.041, + "step": 5977 + }, + { + "epoch": 4.774760383386582, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 5978 + }, + { + "epoch": 4.77555910543131, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0405, + "step": 5979 + }, + { + "epoch": 4.776357827476039, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 5980 + }, + { + "epoch": 4.777156549520766, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 5981 + }, + { + "epoch": 4.777955271565495, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 5982 + }, + { + "epoch": 4.7787539936102235, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0413, + "step": 5983 + }, + { + "epoch": 4.779552715654952, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0454, + "step": 5984 + }, + { + "epoch": 4.7803514376996805, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 5985 + }, + { + "epoch": 4.781150159744409, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.0417, + "step": 5986 + }, + { + "epoch": 4.781948881789138, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0405, + "step": 5987 + }, + { + "epoch": 4.782747603833866, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0407, + "step": 5988 + }, + { + "epoch": 4.783546325878595, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 5989 + }, + { + "epoch": 4.784345047923322, + "grad_norm": 0.2275390625, + "learning_rate": 0.0005, + "loss": 1.0452, + "step": 5990 + }, + { + "epoch": 4.785143769968051, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0449, + "step": 5991 + }, + { + "epoch": 4.785942492012779, + "grad_norm": 0.3125, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 5992 + }, + { + "epoch": 4.786741214057508, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0417, + "step": 5993 + }, + { + "epoch": 4.787539936102236, + "grad_norm": 0.2041015625, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 5994 + }, + { + "epoch": 4.788338658146965, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0542, + "step": 5995 + }, + { + "epoch": 4.789137380191693, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 5996 + }, + { + "epoch": 4.789936102236422, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0389, + "step": 5997 + }, + { + "epoch": 4.7907348242811505, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0417, + "step": 5998 + }, + { + "epoch": 4.791533546325878, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0425, + "step": 5999 + }, + { + "epoch": 4.792332268370607, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0405, + "step": 6000 + }, + { + "epoch": 4.793130990415335, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 6001 + }, + { + "epoch": 4.793929712460064, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.04, + "step": 6002 + }, + { + "epoch": 4.794728434504792, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 6003 + }, + { + "epoch": 4.795527156549521, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 6004 + }, + { + "epoch": 4.796325878594249, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 6005 + }, + { + "epoch": 4.797124600638978, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0394, + "step": 6006 + }, + { + "epoch": 4.797923322683706, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 6007 + }, + { + "epoch": 4.798722044728435, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0468, + "step": 6008 + }, + { + "epoch": 4.799520766773163, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0388, + "step": 6009 + }, + { + "epoch": 4.800319488817891, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.043, + "step": 6010 + }, + { + "epoch": 4.80111821086262, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0424, + "step": 6011 + }, + { + "epoch": 4.801916932907348, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0414, + "step": 6012 + }, + { + "epoch": 4.802715654952077, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.034, + "step": 6013 + }, + { + "epoch": 4.803514376996805, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0395, + "step": 6014 + }, + { + "epoch": 4.804313099041534, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.042, + "step": 6015 + }, + { + "epoch": 4.805111821086262, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 6016 + }, + { + "epoch": 4.805910543130991, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0476, + "step": 6017 + }, + { + "epoch": 4.806709265175719, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0403, + "step": 6018 + }, + { + "epoch": 4.807507987220447, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 6019 + }, + { + "epoch": 4.8083067092651754, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0375, + "step": 6020 + }, + { + "epoch": 4.809105431309904, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0435, + "step": 6021 + }, + { + "epoch": 4.8099041533546325, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0361, + "step": 6022 + }, + { + "epoch": 4.810702875399361, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0453, + "step": 6023 + }, + { + "epoch": 4.81150159744409, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0361, + "step": 6024 + }, + { + "epoch": 4.812300319488818, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0446, + "step": 6025 + }, + { + "epoch": 4.813099041533547, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0425, + "step": 6026 + }, + { + "epoch": 4.813897763578275, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0394, + "step": 6027 + }, + { + "epoch": 4.814696485623003, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 6028 + }, + { + "epoch": 4.815495207667731, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0427, + "step": 6029 + }, + { + "epoch": 4.81629392971246, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0455, + "step": 6030 + }, + { + "epoch": 4.817092651757188, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0412, + "step": 6031 + }, + { + "epoch": 4.817891373801917, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0386, + "step": 6032 + }, + { + "epoch": 4.818690095846645, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 6033 + }, + { + "epoch": 4.819488817891374, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0488, + "step": 6034 + }, + { + "epoch": 4.8202875399361025, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0395, + "step": 6035 + }, + { + "epoch": 4.821086261980831, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 6036 + }, + { + "epoch": 4.821884984025559, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 6037 + }, + { + "epoch": 4.822683706070287, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0372, + "step": 6038 + }, + { + "epoch": 4.823482428115016, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0428, + "step": 6039 + }, + { + "epoch": 4.824281150159744, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0375, + "step": 6040 + }, + { + "epoch": 4.825079872204473, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0441, + "step": 6041 + }, + { + "epoch": 4.825878594249201, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 6042 + }, + { + "epoch": 4.82667731629393, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 6043 + }, + { + "epoch": 4.827476038338658, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0409, + "step": 6044 + }, + { + "epoch": 4.828274760383387, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0433, + "step": 6045 + }, + { + "epoch": 4.8290734824281145, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0406, + "step": 6046 + }, + { + "epoch": 4.829872204472844, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0419, + "step": 6047 + }, + { + "epoch": 4.830670926517572, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 6048 + }, + { + "epoch": 4.8314696485623, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0428, + "step": 6049 + }, + { + "epoch": 4.832268370607029, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 6050 + }, + { + "epoch": 4.833067092651757, + "grad_norm": 0.17578125, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 6051 + }, + { + "epoch": 4.833865814696486, + "grad_norm": 0.1923828125, + "learning_rate": 0.0005, + "loss": 1.0402, + "step": 6052 + }, + { + "epoch": 4.834664536741214, + "grad_norm": 0.1923828125, + "learning_rate": 0.0005, + "loss": 1.044, + "step": 6053 + }, + { + "epoch": 4.835463258785943, + "grad_norm": 0.193359375, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 6054 + }, + { + "epoch": 4.836261980830671, + "grad_norm": 0.2158203125, + "learning_rate": 0.0005, + "loss": 1.0419, + "step": 6055 + }, + { + "epoch": 4.8370607028754, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0379, + "step": 6056 + }, + { + "epoch": 4.837859424920127, + "grad_norm": 0.2353515625, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 6057 + }, + { + "epoch": 4.838658146964856, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 6058 + }, + { + "epoch": 4.8394568690095845, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 6059 + }, + { + "epoch": 4.840255591054313, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.04, + "step": 6060 + }, + { + "epoch": 4.8410543130990416, + "grad_norm": 0.1669921875, + "learning_rate": 0.0005, + "loss": 1.0422, + "step": 6061 + }, + { + "epoch": 4.84185303514377, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 6062 + }, + { + "epoch": 4.842651757188499, + "grad_norm": 0.203125, + "learning_rate": 0.0005, + "loss": 1.0451, + "step": 6063 + }, + { + "epoch": 4.843450479233227, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 6064 + }, + { + "epoch": 4.844249201277956, + "grad_norm": 0.287109375, + "learning_rate": 0.0005, + "loss": 1.0412, + "step": 6065 + }, + { + "epoch": 4.845047923322683, + "grad_norm": 0.1669921875, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 6066 + }, + { + "epoch": 4.845846645367412, + "grad_norm": 0.267578125, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 6067 + }, + { + "epoch": 4.84664536741214, + "grad_norm": 0.173828125, + "learning_rate": 0.0005, + "loss": 1.0393, + "step": 6068 + }, + { + "epoch": 4.847444089456869, + "grad_norm": 0.263671875, + "learning_rate": 0.0005, + "loss": 1.0337, + "step": 6069 + }, + { + "epoch": 4.848242811501597, + "grad_norm": 0.2451171875, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 6070 + }, + { + "epoch": 4.849041533546326, + "grad_norm": 0.2138671875, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 6071 + }, + { + "epoch": 4.8498402555910545, + "grad_norm": 0.2890625, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 6072 + }, + { + "epoch": 4.850638977635783, + "grad_norm": 0.1689453125, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 6073 + }, + { + "epoch": 4.8514376996805115, + "grad_norm": 0.314453125, + "learning_rate": 0.0005, + "loss": 1.0451, + "step": 6074 + }, + { + "epoch": 4.852236421725239, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 6075 + }, + { + "epoch": 4.853035143769968, + "grad_norm": 0.306640625, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 6076 + }, + { + "epoch": 4.853833865814696, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0393, + "step": 6077 + }, + { + "epoch": 4.854632587859425, + "grad_norm": 0.2080078125, + "learning_rate": 0.0005, + "loss": 1.0437, + "step": 6078 + }, + { + "epoch": 4.855431309904153, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.04, + "step": 6079 + }, + { + "epoch": 4.856230031948882, + "grad_norm": 0.1875, + "learning_rate": 0.0005, + "loss": 1.0444, + "step": 6080 + }, + { + "epoch": 4.85702875399361, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 6081 + }, + { + "epoch": 4.857827476038339, + "grad_norm": 0.248046875, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 6082 + }, + { + "epoch": 4.858626198083067, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 6083 + }, + { + "epoch": 4.859424920127795, + "grad_norm": 0.32421875, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 6084 + }, + { + "epoch": 4.8602236421725244, + "grad_norm": 0.1640625, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 6085 + }, + { + "epoch": 4.861022364217252, + "grad_norm": 0.2412109375, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 6086 + }, + { + "epoch": 4.861821086261981, + "grad_norm": 0.1884765625, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 6087 + }, + { + "epoch": 4.862619808306709, + "grad_norm": 0.1884765625, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 6088 + }, + { + "epoch": 4.863418530351438, + "grad_norm": 0.181640625, + "learning_rate": 0.0005, + "loss": 1.0388, + "step": 6089 + }, + { + "epoch": 4.864217252396166, + "grad_norm": 0.19140625, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 6090 + }, + { + "epoch": 4.865015974440895, + "grad_norm": 0.205078125, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 6091 + }, + { + "epoch": 4.865814696485623, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0449, + "step": 6092 + }, + { + "epoch": 4.866613418530352, + "grad_norm": 0.2109375, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 6093 + }, + { + "epoch": 4.86741214057508, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0436, + "step": 6094 + }, + { + "epoch": 4.868210862619808, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 6095 + }, + { + "epoch": 4.8690095846645365, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.0445, + "step": 6096 + }, + { + "epoch": 4.869808306709265, + "grad_norm": 0.193359375, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 6097 + }, + { + "epoch": 4.8706070287539935, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 6098 + }, + { + "epoch": 4.871405750798722, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 6099 + }, + { + "epoch": 4.872204472843451, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 6100 + }, + { + "epoch": 4.873003194888179, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0427, + "step": 6101 + }, + { + "epoch": 4.873801916932908, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0469, + "step": 6102 + }, + { + "epoch": 4.874600638977636, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0391, + "step": 6103 + }, + { + "epoch": 4.875399361022364, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 6104 + }, + { + "epoch": 4.876198083067092, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0476, + "step": 6105 + }, + { + "epoch": 4.876996805111821, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 6106 + }, + { + "epoch": 4.877795527156549, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 6107 + }, + { + "epoch": 4.878594249201278, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0375, + "step": 6108 + }, + { + "epoch": 4.8793929712460065, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0436, + "step": 6109 + }, + { + "epoch": 4.880191693290735, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0362, + "step": 6110 + }, + { + "epoch": 4.8809904153354635, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0566, + "step": 6111 + }, + { + "epoch": 4.881789137380192, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 6112 + }, + { + "epoch": 4.88258785942492, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 6113 + }, + { + "epoch": 4.883386581469648, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 6114 + }, + { + "epoch": 4.884185303514377, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0376, + "step": 6115 + }, + { + "epoch": 4.884984025559105, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 6116 + }, + { + "epoch": 4.885782747603834, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0361, + "step": 6117 + }, + { + "epoch": 4.886581469648562, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0426, + "step": 6118 + }, + { + "epoch": 4.887380191693291, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0384, + "step": 6119 + }, + { + "epoch": 4.888178913738019, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 6120 + }, + { + "epoch": 4.888977635782748, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 6121 + }, + { + "epoch": 4.8897763578274756, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0444, + "step": 6122 + }, + { + "epoch": 4.890575079872205, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 6123 + }, + { + "epoch": 4.891373801916933, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.046, + "step": 6124 + }, + { + "epoch": 4.892172523961661, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0423, + "step": 6125 + }, + { + "epoch": 4.89297124600639, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 6126 + }, + { + "epoch": 4.893769968051118, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0459, + "step": 6127 + }, + { + "epoch": 4.894568690095847, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0354, + "step": 6128 + }, + { + "epoch": 4.895367412140575, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0424, + "step": 6129 + }, + { + "epoch": 4.896166134185304, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 6130 + }, + { + "epoch": 4.896964856230032, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0391, + "step": 6131 + }, + { + "epoch": 4.897763578274761, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.036, + "step": 6132 + }, + { + "epoch": 4.8985623003194885, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 6133 + }, + { + "epoch": 4.899361022364217, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0421, + "step": 6134 + }, + { + "epoch": 4.9001597444089455, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0412, + "step": 6135 + }, + { + "epoch": 4.900958466453674, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0415, + "step": 6136 + }, + { + "epoch": 4.901757188498403, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 6137 + }, + { + "epoch": 4.902555910543131, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 6138 + }, + { + "epoch": 4.90335463258786, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0385, + "step": 6139 + }, + { + "epoch": 4.904153354632588, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0338, + "step": 6140 + }, + { + "epoch": 4.904952076677317, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0384, + "step": 6141 + }, + { + "epoch": 4.905750798722044, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 6142 + }, + { + "epoch": 4.906549520766773, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 6143 + }, + { + "epoch": 4.907348242811501, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0448, + "step": 6144 + }, + { + "epoch": 4.90814696485623, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 6145 + }, + { + "epoch": 4.9089456869009584, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0404, + "step": 6146 + }, + { + "epoch": 4.909744408945687, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 6147 + }, + { + "epoch": 4.9105431309904155, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 6148 + }, + { + "epoch": 4.911341853035144, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 6149 + }, + { + "epoch": 4.912140575079873, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0413, + "step": 6150 + }, + { + "epoch": 4.9129392971246, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0455, + "step": 6151 + }, + { + "epoch": 4.913738019169329, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0488, + "step": 6152 + }, + { + "epoch": 4.914536741214057, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 6153 + }, + { + "epoch": 4.915335463258786, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 6154 + }, + { + "epoch": 4.916134185303514, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 6155 + }, + { + "epoch": 4.916932907348243, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 6156 + }, + { + "epoch": 4.917731629392971, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.043, + "step": 6157 + }, + { + "epoch": 4.9185303514377, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.045, + "step": 6158 + }, + { + "epoch": 4.919329073482428, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 6159 + }, + { + "epoch": 4.920127795527156, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 6160 + }, + { + "epoch": 4.9209265175718855, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 6161 + }, + { + "epoch": 4.921725239616613, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 6162 + }, + { + "epoch": 4.922523961661342, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 6163 + }, + { + "epoch": 4.92332268370607, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0569, + "step": 6164 + }, + { + "epoch": 4.924121405750799, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0405, + "step": 6165 + }, + { + "epoch": 4.924920127795527, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0457, + "step": 6166 + }, + { + "epoch": 4.925718849840256, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 6167 + }, + { + "epoch": 4.926517571884984, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 6168 + }, + { + "epoch": 4.927316293929713, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 6169 + }, + { + "epoch": 4.928115015974441, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0388, + "step": 6170 + }, + { + "epoch": 4.928913738019169, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 6171 + }, + { + "epoch": 4.9297124600638975, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0472, + "step": 6172 + }, + { + "epoch": 4.930511182108626, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 6173 + }, + { + "epoch": 4.931309904153355, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 6174 + }, + { + "epoch": 4.932108626198083, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0436, + "step": 6175 + }, + { + "epoch": 4.932907348242812, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0393, + "step": 6176 + }, + { + "epoch": 4.93370607028754, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 6177 + }, + { + "epoch": 4.934504792332269, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0395, + "step": 6178 + }, + { + "epoch": 4.935303514376997, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0378, + "step": 6179 + }, + { + "epoch": 4.936102236421725, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 6180 + }, + { + "epoch": 4.936900958466453, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0411, + "step": 6181 + }, + { + "epoch": 4.937699680511182, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0409, + "step": 6182 + }, + { + "epoch": 4.93849840255591, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0375, + "step": 6183 + }, + { + "epoch": 4.939297124600639, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 6184 + }, + { + "epoch": 4.9400958466453675, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 6185 + }, + { + "epoch": 4.940894568690096, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 6186 + }, + { + "epoch": 4.9416932907348246, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0408, + "step": 6187 + }, + { + "epoch": 4.942492012779553, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 6188 + }, + { + "epoch": 4.943290734824281, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0459, + "step": 6189 + }, + { + "epoch": 4.944089456869009, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 6190 + }, + { + "epoch": 4.944888178913738, + "grad_norm": 0.1640625, + "learning_rate": 0.0005, + "loss": 1.0397, + "step": 6191 + }, + { + "epoch": 4.945686900958466, + "grad_norm": 0.2060546875, + "learning_rate": 0.0005, + "loss": 1.0423, + "step": 6192 + }, + { + "epoch": 4.946485623003195, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0361, + "step": 6193 + }, + { + "epoch": 4.947284345047923, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 6194 + }, + { + "epoch": 4.948083067092652, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0436, + "step": 6195 + }, + { + "epoch": 4.94888178913738, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 6196 + }, + { + "epoch": 4.949680511182109, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 6197 + }, + { + "epoch": 4.950479233226837, + "grad_norm": 0.17578125, + "learning_rate": 0.0005, + "loss": 1.0431, + "step": 6198 + }, + { + "epoch": 4.951277955271565, + "grad_norm": 0.2470703125, + "learning_rate": 0.0005, + "loss": 1.042, + "step": 6199 + }, + { + "epoch": 4.952076677316294, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 6200 + }, + { + "epoch": 4.952875399361022, + "grad_norm": 0.30078125, + "learning_rate": 0.0005, + "loss": 1.044, + "step": 6201 + }, + { + "epoch": 4.953674121405751, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0378, + "step": 6202 + }, + { + "epoch": 4.954472843450479, + "grad_norm": 0.291015625, + "learning_rate": 0.0005, + "loss": 1.0472, + "step": 6203 + }, + { + "epoch": 4.955271565495208, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0353, + "step": 6204 + }, + { + "epoch": 4.956070287539936, + "grad_norm": 0.251953125, + "learning_rate": 0.0005, + "loss": 1.0429, + "step": 6205 + }, + { + "epoch": 4.956869009584665, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0369, + "step": 6206 + }, + { + "epoch": 4.957667731629393, + "grad_norm": 0.263671875, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 6207 + }, + { + "epoch": 4.958466453674122, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.041, + "step": 6208 + }, + { + "epoch": 4.9592651757188495, + "grad_norm": 0.3125, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 6209 + }, + { + "epoch": 4.960063897763578, + "grad_norm": 0.1669921875, + "learning_rate": 0.0005, + "loss": 1.0419, + "step": 6210 + }, + { + "epoch": 4.960862619808307, + "grad_norm": 0.390625, + "learning_rate": 0.0005, + "loss": 1.0523, + "step": 6211 + }, + { + "epoch": 4.961661341853035, + "grad_norm": 0.1787109375, + "learning_rate": 0.0005, + "loss": 1.0409, + "step": 6212 + }, + { + "epoch": 4.962460063897764, + "grad_norm": 0.62890625, + "learning_rate": 0.0005, + "loss": 1.0386, + "step": 6213 + }, + { + "epoch": 4.963258785942492, + "grad_norm": 0.392578125, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 6214 + }, + { + "epoch": 4.964057507987221, + "grad_norm": 0.36328125, + "learning_rate": 0.0005, + "loss": 1.0448, + "step": 6215 + }, + { + "epoch": 4.964856230031949, + "grad_norm": 0.88671875, + "learning_rate": 0.0005, + "loss": 1.0449, + "step": 6216 + }, + { + "epoch": 4.965654952076678, + "grad_norm": 0.59765625, + "learning_rate": 0.0005, + "loss": 1.0536, + "step": 6217 + }, + { + "epoch": 4.966453674121405, + "grad_norm": 0.376953125, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 6218 + }, + { + "epoch": 4.967252396166134, + "grad_norm": 0.9453125, + "learning_rate": 0.0005, + "loss": 1.0512, + "step": 6219 + }, + { + "epoch": 4.968051118210862, + "grad_norm": 0.359375, + "learning_rate": 0.0005, + "loss": 1.0534, + "step": 6220 + }, + { + "epoch": 4.968849840255591, + "grad_norm": 0.484375, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 6221 + }, + { + "epoch": 4.9696485623003195, + "grad_norm": 0.3671875, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 6222 + }, + { + "epoch": 4.970447284345048, + "grad_norm": 0.3828125, + "learning_rate": 0.0005, + "loss": 1.0417, + "step": 6223 + }, + { + "epoch": 4.9712460063897765, + "grad_norm": 0.55859375, + "learning_rate": 0.0005, + "loss": 1.0407, + "step": 6224 + }, + { + "epoch": 4.972044728434505, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0441, + "step": 6225 + }, + { + "epoch": 4.972843450479234, + "grad_norm": 0.4609375, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 6226 + }, + { + "epoch": 4.973642172523961, + "grad_norm": 0.162109375, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 6227 + }, + { + "epoch": 4.97444089456869, + "grad_norm": 0.52734375, + "learning_rate": 0.0005, + "loss": 1.0492, + "step": 6228 + }, + { + "epoch": 4.975239616613418, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0443, + "step": 6229 + }, + { + "epoch": 4.976038338658147, + "grad_norm": 0.56640625, + "learning_rate": 0.0005, + "loss": 1.0556, + "step": 6230 + }, + { + "epoch": 4.976837060702875, + "grad_norm": 0.16015625, + "learning_rate": 0.0005, + "loss": 1.044, + "step": 6231 + }, + { + "epoch": 4.977635782747604, + "grad_norm": 0.515625, + "learning_rate": 0.0005, + "loss": 1.0551, + "step": 6232 + }, + { + "epoch": 4.978434504792332, + "grad_norm": 0.18359375, + "learning_rate": 0.0005, + "loss": 1.0423, + "step": 6233 + }, + { + "epoch": 4.979233226837061, + "grad_norm": 0.470703125, + "learning_rate": 0.0005, + "loss": 1.0458, + "step": 6234 + }, + { + "epoch": 4.9800319488817895, + "grad_norm": 0.1630859375, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 6235 + }, + { + "epoch": 4.980830670926517, + "grad_norm": 0.34765625, + "learning_rate": 0.0005, + "loss": 1.0468, + "step": 6236 + }, + { + "epoch": 4.981629392971246, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.0437, + "step": 6237 + }, + { + "epoch": 4.982428115015974, + "grad_norm": 0.248046875, + "learning_rate": 0.0005, + "loss": 1.0396, + "step": 6238 + }, + { + "epoch": 4.983226837060703, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0373, + "step": 6239 + }, + { + "epoch": 4.984025559105431, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0428, + "step": 6240 + }, + { + "epoch": 4.98482428115016, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.0402, + "step": 6241 + }, + { + "epoch": 4.985623003194888, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0397, + "step": 6242 + }, + { + "epoch": 4.986421725239617, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 6243 + }, + { + "epoch": 4.987220447284345, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 6244 + }, + { + "epoch": 4.988019169329074, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 6245 + }, + { + "epoch": 4.988817891373802, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 6246 + }, + { + "epoch": 4.98961661341853, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0389, + "step": 6247 + }, + { + "epoch": 4.9904153354632586, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0389, + "step": 6248 + }, + { + "epoch": 4.991214057507987, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 6249 + }, + { + "epoch": 4.992012779552716, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0545, + "step": 6250 + }, + { + "epoch": 4.992811501597444, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0485, + "step": 6251 + }, + { + "epoch": 4.993610223642173, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0419, + "step": 6252 + }, + { + "epoch": 4.994408945686901, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 6253 + }, + { + "epoch": 4.99520766773163, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0436, + "step": 6254 + }, + { + "epoch": 4.996006389776358, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0375, + "step": 6255 + }, + { + "epoch": 4.996805111821086, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0454, + "step": 6256 + }, + { + "epoch": 4.997603833865814, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0448, + "step": 6257 + }, + { + "epoch": 4.998402555910543, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 6258 + }, + { + "epoch": 4.9992012779552715, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 6259 + }, + { + "epoch": 5.0, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0401, + "step": 6260 + }, + { + "epoch": 5.0007987220447285, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 6261 + }, + { + "epoch": 5.001597444089457, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0348, + "step": 6262 + }, + { + "epoch": 5.002396166134186, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.039, + "step": 6263 + }, + { + "epoch": 5.003194888178914, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0393, + "step": 6264 + }, + { + "epoch": 5.003993610223642, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0337, + "step": 6265 + }, + { + "epoch": 5.00479233226837, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0384, + "step": 6266 + }, + { + "epoch": 5.005591054313099, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 6267 + }, + { + "epoch": 5.006389776357827, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.0395, + "step": 6268 + }, + { + "epoch": 5.007188498402556, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 6269 + }, + { + "epoch": 5.007987220447284, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0376, + "step": 6270 + }, + { + "epoch": 5.008785942492013, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 6271 + }, + { + "epoch": 5.0095846645367414, + "grad_norm": 0.16796875, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 6272 + }, + { + "epoch": 5.01038338658147, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.04, + "step": 6273 + }, + { + "epoch": 5.0111821086261985, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0407, + "step": 6274 + }, + { + "epoch": 5.011980830670926, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0407, + "step": 6275 + }, + { + "epoch": 5.012779552715655, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0421, + "step": 6276 + }, + { + "epoch": 5.013578274760383, + "grad_norm": 0.21484375, + "learning_rate": 0.0005, + "loss": 1.0433, + "step": 6277 + }, + { + "epoch": 5.014376996805112, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0447, + "step": 6278 + }, + { + "epoch": 5.01517571884984, + "grad_norm": 0.1982421875, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 6279 + }, + { + "epoch": 5.015974440894569, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 6280 + }, + { + "epoch": 5.016773162939297, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 6281 + }, + { + "epoch": 5.017571884984026, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 6282 + }, + { + "epoch": 5.018370607028754, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 6283 + }, + { + "epoch": 5.019169329073482, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0396, + "step": 6284 + }, + { + "epoch": 5.0199680511182105, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.048, + "step": 6285 + }, + { + "epoch": 5.020766773162939, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 6286 + }, + { + "epoch": 5.021565495207668, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 6287 + }, + { + "epoch": 5.022364217252396, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 6288 + }, + { + "epoch": 5.023162939297125, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 6289 + }, + { + "epoch": 5.023961661341853, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 6290 + }, + { + "epoch": 5.024760383386582, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0405, + "step": 6291 + }, + { + "epoch": 5.02555910543131, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 6292 + }, + { + "epoch": 5.026357827476039, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 6293 + }, + { + "epoch": 5.027156549520766, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 6294 + }, + { + "epoch": 5.027955271565495, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 6295 + }, + { + "epoch": 5.0287539936102235, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0465, + "step": 6296 + }, + { + "epoch": 5.029552715654952, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0466, + "step": 6297 + }, + { + "epoch": 5.0303514376996805, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 6298 + }, + { + "epoch": 5.031150159744409, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.042, + "step": 6299 + }, + { + "epoch": 5.031948881789138, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 6300 + }, + { + "epoch": 5.032747603833866, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0375, + "step": 6301 + }, + { + "epoch": 5.033546325878595, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 6302 + }, + { + "epoch": 5.034345047923322, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0428, + "step": 6303 + }, + { + "epoch": 5.035143769968051, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0412, + "step": 6304 + }, + { + "epoch": 5.035942492012779, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0399, + "step": 6305 + }, + { + "epoch": 5.036741214057508, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0436, + "step": 6306 + }, + { + "epoch": 5.037539936102236, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0396, + "step": 6307 + }, + { + "epoch": 5.038338658146965, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.04, + "step": 6308 + }, + { + "epoch": 5.039137380191693, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 6309 + }, + { + "epoch": 5.039936102236422, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 6310 + }, + { + "epoch": 5.0407348242811505, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0451, + "step": 6311 + }, + { + "epoch": 5.041533546325879, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0445, + "step": 6312 + }, + { + "epoch": 5.042332268370607, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0459, + "step": 6313 + }, + { + "epoch": 5.043130990415335, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 6314 + }, + { + "epoch": 5.043929712460064, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 6315 + }, + { + "epoch": 5.044728434504792, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 6316 + }, + { + "epoch": 5.045527156549521, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0416, + "step": 6317 + }, + { + "epoch": 5.046325878594249, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0524, + "step": 6318 + }, + { + "epoch": 5.047124600638978, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 6319 + }, + { + "epoch": 5.047923322683706, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0394, + "step": 6320 + }, + { + "epoch": 5.048722044728435, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0435, + "step": 6321 + }, + { + "epoch": 5.0495207667731625, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.038, + "step": 6322 + }, + { + "epoch": 5.050319488817891, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0402, + "step": 6323 + }, + { + "epoch": 5.05111821086262, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0388, + "step": 6324 + }, + { + "epoch": 5.051916932907348, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0421, + "step": 6325 + }, + { + "epoch": 5.052715654952077, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0399, + "step": 6326 + }, + { + "epoch": 5.053514376996805, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0354, + "step": 6327 + }, + { + "epoch": 5.054313099041534, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0442, + "step": 6328 + }, + { + "epoch": 5.055111821086262, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 6329 + }, + { + "epoch": 5.055910543130991, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0433, + "step": 6330 + }, + { + "epoch": 5.056709265175719, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 6331 + }, + { + "epoch": 5.057507987220447, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.039, + "step": 6332 + }, + { + "epoch": 5.0583067092651754, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 6333 + }, + { + "epoch": 5.059105431309904, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0385, + "step": 6334 + }, + { + "epoch": 5.0599041533546325, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 6335 + }, + { + "epoch": 5.060702875399361, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 6336 + }, + { + "epoch": 5.06150159744409, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0441, + "step": 6337 + }, + { + "epoch": 5.062300319488818, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0428, + "step": 6338 + }, + { + "epoch": 5.063099041533547, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0467, + "step": 6339 + }, + { + "epoch": 5.063897763578275, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0454, + "step": 6340 + }, + { + "epoch": 5.064696485623003, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0426, + "step": 6341 + }, + { + "epoch": 5.065495207667731, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0372, + "step": 6342 + }, + { + "epoch": 5.06629392971246, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0394, + "step": 6343 + }, + { + "epoch": 5.067092651757188, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0415, + "step": 6344 + }, + { + "epoch": 5.067891373801917, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 6345 + }, + { + "epoch": 5.068690095846645, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 6346 + }, + { + "epoch": 5.069488817891374, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.039, + "step": 6347 + }, + { + "epoch": 5.0702875399361025, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0408, + "step": 6348 + }, + { + "epoch": 5.071086261980831, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 6349 + }, + { + "epoch": 5.0718849840255595, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.04, + "step": 6350 + }, + { + "epoch": 5.072683706070287, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 6351 + }, + { + "epoch": 5.073482428115016, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0401, + "step": 6352 + }, + { + "epoch": 5.074281150159744, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0466, + "step": 6353 + }, + { + "epoch": 5.075079872204473, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 6354 + }, + { + "epoch": 5.075878594249201, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0416, + "step": 6355 + }, + { + "epoch": 5.07667731629393, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0394, + "step": 6356 + }, + { + "epoch": 5.077476038338658, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 6357 + }, + { + "epoch": 5.078274760383387, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0433, + "step": 6358 + }, + { + "epoch": 5.079073482428115, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0459, + "step": 6359 + }, + { + "epoch": 5.079872204472843, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0426, + "step": 6360 + }, + { + "epoch": 5.080670926517572, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 6361 + }, + { + "epoch": 5.0814696485623, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 6362 + }, + { + "epoch": 5.082268370607029, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 6363 + }, + { + "epoch": 5.083067092651757, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0432, + "step": 6364 + }, + { + "epoch": 5.083865814696486, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 6365 + }, + { + "epoch": 5.084664536741214, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 6366 + }, + { + "epoch": 5.085463258785943, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0424, + "step": 6367 + }, + { + "epoch": 5.086261980830671, + "grad_norm": 0.240234375, + "learning_rate": 0.0005, + "loss": 1.0453, + "step": 6368 + }, + { + "epoch": 5.0870607028754, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0411, + "step": 6369 + }, + { + "epoch": 5.087859424920127, + "grad_norm": 0.1572265625, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 6370 + }, + { + "epoch": 5.088658146964856, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 6371 + }, + { + "epoch": 5.0894568690095845, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.0469, + "step": 6372 + }, + { + "epoch": 5.090255591054313, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 6373 + }, + { + "epoch": 5.0910543130990416, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0399, + "step": 6374 + }, + { + "epoch": 5.09185303514377, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 6375 + }, + { + "epoch": 5.092651757188499, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 6376 + }, + { + "epoch": 5.093450479233227, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0577, + "step": 6377 + }, + { + "epoch": 5.094249201277956, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0443, + "step": 6378 + }, + { + "epoch": 5.095047923322683, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 6379 + }, + { + "epoch": 5.095846645367412, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 6380 + }, + { + "epoch": 5.09664536741214, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.039, + "step": 6381 + }, + { + "epoch": 5.097444089456869, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0411, + "step": 6382 + }, + { + "epoch": 5.098242811501597, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0414, + "step": 6383 + }, + { + "epoch": 5.099041533546326, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0424, + "step": 6384 + }, + { + "epoch": 5.0998402555910545, + "grad_norm": 0.1640625, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 6385 + }, + { + "epoch": 5.100638977635783, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 6386 + }, + { + "epoch": 5.1014376996805115, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0401, + "step": 6387 + }, + { + "epoch": 5.102236421725239, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0411, + "step": 6388 + }, + { + "epoch": 5.103035143769968, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0386, + "step": 6389 + }, + { + "epoch": 5.103833865814696, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 6390 + }, + { + "epoch": 5.104632587859425, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0391, + "step": 6391 + }, + { + "epoch": 5.105431309904153, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 6392 + }, + { + "epoch": 5.106230031948882, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0423, + "step": 6393 + }, + { + "epoch": 5.10702875399361, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0372, + "step": 6394 + }, + { + "epoch": 5.107827476038339, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0396, + "step": 6395 + }, + { + "epoch": 5.108626198083067, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 6396 + }, + { + "epoch": 5.109424920127796, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0468, + "step": 6397 + }, + { + "epoch": 5.110223642172524, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0429, + "step": 6398 + }, + { + "epoch": 5.111022364217252, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0447, + "step": 6399 + }, + { + "epoch": 5.111821086261981, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 6400 + }, + { + "epoch": 5.112619808306709, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0406, + "step": 6401 + }, + { + "epoch": 5.113418530351438, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0459, + "step": 6402 + }, + { + "epoch": 5.114217252396166, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 6403 + }, + { + "epoch": 5.115015974440895, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0372, + "step": 6404 + }, + { + "epoch": 5.115814696485623, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0399, + "step": 6405 + }, + { + "epoch": 5.116613418530352, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0434, + "step": 6406 + }, + { + "epoch": 5.11741214057508, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0425, + "step": 6407 + }, + { + "epoch": 5.118210862619808, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0372, + "step": 6408 + }, + { + "epoch": 5.1190095846645365, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 6409 + }, + { + "epoch": 5.119808306709265, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 6410 + }, + { + "epoch": 5.1206070287539935, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0426, + "step": 6411 + }, + { + "epoch": 5.121405750798722, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 6412 + }, + { + "epoch": 5.122204472843451, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 6413 + }, + { + "epoch": 5.123003194888179, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0439, + "step": 6414 + }, + { + "epoch": 5.123801916932908, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 6415 + }, + { + "epoch": 5.124600638977636, + "grad_norm": 0.185546875, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 6416 + }, + { + "epoch": 5.125399361022364, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0388, + "step": 6417 + }, + { + "epoch": 5.126198083067092, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0385, + "step": 6418 + }, + { + "epoch": 5.126996805111821, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 6419 + }, + { + "epoch": 5.127795527156549, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 6420 + }, + { + "epoch": 5.128594249201278, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 6421 + }, + { + "epoch": 5.1293929712460065, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 6422 + }, + { + "epoch": 5.130191693290735, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 6423 + }, + { + "epoch": 5.1309904153354635, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0401, + "step": 6424 + }, + { + "epoch": 5.131789137380192, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0403, + "step": 6425 + }, + { + "epoch": 5.13258785942492, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0414, + "step": 6426 + }, + { + "epoch": 5.133386581469648, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0433, + "step": 6427 + }, + { + "epoch": 5.134185303514377, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 6428 + }, + { + "epoch": 5.134984025559105, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0373, + "step": 6429 + }, + { + "epoch": 5.135782747603834, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 6430 + }, + { + "epoch": 5.136581469648562, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0448, + "step": 6431 + }, + { + "epoch": 5.137380191693291, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0409, + "step": 6432 + }, + { + "epoch": 5.138178913738019, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 6433 + }, + { + "epoch": 5.138977635782748, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0378, + "step": 6434 + }, + { + "epoch": 5.139776357827476, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0408, + "step": 6435 + }, + { + "epoch": 5.140575079872204, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0408, + "step": 6436 + }, + { + "epoch": 5.141373801916933, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 6437 + }, + { + "epoch": 5.142172523961661, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0393, + "step": 6438 + }, + { + "epoch": 5.14297124600639, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0369, + "step": 6439 + }, + { + "epoch": 5.143769968051118, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0386, + "step": 6440 + }, + { + "epoch": 5.144568690095847, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 6441 + }, + { + "epoch": 5.145367412140575, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0431, + "step": 6442 + }, + { + "epoch": 5.146166134185304, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.046, + "step": 6443 + }, + { + "epoch": 5.146964856230032, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 6444 + }, + { + "epoch": 5.147763578274761, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0421, + "step": 6445 + }, + { + "epoch": 5.1485623003194885, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 6446 + }, + { + "epoch": 5.149361022364217, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 6447 + }, + { + "epoch": 5.1501597444089455, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0543, + "step": 6448 + }, + { + "epoch": 5.150958466453674, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 6449 + }, + { + "epoch": 5.151757188498403, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 6450 + }, + { + "epoch": 5.152555910543131, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0375, + "step": 6451 + }, + { + "epoch": 5.15335463258786, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0395, + "step": 6452 + }, + { + "epoch": 5.154153354632588, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0403, + "step": 6453 + }, + { + "epoch": 5.154952076677317, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 6454 + }, + { + "epoch": 5.155750798722044, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 6455 + }, + { + "epoch": 5.156549520766773, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0376, + "step": 6456 + }, + { + "epoch": 5.157348242811501, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0407, + "step": 6457 + }, + { + "epoch": 5.15814696485623, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0408, + "step": 6458 + }, + { + "epoch": 5.1589456869009584, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0404, + "step": 6459 + }, + { + "epoch": 5.159744408945687, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 6460 + }, + { + "epoch": 5.1605431309904155, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0435, + "step": 6461 + }, + { + "epoch": 5.161341853035144, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.042, + "step": 6462 + }, + { + "epoch": 5.162140575079873, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0428, + "step": 6463 + }, + { + "epoch": 5.1629392971246, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.042, + "step": 6464 + }, + { + "epoch": 5.163738019169329, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0416, + "step": 6465 + }, + { + "epoch": 5.164536741214057, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0396, + "step": 6466 + }, + { + "epoch": 5.165335463258786, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.042, + "step": 6467 + }, + { + "epoch": 5.166134185303514, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0386, + "step": 6468 + }, + { + "epoch": 5.166932907348243, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0421, + "step": 6469 + }, + { + "epoch": 5.167731629392971, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 6470 + }, + { + "epoch": 5.1685303514377, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 6471 + }, + { + "epoch": 5.169329073482428, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0453, + "step": 6472 + }, + { + "epoch": 5.170127795527157, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0424, + "step": 6473 + }, + { + "epoch": 5.170926517571885, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 6474 + }, + { + "epoch": 5.171725239616613, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0397, + "step": 6475 + }, + { + "epoch": 5.172523961661342, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 6476 + }, + { + "epoch": 5.17332268370607, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.039, + "step": 6477 + }, + { + "epoch": 5.174121405750799, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0459, + "step": 6478 + }, + { + "epoch": 5.174920127795527, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 6479 + }, + { + "epoch": 5.175718849840256, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 6480 + }, + { + "epoch": 5.176517571884984, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0494, + "step": 6481 + }, + { + "epoch": 5.177316293929713, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 6482 + }, + { + "epoch": 5.178115015974441, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0514, + "step": 6483 + }, + { + "epoch": 5.178913738019169, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0518, + "step": 6484 + }, + { + "epoch": 5.1797124600638975, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0444, + "step": 6485 + }, + { + "epoch": 5.180511182108626, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 6486 + }, + { + "epoch": 5.181309904153355, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0348, + "step": 6487 + }, + { + "epoch": 5.182108626198083, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0439, + "step": 6488 + }, + { + "epoch": 5.182907348242812, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 6489 + }, + { + "epoch": 5.18370607028754, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0412, + "step": 6490 + }, + { + "epoch": 5.184504792332269, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0436, + "step": 6491 + }, + { + "epoch": 5.185303514376997, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0481, + "step": 6492 + }, + { + "epoch": 5.186102236421725, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.041, + "step": 6493 + }, + { + "epoch": 5.186900958466453, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 6494 + }, + { + "epoch": 5.187699680511182, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0389, + "step": 6495 + }, + { + "epoch": 5.18849840255591, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.037, + "step": 6496 + }, + { + "epoch": 5.189297124600639, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0447, + "step": 6497 + }, + { + "epoch": 5.1900958466453675, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 6498 + }, + { + "epoch": 5.190894568690096, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0446, + "step": 6499 + }, + { + "epoch": 5.1916932907348246, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0415, + "step": 6500 + }, + { + "epoch": 5.192492012779553, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 6501 + }, + { + "epoch": 5.193290734824281, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0361, + "step": 6502 + }, + { + "epoch": 5.194089456869009, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0376, + "step": 6503 + }, + { + "epoch": 5.194888178913738, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0467, + "step": 6504 + }, + { + "epoch": 5.195686900958466, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0452, + "step": 6505 + }, + { + "epoch": 5.196485623003195, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 6506 + }, + { + "epoch": 5.197284345047923, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0377, + "step": 6507 + }, + { + "epoch": 5.198083067092652, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.043, + "step": 6508 + }, + { + "epoch": 5.19888178913738, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0464, + "step": 6509 + }, + { + "epoch": 5.199680511182109, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 6510 + }, + { + "epoch": 5.2004792332268375, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0384, + "step": 6511 + }, + { + "epoch": 5.201277955271565, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0338, + "step": 6512 + }, + { + "epoch": 5.202076677316294, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0353, + "step": 6513 + }, + { + "epoch": 5.202875399361022, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 6514 + }, + { + "epoch": 5.203674121405751, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0416, + "step": 6515 + }, + { + "epoch": 5.204472843450479, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 6516 + }, + { + "epoch": 5.205271565495208, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0389, + "step": 6517 + }, + { + "epoch": 5.206070287539936, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0399, + "step": 6518 + }, + { + "epoch": 5.206869009584665, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 6519 + }, + { + "epoch": 5.207667731629393, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0414, + "step": 6520 + }, + { + "epoch": 5.208466453674121, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 6521 + }, + { + "epoch": 5.2092651757188495, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 6522 + }, + { + "epoch": 5.210063897763578, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 6523 + }, + { + "epoch": 5.210862619808307, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0386, + "step": 6524 + }, + { + "epoch": 5.211661341853035, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 6525 + }, + { + "epoch": 5.212460063897764, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0407, + "step": 6526 + }, + { + "epoch": 5.213258785942492, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0405, + "step": 6527 + }, + { + "epoch": 5.214057507987221, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0426, + "step": 6528 + }, + { + "epoch": 5.214856230031949, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0496, + "step": 6529 + }, + { + "epoch": 5.215654952076678, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0468, + "step": 6530 + }, + { + "epoch": 5.216453674121405, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 6531 + }, + { + "epoch": 5.217252396166134, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0411, + "step": 6532 + }, + { + "epoch": 5.218051118210862, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 6533 + }, + { + "epoch": 5.218849840255591, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0384, + "step": 6534 + }, + { + "epoch": 5.2196485623003195, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0394, + "step": 6535 + }, + { + "epoch": 5.220447284345048, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0406, + "step": 6536 + }, + { + "epoch": 5.2212460063897765, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0436, + "step": 6537 + }, + { + "epoch": 5.222044728434505, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 6538 + }, + { + "epoch": 5.222843450479234, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 6539 + }, + { + "epoch": 5.223642172523961, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0354, + "step": 6540 + }, + { + "epoch": 5.22444089456869, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.042, + "step": 6541 + }, + { + "epoch": 5.225239616613418, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 6542 + }, + { + "epoch": 5.226038338658147, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 6543 + }, + { + "epoch": 5.226837060702875, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 6544 + }, + { + "epoch": 5.227635782747604, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 6545 + }, + { + "epoch": 5.228434504792332, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 6546 + }, + { + "epoch": 5.229233226837061, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.044, + "step": 6547 + }, + { + "epoch": 5.2300319488817895, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0431, + "step": 6548 + }, + { + "epoch": 5.230830670926518, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 6549 + }, + { + "epoch": 5.231629392971246, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 6550 + }, + { + "epoch": 5.232428115015974, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0338, + "step": 6551 + }, + { + "epoch": 5.233226837060703, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0377, + "step": 6552 + }, + { + "epoch": 5.234025559105431, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 6553 + }, + { + "epoch": 5.23482428115016, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 6554 + }, + { + "epoch": 5.235623003194888, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 6555 + }, + { + "epoch": 5.236421725239617, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 6556 + }, + { + "epoch": 5.237220447284345, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 6557 + }, + { + "epoch": 5.238019169329074, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0406, + "step": 6558 + }, + { + "epoch": 5.2388178913738015, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 6559 + }, + { + "epoch": 5.23961661341853, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 6560 + }, + { + "epoch": 5.2404153354632586, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 6561 + }, + { + "epoch": 5.241214057507987, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0421, + "step": 6562 + }, + { + "epoch": 5.242012779552716, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0421, + "step": 6563 + }, + { + "epoch": 5.242811501597444, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0468, + "step": 6564 + }, + { + "epoch": 5.243610223642173, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0396, + "step": 6565 + }, + { + "epoch": 5.244408945686901, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0444, + "step": 6566 + }, + { + "epoch": 5.24520766773163, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 6567 + }, + { + "epoch": 5.246006389776358, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 6568 + }, + { + "epoch": 5.246805111821086, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 6569 + }, + { + "epoch": 5.247603833865814, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0454, + "step": 6570 + }, + { + "epoch": 5.248402555910543, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0517, + "step": 6571 + }, + { + "epoch": 5.2492012779552715, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 6572 + }, + { + "epoch": 5.25, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0402, + "step": 6573 + }, + { + "epoch": 5.2507987220447285, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0389, + "step": 6574 + }, + { + "epoch": 5.251597444089457, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 6575 + }, + { + "epoch": 5.252396166134186, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0337, + "step": 6576 + }, + { + "epoch": 5.253194888178914, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 6577 + }, + { + "epoch": 5.253993610223642, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0417, + "step": 6578 + }, + { + "epoch": 5.25479233226837, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0372, + "step": 6579 + }, + { + "epoch": 5.255591054313099, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 6580 + }, + { + "epoch": 5.256389776357827, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0434, + "step": 6581 + }, + { + "epoch": 5.257188498402556, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 6582 + }, + { + "epoch": 5.257987220447284, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0441, + "step": 6583 + }, + { + "epoch": 5.258785942492013, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 6584 + }, + { + "epoch": 5.2595846645367414, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0393, + "step": 6585 + }, + { + "epoch": 5.26038338658147, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 6586 + }, + { + "epoch": 5.261182108626198, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0429, + "step": 6587 + }, + { + "epoch": 5.261980830670926, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.0386, + "step": 6588 + }, + { + "epoch": 5.262779552715655, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0454, + "step": 6589 + }, + { + "epoch": 5.263578274760383, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 6590 + }, + { + "epoch": 5.264376996805112, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 6591 + }, + { + "epoch": 5.26517571884984, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0406, + "step": 6592 + }, + { + "epoch": 5.265974440894569, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.0427, + "step": 6593 + }, + { + "epoch": 5.266773162939297, + "grad_norm": 0.1767578125, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 6594 + }, + { + "epoch": 5.267571884984026, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 6595 + }, + { + "epoch": 5.268370607028754, + "grad_norm": 0.248046875, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 6596 + }, + { + "epoch": 5.269169329073483, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 6597 + }, + { + "epoch": 5.2699680511182105, + "grad_norm": 0.2255859375, + "learning_rate": 0.0005, + "loss": 1.0408, + "step": 6598 + }, + { + "epoch": 5.270766773162939, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 6599 + }, + { + "epoch": 5.271565495207668, + "grad_norm": 0.296875, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 6600 + }, + { + "epoch": 5.272364217252396, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0362, + "step": 6601 + }, + { + "epoch": 5.273162939297125, + "grad_norm": 0.3359375, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 6602 + }, + { + "epoch": 5.273961661341853, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.043, + "step": 6603 + }, + { + "epoch": 5.274760383386582, + "grad_norm": 0.28125, + "learning_rate": 0.0005, + "loss": 1.052, + "step": 6604 + }, + { + "epoch": 5.27555910543131, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0406, + "step": 6605 + }, + { + "epoch": 5.276357827476039, + "grad_norm": 0.2314453125, + "learning_rate": 0.0005, + "loss": 1.0487, + "step": 6606 + }, + { + "epoch": 5.277156549520766, + "grad_norm": 0.2138671875, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 6607 + }, + { + "epoch": 5.277955271565495, + "grad_norm": 0.32421875, + "learning_rate": 0.0005, + "loss": 1.0419, + "step": 6608 + }, + { + "epoch": 5.2787539936102235, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.0467, + "step": 6609 + }, + { + "epoch": 5.279552715654952, + "grad_norm": 0.37109375, + "learning_rate": 0.0005, + "loss": 1.0519, + "step": 6610 + }, + { + "epoch": 5.2803514376996805, + "grad_norm": 0.1533203125, + "learning_rate": 0.0005, + "loss": 1.038, + "step": 6611 + }, + { + "epoch": 5.281150159744409, + "grad_norm": 0.486328125, + "learning_rate": 0.0005, + "loss": 1.0437, + "step": 6612 + }, + { + "epoch": 5.281948881789138, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 6613 + }, + { + "epoch": 5.282747603833866, + "grad_norm": 0.41796875, + "learning_rate": 0.0005, + "loss": 1.0435, + "step": 6614 + }, + { + "epoch": 5.283546325878595, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 6615 + }, + { + "epoch": 5.284345047923322, + "grad_norm": 0.46875, + "learning_rate": 0.0005, + "loss": 1.0411, + "step": 6616 + }, + { + "epoch": 5.285143769968051, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0469, + "step": 6617 + }, + { + "epoch": 5.285942492012779, + "grad_norm": 0.328125, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 6618 + }, + { + "epoch": 5.286741214057508, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 6619 + }, + { + "epoch": 5.287539936102236, + "grad_norm": 0.1689453125, + "learning_rate": 0.0005, + "loss": 1.0407, + "step": 6620 + }, + { + "epoch": 5.288338658146965, + "grad_norm": 0.1845703125, + "learning_rate": 0.0005, + "loss": 1.0443, + "step": 6621 + }, + { + "epoch": 5.289137380191693, + "grad_norm": 0.2431640625, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 6622 + }, + { + "epoch": 5.289936102236422, + "grad_norm": 0.240234375, + "learning_rate": 0.0005, + "loss": 1.0414, + "step": 6623 + }, + { + "epoch": 5.2907348242811505, + "grad_norm": 0.373046875, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 6624 + }, + { + "epoch": 5.291533546325878, + "grad_norm": 0.1826171875, + "learning_rate": 0.0005, + "loss": 1.0458, + "step": 6625 + }, + { + "epoch": 5.292332268370607, + "grad_norm": 0.431640625, + "learning_rate": 0.0005, + "loss": 1.0482, + "step": 6626 + }, + { + "epoch": 5.293130990415335, + "grad_norm": 0.16015625, + "learning_rate": 0.0005, + "loss": 1.0377, + "step": 6627 + }, + { + "epoch": 5.293929712460064, + "grad_norm": 0.55078125, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 6628 + }, + { + "epoch": 5.294728434504792, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0409, + "step": 6629 + }, + { + "epoch": 5.295527156549521, + "grad_norm": 0.40625, + "learning_rate": 0.0005, + "loss": 1.0422, + "step": 6630 + }, + { + "epoch": 5.296325878594249, + "grad_norm": 0.1982421875, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 6631 + }, + { + "epoch": 5.297124600638978, + "grad_norm": 0.578125, + "learning_rate": 0.0005, + "loss": 1.04, + "step": 6632 + }, + { + "epoch": 5.297923322683706, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0445, + "step": 6633 + }, + { + "epoch": 5.298722044728435, + "grad_norm": 0.61328125, + "learning_rate": 0.0005, + "loss": 1.0428, + "step": 6634 + }, + { + "epoch": 5.2995207667731625, + "grad_norm": 0.216796875, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 6635 + }, + { + "epoch": 5.300319488817891, + "grad_norm": 0.30859375, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 6636 + }, + { + "epoch": 5.30111821086262, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 6637 + }, + { + "epoch": 5.301916932907348, + "grad_norm": 0.19921875, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 6638 + }, + { + "epoch": 5.302715654952077, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 6639 + }, + { + "epoch": 5.303514376996805, + "grad_norm": 0.1982421875, + "learning_rate": 0.0005, + "loss": 1.0427, + "step": 6640 + }, + { + "epoch": 5.304313099041534, + "grad_norm": 0.1640625, + "learning_rate": 0.0005, + "loss": 1.0403, + "step": 6641 + }, + { + "epoch": 5.305111821086262, + "grad_norm": 0.1533203125, + "learning_rate": 0.0005, + "loss": 1.0401, + "step": 6642 + }, + { + "epoch": 5.305910543130991, + "grad_norm": 0.1865234375, + "learning_rate": 0.0005, + "loss": 1.0395, + "step": 6643 + }, + { + "epoch": 5.306709265175719, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 6644 + }, + { + "epoch": 5.307507987220447, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0362, + "step": 6645 + }, + { + "epoch": 5.3083067092651754, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 6646 + }, + { + "epoch": 5.309105431309904, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0361, + "step": 6647 + }, + { + "epoch": 5.3099041533546325, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0385, + "step": 6648 + }, + { + "epoch": 5.310702875399361, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0423, + "step": 6649 + }, + { + "epoch": 5.31150159744409, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 6650 + }, + { + "epoch": 5.312300319488818, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0448, + "step": 6651 + }, + { + "epoch": 5.313099041533547, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0385, + "step": 6652 + }, + { + "epoch": 5.313897763578275, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 6653 + }, + { + "epoch": 5.314696485623003, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0442, + "step": 6654 + }, + { + "epoch": 5.315495207667731, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0403, + "step": 6655 + }, + { + "epoch": 5.31629392971246, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 6656 + }, + { + "epoch": 5.317092651757188, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 6657 + }, + { + "epoch": 5.317891373801917, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 6658 + }, + { + "epoch": 5.318690095846645, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0433, + "step": 6659 + }, + { + "epoch": 5.319488817891374, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0379, + "step": 6660 + }, + { + "epoch": 5.3202875399361025, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 6661 + }, + { + "epoch": 5.321086261980831, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0407, + "step": 6662 + }, + { + "epoch": 5.321884984025559, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0443, + "step": 6663 + }, + { + "epoch": 5.322683706070287, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.042, + "step": 6664 + }, + { + "epoch": 5.323482428115016, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0373, + "step": 6665 + }, + { + "epoch": 5.324281150159744, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0421, + "step": 6666 + }, + { + "epoch": 5.325079872204473, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0378, + "step": 6667 + }, + { + "epoch": 5.325878594249201, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 6668 + }, + { + "epoch": 5.32667731629393, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0442, + "step": 6669 + }, + { + "epoch": 5.327476038338658, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 6670 + }, + { + "epoch": 5.328274760383387, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 6671 + }, + { + "epoch": 5.329073482428115, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 6672 + }, + { + "epoch": 5.329872204472843, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0391, + "step": 6673 + }, + { + "epoch": 5.330670926517572, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0399, + "step": 6674 + }, + { + "epoch": 5.3314696485623, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0431, + "step": 6675 + }, + { + "epoch": 5.332268370607029, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 6676 + }, + { + "epoch": 5.333067092651757, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0424, + "step": 6677 + }, + { + "epoch": 5.333865814696486, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 6678 + }, + { + "epoch": 5.334664536741214, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 6679 + }, + { + "epoch": 5.335463258785943, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 6680 + }, + { + "epoch": 5.336261980830671, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0369, + "step": 6681 + }, + { + "epoch": 5.3370607028754, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 6682 + }, + { + "epoch": 5.337859424920127, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 6683 + }, + { + "epoch": 5.338658146964856, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 6684 + }, + { + "epoch": 5.3394568690095845, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 6685 + }, + { + "epoch": 5.340255591054313, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.044, + "step": 6686 + }, + { + "epoch": 5.3410543130990416, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 6687 + }, + { + "epoch": 5.34185303514377, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0402, + "step": 6688 + }, + { + "epoch": 5.342651757188499, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0419, + "step": 6689 + }, + { + "epoch": 5.343450479233227, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 6690 + }, + { + "epoch": 5.344249201277956, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 6691 + }, + { + "epoch": 5.345047923322683, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 6692 + }, + { + "epoch": 5.345846645367412, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0375, + "step": 6693 + }, + { + "epoch": 5.34664536741214, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0432, + "step": 6694 + }, + { + "epoch": 5.347444089456869, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0409, + "step": 6695 + }, + { + "epoch": 5.348242811501597, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0388, + "step": 6696 + }, + { + "epoch": 5.349041533546326, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0457, + "step": 6697 + }, + { + "epoch": 5.3498402555910545, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.04, + "step": 6698 + }, + { + "epoch": 5.350638977635783, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0439, + "step": 6699 + }, + { + "epoch": 5.3514376996805115, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0469, + "step": 6700 + }, + { + "epoch": 5.352236421725239, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 6701 + }, + { + "epoch": 5.353035143769968, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0408, + "step": 6702 + }, + { + "epoch": 5.353833865814696, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0426, + "step": 6703 + }, + { + "epoch": 5.354632587859425, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0444, + "step": 6704 + }, + { + "epoch": 5.355431309904153, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0448, + "step": 6705 + }, + { + "epoch": 5.356230031948882, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0435, + "step": 6706 + }, + { + "epoch": 5.35702875399361, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 6707 + }, + { + "epoch": 5.357827476038339, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0354, + "step": 6708 + }, + { + "epoch": 5.358626198083067, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0405, + "step": 6709 + }, + { + "epoch": 5.359424920127796, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 6710 + }, + { + "epoch": 5.360223642172524, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 6711 + }, + { + "epoch": 5.361022364217252, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0432, + "step": 6712 + }, + { + "epoch": 5.361821086261981, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 6713 + }, + { + "epoch": 5.362619808306709, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0443, + "step": 6714 + }, + { + "epoch": 5.363418530351438, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0416, + "step": 6715 + }, + { + "epoch": 5.364217252396166, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 6716 + }, + { + "epoch": 5.365015974440895, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0402, + "step": 6717 + }, + { + "epoch": 5.365814696485623, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0445, + "step": 6718 + }, + { + "epoch": 5.366613418530352, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 6719 + }, + { + "epoch": 5.36741214057508, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0406, + "step": 6720 + }, + { + "epoch": 5.368210862619808, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0425, + "step": 6721 + }, + { + "epoch": 5.3690095846645365, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 6722 + }, + { + "epoch": 5.369808306709265, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0506, + "step": 6723 + }, + { + "epoch": 5.3706070287539935, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0432, + "step": 6724 + }, + { + "epoch": 5.371405750798722, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 6725 + }, + { + "epoch": 5.372204472843451, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 6726 + }, + { + "epoch": 5.373003194888179, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0482, + "step": 6727 + }, + { + "epoch": 5.373801916932908, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0354, + "step": 6728 + }, + { + "epoch": 5.374600638977636, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 6729 + }, + { + "epoch": 5.375399361022364, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 6730 + }, + { + "epoch": 5.376198083067092, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 6731 + }, + { + "epoch": 5.376996805111821, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0416, + "step": 6732 + }, + { + "epoch": 5.377795527156549, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 6733 + }, + { + "epoch": 5.378594249201278, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 6734 + }, + { + "epoch": 5.3793929712460065, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 6735 + }, + { + "epoch": 5.380191693290735, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.042, + "step": 6736 + }, + { + "epoch": 5.3809904153354635, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0411, + "step": 6737 + }, + { + "epoch": 5.381789137380192, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 6738 + }, + { + "epoch": 5.38258785942492, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0402, + "step": 6739 + }, + { + "epoch": 5.383386581469648, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 6740 + }, + { + "epoch": 5.384185303514377, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0442, + "step": 6741 + }, + { + "epoch": 5.384984025559105, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0386, + "step": 6742 + }, + { + "epoch": 5.385782747603834, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 6743 + }, + { + "epoch": 5.386581469648562, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0447, + "step": 6744 + }, + { + "epoch": 5.387380191693291, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0393, + "step": 6745 + }, + { + "epoch": 5.388178913738019, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 6746 + }, + { + "epoch": 5.388977635782748, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0424, + "step": 6747 + }, + { + "epoch": 5.389776357827476, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0399, + "step": 6748 + }, + { + "epoch": 5.390575079872204, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0397, + "step": 6749 + }, + { + "epoch": 5.391373801916933, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0445, + "step": 6750 + }, + { + "epoch": 5.392172523961661, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 6751 + }, + { + "epoch": 5.39297124600639, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 6752 + }, + { + "epoch": 5.393769968051118, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.039, + "step": 6753 + }, + { + "epoch": 5.394568690095847, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0375, + "step": 6754 + }, + { + "epoch": 5.395367412140575, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 6755 + }, + { + "epoch": 5.396166134185304, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0372, + "step": 6756 + }, + { + "epoch": 5.396964856230032, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 6757 + }, + { + "epoch": 5.397763578274761, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 6758 + }, + { + "epoch": 5.3985623003194885, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0408, + "step": 6759 + }, + { + "epoch": 5.399361022364217, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0385, + "step": 6760 + }, + { + "epoch": 5.4001597444089455, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 6761 + }, + { + "epoch": 5.400958466453674, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 6762 + }, + { + "epoch": 5.401757188498403, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0428, + "step": 6763 + }, + { + "epoch": 5.402555910543131, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 6764 + }, + { + "epoch": 5.40335463258786, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 6765 + }, + { + "epoch": 5.404153354632588, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 6766 + }, + { + "epoch": 5.404952076677317, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0376, + "step": 6767 + }, + { + "epoch": 5.405750798722044, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 6768 + }, + { + "epoch": 5.406549520766773, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0423, + "step": 6769 + }, + { + "epoch": 5.407348242811501, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 6770 + }, + { + "epoch": 5.40814696485623, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 6771 + }, + { + "epoch": 5.4089456869009584, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0423, + "step": 6772 + }, + { + "epoch": 5.409744408945687, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.036, + "step": 6773 + }, + { + "epoch": 5.4105431309904155, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0425, + "step": 6774 + }, + { + "epoch": 5.411341853035144, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0412, + "step": 6775 + }, + { + "epoch": 5.412140575079873, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 6776 + }, + { + "epoch": 5.4129392971246, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 6777 + }, + { + "epoch": 5.413738019169329, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.038, + "step": 6778 + }, + { + "epoch": 5.414536741214057, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 6779 + }, + { + "epoch": 5.415335463258786, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0429, + "step": 6780 + }, + { + "epoch": 5.416134185303514, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 6781 + }, + { + "epoch": 5.416932907348243, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 6782 + }, + { + "epoch": 5.417731629392971, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 6783 + }, + { + "epoch": 5.4185303514377, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.04, + "step": 6784 + }, + { + "epoch": 5.419329073482428, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 6785 + }, + { + "epoch": 5.420127795527157, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 6786 + }, + { + "epoch": 5.420926517571885, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 6787 + }, + { + "epoch": 5.421725239616613, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.042, + "step": 6788 + }, + { + "epoch": 5.422523961661342, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 6789 + }, + { + "epoch": 5.42332268370607, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 6790 + }, + { + "epoch": 5.424121405750799, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 6791 + }, + { + "epoch": 5.424920127795527, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0399, + "step": 6792 + }, + { + "epoch": 5.425718849840256, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 6793 + }, + { + "epoch": 5.426517571884984, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0443, + "step": 6794 + }, + { + "epoch": 5.427316293929713, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 6795 + }, + { + "epoch": 5.428115015974441, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 6796 + }, + { + "epoch": 5.428913738019169, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0464, + "step": 6797 + }, + { + "epoch": 5.4297124600638975, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 6798 + }, + { + "epoch": 5.430511182108626, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0404, + "step": 6799 + }, + { + "epoch": 5.431309904153355, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 6800 + }, + { + "epoch": 5.432108626198083, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 6801 + }, + { + "epoch": 5.432907348242812, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0385, + "step": 6802 + }, + { + "epoch": 5.43370607028754, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 6803 + }, + { + "epoch": 5.434504792332269, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 6804 + }, + { + "epoch": 5.435303514376997, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 6805 + }, + { + "epoch": 5.436102236421725, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 6806 + }, + { + "epoch": 5.436900958466453, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0426, + "step": 6807 + }, + { + "epoch": 5.437699680511182, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0406, + "step": 6808 + }, + { + "epoch": 5.43849840255591, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 6809 + }, + { + "epoch": 5.439297124600639, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0384, + "step": 6810 + }, + { + "epoch": 5.4400958466453675, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0419, + "step": 6811 + }, + { + "epoch": 5.440894568690096, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0447, + "step": 6812 + }, + { + "epoch": 5.4416932907348246, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 6813 + }, + { + "epoch": 5.442492012779553, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0403, + "step": 6814 + }, + { + "epoch": 5.443290734824281, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.042, + "step": 6815 + }, + { + "epoch": 5.444089456869009, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0369, + "step": 6816 + }, + { + "epoch": 5.444888178913738, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 6817 + }, + { + "epoch": 5.445686900958466, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.045, + "step": 6818 + }, + { + "epoch": 5.446485623003195, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 6819 + }, + { + "epoch": 5.447284345047923, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0425, + "step": 6820 + }, + { + "epoch": 5.448083067092652, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 6821 + }, + { + "epoch": 5.44888178913738, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0379, + "step": 6822 + }, + { + "epoch": 5.449680511182109, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.042, + "step": 6823 + }, + { + "epoch": 5.4504792332268375, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0389, + "step": 6824 + }, + { + "epoch": 5.451277955271565, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 6825 + }, + { + "epoch": 5.452076677316294, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 6826 + }, + { + "epoch": 5.452875399361022, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 6827 + }, + { + "epoch": 5.453674121405751, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 6828 + }, + { + "epoch": 5.454472843450479, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0453, + "step": 6829 + }, + { + "epoch": 5.455271565495208, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 6830 + }, + { + "epoch": 5.456070287539936, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 6831 + }, + { + "epoch": 5.456869009584665, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 6832 + }, + { + "epoch": 5.457667731629393, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 6833 + }, + { + "epoch": 5.458466453674122, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 6834 + }, + { + "epoch": 5.4592651757188495, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 6835 + }, + { + "epoch": 5.460063897763578, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 6836 + }, + { + "epoch": 5.460862619808307, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 6837 + }, + { + "epoch": 5.461661341853035, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 6838 + }, + { + "epoch": 5.462460063897764, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0413, + "step": 6839 + }, + { + "epoch": 5.463258785942492, + "grad_norm": 0.166015625, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 6840 + }, + { + "epoch": 5.464057507987221, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 6841 + }, + { + "epoch": 5.464856230031949, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0396, + "step": 6842 + }, + { + "epoch": 5.465654952076678, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 6843 + }, + { + "epoch": 5.466453674121405, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 6844 + }, + { + "epoch": 5.467252396166134, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 6845 + }, + { + "epoch": 5.468051118210862, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0389, + "step": 6846 + }, + { + "epoch": 5.468849840255591, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0413, + "step": 6847 + }, + { + "epoch": 5.4696485623003195, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 6848 + }, + { + "epoch": 5.470447284345048, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0395, + "step": 6849 + }, + { + "epoch": 5.4712460063897765, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 6850 + }, + { + "epoch": 5.472044728434505, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0408, + "step": 6851 + }, + { + "epoch": 5.472843450479234, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0529, + "step": 6852 + }, + { + "epoch": 5.473642172523961, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0391, + "step": 6853 + }, + { + "epoch": 5.47444089456869, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.042, + "step": 6854 + }, + { + "epoch": 5.475239616613418, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 6855 + }, + { + "epoch": 5.476038338658147, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 6856 + }, + { + "epoch": 5.476837060702875, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0388, + "step": 6857 + }, + { + "epoch": 5.477635782747604, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0423, + "step": 6858 + }, + { + "epoch": 5.478434504792332, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 6859 + }, + { + "epoch": 5.479233226837061, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0454, + "step": 6860 + }, + { + "epoch": 5.4800319488817895, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0377, + "step": 6861 + }, + { + "epoch": 5.480830670926517, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0372, + "step": 6862 + }, + { + "epoch": 5.481629392971246, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0439, + "step": 6863 + }, + { + "epoch": 5.482428115015974, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 6864 + }, + { + "epoch": 5.483226837060703, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0384, + "step": 6865 + }, + { + "epoch": 5.484025559105431, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0455, + "step": 6866 + }, + { + "epoch": 5.48482428115016, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 6867 + }, + { + "epoch": 5.485623003194888, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 6868 + }, + { + "epoch": 5.486421725239617, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 6869 + }, + { + "epoch": 5.487220447284345, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0375, + "step": 6870 + }, + { + "epoch": 5.488019169329074, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 6871 + }, + { + "epoch": 5.488817891373802, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 6872 + }, + { + "epoch": 5.48961661341853, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0412, + "step": 6873 + }, + { + "epoch": 5.4904153354632586, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0412, + "step": 6874 + }, + { + "epoch": 5.491214057507987, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0391, + "step": 6875 + }, + { + "epoch": 5.492012779552716, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0402, + "step": 6876 + }, + { + "epoch": 5.492811501597444, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0361, + "step": 6877 + }, + { + "epoch": 5.493610223642173, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 6878 + }, + { + "epoch": 5.494408945686901, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0373, + "step": 6879 + }, + { + "epoch": 5.49520766773163, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0415, + "step": 6880 + }, + { + "epoch": 5.496006389776358, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0403, + "step": 6881 + }, + { + "epoch": 5.496805111821086, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0432, + "step": 6882 + }, + { + "epoch": 5.497603833865814, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0521, + "step": 6883 + }, + { + "epoch": 5.498402555910543, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 6884 + }, + { + "epoch": 5.4992012779552715, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0377, + "step": 6885 + }, + { + "epoch": 5.5, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0338, + "step": 6886 + }, + { + "epoch": 5.5007987220447285, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0348, + "step": 6887 + }, + { + "epoch": 5.501597444089457, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0427, + "step": 6888 + }, + { + "epoch": 5.502396166134186, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0389, + "step": 6889 + }, + { + "epoch": 5.503194888178914, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0407, + "step": 6890 + }, + { + "epoch": 5.503993610223642, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.041, + "step": 6891 + }, + { + "epoch": 5.50479233226837, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0446, + "step": 6892 + }, + { + "epoch": 5.505591054313099, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0406, + "step": 6893 + }, + { + "epoch": 5.506389776357827, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 6894 + }, + { + "epoch": 5.507188498402556, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 6895 + }, + { + "epoch": 5.507987220447284, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 6896 + }, + { + "epoch": 5.508785942492013, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 6897 + }, + { + "epoch": 5.5095846645367414, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0411, + "step": 6898 + }, + { + "epoch": 5.51038338658147, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0504, + "step": 6899 + }, + { + "epoch": 5.511182108626198, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0348, + "step": 6900 + }, + { + "epoch": 5.511980830670926, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0391, + "step": 6901 + }, + { + "epoch": 5.512779552715655, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0448, + "step": 6902 + }, + { + "epoch": 5.513578274760383, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 6903 + }, + { + "epoch": 5.514376996805112, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0509, + "step": 6904 + }, + { + "epoch": 5.51517571884984, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0455, + "step": 6905 + }, + { + "epoch": 5.515974440894569, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 6906 + }, + { + "epoch": 5.516773162939297, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 6907 + }, + { + "epoch": 5.517571884984026, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 6908 + }, + { + "epoch": 5.518370607028754, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 6909 + }, + { + "epoch": 5.519169329073483, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.038, + "step": 6910 + }, + { + "epoch": 5.5199680511182105, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0405, + "step": 6911 + }, + { + "epoch": 5.520766773162939, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0396, + "step": 6912 + }, + { + "epoch": 5.521565495207668, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 6913 + }, + { + "epoch": 5.522364217252396, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 6914 + }, + { + "epoch": 5.523162939297125, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0401, + "step": 6915 + }, + { + "epoch": 5.523961661341853, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 6916 + }, + { + "epoch": 5.524760383386582, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 6917 + }, + { + "epoch": 5.52555910543131, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.04, + "step": 6918 + }, + { + "epoch": 5.526357827476039, + "grad_norm": 0.169921875, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 6919 + }, + { + "epoch": 5.527156549520766, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0429, + "step": 6920 + }, + { + "epoch": 5.527955271565495, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0408, + "step": 6921 + }, + { + "epoch": 5.5287539936102235, + "grad_norm": 0.162109375, + "learning_rate": 0.0005, + "loss": 1.0415, + "step": 6922 + }, + { + "epoch": 5.529552715654952, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0361, + "step": 6923 + }, + { + "epoch": 5.5303514376996805, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0414, + "step": 6924 + }, + { + "epoch": 5.531150159744409, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 6925 + }, + { + "epoch": 5.531948881789138, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0376, + "step": 6926 + }, + { + "epoch": 5.532747603833866, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0361, + "step": 6927 + }, + { + "epoch": 5.533546325878595, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 6928 + }, + { + "epoch": 5.534345047923322, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 6929 + }, + { + "epoch": 5.535143769968051, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 6930 + }, + { + "epoch": 5.535942492012779, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0501, + "step": 6931 + }, + { + "epoch": 5.536741214057508, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0422, + "step": 6932 + }, + { + "epoch": 5.537539936102236, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0421, + "step": 6933 + }, + { + "epoch": 5.538338658146965, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0435, + "step": 6934 + }, + { + "epoch": 5.539137380191693, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0402, + "step": 6935 + }, + { + "epoch": 5.539936102236422, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0456, + "step": 6936 + }, + { + "epoch": 5.5407348242811505, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 6937 + }, + { + "epoch": 5.541533546325878, + "grad_norm": 0.1748046875, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 6938 + }, + { + "epoch": 5.542332268370607, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0372, + "step": 6939 + }, + { + "epoch": 5.543130990415335, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0439, + "step": 6940 + }, + { + "epoch": 5.543929712460064, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0454, + "step": 6941 + }, + { + "epoch": 5.544728434504792, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0468, + "step": 6942 + }, + { + "epoch": 5.545527156549521, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0446, + "step": 6943 + }, + { + "epoch": 5.546325878594249, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0396, + "step": 6944 + }, + { + "epoch": 5.547124600638978, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 6945 + }, + { + "epoch": 5.547923322683706, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 6946 + }, + { + "epoch": 5.548722044728435, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 6947 + }, + { + "epoch": 5.549520766773163, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 6948 + }, + { + "epoch": 5.550319488817891, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 6949 + }, + { + "epoch": 5.55111821086262, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0412, + "step": 6950 + }, + { + "epoch": 5.551916932907348, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0486, + "step": 6951 + }, + { + "epoch": 5.552715654952077, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 6952 + }, + { + "epoch": 5.553514376996805, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 6953 + }, + { + "epoch": 5.554313099041534, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0399, + "step": 6954 + }, + { + "epoch": 5.555111821086262, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 6955 + }, + { + "epoch": 5.555910543130991, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0404, + "step": 6956 + }, + { + "epoch": 5.556709265175719, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 6957 + }, + { + "epoch": 5.557507987220447, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 6958 + }, + { + "epoch": 5.5583067092651754, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 6959 + }, + { + "epoch": 5.559105431309904, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 6960 + }, + { + "epoch": 5.5599041533546325, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0391, + "step": 6961 + }, + { + "epoch": 5.560702875399361, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.04, + "step": 6962 + }, + { + "epoch": 5.56150159744409, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 6963 + }, + { + "epoch": 5.562300319488818, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0338, + "step": 6964 + }, + { + "epoch": 5.563099041533547, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0439, + "step": 6965 + }, + { + "epoch": 5.563897763578275, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0439, + "step": 6966 + }, + { + "epoch": 5.564696485623003, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0393, + "step": 6967 + }, + { + "epoch": 5.565495207667731, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 6968 + }, + { + "epoch": 5.56629392971246, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 6969 + }, + { + "epoch": 5.567092651757188, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 6970 + }, + { + "epoch": 5.567891373801917, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 6971 + }, + { + "epoch": 5.568690095846645, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 6972 + }, + { + "epoch": 5.569488817891374, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.045, + "step": 6973 + }, + { + "epoch": 5.5702875399361025, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0361, + "step": 6974 + }, + { + "epoch": 5.571086261980831, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 6975 + }, + { + "epoch": 5.571884984025559, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0406, + "step": 6976 + }, + { + "epoch": 5.572683706070287, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 6977 + }, + { + "epoch": 5.573482428115016, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0377, + "step": 6978 + }, + { + "epoch": 5.574281150159744, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 6979 + }, + { + "epoch": 5.575079872204473, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0377, + "step": 6980 + }, + { + "epoch": 5.575878594249201, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 6981 + }, + { + "epoch": 5.57667731629393, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 6982 + }, + { + "epoch": 5.577476038338658, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 6983 + }, + { + "epoch": 5.578274760383387, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0401, + "step": 6984 + }, + { + "epoch": 5.5790734824281145, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 6985 + }, + { + "epoch": 5.579872204472844, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 6986 + }, + { + "epoch": 5.580670926517572, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 6987 + }, + { + "epoch": 5.5814696485623, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0408, + "step": 6988 + }, + { + "epoch": 5.582268370607029, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0459, + "step": 6989 + }, + { + "epoch": 5.583067092651757, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 6990 + }, + { + "epoch": 5.583865814696486, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 6991 + }, + { + "epoch": 5.584664536741214, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 6992 + }, + { + "epoch": 5.585463258785943, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0419, + "step": 6993 + }, + { + "epoch": 5.586261980830671, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.034, + "step": 6994 + }, + { + "epoch": 5.5870607028754, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0353, + "step": 6995 + }, + { + "epoch": 5.587859424920127, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0446, + "step": 6996 + }, + { + "epoch": 5.588658146964856, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 6997 + }, + { + "epoch": 5.5894568690095845, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0421, + "step": 6998 + }, + { + "epoch": 5.590255591054313, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 6999 + }, + { + "epoch": 5.5910543130990416, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 7000 + }, + { + "epoch": 5.59185303514377, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0372, + "step": 7001 + }, + { + "epoch": 5.592651757188499, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 7002 + }, + { + "epoch": 5.593450479233227, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 7003 + }, + { + "epoch": 5.594249201277956, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0433, + "step": 7004 + }, + { + "epoch": 5.595047923322683, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.0377, + "step": 7005 + }, + { + "epoch": 5.595846645367412, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0434, + "step": 7006 + }, + { + "epoch": 5.59664536741214, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.0426, + "step": 7007 + }, + { + "epoch": 5.597444089456869, + "grad_norm": 0.205078125, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 7008 + }, + { + "epoch": 5.598242811501597, + "grad_norm": 0.1640625, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 7009 + }, + { + "epoch": 5.599041533546326, + "grad_norm": 0.2080078125, + "learning_rate": 0.0005, + "loss": 1.0446, + "step": 7010 + }, + { + "epoch": 5.5998402555910545, + "grad_norm": 0.251953125, + "learning_rate": 0.0005, + "loss": 1.0437, + "step": 7011 + }, + { + "epoch": 5.600638977635783, + "grad_norm": 0.3984375, + "learning_rate": 0.0005, + "loss": 1.0484, + "step": 7012 + }, + { + "epoch": 5.6014376996805115, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 7013 + }, + { + "epoch": 5.602236421725239, + "grad_norm": 0.302734375, + "learning_rate": 0.0005, + "loss": 1.0388, + "step": 7014 + }, + { + "epoch": 5.603035143769968, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0416, + "step": 7015 + }, + { + "epoch": 5.603833865814696, + "grad_norm": 0.361328125, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 7016 + }, + { + "epoch": 5.604632587859425, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 7017 + }, + { + "epoch": 5.605431309904153, + "grad_norm": 0.30078125, + "learning_rate": 0.0005, + "loss": 1.0468, + "step": 7018 + }, + { + "epoch": 5.606230031948882, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0401, + "step": 7019 + }, + { + "epoch": 5.60702875399361, + "grad_norm": 0.2353515625, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 7020 + }, + { + "epoch": 5.607827476038339, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 7021 + }, + { + "epoch": 5.608626198083067, + "grad_norm": 0.1611328125, + "learning_rate": 0.0005, + "loss": 1.0443, + "step": 7022 + }, + { + "epoch": 5.609424920127795, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0465, + "step": 7023 + }, + { + "epoch": 5.6102236421725244, + "grad_norm": 0.1630859375, + "learning_rate": 0.0005, + "loss": 1.0429, + "step": 7024 + }, + { + "epoch": 5.611022364217252, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.041, + "step": 7025 + }, + { + "epoch": 5.611821086261981, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0391, + "step": 7026 + }, + { + "epoch": 5.612619808306709, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0482, + "step": 7027 + }, + { + "epoch": 5.613418530351438, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 7028 + }, + { + "epoch": 5.614217252396166, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 7029 + }, + { + "epoch": 5.615015974440895, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 7030 + }, + { + "epoch": 5.615814696485623, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 7031 + }, + { + "epoch": 5.616613418530352, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 7032 + }, + { + "epoch": 5.61741214057508, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0439, + "step": 7033 + }, + { + "epoch": 5.618210862619808, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 7034 + }, + { + "epoch": 5.6190095846645365, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0408, + "step": 7035 + }, + { + "epoch": 5.619808306709265, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0396, + "step": 7036 + }, + { + "epoch": 5.6206070287539935, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0396, + "step": 7037 + }, + { + "epoch": 5.621405750798722, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 7038 + }, + { + "epoch": 5.622204472843451, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0415, + "step": 7039 + }, + { + "epoch": 5.623003194888179, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 7040 + }, + { + "epoch": 5.623801916932908, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0399, + "step": 7041 + }, + { + "epoch": 5.624600638977636, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0376, + "step": 7042 + }, + { + "epoch": 5.625399361022364, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 7043 + }, + { + "epoch": 5.626198083067092, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 7044 + }, + { + "epoch": 5.626996805111821, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0454, + "step": 7045 + }, + { + "epoch": 5.627795527156549, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0376, + "step": 7046 + }, + { + "epoch": 5.628594249201278, + "grad_norm": 0.1708984375, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 7047 + }, + { + "epoch": 5.6293929712460065, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 7048 + }, + { + "epoch": 5.630191693290735, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 7049 + }, + { + "epoch": 5.6309904153354635, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 7050 + }, + { + "epoch": 5.631789137380192, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0406, + "step": 7051 + }, + { + "epoch": 5.63258785942492, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.0375, + "step": 7052 + }, + { + "epoch": 5.633386581469648, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 7053 + }, + { + "epoch": 5.634185303514377, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.0384, + "step": 7054 + }, + { + "epoch": 5.634984025559105, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0425, + "step": 7055 + }, + { + "epoch": 5.635782747603834, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 7056 + }, + { + "epoch": 5.636581469648562, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0403, + "step": 7057 + }, + { + "epoch": 5.637380191693291, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0415, + "step": 7058 + }, + { + "epoch": 5.638178913738019, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 7059 + }, + { + "epoch": 5.638977635782748, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0348, + "step": 7060 + }, + { + "epoch": 5.6397763578274756, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 7061 + }, + { + "epoch": 5.640575079872205, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 7062 + }, + { + "epoch": 5.641373801916933, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0414, + "step": 7063 + }, + { + "epoch": 5.642172523961661, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0457, + "step": 7064 + }, + { + "epoch": 5.64297124600639, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0337, + "step": 7065 + }, + { + "epoch": 5.643769968051118, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.0379, + "step": 7066 + }, + { + "epoch": 5.644568690095847, + "grad_norm": 0.1669921875, + "learning_rate": 0.0005, + "loss": 1.039, + "step": 7067 + }, + { + "epoch": 5.645367412140575, + "grad_norm": 0.205078125, + "learning_rate": 0.0005, + "loss": 1.0384, + "step": 7068 + }, + { + "epoch": 5.646166134185304, + "grad_norm": 0.1923828125, + "learning_rate": 0.0005, + "loss": 1.0401, + "step": 7069 + }, + { + "epoch": 5.646964856230032, + "grad_norm": 0.275390625, + "learning_rate": 0.0005, + "loss": 1.0402, + "step": 7070 + }, + { + "epoch": 5.647763578274761, + "grad_norm": 0.169921875, + "learning_rate": 0.0005, + "loss": 1.0394, + "step": 7071 + }, + { + "epoch": 5.6485623003194885, + "grad_norm": 0.287109375, + "learning_rate": 0.0005, + "loss": 1.0389, + "step": 7072 + }, + { + "epoch": 5.649361022364217, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 7073 + }, + { + "epoch": 5.6501597444089455, + "grad_norm": 0.35546875, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 7074 + }, + { + "epoch": 5.650958466453674, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0399, + "step": 7075 + }, + { + "epoch": 5.651757188498403, + "grad_norm": 0.423828125, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 7076 + }, + { + "epoch": 5.652555910543131, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 7077 + }, + { + "epoch": 5.65335463258786, + "grad_norm": 0.3828125, + "learning_rate": 0.0005, + "loss": 1.0432, + "step": 7078 + }, + { + "epoch": 5.654153354632588, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0416, + "step": 7079 + }, + { + "epoch": 5.654952076677317, + "grad_norm": 0.251953125, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 7080 + }, + { + "epoch": 5.655750798722044, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0411, + "step": 7081 + }, + { + "epoch": 5.656549520766773, + "grad_norm": 0.224609375, + "learning_rate": 0.0005, + "loss": 1.0441, + "step": 7082 + }, + { + "epoch": 5.657348242811501, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 7083 + }, + { + "epoch": 5.65814696485623, + "grad_norm": 0.25, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 7084 + }, + { + "epoch": 5.6589456869009584, + "grad_norm": 0.1806640625, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 7085 + }, + { + "epoch": 5.659744408945687, + "grad_norm": 0.291015625, + "learning_rate": 0.0005, + "loss": 1.0405, + "step": 7086 + }, + { + "epoch": 5.6605431309904155, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0426, + "step": 7087 + }, + { + "epoch": 5.661341853035144, + "grad_norm": 0.1962890625, + "learning_rate": 0.0005, + "loss": 1.0397, + "step": 7088 + }, + { + "epoch": 5.662140575079873, + "grad_norm": 0.1826171875, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 7089 + }, + { + "epoch": 5.6629392971246, + "grad_norm": 0.1875, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 7090 + }, + { + "epoch": 5.663738019169329, + "grad_norm": 0.2109375, + "learning_rate": 0.0005, + "loss": 1.0472, + "step": 7091 + }, + { + "epoch": 5.664536741214057, + "grad_norm": 0.18359375, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 7092 + }, + { + "epoch": 5.665335463258786, + "grad_norm": 0.1796875, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 7093 + }, + { + "epoch": 5.666134185303514, + "grad_norm": 0.1845703125, + "learning_rate": 0.0005, + "loss": 1.0389, + "step": 7094 + }, + { + "epoch": 5.666932907348243, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0377, + "step": 7095 + }, + { + "epoch": 5.667731629392971, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 7096 + }, + { + "epoch": 5.6685303514377, + "grad_norm": 0.193359375, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 7097 + }, + { + "epoch": 5.669329073482428, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 7098 + }, + { + "epoch": 5.670127795527156, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 7099 + }, + { + "epoch": 5.6709265175718855, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.037, + "step": 7100 + }, + { + "epoch": 5.671725239616613, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 7101 + }, + { + "epoch": 5.672523961661342, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0399, + "step": 7102 + }, + { + "epoch": 5.67332268370607, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0455, + "step": 7103 + }, + { + "epoch": 5.674121405750799, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0405, + "step": 7104 + }, + { + "epoch": 5.674920127795527, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0348, + "step": 7105 + }, + { + "epoch": 5.675718849840256, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 7106 + }, + { + "epoch": 5.676517571884984, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 7107 + }, + { + "epoch": 5.677316293929713, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 7108 + }, + { + "epoch": 5.678115015974441, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.037, + "step": 7109 + }, + { + "epoch": 5.678913738019169, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0421, + "step": 7110 + }, + { + "epoch": 5.6797124600638975, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 7111 + }, + { + "epoch": 5.680511182108626, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 7112 + }, + { + "epoch": 5.681309904153355, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0404, + "step": 7113 + }, + { + "epoch": 5.682108626198083, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 7114 + }, + { + "epoch": 5.682907348242812, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 7115 + }, + { + "epoch": 5.68370607028754, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0432, + "step": 7116 + }, + { + "epoch": 5.684504792332269, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0393, + "step": 7117 + }, + { + "epoch": 5.685303514376997, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0404, + "step": 7118 + }, + { + "epoch": 5.686102236421725, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0429, + "step": 7119 + }, + { + "epoch": 5.686900958466453, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 7120 + }, + { + "epoch": 5.687699680511182, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0397, + "step": 7121 + }, + { + "epoch": 5.68849840255591, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0396, + "step": 7122 + }, + { + "epoch": 5.689297124600639, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 7123 + }, + { + "epoch": 5.6900958466453675, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 7124 + }, + { + "epoch": 5.690894568690096, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0409, + "step": 7125 + }, + { + "epoch": 5.6916932907348246, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0472, + "step": 7126 + }, + { + "epoch": 5.692492012779553, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 7127 + }, + { + "epoch": 5.693290734824281, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 7128 + }, + { + "epoch": 5.694089456869009, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0397, + "step": 7129 + }, + { + "epoch": 5.694888178913738, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.036, + "step": 7130 + }, + { + "epoch": 5.695686900958466, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 7131 + }, + { + "epoch": 5.696485623003195, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 7132 + }, + { + "epoch": 5.697284345047923, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 7133 + }, + { + "epoch": 5.698083067092652, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.0403, + "step": 7134 + }, + { + "epoch": 5.69888178913738, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0395, + "step": 7135 + }, + { + "epoch": 5.699680511182109, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 7136 + }, + { + "epoch": 5.700479233226837, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 7137 + }, + { + "epoch": 5.701277955271565, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 7138 + }, + { + "epoch": 5.702076677316294, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 7139 + }, + { + "epoch": 5.702875399361022, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0408, + "step": 7140 + }, + { + "epoch": 5.703674121405751, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0361, + "step": 7141 + }, + { + "epoch": 5.704472843450479, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0444, + "step": 7142 + }, + { + "epoch": 5.705271565495208, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0427, + "step": 7143 + }, + { + "epoch": 5.706070287539936, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 7144 + }, + { + "epoch": 5.706869009584665, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0399, + "step": 7145 + }, + { + "epoch": 5.707667731629393, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 7146 + }, + { + "epoch": 5.708466453674122, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 7147 + }, + { + "epoch": 5.7092651757188495, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0361, + "step": 7148 + }, + { + "epoch": 5.710063897763578, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.037, + "step": 7149 + }, + { + "epoch": 5.710862619808307, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0423, + "step": 7150 + }, + { + "epoch": 5.711661341853035, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0406, + "step": 7151 + }, + { + "epoch": 5.712460063897764, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 7152 + }, + { + "epoch": 5.713258785942492, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 7153 + }, + { + "epoch": 5.714057507987221, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 7154 + }, + { + "epoch": 5.714856230031949, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 7155 + }, + { + "epoch": 5.715654952076678, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0436, + "step": 7156 + }, + { + "epoch": 5.716453674121405, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 7157 + }, + { + "epoch": 5.717252396166134, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.034, + "step": 7158 + }, + { + "epoch": 5.718051118210862, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 7159 + }, + { + "epoch": 5.718849840255591, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 7160 + }, + { + "epoch": 5.7196485623003195, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 7161 + }, + { + "epoch": 5.720447284345048, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 7162 + }, + { + "epoch": 5.7212460063897765, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0384, + "step": 7163 + }, + { + "epoch": 5.722044728434505, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0443, + "step": 7164 + }, + { + "epoch": 5.722843450479234, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 7165 + }, + { + "epoch": 5.723642172523961, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0369, + "step": 7166 + }, + { + "epoch": 5.72444089456869, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 7167 + }, + { + "epoch": 5.725239616613418, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 7168 + }, + { + "epoch": 5.726038338658147, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 7169 + }, + { + "epoch": 5.726837060702875, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0391, + "step": 7170 + }, + { + "epoch": 5.727635782747604, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 7171 + }, + { + "epoch": 5.728434504792332, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 7172 + }, + { + "epoch": 5.729233226837061, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0446, + "step": 7173 + }, + { + "epoch": 5.7300319488817895, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.037, + "step": 7174 + }, + { + "epoch": 5.730830670926517, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0388, + "step": 7175 + }, + { + "epoch": 5.731629392971246, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 7176 + }, + { + "epoch": 5.732428115015974, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0445, + "step": 7177 + }, + { + "epoch": 5.733226837060703, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 7178 + }, + { + "epoch": 5.734025559105431, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 7179 + }, + { + "epoch": 5.73482428115016, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0414, + "step": 7180 + }, + { + "epoch": 5.735623003194888, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0441, + "step": 7181 + }, + { + "epoch": 5.736421725239617, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 7182 + }, + { + "epoch": 5.737220447284345, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 7183 + }, + { + "epoch": 5.738019169329074, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0402, + "step": 7184 + }, + { + "epoch": 5.738817891373802, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 7185 + }, + { + "epoch": 5.73961661341853, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0442, + "step": 7186 + }, + { + "epoch": 5.7404153354632586, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 7187 + }, + { + "epoch": 5.741214057507987, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0447, + "step": 7188 + }, + { + "epoch": 5.742012779552716, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0448, + "step": 7189 + }, + { + "epoch": 5.742811501597444, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0412, + "step": 7190 + }, + { + "epoch": 5.743610223642173, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.034, + "step": 7191 + }, + { + "epoch": 5.744408945686901, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.042, + "step": 7192 + }, + { + "epoch": 5.74520766773163, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 7193 + }, + { + "epoch": 5.746006389776358, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 7194 + }, + { + "epoch": 5.746805111821086, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0436, + "step": 7195 + }, + { + "epoch": 5.747603833865814, + "grad_norm": 0.1669921875, + "learning_rate": 0.0005, + "loss": 1.0428, + "step": 7196 + }, + { + "epoch": 5.748402555910543, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 7197 + }, + { + "epoch": 5.7492012779552715, + "grad_norm": 0.1787109375, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 7198 + }, + { + "epoch": 5.75, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0386, + "step": 7199 + }, + { + "epoch": 5.7507987220447285, + "grad_norm": 0.17578125, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 7200 + }, + { + "epoch": 5.751597444089457, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 7201 + }, + { + "epoch": 5.752396166134186, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 7202 + }, + { + "epoch": 5.753194888178914, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 7203 + }, + { + "epoch": 5.753993610223642, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0402, + "step": 7204 + }, + { + "epoch": 5.75479233226837, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0459, + "step": 7205 + }, + { + "epoch": 5.755591054313099, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 7206 + }, + { + "epoch": 5.756389776357827, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.038, + "step": 7207 + }, + { + "epoch": 5.757188498402556, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 7208 + }, + { + "epoch": 5.757987220447284, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 7209 + }, + { + "epoch": 5.758785942492013, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0419, + "step": 7210 + }, + { + "epoch": 5.7595846645367414, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0373, + "step": 7211 + }, + { + "epoch": 5.76038338658147, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 7212 + }, + { + "epoch": 5.761182108626198, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 7213 + }, + { + "epoch": 5.761980830670926, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0353, + "step": 7214 + }, + { + "epoch": 5.762779552715655, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0393, + "step": 7215 + }, + { + "epoch": 5.763578274760383, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.043, + "step": 7216 + }, + { + "epoch": 5.764376996805112, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.041, + "step": 7217 + }, + { + "epoch": 5.76517571884984, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 7218 + }, + { + "epoch": 5.765974440894569, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 7219 + }, + { + "epoch": 5.766773162939297, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 7220 + }, + { + "epoch": 5.767571884984026, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0432, + "step": 7221 + }, + { + "epoch": 5.768370607028754, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 7222 + }, + { + "epoch": 5.769169329073483, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0397, + "step": 7223 + }, + { + "epoch": 5.7699680511182105, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 7224 + }, + { + "epoch": 5.770766773162939, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 7225 + }, + { + "epoch": 5.771565495207668, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0376, + "step": 7226 + }, + { + "epoch": 5.772364217252396, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 7227 + }, + { + "epoch": 5.773162939297125, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 7228 + }, + { + "epoch": 5.773961661341853, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0419, + "step": 7229 + }, + { + "epoch": 5.774760383386582, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0454, + "step": 7230 + }, + { + "epoch": 5.77555910543131, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 7231 + }, + { + "epoch": 5.776357827476039, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 7232 + }, + { + "epoch": 5.777156549520766, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.0416, + "step": 7233 + }, + { + "epoch": 5.777955271565495, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 7234 + }, + { + "epoch": 5.7787539936102235, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 7235 + }, + { + "epoch": 5.779552715654952, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 7236 + }, + { + "epoch": 5.7803514376996805, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0385, + "step": 7237 + }, + { + "epoch": 5.781150159744409, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 7238 + }, + { + "epoch": 5.781948881789138, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 7239 + }, + { + "epoch": 5.782747603833866, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0377, + "step": 7240 + }, + { + "epoch": 5.783546325878595, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0406, + "step": 7241 + }, + { + "epoch": 5.784345047923322, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0396, + "step": 7242 + }, + { + "epoch": 5.785143769968051, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 7243 + }, + { + "epoch": 5.785942492012779, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 7244 + }, + { + "epoch": 5.786741214057508, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0369, + "step": 7245 + }, + { + "epoch": 5.787539936102236, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 7246 + }, + { + "epoch": 5.788338658146965, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 7247 + }, + { + "epoch": 5.789137380191693, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 7248 + }, + { + "epoch": 5.789936102236422, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0476, + "step": 7249 + }, + { + "epoch": 5.7907348242811505, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 7250 + }, + { + "epoch": 5.791533546325878, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0372, + "step": 7251 + }, + { + "epoch": 5.792332268370607, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0409, + "step": 7252 + }, + { + "epoch": 5.793130990415335, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 7253 + }, + { + "epoch": 5.793929712460064, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 7254 + }, + { + "epoch": 5.794728434504792, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 7255 + }, + { + "epoch": 5.795527156549521, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0386, + "step": 7256 + }, + { + "epoch": 5.796325878594249, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 7257 + }, + { + "epoch": 5.797124600638978, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0407, + "step": 7258 + }, + { + "epoch": 5.797923322683706, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 7259 + }, + { + "epoch": 5.798722044728435, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0439, + "step": 7260 + }, + { + "epoch": 5.799520766773163, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0427, + "step": 7261 + }, + { + "epoch": 5.800319488817891, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0451, + "step": 7262 + }, + { + "epoch": 5.80111821086262, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.041, + "step": 7263 + }, + { + "epoch": 5.801916932907348, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 7264 + }, + { + "epoch": 5.802715654952077, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 7265 + }, + { + "epoch": 5.803514376996805, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0353, + "step": 7266 + }, + { + "epoch": 5.804313099041534, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 7267 + }, + { + "epoch": 5.805111821086262, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 7268 + }, + { + "epoch": 5.805910543130991, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.036, + "step": 7269 + }, + { + "epoch": 5.806709265175719, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 7270 + }, + { + "epoch": 5.807507987220447, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.045, + "step": 7271 + }, + { + "epoch": 5.8083067092651754, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 7272 + }, + { + "epoch": 5.809105431309904, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 7273 + }, + { + "epoch": 5.8099041533546325, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0353, + "step": 7274 + }, + { + "epoch": 5.810702875399361, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 7275 + }, + { + "epoch": 5.81150159744409, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0378, + "step": 7276 + }, + { + "epoch": 5.812300319488818, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 7277 + }, + { + "epoch": 5.813099041533547, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0419, + "step": 7278 + }, + { + "epoch": 5.813897763578275, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0457, + "step": 7279 + }, + { + "epoch": 5.814696485623003, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0435, + "step": 7280 + }, + { + "epoch": 5.815495207667731, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0395, + "step": 7281 + }, + { + "epoch": 5.81629392971246, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0377, + "step": 7282 + }, + { + "epoch": 5.817092651757188, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 7283 + }, + { + "epoch": 5.817891373801917, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 7284 + }, + { + "epoch": 5.818690095846645, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 7285 + }, + { + "epoch": 5.819488817891374, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 7286 + }, + { + "epoch": 5.8202875399361025, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 7287 + }, + { + "epoch": 5.821086261980831, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 7288 + }, + { + "epoch": 5.821884984025559, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 7289 + }, + { + "epoch": 5.822683706070287, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0375, + "step": 7290 + }, + { + "epoch": 5.823482428115016, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 7291 + }, + { + "epoch": 5.824281150159744, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0447, + "step": 7292 + }, + { + "epoch": 5.825079872204473, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0369, + "step": 7293 + }, + { + "epoch": 5.825878594249201, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0353, + "step": 7294 + }, + { + "epoch": 5.82667731629393, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 7295 + }, + { + "epoch": 5.827476038338658, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0389, + "step": 7296 + }, + { + "epoch": 5.828274760383387, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0337, + "step": 7297 + }, + { + "epoch": 5.8290734824281145, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 7298 + }, + { + "epoch": 5.829872204472844, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0402, + "step": 7299 + }, + { + "epoch": 5.830670926517572, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 7300 + }, + { + "epoch": 5.8314696485623, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 7301 + }, + { + "epoch": 5.832268370607029, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 7302 + }, + { + "epoch": 5.833067092651757, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 7303 + }, + { + "epoch": 5.833865814696486, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0373, + "step": 7304 + }, + { + "epoch": 5.834664536741214, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0401, + "step": 7305 + }, + { + "epoch": 5.835463258785943, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0427, + "step": 7306 + }, + { + "epoch": 5.836261980830671, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 7307 + }, + { + "epoch": 5.8370607028754, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 7308 + }, + { + "epoch": 5.837859424920127, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0436, + "step": 7309 + }, + { + "epoch": 5.838658146964856, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 7310 + }, + { + "epoch": 5.8394568690095845, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 7311 + }, + { + "epoch": 5.840255591054313, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0403, + "step": 7312 + }, + { + "epoch": 5.8410543130990416, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 7313 + }, + { + "epoch": 5.84185303514377, + "grad_norm": 0.17578125, + "learning_rate": 0.0005, + "loss": 1.0386, + "step": 7314 + }, + { + "epoch": 5.842651757188499, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 7315 + }, + { + "epoch": 5.843450479233227, + "grad_norm": 0.162109375, + "learning_rate": 0.0005, + "loss": 1.0369, + "step": 7316 + }, + { + "epoch": 5.844249201277956, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0388, + "step": 7317 + }, + { + "epoch": 5.845047923322683, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 7318 + }, + { + "epoch": 5.845846645367412, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0376, + "step": 7319 + }, + { + "epoch": 5.84664536741214, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0375, + "step": 7320 + }, + { + "epoch": 5.847444089456869, + "grad_norm": 0.2109375, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 7321 + }, + { + "epoch": 5.848242811501597, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0446, + "step": 7322 + }, + { + "epoch": 5.849041533546326, + "grad_norm": 0.1884765625, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 7323 + }, + { + "epoch": 5.8498402555910545, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.042, + "step": 7324 + }, + { + "epoch": 5.850638977635783, + "grad_norm": 0.197265625, + "learning_rate": 0.0005, + "loss": 1.0405, + "step": 7325 + }, + { + "epoch": 5.8514376996805115, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 7326 + }, + { + "epoch": 5.852236421725239, + "grad_norm": 0.197265625, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 7327 + }, + { + "epoch": 5.853035143769968, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 7328 + }, + { + "epoch": 5.853833865814696, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 7329 + }, + { + "epoch": 5.854632587859425, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 7330 + }, + { + "epoch": 5.855431309904153, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 7331 + }, + { + "epoch": 5.856230031948882, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 7332 + }, + { + "epoch": 5.85702875399361, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 7333 + }, + { + "epoch": 5.857827476038339, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 7334 + }, + { + "epoch": 5.858626198083067, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0391, + "step": 7335 + }, + { + "epoch": 5.859424920127795, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0361, + "step": 7336 + }, + { + "epoch": 5.8602236421725244, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0421, + "step": 7337 + }, + { + "epoch": 5.861022364217252, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 7338 + }, + { + "epoch": 5.861821086261981, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 7339 + }, + { + "epoch": 5.862619808306709, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 7340 + }, + { + "epoch": 5.863418530351438, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0428, + "step": 7341 + }, + { + "epoch": 5.864217252396166, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.046, + "step": 7342 + }, + { + "epoch": 5.865015974440895, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 7343 + }, + { + "epoch": 5.865814696485623, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 7344 + }, + { + "epoch": 5.866613418530352, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 7345 + }, + { + "epoch": 5.86741214057508, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 7346 + }, + { + "epoch": 5.868210862619808, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 7347 + }, + { + "epoch": 5.8690095846645365, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 7348 + }, + { + "epoch": 5.869808306709265, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 7349 + }, + { + "epoch": 5.8706070287539935, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 7350 + }, + { + "epoch": 5.871405750798722, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0393, + "step": 7351 + }, + { + "epoch": 5.872204472843451, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0348, + "step": 7352 + }, + { + "epoch": 5.873003194888179, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 7353 + }, + { + "epoch": 5.873801916932908, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 7354 + }, + { + "epoch": 5.874600638977636, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 7355 + }, + { + "epoch": 5.875399361022364, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0467, + "step": 7356 + }, + { + "epoch": 5.876198083067092, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0423, + "step": 7357 + }, + { + "epoch": 5.876996805111821, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0353, + "step": 7358 + }, + { + "epoch": 5.877795527156549, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 7359 + }, + { + "epoch": 5.878594249201278, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0372, + "step": 7360 + }, + { + "epoch": 5.8793929712460065, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 7361 + }, + { + "epoch": 5.880191693290735, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0408, + "step": 7362 + }, + { + "epoch": 5.8809904153354635, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 7363 + }, + { + "epoch": 5.881789137380192, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 7364 + }, + { + "epoch": 5.88258785942492, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 7365 + }, + { + "epoch": 5.883386581469648, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 7366 + }, + { + "epoch": 5.884185303514377, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0413, + "step": 7367 + }, + { + "epoch": 5.884984025559105, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0372, + "step": 7368 + }, + { + "epoch": 5.885782747603834, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 7369 + }, + { + "epoch": 5.886581469648562, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 7370 + }, + { + "epoch": 5.887380191693291, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.043, + "step": 7371 + }, + { + "epoch": 5.888178913738019, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0397, + "step": 7372 + }, + { + "epoch": 5.888977635782748, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.04, + "step": 7373 + }, + { + "epoch": 5.8897763578274756, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 7374 + }, + { + "epoch": 5.890575079872205, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0375, + "step": 7375 + }, + { + "epoch": 5.891373801916933, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0401, + "step": 7376 + }, + { + "epoch": 5.892172523961661, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 7377 + }, + { + "epoch": 5.89297124600639, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0384, + "step": 7378 + }, + { + "epoch": 5.893769968051118, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 7379 + }, + { + "epoch": 5.894568690095847, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 7380 + }, + { + "epoch": 5.895367412140575, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 7381 + }, + { + "epoch": 5.896166134185304, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 7382 + }, + { + "epoch": 5.896964856230032, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 7383 + }, + { + "epoch": 5.897763578274761, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 7384 + }, + { + "epoch": 5.8985623003194885, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.039, + "step": 7385 + }, + { + "epoch": 5.899361022364217, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 7386 + }, + { + "epoch": 5.9001597444089455, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 7387 + }, + { + "epoch": 5.900958466453674, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 7388 + }, + { + "epoch": 5.901757188498403, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 7389 + }, + { + "epoch": 5.902555910543131, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 7390 + }, + { + "epoch": 5.90335463258786, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0464, + "step": 7391 + }, + { + "epoch": 5.904153354632588, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0429, + "step": 7392 + }, + { + "epoch": 5.904952076677317, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 7393 + }, + { + "epoch": 5.905750798722044, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.046, + "step": 7394 + }, + { + "epoch": 5.906549520766773, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0402, + "step": 7395 + }, + { + "epoch": 5.907348242811501, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 7396 + }, + { + "epoch": 5.90814696485623, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.039, + "step": 7397 + }, + { + "epoch": 5.9089456869009584, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 7398 + }, + { + "epoch": 5.909744408945687, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 7399 + }, + { + "epoch": 5.9105431309904155, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0425, + "step": 7400 + }, + { + "epoch": 5.911341853035144, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0431, + "step": 7401 + }, + { + "epoch": 5.912140575079873, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 7402 + }, + { + "epoch": 5.9129392971246, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 7403 + }, + { + "epoch": 5.913738019169329, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0379, + "step": 7404 + }, + { + "epoch": 5.914536741214057, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0416, + "step": 7405 + }, + { + "epoch": 5.915335463258786, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0434, + "step": 7406 + }, + { + "epoch": 5.916134185303514, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 7407 + }, + { + "epoch": 5.916932907348243, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 7408 + }, + { + "epoch": 5.917731629392971, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0428, + "step": 7409 + }, + { + "epoch": 5.9185303514377, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 7410 + }, + { + "epoch": 5.919329073482428, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 7411 + }, + { + "epoch": 5.920127795527156, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 7412 + }, + { + "epoch": 5.9209265175718855, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0386, + "step": 7413 + }, + { + "epoch": 5.921725239616613, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0443, + "step": 7414 + }, + { + "epoch": 5.922523961661342, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0396, + "step": 7415 + }, + { + "epoch": 5.92332268370607, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 7416 + }, + { + "epoch": 5.924121405750799, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0434, + "step": 7417 + }, + { + "epoch": 5.924920127795527, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 7418 + }, + { + "epoch": 5.925718849840256, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 7419 + }, + { + "epoch": 5.926517571884984, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 7420 + }, + { + "epoch": 5.927316293929713, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 7421 + }, + { + "epoch": 5.928115015974441, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 7422 + }, + { + "epoch": 5.928913738019169, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 7423 + }, + { + "epoch": 5.9297124600638975, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.0428, + "step": 7424 + }, + { + "epoch": 5.930511182108626, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 7425 + }, + { + "epoch": 5.931309904153355, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0489, + "step": 7426 + }, + { + "epoch": 5.932108626198083, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0396, + "step": 7427 + }, + { + "epoch": 5.932907348242812, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 7428 + }, + { + "epoch": 5.93370607028754, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0402, + "step": 7429 + }, + { + "epoch": 5.934504792332269, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0388, + "step": 7430 + }, + { + "epoch": 5.935303514376997, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 7431 + }, + { + "epoch": 5.936102236421725, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0434, + "step": 7432 + }, + { + "epoch": 5.936900958466453, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0457, + "step": 7433 + }, + { + "epoch": 5.937699680511182, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 7434 + }, + { + "epoch": 5.93849840255591, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0373, + "step": 7435 + }, + { + "epoch": 5.939297124600639, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 7436 + }, + { + "epoch": 5.9400958466453675, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0422, + "step": 7437 + }, + { + "epoch": 5.940894568690096, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 7438 + }, + { + "epoch": 5.9416932907348246, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 7439 + }, + { + "epoch": 5.942492012779553, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.039, + "step": 7440 + }, + { + "epoch": 5.943290734824281, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 7441 + }, + { + "epoch": 5.944089456869009, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0469, + "step": 7442 + }, + { + "epoch": 5.944888178913738, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 7443 + }, + { + "epoch": 5.945686900958466, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0376, + "step": 7444 + }, + { + "epoch": 5.946485623003195, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 7445 + }, + { + "epoch": 5.947284345047923, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 7446 + }, + { + "epoch": 5.948083067092652, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 7447 + }, + { + "epoch": 5.94888178913738, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.041, + "step": 7448 + }, + { + "epoch": 5.949680511182109, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 7449 + }, + { + "epoch": 5.950479233226837, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 7450 + }, + { + "epoch": 5.951277955271565, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.036, + "step": 7451 + }, + { + "epoch": 5.952076677316294, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 7452 + }, + { + "epoch": 5.952875399361022, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 7453 + }, + { + "epoch": 5.953674121405751, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0412, + "step": 7454 + }, + { + "epoch": 5.954472843450479, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0528, + "step": 7455 + }, + { + "epoch": 5.955271565495208, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0408, + "step": 7456 + }, + { + "epoch": 5.956070287539936, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 7457 + }, + { + "epoch": 5.956869009584665, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0396, + "step": 7458 + }, + { + "epoch": 5.957667731629393, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 7459 + }, + { + "epoch": 5.958466453674122, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 7460 + }, + { + "epoch": 5.9592651757188495, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0425, + "step": 7461 + }, + { + "epoch": 5.960063897763578, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0401, + "step": 7462 + }, + { + "epoch": 5.960862619808307, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 7463 + }, + { + "epoch": 5.961661341853035, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 7464 + }, + { + "epoch": 5.962460063897764, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 7465 + }, + { + "epoch": 5.963258785942492, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0375, + "step": 7466 + }, + { + "epoch": 5.964057507987221, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.0393, + "step": 7467 + }, + { + "epoch": 5.964856230031949, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 7468 + }, + { + "epoch": 5.965654952076678, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 7469 + }, + { + "epoch": 5.966453674121405, + "grad_norm": 0.171875, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 7470 + }, + { + "epoch": 5.967252396166134, + "grad_norm": 0.255859375, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 7471 + }, + { + "epoch": 5.968051118210862, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 7472 + }, + { + "epoch": 5.968849840255591, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 7473 + }, + { + "epoch": 5.9696485623003195, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 7474 + }, + { + "epoch": 5.970447284345048, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0447, + "step": 7475 + }, + { + "epoch": 5.9712460063897765, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 7476 + }, + { + "epoch": 5.972044728434505, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 7477 + }, + { + "epoch": 5.972843450479234, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.041, + "step": 7478 + }, + { + "epoch": 5.973642172523961, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.038, + "step": 7479 + }, + { + "epoch": 5.97444089456869, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0419, + "step": 7480 + }, + { + "epoch": 5.975239616613418, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0503, + "step": 7481 + }, + { + "epoch": 5.976038338658147, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 7482 + }, + { + "epoch": 5.976837060702875, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 7483 + }, + { + "epoch": 5.977635782747604, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 7484 + }, + { + "epoch": 5.978434504792332, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 7485 + }, + { + "epoch": 5.979233226837061, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 7486 + }, + { + "epoch": 5.9800319488817895, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0482, + "step": 7487 + }, + { + "epoch": 5.980830670926517, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0403, + "step": 7488 + }, + { + "epoch": 5.981629392971246, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 7489 + }, + { + "epoch": 5.982428115015974, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 7490 + }, + { + "epoch": 5.983226837060703, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 7491 + }, + { + "epoch": 5.984025559105431, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 7492 + }, + { + "epoch": 5.98482428115016, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 7493 + }, + { + "epoch": 5.985623003194888, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 7494 + }, + { + "epoch": 5.986421725239617, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 7495 + }, + { + "epoch": 5.987220447284345, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 7496 + }, + { + "epoch": 5.988019169329074, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0429, + "step": 7497 + }, + { + "epoch": 5.988817891373802, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 7498 + }, + { + "epoch": 5.98961661341853, + "grad_norm": 0.177734375, + "learning_rate": 0.0005, + "loss": 1.0389, + "step": 7499 + }, + { + "epoch": 5.9904153354632586, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0409, + "step": 7500 + }, + { + "epoch": 5.991214057507987, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0405, + "step": 7501 + }, + { + "epoch": 5.992012779552716, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 7502 + }, + { + "epoch": 5.992811501597444, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 7503 + }, + { + "epoch": 5.993610223642173, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0407, + "step": 7504 + }, + { + "epoch": 5.994408945686901, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 7505 + }, + { + "epoch": 5.99520766773163, + "grad_norm": 0.2431640625, + "learning_rate": 0.0005, + "loss": 1.0468, + "step": 7506 + }, + { + "epoch": 5.996006389776358, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 7507 + }, + { + "epoch": 5.996805111821086, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0427, + "step": 7508 + }, + { + "epoch": 5.997603833865814, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 7509 + }, + { + "epoch": 5.998402555910543, + "grad_norm": 0.1640625, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 7510 + }, + { + "epoch": 5.9992012779552715, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0429, + "step": 7511 + }, + { + "epoch": 6.0, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 7512 + }, + { + "epoch": 6.0007987220447285, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0348, + "step": 7513 + }, + { + "epoch": 6.001597444089457, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0354, + "step": 7514 + }, + { + "epoch": 6.002396166134186, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0452, + "step": 7515 + }, + { + "epoch": 6.003194888178914, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0426, + "step": 7516 + }, + { + "epoch": 6.003993610223642, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0417, + "step": 7517 + }, + { + "epoch": 6.00479233226837, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0429, + "step": 7518 + }, + { + "epoch": 6.005591054313099, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 7519 + }, + { + "epoch": 6.006389776357827, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0442, + "step": 7520 + }, + { + "epoch": 6.007188498402556, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0408, + "step": 7521 + }, + { + "epoch": 6.007987220447284, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0388, + "step": 7522 + }, + { + "epoch": 6.008785942492013, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 7523 + }, + { + "epoch": 6.0095846645367414, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0411, + "step": 7524 + }, + { + "epoch": 6.01038338658147, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 7525 + }, + { + "epoch": 6.0111821086261985, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0372, + "step": 7526 + }, + { + "epoch": 6.011980830670926, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0411, + "step": 7527 + }, + { + "epoch": 6.012779552715655, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.045, + "step": 7528 + }, + { + "epoch": 6.013578274760383, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 7529 + }, + { + "epoch": 6.014376996805112, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0429, + "step": 7530 + }, + { + "epoch": 6.01517571884984, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 7531 + }, + { + "epoch": 6.015974440894569, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0397, + "step": 7532 + }, + { + "epoch": 6.016773162939297, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 7533 + }, + { + "epoch": 6.017571884984026, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0426, + "step": 7534 + }, + { + "epoch": 6.018370607028754, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 7535 + }, + { + "epoch": 6.019169329073482, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0362, + "step": 7536 + }, + { + "epoch": 6.0199680511182105, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0405, + "step": 7537 + }, + { + "epoch": 6.020766773162939, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 7538 + }, + { + "epoch": 6.021565495207668, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 7539 + }, + { + "epoch": 6.022364217252396, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 7540 + }, + { + "epoch": 6.023162939297125, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.042, + "step": 7541 + }, + { + "epoch": 6.023961661341853, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 7542 + }, + { + "epoch": 6.024760383386582, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 7543 + }, + { + "epoch": 6.02555910543131, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0412, + "step": 7544 + }, + { + "epoch": 6.026357827476039, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 7545 + }, + { + "epoch": 6.027156549520766, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0389, + "step": 7546 + }, + { + "epoch": 6.027955271565495, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0435, + "step": 7547 + }, + { + "epoch": 6.0287539936102235, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 7548 + }, + { + "epoch": 6.029552715654952, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0394, + "step": 7549 + }, + { + "epoch": 6.0303514376996805, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.034, + "step": 7550 + }, + { + "epoch": 6.031150159744409, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0435, + "step": 7551 + }, + { + "epoch": 6.031948881789138, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0424, + "step": 7552 + }, + { + "epoch": 6.032747603833866, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.036, + "step": 7553 + }, + { + "epoch": 6.033546325878595, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 7554 + }, + { + "epoch": 6.034345047923322, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 7555 + }, + { + "epoch": 6.035143769968051, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0414, + "step": 7556 + }, + { + "epoch": 6.035942492012779, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0442, + "step": 7557 + }, + { + "epoch": 6.036741214057508, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0376, + "step": 7558 + }, + { + "epoch": 6.037539936102236, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0397, + "step": 7559 + }, + { + "epoch": 6.038338658146965, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0362, + "step": 7560 + }, + { + "epoch": 6.039137380191693, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 7561 + }, + { + "epoch": 6.039936102236422, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 7562 + }, + { + "epoch": 6.0407348242811505, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 7563 + }, + { + "epoch": 6.041533546325879, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 7564 + }, + { + "epoch": 6.042332268370607, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 7565 + }, + { + "epoch": 6.043130990415335, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 7566 + }, + { + "epoch": 6.043929712460064, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0379, + "step": 7567 + }, + { + "epoch": 6.044728434504792, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 7568 + }, + { + "epoch": 6.045527156549521, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0409, + "step": 7569 + }, + { + "epoch": 6.046325878594249, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0361, + "step": 7570 + }, + { + "epoch": 6.047124600638978, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0386, + "step": 7571 + }, + { + "epoch": 6.047923322683706, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0399, + "step": 7572 + }, + { + "epoch": 6.048722044728435, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 7573 + }, + { + "epoch": 6.0495207667731625, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 7574 + }, + { + "epoch": 6.050319488817891, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 7575 + }, + { + "epoch": 6.05111821086262, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 7576 + }, + { + "epoch": 6.051916932907348, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 7577 + }, + { + "epoch": 6.052715654952077, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0389, + "step": 7578 + }, + { + "epoch": 6.053514376996805, + "grad_norm": 0.1728515625, + "learning_rate": 0.0005, + "loss": 1.0409, + "step": 7579 + }, + { + "epoch": 6.054313099041534, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.041, + "step": 7580 + }, + { + "epoch": 6.055111821086262, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.0404, + "step": 7581 + }, + { + "epoch": 6.055910543130991, + "grad_norm": 0.1533203125, + "learning_rate": 0.0005, + "loss": 1.0448, + "step": 7582 + }, + { + "epoch": 6.056709265175719, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 7583 + }, + { + "epoch": 6.057507987220447, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 7584 + }, + { + "epoch": 6.0583067092651754, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0391, + "step": 7585 + }, + { + "epoch": 6.059105431309904, + "grad_norm": 0.1533203125, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 7586 + }, + { + "epoch": 6.0599041533546325, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 7587 + }, + { + "epoch": 6.060702875399361, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0439, + "step": 7588 + }, + { + "epoch": 6.06150159744409, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 7589 + }, + { + "epoch": 6.062300319488818, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 7590 + }, + { + "epoch": 6.063099041533547, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0428, + "step": 7591 + }, + { + "epoch": 6.063897763578275, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0422, + "step": 7592 + }, + { + "epoch": 6.064696485623003, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0432, + "step": 7593 + }, + { + "epoch": 6.065495207667731, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0397, + "step": 7594 + }, + { + "epoch": 6.06629392971246, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 7595 + }, + { + "epoch": 6.067092651757188, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 7596 + }, + { + "epoch": 6.067891373801917, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 7597 + }, + { + "epoch": 6.068690095846645, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 7598 + }, + { + "epoch": 6.069488817891374, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 7599 + }, + { + "epoch": 6.0702875399361025, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 7600 + }, + { + "epoch": 6.071086261980831, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 7601 + }, + { + "epoch": 6.0718849840255595, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 7602 + }, + { + "epoch": 6.072683706070287, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0425, + "step": 7603 + }, + { + "epoch": 6.073482428115016, + "grad_norm": 0.17578125, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 7604 + }, + { + "epoch": 6.074281150159744, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0337, + "step": 7605 + }, + { + "epoch": 6.075079872204473, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 7606 + }, + { + "epoch": 6.075878594249201, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 7607 + }, + { + "epoch": 6.07667731629393, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0405, + "step": 7608 + }, + { + "epoch": 6.077476038338658, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 7609 + }, + { + "epoch": 6.078274760383387, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0412, + "step": 7610 + }, + { + "epoch": 6.079073482428115, + "grad_norm": 0.1826171875, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 7611 + }, + { + "epoch": 6.079872204472843, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0453, + "step": 7612 + }, + { + "epoch": 6.080670926517572, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 7613 + }, + { + "epoch": 6.0814696485623, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0413, + "step": 7614 + }, + { + "epoch": 6.082268370607029, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 7615 + }, + { + "epoch": 6.083067092651757, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 7616 + }, + { + "epoch": 6.083865814696486, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0422, + "step": 7617 + }, + { + "epoch": 6.084664536741214, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 7618 + }, + { + "epoch": 6.085463258785943, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 7619 + }, + { + "epoch": 6.086261980830671, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0391, + "step": 7620 + }, + { + "epoch": 6.0870607028754, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 7621 + }, + { + "epoch": 6.087859424920127, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0393, + "step": 7622 + }, + { + "epoch": 6.088658146964856, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0348, + "step": 7623 + }, + { + "epoch": 6.0894568690095845, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0391, + "step": 7624 + }, + { + "epoch": 6.090255591054313, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0397, + "step": 7625 + }, + { + "epoch": 6.0910543130990416, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0362, + "step": 7626 + }, + { + "epoch": 6.09185303514377, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0414, + "step": 7627 + }, + { + "epoch": 6.092651757188499, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0362, + "step": 7628 + }, + { + "epoch": 6.093450479233227, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 7629 + }, + { + "epoch": 6.094249201277956, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 7630 + }, + { + "epoch": 6.095047923322683, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 7631 + }, + { + "epoch": 6.095846645367412, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0379, + "step": 7632 + }, + { + "epoch": 6.09664536741214, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0348, + "step": 7633 + }, + { + "epoch": 6.097444089456869, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.044, + "step": 7634 + }, + { + "epoch": 6.098242811501597, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 7635 + }, + { + "epoch": 6.099041533546326, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0372, + "step": 7636 + }, + { + "epoch": 6.0998402555910545, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 7637 + }, + { + "epoch": 6.100638977635783, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 7638 + }, + { + "epoch": 6.1014376996805115, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 7639 + }, + { + "epoch": 6.102236421725239, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 7640 + }, + { + "epoch": 6.103035143769968, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0378, + "step": 7641 + }, + { + "epoch": 6.103833865814696, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 7642 + }, + { + "epoch": 6.104632587859425, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0375, + "step": 7643 + }, + { + "epoch": 6.105431309904153, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0376, + "step": 7644 + }, + { + "epoch": 6.106230031948882, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 7645 + }, + { + "epoch": 6.10702875399361, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 7646 + }, + { + "epoch": 6.107827476038339, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 7647 + }, + { + "epoch": 6.108626198083067, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0353, + "step": 7648 + }, + { + "epoch": 6.109424920127796, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 7649 + }, + { + "epoch": 6.110223642172524, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 7650 + }, + { + "epoch": 6.111022364217252, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 7651 + }, + { + "epoch": 6.111821086261981, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0372, + "step": 7652 + }, + { + "epoch": 6.112619808306709, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 7653 + }, + { + "epoch": 6.113418530351438, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.039, + "step": 7654 + }, + { + "epoch": 6.114217252396166, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0459, + "step": 7655 + }, + { + "epoch": 6.115015974440895, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 7656 + }, + { + "epoch": 6.115814696485623, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 7657 + }, + { + "epoch": 6.116613418530352, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 7658 + }, + { + "epoch": 6.11741214057508, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0396, + "step": 7659 + }, + { + "epoch": 6.118210862619808, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 7660 + }, + { + "epoch": 6.1190095846645365, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.044, + "step": 7661 + }, + { + "epoch": 6.119808306709265, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 7662 + }, + { + "epoch": 6.1206070287539935, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.04, + "step": 7663 + }, + { + "epoch": 6.121405750798722, + "grad_norm": 0.1669921875, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 7664 + }, + { + "epoch": 6.122204472843451, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 7665 + }, + { + "epoch": 6.123003194888179, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0373, + "step": 7666 + }, + { + "epoch": 6.123801916932908, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.037, + "step": 7667 + }, + { + "epoch": 6.124600638977636, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 7668 + }, + { + "epoch": 6.125399361022364, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 7669 + }, + { + "epoch": 6.126198083067092, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 7670 + }, + { + "epoch": 6.126996805111821, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0377, + "step": 7671 + }, + { + "epoch": 6.127795527156549, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 7672 + }, + { + "epoch": 6.128594249201278, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 7673 + }, + { + "epoch": 6.1293929712460065, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0406, + "step": 7674 + }, + { + "epoch": 6.130191693290735, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 7675 + }, + { + "epoch": 6.1309904153354635, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0434, + "step": 7676 + }, + { + "epoch": 6.131789137380192, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 7677 + }, + { + "epoch": 6.13258785942492, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0353, + "step": 7678 + }, + { + "epoch": 6.133386581469648, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0373, + "step": 7679 + }, + { + "epoch": 6.134185303514377, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.039, + "step": 7680 + }, + { + "epoch": 6.134984025559105, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0441, + "step": 7681 + }, + { + "epoch": 6.135782747603834, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 7682 + }, + { + "epoch": 6.136581469648562, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 7683 + }, + { + "epoch": 6.137380191693291, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 7684 + }, + { + "epoch": 6.138178913738019, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 7685 + }, + { + "epoch": 6.138977635782748, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 7686 + }, + { + "epoch": 6.139776357827476, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0426, + "step": 7687 + }, + { + "epoch": 6.140575079872204, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 7688 + }, + { + "epoch": 6.141373801916933, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 7689 + }, + { + "epoch": 6.142172523961661, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 7690 + }, + { + "epoch": 6.14297124600639, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 7691 + }, + { + "epoch": 6.143769968051118, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 7692 + }, + { + "epoch": 6.144568690095847, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 7693 + }, + { + "epoch": 6.145367412140575, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 7694 + }, + { + "epoch": 6.146166134185304, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0412, + "step": 7695 + }, + { + "epoch": 6.146964856230032, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.036, + "step": 7696 + }, + { + "epoch": 6.147763578274761, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0426, + "step": 7697 + }, + { + "epoch": 6.1485623003194885, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0386, + "step": 7698 + }, + { + "epoch": 6.149361022364217, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 7699 + }, + { + "epoch": 6.1501597444089455, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0432, + "step": 7700 + }, + { + "epoch": 6.150958466453674, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0411, + "step": 7701 + }, + { + "epoch": 6.151757188498403, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0411, + "step": 7702 + }, + { + "epoch": 6.152555910543131, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 7703 + }, + { + "epoch": 6.15335463258786, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 7704 + }, + { + "epoch": 6.154153354632588, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 7705 + }, + { + "epoch": 6.154952076677317, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 7706 + }, + { + "epoch": 6.155750798722044, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.041, + "step": 7707 + }, + { + "epoch": 6.156549520766773, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 7708 + }, + { + "epoch": 6.157348242811501, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 7709 + }, + { + "epoch": 6.15814696485623, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 7710 + }, + { + "epoch": 6.1589456869009584, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 7711 + }, + { + "epoch": 6.159744408945687, + "grad_norm": 0.162109375, + "learning_rate": 0.0005, + "loss": 1.0416, + "step": 7712 + }, + { + "epoch": 6.1605431309904155, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 7713 + }, + { + "epoch": 6.161341853035144, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0446, + "step": 7714 + }, + { + "epoch": 6.162140575079873, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 7715 + }, + { + "epoch": 6.1629392971246, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 7716 + }, + { + "epoch": 6.163738019169329, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0395, + "step": 7717 + }, + { + "epoch": 6.164536741214057, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0164, + "step": 7718 + }, + { + "epoch": 6.165335463258786, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0408, + "step": 7719 + }, + { + "epoch": 6.166134185303514, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.0412, + "step": 7720 + }, + { + "epoch": 6.166932907348243, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0409, + "step": 7721 + }, + { + "epoch": 6.167731629392971, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 7722 + }, + { + "epoch": 6.1685303514377, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 7723 + }, + { + "epoch": 6.169329073482428, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 7724 + }, + { + "epoch": 6.170127795527157, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 7725 + }, + { + "epoch": 6.170926517571885, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 7726 + }, + { + "epoch": 6.171725239616613, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0434, + "step": 7727 + }, + { + "epoch": 6.172523961661342, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0389, + "step": 7728 + }, + { + "epoch": 6.17332268370607, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 7729 + }, + { + "epoch": 6.174121405750799, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 7730 + }, + { + "epoch": 6.174920127795527, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 7731 + }, + { + "epoch": 6.175718849840256, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 7732 + }, + { + "epoch": 6.176517571884984, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 7733 + }, + { + "epoch": 6.177316293929713, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0466, + "step": 7734 + }, + { + "epoch": 6.178115015974441, + "grad_norm": 0.1640625, + "learning_rate": 0.0005, + "loss": 1.0377, + "step": 7735 + }, + { + "epoch": 6.178913738019169, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 7736 + }, + { + "epoch": 6.1797124600638975, + "grad_norm": 0.1875, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 7737 + }, + { + "epoch": 6.180511182108626, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0397, + "step": 7738 + }, + { + "epoch": 6.181309904153355, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 7739 + }, + { + "epoch": 6.182108626198083, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0388, + "step": 7740 + }, + { + "epoch": 6.182907348242812, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.039, + "step": 7741 + }, + { + "epoch": 6.18370607028754, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 7742 + }, + { + "epoch": 6.184504792332269, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 7743 + }, + { + "epoch": 6.185303514376997, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 7744 + }, + { + "epoch": 6.186102236421725, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 7745 + }, + { + "epoch": 6.186900958466453, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.041, + "step": 7746 + }, + { + "epoch": 6.187699680511182, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0463, + "step": 7747 + }, + { + "epoch": 6.18849840255591, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0378, + "step": 7748 + }, + { + "epoch": 6.189297124600639, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0394, + "step": 7749 + }, + { + "epoch": 6.1900958466453675, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 7750 + }, + { + "epoch": 6.190894568690096, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0451, + "step": 7751 + }, + { + "epoch": 6.1916932907348246, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 7752 + }, + { + "epoch": 6.192492012779553, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0468, + "step": 7753 + }, + { + "epoch": 6.193290734824281, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 7754 + }, + { + "epoch": 6.194089456869009, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 7755 + }, + { + "epoch": 6.194888178913738, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0431, + "step": 7756 + }, + { + "epoch": 6.195686900958466, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 7757 + }, + { + "epoch": 6.196485623003195, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 7758 + }, + { + "epoch": 6.197284345047923, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 7759 + }, + { + "epoch": 6.198083067092652, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0396, + "step": 7760 + }, + { + "epoch": 6.19888178913738, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0376, + "step": 7761 + }, + { + "epoch": 6.199680511182109, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0432, + "step": 7762 + }, + { + "epoch": 6.2004792332268375, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0413, + "step": 7763 + }, + { + "epoch": 6.201277955271565, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 7764 + }, + { + "epoch": 6.202076677316294, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 7765 + }, + { + "epoch": 6.202875399361022, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 7766 + }, + { + "epoch": 6.203674121405751, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 7767 + }, + { + "epoch": 6.204472843450479, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0402, + "step": 7768 + }, + { + "epoch": 6.205271565495208, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 7769 + }, + { + "epoch": 6.206070287539936, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0397, + "step": 7770 + }, + { + "epoch": 6.206869009584665, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 7771 + }, + { + "epoch": 6.207667731629393, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 7772 + }, + { + "epoch": 6.208466453674121, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 7773 + }, + { + "epoch": 6.2092651757188495, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 7774 + }, + { + "epoch": 6.210063897763578, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 7775 + }, + { + "epoch": 6.210862619808307, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 7776 + }, + { + "epoch": 6.211661341853035, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 7777 + }, + { + "epoch": 6.212460063897764, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0396, + "step": 7778 + }, + { + "epoch": 6.213258785942492, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 7779 + }, + { + "epoch": 6.214057507987221, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0397, + "step": 7780 + }, + { + "epoch": 6.214856230031949, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 7781 + }, + { + "epoch": 6.215654952076678, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 7782 + }, + { + "epoch": 6.216453674121405, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0394, + "step": 7783 + }, + { + "epoch": 6.217252396166134, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 7784 + }, + { + "epoch": 6.218051118210862, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 7785 + }, + { + "epoch": 6.218849840255591, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0407, + "step": 7786 + }, + { + "epoch": 6.2196485623003195, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 7787 + }, + { + "epoch": 6.220447284345048, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 7788 + }, + { + "epoch": 6.2212460063897765, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.037, + "step": 7789 + }, + { + "epoch": 6.222044728434505, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0385, + "step": 7790 + }, + { + "epoch": 6.222843450479234, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 7791 + }, + { + "epoch": 6.223642172523961, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0338, + "step": 7792 + }, + { + "epoch": 6.22444089456869, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 7793 + }, + { + "epoch": 6.225239616613418, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 7794 + }, + { + "epoch": 6.226038338658147, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 7795 + }, + { + "epoch": 6.226837060702875, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 7796 + }, + { + "epoch": 6.227635782747604, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0403, + "step": 7797 + }, + { + "epoch": 6.228434504792332, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 7798 + }, + { + "epoch": 6.229233226837061, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0413, + "step": 7799 + }, + { + "epoch": 6.2300319488817895, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0419, + "step": 7800 + }, + { + "epoch": 6.230830670926518, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.0406, + "step": 7801 + }, + { + "epoch": 6.231629392971246, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0419, + "step": 7802 + }, + { + "epoch": 6.232428115015974, + "grad_norm": 0.181640625, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 7803 + }, + { + "epoch": 6.233226837060703, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 7804 + }, + { + "epoch": 6.234025559105431, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 7805 + }, + { + "epoch": 6.23482428115016, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 7806 + }, + { + "epoch": 6.235623003194888, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 7807 + }, + { + "epoch": 6.236421725239617, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 7808 + }, + { + "epoch": 6.237220447284345, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 7809 + }, + { + "epoch": 6.238019169329074, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 7810 + }, + { + "epoch": 6.2388178913738015, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 7811 + }, + { + "epoch": 6.23961661341853, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.038, + "step": 7812 + }, + { + "epoch": 6.2404153354632586, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0412, + "step": 7813 + }, + { + "epoch": 6.241214057507987, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 7814 + }, + { + "epoch": 6.242012779552716, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.036, + "step": 7815 + }, + { + "epoch": 6.242811501597444, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 7816 + }, + { + "epoch": 6.243610223642173, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 7817 + }, + { + "epoch": 6.244408945686901, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 7818 + }, + { + "epoch": 6.24520766773163, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 7819 + }, + { + "epoch": 6.246006389776358, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 7820 + }, + { + "epoch": 6.246805111821086, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 7821 + }, + { + "epoch": 6.247603833865814, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0389, + "step": 7822 + }, + { + "epoch": 6.248402555910543, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0409, + "step": 7823 + }, + { + "epoch": 6.2492012779552715, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 7824 + }, + { + "epoch": 6.25, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0378, + "step": 7825 + }, + { + "epoch": 6.2507987220447285, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 7826 + }, + { + "epoch": 6.251597444089457, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 7827 + }, + { + "epoch": 6.252396166134186, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 7828 + }, + { + "epoch": 6.253194888178914, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.044, + "step": 7829 + }, + { + "epoch": 6.253993610223642, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 7830 + }, + { + "epoch": 6.25479233226837, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 7831 + }, + { + "epoch": 6.255591054313099, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 7832 + }, + { + "epoch": 6.256389776357827, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0389, + "step": 7833 + }, + { + "epoch": 6.257188498402556, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0391, + "step": 7834 + }, + { + "epoch": 6.257987220447284, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.037, + "step": 7835 + }, + { + "epoch": 6.258785942492013, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 7836 + }, + { + "epoch": 6.2595846645367414, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0397, + "step": 7837 + }, + { + "epoch": 6.26038338658147, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 7838 + }, + { + "epoch": 6.261182108626198, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 7839 + }, + { + "epoch": 6.261980830670926, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 7840 + }, + { + "epoch": 6.262779552715655, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 7841 + }, + { + "epoch": 6.263578274760383, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.041, + "step": 7842 + }, + { + "epoch": 6.264376996805112, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0413, + "step": 7843 + }, + { + "epoch": 6.26517571884984, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0416, + "step": 7844 + }, + { + "epoch": 6.265974440894569, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 7845 + }, + { + "epoch": 6.266773162939297, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0401, + "step": 7846 + }, + { + "epoch": 6.267571884984026, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 7847 + }, + { + "epoch": 6.268370607028754, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 7848 + }, + { + "epoch": 6.269169329073483, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 7849 + }, + { + "epoch": 6.2699680511182105, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0385, + "step": 7850 + }, + { + "epoch": 6.270766773162939, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 7851 + }, + { + "epoch": 6.271565495207668, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 7852 + }, + { + "epoch": 6.272364217252396, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 7853 + }, + { + "epoch": 6.273162939297125, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 7854 + }, + { + "epoch": 6.273961661341853, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 7855 + }, + { + "epoch": 6.274760383386582, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 7856 + }, + { + "epoch": 6.27555910543131, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 7857 + }, + { + "epoch": 6.276357827476039, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 7858 + }, + { + "epoch": 6.277156549520766, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0412, + "step": 7859 + }, + { + "epoch": 6.277955271565495, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 7860 + }, + { + "epoch": 6.2787539936102235, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 7861 + }, + { + "epoch": 6.279552715654952, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.042, + "step": 7862 + }, + { + "epoch": 6.2803514376996805, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 7863 + }, + { + "epoch": 6.281150159744409, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 7864 + }, + { + "epoch": 6.281948881789138, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0391, + "step": 7865 + }, + { + "epoch": 6.282747603833866, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 7866 + }, + { + "epoch": 6.283546325878595, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0353, + "step": 7867 + }, + { + "epoch": 6.284345047923322, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 7868 + }, + { + "epoch": 6.285143769968051, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 7869 + }, + { + "epoch": 6.285942492012779, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 7870 + }, + { + "epoch": 6.286741214057508, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 7871 + }, + { + "epoch": 6.287539936102236, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0452, + "step": 7872 + }, + { + "epoch": 6.288338658146965, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 7873 + }, + { + "epoch": 6.289137380191693, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.037, + "step": 7874 + }, + { + "epoch": 6.289936102236422, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0372, + "step": 7875 + }, + { + "epoch": 6.2907348242811505, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 7876 + }, + { + "epoch": 6.291533546325878, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 7877 + }, + { + "epoch": 6.292332268370607, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 7878 + }, + { + "epoch": 6.293130990415335, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0377, + "step": 7879 + }, + { + "epoch": 6.293929712460064, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 7880 + }, + { + "epoch": 6.294728434504792, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0423, + "step": 7881 + }, + { + "epoch": 6.295527156549521, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 7882 + }, + { + "epoch": 6.296325878594249, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0404, + "step": 7883 + }, + { + "epoch": 6.297124600638978, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.039, + "step": 7884 + }, + { + "epoch": 6.297923322683706, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0407, + "step": 7885 + }, + { + "epoch": 6.298722044728435, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 7886 + }, + { + "epoch": 6.2995207667731625, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 7887 + }, + { + "epoch": 6.300319488817891, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 7888 + }, + { + "epoch": 6.30111821086262, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 7889 + }, + { + "epoch": 6.301916932907348, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 7890 + }, + { + "epoch": 6.302715654952077, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0375, + "step": 7891 + }, + { + "epoch": 6.303514376996805, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 7892 + }, + { + "epoch": 6.304313099041534, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0426, + "step": 7893 + }, + { + "epoch": 6.305111821086262, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.043, + "step": 7894 + }, + { + "epoch": 6.305910543130991, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 7895 + }, + { + "epoch": 6.306709265175719, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 7896 + }, + { + "epoch": 6.307507987220447, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0389, + "step": 7897 + }, + { + "epoch": 6.3083067092651754, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 7898 + }, + { + "epoch": 6.309105431309904, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0403, + "step": 7899 + }, + { + "epoch": 6.3099041533546325, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0447, + "step": 7900 + }, + { + "epoch": 6.310702875399361, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 7901 + }, + { + "epoch": 6.31150159744409, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 7902 + }, + { + "epoch": 6.312300319488818, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 7903 + }, + { + "epoch": 6.313099041533547, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 7904 + }, + { + "epoch": 6.313897763578275, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 7905 + }, + { + "epoch": 6.314696485623003, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 7906 + }, + { + "epoch": 6.315495207667731, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 7907 + }, + { + "epoch": 6.31629392971246, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 7908 + }, + { + "epoch": 6.317092651757188, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 7909 + }, + { + "epoch": 6.317891373801917, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0433, + "step": 7910 + }, + { + "epoch": 6.318690095846645, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 7911 + }, + { + "epoch": 6.319488817891374, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 7912 + }, + { + "epoch": 6.3202875399361025, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 7913 + }, + { + "epoch": 6.321086261980831, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 7914 + }, + { + "epoch": 6.321884984025559, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 7915 + }, + { + "epoch": 6.322683706070287, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.05, + "step": 7916 + }, + { + "epoch": 6.323482428115016, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 7917 + }, + { + "epoch": 6.324281150159744, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 7918 + }, + { + "epoch": 6.325079872204473, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0396, + "step": 7919 + }, + { + "epoch": 6.325878594249201, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 7920 + }, + { + "epoch": 6.32667731629393, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0416, + "step": 7921 + }, + { + "epoch": 6.327476038338658, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 7922 + }, + { + "epoch": 6.328274760383387, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 7923 + }, + { + "epoch": 6.329073482428115, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 7924 + }, + { + "epoch": 6.329872204472843, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0416, + "step": 7925 + }, + { + "epoch": 6.330670926517572, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 7926 + }, + { + "epoch": 6.3314696485623, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 7927 + }, + { + "epoch": 6.332268370607029, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 7928 + }, + { + "epoch": 6.333067092651757, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.038, + "step": 7929 + }, + { + "epoch": 6.333865814696486, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 7930 + }, + { + "epoch": 6.334664536741214, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0444, + "step": 7931 + }, + { + "epoch": 6.335463258785943, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 7932 + }, + { + "epoch": 6.336261980830671, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 7933 + }, + { + "epoch": 6.3370607028754, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 7934 + }, + { + "epoch": 6.337859424920127, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0377, + "step": 7935 + }, + { + "epoch": 6.338658146964856, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0414, + "step": 7936 + }, + { + "epoch": 6.3394568690095845, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0362, + "step": 7937 + }, + { + "epoch": 6.340255591054313, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 7938 + }, + { + "epoch": 6.3410543130990416, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0455, + "step": 7939 + }, + { + "epoch": 6.34185303514377, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 7940 + }, + { + "epoch": 6.342651757188499, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0424, + "step": 7941 + }, + { + "epoch": 6.343450479233227, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0353, + "step": 7942 + }, + { + "epoch": 6.344249201277956, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 7943 + }, + { + "epoch": 6.345047923322683, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 7944 + }, + { + "epoch": 6.345846645367412, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0354, + "step": 7945 + }, + { + "epoch": 6.34664536741214, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0379, + "step": 7946 + }, + { + "epoch": 6.347444089456869, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 7947 + }, + { + "epoch": 6.348242811501597, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 7948 + }, + { + "epoch": 6.349041533546326, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 7949 + }, + { + "epoch": 6.3498402555910545, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0429, + "step": 7950 + }, + { + "epoch": 6.350638977635783, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0394, + "step": 7951 + }, + { + "epoch": 6.3514376996805115, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 7952 + }, + { + "epoch": 6.352236421725239, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 7953 + }, + { + "epoch": 6.353035143769968, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0373, + "step": 7954 + }, + { + "epoch": 6.353833865814696, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 7955 + }, + { + "epoch": 6.354632587859425, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 7956 + }, + { + "epoch": 6.355431309904153, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0445, + "step": 7957 + }, + { + "epoch": 6.356230031948882, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0389, + "step": 7958 + }, + { + "epoch": 6.35702875399361, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0414, + "step": 7959 + }, + { + "epoch": 6.357827476038339, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.037, + "step": 7960 + }, + { + "epoch": 6.358626198083067, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 7961 + }, + { + "epoch": 6.359424920127796, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0401, + "step": 7962 + }, + { + "epoch": 6.360223642172524, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 7963 + }, + { + "epoch": 6.361022364217252, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 7964 + }, + { + "epoch": 6.361821086261981, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 7965 + }, + { + "epoch": 6.362619808306709, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 7966 + }, + { + "epoch": 6.363418530351438, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 7967 + }, + { + "epoch": 6.364217252396166, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0424, + "step": 7968 + }, + { + "epoch": 6.365015974440895, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0431, + "step": 7969 + }, + { + "epoch": 6.365814696485623, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 7970 + }, + { + "epoch": 6.366613418530352, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 7971 + }, + { + "epoch": 6.36741214057508, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 7972 + }, + { + "epoch": 6.368210862619808, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 7973 + }, + { + "epoch": 6.3690095846645365, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 7974 + }, + { + "epoch": 6.369808306709265, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0448, + "step": 7975 + }, + { + "epoch": 6.3706070287539935, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0436, + "step": 7976 + }, + { + "epoch": 6.371405750798722, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 7977 + }, + { + "epoch": 6.372204472843451, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0372, + "step": 7978 + }, + { + "epoch": 6.373003194888179, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.039, + "step": 7979 + }, + { + "epoch": 6.373801916932908, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0502, + "step": 7980 + }, + { + "epoch": 6.374600638977636, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 7981 + }, + { + "epoch": 6.375399361022364, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 7982 + }, + { + "epoch": 6.376198083067092, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0361, + "step": 7983 + }, + { + "epoch": 6.376996805111821, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0442, + "step": 7984 + }, + { + "epoch": 6.377795527156549, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 7985 + }, + { + "epoch": 6.378594249201278, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 7986 + }, + { + "epoch": 6.3793929712460065, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 7987 + }, + { + "epoch": 6.380191693290735, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0413, + "step": 7988 + }, + { + "epoch": 6.3809904153354635, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 7989 + }, + { + "epoch": 6.381789137380192, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 7990 + }, + { + "epoch": 6.38258785942492, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0362, + "step": 7991 + }, + { + "epoch": 6.383386581469648, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 7992 + }, + { + "epoch": 6.384185303514377, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 7993 + }, + { + "epoch": 6.384984025559105, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 7994 + }, + { + "epoch": 6.385782747603834, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0396, + "step": 7995 + }, + { + "epoch": 6.386581469648562, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 7996 + }, + { + "epoch": 6.387380191693291, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 7997 + }, + { + "epoch": 6.388178913738019, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0375, + "step": 7998 + }, + { + "epoch": 6.388977635782748, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 7999 + }, + { + "epoch": 6.389776357827476, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 8000 + }, + { + "epoch": 6.390575079872204, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 8001 + }, + { + "epoch": 6.391373801916933, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 8002 + }, + { + "epoch": 6.392172523961661, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0389, + "step": 8003 + }, + { + "epoch": 6.39297124600639, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 8004 + }, + { + "epoch": 6.393769968051118, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0401, + "step": 8005 + }, + { + "epoch": 6.394568690095847, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 8006 + }, + { + "epoch": 6.395367412140575, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0394, + "step": 8007 + }, + { + "epoch": 6.396166134185304, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.043, + "step": 8008 + }, + { + "epoch": 6.396964856230032, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0401, + "step": 8009 + }, + { + "epoch": 6.397763578274761, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0413, + "step": 8010 + }, + { + "epoch": 6.3985623003194885, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 8011 + }, + { + "epoch": 6.399361022364217, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0391, + "step": 8012 + }, + { + "epoch": 6.4001597444089455, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 8013 + }, + { + "epoch": 6.400958466453674, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 8014 + }, + { + "epoch": 6.401757188498403, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 8015 + }, + { + "epoch": 6.402555910543131, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 8016 + }, + { + "epoch": 6.40335463258786, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 8017 + }, + { + "epoch": 6.404153354632588, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 8018 + }, + { + "epoch": 6.404952076677317, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0377, + "step": 8019 + }, + { + "epoch": 6.405750798722044, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 8020 + }, + { + "epoch": 6.406549520766773, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0411, + "step": 8021 + }, + { + "epoch": 6.407348242811501, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.0416, + "step": 8022 + }, + { + "epoch": 6.40814696485623, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 8023 + }, + { + "epoch": 6.4089456869009584, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 8024 + }, + { + "epoch": 6.409744408945687, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 8025 + }, + { + "epoch": 6.4105431309904155, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 8026 + }, + { + "epoch": 6.411341853035144, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.045, + "step": 8027 + }, + { + "epoch": 6.412140575079873, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 8028 + }, + { + "epoch": 6.4129392971246, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0369, + "step": 8029 + }, + { + "epoch": 6.413738019169329, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0386, + "step": 8030 + }, + { + "epoch": 6.414536741214057, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 8031 + }, + { + "epoch": 6.415335463258786, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0372, + "step": 8032 + }, + { + "epoch": 6.416134185303514, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 8033 + }, + { + "epoch": 6.416932907348243, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0389, + "step": 8034 + }, + { + "epoch": 6.417731629392971, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0377, + "step": 8035 + }, + { + "epoch": 6.4185303514377, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0393, + "step": 8036 + }, + { + "epoch": 6.419329073482428, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 8037 + }, + { + "epoch": 6.420127795527157, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 8038 + }, + { + "epoch": 6.420926517571885, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 8039 + }, + { + "epoch": 6.421725239616613, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 8040 + }, + { + "epoch": 6.422523961661342, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 8041 + }, + { + "epoch": 6.42332268370607, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 8042 + }, + { + "epoch": 6.424121405750799, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 8043 + }, + { + "epoch": 6.424920127795527, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 8044 + }, + { + "epoch": 6.425718849840256, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0408, + "step": 8045 + }, + { + "epoch": 6.426517571884984, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 8046 + }, + { + "epoch": 6.427316293929713, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0377, + "step": 8047 + }, + { + "epoch": 6.428115015974441, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0338, + "step": 8048 + }, + { + "epoch": 6.428913738019169, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0393, + "step": 8049 + }, + { + "epoch": 6.4297124600638975, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 8050 + }, + { + "epoch": 6.430511182108626, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0389, + "step": 8051 + }, + { + "epoch": 6.431309904153355, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 8052 + }, + { + "epoch": 6.432108626198083, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 8053 + }, + { + "epoch": 6.432907348242812, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0426, + "step": 8054 + }, + { + "epoch": 6.43370607028754, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 8055 + }, + { + "epoch": 6.434504792332269, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0424, + "step": 8056 + }, + { + "epoch": 6.435303514376997, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 8057 + }, + { + "epoch": 6.436102236421725, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 8058 + }, + { + "epoch": 6.436900958466453, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0373, + "step": 8059 + }, + { + "epoch": 6.437699680511182, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 8060 + }, + { + "epoch": 6.43849840255591, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.045, + "step": 8061 + }, + { + "epoch": 6.439297124600639, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 8062 + }, + { + "epoch": 6.4400958466453675, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 8063 + }, + { + "epoch": 6.440894568690096, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0362, + "step": 8064 + }, + { + "epoch": 6.4416932907348246, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 8065 + }, + { + "epoch": 6.442492012779553, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0376, + "step": 8066 + }, + { + "epoch": 6.443290734824281, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 8067 + }, + { + "epoch": 6.444089456869009, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0372, + "step": 8068 + }, + { + "epoch": 6.444888178913738, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 8069 + }, + { + "epoch": 6.445686900958466, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 8070 + }, + { + "epoch": 6.446485623003195, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 8071 + }, + { + "epoch": 6.447284345047923, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 8072 + }, + { + "epoch": 6.448083067092652, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 8073 + }, + { + "epoch": 6.44888178913738, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 8074 + }, + { + "epoch": 6.449680511182109, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 8075 + }, + { + "epoch": 6.4504792332268375, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 8076 + }, + { + "epoch": 6.451277955271565, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 8077 + }, + { + "epoch": 6.452076677316294, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0474, + "step": 8078 + }, + { + "epoch": 6.452875399361022, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0455, + "step": 8079 + }, + { + "epoch": 6.453674121405751, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 8080 + }, + { + "epoch": 6.454472843450479, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 8081 + }, + { + "epoch": 6.455271565495208, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 8082 + }, + { + "epoch": 6.456070287539936, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0427, + "step": 8083 + }, + { + "epoch": 6.456869009584665, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 8084 + }, + { + "epoch": 6.457667731629393, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 8085 + }, + { + "epoch": 6.458466453674122, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0372, + "step": 8086 + }, + { + "epoch": 6.4592651757188495, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0429, + "step": 8087 + }, + { + "epoch": 6.460063897763578, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0397, + "step": 8088 + }, + { + "epoch": 6.460862619808307, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.047, + "step": 8089 + }, + { + "epoch": 6.461661341853035, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 8090 + }, + { + "epoch": 6.462460063897764, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0388, + "step": 8091 + }, + { + "epoch": 6.463258785942492, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 8092 + }, + { + "epoch": 6.464057507987221, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 8093 + }, + { + "epoch": 6.464856230031949, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 8094 + }, + { + "epoch": 6.465654952076678, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 8095 + }, + { + "epoch": 6.466453674121405, + "grad_norm": 0.16015625, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 8096 + }, + { + "epoch": 6.467252396166134, + "grad_norm": 0.2197265625, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 8097 + }, + { + "epoch": 6.468051118210862, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 8098 + }, + { + "epoch": 6.468849840255591, + "grad_norm": 0.251953125, + "learning_rate": 0.0005, + "loss": 1.0428, + "step": 8099 + }, + { + "epoch": 6.4696485623003195, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 8100 + }, + { + "epoch": 6.470447284345048, + "grad_norm": 0.26171875, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 8101 + }, + { + "epoch": 6.4712460063897765, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 8102 + }, + { + "epoch": 6.472044728434505, + "grad_norm": 0.28515625, + "learning_rate": 0.0005, + "loss": 1.0385, + "step": 8103 + }, + { + "epoch": 6.472843450479234, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 8104 + }, + { + "epoch": 6.473642172523961, + "grad_norm": 0.265625, + "learning_rate": 0.0005, + "loss": 1.0412, + "step": 8105 + }, + { + "epoch": 6.47444089456869, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 8106 + }, + { + "epoch": 6.475239616613418, + "grad_norm": 0.29296875, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 8107 + }, + { + "epoch": 6.476038338658147, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0404, + "step": 8108 + }, + { + "epoch": 6.476837060702875, + "grad_norm": 0.31640625, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 8109 + }, + { + "epoch": 6.477635782747604, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0396, + "step": 8110 + }, + { + "epoch": 6.478434504792332, + "grad_norm": 0.3046875, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 8111 + }, + { + "epoch": 6.479233226837061, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 8112 + }, + { + "epoch": 6.4800319488817895, + "grad_norm": 0.3203125, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 8113 + }, + { + "epoch": 6.480830670926517, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 8114 + }, + { + "epoch": 6.481629392971246, + "grad_norm": 0.291015625, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 8115 + }, + { + "epoch": 6.482428115015974, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.034, + "step": 8116 + }, + { + "epoch": 6.483226837060703, + "grad_norm": 0.373046875, + "learning_rate": 0.0005, + "loss": 1.0373, + "step": 8117 + }, + { + "epoch": 6.484025559105431, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 8118 + }, + { + "epoch": 6.48482428115016, + "grad_norm": 0.423828125, + "learning_rate": 0.0005, + "loss": 1.0403, + "step": 8119 + }, + { + "epoch": 6.485623003194888, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 8120 + }, + { + "epoch": 6.486421725239617, + "grad_norm": 0.376953125, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 8121 + }, + { + "epoch": 6.487220447284345, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0414, + "step": 8122 + }, + { + "epoch": 6.488019169329074, + "grad_norm": 0.375, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 8123 + }, + { + "epoch": 6.488817891373802, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 8124 + }, + { + "epoch": 6.48961661341853, + "grad_norm": 0.33203125, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 8125 + }, + { + "epoch": 6.4904153354632586, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 8126 + }, + { + "epoch": 6.491214057507987, + "grad_norm": 0.38671875, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 8127 + }, + { + "epoch": 6.492012779552716, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 8128 + }, + { + "epoch": 6.492811501597444, + "grad_norm": 0.2890625, + "learning_rate": 0.0005, + "loss": 1.0479, + "step": 8129 + }, + { + "epoch": 6.493610223642173, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 8130 + }, + { + "epoch": 6.494408945686901, + "grad_norm": 0.349609375, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 8131 + }, + { + "epoch": 6.49520766773163, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 8132 + }, + { + "epoch": 6.496006389776358, + "grad_norm": 0.31640625, + "learning_rate": 0.0005, + "loss": 1.0411, + "step": 8133 + }, + { + "epoch": 6.496805111821086, + "grad_norm": 0.166015625, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 8134 + }, + { + "epoch": 6.497603833865814, + "grad_norm": 0.369140625, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 8135 + }, + { + "epoch": 6.498402555910543, + "grad_norm": 0.2158203125, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 8136 + }, + { + "epoch": 6.4992012779552715, + "grad_norm": 0.54296875, + "learning_rate": 0.0005, + "loss": 1.0465, + "step": 8137 + }, + { + "epoch": 6.5, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.0416, + "step": 8138 + }, + { + "epoch": 6.5007987220447285, + "grad_norm": 0.76171875, + "learning_rate": 0.0005, + "loss": 1.0409, + "step": 8139 + }, + { + "epoch": 6.501597444089457, + "grad_norm": 0.30078125, + "learning_rate": 0.0005, + "loss": 1.0379, + "step": 8140 + }, + { + "epoch": 6.502396166134186, + "grad_norm": 0.396484375, + "learning_rate": 0.0005, + "loss": 1.0386, + "step": 8141 + }, + { + "epoch": 6.503194888178914, + "grad_norm": 0.29296875, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 8142 + }, + { + "epoch": 6.503993610223642, + "grad_norm": 0.318359375, + "learning_rate": 0.0005, + "loss": 1.036, + "step": 8143 + }, + { + "epoch": 6.50479233226837, + "grad_norm": 0.36328125, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 8144 + }, + { + "epoch": 6.505591054313099, + "grad_norm": 0.287109375, + "learning_rate": 0.0005, + "loss": 1.0469, + "step": 8145 + }, + { + "epoch": 6.506389776357827, + "grad_norm": 0.546875, + "learning_rate": 0.0005, + "loss": 1.0406, + "step": 8146 + }, + { + "epoch": 6.507188498402556, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0353, + "step": 8147 + }, + { + "epoch": 6.507987220447284, + "grad_norm": 0.373046875, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 8148 + }, + { + "epoch": 6.508785942492013, + "grad_norm": 0.2294921875, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 8149 + }, + { + "epoch": 6.5095846645367414, + "grad_norm": 0.53515625, + "learning_rate": 0.0005, + "loss": 1.0425, + "step": 8150 + }, + { + "epoch": 6.51038338658147, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 8151 + }, + { + "epoch": 6.511182108626198, + "grad_norm": 0.59375, + "learning_rate": 0.0005, + "loss": 1.0436, + "step": 8152 + }, + { + "epoch": 6.511980830670926, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0419, + "step": 8153 + }, + { + "epoch": 6.512779552715655, + "grad_norm": 0.578125, + "learning_rate": 0.0005, + "loss": 1.0466, + "step": 8154 + }, + { + "epoch": 6.513578274760383, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 8155 + }, + { + "epoch": 6.514376996805112, + "grad_norm": 0.5, + "learning_rate": 0.0005, + "loss": 1.0409, + "step": 8156 + }, + { + "epoch": 6.51517571884984, + "grad_norm": 0.16015625, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 8157 + }, + { + "epoch": 6.515974440894569, + "grad_norm": 0.41796875, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 8158 + }, + { + "epoch": 6.516773162939297, + "grad_norm": 0.1865234375, + "learning_rate": 0.0005, + "loss": 1.0404, + "step": 8159 + }, + { + "epoch": 6.517571884984026, + "grad_norm": 0.43359375, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 8160 + }, + { + "epoch": 6.518370607028754, + "grad_norm": 0.177734375, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 8161 + }, + { + "epoch": 6.519169329073483, + "grad_norm": 0.2451171875, + "learning_rate": 0.0005, + "loss": 1.0423, + "step": 8162 + }, + { + "epoch": 6.5199680511182105, + "grad_norm": 0.173828125, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 8163 + }, + { + "epoch": 6.520766773162939, + "grad_norm": 0.212890625, + "learning_rate": 0.0005, + "loss": 1.036, + "step": 8164 + }, + { + "epoch": 6.521565495207668, + "grad_norm": 0.189453125, + "learning_rate": 0.0005, + "loss": 1.0409, + "step": 8165 + }, + { + "epoch": 6.522364217252396, + "grad_norm": 0.1953125, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 8166 + }, + { + "epoch": 6.523162939297125, + "grad_norm": 0.2021484375, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 8167 + }, + { + "epoch": 6.523961661341853, + "grad_norm": 0.2021484375, + "learning_rate": 0.0005, + "loss": 1.0379, + "step": 8168 + }, + { + "epoch": 6.524760383386582, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 8169 + }, + { + "epoch": 6.52555910543131, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0446, + "step": 8170 + }, + { + "epoch": 6.526357827476039, + "grad_norm": 0.1669921875, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 8171 + }, + { + "epoch": 6.527156549520766, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 8172 + }, + { + "epoch": 6.527955271565495, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.0413, + "step": 8173 + }, + { + "epoch": 6.5287539936102235, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 8174 + }, + { + "epoch": 6.529552715654952, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0406, + "step": 8175 + }, + { + "epoch": 6.5303514376996805, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0452, + "step": 8176 + }, + { + "epoch": 6.531150159744409, + "grad_norm": 0.1533203125, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 8177 + }, + { + "epoch": 6.531948881789138, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.039, + "step": 8178 + }, + { + "epoch": 6.532747603833866, + "grad_norm": 0.1572265625, + "learning_rate": 0.0005, + "loss": 1.0384, + "step": 8179 + }, + { + "epoch": 6.533546325878595, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0441, + "step": 8180 + }, + { + "epoch": 6.534345047923322, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 8181 + }, + { + "epoch": 6.535143769968051, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 8182 + }, + { + "epoch": 6.535942492012779, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 8183 + }, + { + "epoch": 6.536741214057508, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0408, + "step": 8184 + }, + { + "epoch": 6.537539936102236, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 8185 + }, + { + "epoch": 6.538338658146965, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 8186 + }, + { + "epoch": 6.539137380191693, + "grad_norm": 0.1669921875, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 8187 + }, + { + "epoch": 6.539936102236422, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 8188 + }, + { + "epoch": 6.5407348242811505, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.0386, + "step": 8189 + }, + { + "epoch": 6.541533546325878, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0409, + "step": 8190 + }, + { + "epoch": 6.542332268370607, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.04, + "step": 8191 + }, + { + "epoch": 6.543130990415335, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 8192 + }, + { + "epoch": 6.543929712460064, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 8193 + }, + { + "epoch": 6.544728434504792, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0405, + "step": 8194 + }, + { + "epoch": 6.545527156549521, + "grad_norm": 0.1708984375, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 8195 + }, + { + "epoch": 6.546325878594249, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0373, + "step": 8196 + }, + { + "epoch": 6.547124600638978, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 8197 + }, + { + "epoch": 6.547923322683706, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 8198 + }, + { + "epoch": 6.548722044728435, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0415, + "step": 8199 + }, + { + "epoch": 6.549520766773163, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 8200 + }, + { + "epoch": 6.550319488817891, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 8201 + }, + { + "epoch": 6.55111821086262, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0475, + "step": 8202 + }, + { + "epoch": 6.551916932907348, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 8203 + }, + { + "epoch": 6.552715654952077, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 8204 + }, + { + "epoch": 6.553514376996805, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0429, + "step": 8205 + }, + { + "epoch": 6.554313099041534, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0372, + "step": 8206 + }, + { + "epoch": 6.555111821086262, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 8207 + }, + { + "epoch": 6.555910543130991, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 8208 + }, + { + "epoch": 6.556709265175719, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0405, + "step": 8209 + }, + { + "epoch": 6.557507987220447, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 8210 + }, + { + "epoch": 6.5583067092651754, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0394, + "step": 8211 + }, + { + "epoch": 6.559105431309904, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 8212 + }, + { + "epoch": 6.5599041533546325, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0401, + "step": 8213 + }, + { + "epoch": 6.560702875399361, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0412, + "step": 8214 + }, + { + "epoch": 6.56150159744409, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 8215 + }, + { + "epoch": 6.562300319488818, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 8216 + }, + { + "epoch": 6.563099041533547, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0354, + "step": 8217 + }, + { + "epoch": 6.563897763578275, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 8218 + }, + { + "epoch": 6.564696485623003, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.034, + "step": 8219 + }, + { + "epoch": 6.565495207667731, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 8220 + }, + { + "epoch": 6.56629392971246, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 8221 + }, + { + "epoch": 6.567092651757188, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 8222 + }, + { + "epoch": 6.567891373801917, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 8223 + }, + { + "epoch": 6.568690095846645, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 8224 + }, + { + "epoch": 6.569488817891374, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 8225 + }, + { + "epoch": 6.5702875399361025, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0417, + "step": 8226 + }, + { + "epoch": 6.571086261980831, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0372, + "step": 8227 + }, + { + "epoch": 6.571884984025559, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.036, + "step": 8228 + }, + { + "epoch": 6.572683706070287, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0389, + "step": 8229 + }, + { + "epoch": 6.573482428115016, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 8230 + }, + { + "epoch": 6.574281150159744, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.04, + "step": 8231 + }, + { + "epoch": 6.575079872204473, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 8232 + }, + { + "epoch": 6.575878594249201, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 8233 + }, + { + "epoch": 6.57667731629393, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.036, + "step": 8234 + }, + { + "epoch": 6.577476038338658, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0393, + "step": 8235 + }, + { + "epoch": 6.578274760383387, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 8236 + }, + { + "epoch": 6.5790734824281145, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0373, + "step": 8237 + }, + { + "epoch": 6.579872204472844, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 8238 + }, + { + "epoch": 6.580670926517572, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0412, + "step": 8239 + }, + { + "epoch": 6.5814696485623, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 8240 + }, + { + "epoch": 6.582268370607029, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 8241 + }, + { + "epoch": 6.583067092651757, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 8242 + }, + { + "epoch": 6.583865814696486, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 8243 + }, + { + "epoch": 6.584664536741214, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 8244 + }, + { + "epoch": 6.585463258785943, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 8245 + }, + { + "epoch": 6.586261980830671, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.037, + "step": 8246 + }, + { + "epoch": 6.5870607028754, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 8247 + }, + { + "epoch": 6.587859424920127, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 8248 + }, + { + "epoch": 6.588658146964856, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 8249 + }, + { + "epoch": 6.5894568690095845, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 8250 + }, + { + "epoch": 6.590255591054313, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 8251 + }, + { + "epoch": 6.5910543130990416, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0416, + "step": 8252 + }, + { + "epoch": 6.59185303514377, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 8253 + }, + { + "epoch": 6.592651757188499, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 8254 + }, + { + "epoch": 6.593450479233227, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 8255 + }, + { + "epoch": 6.594249201277956, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0384, + "step": 8256 + }, + { + "epoch": 6.595047923322683, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 8257 + }, + { + "epoch": 6.595846645367412, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.04, + "step": 8258 + }, + { + "epoch": 6.59664536741214, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0412, + "step": 8259 + }, + { + "epoch": 6.597444089456869, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 8260 + }, + { + "epoch": 6.598242811501597, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0424, + "step": 8261 + }, + { + "epoch": 6.599041533546326, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0483, + "step": 8262 + }, + { + "epoch": 6.5998402555910545, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 8263 + }, + { + "epoch": 6.600638977635783, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 8264 + }, + { + "epoch": 6.6014376996805115, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 8265 + }, + { + "epoch": 6.602236421725239, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0379, + "step": 8266 + }, + { + "epoch": 6.603035143769968, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 8267 + }, + { + "epoch": 6.603833865814696, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 8268 + }, + { + "epoch": 6.604632587859425, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 8269 + }, + { + "epoch": 6.605431309904153, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 8270 + }, + { + "epoch": 6.606230031948882, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0432, + "step": 8271 + }, + { + "epoch": 6.60702875399361, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 8272 + }, + { + "epoch": 6.607827476038339, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 8273 + }, + { + "epoch": 6.608626198083067, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0354, + "step": 8274 + }, + { + "epoch": 6.609424920127795, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 8275 + }, + { + "epoch": 6.6102236421725244, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0375, + "step": 8276 + }, + { + "epoch": 6.611022364217252, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0378, + "step": 8277 + }, + { + "epoch": 6.611821086261981, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 8278 + }, + { + "epoch": 6.612619808306709, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 8279 + }, + { + "epoch": 6.613418530351438, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.0477, + "step": 8280 + }, + { + "epoch": 6.614217252396166, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0397, + "step": 8281 + }, + { + "epoch": 6.615015974440895, + "grad_norm": 0.2431640625, + "learning_rate": 0.0005, + "loss": 1.0384, + "step": 8282 + }, + { + "epoch": 6.615814696485623, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 8283 + }, + { + "epoch": 6.616613418530352, + "grad_norm": 0.2119140625, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 8284 + }, + { + "epoch": 6.61741214057508, + "grad_norm": 0.16015625, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 8285 + }, + { + "epoch": 6.618210862619808, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 8286 + }, + { + "epoch": 6.6190095846645365, + "grad_norm": 0.2080078125, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 8287 + }, + { + "epoch": 6.619808306709265, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 8288 + }, + { + "epoch": 6.6206070287539935, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0386, + "step": 8289 + }, + { + "epoch": 6.621405750798722, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 8290 + }, + { + "epoch": 6.622204472843451, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0456, + "step": 8291 + }, + { + "epoch": 6.623003194888179, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 8292 + }, + { + "epoch": 6.623801916932908, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 8293 + }, + { + "epoch": 6.624600638977636, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0388, + "step": 8294 + }, + { + "epoch": 6.625399361022364, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 8295 + }, + { + "epoch": 6.626198083067092, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 8296 + }, + { + "epoch": 6.626996805111821, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 8297 + }, + { + "epoch": 6.627795527156549, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 8298 + }, + { + "epoch": 6.628594249201278, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 8299 + }, + { + "epoch": 6.6293929712460065, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 8300 + }, + { + "epoch": 6.630191693290735, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 8301 + }, + { + "epoch": 6.6309904153354635, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0373, + "step": 8302 + }, + { + "epoch": 6.631789137380192, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.043, + "step": 8303 + }, + { + "epoch": 6.63258785942492, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0434, + "step": 8304 + }, + { + "epoch": 6.633386581469648, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 8305 + }, + { + "epoch": 6.634185303514377, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0414, + "step": 8306 + }, + { + "epoch": 6.634984025559105, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 8307 + }, + { + "epoch": 6.635782747603834, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.039, + "step": 8308 + }, + { + "epoch": 6.636581469648562, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0459, + "step": 8309 + }, + { + "epoch": 6.637380191693291, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 8310 + }, + { + "epoch": 6.638178913738019, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0411, + "step": 8311 + }, + { + "epoch": 6.638977635782748, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0478, + "step": 8312 + }, + { + "epoch": 6.6397763578274756, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0377, + "step": 8313 + }, + { + "epoch": 6.640575079872205, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 8314 + }, + { + "epoch": 6.641373801916933, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0433, + "step": 8315 + }, + { + "epoch": 6.642172523961661, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 8316 + }, + { + "epoch": 6.64297124600639, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0369, + "step": 8317 + }, + { + "epoch": 6.643769968051118, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 8318 + }, + { + "epoch": 6.644568690095847, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 8319 + }, + { + "epoch": 6.645367412140575, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 8320 + }, + { + "epoch": 6.646166134185304, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0423, + "step": 8321 + }, + { + "epoch": 6.646964856230032, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 8322 + }, + { + "epoch": 6.647763578274761, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 8323 + }, + { + "epoch": 6.6485623003194885, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 8324 + }, + { + "epoch": 6.649361022364217, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 8325 + }, + { + "epoch": 6.6501597444089455, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 8326 + }, + { + "epoch": 6.650958466453674, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.039, + "step": 8327 + }, + { + "epoch": 6.651757188498403, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0369, + "step": 8328 + }, + { + "epoch": 6.652555910543131, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 8329 + }, + { + "epoch": 6.65335463258786, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 8330 + }, + { + "epoch": 6.654153354632588, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 8331 + }, + { + "epoch": 6.654952076677317, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0394, + "step": 8332 + }, + { + "epoch": 6.655750798722044, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 8333 + }, + { + "epoch": 6.656549520766773, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0385, + "step": 8334 + }, + { + "epoch": 6.657348242811501, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0403, + "step": 8335 + }, + { + "epoch": 6.65814696485623, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 8336 + }, + { + "epoch": 6.6589456869009584, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 8337 + }, + { + "epoch": 6.659744408945687, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 8338 + }, + { + "epoch": 6.6605431309904155, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 8339 + }, + { + "epoch": 6.661341853035144, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 8340 + }, + { + "epoch": 6.662140575079873, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0378, + "step": 8341 + }, + { + "epoch": 6.6629392971246, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0384, + "step": 8342 + }, + { + "epoch": 6.663738019169329, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 8343 + }, + { + "epoch": 6.664536741214057, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0162, + "step": 8344 + }, + { + "epoch": 6.665335463258786, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0338, + "step": 8345 + }, + { + "epoch": 6.666134185303514, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 8346 + }, + { + "epoch": 6.666932907348243, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 8347 + }, + { + "epoch": 6.667731629392971, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 8348 + }, + { + "epoch": 6.6685303514377, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 8349 + }, + { + "epoch": 6.669329073482428, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 8350 + }, + { + "epoch": 6.670127795527156, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0403, + "step": 8351 + }, + { + "epoch": 6.6709265175718855, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 8352 + }, + { + "epoch": 6.671725239616613, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.036, + "step": 8353 + }, + { + "epoch": 6.672523961661342, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 8354 + }, + { + "epoch": 6.67332268370607, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 8355 + }, + { + "epoch": 6.674121405750799, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 8356 + }, + { + "epoch": 6.674920127795527, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 8357 + }, + { + "epoch": 6.675718849840256, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 8358 + }, + { + "epoch": 6.676517571884984, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 8359 + }, + { + "epoch": 6.677316293929713, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 8360 + }, + { + "epoch": 6.678115015974441, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 8361 + }, + { + "epoch": 6.678913738019169, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 8362 + }, + { + "epoch": 6.6797124600638975, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 8363 + }, + { + "epoch": 6.680511182108626, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 8364 + }, + { + "epoch": 6.681309904153355, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 8365 + }, + { + "epoch": 6.682108626198083, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0372, + "step": 8366 + }, + { + "epoch": 6.682907348242812, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 8367 + }, + { + "epoch": 6.68370607028754, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 8368 + }, + { + "epoch": 6.684504792332269, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 8369 + }, + { + "epoch": 6.685303514376997, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0377, + "step": 8370 + }, + { + "epoch": 6.686102236421725, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0413, + "step": 8371 + }, + { + "epoch": 6.686900958466453, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0432, + "step": 8372 + }, + { + "epoch": 6.687699680511182, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.036, + "step": 8373 + }, + { + "epoch": 6.68849840255591, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 8374 + }, + { + "epoch": 6.689297124600639, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 8375 + }, + { + "epoch": 6.6900958466453675, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 8376 + }, + { + "epoch": 6.690894568690096, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 8377 + }, + { + "epoch": 6.6916932907348246, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 8378 + }, + { + "epoch": 6.692492012779553, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 8379 + }, + { + "epoch": 6.693290734824281, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0353, + "step": 8380 + }, + { + "epoch": 6.694089456869009, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 8381 + }, + { + "epoch": 6.694888178913738, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 8382 + }, + { + "epoch": 6.695686900958466, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0378, + "step": 8383 + }, + { + "epoch": 6.696485623003195, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 8384 + }, + { + "epoch": 6.697284345047923, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0376, + "step": 8385 + }, + { + "epoch": 6.698083067092652, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 8386 + }, + { + "epoch": 6.69888178913738, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 8387 + }, + { + "epoch": 6.699680511182109, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 8388 + }, + { + "epoch": 6.700479233226837, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 8389 + }, + { + "epoch": 6.701277955271565, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 8390 + }, + { + "epoch": 6.702076677316294, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0444, + "step": 8391 + }, + { + "epoch": 6.702875399361022, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0427, + "step": 8392 + }, + { + "epoch": 6.703674121405751, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 8393 + }, + { + "epoch": 6.704472843450479, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 8394 + }, + { + "epoch": 6.705271565495208, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 8395 + }, + { + "epoch": 6.706070287539936, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 8396 + }, + { + "epoch": 6.706869009584665, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 8397 + }, + { + "epoch": 6.707667731629393, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0448, + "step": 8398 + }, + { + "epoch": 6.708466453674122, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 8399 + }, + { + "epoch": 6.7092651757188495, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.037, + "step": 8400 + }, + { + "epoch": 6.710063897763578, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0384, + "step": 8401 + }, + { + "epoch": 6.710862619808307, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0442, + "step": 8402 + }, + { + "epoch": 6.711661341853035, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 8403 + }, + { + "epoch": 6.712460063897764, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 8404 + }, + { + "epoch": 6.713258785942492, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 8405 + }, + { + "epoch": 6.714057507987221, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.038, + "step": 8406 + }, + { + "epoch": 6.714856230031949, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0454, + "step": 8407 + }, + { + "epoch": 6.715654952076678, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 8408 + }, + { + "epoch": 6.716453674121405, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0362, + "step": 8409 + }, + { + "epoch": 6.717252396166134, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 8410 + }, + { + "epoch": 6.718051118210862, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 8411 + }, + { + "epoch": 6.718849840255591, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0397, + "step": 8412 + }, + { + "epoch": 6.7196485623003195, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 8413 + }, + { + "epoch": 6.720447284345048, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 8414 + }, + { + "epoch": 6.7212460063897765, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0419, + "step": 8415 + }, + { + "epoch": 6.722044728434505, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 8416 + }, + { + "epoch": 6.722843450479234, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 8417 + }, + { + "epoch": 6.723642172523961, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 8418 + }, + { + "epoch": 6.72444089456869, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0372, + "step": 8419 + }, + { + "epoch": 6.725239616613418, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 8420 + }, + { + "epoch": 6.726038338658147, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 8421 + }, + { + "epoch": 6.726837060702875, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0361, + "step": 8422 + }, + { + "epoch": 6.727635782747604, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0403, + "step": 8423 + }, + { + "epoch": 6.728434504792332, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 8424 + }, + { + "epoch": 6.729233226837061, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 8425 + }, + { + "epoch": 6.7300319488817895, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 8426 + }, + { + "epoch": 6.730830670926517, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 8427 + }, + { + "epoch": 6.731629392971246, + "grad_norm": 0.166015625, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 8428 + }, + { + "epoch": 6.732428115015974, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 8429 + }, + { + "epoch": 6.733226837060703, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0434, + "step": 8430 + }, + { + "epoch": 6.734025559105431, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 8431 + }, + { + "epoch": 6.73482428115016, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 8432 + }, + { + "epoch": 6.735623003194888, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 8433 + }, + { + "epoch": 6.736421725239617, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0369, + "step": 8434 + }, + { + "epoch": 6.737220447284345, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 8435 + }, + { + "epoch": 6.738019169329074, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 8436 + }, + { + "epoch": 6.738817891373802, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 8437 + }, + { + "epoch": 6.73961661341853, + "grad_norm": 0.18359375, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 8438 + }, + { + "epoch": 6.7404153354632586, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 8439 + }, + { + "epoch": 6.741214057507987, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.044, + "step": 8440 + }, + { + "epoch": 6.742012779552716, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.039, + "step": 8441 + }, + { + "epoch": 6.742811501597444, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 8442 + }, + { + "epoch": 6.743610223642173, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 8443 + }, + { + "epoch": 6.744408945686901, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 8444 + }, + { + "epoch": 6.74520766773163, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0403, + "step": 8445 + }, + { + "epoch": 6.746006389776358, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0488, + "step": 8446 + }, + { + "epoch": 6.746805111821086, + "grad_norm": 0.162109375, + "learning_rate": 0.0005, + "loss": 1.0391, + "step": 8447 + }, + { + "epoch": 6.747603833865814, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 8448 + }, + { + "epoch": 6.748402555910543, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 8449 + }, + { + "epoch": 6.7492012779552715, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 8450 + }, + { + "epoch": 6.75, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 8451 + }, + { + "epoch": 6.7507987220447285, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 8452 + }, + { + "epoch": 6.751597444089457, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 8453 + }, + { + "epoch": 6.752396166134186, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 8454 + }, + { + "epoch": 6.753194888178914, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.041, + "step": 8455 + }, + { + "epoch": 6.753993610223642, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0439, + "step": 8456 + }, + { + "epoch": 6.75479233226837, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 8457 + }, + { + "epoch": 6.755591054313099, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 8458 + }, + { + "epoch": 6.756389776357827, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 8459 + }, + { + "epoch": 6.757188498402556, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 8460 + }, + { + "epoch": 6.757987220447284, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 8461 + }, + { + "epoch": 6.758785942492013, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 8462 + }, + { + "epoch": 6.7595846645367414, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0376, + "step": 8463 + }, + { + "epoch": 6.76038338658147, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 8464 + }, + { + "epoch": 6.761182108626198, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 8465 + }, + { + "epoch": 6.761980830670926, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 8466 + }, + { + "epoch": 6.762779552715655, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0495, + "step": 8467 + }, + { + "epoch": 6.763578274760383, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.036, + "step": 8468 + }, + { + "epoch": 6.764376996805112, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0388, + "step": 8469 + }, + { + "epoch": 6.76517571884984, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 8470 + }, + { + "epoch": 6.765974440894569, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0404, + "step": 8471 + }, + { + "epoch": 6.766773162939297, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 8472 + }, + { + "epoch": 6.767571884984026, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0353, + "step": 8473 + }, + { + "epoch": 6.768370607028754, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0423, + "step": 8474 + }, + { + "epoch": 6.769169329073483, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 8475 + }, + { + "epoch": 6.7699680511182105, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 8476 + }, + { + "epoch": 6.770766773162939, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 8477 + }, + { + "epoch": 6.771565495207668, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 8478 + }, + { + "epoch": 6.772364217252396, + "grad_norm": 0.2001953125, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 8479 + }, + { + "epoch": 6.773162939297125, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0369, + "step": 8480 + }, + { + "epoch": 6.773961661341853, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.037, + "step": 8481 + }, + { + "epoch": 6.774760383386582, + "grad_norm": 0.1533203125, + "learning_rate": 0.0005, + "loss": 1.0493, + "step": 8482 + }, + { + "epoch": 6.77555910543131, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 8483 + }, + { + "epoch": 6.776357827476039, + "grad_norm": 0.173828125, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 8484 + }, + { + "epoch": 6.777156549520766, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0416, + "step": 8485 + }, + { + "epoch": 6.777955271565495, + "grad_norm": 0.1689453125, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 8486 + }, + { + "epoch": 6.7787539936102235, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 8487 + }, + { + "epoch": 6.779552715654952, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 8488 + }, + { + "epoch": 6.7803514376996805, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0393, + "step": 8489 + }, + { + "epoch": 6.781150159744409, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 8490 + }, + { + "epoch": 6.781948881789138, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0402, + "step": 8491 + }, + { + "epoch": 6.782747603833866, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 8492 + }, + { + "epoch": 6.783546325878595, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 8493 + }, + { + "epoch": 6.784345047923322, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 8494 + }, + { + "epoch": 6.785143769968051, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.041, + "step": 8495 + }, + { + "epoch": 6.785942492012779, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 8496 + }, + { + "epoch": 6.786741214057508, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 8497 + }, + { + "epoch": 6.787539936102236, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 8498 + }, + { + "epoch": 6.788338658146965, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0373, + "step": 8499 + }, + { + "epoch": 6.789137380191693, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0354, + "step": 8500 + }, + { + "epoch": 6.789936102236422, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 8501 + }, + { + "epoch": 6.7907348242811505, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 8502 + }, + { + "epoch": 6.791533546325878, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 8503 + }, + { + "epoch": 6.792332268370607, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0409, + "step": 8504 + }, + { + "epoch": 6.793130990415335, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 8505 + }, + { + "epoch": 6.793929712460064, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 8506 + }, + { + "epoch": 6.794728434504792, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 8507 + }, + { + "epoch": 6.795527156549521, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 8508 + }, + { + "epoch": 6.796325878594249, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 8509 + }, + { + "epoch": 6.797124600638978, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 8510 + }, + { + "epoch": 6.797923322683706, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 8511 + }, + { + "epoch": 6.798722044728435, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0407, + "step": 8512 + }, + { + "epoch": 6.799520766773163, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0404, + "step": 8513 + }, + { + "epoch": 6.800319488817891, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0457, + "step": 8514 + }, + { + "epoch": 6.80111821086262, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 8515 + }, + { + "epoch": 6.801916932907348, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 8516 + }, + { + "epoch": 6.802715654952077, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 8517 + }, + { + "epoch": 6.803514376996805, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 8518 + }, + { + "epoch": 6.804313099041534, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0372, + "step": 8519 + }, + { + "epoch": 6.805111821086262, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 8520 + }, + { + "epoch": 6.805910543130991, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0409, + "step": 8521 + }, + { + "epoch": 6.806709265175719, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 8522 + }, + { + "epoch": 6.807507987220447, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 8523 + }, + { + "epoch": 6.8083067092651754, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0373, + "step": 8524 + }, + { + "epoch": 6.809105431309904, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 8525 + }, + { + "epoch": 6.8099041533546325, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 8526 + }, + { + "epoch": 6.810702875399361, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0375, + "step": 8527 + }, + { + "epoch": 6.81150159744409, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 8528 + }, + { + "epoch": 6.812300319488818, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.046, + "step": 8529 + }, + { + "epoch": 6.813099041533547, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0369, + "step": 8530 + }, + { + "epoch": 6.813897763578275, + "grad_norm": 0.18359375, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 8531 + }, + { + "epoch": 6.814696485623003, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 8532 + }, + { + "epoch": 6.815495207667731, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 8533 + }, + { + "epoch": 6.81629392971246, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 8534 + }, + { + "epoch": 6.817092651757188, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 8535 + }, + { + "epoch": 6.817891373801917, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 8536 + }, + { + "epoch": 6.818690095846645, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 8537 + }, + { + "epoch": 6.819488817891374, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 8538 + }, + { + "epoch": 6.8202875399361025, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0386, + "step": 8539 + }, + { + "epoch": 6.821086261980831, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0401, + "step": 8540 + }, + { + "epoch": 6.821884984025559, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0404, + "step": 8541 + }, + { + "epoch": 6.822683706070287, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0361, + "step": 8542 + }, + { + "epoch": 6.823482428115016, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 8543 + }, + { + "epoch": 6.824281150159744, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 8544 + }, + { + "epoch": 6.825079872204473, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 8545 + }, + { + "epoch": 6.825878594249201, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 8546 + }, + { + "epoch": 6.82667731629393, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 8547 + }, + { + "epoch": 6.827476038338658, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 8548 + }, + { + "epoch": 6.828274760383387, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 8549 + }, + { + "epoch": 6.8290734824281145, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0379, + "step": 8550 + }, + { + "epoch": 6.829872204472844, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.034, + "step": 8551 + }, + { + "epoch": 6.830670926517572, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0407, + "step": 8552 + }, + { + "epoch": 6.8314696485623, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0384, + "step": 8553 + }, + { + "epoch": 6.832268370607029, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 8554 + }, + { + "epoch": 6.833067092651757, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 8555 + }, + { + "epoch": 6.833865814696486, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 8556 + }, + { + "epoch": 6.834664536741214, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.041, + "step": 8557 + }, + { + "epoch": 6.835463258785943, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 8558 + }, + { + "epoch": 6.836261980830671, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 8559 + }, + { + "epoch": 6.8370607028754, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 8560 + }, + { + "epoch": 6.837859424920127, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 8561 + }, + { + "epoch": 6.838658146964856, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 8562 + }, + { + "epoch": 6.8394568690095845, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 8563 + }, + { + "epoch": 6.840255591054313, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 8564 + }, + { + "epoch": 6.8410543130990416, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 8565 + }, + { + "epoch": 6.84185303514377, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 8566 + }, + { + "epoch": 6.842651757188499, + "grad_norm": 0.1689453125, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 8567 + }, + { + "epoch": 6.843450479233227, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 8568 + }, + { + "epoch": 6.844249201277956, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 8569 + }, + { + "epoch": 6.845047923322683, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 8570 + }, + { + "epoch": 6.845846645367412, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0459, + "step": 8571 + }, + { + "epoch": 6.84664536741214, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 8572 + }, + { + "epoch": 6.847444089456869, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 8573 + }, + { + "epoch": 6.848242811501597, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 8574 + }, + { + "epoch": 6.849041533546326, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0338, + "step": 8575 + }, + { + "epoch": 6.8498402555910545, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 8576 + }, + { + "epoch": 6.850638977635783, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 8577 + }, + { + "epoch": 6.8514376996805115, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 8578 + }, + { + "epoch": 6.852236421725239, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 8579 + }, + { + "epoch": 6.853035143769968, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 8580 + }, + { + "epoch": 6.853833865814696, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 8581 + }, + { + "epoch": 6.854632587859425, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 8582 + }, + { + "epoch": 6.855431309904153, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 8583 + }, + { + "epoch": 6.856230031948882, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 8584 + }, + { + "epoch": 6.85702875399361, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0408, + "step": 8585 + }, + { + "epoch": 6.857827476038339, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 8586 + }, + { + "epoch": 6.858626198083067, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 8587 + }, + { + "epoch": 6.859424920127795, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0401, + "step": 8588 + }, + { + "epoch": 6.8602236421725244, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 8589 + }, + { + "epoch": 6.861022364217252, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0385, + "step": 8590 + }, + { + "epoch": 6.861821086261981, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0393, + "step": 8591 + }, + { + "epoch": 6.862619808306709, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0354, + "step": 8592 + }, + { + "epoch": 6.863418530351438, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 8593 + }, + { + "epoch": 6.864217252396166, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 8594 + }, + { + "epoch": 6.865015974440895, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.038, + "step": 8595 + }, + { + "epoch": 6.865814696485623, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0416, + "step": 8596 + }, + { + "epoch": 6.866613418530352, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0337, + "step": 8597 + }, + { + "epoch": 6.86741214057508, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0354, + "step": 8598 + }, + { + "epoch": 6.868210862619808, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0373, + "step": 8599 + }, + { + "epoch": 6.8690095846645365, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 8600 + }, + { + "epoch": 6.869808306709265, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 8601 + }, + { + "epoch": 6.8706070287539935, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 8602 + }, + { + "epoch": 6.871405750798722, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0377, + "step": 8603 + }, + { + "epoch": 6.872204472843451, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0427, + "step": 8604 + }, + { + "epoch": 6.873003194888179, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0394, + "step": 8605 + }, + { + "epoch": 6.873801916932908, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0354, + "step": 8606 + }, + { + "epoch": 6.874600638977636, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 8607 + }, + { + "epoch": 6.875399361022364, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.041, + "step": 8608 + }, + { + "epoch": 6.876198083067092, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 8609 + }, + { + "epoch": 6.876996805111821, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 8610 + }, + { + "epoch": 6.877795527156549, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 8611 + }, + { + "epoch": 6.878594249201278, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 8612 + }, + { + "epoch": 6.8793929712460065, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 8613 + }, + { + "epoch": 6.880191693290735, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 8614 + }, + { + "epoch": 6.8809904153354635, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 8615 + }, + { + "epoch": 6.881789137380192, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 8616 + }, + { + "epoch": 6.88258785942492, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 8617 + }, + { + "epoch": 6.883386581469648, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 8618 + }, + { + "epoch": 6.884185303514377, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0436, + "step": 8619 + }, + { + "epoch": 6.884984025559105, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 8620 + }, + { + "epoch": 6.885782747603834, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 8621 + }, + { + "epoch": 6.886581469648562, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 8622 + }, + { + "epoch": 6.887380191693291, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0348, + "step": 8623 + }, + { + "epoch": 6.888178913738019, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 8624 + }, + { + "epoch": 6.888977635782748, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0361, + "step": 8625 + }, + { + "epoch": 6.8897763578274756, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 8626 + }, + { + "epoch": 6.890575079872205, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 8627 + }, + { + "epoch": 6.891373801916933, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 8628 + }, + { + "epoch": 6.892172523961661, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 8629 + }, + { + "epoch": 6.89297124600639, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0348, + "step": 8630 + }, + { + "epoch": 6.893769968051118, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 8631 + }, + { + "epoch": 6.894568690095847, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 8632 + }, + { + "epoch": 6.895367412140575, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0338, + "step": 8633 + }, + { + "epoch": 6.896166134185304, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0353, + "step": 8634 + }, + { + "epoch": 6.896964856230032, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 8635 + }, + { + "epoch": 6.897763578274761, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 8636 + }, + { + "epoch": 6.8985623003194885, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 8637 + }, + { + "epoch": 6.899361022364217, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 8638 + }, + { + "epoch": 6.9001597444089455, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0338, + "step": 8639 + }, + { + "epoch": 6.900958466453674, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 8640 + }, + { + "epoch": 6.901757188498403, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 8641 + }, + { + "epoch": 6.902555910543131, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 8642 + }, + { + "epoch": 6.90335463258786, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 8643 + }, + { + "epoch": 6.904153354632588, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 8644 + }, + { + "epoch": 6.904952076677317, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 8645 + }, + { + "epoch": 6.905750798722044, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 8646 + }, + { + "epoch": 6.906549520766773, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0354, + "step": 8647 + }, + { + "epoch": 6.907348242811501, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 8648 + }, + { + "epoch": 6.90814696485623, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0431, + "step": 8649 + }, + { + "epoch": 6.9089456869009584, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 8650 + }, + { + "epoch": 6.909744408945687, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 8651 + }, + { + "epoch": 6.9105431309904155, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 8652 + }, + { + "epoch": 6.911341853035144, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 8653 + }, + { + "epoch": 6.912140575079873, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 8654 + }, + { + "epoch": 6.9129392971246, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0422, + "step": 8655 + }, + { + "epoch": 6.913738019169329, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 8656 + }, + { + "epoch": 6.914536741214057, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 8657 + }, + { + "epoch": 6.915335463258786, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 8658 + }, + { + "epoch": 6.916134185303514, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0369, + "step": 8659 + }, + { + "epoch": 6.916932907348243, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0354, + "step": 8660 + }, + { + "epoch": 6.917731629392971, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0453, + "step": 8661 + }, + { + "epoch": 6.9185303514377, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 8662 + }, + { + "epoch": 6.919329073482428, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0453, + "step": 8663 + }, + { + "epoch": 6.920127795527156, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 8664 + }, + { + "epoch": 6.9209265175718855, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0472, + "step": 8665 + }, + { + "epoch": 6.921725239616613, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0361, + "step": 8666 + }, + { + "epoch": 6.922523961661342, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.036, + "step": 8667 + }, + { + "epoch": 6.92332268370607, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 8668 + }, + { + "epoch": 6.924121405750799, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 8669 + }, + { + "epoch": 6.924920127795527, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0412, + "step": 8670 + }, + { + "epoch": 6.925718849840256, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 8671 + }, + { + "epoch": 6.926517571884984, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0353, + "step": 8672 + }, + { + "epoch": 6.927316293929713, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0394, + "step": 8673 + }, + { + "epoch": 6.928115015974441, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 8674 + }, + { + "epoch": 6.928913738019169, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 8675 + }, + { + "epoch": 6.9297124600638975, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0361, + "step": 8676 + }, + { + "epoch": 6.930511182108626, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 8677 + }, + { + "epoch": 6.931309904153355, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 8678 + }, + { + "epoch": 6.932108626198083, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 8679 + }, + { + "epoch": 6.932907348242812, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0378, + "step": 8680 + }, + { + "epoch": 6.93370607028754, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0362, + "step": 8681 + }, + { + "epoch": 6.934504792332269, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0434, + "step": 8682 + }, + { + "epoch": 6.935303514376997, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 8683 + }, + { + "epoch": 6.936102236421725, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0391, + "step": 8684 + }, + { + "epoch": 6.936900958466453, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0369, + "step": 8685 + }, + { + "epoch": 6.937699680511182, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 8686 + }, + { + "epoch": 6.93849840255591, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0419, + "step": 8687 + }, + { + "epoch": 6.939297124600639, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 8688 + }, + { + "epoch": 6.9400958466453675, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 8689 + }, + { + "epoch": 6.940894568690096, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 8690 + }, + { + "epoch": 6.9416932907348246, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 8691 + }, + { + "epoch": 6.942492012779553, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0453, + "step": 8692 + }, + { + "epoch": 6.943290734824281, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 8693 + }, + { + "epoch": 6.944089456869009, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 8694 + }, + { + "epoch": 6.944888178913738, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 8695 + }, + { + "epoch": 6.945686900958466, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0431, + "step": 8696 + }, + { + "epoch": 6.946485623003195, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 8697 + }, + { + "epoch": 6.947284345047923, + "grad_norm": 0.1669921875, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 8698 + }, + { + "epoch": 6.948083067092652, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 8699 + }, + { + "epoch": 6.94888178913738, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 8700 + }, + { + "epoch": 6.949680511182109, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 8701 + }, + { + "epoch": 6.950479233226837, + "grad_norm": 0.177734375, + "learning_rate": 0.0005, + "loss": 1.0399, + "step": 8702 + }, + { + "epoch": 6.951277955271565, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0386, + "step": 8703 + }, + { + "epoch": 6.952076677316294, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.04, + "step": 8704 + }, + { + "epoch": 6.952875399361022, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 8705 + }, + { + "epoch": 6.953674121405751, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 8706 + }, + { + "epoch": 6.954472843450479, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 8707 + }, + { + "epoch": 6.955271565495208, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0404, + "step": 8708 + }, + { + "epoch": 6.956070287539936, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 8709 + }, + { + "epoch": 6.956869009584665, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0379, + "step": 8710 + }, + { + "epoch": 6.957667731629393, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 8711 + }, + { + "epoch": 6.958466453674122, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 8712 + }, + { + "epoch": 6.9592651757188495, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 8713 + }, + { + "epoch": 6.960063897763578, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 8714 + }, + { + "epoch": 6.960862619808307, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 8715 + }, + { + "epoch": 6.961661341853035, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 8716 + }, + { + "epoch": 6.962460063897764, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0386, + "step": 8717 + }, + { + "epoch": 6.963258785942492, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 8718 + }, + { + "epoch": 6.964057507987221, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 8719 + }, + { + "epoch": 6.964856230031949, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0461, + "step": 8720 + }, + { + "epoch": 6.965654952076678, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0378, + "step": 8721 + }, + { + "epoch": 6.966453674121405, + "grad_norm": 0.1748046875, + "learning_rate": 0.0005, + "loss": 1.0411, + "step": 8722 + }, + { + "epoch": 6.967252396166134, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 8723 + }, + { + "epoch": 6.968051118210862, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 8724 + }, + { + "epoch": 6.968849840255591, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 8725 + }, + { + "epoch": 6.9696485623003195, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0388, + "step": 8726 + }, + { + "epoch": 6.970447284345048, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0376, + "step": 8727 + }, + { + "epoch": 6.9712460063897765, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 8728 + }, + { + "epoch": 6.972044728434505, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 8729 + }, + { + "epoch": 6.972843450479234, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0354, + "step": 8730 + }, + { + "epoch": 6.973642172523961, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 8731 + }, + { + "epoch": 6.97444089456869, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 8732 + }, + { + "epoch": 6.975239616613418, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 8733 + }, + { + "epoch": 6.976038338658147, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 8734 + }, + { + "epoch": 6.976837060702875, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0353, + "step": 8735 + }, + { + "epoch": 6.977635782747604, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0424, + "step": 8736 + }, + { + "epoch": 6.978434504792332, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 8737 + }, + { + "epoch": 6.979233226837061, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 8738 + }, + { + "epoch": 6.9800319488817895, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 8739 + }, + { + "epoch": 6.980830670926517, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 8740 + }, + { + "epoch": 6.981629392971246, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 8741 + }, + { + "epoch": 6.982428115015974, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 8742 + }, + { + "epoch": 6.983226837060703, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 8743 + }, + { + "epoch": 6.984025559105431, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.044, + "step": 8744 + }, + { + "epoch": 6.98482428115016, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 8745 + }, + { + "epoch": 6.985623003194888, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 8746 + }, + { + "epoch": 6.986421725239617, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 8747 + }, + { + "epoch": 6.987220447284345, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 8748 + }, + { + "epoch": 6.988019169329074, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 8749 + }, + { + "epoch": 6.988817891373802, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 8750 + }, + { + "epoch": 6.98961661341853, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 8751 + }, + { + "epoch": 6.9904153354632586, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 8752 + }, + { + "epoch": 6.991214057507987, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 8753 + }, + { + "epoch": 6.992012779552716, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 8754 + }, + { + "epoch": 6.992811501597444, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 8755 + }, + { + "epoch": 6.993610223642173, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 8756 + }, + { + "epoch": 6.994408945686901, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0338, + "step": 8757 + }, + { + "epoch": 6.99520766773163, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 8758 + }, + { + "epoch": 6.996006389776358, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.04, + "step": 8759 + }, + { + "epoch": 6.996805111821086, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 8760 + }, + { + "epoch": 6.997603833865814, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0394, + "step": 8761 + }, + { + "epoch": 6.998402555910543, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0379, + "step": 8762 + }, + { + "epoch": 6.9992012779552715, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 8763 + }, + { + "epoch": 7.0, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 8764 + }, + { + "epoch": 7.0007987220447285, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 8765 + }, + { + "epoch": 7.001597444089457, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0386, + "step": 8766 + }, + { + "epoch": 7.002396166134186, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.051, + "step": 8767 + }, + { + "epoch": 7.003194888178914, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 8768 + }, + { + "epoch": 7.003993610223642, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 8769 + }, + { + "epoch": 7.00479233226837, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0404, + "step": 8770 + }, + { + "epoch": 7.005591054313099, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 8771 + }, + { + "epoch": 7.006389776357827, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0372, + "step": 8772 + }, + { + "epoch": 7.007188498402556, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 8773 + }, + { + "epoch": 7.007987220447284, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 8774 + }, + { + "epoch": 7.008785942492013, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 8775 + }, + { + "epoch": 7.0095846645367414, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 8776 + }, + { + "epoch": 7.01038338658147, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 8777 + }, + { + "epoch": 7.0111821086261985, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0424, + "step": 8778 + }, + { + "epoch": 7.011980830670926, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 8779 + }, + { + "epoch": 7.012779552715655, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 8780 + }, + { + "epoch": 7.013578274760383, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 8781 + }, + { + "epoch": 7.014376996805112, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 8782 + }, + { + "epoch": 7.01517571884984, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 8783 + }, + { + "epoch": 7.015974440894569, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.044, + "step": 8784 + }, + { + "epoch": 7.016773162939297, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0419, + "step": 8785 + }, + { + "epoch": 7.017571884984026, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0421, + "step": 8786 + }, + { + "epoch": 7.018370607028754, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 8787 + }, + { + "epoch": 7.019169329073482, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 8788 + }, + { + "epoch": 7.0199680511182105, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 8789 + }, + { + "epoch": 7.020766773162939, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 8790 + }, + { + "epoch": 7.021565495207668, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0438, + "step": 8791 + }, + { + "epoch": 7.022364217252396, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 8792 + }, + { + "epoch": 7.023162939297125, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 8793 + }, + { + "epoch": 7.023961661341853, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 8794 + }, + { + "epoch": 7.024760383386582, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 8795 + }, + { + "epoch": 7.02555910543131, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 8796 + }, + { + "epoch": 7.026357827476039, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 8797 + }, + { + "epoch": 7.027156549520766, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 8798 + }, + { + "epoch": 7.027955271565495, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 8799 + }, + { + "epoch": 7.0287539936102235, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 8800 + }, + { + "epoch": 7.029552715654952, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 8801 + }, + { + "epoch": 7.0303514376996805, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 8802 + }, + { + "epoch": 7.031150159744409, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0419, + "step": 8803 + }, + { + "epoch": 7.031948881789138, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 8804 + }, + { + "epoch": 7.032747603833866, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 8805 + }, + { + "epoch": 7.033546325878595, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 8806 + }, + { + "epoch": 7.034345047923322, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.041, + "step": 8807 + }, + { + "epoch": 7.035143769968051, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 8808 + }, + { + "epoch": 7.035942492012779, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 8809 + }, + { + "epoch": 7.036741214057508, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 8810 + }, + { + "epoch": 7.037539936102236, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 8811 + }, + { + "epoch": 7.038338658146965, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.042, + "step": 8812 + }, + { + "epoch": 7.039137380191693, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 8813 + }, + { + "epoch": 7.039936102236422, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 8814 + }, + { + "epoch": 7.0407348242811505, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 8815 + }, + { + "epoch": 7.041533546325879, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 8816 + }, + { + "epoch": 7.042332268370607, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 8817 + }, + { + "epoch": 7.043130990415335, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 8818 + }, + { + "epoch": 7.043929712460064, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 8819 + }, + { + "epoch": 7.044728434504792, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 8820 + }, + { + "epoch": 7.045527156549521, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 8821 + }, + { + "epoch": 7.046325878594249, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 8822 + }, + { + "epoch": 7.047124600638978, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0394, + "step": 8823 + }, + { + "epoch": 7.047923322683706, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0385, + "step": 8824 + }, + { + "epoch": 7.048722044728435, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0407, + "step": 8825 + }, + { + "epoch": 7.0495207667731625, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 8826 + }, + { + "epoch": 7.050319488817891, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 8827 + }, + { + "epoch": 7.05111821086262, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 8828 + }, + { + "epoch": 7.051916932907348, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 8829 + }, + { + "epoch": 7.052715654952077, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 8830 + }, + { + "epoch": 7.053514376996805, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 8831 + }, + { + "epoch": 7.054313099041534, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 8832 + }, + { + "epoch": 7.055111821086262, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 8833 + }, + { + "epoch": 7.055910543130991, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0394, + "step": 8834 + }, + { + "epoch": 7.056709265175719, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 8835 + }, + { + "epoch": 7.057507987220447, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 8836 + }, + { + "epoch": 7.0583067092651754, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 8837 + }, + { + "epoch": 7.059105431309904, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0388, + "step": 8838 + }, + { + "epoch": 7.0599041533546325, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 8839 + }, + { + "epoch": 7.060702875399361, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 8840 + }, + { + "epoch": 7.06150159744409, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 8841 + }, + { + "epoch": 7.062300319488818, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.034, + "step": 8842 + }, + { + "epoch": 7.063099041533547, + "grad_norm": 0.16015625, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 8843 + }, + { + "epoch": 7.063897763578275, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 8844 + }, + { + "epoch": 7.064696485623003, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 8845 + }, + { + "epoch": 7.065495207667731, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 8846 + }, + { + "epoch": 7.06629392971246, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 8847 + }, + { + "epoch": 7.067092651757188, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0406, + "step": 8848 + }, + { + "epoch": 7.067891373801917, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 8849 + }, + { + "epoch": 7.068690095846645, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 8850 + }, + { + "epoch": 7.069488817891374, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 8851 + }, + { + "epoch": 7.0702875399361025, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 8852 + }, + { + "epoch": 7.071086261980831, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0441, + "step": 8853 + }, + { + "epoch": 7.0718849840255595, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.036, + "step": 8854 + }, + { + "epoch": 7.072683706070287, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 8855 + }, + { + "epoch": 7.073482428115016, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 8856 + }, + { + "epoch": 7.074281150159744, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 8857 + }, + { + "epoch": 7.075079872204473, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 8858 + }, + { + "epoch": 7.075878594249201, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 8859 + }, + { + "epoch": 7.07667731629393, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0379, + "step": 8860 + }, + { + "epoch": 7.077476038338658, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 8861 + }, + { + "epoch": 7.078274760383387, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 8862 + }, + { + "epoch": 7.079073482428115, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 8863 + }, + { + "epoch": 7.079872204472843, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 8864 + }, + { + "epoch": 7.080670926517572, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0354, + "step": 8865 + }, + { + "epoch": 7.0814696485623, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 8866 + }, + { + "epoch": 7.082268370607029, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0373, + "step": 8867 + }, + { + "epoch": 7.083067092651757, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 8868 + }, + { + "epoch": 7.083865814696486, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0205, + "step": 8869 + }, + { + "epoch": 7.084664536741214, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 8870 + }, + { + "epoch": 7.085463258785943, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0362, + "step": 8871 + }, + { + "epoch": 7.086261980830671, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 8872 + }, + { + "epoch": 7.0870607028754, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 8873 + }, + { + "epoch": 7.087859424920127, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 8874 + }, + { + "epoch": 7.088658146964856, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0369, + "step": 8875 + }, + { + "epoch": 7.0894568690095845, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 8876 + }, + { + "epoch": 7.090255591054313, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 8877 + }, + { + "epoch": 7.0910543130990416, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 8878 + }, + { + "epoch": 7.09185303514377, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 8879 + }, + { + "epoch": 7.092651757188499, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 8880 + }, + { + "epoch": 7.093450479233227, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 8881 + }, + { + "epoch": 7.094249201277956, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 8882 + }, + { + "epoch": 7.095047923322683, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 8883 + }, + { + "epoch": 7.095846645367412, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0379, + "step": 8884 + }, + { + "epoch": 7.09664536741214, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 8885 + }, + { + "epoch": 7.097444089456869, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 8886 + }, + { + "epoch": 7.098242811501597, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 8887 + }, + { + "epoch": 7.099041533546326, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0405, + "step": 8888 + }, + { + "epoch": 7.0998402555910545, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 8889 + }, + { + "epoch": 7.100638977635783, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 8890 + }, + { + "epoch": 7.1014376996805115, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 8891 + }, + { + "epoch": 7.102236421725239, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.036, + "step": 8892 + }, + { + "epoch": 7.103035143769968, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 8893 + }, + { + "epoch": 7.103833865814696, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 8894 + }, + { + "epoch": 7.104632587859425, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 8895 + }, + { + "epoch": 7.105431309904153, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 8896 + }, + { + "epoch": 7.106230031948882, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 8897 + }, + { + "epoch": 7.10702875399361, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 8898 + }, + { + "epoch": 7.107827476038339, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 8899 + }, + { + "epoch": 7.108626198083067, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 8900 + }, + { + "epoch": 7.109424920127796, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 8901 + }, + { + "epoch": 7.110223642172524, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 8902 + }, + { + "epoch": 7.111022364217252, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 8903 + }, + { + "epoch": 7.111821086261981, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.037, + "step": 8904 + }, + { + "epoch": 7.112619808306709, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 8905 + }, + { + "epoch": 7.113418530351438, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 8906 + }, + { + "epoch": 7.114217252396166, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0378, + "step": 8907 + }, + { + "epoch": 7.115015974440895, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 8908 + }, + { + "epoch": 7.115814696485623, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 8909 + }, + { + "epoch": 7.116613418530352, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.042, + "step": 8910 + }, + { + "epoch": 7.11741214057508, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0337, + "step": 8911 + }, + { + "epoch": 7.118210862619808, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 8912 + }, + { + "epoch": 7.1190095846645365, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 8913 + }, + { + "epoch": 7.119808306709265, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 8914 + }, + { + "epoch": 7.1206070287539935, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 8915 + }, + { + "epoch": 7.121405750798722, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 8916 + }, + { + "epoch": 7.122204472843451, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0455, + "step": 8917 + }, + { + "epoch": 7.123003194888179, + "grad_norm": 0.16015625, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 8918 + }, + { + "epoch": 7.123801916932908, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 8919 + }, + { + "epoch": 7.124600638977636, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 8920 + }, + { + "epoch": 7.125399361022364, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 8921 + }, + { + "epoch": 7.126198083067092, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.038, + "step": 8922 + }, + { + "epoch": 7.126996805111821, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0378, + "step": 8923 + }, + { + "epoch": 7.127795527156549, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0407, + "step": 8924 + }, + { + "epoch": 7.128594249201278, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 8925 + }, + { + "epoch": 7.1293929712460065, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 8926 + }, + { + "epoch": 7.130191693290735, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0377, + "step": 8927 + }, + { + "epoch": 7.1309904153354635, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 8928 + }, + { + "epoch": 7.131789137380192, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 8929 + }, + { + "epoch": 7.13258785942492, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 8930 + }, + { + "epoch": 7.133386581469648, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 8931 + }, + { + "epoch": 7.134185303514377, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 8932 + }, + { + "epoch": 7.134984025559105, + "grad_norm": 0.1640625, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 8933 + }, + { + "epoch": 7.135782747603834, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 8934 + }, + { + "epoch": 7.136581469648562, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 8935 + }, + { + "epoch": 7.137380191693291, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 8936 + }, + { + "epoch": 7.138178913738019, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 8937 + }, + { + "epoch": 7.138977635782748, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 8938 + }, + { + "epoch": 7.139776357827476, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 8939 + }, + { + "epoch": 7.140575079872204, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 8940 + }, + { + "epoch": 7.141373801916933, + "grad_norm": 0.16015625, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 8941 + }, + { + "epoch": 7.142172523961661, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 8942 + }, + { + "epoch": 7.14297124600639, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 8943 + }, + { + "epoch": 7.143769968051118, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 8944 + }, + { + "epoch": 7.144568690095847, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 8945 + }, + { + "epoch": 7.145367412140575, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 8946 + }, + { + "epoch": 7.146166134185304, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 8947 + }, + { + "epoch": 7.146964856230032, + "grad_norm": 0.2099609375, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 8948 + }, + { + "epoch": 7.147763578274761, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 8949 + }, + { + "epoch": 7.1485623003194885, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 8950 + }, + { + "epoch": 7.149361022364217, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.034, + "step": 8951 + }, + { + "epoch": 7.1501597444089455, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 8952 + }, + { + "epoch": 7.150958466453674, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 8953 + }, + { + "epoch": 7.151757188498403, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 8954 + }, + { + "epoch": 7.152555910543131, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0427, + "step": 8955 + }, + { + "epoch": 7.15335463258786, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 8956 + }, + { + "epoch": 7.154153354632588, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 8957 + }, + { + "epoch": 7.154952076677317, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 8958 + }, + { + "epoch": 7.155750798722044, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 8959 + }, + { + "epoch": 7.156549520766773, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0404, + "step": 8960 + }, + { + "epoch": 7.157348242811501, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0378, + "step": 8961 + }, + { + "epoch": 7.15814696485623, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0378, + "step": 8962 + }, + { + "epoch": 7.1589456869009584, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0373, + "step": 8963 + }, + { + "epoch": 7.159744408945687, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0362, + "step": 8964 + }, + { + "epoch": 7.1605431309904155, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 8965 + }, + { + "epoch": 7.161341853035144, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 8966 + }, + { + "epoch": 7.162140575079873, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 8967 + }, + { + "epoch": 7.1629392971246, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.037, + "step": 8968 + }, + { + "epoch": 7.163738019169329, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 8969 + }, + { + "epoch": 7.164536741214057, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 8970 + }, + { + "epoch": 7.165335463258786, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 8971 + }, + { + "epoch": 7.166134185303514, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 8972 + }, + { + "epoch": 7.166932907348243, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 8973 + }, + { + "epoch": 7.167731629392971, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 8974 + }, + { + "epoch": 7.1685303514377, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 8975 + }, + { + "epoch": 7.169329073482428, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 8976 + }, + { + "epoch": 7.170127795527157, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.043, + "step": 8977 + }, + { + "epoch": 7.170926517571885, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 8978 + }, + { + "epoch": 7.171725239616613, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 8979 + }, + { + "epoch": 7.172523961661342, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 8980 + }, + { + "epoch": 7.17332268370607, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 8981 + }, + { + "epoch": 7.174121405750799, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0338, + "step": 8982 + }, + { + "epoch": 7.174920127795527, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 8983 + }, + { + "epoch": 7.175718849840256, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 8984 + }, + { + "epoch": 7.176517571884984, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 8985 + }, + { + "epoch": 7.177316293929713, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0338, + "step": 8986 + }, + { + "epoch": 7.178115015974441, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 8987 + }, + { + "epoch": 7.178913738019169, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 8988 + }, + { + "epoch": 7.1797124600638975, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 8989 + }, + { + "epoch": 7.180511182108626, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 8990 + }, + { + "epoch": 7.181309904153355, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 8991 + }, + { + "epoch": 7.182108626198083, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 8992 + }, + { + "epoch": 7.182907348242812, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 8993 + }, + { + "epoch": 7.18370607028754, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 8994 + }, + { + "epoch": 7.184504792332269, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 8995 + }, + { + "epoch": 7.185303514376997, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 8996 + }, + { + "epoch": 7.186102236421725, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 8997 + }, + { + "epoch": 7.186900958466453, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0401, + "step": 8998 + }, + { + "epoch": 7.187699680511182, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 8999 + }, + { + "epoch": 7.18849840255591, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0404, + "step": 9000 + }, + { + "epoch": 7.189297124600639, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 9001 + }, + { + "epoch": 7.1900958466453675, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0354, + "step": 9002 + }, + { + "epoch": 7.190894568690096, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 9003 + }, + { + "epoch": 7.1916932907348246, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 9004 + }, + { + "epoch": 7.192492012779553, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 9005 + }, + { + "epoch": 7.193290734824281, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 9006 + }, + { + "epoch": 7.194089456869009, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 9007 + }, + { + "epoch": 7.194888178913738, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 9008 + }, + { + "epoch": 7.195686900958466, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.036, + "step": 9009 + }, + { + "epoch": 7.196485623003195, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 9010 + }, + { + "epoch": 7.197284345047923, + "grad_norm": 0.169921875, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 9011 + }, + { + "epoch": 7.198083067092652, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0369, + "step": 9012 + }, + { + "epoch": 7.19888178913738, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 9013 + }, + { + "epoch": 7.199680511182109, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 9014 + }, + { + "epoch": 7.2004792332268375, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0373, + "step": 9015 + }, + { + "epoch": 7.201277955271565, + "grad_norm": 0.1669921875, + "learning_rate": 0.0005, + "loss": 1.0377, + "step": 9016 + }, + { + "epoch": 7.202076677316294, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 9017 + }, + { + "epoch": 7.202875399361022, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 9018 + }, + { + "epoch": 7.203674121405751, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0407, + "step": 9019 + }, + { + "epoch": 7.204472843450479, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 9020 + }, + { + "epoch": 7.205271565495208, + "grad_norm": 0.1640625, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 9021 + }, + { + "epoch": 7.206070287539936, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0354, + "step": 9022 + }, + { + "epoch": 7.206869009584665, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 9023 + }, + { + "epoch": 7.207667731629393, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 9024 + }, + { + "epoch": 7.208466453674121, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 9025 + }, + { + "epoch": 7.2092651757188495, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 9026 + }, + { + "epoch": 7.210063897763578, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0348, + "step": 9027 + }, + { + "epoch": 7.210862619808307, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 9028 + }, + { + "epoch": 7.211661341853035, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 9029 + }, + { + "epoch": 7.212460063897764, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 9030 + }, + { + "epoch": 7.213258785942492, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0378, + "step": 9031 + }, + { + "epoch": 7.214057507987221, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 9032 + }, + { + "epoch": 7.214856230031949, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 9033 + }, + { + "epoch": 7.215654952076678, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 9034 + }, + { + "epoch": 7.216453674121405, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 9035 + }, + { + "epoch": 7.217252396166134, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 9036 + }, + { + "epoch": 7.218051118210862, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 9037 + }, + { + "epoch": 7.218849840255591, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 9038 + }, + { + "epoch": 7.2196485623003195, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 9039 + }, + { + "epoch": 7.220447284345048, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0394, + "step": 9040 + }, + { + "epoch": 7.2212460063897765, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 9041 + }, + { + "epoch": 7.222044728434505, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 9042 + }, + { + "epoch": 7.222843450479234, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 9043 + }, + { + "epoch": 7.223642172523961, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 9044 + }, + { + "epoch": 7.22444089456869, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0384, + "step": 9045 + }, + { + "epoch": 7.225239616613418, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0353, + "step": 9046 + }, + { + "epoch": 7.226038338658147, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 9047 + }, + { + "epoch": 7.226837060702875, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0419, + "step": 9048 + }, + { + "epoch": 7.227635782747604, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 9049 + }, + { + "epoch": 7.228434504792332, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 9050 + }, + { + "epoch": 7.229233226837061, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 9051 + }, + { + "epoch": 7.2300319488817895, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 9052 + }, + { + "epoch": 7.230830670926518, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 9053 + }, + { + "epoch": 7.231629392971246, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 9054 + }, + { + "epoch": 7.232428115015974, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0378, + "step": 9055 + }, + { + "epoch": 7.233226837060703, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 9056 + }, + { + "epoch": 7.234025559105431, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 9057 + }, + { + "epoch": 7.23482428115016, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 9058 + }, + { + "epoch": 7.235623003194888, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 9059 + }, + { + "epoch": 7.236421725239617, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 9060 + }, + { + "epoch": 7.237220447284345, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 9061 + }, + { + "epoch": 7.238019169329074, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 9062 + }, + { + "epoch": 7.2388178913738015, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 9063 + }, + { + "epoch": 7.23961661341853, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 9064 + }, + { + "epoch": 7.2404153354632586, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0413, + "step": 9065 + }, + { + "epoch": 7.241214057507987, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 9066 + }, + { + "epoch": 7.242012779552716, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 9067 + }, + { + "epoch": 7.242811501597444, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 9068 + }, + { + "epoch": 7.243610223642173, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0391, + "step": 9069 + }, + { + "epoch": 7.244408945686901, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 9070 + }, + { + "epoch": 7.24520766773163, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 9071 + }, + { + "epoch": 7.246006389776358, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0379, + "step": 9072 + }, + { + "epoch": 7.246805111821086, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 9073 + }, + { + "epoch": 7.247603833865814, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 9074 + }, + { + "epoch": 7.248402555910543, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 9075 + }, + { + "epoch": 7.2492012779552715, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0379, + "step": 9076 + }, + { + "epoch": 7.25, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 9077 + }, + { + "epoch": 7.2507987220447285, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 9078 + }, + { + "epoch": 7.251597444089457, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.036, + "step": 9079 + }, + { + "epoch": 7.252396166134186, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 9080 + }, + { + "epoch": 7.253194888178914, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 9081 + }, + { + "epoch": 7.253993610223642, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 9082 + }, + { + "epoch": 7.25479233226837, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 9083 + }, + { + "epoch": 7.255591054313099, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 9084 + }, + { + "epoch": 7.256389776357827, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 9085 + }, + { + "epoch": 7.257188498402556, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 9086 + }, + { + "epoch": 7.257987220447284, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 9087 + }, + { + "epoch": 7.258785942492013, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 9088 + }, + { + "epoch": 7.2595846645367414, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 9089 + }, + { + "epoch": 7.26038338658147, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 9090 + }, + { + "epoch": 7.261182108626198, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 9091 + }, + { + "epoch": 7.261980830670926, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 9092 + }, + { + "epoch": 7.262779552715655, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 9093 + }, + { + "epoch": 7.263578274760383, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 9094 + }, + { + "epoch": 7.264376996805112, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 9095 + }, + { + "epoch": 7.26517571884984, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0353, + "step": 9096 + }, + { + "epoch": 7.265974440894569, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 9097 + }, + { + "epoch": 7.266773162939297, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 9098 + }, + { + "epoch": 7.267571884984026, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 9099 + }, + { + "epoch": 7.268370607028754, + "grad_norm": 0.185546875, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 9100 + }, + { + "epoch": 7.269169329073483, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 9101 + }, + { + "epoch": 7.2699680511182105, + "grad_norm": 0.16796875, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 9102 + }, + { + "epoch": 7.270766773162939, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 9103 + }, + { + "epoch": 7.271565495207668, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.0401, + "step": 9104 + }, + { + "epoch": 7.272364217252396, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 9105 + }, + { + "epoch": 7.273162939297125, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 9106 + }, + { + "epoch": 7.273961661341853, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 9107 + }, + { + "epoch": 7.274760383386582, + "grad_norm": 0.16015625, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 9108 + }, + { + "epoch": 7.27555910543131, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 9109 + }, + { + "epoch": 7.276357827476039, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 9110 + }, + { + "epoch": 7.277156549520766, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0419, + "step": 9111 + }, + { + "epoch": 7.277955271565495, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 9112 + }, + { + "epoch": 7.2787539936102235, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 9113 + }, + { + "epoch": 7.279552715654952, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 9114 + }, + { + "epoch": 7.2803514376996805, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0473, + "step": 9115 + }, + { + "epoch": 7.281150159744409, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 9116 + }, + { + "epoch": 7.281948881789138, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 9117 + }, + { + "epoch": 7.282747603833866, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 9118 + }, + { + "epoch": 7.283546325878595, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 9119 + }, + { + "epoch": 7.284345047923322, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 9120 + }, + { + "epoch": 7.285143769968051, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 9121 + }, + { + "epoch": 7.285942492012779, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 9122 + }, + { + "epoch": 7.286741214057508, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 9123 + }, + { + "epoch": 7.287539936102236, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 9124 + }, + { + "epoch": 7.288338658146965, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 9125 + }, + { + "epoch": 7.289137380191693, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 9126 + }, + { + "epoch": 7.289936102236422, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 9127 + }, + { + "epoch": 7.2907348242811505, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 9128 + }, + { + "epoch": 7.291533546325878, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.045, + "step": 9129 + }, + { + "epoch": 7.292332268370607, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 9130 + }, + { + "epoch": 7.293130990415335, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 9131 + }, + { + "epoch": 7.293929712460064, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 9132 + }, + { + "epoch": 7.294728434504792, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 9133 + }, + { + "epoch": 7.295527156549521, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 9134 + }, + { + "epoch": 7.296325878594249, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0431, + "step": 9135 + }, + { + "epoch": 7.297124600638978, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 9136 + }, + { + "epoch": 7.297923322683706, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 9137 + }, + { + "epoch": 7.298722044728435, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 9138 + }, + { + "epoch": 7.2995207667731625, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 9139 + }, + { + "epoch": 7.300319488817891, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0337, + "step": 9140 + }, + { + "epoch": 7.30111821086262, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 9141 + }, + { + "epoch": 7.301916932907348, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 9142 + }, + { + "epoch": 7.302715654952077, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 9143 + }, + { + "epoch": 7.303514376996805, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 9144 + }, + { + "epoch": 7.304313099041534, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0423, + "step": 9145 + }, + { + "epoch": 7.305111821086262, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0459, + "step": 9146 + }, + { + "epoch": 7.305910543130991, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 9147 + }, + { + "epoch": 7.306709265175719, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 9148 + }, + { + "epoch": 7.307507987220447, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 9149 + }, + { + "epoch": 7.3083067092651754, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 9150 + }, + { + "epoch": 7.309105431309904, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 9151 + }, + { + "epoch": 7.3099041533546325, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 9152 + }, + { + "epoch": 7.310702875399361, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 9153 + }, + { + "epoch": 7.31150159744409, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 9154 + }, + { + "epoch": 7.312300319488818, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 9155 + }, + { + "epoch": 7.313099041533547, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0338, + "step": 9156 + }, + { + "epoch": 7.313897763578275, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 9157 + }, + { + "epoch": 7.314696485623003, + "grad_norm": 0.16015625, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 9158 + }, + { + "epoch": 7.315495207667731, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0393, + "step": 9159 + }, + { + "epoch": 7.31629392971246, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.04, + "step": 9160 + }, + { + "epoch": 7.317092651757188, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 9161 + }, + { + "epoch": 7.317891373801917, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 9162 + }, + { + "epoch": 7.318690095846645, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 9163 + }, + { + "epoch": 7.319488817891374, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 9164 + }, + { + "epoch": 7.3202875399361025, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 9165 + }, + { + "epoch": 7.321086261980831, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 9166 + }, + { + "epoch": 7.321884984025559, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 9167 + }, + { + "epoch": 7.322683706070287, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 9168 + }, + { + "epoch": 7.323482428115016, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 9169 + }, + { + "epoch": 7.324281150159744, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 9170 + }, + { + "epoch": 7.325079872204473, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 9171 + }, + { + "epoch": 7.325878594249201, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 9172 + }, + { + "epoch": 7.32667731629393, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 9173 + }, + { + "epoch": 7.327476038338658, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 9174 + }, + { + "epoch": 7.328274760383387, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 9175 + }, + { + "epoch": 7.329073482428115, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 9176 + }, + { + "epoch": 7.329872204472843, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0407, + "step": 9177 + }, + { + "epoch": 7.330670926517572, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 9178 + }, + { + "epoch": 7.3314696485623, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0379, + "step": 9179 + }, + { + "epoch": 7.332268370607029, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0405, + "step": 9180 + }, + { + "epoch": 7.333067092651757, + "grad_norm": 0.1572265625, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 9181 + }, + { + "epoch": 7.333865814696486, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0354, + "step": 9182 + }, + { + "epoch": 7.334664536741214, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 9183 + }, + { + "epoch": 7.335463258785943, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 9184 + }, + { + "epoch": 7.336261980830671, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 9185 + }, + { + "epoch": 7.3370607028754, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 9186 + }, + { + "epoch": 7.337859424920127, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 9187 + }, + { + "epoch": 7.338658146964856, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 9188 + }, + { + "epoch": 7.3394568690095845, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 9189 + }, + { + "epoch": 7.340255591054313, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 9190 + }, + { + "epoch": 7.3410543130990416, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 9191 + }, + { + "epoch": 7.34185303514377, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 9192 + }, + { + "epoch": 7.342651757188499, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 9193 + }, + { + "epoch": 7.343450479233227, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 9194 + }, + { + "epoch": 7.344249201277956, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 9195 + }, + { + "epoch": 7.345047923322683, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 9196 + }, + { + "epoch": 7.345846645367412, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 9197 + }, + { + "epoch": 7.34664536741214, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.042, + "step": 9198 + }, + { + "epoch": 7.347444089456869, + "grad_norm": 0.162109375, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 9199 + }, + { + "epoch": 7.348242811501597, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0147, + "step": 9200 + }, + { + "epoch": 7.349041533546326, + "grad_norm": 0.1875, + "learning_rate": 0.0005, + "loss": 1.036, + "step": 9201 + }, + { + "epoch": 7.3498402555910545, + "grad_norm": 0.16796875, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 9202 + }, + { + "epoch": 7.350638977635783, + "grad_norm": 0.19921875, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 9203 + }, + { + "epoch": 7.3514376996805115, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.0395, + "step": 9204 + }, + { + "epoch": 7.352236421725239, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0412, + "step": 9205 + }, + { + "epoch": 7.353035143769968, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0397, + "step": 9206 + }, + { + "epoch": 7.353833865814696, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 9207 + }, + { + "epoch": 7.354632587859425, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0337, + "step": 9208 + }, + { + "epoch": 7.355431309904153, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 9209 + }, + { + "epoch": 7.356230031948882, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 9210 + }, + { + "epoch": 7.35702875399361, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0375, + "step": 9211 + }, + { + "epoch": 7.357827476038339, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 9212 + }, + { + "epoch": 7.358626198083067, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 9213 + }, + { + "epoch": 7.359424920127796, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 9214 + }, + { + "epoch": 7.360223642172524, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 9215 + }, + { + "epoch": 7.361022364217252, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0412, + "step": 9216 + }, + { + "epoch": 7.361821086261981, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 9217 + }, + { + "epoch": 7.362619808306709, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0375, + "step": 9218 + }, + { + "epoch": 7.363418530351438, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0406, + "step": 9219 + }, + { + "epoch": 7.364217252396166, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0435, + "step": 9220 + }, + { + "epoch": 7.365015974440895, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0348, + "step": 9221 + }, + { + "epoch": 7.365814696485623, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 9222 + }, + { + "epoch": 7.366613418530352, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 9223 + }, + { + "epoch": 7.36741214057508, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 9224 + }, + { + "epoch": 7.368210862619808, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 9225 + }, + { + "epoch": 7.3690095846645365, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 9226 + }, + { + "epoch": 7.369808306709265, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 9227 + }, + { + "epoch": 7.3706070287539935, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 9228 + }, + { + "epoch": 7.371405750798722, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 9229 + }, + { + "epoch": 7.372204472843451, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 9230 + }, + { + "epoch": 7.373003194888179, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0337, + "step": 9231 + }, + { + "epoch": 7.373801916932908, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 9232 + }, + { + "epoch": 7.374600638977636, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0379, + "step": 9233 + }, + { + "epoch": 7.375399361022364, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 9234 + }, + { + "epoch": 7.376198083067092, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.042, + "step": 9235 + }, + { + "epoch": 7.376996805111821, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 9236 + }, + { + "epoch": 7.377795527156549, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0378, + "step": 9237 + }, + { + "epoch": 7.378594249201278, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 9238 + }, + { + "epoch": 7.3793929712460065, + "grad_norm": 0.2041015625, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 9239 + }, + { + "epoch": 7.380191693290735, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0353, + "step": 9240 + }, + { + "epoch": 7.3809904153354635, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 9241 + }, + { + "epoch": 7.381789137380192, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 9242 + }, + { + "epoch": 7.38258785942492, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 9243 + }, + { + "epoch": 7.383386581469648, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 9244 + }, + { + "epoch": 7.384185303514377, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 9245 + }, + { + "epoch": 7.384984025559105, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0389, + "step": 9246 + }, + { + "epoch": 7.385782747603834, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 9247 + }, + { + "epoch": 7.386581469648562, + "grad_norm": 0.212890625, + "learning_rate": 0.0005, + "loss": 1.0404, + "step": 9248 + }, + { + "epoch": 7.387380191693291, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0385, + "step": 9249 + }, + { + "epoch": 7.388178913738019, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 9250 + }, + { + "epoch": 7.388977635782748, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 9251 + }, + { + "epoch": 7.389776357827476, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 9252 + }, + { + "epoch": 7.390575079872204, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 9253 + }, + { + "epoch": 7.391373801916933, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0394, + "step": 9254 + }, + { + "epoch": 7.392172523961661, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 9255 + }, + { + "epoch": 7.39297124600639, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 9256 + }, + { + "epoch": 7.393769968051118, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0405, + "step": 9257 + }, + { + "epoch": 7.394568690095847, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 9258 + }, + { + "epoch": 7.395367412140575, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 9259 + }, + { + "epoch": 7.396166134185304, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0373, + "step": 9260 + }, + { + "epoch": 7.396964856230032, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 9261 + }, + { + "epoch": 7.397763578274761, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 9262 + }, + { + "epoch": 7.3985623003194885, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 9263 + }, + { + "epoch": 7.399361022364217, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 9264 + }, + { + "epoch": 7.4001597444089455, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 9265 + }, + { + "epoch": 7.400958466453674, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 9266 + }, + { + "epoch": 7.401757188498403, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0375, + "step": 9267 + }, + { + "epoch": 7.402555910543131, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 9268 + }, + { + "epoch": 7.40335463258786, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 9269 + }, + { + "epoch": 7.404153354632588, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 9270 + }, + { + "epoch": 7.404952076677317, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 9271 + }, + { + "epoch": 7.405750798722044, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 9272 + }, + { + "epoch": 7.406549520766773, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 9273 + }, + { + "epoch": 7.407348242811501, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 9274 + }, + { + "epoch": 7.40814696485623, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 9275 + }, + { + "epoch": 7.4089456869009584, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 9276 + }, + { + "epoch": 7.409744408945687, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0433, + "step": 9277 + }, + { + "epoch": 7.4105431309904155, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 9278 + }, + { + "epoch": 7.411341853035144, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 9279 + }, + { + "epoch": 7.412140575079873, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 9280 + }, + { + "epoch": 7.4129392971246, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 9281 + }, + { + "epoch": 7.413738019169329, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0354, + "step": 9282 + }, + { + "epoch": 7.414536741214057, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 9283 + }, + { + "epoch": 7.415335463258786, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0372, + "step": 9284 + }, + { + "epoch": 7.416134185303514, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 9285 + }, + { + "epoch": 7.416932907348243, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 9286 + }, + { + "epoch": 7.417731629392971, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 9287 + }, + { + "epoch": 7.4185303514377, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.036, + "step": 9288 + }, + { + "epoch": 7.419329073482428, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 9289 + }, + { + "epoch": 7.420127795527157, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 9290 + }, + { + "epoch": 7.420926517571885, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 9291 + }, + { + "epoch": 7.421725239616613, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 9292 + }, + { + "epoch": 7.422523961661342, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.037, + "step": 9293 + }, + { + "epoch": 7.42332268370607, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 9294 + }, + { + "epoch": 7.424121405750799, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 9295 + }, + { + "epoch": 7.424920127795527, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 9296 + }, + { + "epoch": 7.425718849840256, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 9297 + }, + { + "epoch": 7.426517571884984, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0362, + "step": 9298 + }, + { + "epoch": 7.427316293929713, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 9299 + }, + { + "epoch": 7.428115015974441, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0377, + "step": 9300 + }, + { + "epoch": 7.428913738019169, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 9301 + }, + { + "epoch": 7.4297124600638975, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0426, + "step": 9302 + }, + { + "epoch": 7.430511182108626, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 9303 + }, + { + "epoch": 7.431309904153355, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0378, + "step": 9304 + }, + { + "epoch": 7.432108626198083, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 9305 + }, + { + "epoch": 7.432907348242812, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 9306 + }, + { + "epoch": 7.43370607028754, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 9307 + }, + { + "epoch": 7.434504792332269, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 9308 + }, + { + "epoch": 7.435303514376997, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 9309 + }, + { + "epoch": 7.436102236421725, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.018, + "step": 9310 + }, + { + "epoch": 7.436900958466453, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 9311 + }, + { + "epoch": 7.437699680511182, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 9312 + }, + { + "epoch": 7.43849840255591, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 9313 + }, + { + "epoch": 7.439297124600639, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.04, + "step": 9314 + }, + { + "epoch": 7.4400958466453675, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 9315 + }, + { + "epoch": 7.440894568690096, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 9316 + }, + { + "epoch": 7.4416932907348246, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 9317 + }, + { + "epoch": 7.442492012779553, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 9318 + }, + { + "epoch": 7.443290734824281, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.037, + "step": 9319 + }, + { + "epoch": 7.444089456869009, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 9320 + }, + { + "epoch": 7.444888178913738, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0379, + "step": 9321 + }, + { + "epoch": 7.445686900958466, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 9322 + }, + { + "epoch": 7.446485623003195, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 9323 + }, + { + "epoch": 7.447284345047923, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 9324 + }, + { + "epoch": 7.448083067092652, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 9325 + }, + { + "epoch": 7.44888178913738, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 9326 + }, + { + "epoch": 7.449680511182109, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 9327 + }, + { + "epoch": 7.4504792332268375, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 9328 + }, + { + "epoch": 7.451277955271565, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0338, + "step": 9329 + }, + { + "epoch": 7.452076677316294, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 9330 + }, + { + "epoch": 7.452875399361022, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0338, + "step": 9331 + }, + { + "epoch": 7.453674121405751, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0361, + "step": 9332 + }, + { + "epoch": 7.454472843450479, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 9333 + }, + { + "epoch": 7.455271565495208, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 9334 + }, + { + "epoch": 7.456070287539936, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 9335 + }, + { + "epoch": 7.456869009584665, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 9336 + }, + { + "epoch": 7.457667731629393, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 9337 + }, + { + "epoch": 7.458466453674122, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 9338 + }, + { + "epoch": 7.4592651757188495, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 9339 + }, + { + "epoch": 7.460063897763578, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 9340 + }, + { + "epoch": 7.460862619808307, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 9341 + }, + { + "epoch": 7.461661341853035, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 9342 + }, + { + "epoch": 7.462460063897764, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 9343 + }, + { + "epoch": 7.463258785942492, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.04, + "step": 9344 + }, + { + "epoch": 7.464057507987221, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 9345 + }, + { + "epoch": 7.464856230031949, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0348, + "step": 9346 + }, + { + "epoch": 7.465654952076678, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 9347 + }, + { + "epoch": 7.466453674121405, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 9348 + }, + { + "epoch": 7.467252396166134, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 9349 + }, + { + "epoch": 7.468051118210862, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 9350 + }, + { + "epoch": 7.468849840255591, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 9351 + }, + { + "epoch": 7.4696485623003195, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 9352 + }, + { + "epoch": 7.470447284345048, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 9353 + }, + { + "epoch": 7.4712460063897765, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0372, + "step": 9354 + }, + { + "epoch": 7.472044728434505, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 9355 + }, + { + "epoch": 7.472843450479234, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0384, + "step": 9356 + }, + { + "epoch": 7.473642172523961, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 9357 + }, + { + "epoch": 7.47444089456869, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 9358 + }, + { + "epoch": 7.475239616613418, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0379, + "step": 9359 + }, + { + "epoch": 7.476038338658147, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 9360 + }, + { + "epoch": 7.476837060702875, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 9361 + }, + { + "epoch": 7.477635782747604, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 9362 + }, + { + "epoch": 7.478434504792332, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 9363 + }, + { + "epoch": 7.479233226837061, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.0377, + "step": 9364 + }, + { + "epoch": 7.4800319488817895, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 9365 + }, + { + "epoch": 7.480830670926517, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.0376, + "step": 9366 + }, + { + "epoch": 7.481629392971246, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 9367 + }, + { + "epoch": 7.482428115015974, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 9368 + }, + { + "epoch": 7.483226837060703, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 9369 + }, + { + "epoch": 7.484025559105431, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.043, + "step": 9370 + }, + { + "epoch": 7.48482428115016, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 9371 + }, + { + "epoch": 7.485623003194888, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 9372 + }, + { + "epoch": 7.486421725239617, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 9373 + }, + { + "epoch": 7.487220447284345, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0338, + "step": 9374 + }, + { + "epoch": 7.488019169329074, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 9375 + }, + { + "epoch": 7.488817891373802, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 9376 + }, + { + "epoch": 7.48961661341853, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0377, + "step": 9377 + }, + { + "epoch": 7.4904153354632586, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 9378 + }, + { + "epoch": 7.491214057507987, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 9379 + }, + { + "epoch": 7.492012779552716, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 9380 + }, + { + "epoch": 7.492811501597444, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 9381 + }, + { + "epoch": 7.493610223642173, + "grad_norm": 0.201171875, + "learning_rate": 0.0005, + "loss": 1.0338, + "step": 9382 + }, + { + "epoch": 7.494408945686901, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 9383 + }, + { + "epoch": 7.49520766773163, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 9384 + }, + { + "epoch": 7.496006389776358, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 9385 + }, + { + "epoch": 7.496805111821086, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 9386 + }, + { + "epoch": 7.497603833865814, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 9387 + }, + { + "epoch": 7.498402555910543, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 9388 + }, + { + "epoch": 7.4992012779552715, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 9389 + }, + { + "epoch": 7.5, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 9390 + }, + { + "epoch": 7.5007987220447285, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 9391 + }, + { + "epoch": 7.501597444089457, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 9392 + }, + { + "epoch": 7.502396166134186, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 9393 + }, + { + "epoch": 7.503194888178914, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 9394 + }, + { + "epoch": 7.503993610223642, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 9395 + }, + { + "epoch": 7.50479233226837, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 9396 + }, + { + "epoch": 7.505591054313099, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 9397 + }, + { + "epoch": 7.506389776357827, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 9398 + }, + { + "epoch": 7.507188498402556, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 9399 + }, + { + "epoch": 7.507987220447284, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 9400 + }, + { + "epoch": 7.508785942492013, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.036, + "step": 9401 + }, + { + "epoch": 7.5095846645367414, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 9402 + }, + { + "epoch": 7.51038338658147, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 9403 + }, + { + "epoch": 7.511182108626198, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0379, + "step": 9404 + }, + { + "epoch": 7.511980830670926, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0337, + "step": 9405 + }, + { + "epoch": 7.512779552715655, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 9406 + }, + { + "epoch": 7.513578274760383, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0384, + "step": 9407 + }, + { + "epoch": 7.514376996805112, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0386, + "step": 9408 + }, + { + "epoch": 7.51517571884984, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 9409 + }, + { + "epoch": 7.515974440894569, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 9410 + }, + { + "epoch": 7.516773162939297, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0369, + "step": 9411 + }, + { + "epoch": 7.517571884984026, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 9412 + }, + { + "epoch": 7.518370607028754, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 9413 + }, + { + "epoch": 7.519169329073483, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 9414 + }, + { + "epoch": 7.5199680511182105, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 9415 + }, + { + "epoch": 7.520766773162939, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 9416 + }, + { + "epoch": 7.521565495207668, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 9417 + }, + { + "epoch": 7.522364217252396, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 9418 + }, + { + "epoch": 7.523162939297125, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 9419 + }, + { + "epoch": 7.523961661341853, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.034, + "step": 9420 + }, + { + "epoch": 7.524760383386582, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 9421 + }, + { + "epoch": 7.52555910543131, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 9422 + }, + { + "epoch": 7.526357827476039, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0337, + "step": 9423 + }, + { + "epoch": 7.527156549520766, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0424, + "step": 9424 + }, + { + "epoch": 7.527955271565495, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 9425 + }, + { + "epoch": 7.5287539936102235, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 9426 + }, + { + "epoch": 7.529552715654952, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0451, + "step": 9427 + }, + { + "epoch": 7.5303514376996805, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0399, + "step": 9428 + }, + { + "epoch": 7.531150159744409, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 9429 + }, + { + "epoch": 7.531948881789138, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 9430 + }, + { + "epoch": 7.532747603833866, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 9431 + }, + { + "epoch": 7.533546325878595, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 9432 + }, + { + "epoch": 7.534345047923322, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 9433 + }, + { + "epoch": 7.535143769968051, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 9434 + }, + { + "epoch": 7.535942492012779, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 9435 + }, + { + "epoch": 7.536741214057508, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 9436 + }, + { + "epoch": 7.537539936102236, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 9437 + }, + { + "epoch": 7.538338658146965, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 9438 + }, + { + "epoch": 7.539137380191693, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0402, + "step": 9439 + }, + { + "epoch": 7.539936102236422, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 9440 + }, + { + "epoch": 7.5407348242811505, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 9441 + }, + { + "epoch": 7.541533546325878, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0404, + "step": 9442 + }, + { + "epoch": 7.542332268370607, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 9443 + }, + { + "epoch": 7.543130990415335, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 9444 + }, + { + "epoch": 7.543929712460064, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 9445 + }, + { + "epoch": 7.544728434504792, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 9446 + }, + { + "epoch": 7.545527156549521, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0362, + "step": 9447 + }, + { + "epoch": 7.546325878594249, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 9448 + }, + { + "epoch": 7.547124600638978, + "grad_norm": 0.1787109375, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 9449 + }, + { + "epoch": 7.547923322683706, + "grad_norm": 0.1669921875, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 9450 + }, + { + "epoch": 7.548722044728435, + "grad_norm": 0.17578125, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 9451 + }, + { + "epoch": 7.549520766773163, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 9452 + }, + { + "epoch": 7.550319488817891, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 9453 + }, + { + "epoch": 7.55111821086262, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 9454 + }, + { + "epoch": 7.551916932907348, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 9455 + }, + { + "epoch": 7.552715654952077, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0388, + "step": 9456 + }, + { + "epoch": 7.553514376996805, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 9457 + }, + { + "epoch": 7.554313099041534, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 9458 + }, + { + "epoch": 7.555111821086262, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 9459 + }, + { + "epoch": 7.555910543130991, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0393, + "step": 9460 + }, + { + "epoch": 7.556709265175719, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 9461 + }, + { + "epoch": 7.557507987220447, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 9462 + }, + { + "epoch": 7.5583067092651754, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 9463 + }, + { + "epoch": 7.559105431309904, + "grad_norm": 0.171875, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 9464 + }, + { + "epoch": 7.5599041533546325, + "grad_norm": 0.2265625, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 9465 + }, + { + "epoch": 7.560702875399361, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0386, + "step": 9466 + }, + { + "epoch": 7.56150159744409, + "grad_norm": 0.1982421875, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 9467 + }, + { + "epoch": 7.562300319488818, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.037, + "step": 9468 + }, + { + "epoch": 7.563099041533547, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.0376, + "step": 9469 + }, + { + "epoch": 7.563897763578275, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0388, + "step": 9470 + }, + { + "epoch": 7.564696485623003, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0362, + "step": 9471 + }, + { + "epoch": 7.565495207667731, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 9472 + }, + { + "epoch": 7.56629392971246, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 9473 + }, + { + "epoch": 7.567092651757188, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0427, + "step": 9474 + }, + { + "epoch": 7.567891373801917, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 9475 + }, + { + "epoch": 7.568690095846645, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 9476 + }, + { + "epoch": 7.569488817891374, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 9477 + }, + { + "epoch": 7.5702875399361025, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 9478 + }, + { + "epoch": 7.571086261980831, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 9479 + }, + { + "epoch": 7.571884984025559, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 9480 + }, + { + "epoch": 7.572683706070287, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 9481 + }, + { + "epoch": 7.573482428115016, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 9482 + }, + { + "epoch": 7.574281150159744, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 9483 + }, + { + "epoch": 7.575079872204473, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 9484 + }, + { + "epoch": 7.575878594249201, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0403, + "step": 9485 + }, + { + "epoch": 7.57667731629393, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 9486 + }, + { + "epoch": 7.577476038338658, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 9487 + }, + { + "epoch": 7.578274760383387, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.038, + "step": 9488 + }, + { + "epoch": 7.5790734824281145, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 9489 + }, + { + "epoch": 7.579872204472844, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0427, + "step": 9490 + }, + { + "epoch": 7.580670926517572, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0384, + "step": 9491 + }, + { + "epoch": 7.5814696485623, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0192, + "step": 9492 + }, + { + "epoch": 7.582268370607029, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 9493 + }, + { + "epoch": 7.583067092651757, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 9494 + }, + { + "epoch": 7.583865814696486, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0388, + "step": 9495 + }, + { + "epoch": 7.584664536741214, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 9496 + }, + { + "epoch": 7.585463258785943, + "grad_norm": 0.1689453125, + "learning_rate": 0.0005, + "loss": 1.0406, + "step": 9497 + }, + { + "epoch": 7.586261980830671, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 9498 + }, + { + "epoch": 7.5870607028754, + "grad_norm": 0.1767578125, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 9499 + }, + { + "epoch": 7.587859424920127, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 9500 + }, + { + "epoch": 7.588658146964856, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 9501 + }, + { + "epoch": 7.5894568690095845, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 9502 + }, + { + "epoch": 7.590255591054313, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 9503 + }, + { + "epoch": 7.5910543130990416, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 9504 + }, + { + "epoch": 7.59185303514377, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 9505 + }, + { + "epoch": 7.592651757188499, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 9506 + }, + { + "epoch": 7.593450479233227, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 9507 + }, + { + "epoch": 7.594249201277956, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 9508 + }, + { + "epoch": 7.595047923322683, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 9509 + }, + { + "epoch": 7.595846645367412, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0377, + "step": 9510 + }, + { + "epoch": 7.59664536741214, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 9511 + }, + { + "epoch": 7.597444089456869, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 9512 + }, + { + "epoch": 7.598242811501597, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0375, + "step": 9513 + }, + { + "epoch": 7.599041533546326, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 9514 + }, + { + "epoch": 7.5998402555910545, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.034, + "step": 9515 + }, + { + "epoch": 7.600638977635783, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 9516 + }, + { + "epoch": 7.6014376996805115, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 9517 + }, + { + "epoch": 7.602236421725239, + "grad_norm": 0.166015625, + "learning_rate": 0.0005, + "loss": 1.0405, + "step": 9518 + }, + { + "epoch": 7.603035143769968, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 9519 + }, + { + "epoch": 7.603833865814696, + "grad_norm": 0.234375, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 9520 + }, + { + "epoch": 7.604632587859425, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 9521 + }, + { + "epoch": 7.605431309904153, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 9522 + }, + { + "epoch": 7.606230031948882, + "grad_norm": 0.17578125, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 9523 + }, + { + "epoch": 7.60702875399361, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0348, + "step": 9524 + }, + { + "epoch": 7.607827476038339, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 9525 + }, + { + "epoch": 7.608626198083067, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 9526 + }, + { + "epoch": 7.609424920127795, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 9527 + }, + { + "epoch": 7.6102236421725244, + "grad_norm": 0.1708984375, + "learning_rate": 0.0005, + "loss": 1.0376, + "step": 9528 + }, + { + "epoch": 7.611022364217252, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 9529 + }, + { + "epoch": 7.611821086261981, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0449, + "step": 9530 + }, + { + "epoch": 7.612619808306709, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 9531 + }, + { + "epoch": 7.613418530351438, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 9532 + }, + { + "epoch": 7.614217252396166, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 9533 + }, + { + "epoch": 7.615015974440895, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 9534 + }, + { + "epoch": 7.615814696485623, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 9535 + }, + { + "epoch": 7.616613418530352, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0354, + "step": 9536 + }, + { + "epoch": 7.61741214057508, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0413, + "step": 9537 + }, + { + "epoch": 7.618210862619808, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 9538 + }, + { + "epoch": 7.6190095846645365, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 9539 + }, + { + "epoch": 7.619808306709265, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 9540 + }, + { + "epoch": 7.6206070287539935, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 9541 + }, + { + "epoch": 7.621405750798722, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 9542 + }, + { + "epoch": 7.622204472843451, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.034, + "step": 9543 + }, + { + "epoch": 7.623003194888179, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 9544 + }, + { + "epoch": 7.623801916932908, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 9545 + }, + { + "epoch": 7.624600638977636, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 9546 + }, + { + "epoch": 7.625399361022364, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 9547 + }, + { + "epoch": 7.626198083067092, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0401, + "step": 9548 + }, + { + "epoch": 7.626996805111821, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0373, + "step": 9549 + }, + { + "epoch": 7.627795527156549, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 9550 + }, + { + "epoch": 7.628594249201278, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 9551 + }, + { + "epoch": 7.6293929712460065, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 9552 + }, + { + "epoch": 7.630191693290735, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 9553 + }, + { + "epoch": 7.6309904153354635, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0353, + "step": 9554 + }, + { + "epoch": 7.631789137380192, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.039, + "step": 9555 + }, + { + "epoch": 7.63258785942492, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 9556 + }, + { + "epoch": 7.633386581469648, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 9557 + }, + { + "epoch": 7.634185303514377, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 9558 + }, + { + "epoch": 7.634984025559105, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 9559 + }, + { + "epoch": 7.635782747603834, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 9560 + }, + { + "epoch": 7.636581469648562, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0412, + "step": 9561 + }, + { + "epoch": 7.637380191693291, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 9562 + }, + { + "epoch": 7.638178913738019, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 9563 + }, + { + "epoch": 7.638977635782748, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0471, + "step": 9564 + }, + { + "epoch": 7.6397763578274756, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 9565 + }, + { + "epoch": 7.640575079872205, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 9566 + }, + { + "epoch": 7.641373801916933, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 9567 + }, + { + "epoch": 7.642172523961661, + "grad_norm": 0.1796875, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 9568 + }, + { + "epoch": 7.64297124600639, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 9569 + }, + { + "epoch": 7.643769968051118, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 9570 + }, + { + "epoch": 7.644568690095847, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 9571 + }, + { + "epoch": 7.645367412140575, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 9572 + }, + { + "epoch": 7.646166134185304, + "grad_norm": 0.1533203125, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 9573 + }, + { + "epoch": 7.646964856230032, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 9574 + }, + { + "epoch": 7.647763578274761, + "grad_norm": 0.181640625, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 9575 + }, + { + "epoch": 7.6485623003194885, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 9576 + }, + { + "epoch": 7.649361022364217, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 9577 + }, + { + "epoch": 7.6501597444089455, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 9578 + }, + { + "epoch": 7.650958466453674, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 9579 + }, + { + "epoch": 7.651757188498403, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 9580 + }, + { + "epoch": 7.652555910543131, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 9581 + }, + { + "epoch": 7.65335463258786, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 9582 + }, + { + "epoch": 7.654153354632588, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0417, + "step": 9583 + }, + { + "epoch": 7.654952076677317, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 9584 + }, + { + "epoch": 7.655750798722044, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 9585 + }, + { + "epoch": 7.656549520766773, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 9586 + }, + { + "epoch": 7.657348242811501, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 9587 + }, + { + "epoch": 7.65814696485623, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 9588 + }, + { + "epoch": 7.6589456869009584, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 9589 + }, + { + "epoch": 7.659744408945687, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.042, + "step": 9590 + }, + { + "epoch": 7.6605431309904155, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 9591 + }, + { + "epoch": 7.661341853035144, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 9592 + }, + { + "epoch": 7.662140575079873, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 9593 + }, + { + "epoch": 7.6629392971246, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0403, + "step": 9594 + }, + { + "epoch": 7.663738019169329, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 9595 + }, + { + "epoch": 7.664536741214057, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 9596 + }, + { + "epoch": 7.665335463258786, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 9597 + }, + { + "epoch": 7.666134185303514, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 9598 + }, + { + "epoch": 7.666932907348243, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 9599 + }, + { + "epoch": 7.667731629392971, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 9600 + }, + { + "epoch": 7.6685303514377, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 9601 + }, + { + "epoch": 7.669329073482428, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 9602 + }, + { + "epoch": 7.670127795527156, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 9603 + }, + { + "epoch": 7.6709265175718855, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 9604 + }, + { + "epoch": 7.671725239616613, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 9605 + }, + { + "epoch": 7.672523961661342, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 9606 + }, + { + "epoch": 7.67332268370607, + "grad_norm": 0.1728515625, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 9607 + }, + { + "epoch": 7.674121405750799, + "grad_norm": 0.1748046875, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 9608 + }, + { + "epoch": 7.674920127795527, + "grad_norm": 0.314453125, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 9609 + }, + { + "epoch": 7.675718849840256, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 9610 + }, + { + "epoch": 7.676517571884984, + "grad_norm": 0.26953125, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 9611 + }, + { + "epoch": 7.677316293929713, + "grad_norm": 0.1982421875, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 9612 + }, + { + "epoch": 7.678115015974441, + "grad_norm": 0.4140625, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 9613 + }, + { + "epoch": 7.678913738019169, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 9614 + }, + { + "epoch": 7.6797124600638975, + "grad_norm": 0.25390625, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 9615 + }, + { + "epoch": 7.680511182108626, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 9616 + }, + { + "epoch": 7.681309904153355, + "grad_norm": 0.2890625, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 9617 + }, + { + "epoch": 7.682108626198083, + "grad_norm": 0.16015625, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 9618 + }, + { + "epoch": 7.682907348242812, + "grad_norm": 0.330078125, + "learning_rate": 0.0005, + "loss": 1.0378, + "step": 9619 + }, + { + "epoch": 7.68370607028754, + "grad_norm": 0.20703125, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 9620 + }, + { + "epoch": 7.684504792332269, + "grad_norm": 0.51171875, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 9621 + }, + { + "epoch": 7.685303514376997, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 9622 + }, + { + "epoch": 7.686102236421725, + "grad_norm": 0.404296875, + "learning_rate": 0.0005, + "loss": 1.0338, + "step": 9623 + }, + { + "epoch": 7.686900958466453, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 9624 + }, + { + "epoch": 7.687699680511182, + "grad_norm": 0.39453125, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 9625 + }, + { + "epoch": 7.68849840255591, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 9626 + }, + { + "epoch": 7.689297124600639, + "grad_norm": 0.3671875, + "learning_rate": 0.0005, + "loss": 1.0396, + "step": 9627 + }, + { + "epoch": 7.6900958466453675, + "grad_norm": 0.212890625, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 9628 + }, + { + "epoch": 7.690894568690096, + "grad_norm": 0.515625, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 9629 + }, + { + "epoch": 7.6916932907348246, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0386, + "step": 9630 + }, + { + "epoch": 7.692492012779553, + "grad_norm": 0.51953125, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 9631 + }, + { + "epoch": 7.693290734824281, + "grad_norm": 0.1669921875, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 9632 + }, + { + "epoch": 7.694089456869009, + "grad_norm": 0.7578125, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 9633 + }, + { + "epoch": 7.694888178913738, + "grad_norm": 0.357421875, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 9634 + }, + { + "epoch": 7.695686900958466, + "grad_norm": 0.376953125, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 9635 + }, + { + "epoch": 7.696485623003195, + "grad_norm": 0.353515625, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 9636 + }, + { + "epoch": 7.697284345047923, + "grad_norm": 0.267578125, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 9637 + }, + { + "epoch": 7.698083067092652, + "grad_norm": 0.357421875, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 9638 + }, + { + "epoch": 7.69888178913738, + "grad_norm": 0.294921875, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 9639 + }, + { + "epoch": 7.699680511182109, + "grad_norm": 0.494140625, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 9640 + }, + { + "epoch": 7.700479233226837, + "grad_norm": 0.2080078125, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 9641 + }, + { + "epoch": 7.701277955271565, + "grad_norm": 0.50390625, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 9642 + }, + { + "epoch": 7.702076677316294, + "grad_norm": 0.2060546875, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 9643 + }, + { + "epoch": 7.702875399361022, + "grad_norm": 0.59765625, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 9644 + }, + { + "epoch": 7.703674121405751, + "grad_norm": 0.1640625, + "learning_rate": 0.0005, + "loss": 1.0348, + "step": 9645 + }, + { + "epoch": 7.704472843450479, + "grad_norm": 0.357421875, + "learning_rate": 0.0005, + "loss": 1.0434, + "step": 9646 + }, + { + "epoch": 7.705271565495208, + "grad_norm": 0.224609375, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 9647 + }, + { + "epoch": 7.706070287539936, + "grad_norm": 0.376953125, + "learning_rate": 0.0005, + "loss": 1.0416, + "step": 9648 + }, + { + "epoch": 7.706869009584665, + "grad_norm": 0.1962890625, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 9649 + }, + { + "epoch": 7.707667731629393, + "grad_norm": 0.2578125, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 9650 + }, + { + "epoch": 7.708466453674122, + "grad_norm": 0.26171875, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 9651 + }, + { + "epoch": 7.7092651757188495, + "grad_norm": 0.322265625, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 9652 + }, + { + "epoch": 7.710063897763578, + "grad_norm": 0.1708984375, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 9653 + }, + { + "epoch": 7.710862619808307, + "grad_norm": 0.28125, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 9654 + }, + { + "epoch": 7.711661341853035, + "grad_norm": 0.1884765625, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 9655 + }, + { + "epoch": 7.712460063897764, + "grad_norm": 0.267578125, + "learning_rate": 0.0005, + "loss": 1.043, + "step": 9656 + }, + { + "epoch": 7.713258785942492, + "grad_norm": 0.234375, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 9657 + }, + { + "epoch": 7.714057507987221, + "grad_norm": 0.3046875, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 9658 + }, + { + "epoch": 7.714856230031949, + "grad_norm": 0.2197265625, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 9659 + }, + { + "epoch": 7.715654952076678, + "grad_norm": 0.322265625, + "learning_rate": 0.0005, + "loss": 1.034, + "step": 9660 + }, + { + "epoch": 7.716453674121405, + "grad_norm": 0.201171875, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 9661 + }, + { + "epoch": 7.717252396166134, + "grad_norm": 0.25390625, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 9662 + }, + { + "epoch": 7.718051118210862, + "grad_norm": 0.2412109375, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 9663 + }, + { + "epoch": 7.718849840255591, + "grad_norm": 0.2138671875, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 9664 + }, + { + "epoch": 7.7196485623003195, + "grad_norm": 0.236328125, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 9665 + }, + { + "epoch": 7.720447284345048, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0384, + "step": 9666 + }, + { + "epoch": 7.7212460063897765, + "grad_norm": 0.189453125, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 9667 + }, + { + "epoch": 7.722044728434505, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0379, + "step": 9668 + }, + { + "epoch": 7.722843450479234, + "grad_norm": 0.1875, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 9669 + }, + { + "epoch": 7.723642172523961, + "grad_norm": 0.1943359375, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 9670 + }, + { + "epoch": 7.72444089456869, + "grad_norm": 0.1875, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 9671 + }, + { + "epoch": 7.725239616613418, + "grad_norm": 0.16796875, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 9672 + }, + { + "epoch": 7.726038338658147, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 9673 + }, + { + "epoch": 7.726837060702875, + "grad_norm": 0.1572265625, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 9674 + }, + { + "epoch": 7.727635782747604, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 9675 + }, + { + "epoch": 7.728434504792332, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0401, + "step": 9676 + }, + { + "epoch": 7.729233226837061, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 9677 + }, + { + "epoch": 7.7300319488817895, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 9678 + }, + { + "epoch": 7.730830670926517, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 9679 + }, + { + "epoch": 7.731629392971246, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 9680 + }, + { + "epoch": 7.732428115015974, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 9681 + }, + { + "epoch": 7.733226837060703, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 9682 + }, + { + "epoch": 7.734025559105431, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0411, + "step": 9683 + }, + { + "epoch": 7.73482428115016, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 9684 + }, + { + "epoch": 7.735623003194888, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 9685 + }, + { + "epoch": 7.736421725239617, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.039, + "step": 9686 + }, + { + "epoch": 7.737220447284345, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0441, + "step": 9687 + }, + { + "epoch": 7.738019169329074, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 9688 + }, + { + "epoch": 7.738817891373802, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 9689 + }, + { + "epoch": 7.73961661341853, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 9690 + }, + { + "epoch": 7.7404153354632586, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 9691 + }, + { + "epoch": 7.741214057507987, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 9692 + }, + { + "epoch": 7.742012779552716, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 9693 + }, + { + "epoch": 7.742811501597444, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 9694 + }, + { + "epoch": 7.743610223642173, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 9695 + }, + { + "epoch": 7.744408945686901, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 9696 + }, + { + "epoch": 7.74520766773163, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 9697 + }, + { + "epoch": 7.746006389776358, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 9698 + }, + { + "epoch": 7.746805111821086, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0362, + "step": 9699 + }, + { + "epoch": 7.747603833865814, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 9700 + }, + { + "epoch": 7.748402555910543, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0353, + "step": 9701 + }, + { + "epoch": 7.7492012779552715, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0378, + "step": 9702 + }, + { + "epoch": 7.75, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 9703 + }, + { + "epoch": 7.7507987220447285, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0465, + "step": 9704 + }, + { + "epoch": 7.751597444089457, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 9705 + }, + { + "epoch": 7.752396166134186, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 9706 + }, + { + "epoch": 7.753194888178914, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 9707 + }, + { + "epoch": 7.753993610223642, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 9708 + }, + { + "epoch": 7.75479233226837, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 9709 + }, + { + "epoch": 7.755591054313099, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 9710 + }, + { + "epoch": 7.756389776357827, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 9711 + }, + { + "epoch": 7.757188498402556, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 9712 + }, + { + "epoch": 7.757987220447284, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 9713 + }, + { + "epoch": 7.758785942492013, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 9714 + }, + { + "epoch": 7.7595846645367414, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 9715 + }, + { + "epoch": 7.76038338658147, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.034, + "step": 9716 + }, + { + "epoch": 7.761182108626198, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 9717 + }, + { + "epoch": 7.761980830670926, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 9718 + }, + { + "epoch": 7.762779552715655, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0375, + "step": 9719 + }, + { + "epoch": 7.763578274760383, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 9720 + }, + { + "epoch": 7.764376996805112, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0379, + "step": 9721 + }, + { + "epoch": 7.76517571884984, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 9722 + }, + { + "epoch": 7.765974440894569, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0354, + "step": 9723 + }, + { + "epoch": 7.766773162939297, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0411, + "step": 9724 + }, + { + "epoch": 7.767571884984026, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0376, + "step": 9725 + }, + { + "epoch": 7.768370607028754, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 9726 + }, + { + "epoch": 7.769169329073483, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0405, + "step": 9727 + }, + { + "epoch": 7.7699680511182105, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 9728 + }, + { + "epoch": 7.770766773162939, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0411, + "step": 9729 + }, + { + "epoch": 7.771565495207668, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 9730 + }, + { + "epoch": 7.772364217252396, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0396, + "step": 9731 + }, + { + "epoch": 7.773162939297125, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 9732 + }, + { + "epoch": 7.773961661341853, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0424, + "step": 9733 + }, + { + "epoch": 7.774760383386582, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 9734 + }, + { + "epoch": 7.77555910543131, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 9735 + }, + { + "epoch": 7.776357827476039, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 9736 + }, + { + "epoch": 7.777156549520766, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.038, + "step": 9737 + }, + { + "epoch": 7.777955271565495, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 9738 + }, + { + "epoch": 7.7787539936102235, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 9739 + }, + { + "epoch": 7.779552715654952, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0361, + "step": 9740 + }, + { + "epoch": 7.7803514376996805, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 9741 + }, + { + "epoch": 7.781150159744409, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 9742 + }, + { + "epoch": 7.781948881789138, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.034, + "step": 9743 + }, + { + "epoch": 7.782747603833866, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 9744 + }, + { + "epoch": 7.783546325878595, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 9745 + }, + { + "epoch": 7.784345047923322, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 9746 + }, + { + "epoch": 7.785143769968051, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 9747 + }, + { + "epoch": 7.785942492012779, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 9748 + }, + { + "epoch": 7.786741214057508, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 9749 + }, + { + "epoch": 7.787539936102236, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 9750 + }, + { + "epoch": 7.788338658146965, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 9751 + }, + { + "epoch": 7.789137380191693, + "grad_norm": 0.1708984375, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 9752 + }, + { + "epoch": 7.789936102236422, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 9753 + }, + { + "epoch": 7.7907348242811505, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 9754 + }, + { + "epoch": 7.791533546325878, + "grad_norm": 0.1611328125, + "learning_rate": 0.0005, + "loss": 1.0361, + "step": 9755 + }, + { + "epoch": 7.792332268370607, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 9756 + }, + { + "epoch": 7.793130990415335, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 9757 + }, + { + "epoch": 7.793929712460064, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 9758 + }, + { + "epoch": 7.794728434504792, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 9759 + }, + { + "epoch": 7.795527156549521, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 9760 + }, + { + "epoch": 7.796325878594249, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 9761 + }, + { + "epoch": 7.797124600638978, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 9762 + }, + { + "epoch": 7.797923322683706, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 9763 + }, + { + "epoch": 7.798722044728435, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.037, + "step": 9764 + }, + { + "epoch": 7.799520766773163, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0376, + "step": 9765 + }, + { + "epoch": 7.800319488817891, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 9766 + }, + { + "epoch": 7.80111821086262, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 9767 + }, + { + "epoch": 7.801916932907348, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 9768 + }, + { + "epoch": 7.802715654952077, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0361, + "step": 9769 + }, + { + "epoch": 7.803514376996805, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 9770 + }, + { + "epoch": 7.804313099041534, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 9771 + }, + { + "epoch": 7.805111821086262, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 9772 + }, + { + "epoch": 7.805910543130991, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 9773 + }, + { + "epoch": 7.806709265175719, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 9774 + }, + { + "epoch": 7.807507987220447, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 9775 + }, + { + "epoch": 7.8083067092651754, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 9776 + }, + { + "epoch": 7.809105431309904, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0177, + "step": 9777 + }, + { + "epoch": 7.8099041533546325, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 9778 + }, + { + "epoch": 7.810702875399361, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 9779 + }, + { + "epoch": 7.81150159744409, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 9780 + }, + { + "epoch": 7.812300319488818, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0378, + "step": 9781 + }, + { + "epoch": 7.813099041533547, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 9782 + }, + { + "epoch": 7.813897763578275, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 9783 + }, + { + "epoch": 7.814696485623003, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 9784 + }, + { + "epoch": 7.815495207667731, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 9785 + }, + { + "epoch": 7.81629392971246, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 9786 + }, + { + "epoch": 7.817092651757188, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 9787 + }, + { + "epoch": 7.817891373801917, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 9788 + }, + { + "epoch": 7.818690095846645, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 9789 + }, + { + "epoch": 7.819488817891374, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0171, + "step": 9790 + }, + { + "epoch": 7.8202875399361025, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 9791 + }, + { + "epoch": 7.821086261980831, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 9792 + }, + { + "epoch": 7.821884984025559, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0402, + "step": 9793 + }, + { + "epoch": 7.822683706070287, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 9794 + }, + { + "epoch": 7.823482428115016, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0361, + "step": 9795 + }, + { + "epoch": 7.824281150159744, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 9796 + }, + { + "epoch": 7.825079872204473, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 9797 + }, + { + "epoch": 7.825878594249201, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 9798 + }, + { + "epoch": 7.82667731629393, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0408, + "step": 9799 + }, + { + "epoch": 7.827476038338658, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 9800 + }, + { + "epoch": 7.828274760383387, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 9801 + }, + { + "epoch": 7.8290734824281145, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 9802 + }, + { + "epoch": 7.829872204472844, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0354, + "step": 9803 + }, + { + "epoch": 7.830670926517572, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0353, + "step": 9804 + }, + { + "epoch": 7.8314696485623, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 9805 + }, + { + "epoch": 7.832268370607029, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 9806 + }, + { + "epoch": 7.833067092651757, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0348, + "step": 9807 + }, + { + "epoch": 7.833865814696486, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0411, + "step": 9808 + }, + { + "epoch": 7.834664536741214, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 9809 + }, + { + "epoch": 7.835463258785943, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 9810 + }, + { + "epoch": 7.836261980830671, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 9811 + }, + { + "epoch": 7.8370607028754, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 9812 + }, + { + "epoch": 7.837859424920127, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 9813 + }, + { + "epoch": 7.838658146964856, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 9814 + }, + { + "epoch": 7.8394568690095845, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 9815 + }, + { + "epoch": 7.840255591054313, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 9816 + }, + { + "epoch": 7.8410543130990416, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 9817 + }, + { + "epoch": 7.84185303514377, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 9818 + }, + { + "epoch": 7.842651757188499, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0397, + "step": 9819 + }, + { + "epoch": 7.843450479233227, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 9820 + }, + { + "epoch": 7.844249201277956, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 9821 + }, + { + "epoch": 7.845047923322683, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 9822 + }, + { + "epoch": 7.845846645367412, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 9823 + }, + { + "epoch": 7.84664536741214, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 9824 + }, + { + "epoch": 7.847444089456869, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 9825 + }, + { + "epoch": 7.848242811501597, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 9826 + }, + { + "epoch": 7.849041533546326, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 9827 + }, + { + "epoch": 7.8498402555910545, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 9828 + }, + { + "epoch": 7.850638977635783, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 9829 + }, + { + "epoch": 7.8514376996805115, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 9830 + }, + { + "epoch": 7.852236421725239, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 9831 + }, + { + "epoch": 7.853035143769968, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 9832 + }, + { + "epoch": 7.853833865814696, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 9833 + }, + { + "epoch": 7.854632587859425, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 9834 + }, + { + "epoch": 7.855431309904153, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 9835 + }, + { + "epoch": 7.856230031948882, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0362, + "step": 9836 + }, + { + "epoch": 7.85702875399361, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 9837 + }, + { + "epoch": 7.857827476038339, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 9838 + }, + { + "epoch": 7.858626198083067, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 9839 + }, + { + "epoch": 7.859424920127795, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 9840 + }, + { + "epoch": 7.8602236421725244, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 9841 + }, + { + "epoch": 7.861022364217252, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 9842 + }, + { + "epoch": 7.861821086261981, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 9843 + }, + { + "epoch": 7.862619808306709, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 9844 + }, + { + "epoch": 7.863418530351438, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0354, + "step": 9845 + }, + { + "epoch": 7.864217252396166, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 9846 + }, + { + "epoch": 7.865015974440895, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 9847 + }, + { + "epoch": 7.865814696485623, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.034, + "step": 9848 + }, + { + "epoch": 7.866613418530352, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 9849 + }, + { + "epoch": 7.86741214057508, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 9850 + }, + { + "epoch": 7.868210862619808, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 9851 + }, + { + "epoch": 7.8690095846645365, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 9852 + }, + { + "epoch": 7.869808306709265, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0378, + "step": 9853 + }, + { + "epoch": 7.8706070287539935, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 9854 + }, + { + "epoch": 7.871405750798722, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 9855 + }, + { + "epoch": 7.872204472843451, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.038, + "step": 9856 + }, + { + "epoch": 7.873003194888179, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 9857 + }, + { + "epoch": 7.873801916932908, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 9858 + }, + { + "epoch": 7.874600638977636, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 9859 + }, + { + "epoch": 7.875399361022364, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 9860 + }, + { + "epoch": 7.876198083067092, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 9861 + }, + { + "epoch": 7.876996805111821, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 9862 + }, + { + "epoch": 7.877795527156549, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0389, + "step": 9863 + }, + { + "epoch": 7.878594249201278, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 9864 + }, + { + "epoch": 7.8793929712460065, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0414, + "step": 9865 + }, + { + "epoch": 7.880191693290735, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 9866 + }, + { + "epoch": 7.8809904153354635, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 9867 + }, + { + "epoch": 7.881789137380192, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 9868 + }, + { + "epoch": 7.88258785942492, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 9869 + }, + { + "epoch": 7.883386581469648, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 9870 + }, + { + "epoch": 7.884185303514377, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 9871 + }, + { + "epoch": 7.884984025559105, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 9872 + }, + { + "epoch": 7.885782747603834, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 9873 + }, + { + "epoch": 7.886581469648562, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 9874 + }, + { + "epoch": 7.887380191693291, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 9875 + }, + { + "epoch": 7.888178913738019, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 9876 + }, + { + "epoch": 7.888977635782748, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.0348, + "step": 9877 + }, + { + "epoch": 7.8897763578274756, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 9878 + }, + { + "epoch": 7.890575079872205, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 9879 + }, + { + "epoch": 7.891373801916933, + "grad_norm": 0.1748046875, + "learning_rate": 0.0005, + "loss": 1.0397, + "step": 9880 + }, + { + "epoch": 7.892172523961661, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0353, + "step": 9881 + }, + { + "epoch": 7.89297124600639, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 9882 + }, + { + "epoch": 7.893769968051118, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 9883 + }, + { + "epoch": 7.894568690095847, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 9884 + }, + { + "epoch": 7.895367412140575, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0431, + "step": 9885 + }, + { + "epoch": 7.896166134185304, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0337, + "step": 9886 + }, + { + "epoch": 7.896964856230032, + "grad_norm": 0.1806640625, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 9887 + }, + { + "epoch": 7.897763578274761, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0391, + "step": 9888 + }, + { + "epoch": 7.8985623003194885, + "grad_norm": 0.1611328125, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 9889 + }, + { + "epoch": 7.899361022364217, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 9890 + }, + { + "epoch": 7.9001597444089455, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 9891 + }, + { + "epoch": 7.900958466453674, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 9892 + }, + { + "epoch": 7.901757188498403, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0386, + "step": 9893 + }, + { + "epoch": 7.902555910543131, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0378, + "step": 9894 + }, + { + "epoch": 7.90335463258786, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 9895 + }, + { + "epoch": 7.904153354632588, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 9896 + }, + { + "epoch": 7.904952076677317, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 9897 + }, + { + "epoch": 7.905750798722044, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 9898 + }, + { + "epoch": 7.906549520766773, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 9899 + }, + { + "epoch": 7.907348242811501, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0361, + "step": 9900 + }, + { + "epoch": 7.90814696485623, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 9901 + }, + { + "epoch": 7.9089456869009584, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 9902 + }, + { + "epoch": 7.909744408945687, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 9903 + }, + { + "epoch": 7.9105431309904155, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 9904 + }, + { + "epoch": 7.911341853035144, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 9905 + }, + { + "epoch": 7.912140575079873, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0362, + "step": 9906 + }, + { + "epoch": 7.9129392971246, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 9907 + }, + { + "epoch": 7.913738019169329, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 9908 + }, + { + "epoch": 7.914536741214057, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 9909 + }, + { + "epoch": 7.915335463258786, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 9910 + }, + { + "epoch": 7.916134185303514, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 9911 + }, + { + "epoch": 7.916932907348243, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0409, + "step": 9912 + }, + { + "epoch": 7.917731629392971, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 9913 + }, + { + "epoch": 7.9185303514377, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0405, + "step": 9914 + }, + { + "epoch": 7.919329073482428, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 9915 + }, + { + "epoch": 7.920127795527156, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 9916 + }, + { + "epoch": 7.9209265175718855, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 9917 + }, + { + "epoch": 7.921725239616613, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 9918 + }, + { + "epoch": 7.922523961661342, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 9919 + }, + { + "epoch": 7.92332268370607, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 9920 + }, + { + "epoch": 7.924121405750799, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0354, + "step": 9921 + }, + { + "epoch": 7.924920127795527, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 9922 + }, + { + "epoch": 7.925718849840256, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 9923 + }, + { + "epoch": 7.926517571884984, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0423, + "step": 9924 + }, + { + "epoch": 7.927316293929713, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 9925 + }, + { + "epoch": 7.928115015974441, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 9926 + }, + { + "epoch": 7.928913738019169, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 9927 + }, + { + "epoch": 7.9297124600638975, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 9928 + }, + { + "epoch": 7.930511182108626, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 9929 + }, + { + "epoch": 7.931309904153355, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 9930 + }, + { + "epoch": 7.932108626198083, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 9931 + }, + { + "epoch": 7.932907348242812, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 9932 + }, + { + "epoch": 7.93370607028754, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 9933 + }, + { + "epoch": 7.934504792332269, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 9934 + }, + { + "epoch": 7.935303514376997, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 9935 + }, + { + "epoch": 7.936102236421725, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 9936 + }, + { + "epoch": 7.936900958466453, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 9937 + }, + { + "epoch": 7.937699680511182, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 9938 + }, + { + "epoch": 7.93849840255591, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 9939 + }, + { + "epoch": 7.939297124600639, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 9940 + }, + { + "epoch": 7.9400958466453675, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 9941 + }, + { + "epoch": 7.940894568690096, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0205, + "step": 9942 + }, + { + "epoch": 7.9416932907348246, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 9943 + }, + { + "epoch": 7.942492012779553, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 9944 + }, + { + "epoch": 7.943290734824281, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 9945 + }, + { + "epoch": 7.944089456869009, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 9946 + }, + { + "epoch": 7.944888178913738, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 9947 + }, + { + "epoch": 7.945686900958466, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 9948 + }, + { + "epoch": 7.946485623003195, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 9949 + }, + { + "epoch": 7.947284345047923, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 9950 + }, + { + "epoch": 7.948083067092652, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 9951 + }, + { + "epoch": 7.94888178913738, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 9952 + }, + { + "epoch": 7.949680511182109, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 9953 + }, + { + "epoch": 7.950479233226837, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 9954 + }, + { + "epoch": 7.951277955271565, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 9955 + }, + { + "epoch": 7.952076677316294, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0412, + "step": 9956 + }, + { + "epoch": 7.952875399361022, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0369, + "step": 9957 + }, + { + "epoch": 7.953674121405751, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 9958 + }, + { + "epoch": 7.954472843450479, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 9959 + }, + { + "epoch": 7.955271565495208, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 9960 + }, + { + "epoch": 7.956070287539936, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 9961 + }, + { + "epoch": 7.956869009584665, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 9962 + }, + { + "epoch": 7.957667731629393, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 9963 + }, + { + "epoch": 7.958466453674122, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 9964 + }, + { + "epoch": 7.9592651757188495, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0377, + "step": 9965 + }, + { + "epoch": 7.960063897763578, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 9966 + }, + { + "epoch": 7.960862619808307, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 9967 + }, + { + "epoch": 7.961661341853035, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 9968 + }, + { + "epoch": 7.962460063897764, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.037, + "step": 9969 + }, + { + "epoch": 7.963258785942492, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 9970 + }, + { + "epoch": 7.964057507987221, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 9971 + }, + { + "epoch": 7.964856230031949, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 9972 + }, + { + "epoch": 7.965654952076678, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0361, + "step": 9973 + }, + { + "epoch": 7.966453674121405, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 9974 + }, + { + "epoch": 7.967252396166134, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 9975 + }, + { + "epoch": 7.968051118210862, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 9976 + }, + { + "epoch": 7.968849840255591, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 9977 + }, + { + "epoch": 7.9696485623003195, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 9978 + }, + { + "epoch": 7.970447284345048, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 9979 + }, + { + "epoch": 7.9712460063897765, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 9980 + }, + { + "epoch": 7.972044728434505, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 9981 + }, + { + "epoch": 7.972843450479234, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 9982 + }, + { + "epoch": 7.973642172523961, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.039, + "step": 9983 + }, + { + "epoch": 7.97444089456869, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 9984 + }, + { + "epoch": 7.975239616613418, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 9985 + }, + { + "epoch": 7.976038338658147, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 9986 + }, + { + "epoch": 7.976837060702875, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 9987 + }, + { + "epoch": 7.977635782747604, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 9988 + }, + { + "epoch": 7.978434504792332, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0372, + "step": 9989 + }, + { + "epoch": 7.979233226837061, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 9990 + }, + { + "epoch": 7.9800319488817895, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 9991 + }, + { + "epoch": 7.980830670926517, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 9992 + }, + { + "epoch": 7.981629392971246, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 9993 + }, + { + "epoch": 7.982428115015974, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 9994 + }, + { + "epoch": 7.983226837060703, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 9995 + }, + { + "epoch": 7.984025559105431, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 9996 + }, + { + "epoch": 7.98482428115016, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 9997 + }, + { + "epoch": 7.985623003194888, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 9998 + }, + { + "epoch": 7.986421725239617, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 9999 + }, + { + "epoch": 7.987220447284345, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.037, + "step": 10000 + }, + { + "epoch": 7.988019169329074, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 10001 + }, + { + "epoch": 7.988817891373802, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0386, + "step": 10002 + }, + { + "epoch": 7.98961661341853, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 10003 + }, + { + "epoch": 7.9904153354632586, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 10004 + }, + { + "epoch": 7.991214057507987, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0369, + "step": 10005 + }, + { + "epoch": 7.992012779552716, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 10006 + }, + { + "epoch": 7.992811501597444, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 10007 + }, + { + "epoch": 7.993610223642173, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 10008 + }, + { + "epoch": 7.994408945686901, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 10009 + }, + { + "epoch": 7.99520766773163, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 10010 + }, + { + "epoch": 7.996006389776358, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0408, + "step": 10011 + }, + { + "epoch": 7.996805111821086, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 10012 + }, + { + "epoch": 7.997603833865814, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0195, + "step": 10013 + }, + { + "epoch": 7.998402555910543, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 10014 + }, + { + "epoch": 7.9992012779552715, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 10015 + }, + { + "epoch": 8.0, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 10016 + }, + { + "epoch": 8.000798722044728, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 10017 + }, + { + "epoch": 8.001597444089457, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 10018 + }, + { + "epoch": 8.002396166134185, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 10019 + }, + { + "epoch": 8.003194888178914, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 10020 + }, + { + "epoch": 8.003993610223642, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 10021 + }, + { + "epoch": 8.004792332268371, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 10022 + }, + { + "epoch": 8.005591054313099, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 10023 + }, + { + "epoch": 8.006389776357828, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 10024 + }, + { + "epoch": 8.007188498402556, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 10025 + }, + { + "epoch": 8.007987220447284, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 10026 + }, + { + "epoch": 8.008785942492013, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 10027 + }, + { + "epoch": 8.00958466453674, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 10028 + }, + { + "epoch": 8.01038338658147, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 10029 + }, + { + "epoch": 8.011182108626198, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 10030 + }, + { + "epoch": 8.011980830670927, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0399, + "step": 10031 + }, + { + "epoch": 8.012779552715655, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 10032 + }, + { + "epoch": 8.013578274760384, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.034, + "step": 10033 + }, + { + "epoch": 8.014376996805112, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 10034 + }, + { + "epoch": 8.01517571884984, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0385, + "step": 10035 + }, + { + "epoch": 8.015974440894569, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 10036 + }, + { + "epoch": 8.016773162939296, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0385, + "step": 10037 + }, + { + "epoch": 8.017571884984026, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 10038 + }, + { + "epoch": 8.018370607028753, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 10039 + }, + { + "epoch": 8.019169329073483, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 10040 + }, + { + "epoch": 8.01996805111821, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 10041 + }, + { + "epoch": 8.02076677316294, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 10042 + }, + { + "epoch": 8.021565495207668, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0388, + "step": 10043 + }, + { + "epoch": 8.022364217252397, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 10044 + }, + { + "epoch": 8.023162939297125, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 10045 + }, + { + "epoch": 8.023961661341852, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 10046 + }, + { + "epoch": 8.024760383386582, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 10047 + }, + { + "epoch": 8.02555910543131, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 10048 + }, + { + "epoch": 8.026357827476039, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.039, + "step": 10049 + }, + { + "epoch": 8.027156549520766, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0372, + "step": 10050 + }, + { + "epoch": 8.027955271565496, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 10051 + }, + { + "epoch": 8.028753993610223, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0196, + "step": 10052 + }, + { + "epoch": 8.029552715654953, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 10053 + }, + { + "epoch": 8.03035143769968, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 10054 + }, + { + "epoch": 8.031150159744408, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 10055 + }, + { + "epoch": 8.031948881789138, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 10056 + }, + { + "epoch": 8.032747603833865, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 10057 + }, + { + "epoch": 8.033546325878595, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 10058 + }, + { + "epoch": 8.034345047923322, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 10059 + }, + { + "epoch": 8.035143769968052, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 10060 + }, + { + "epoch": 8.03594249201278, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 10061 + }, + { + "epoch": 8.036741214057509, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 10062 + }, + { + "epoch": 8.037539936102236, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 10063 + }, + { + "epoch": 8.038338658146964, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 10064 + }, + { + "epoch": 8.039137380191693, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.0414, + "step": 10065 + }, + { + "epoch": 8.039936102236421, + "grad_norm": 0.16796875, + "learning_rate": 0.0005, + "loss": 1.034, + "step": 10066 + }, + { + "epoch": 8.04073482428115, + "grad_norm": 0.20703125, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 10067 + }, + { + "epoch": 8.041533546325878, + "grad_norm": 0.255859375, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 10068 + }, + { + "epoch": 8.042332268370608, + "grad_norm": 0.1640625, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 10069 + }, + { + "epoch": 8.043130990415335, + "grad_norm": 0.244140625, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 10070 + }, + { + "epoch": 8.043929712460065, + "grad_norm": 0.1962890625, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 10071 + }, + { + "epoch": 8.044728434504792, + "grad_norm": 0.333984375, + "learning_rate": 0.0005, + "loss": 1.0412, + "step": 10072 + }, + { + "epoch": 8.04552715654952, + "grad_norm": 0.1806640625, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 10073 + }, + { + "epoch": 8.04632587859425, + "grad_norm": 0.39453125, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 10074 + }, + { + "epoch": 8.047124600638977, + "grad_norm": 0.1787109375, + "learning_rate": 0.0005, + "loss": 1.0427, + "step": 10075 + }, + { + "epoch": 8.047923322683706, + "grad_norm": 0.494140625, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 10076 + }, + { + "epoch": 8.048722044728434, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0377, + "step": 10077 + }, + { + "epoch": 8.049520766773163, + "grad_norm": 0.60546875, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 10078 + }, + { + "epoch": 8.050319488817891, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0413, + "step": 10079 + }, + { + "epoch": 8.05111821086262, + "grad_norm": 0.435546875, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 10080 + }, + { + "epoch": 8.051916932907348, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.043, + "step": 10081 + }, + { + "epoch": 8.052715654952078, + "grad_norm": 0.42578125, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 10082 + }, + { + "epoch": 8.053514376996805, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 10083 + }, + { + "epoch": 8.054313099041533, + "grad_norm": 0.35546875, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 10084 + }, + { + "epoch": 8.055111821086262, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 10085 + }, + { + "epoch": 8.05591054313099, + "grad_norm": 0.30078125, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 10086 + }, + { + "epoch": 8.05670926517572, + "grad_norm": 0.1875, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 10087 + }, + { + "epoch": 8.057507987220447, + "grad_norm": 0.28125, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 10088 + }, + { + "epoch": 8.058306709265176, + "grad_norm": 0.2294921875, + "learning_rate": 0.0005, + "loss": 1.0431, + "step": 10089 + }, + { + "epoch": 8.059105431309904, + "grad_norm": 0.2431640625, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 10090 + }, + { + "epoch": 8.059904153354633, + "grad_norm": 0.1806640625, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 10091 + }, + { + "epoch": 8.060702875399361, + "grad_norm": 0.1708984375, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 10092 + }, + { + "epoch": 8.061501597444089, + "grad_norm": 0.1533203125, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 10093 + }, + { + "epoch": 8.062300319488818, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 10094 + }, + { + "epoch": 8.063099041533546, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 10095 + }, + { + "epoch": 8.063897763578275, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 10096 + }, + { + "epoch": 8.064696485623003, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 10097 + }, + { + "epoch": 8.065495207667732, + "grad_norm": 0.1806640625, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 10098 + }, + { + "epoch": 8.06629392971246, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 10099 + }, + { + "epoch": 8.06709265175719, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 10100 + }, + { + "epoch": 8.067891373801917, + "grad_norm": 0.1748046875, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 10101 + }, + { + "epoch": 8.068690095846645, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 10102 + }, + { + "epoch": 8.069488817891374, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 10103 + }, + { + "epoch": 8.070287539936102, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 10104 + }, + { + "epoch": 8.071086261980831, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 10105 + }, + { + "epoch": 8.071884984025559, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0384, + "step": 10106 + }, + { + "epoch": 8.072683706070288, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 10107 + }, + { + "epoch": 8.073482428115016, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0376, + "step": 10108 + }, + { + "epoch": 8.074281150159745, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 10109 + }, + { + "epoch": 8.075079872204473, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 10110 + }, + { + "epoch": 8.0758785942492, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 10111 + }, + { + "epoch": 8.07667731629393, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 10112 + }, + { + "epoch": 8.077476038338657, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 10113 + }, + { + "epoch": 8.078274760383387, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 10114 + }, + { + "epoch": 8.079073482428115, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 10115 + }, + { + "epoch": 8.079872204472844, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 10116 + }, + { + "epoch": 8.080670926517572, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 10117 + }, + { + "epoch": 8.081469648562301, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 10118 + }, + { + "epoch": 8.082268370607029, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 10119 + }, + { + "epoch": 8.083067092651758, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 10120 + }, + { + "epoch": 8.083865814696486, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 10121 + }, + { + "epoch": 8.084664536741213, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 10122 + }, + { + "epoch": 8.085463258785943, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0378, + "step": 10123 + }, + { + "epoch": 8.08626198083067, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0338, + "step": 10124 + }, + { + "epoch": 8.0870607028754, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 10125 + }, + { + "epoch": 8.087859424920127, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 10126 + }, + { + "epoch": 8.088658146964857, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 10127 + }, + { + "epoch": 8.089456869009584, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 10128 + }, + { + "epoch": 8.090255591054314, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 10129 + }, + { + "epoch": 8.091054313099042, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 10130 + }, + { + "epoch": 8.09185303514377, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 10131 + }, + { + "epoch": 8.092651757188499, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 10132 + }, + { + "epoch": 8.093450479233226, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 10133 + }, + { + "epoch": 8.094249201277956, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 10134 + }, + { + "epoch": 8.095047923322683, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0378, + "step": 10135 + }, + { + "epoch": 8.095846645367413, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 10136 + }, + { + "epoch": 8.09664536741214, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 10137 + }, + { + "epoch": 8.09744408945687, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 10138 + }, + { + "epoch": 8.098242811501597, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 10139 + }, + { + "epoch": 8.099041533546325, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 10140 + }, + { + "epoch": 8.099840255591054, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 10141 + }, + { + "epoch": 8.100638977635782, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 10142 + }, + { + "epoch": 8.101437699680512, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 10143 + }, + { + "epoch": 8.10223642172524, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 10144 + }, + { + "epoch": 8.103035143769969, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 10145 + }, + { + "epoch": 8.103833865814696, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 10146 + }, + { + "epoch": 8.104632587859426, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 10147 + }, + { + "epoch": 8.105431309904153, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 10148 + }, + { + "epoch": 8.106230031948881, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 10149 + }, + { + "epoch": 8.10702875399361, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 10150 + }, + { + "epoch": 8.107827476038338, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 10151 + }, + { + "epoch": 8.108626198083067, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 10152 + }, + { + "epoch": 8.109424920127795, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 10153 + }, + { + "epoch": 8.110223642172524, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 10154 + }, + { + "epoch": 8.111022364217252, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0422, + "step": 10155 + }, + { + "epoch": 8.111821086261982, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 10156 + }, + { + "epoch": 8.11261980830671, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 10157 + }, + { + "epoch": 8.113418530351439, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 10158 + }, + { + "epoch": 8.114217252396166, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 10159 + }, + { + "epoch": 8.115015974440894, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 10160 + }, + { + "epoch": 8.115814696485623, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 10161 + }, + { + "epoch": 8.116613418530351, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 10162 + }, + { + "epoch": 8.11741214057508, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0451, + "step": 10163 + }, + { + "epoch": 8.118210862619808, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 10164 + }, + { + "epoch": 8.119009584664537, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 10165 + }, + { + "epoch": 8.119808306709265, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 10166 + }, + { + "epoch": 8.120607028753994, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 10167 + }, + { + "epoch": 8.121405750798722, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 10168 + }, + { + "epoch": 8.12220447284345, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 10169 + }, + { + "epoch": 8.12300319488818, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 10170 + }, + { + "epoch": 8.123801916932907, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 10171 + }, + { + "epoch": 8.124600638977636, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0378, + "step": 10172 + }, + { + "epoch": 8.125399361022364, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 10173 + }, + { + "epoch": 8.126198083067093, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0199, + "step": 10174 + }, + { + "epoch": 8.12699680511182, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 10175 + }, + { + "epoch": 8.12779552715655, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 10176 + }, + { + "epoch": 8.128594249201278, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 10177 + }, + { + "epoch": 8.129392971246006, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 10178 + }, + { + "epoch": 8.130191693290735, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 10179 + }, + { + "epoch": 8.130990415335463, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 10180 + }, + { + "epoch": 8.131789137380192, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 10181 + }, + { + "epoch": 8.13258785942492, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 10182 + }, + { + "epoch": 8.133386581469649, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 10183 + }, + { + "epoch": 8.134185303514377, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 10184 + }, + { + "epoch": 8.134984025559106, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 10185 + }, + { + "epoch": 8.135782747603834, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 10186 + }, + { + "epoch": 8.136581469648561, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 10187 + }, + { + "epoch": 8.13738019169329, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 10188 + }, + { + "epoch": 8.138178913738018, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 10189 + }, + { + "epoch": 8.138977635782748, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 10190 + }, + { + "epoch": 8.139776357827476, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 10191 + }, + { + "epoch": 8.140575079872205, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0361, + "step": 10192 + }, + { + "epoch": 8.141373801916933, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 10193 + }, + { + "epoch": 8.142172523961662, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 10194 + }, + { + "epoch": 8.14297124600639, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 10195 + }, + { + "epoch": 8.143769968051119, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 10196 + }, + { + "epoch": 8.144568690095847, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 10197 + }, + { + "epoch": 8.145367412140574, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 10198 + }, + { + "epoch": 8.146166134185304, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0369, + "step": 10199 + }, + { + "epoch": 8.146964856230031, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 10200 + }, + { + "epoch": 8.14776357827476, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 10201 + }, + { + "epoch": 8.148562300319488, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0373, + "step": 10202 + }, + { + "epoch": 8.149361022364218, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 10203 + }, + { + "epoch": 8.150159744408946, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.043, + "step": 10204 + }, + { + "epoch": 8.150958466453675, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 10205 + }, + { + "epoch": 8.151757188498403, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 10206 + }, + { + "epoch": 8.15255591054313, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 10207 + }, + { + "epoch": 8.15335463258786, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0369, + "step": 10208 + }, + { + "epoch": 8.154153354632587, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 10209 + }, + { + "epoch": 8.154952076677317, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 10210 + }, + { + "epoch": 8.155750798722044, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0354, + "step": 10211 + }, + { + "epoch": 8.156549520766774, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0187, + "step": 10212 + }, + { + "epoch": 8.157348242811501, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 10213 + }, + { + "epoch": 8.15814696485623, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 10214 + }, + { + "epoch": 8.158945686900958, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 10215 + }, + { + "epoch": 8.159744408945686, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 10216 + }, + { + "epoch": 8.160543130990416, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.036, + "step": 10217 + }, + { + "epoch": 8.161341853035143, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 10218 + }, + { + "epoch": 8.162140575079873, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 10219 + }, + { + "epoch": 8.1629392971246, + "grad_norm": 0.1767578125, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 10220 + }, + { + "epoch": 8.16373801916933, + "grad_norm": 0.1611328125, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 10221 + }, + { + "epoch": 8.164536741214057, + "grad_norm": 0.19921875, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 10222 + }, + { + "epoch": 8.165335463258787, + "grad_norm": 0.1865234375, + "learning_rate": 0.0005, + "loss": 1.0362, + "step": 10223 + }, + { + "epoch": 8.166134185303514, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 10224 + }, + { + "epoch": 8.166932907348242, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 10225 + }, + { + "epoch": 8.167731629392971, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0354, + "step": 10226 + }, + { + "epoch": 8.168530351437699, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 10227 + }, + { + "epoch": 8.169329073482428, + "grad_norm": 0.2060546875, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 10228 + }, + { + "epoch": 8.170127795527156, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 10229 + }, + { + "epoch": 8.170926517571885, + "grad_norm": 0.1787109375, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 10230 + }, + { + "epoch": 8.171725239616613, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 10231 + }, + { + "epoch": 8.172523961661343, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 10232 + }, + { + "epoch": 8.17332268370607, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 10233 + }, + { + "epoch": 8.1741214057508, + "grad_norm": 0.16796875, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 10234 + }, + { + "epoch": 8.174920127795527, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 10235 + }, + { + "epoch": 8.175718849840255, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.034, + "step": 10236 + }, + { + "epoch": 8.176517571884984, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0373, + "step": 10237 + }, + { + "epoch": 8.177316293929712, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 10238 + }, + { + "epoch": 8.178115015974441, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 10239 + }, + { + "epoch": 8.178913738019169, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 10240 + }, + { + "epoch": 8.179712460063898, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 10241 + }, + { + "epoch": 8.180511182108626, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 10242 + }, + { + "epoch": 8.181309904153355, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 10243 + }, + { + "epoch": 8.182108626198083, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 10244 + }, + { + "epoch": 8.18290734824281, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 10245 + }, + { + "epoch": 8.18370607028754, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 10246 + }, + { + "epoch": 8.184504792332268, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0424, + "step": 10247 + }, + { + "epoch": 8.185303514376997, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0337, + "step": 10248 + }, + { + "epoch": 8.186102236421725, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 10249 + }, + { + "epoch": 8.186900958466454, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 10250 + }, + { + "epoch": 8.187699680511182, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 10251 + }, + { + "epoch": 8.188498402555911, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 10252 + }, + { + "epoch": 8.189297124600639, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 10253 + }, + { + "epoch": 8.190095846645367, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 10254 + }, + { + "epoch": 8.190894568690096, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 10255 + }, + { + "epoch": 8.191693290734824, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 10256 + }, + { + "epoch": 8.192492012779553, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 10257 + }, + { + "epoch": 8.19329073482428, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 10258 + }, + { + "epoch": 8.19408945686901, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 10259 + }, + { + "epoch": 8.194888178913738, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 10260 + }, + { + "epoch": 8.195686900958467, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 10261 + }, + { + "epoch": 8.196485623003195, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 10262 + }, + { + "epoch": 8.197284345047922, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 10263 + }, + { + "epoch": 8.198083067092652, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 10264 + }, + { + "epoch": 8.19888178913738, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 10265 + }, + { + "epoch": 8.199680511182109, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 10266 + }, + { + "epoch": 8.200479233226837, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.034, + "step": 10267 + }, + { + "epoch": 8.201277955271566, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 10268 + }, + { + "epoch": 8.202076677316294, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 10269 + }, + { + "epoch": 8.202875399361023, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 10270 + }, + { + "epoch": 8.20367412140575, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 10271 + }, + { + "epoch": 8.204472843450478, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 10272 + }, + { + "epoch": 8.205271565495208, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 10273 + }, + { + "epoch": 8.206070287539935, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 10274 + }, + { + "epoch": 8.206869009584665, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 10275 + }, + { + "epoch": 8.207667731629392, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 10276 + }, + { + "epoch": 8.208466453674122, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 10277 + }, + { + "epoch": 8.20926517571885, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0455, + "step": 10278 + }, + { + "epoch": 8.210063897763579, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 10279 + }, + { + "epoch": 8.210862619808307, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 10280 + }, + { + "epoch": 8.211661341853036, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 10281 + }, + { + "epoch": 8.212460063897764, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 10282 + }, + { + "epoch": 8.213258785942491, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0362, + "step": 10283 + }, + { + "epoch": 8.21405750798722, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 10284 + }, + { + "epoch": 8.214856230031948, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 10285 + }, + { + "epoch": 8.215654952076678, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 10286 + }, + { + "epoch": 8.216453674121405, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 10287 + }, + { + "epoch": 8.217252396166135, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 10288 + }, + { + "epoch": 8.218051118210862, + "grad_norm": 0.1669921875, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 10289 + }, + { + "epoch": 8.218849840255592, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 10290 + }, + { + "epoch": 8.21964856230032, + "grad_norm": 0.1865234375, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 10291 + }, + { + "epoch": 8.220447284345047, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.0417, + "step": 10292 + }, + { + "epoch": 8.221246006389777, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 10293 + }, + { + "epoch": 8.222044728434504, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 10294 + }, + { + "epoch": 8.222843450479234, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 10295 + }, + { + "epoch": 8.223642172523961, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 10296 + }, + { + "epoch": 8.22444089456869, + "grad_norm": 0.1796875, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 10297 + }, + { + "epoch": 8.225239616613418, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 10298 + }, + { + "epoch": 8.226038338658148, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 10299 + }, + { + "epoch": 8.226837060702875, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 10300 + }, + { + "epoch": 8.227635782747603, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0411, + "step": 10301 + }, + { + "epoch": 8.228434504792332, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 10302 + }, + { + "epoch": 8.22923322683706, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 10303 + }, + { + "epoch": 8.23003194888179, + "grad_norm": 0.1533203125, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 10304 + }, + { + "epoch": 8.230830670926517, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 10305 + }, + { + "epoch": 8.231629392971247, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 10306 + }, + { + "epoch": 8.232428115015974, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 10307 + }, + { + "epoch": 8.233226837060704, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 10308 + }, + { + "epoch": 8.234025559105431, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 10309 + }, + { + "epoch": 8.23482428115016, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 10310 + }, + { + "epoch": 8.235623003194888, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 10311 + }, + { + "epoch": 8.236421725239616, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 10312 + }, + { + "epoch": 8.237220447284345, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 10313 + }, + { + "epoch": 8.238019169329073, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 10314 + }, + { + "epoch": 8.238817891373802, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 10315 + }, + { + "epoch": 8.23961661341853, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0395, + "step": 10316 + }, + { + "epoch": 8.24041533546326, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 10317 + }, + { + "epoch": 8.241214057507987, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 10318 + }, + { + "epoch": 8.242012779552716, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 10319 + }, + { + "epoch": 8.242811501597444, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 10320 + }, + { + "epoch": 8.243610223642172, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 10321 + }, + { + "epoch": 8.244408945686901, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 10322 + }, + { + "epoch": 8.245207667731629, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 10323 + }, + { + "epoch": 8.246006389776358, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 10324 + }, + { + "epoch": 8.246805111821086, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 10325 + }, + { + "epoch": 8.247603833865815, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 10326 + }, + { + "epoch": 8.248402555910543, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 10327 + }, + { + "epoch": 8.249201277955272, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.039, + "step": 10328 + }, + { + "epoch": 8.25, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 10329 + }, + { + "epoch": 8.250798722044728, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0384, + "step": 10330 + }, + { + "epoch": 8.251597444089457, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 10331 + }, + { + "epoch": 8.252396166134185, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 10332 + }, + { + "epoch": 8.253194888178914, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 10333 + }, + { + "epoch": 8.253993610223642, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 10334 + }, + { + "epoch": 8.254792332268371, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 10335 + }, + { + "epoch": 8.255591054313099, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.044, + "step": 10336 + }, + { + "epoch": 8.256389776357828, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 10337 + }, + { + "epoch": 8.257188498402556, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 10338 + }, + { + "epoch": 8.257987220447284, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 10339 + }, + { + "epoch": 8.258785942492013, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 10340 + }, + { + "epoch": 8.25958466453674, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 10341 + }, + { + "epoch": 8.26038338658147, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 10342 + }, + { + "epoch": 8.261182108626198, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0425, + "step": 10343 + }, + { + "epoch": 8.261980830670927, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 10344 + }, + { + "epoch": 8.262779552715655, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 10345 + }, + { + "epoch": 8.263578274760384, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 10346 + }, + { + "epoch": 8.264376996805112, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 10347 + }, + { + "epoch": 8.26517571884984, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 10348 + }, + { + "epoch": 8.265974440894569, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 10349 + }, + { + "epoch": 8.266773162939296, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 10350 + }, + { + "epoch": 8.267571884984026, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 10351 + }, + { + "epoch": 8.268370607028753, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 10352 + }, + { + "epoch": 8.269169329073483, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0389, + "step": 10353 + }, + { + "epoch": 8.26996805111821, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 10354 + }, + { + "epoch": 8.27076677316294, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 10355 + }, + { + "epoch": 8.271565495207668, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0414, + "step": 10356 + }, + { + "epoch": 8.272364217252397, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 10357 + }, + { + "epoch": 8.273162939297125, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 10358 + }, + { + "epoch": 8.273961661341852, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 10359 + }, + { + "epoch": 8.274760383386582, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 10360 + }, + { + "epoch": 8.27555910543131, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0369, + "step": 10361 + }, + { + "epoch": 8.276357827476039, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 10362 + }, + { + "epoch": 8.277156549520766, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 10363 + }, + { + "epoch": 8.277955271565496, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 10364 + }, + { + "epoch": 8.278753993610223, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 10365 + }, + { + "epoch": 8.279552715654953, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 10366 + }, + { + "epoch": 8.28035143769968, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 10367 + }, + { + "epoch": 8.281150159744408, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 10368 + }, + { + "epoch": 8.281948881789138, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 10369 + }, + { + "epoch": 8.282747603833865, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 10370 + }, + { + "epoch": 8.283546325878595, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 10371 + }, + { + "epoch": 8.284345047923322, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 10372 + }, + { + "epoch": 8.285143769968052, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 10373 + }, + { + "epoch": 8.28594249201278, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 10374 + }, + { + "epoch": 8.286741214057509, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 10375 + }, + { + "epoch": 8.287539936102236, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 10376 + }, + { + "epoch": 8.288338658146964, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 10377 + }, + { + "epoch": 8.289137380191693, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 10378 + }, + { + "epoch": 8.289936102236421, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 10379 + }, + { + "epoch": 8.29073482428115, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0401, + "step": 10380 + }, + { + "epoch": 8.291533546325878, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 10381 + }, + { + "epoch": 8.292332268370608, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0401, + "step": 10382 + }, + { + "epoch": 8.293130990415335, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 10383 + }, + { + "epoch": 8.293929712460065, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 10384 + }, + { + "epoch": 8.294728434504792, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 10385 + }, + { + "epoch": 8.295527156549522, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 10386 + }, + { + "epoch": 8.29632587859425, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0399, + "step": 10387 + }, + { + "epoch": 8.297124600638977, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 10388 + }, + { + "epoch": 8.297923322683706, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0389, + "step": 10389 + }, + { + "epoch": 8.298722044728434, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 10390 + }, + { + "epoch": 8.299520766773163, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 10391 + }, + { + "epoch": 8.300319488817891, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 10392 + }, + { + "epoch": 8.30111821086262, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 10393 + }, + { + "epoch": 8.301916932907348, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 10394 + }, + { + "epoch": 8.302715654952078, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 10395 + }, + { + "epoch": 8.303514376996805, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 10396 + }, + { + "epoch": 8.304313099041533, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 10397 + }, + { + "epoch": 8.305111821086262, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 10398 + }, + { + "epoch": 8.30591054313099, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 10399 + }, + { + "epoch": 8.30670926517572, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 10400 + }, + { + "epoch": 8.307507987220447, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0397, + "step": 10401 + }, + { + "epoch": 8.308306709265176, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 10402 + }, + { + "epoch": 8.309105431309904, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 10403 + }, + { + "epoch": 8.309904153354633, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 10404 + }, + { + "epoch": 8.310702875399361, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0404, + "step": 10405 + }, + { + "epoch": 8.311501597444089, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 10406 + }, + { + "epoch": 8.312300319488818, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 10407 + }, + { + "epoch": 8.313099041533546, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 10408 + }, + { + "epoch": 8.313897763578275, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 10409 + }, + { + "epoch": 8.314696485623003, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 10410 + }, + { + "epoch": 8.315495207667732, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 10411 + }, + { + "epoch": 8.31629392971246, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 10412 + }, + { + "epoch": 8.31709265175719, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 10413 + }, + { + "epoch": 8.317891373801917, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 10414 + }, + { + "epoch": 8.318690095846645, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 10415 + }, + { + "epoch": 8.319488817891374, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0338, + "step": 10416 + }, + { + "epoch": 8.320287539936102, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 10417 + }, + { + "epoch": 8.321086261980831, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0413, + "step": 10418 + }, + { + "epoch": 8.321884984025559, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 10419 + }, + { + "epoch": 8.322683706070288, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 10420 + }, + { + "epoch": 8.323482428115016, + "grad_norm": 0.1689453125, + "learning_rate": 0.0005, + "loss": 1.0403, + "step": 10421 + }, + { + "epoch": 8.324281150159745, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 10422 + }, + { + "epoch": 8.325079872204473, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 10423 + }, + { + "epoch": 8.3258785942492, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 10424 + }, + { + "epoch": 8.32667731629393, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0354, + "step": 10425 + }, + { + "epoch": 8.327476038338657, + "grad_norm": 0.197265625, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 10426 + }, + { + "epoch": 8.328274760383387, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 10427 + }, + { + "epoch": 8.329073482428115, + "grad_norm": 0.16796875, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 10428 + }, + { + "epoch": 8.329872204472844, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 10429 + }, + { + "epoch": 8.330670926517572, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 10430 + }, + { + "epoch": 8.331469648562301, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 10431 + }, + { + "epoch": 8.332268370607029, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 10432 + }, + { + "epoch": 8.333067092651758, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 10433 + }, + { + "epoch": 8.333865814696486, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 10434 + }, + { + "epoch": 8.334664536741213, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 10435 + }, + { + "epoch": 8.335463258785943, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 10436 + }, + { + "epoch": 8.33626198083067, + "grad_norm": 0.1689453125, + "learning_rate": 0.0005, + "loss": 1.0406, + "step": 10437 + }, + { + "epoch": 8.3370607028754, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 10438 + }, + { + "epoch": 8.337859424920127, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 10439 + }, + { + "epoch": 8.338658146964857, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 10440 + }, + { + "epoch": 8.339456869009584, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 10441 + }, + { + "epoch": 8.340255591054314, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0369, + "step": 10442 + }, + { + "epoch": 8.341054313099042, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 10443 + }, + { + "epoch": 8.34185303514377, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 10444 + }, + { + "epoch": 8.342651757188499, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 10445 + }, + { + "epoch": 8.343450479233226, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 10446 + }, + { + "epoch": 8.344249201277956, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 10447 + }, + { + "epoch": 8.345047923322683, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 10448 + }, + { + "epoch": 8.345846645367413, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 10449 + }, + { + "epoch": 8.34664536741214, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 10450 + }, + { + "epoch": 8.34744408945687, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 10451 + }, + { + "epoch": 8.348242811501597, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 10452 + }, + { + "epoch": 8.349041533546325, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 10453 + }, + { + "epoch": 8.349840255591054, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 10454 + }, + { + "epoch": 8.350638977635782, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.044, + "step": 10455 + }, + { + "epoch": 8.351437699680512, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0376, + "step": 10456 + }, + { + "epoch": 8.35223642172524, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 10457 + }, + { + "epoch": 8.353035143769969, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 10458 + }, + { + "epoch": 8.353833865814696, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 10459 + }, + { + "epoch": 8.354632587859426, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 10460 + }, + { + "epoch": 8.355431309904153, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 10461 + }, + { + "epoch": 8.356230031948883, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0369, + "step": 10462 + }, + { + "epoch": 8.35702875399361, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0417, + "step": 10463 + }, + { + "epoch": 8.357827476038338, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 10464 + }, + { + "epoch": 8.358626198083067, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 10465 + }, + { + "epoch": 8.359424920127795, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 10466 + }, + { + "epoch": 8.360223642172524, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 10467 + }, + { + "epoch": 8.361022364217252, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 10468 + }, + { + "epoch": 8.361821086261982, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 10469 + }, + { + "epoch": 8.36261980830671, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 10470 + }, + { + "epoch": 8.363418530351439, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 10471 + }, + { + "epoch": 8.364217252396166, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 10472 + }, + { + "epoch": 8.365015974440894, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 10473 + }, + { + "epoch": 8.365814696485623, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 10474 + }, + { + "epoch": 8.366613418530351, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 10475 + }, + { + "epoch": 8.36741214057508, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 10476 + }, + { + "epoch": 8.368210862619808, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 10477 + }, + { + "epoch": 8.369009584664537, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 10478 + }, + { + "epoch": 8.369808306709265, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 10479 + }, + { + "epoch": 8.370607028753994, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 10480 + }, + { + "epoch": 8.371405750798722, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 10481 + }, + { + "epoch": 8.37220447284345, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 10482 + }, + { + "epoch": 8.37300319488818, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 10483 + }, + { + "epoch": 8.373801916932907, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 10484 + }, + { + "epoch": 8.374600638977636, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 10485 + }, + { + "epoch": 8.375399361022364, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 10486 + }, + { + "epoch": 8.376198083067093, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 10487 + }, + { + "epoch": 8.37699680511182, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 10488 + }, + { + "epoch": 8.37779552715655, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 10489 + }, + { + "epoch": 8.378594249201278, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 10490 + }, + { + "epoch": 8.379392971246006, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 10491 + }, + { + "epoch": 8.380191693290735, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0385, + "step": 10492 + }, + { + "epoch": 8.380990415335463, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0369, + "step": 10493 + }, + { + "epoch": 8.381789137380192, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 10494 + }, + { + "epoch": 8.38258785942492, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 10495 + }, + { + "epoch": 8.383386581469649, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 10496 + }, + { + "epoch": 8.384185303514377, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 10497 + }, + { + "epoch": 8.384984025559106, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 10498 + }, + { + "epoch": 8.385782747603834, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 10499 + }, + { + "epoch": 8.386581469648561, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 10500 + }, + { + "epoch": 8.38738019169329, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0388, + "step": 10501 + }, + { + "epoch": 8.388178913738018, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0161, + "step": 10502 + }, + { + "epoch": 8.388977635782748, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 10503 + }, + { + "epoch": 8.389776357827476, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 10504 + }, + { + "epoch": 8.390575079872205, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 10505 + }, + { + "epoch": 8.391373801916933, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 10506 + }, + { + "epoch": 8.392172523961662, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 10507 + }, + { + "epoch": 8.39297124600639, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 10508 + }, + { + "epoch": 8.393769968051117, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 10509 + }, + { + "epoch": 8.394568690095847, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 10510 + }, + { + "epoch": 8.395367412140574, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 10511 + }, + { + "epoch": 8.396166134185304, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 10512 + }, + { + "epoch": 8.396964856230031, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0373, + "step": 10513 + }, + { + "epoch": 8.39776357827476, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0183, + "step": 10514 + }, + { + "epoch": 8.398562300319488, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 10515 + }, + { + "epoch": 8.399361022364218, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 10516 + }, + { + "epoch": 8.400159744408946, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 10517 + }, + { + "epoch": 8.400958466453675, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 10518 + }, + { + "epoch": 8.401757188498403, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 10519 + }, + { + "epoch": 8.40255591054313, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 10520 + }, + { + "epoch": 8.40335463258786, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 10521 + }, + { + "epoch": 8.404153354632587, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.036, + "step": 10522 + }, + { + "epoch": 8.404952076677317, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 10523 + }, + { + "epoch": 8.405750798722044, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 10524 + }, + { + "epoch": 8.406549520766774, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 10525 + }, + { + "epoch": 8.407348242811501, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 10526 + }, + { + "epoch": 8.40814696485623, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 10527 + }, + { + "epoch": 8.408945686900958, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 10528 + }, + { + "epoch": 8.409744408945686, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 10529 + }, + { + "epoch": 8.410543130990416, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 10530 + }, + { + "epoch": 8.411341853035143, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 10531 + }, + { + "epoch": 8.412140575079873, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 10532 + }, + { + "epoch": 8.4129392971246, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 10533 + }, + { + "epoch": 8.41373801916933, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 10534 + }, + { + "epoch": 8.414536741214057, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 10535 + }, + { + "epoch": 8.415335463258787, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 10536 + }, + { + "epoch": 8.416134185303514, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 10537 + }, + { + "epoch": 8.416932907348242, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 10538 + }, + { + "epoch": 8.417731629392971, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 10539 + }, + { + "epoch": 8.418530351437699, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 10540 + }, + { + "epoch": 8.419329073482428, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 10541 + }, + { + "epoch": 8.420127795527156, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 10542 + }, + { + "epoch": 8.420926517571885, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 10543 + }, + { + "epoch": 8.421725239616613, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 10544 + }, + { + "epoch": 8.422523961661343, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 10545 + }, + { + "epoch": 8.42332268370607, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 10546 + }, + { + "epoch": 8.4241214057508, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 10547 + }, + { + "epoch": 8.424920127795527, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 10548 + }, + { + "epoch": 8.425718849840255, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 10549 + }, + { + "epoch": 8.426517571884984, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 10550 + }, + { + "epoch": 8.427316293929712, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 10551 + }, + { + "epoch": 8.428115015974441, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0401, + "step": 10552 + }, + { + "epoch": 8.428913738019169, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 10553 + }, + { + "epoch": 8.429712460063898, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 10554 + }, + { + "epoch": 8.430511182108626, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 10555 + }, + { + "epoch": 8.431309904153355, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 10556 + }, + { + "epoch": 8.432108626198083, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 10557 + }, + { + "epoch": 8.43290734824281, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 10558 + }, + { + "epoch": 8.43370607028754, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 10559 + }, + { + "epoch": 8.434504792332268, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 10560 + }, + { + "epoch": 8.435303514376997, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 10561 + }, + { + "epoch": 8.436102236421725, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 10562 + }, + { + "epoch": 8.436900958466454, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 10563 + }, + { + "epoch": 8.437699680511182, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 10564 + }, + { + "epoch": 8.438498402555911, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 10565 + }, + { + "epoch": 8.439297124600639, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 10566 + }, + { + "epoch": 8.440095846645367, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 10567 + }, + { + "epoch": 8.440894568690096, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 10568 + }, + { + "epoch": 8.441693290734824, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 10569 + }, + { + "epoch": 8.442492012779553, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 10570 + }, + { + "epoch": 8.44329073482428, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0409, + "step": 10571 + }, + { + "epoch": 8.44408945686901, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 10572 + }, + { + "epoch": 8.444888178913738, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0393, + "step": 10573 + }, + { + "epoch": 8.445686900958467, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 10574 + }, + { + "epoch": 8.446485623003195, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 10575 + }, + { + "epoch": 8.447284345047922, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 10576 + }, + { + "epoch": 8.448083067092652, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0421, + "step": 10577 + }, + { + "epoch": 8.44888178913738, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 10578 + }, + { + "epoch": 8.449680511182109, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 10579 + }, + { + "epoch": 8.450479233226837, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 10580 + }, + { + "epoch": 8.451277955271566, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 10581 + }, + { + "epoch": 8.452076677316294, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 10582 + }, + { + "epoch": 8.452875399361023, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0183, + "step": 10583 + }, + { + "epoch": 8.45367412140575, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 10584 + }, + { + "epoch": 8.454472843450478, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 10585 + }, + { + "epoch": 8.455271565495208, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 10586 + }, + { + "epoch": 8.456070287539935, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 10587 + }, + { + "epoch": 8.456869009584665, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0407, + "step": 10588 + }, + { + "epoch": 8.457667731629392, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 10589 + }, + { + "epoch": 8.458466453674122, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 10590 + }, + { + "epoch": 8.45926517571885, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0397, + "step": 10591 + }, + { + "epoch": 8.460063897763579, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 10592 + }, + { + "epoch": 8.460862619808307, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 10593 + }, + { + "epoch": 8.461661341853036, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 10594 + }, + { + "epoch": 8.462460063897764, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 10595 + }, + { + "epoch": 8.463258785942491, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 10596 + }, + { + "epoch": 8.46405750798722, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 10597 + }, + { + "epoch": 8.464856230031948, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 10598 + }, + { + "epoch": 8.465654952076678, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 10599 + }, + { + "epoch": 8.466453674121405, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 10600 + }, + { + "epoch": 8.467252396166135, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 10601 + }, + { + "epoch": 8.468051118210862, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0179, + "step": 10602 + }, + { + "epoch": 8.468849840255592, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0348, + "step": 10603 + }, + { + "epoch": 8.46964856230032, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 10604 + }, + { + "epoch": 8.470447284345047, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 10605 + }, + { + "epoch": 8.471246006389777, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 10606 + }, + { + "epoch": 8.472044728434504, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0348, + "step": 10607 + }, + { + "epoch": 8.472843450479234, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 10608 + }, + { + "epoch": 8.473642172523961, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 10609 + }, + { + "epoch": 8.47444089456869, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 10610 + }, + { + "epoch": 8.475239616613418, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 10611 + }, + { + "epoch": 8.476038338658148, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 10612 + }, + { + "epoch": 8.476837060702875, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 10613 + }, + { + "epoch": 8.477635782747603, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 10614 + }, + { + "epoch": 8.478434504792332, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 10615 + }, + { + "epoch": 8.47923322683706, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 10616 + }, + { + "epoch": 8.48003194888179, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 10617 + }, + { + "epoch": 8.480830670926517, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0414, + "step": 10618 + }, + { + "epoch": 8.481629392971247, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 10619 + }, + { + "epoch": 8.482428115015974, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0377, + "step": 10620 + }, + { + "epoch": 8.483226837060704, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 10621 + }, + { + "epoch": 8.484025559105431, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 10622 + }, + { + "epoch": 8.48482428115016, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 10623 + }, + { + "epoch": 8.485623003194888, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 10624 + }, + { + "epoch": 8.486421725239616, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 10625 + }, + { + "epoch": 8.487220447284345, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 10626 + }, + { + "epoch": 8.488019169329073, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 10627 + }, + { + "epoch": 8.488817891373802, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 10628 + }, + { + "epoch": 8.48961661341853, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 10629 + }, + { + "epoch": 8.49041533546326, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 10630 + }, + { + "epoch": 8.491214057507987, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 10631 + }, + { + "epoch": 8.492012779552716, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 10632 + }, + { + "epoch": 8.492811501597444, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 10633 + }, + { + "epoch": 8.493610223642172, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 10634 + }, + { + "epoch": 8.494408945686901, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 10635 + }, + { + "epoch": 8.495207667731629, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 10636 + }, + { + "epoch": 8.496006389776358, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0337, + "step": 10637 + }, + { + "epoch": 8.496805111821086, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 10638 + }, + { + "epoch": 8.497603833865815, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 10639 + }, + { + "epoch": 8.498402555910543, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0397, + "step": 10640 + }, + { + "epoch": 8.499201277955272, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 10641 + }, + { + "epoch": 8.5, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 10642 + }, + { + "epoch": 8.500798722044728, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 10643 + }, + { + "epoch": 8.501597444089457, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 10644 + }, + { + "epoch": 8.502396166134185, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 10645 + }, + { + "epoch": 8.503194888178914, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 10646 + }, + { + "epoch": 8.503993610223642, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0378, + "step": 10647 + }, + { + "epoch": 8.504792332268371, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 10648 + }, + { + "epoch": 8.505591054313099, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 10649 + }, + { + "epoch": 8.506389776357828, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 10650 + }, + { + "epoch": 8.507188498402556, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 10651 + }, + { + "epoch": 8.507987220447284, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 10652 + }, + { + "epoch": 8.508785942492013, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 10653 + }, + { + "epoch": 8.50958466453674, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 10654 + }, + { + "epoch": 8.51038338658147, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 10655 + }, + { + "epoch": 8.511182108626198, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 10656 + }, + { + "epoch": 8.511980830670927, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 10657 + }, + { + "epoch": 8.512779552715655, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 10658 + }, + { + "epoch": 8.513578274760384, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 10659 + }, + { + "epoch": 8.514376996805112, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 10660 + }, + { + "epoch": 8.51517571884984, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 10661 + }, + { + "epoch": 8.515974440894569, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 10662 + }, + { + "epoch": 8.516773162939296, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 10663 + }, + { + "epoch": 8.517571884984026, + "grad_norm": 0.1826171875, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 10664 + }, + { + "epoch": 8.518370607028753, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 10665 + }, + { + "epoch": 8.519169329073483, + "grad_norm": 0.265625, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 10666 + }, + { + "epoch": 8.51996805111821, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 10667 + }, + { + "epoch": 8.52076677316294, + "grad_norm": 0.236328125, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 10668 + }, + { + "epoch": 8.521565495207668, + "grad_norm": 0.169921875, + "learning_rate": 0.0005, + "loss": 1.036, + "step": 10669 + }, + { + "epoch": 8.522364217252395, + "grad_norm": 0.251953125, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 10670 + }, + { + "epoch": 8.523162939297125, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0377, + "step": 10671 + }, + { + "epoch": 8.523961661341852, + "grad_norm": 0.2001953125, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 10672 + }, + { + "epoch": 8.524760383386582, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 10673 + }, + { + "epoch": 8.52555910543131, + "grad_norm": 0.25390625, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 10674 + }, + { + "epoch": 8.526357827476039, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0353, + "step": 10675 + }, + { + "epoch": 8.527156549520766, + "grad_norm": 0.1982421875, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 10676 + }, + { + "epoch": 8.527955271565496, + "grad_norm": 0.162109375, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 10677 + }, + { + "epoch": 8.528753993610223, + "grad_norm": 0.1865234375, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 10678 + }, + { + "epoch": 8.529552715654953, + "grad_norm": 0.1796875, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 10679 + }, + { + "epoch": 8.53035143769968, + "grad_norm": 0.1689453125, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 10680 + }, + { + "epoch": 8.531150159744408, + "grad_norm": 0.193359375, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 10681 + }, + { + "epoch": 8.531948881789138, + "grad_norm": 0.201171875, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 10682 + }, + { + "epoch": 8.532747603833865, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 10683 + }, + { + "epoch": 8.533546325878595, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 10684 + }, + { + "epoch": 8.534345047923322, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 10685 + }, + { + "epoch": 8.535143769968052, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 10686 + }, + { + "epoch": 8.53594249201278, + "grad_norm": 0.1533203125, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 10687 + }, + { + "epoch": 8.536741214057509, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 10688 + }, + { + "epoch": 8.537539936102236, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 10689 + }, + { + "epoch": 8.538338658146966, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0348, + "step": 10690 + }, + { + "epoch": 8.539137380191693, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0144, + "step": 10691 + }, + { + "epoch": 8.539936102236421, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 10692 + }, + { + "epoch": 8.54073482428115, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 10693 + }, + { + "epoch": 8.541533546325878, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 10694 + }, + { + "epoch": 8.542332268370608, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 10695 + }, + { + "epoch": 8.543130990415335, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 10696 + }, + { + "epoch": 8.543929712460065, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 10697 + }, + { + "epoch": 8.544728434504792, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 10698 + }, + { + "epoch": 8.545527156549522, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 10699 + }, + { + "epoch": 8.54632587859425, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 10700 + }, + { + "epoch": 8.547124600638977, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 10701 + }, + { + "epoch": 8.547923322683706, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 10702 + }, + { + "epoch": 8.548722044728434, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0168, + "step": 10703 + }, + { + "epoch": 8.549520766773163, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 10704 + }, + { + "epoch": 8.550319488817891, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 10705 + }, + { + "epoch": 8.55111821086262, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 10706 + }, + { + "epoch": 8.551916932907348, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 10707 + }, + { + "epoch": 8.552715654952078, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 10708 + }, + { + "epoch": 8.553514376996805, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0397, + "step": 10709 + }, + { + "epoch": 8.554313099041533, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0388, + "step": 10710 + }, + { + "epoch": 8.555111821086262, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0391, + "step": 10711 + }, + { + "epoch": 8.55591054313099, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0386, + "step": 10712 + }, + { + "epoch": 8.55670926517572, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 10713 + }, + { + "epoch": 8.557507987220447, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 10714 + }, + { + "epoch": 8.558306709265176, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 10715 + }, + { + "epoch": 8.559105431309904, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 10716 + }, + { + "epoch": 8.559904153354633, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 10717 + }, + { + "epoch": 8.560702875399361, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 10718 + }, + { + "epoch": 8.561501597444089, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 10719 + }, + { + "epoch": 8.562300319488818, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 10720 + }, + { + "epoch": 8.563099041533546, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 10721 + }, + { + "epoch": 8.563897763578275, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 10722 + }, + { + "epoch": 8.564696485623003, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0422, + "step": 10723 + }, + { + "epoch": 8.565495207667732, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 10724 + }, + { + "epoch": 8.56629392971246, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 10725 + }, + { + "epoch": 8.56709265175719, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.034, + "step": 10726 + }, + { + "epoch": 8.567891373801917, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 10727 + }, + { + "epoch": 8.568690095846645, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 10728 + }, + { + "epoch": 8.569488817891374, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 10729 + }, + { + "epoch": 8.570287539936102, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0401, + "step": 10730 + }, + { + "epoch": 8.571086261980831, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.034, + "step": 10731 + }, + { + "epoch": 8.571884984025559, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 10732 + }, + { + "epoch": 8.572683706070288, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 10733 + }, + { + "epoch": 8.573482428115016, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 10734 + }, + { + "epoch": 8.574281150159745, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 10735 + }, + { + "epoch": 8.575079872204473, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 10736 + }, + { + "epoch": 8.5758785942492, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0192, + "step": 10737 + }, + { + "epoch": 8.57667731629393, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.038, + "step": 10738 + }, + { + "epoch": 8.577476038338657, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 10739 + }, + { + "epoch": 8.578274760383387, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0389, + "step": 10740 + }, + { + "epoch": 8.579073482428115, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 10741 + }, + { + "epoch": 8.579872204472844, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 10742 + }, + { + "epoch": 8.580670926517572, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 10743 + }, + { + "epoch": 8.581469648562301, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 10744 + }, + { + "epoch": 8.582268370607029, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 10745 + }, + { + "epoch": 8.583067092651756, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0177, + "step": 10746 + }, + { + "epoch": 8.583865814696486, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 10747 + }, + { + "epoch": 8.584664536741213, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 10748 + }, + { + "epoch": 8.585463258785943, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0353, + "step": 10749 + }, + { + "epoch": 8.58626198083067, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 10750 + }, + { + "epoch": 8.5870607028754, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 10751 + }, + { + "epoch": 8.587859424920127, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 10752 + }, + { + "epoch": 8.588658146964857, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 10753 + }, + { + "epoch": 8.589456869009584, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 10754 + }, + { + "epoch": 8.590255591054314, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 10755 + }, + { + "epoch": 8.591054313099042, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 10756 + }, + { + "epoch": 8.59185303514377, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 10757 + }, + { + "epoch": 8.592651757188499, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 10758 + }, + { + "epoch": 8.593450479233226, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 10759 + }, + { + "epoch": 8.594249201277956, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 10760 + }, + { + "epoch": 8.595047923322683, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 10761 + }, + { + "epoch": 8.595846645367413, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 10762 + }, + { + "epoch": 8.59664536741214, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 10763 + }, + { + "epoch": 8.59744408945687, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 10764 + }, + { + "epoch": 8.598242811501597, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 10765 + }, + { + "epoch": 8.599041533546325, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 10766 + }, + { + "epoch": 8.599840255591054, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 10767 + }, + { + "epoch": 8.600638977635782, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 10768 + }, + { + "epoch": 8.601437699680512, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 10769 + }, + { + "epoch": 8.60223642172524, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 10770 + }, + { + "epoch": 8.603035143769969, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 10771 + }, + { + "epoch": 8.603833865814696, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0181, + "step": 10772 + }, + { + "epoch": 8.604632587859426, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 10773 + }, + { + "epoch": 8.605431309904153, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 10774 + }, + { + "epoch": 8.606230031948883, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 10775 + }, + { + "epoch": 8.60702875399361, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 10776 + }, + { + "epoch": 8.607827476038338, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 10777 + }, + { + "epoch": 8.608626198083067, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 10778 + }, + { + "epoch": 8.609424920127795, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 10779 + }, + { + "epoch": 8.610223642172524, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 10780 + }, + { + "epoch": 8.611022364217252, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 10781 + }, + { + "epoch": 8.611821086261982, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 10782 + }, + { + "epoch": 8.61261980830671, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 10783 + }, + { + "epoch": 8.613418530351439, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 10784 + }, + { + "epoch": 8.614217252396166, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0417, + "step": 10785 + }, + { + "epoch": 8.615015974440894, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 10786 + }, + { + "epoch": 8.615814696485623, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 10787 + }, + { + "epoch": 8.616613418530351, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 10788 + }, + { + "epoch": 8.61741214057508, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 10789 + }, + { + "epoch": 8.618210862619808, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 10790 + }, + { + "epoch": 8.619009584664537, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 10791 + }, + { + "epoch": 8.619808306709265, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 10792 + }, + { + "epoch": 8.620607028753994, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0405, + "step": 10793 + }, + { + "epoch": 8.621405750798722, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 10794 + }, + { + "epoch": 8.62220447284345, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 10795 + }, + { + "epoch": 8.62300319488818, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 10796 + }, + { + "epoch": 8.623801916932907, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 10797 + }, + { + "epoch": 8.624600638977636, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 10798 + }, + { + "epoch": 8.625399361022364, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 10799 + }, + { + "epoch": 8.626198083067093, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 10800 + }, + { + "epoch": 8.62699680511182, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 10801 + }, + { + "epoch": 8.62779552715655, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0424, + "step": 10802 + }, + { + "epoch": 8.628594249201278, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0426, + "step": 10803 + }, + { + "epoch": 8.629392971246006, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 10804 + }, + { + "epoch": 8.630191693290735, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0405, + "step": 10805 + }, + { + "epoch": 8.630990415335463, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 10806 + }, + { + "epoch": 8.631789137380192, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 10807 + }, + { + "epoch": 8.63258785942492, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 10808 + }, + { + "epoch": 8.633386581469649, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 10809 + }, + { + "epoch": 8.634185303514377, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 10810 + }, + { + "epoch": 8.634984025559106, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 10811 + }, + { + "epoch": 8.635782747603834, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 10812 + }, + { + "epoch": 8.636581469648561, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 10813 + }, + { + "epoch": 8.63738019169329, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 10814 + }, + { + "epoch": 8.638178913738018, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 10815 + }, + { + "epoch": 8.638977635782748, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 10816 + }, + { + "epoch": 8.639776357827476, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 10817 + }, + { + "epoch": 8.640575079872205, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 10818 + }, + { + "epoch": 8.641373801916933, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 10819 + }, + { + "epoch": 8.642172523961662, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 10820 + }, + { + "epoch": 8.64297124600639, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 10821 + }, + { + "epoch": 8.643769968051117, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 10822 + }, + { + "epoch": 8.644568690095847, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 10823 + }, + { + "epoch": 8.645367412140574, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 10824 + }, + { + "epoch": 8.646166134185304, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 10825 + }, + { + "epoch": 8.646964856230031, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0177, + "step": 10826 + }, + { + "epoch": 8.64776357827476, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 10827 + }, + { + "epoch": 8.648562300319488, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 10828 + }, + { + "epoch": 8.649361022364218, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 10829 + }, + { + "epoch": 8.650159744408946, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 10830 + }, + { + "epoch": 8.650958466453675, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 10831 + }, + { + "epoch": 8.651757188498403, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 10832 + }, + { + "epoch": 8.65255591054313, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 10833 + }, + { + "epoch": 8.65335463258786, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 10834 + }, + { + "epoch": 8.654153354632587, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 10835 + }, + { + "epoch": 8.654952076677317, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 10836 + }, + { + "epoch": 8.655750798722044, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 10837 + }, + { + "epoch": 8.656549520766774, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 10838 + }, + { + "epoch": 8.657348242811501, + "grad_norm": 0.181640625, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 10839 + }, + { + "epoch": 8.65814696485623, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 10840 + }, + { + "epoch": 8.658945686900958, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 10841 + }, + { + "epoch": 8.659744408945686, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0168, + "step": 10842 + }, + { + "epoch": 8.660543130990416, + "grad_norm": 0.240234375, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 10843 + }, + { + "epoch": 8.661341853035143, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 10844 + }, + { + "epoch": 8.662140575079873, + "grad_norm": 0.26953125, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 10845 + }, + { + "epoch": 8.6629392971246, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 10846 + }, + { + "epoch": 8.66373801916933, + "grad_norm": 0.279296875, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 10847 + }, + { + "epoch": 8.664536741214057, + "grad_norm": 0.193359375, + "learning_rate": 0.0005, + "loss": 1.0391, + "step": 10848 + }, + { + "epoch": 8.665335463258787, + "grad_norm": 0.330078125, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 10849 + }, + { + "epoch": 8.666134185303514, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 10850 + }, + { + "epoch": 8.666932907348244, + "grad_norm": 0.271484375, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 10851 + }, + { + "epoch": 8.667731629392971, + "grad_norm": 0.22265625, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 10852 + }, + { + "epoch": 8.668530351437699, + "grad_norm": 0.283203125, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 10853 + }, + { + "epoch": 8.669329073482428, + "grad_norm": 0.2490234375, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 10854 + }, + { + "epoch": 8.670127795527156, + "grad_norm": 0.30078125, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 10855 + }, + { + "epoch": 8.670926517571885, + "grad_norm": 0.265625, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 10856 + }, + { + "epoch": 8.671725239616613, + "grad_norm": 0.3515625, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 10857 + }, + { + "epoch": 8.672523961661343, + "grad_norm": 0.30859375, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 10858 + }, + { + "epoch": 8.67332268370607, + "grad_norm": 0.625, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 10859 + }, + { + "epoch": 8.6741214057508, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 10860 + }, + { + "epoch": 8.674920127795527, + "grad_norm": 0.60546875, + "learning_rate": 0.0005, + "loss": 1.0376, + "step": 10861 + }, + { + "epoch": 8.675718849840255, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 10862 + }, + { + "epoch": 8.676517571884984, + "grad_norm": 0.62890625, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 10863 + }, + { + "epoch": 8.677316293929712, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 10864 + }, + { + "epoch": 8.678115015974441, + "grad_norm": 0.41796875, + "learning_rate": 0.0005, + "loss": 1.0376, + "step": 10865 + }, + { + "epoch": 8.678913738019169, + "grad_norm": 0.162109375, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 10866 + }, + { + "epoch": 8.679712460063898, + "grad_norm": 0.42578125, + "learning_rate": 0.0005, + "loss": 1.04, + "step": 10867 + }, + { + "epoch": 8.680511182108626, + "grad_norm": 0.1611328125, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 10868 + }, + { + "epoch": 8.681309904153355, + "grad_norm": 0.310546875, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 10869 + }, + { + "epoch": 8.682108626198083, + "grad_norm": 0.302734375, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 10870 + }, + { + "epoch": 8.68290734824281, + "grad_norm": 0.4921875, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 10871 + }, + { + "epoch": 8.68370607028754, + "grad_norm": 0.201171875, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 10872 + }, + { + "epoch": 8.684504792332268, + "grad_norm": 0.4609375, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 10873 + }, + { + "epoch": 8.685303514376997, + "grad_norm": 0.31640625, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 10874 + }, + { + "epoch": 8.686102236421725, + "grad_norm": 0.7421875, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 10875 + }, + { + "epoch": 8.686900958466454, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0179, + "step": 10876 + }, + { + "epoch": 8.687699680511182, + "grad_norm": 0.5546875, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 10877 + }, + { + "epoch": 8.688498402555911, + "grad_norm": 0.18359375, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 10878 + }, + { + "epoch": 8.689297124600639, + "grad_norm": 0.640625, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 10879 + }, + { + "epoch": 8.690095846645367, + "grad_norm": 0.16796875, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 10880 + }, + { + "epoch": 8.690894568690096, + "grad_norm": 0.7421875, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 10881 + }, + { + "epoch": 8.691693290734824, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0408, + "step": 10882 + }, + { + "epoch": 8.692492012779553, + "grad_norm": 0.5703125, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 10883 + }, + { + "epoch": 8.69329073482428, + "grad_norm": 0.181640625, + "learning_rate": 0.0005, + "loss": 1.0408, + "step": 10884 + }, + { + "epoch": 8.69408945686901, + "grad_norm": 0.51953125, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 10885 + }, + { + "epoch": 8.694888178913738, + "grad_norm": 0.2177734375, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 10886 + }, + { + "epoch": 8.695686900958467, + "grad_norm": 0.50390625, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 10887 + }, + { + "epoch": 8.696485623003195, + "grad_norm": 0.291015625, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 10888 + }, + { + "epoch": 8.697284345047922, + "grad_norm": 0.609375, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 10889 + }, + { + "epoch": 8.698083067092652, + "grad_norm": 0.26171875, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 10890 + }, + { + "epoch": 8.69888178913738, + "grad_norm": 0.55859375, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 10891 + }, + { + "epoch": 8.699680511182109, + "grad_norm": 0.34375, + "learning_rate": 0.0005, + "loss": 1.0393, + "step": 10892 + }, + { + "epoch": 8.700479233226837, + "grad_norm": 0.5234375, + "learning_rate": 0.0005, + "loss": 1.0373, + "step": 10893 + }, + { + "epoch": 8.701277955271566, + "grad_norm": 0.38671875, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 10894 + }, + { + "epoch": 8.702076677316294, + "grad_norm": 0.515625, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 10895 + }, + { + "epoch": 8.702875399361023, + "grad_norm": 0.248046875, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 10896 + }, + { + "epoch": 8.70367412140575, + "grad_norm": 0.298828125, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 10897 + }, + { + "epoch": 8.704472843450478, + "grad_norm": 0.2490234375, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 10898 + }, + { + "epoch": 8.705271565495208, + "grad_norm": 0.3203125, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 10899 + }, + { + "epoch": 8.706070287539935, + "grad_norm": 0.1611328125, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 10900 + }, + { + "epoch": 8.706869009584665, + "grad_norm": 0.2177734375, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 10901 + }, + { + "epoch": 8.707667731629392, + "grad_norm": 0.2080078125, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 10902 + }, + { + "epoch": 8.708466453674122, + "grad_norm": 0.2197265625, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 10903 + }, + { + "epoch": 8.70926517571885, + "grad_norm": 0.203125, + "learning_rate": 0.0005, + "loss": 1.0405, + "step": 10904 + }, + { + "epoch": 8.710063897763579, + "grad_norm": 0.24609375, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 10905 + }, + { + "epoch": 8.710862619808307, + "grad_norm": 0.2265625, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 10906 + }, + { + "epoch": 8.711661341853034, + "grad_norm": 0.17578125, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 10907 + }, + { + "epoch": 8.712460063897764, + "grad_norm": 0.33203125, + "learning_rate": 0.0005, + "loss": 1.0395, + "step": 10908 + }, + { + "epoch": 8.713258785942491, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 10909 + }, + { + "epoch": 8.71405750798722, + "grad_norm": 0.30078125, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 10910 + }, + { + "epoch": 8.714856230031948, + "grad_norm": 0.1826171875, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 10911 + }, + { + "epoch": 8.715654952076678, + "grad_norm": 0.2392578125, + "learning_rate": 0.0005, + "loss": 1.0379, + "step": 10912 + }, + { + "epoch": 8.716453674121405, + "grad_norm": 0.232421875, + "learning_rate": 0.0005, + "loss": 1.036, + "step": 10913 + }, + { + "epoch": 8.717252396166135, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 10914 + }, + { + "epoch": 8.718051118210862, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 10915 + }, + { + "epoch": 8.718849840255592, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 10916 + }, + { + "epoch": 8.71964856230032, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 10917 + }, + { + "epoch": 8.720447284345047, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 10918 + }, + { + "epoch": 8.721246006389777, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 10919 + }, + { + "epoch": 8.722044728434504, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 10920 + }, + { + "epoch": 8.722843450479234, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 10921 + }, + { + "epoch": 8.723642172523961, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 10922 + }, + { + "epoch": 8.72444089456869, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 10923 + }, + { + "epoch": 8.725239616613418, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 10924 + }, + { + "epoch": 8.726038338658148, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0176, + "step": 10925 + }, + { + "epoch": 8.726837060702875, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 10926 + }, + { + "epoch": 8.727635782747605, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 10927 + }, + { + "epoch": 8.728434504792332, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 10928 + }, + { + "epoch": 8.72923322683706, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 10929 + }, + { + "epoch": 8.73003194888179, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 10930 + }, + { + "epoch": 8.730830670926517, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 10931 + }, + { + "epoch": 8.731629392971247, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 10932 + }, + { + "epoch": 8.732428115015974, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 10933 + }, + { + "epoch": 8.733226837060704, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 10934 + }, + { + "epoch": 8.734025559105431, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 10935 + }, + { + "epoch": 8.73482428115016, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 10936 + }, + { + "epoch": 8.735623003194888, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 10937 + }, + { + "epoch": 8.736421725239616, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 10938 + }, + { + "epoch": 8.737220447284345, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 10939 + }, + { + "epoch": 8.738019169329073, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 10940 + }, + { + "epoch": 8.738817891373802, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 10941 + }, + { + "epoch": 8.73961661341853, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 10942 + }, + { + "epoch": 8.74041533546326, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 10943 + }, + { + "epoch": 8.741214057507987, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 10944 + }, + { + "epoch": 8.742012779552716, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0386, + "step": 10945 + }, + { + "epoch": 8.742811501597444, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 10946 + }, + { + "epoch": 8.743610223642172, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 10947 + }, + { + "epoch": 8.744408945686901, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 10948 + }, + { + "epoch": 8.745207667731629, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0376, + "step": 10949 + }, + { + "epoch": 8.746006389776358, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0354, + "step": 10950 + }, + { + "epoch": 8.746805111821086, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 10951 + }, + { + "epoch": 8.747603833865815, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0157, + "step": 10952 + }, + { + "epoch": 8.748402555910543, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 10953 + }, + { + "epoch": 8.749201277955272, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 10954 + }, + { + "epoch": 8.75, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 10955 + }, + { + "epoch": 8.750798722044728, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 10956 + }, + { + "epoch": 8.751597444089457, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 10957 + }, + { + "epoch": 8.752396166134185, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 10958 + }, + { + "epoch": 8.753194888178914, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 10959 + }, + { + "epoch": 8.753993610223642, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 10960 + }, + { + "epoch": 8.754792332268371, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 10961 + }, + { + "epoch": 8.755591054313099, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 10962 + }, + { + "epoch": 8.756389776357828, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 10963 + }, + { + "epoch": 8.757188498402556, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 10964 + }, + { + "epoch": 8.757987220447284, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 10965 + }, + { + "epoch": 8.758785942492013, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 10966 + }, + { + "epoch": 8.75958466453674, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 10967 + }, + { + "epoch": 8.76038338658147, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 10968 + }, + { + "epoch": 8.761182108626198, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 10969 + }, + { + "epoch": 8.761980830670927, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 10970 + }, + { + "epoch": 8.762779552715655, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 10971 + }, + { + "epoch": 8.763578274760384, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 10972 + }, + { + "epoch": 8.764376996805112, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 10973 + }, + { + "epoch": 8.76517571884984, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 10974 + }, + { + "epoch": 8.765974440894569, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 10975 + }, + { + "epoch": 8.766773162939296, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 10976 + }, + { + "epoch": 8.767571884984026, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 10977 + }, + { + "epoch": 8.768370607028753, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0372, + "step": 10978 + }, + { + "epoch": 8.769169329073483, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 10979 + }, + { + "epoch": 8.76996805111821, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 10980 + }, + { + "epoch": 8.77076677316294, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 10981 + }, + { + "epoch": 8.771565495207668, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 10982 + }, + { + "epoch": 8.772364217252395, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 10983 + }, + { + "epoch": 8.773162939297125, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 10984 + }, + { + "epoch": 8.773961661341852, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 10985 + }, + { + "epoch": 8.774760383386582, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 10986 + }, + { + "epoch": 8.77555910543131, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 10987 + }, + { + "epoch": 8.776357827476039, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 10988 + }, + { + "epoch": 8.777156549520766, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 10989 + }, + { + "epoch": 8.777955271565496, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 10990 + }, + { + "epoch": 8.778753993610223, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0195, + "step": 10991 + }, + { + "epoch": 8.779552715654953, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 10992 + }, + { + "epoch": 8.78035143769968, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 10993 + }, + { + "epoch": 8.781150159744408, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 10994 + }, + { + "epoch": 8.781948881789138, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 10995 + }, + { + "epoch": 8.782747603833865, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 10996 + }, + { + "epoch": 8.783546325878595, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 10997 + }, + { + "epoch": 8.784345047923322, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 10998 + }, + { + "epoch": 8.785143769968052, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 10999 + }, + { + "epoch": 8.78594249201278, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 11000 + }, + { + "epoch": 8.786741214057509, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 11001 + }, + { + "epoch": 8.787539936102236, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 11002 + }, + { + "epoch": 8.788338658146966, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 11003 + }, + { + "epoch": 8.789137380191693, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 11004 + }, + { + "epoch": 8.789936102236421, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 11005 + }, + { + "epoch": 8.79073482428115, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 11006 + }, + { + "epoch": 8.791533546325878, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 11007 + }, + { + "epoch": 8.792332268370608, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 11008 + }, + { + "epoch": 8.793130990415335, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 11009 + }, + { + "epoch": 8.793929712460065, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 11010 + }, + { + "epoch": 8.794728434504792, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.034, + "step": 11011 + }, + { + "epoch": 8.795527156549522, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 11012 + }, + { + "epoch": 8.79632587859425, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 11013 + }, + { + "epoch": 8.797124600638977, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 11014 + }, + { + "epoch": 8.797923322683706, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 11015 + }, + { + "epoch": 8.798722044728434, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 11016 + }, + { + "epoch": 8.799520766773163, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 11017 + }, + { + "epoch": 8.800319488817891, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 11018 + }, + { + "epoch": 8.80111821086262, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 11019 + }, + { + "epoch": 8.801916932907348, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 11020 + }, + { + "epoch": 8.802715654952078, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 11021 + }, + { + "epoch": 8.803514376996805, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 11022 + }, + { + "epoch": 8.804313099041533, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 11023 + }, + { + "epoch": 8.805111821086262, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 11024 + }, + { + "epoch": 8.80591054313099, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 11025 + }, + { + "epoch": 8.80670926517572, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0348, + "step": 11026 + }, + { + "epoch": 8.807507987220447, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 11027 + }, + { + "epoch": 8.808306709265176, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 11028 + }, + { + "epoch": 8.809105431309904, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 11029 + }, + { + "epoch": 8.809904153354633, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 11030 + }, + { + "epoch": 8.810702875399361, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 11031 + }, + { + "epoch": 8.811501597444089, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 11032 + }, + { + "epoch": 8.812300319488818, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 11033 + }, + { + "epoch": 8.813099041533546, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 11034 + }, + { + "epoch": 8.813897763578275, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 11035 + }, + { + "epoch": 8.814696485623003, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 11036 + }, + { + "epoch": 8.815495207667732, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 11037 + }, + { + "epoch": 8.81629392971246, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0411, + "step": 11038 + }, + { + "epoch": 8.81709265175719, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 11039 + }, + { + "epoch": 8.817891373801917, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 11040 + }, + { + "epoch": 8.818690095846645, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 11041 + }, + { + "epoch": 8.819488817891374, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 11042 + }, + { + "epoch": 8.820287539936102, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 11043 + }, + { + "epoch": 8.821086261980831, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 11044 + }, + { + "epoch": 8.821884984025559, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 11045 + }, + { + "epoch": 8.822683706070288, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 11046 + }, + { + "epoch": 8.823482428115016, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 11047 + }, + { + "epoch": 8.824281150159745, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 11048 + }, + { + "epoch": 8.825079872204473, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 11049 + }, + { + "epoch": 8.8258785942492, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 11050 + }, + { + "epoch": 8.82667731629393, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 11051 + }, + { + "epoch": 8.827476038338657, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 11052 + }, + { + "epoch": 8.828274760383387, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 11053 + }, + { + "epoch": 8.829073482428115, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 11054 + }, + { + "epoch": 8.829872204472844, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 11055 + }, + { + "epoch": 8.830670926517572, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 11056 + }, + { + "epoch": 8.831469648562301, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 11057 + }, + { + "epoch": 8.832268370607029, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 11058 + }, + { + "epoch": 8.833067092651756, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 11059 + }, + { + "epoch": 8.833865814696486, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 11060 + }, + { + "epoch": 8.834664536741213, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 11061 + }, + { + "epoch": 8.835463258785943, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 11062 + }, + { + "epoch": 8.83626198083067, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 11063 + }, + { + "epoch": 8.8370607028754, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 11064 + }, + { + "epoch": 8.837859424920127, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 11065 + }, + { + "epoch": 8.838658146964857, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 11066 + }, + { + "epoch": 8.839456869009584, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 11067 + }, + { + "epoch": 8.840255591054314, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 11068 + }, + { + "epoch": 8.841054313099042, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 11069 + }, + { + "epoch": 8.84185303514377, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 11070 + }, + { + "epoch": 8.842651757188499, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 11071 + }, + { + "epoch": 8.843450479233226, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 11072 + }, + { + "epoch": 8.844249201277956, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 11073 + }, + { + "epoch": 8.845047923322683, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 11074 + }, + { + "epoch": 8.845846645367413, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 11075 + }, + { + "epoch": 8.84664536741214, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 11076 + }, + { + "epoch": 8.84744408945687, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 11077 + }, + { + "epoch": 8.848242811501597, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 11078 + }, + { + "epoch": 8.849041533546325, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 11079 + }, + { + "epoch": 8.849840255591054, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 11080 + }, + { + "epoch": 8.850638977635782, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0372, + "step": 11081 + }, + { + "epoch": 8.851437699680512, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 11082 + }, + { + "epoch": 8.85223642172524, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 11083 + }, + { + "epoch": 8.853035143769969, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 11084 + }, + { + "epoch": 8.853833865814696, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 11085 + }, + { + "epoch": 8.854632587859426, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 11086 + }, + { + "epoch": 8.855431309904153, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 11087 + }, + { + "epoch": 8.856230031948883, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 11088 + }, + { + "epoch": 8.85702875399361, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 11089 + }, + { + "epoch": 8.857827476038338, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 11090 + }, + { + "epoch": 8.858626198083067, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 11091 + }, + { + "epoch": 8.859424920127795, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0388, + "step": 11092 + }, + { + "epoch": 8.860223642172524, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0373, + "step": 11093 + }, + { + "epoch": 8.861022364217252, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0393, + "step": 11094 + }, + { + "epoch": 8.861821086261982, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0391, + "step": 11095 + }, + { + "epoch": 8.86261980830671, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 11096 + }, + { + "epoch": 8.863418530351439, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 11097 + }, + { + "epoch": 8.864217252396166, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 11098 + }, + { + "epoch": 8.865015974440894, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0348, + "step": 11099 + }, + { + "epoch": 8.865814696485623, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 11100 + }, + { + "epoch": 8.866613418530351, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 11101 + }, + { + "epoch": 8.86741214057508, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 11102 + }, + { + "epoch": 8.868210862619808, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 11103 + }, + { + "epoch": 8.869009584664537, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 11104 + }, + { + "epoch": 8.869808306709265, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 11105 + }, + { + "epoch": 8.870607028753994, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 11106 + }, + { + "epoch": 8.871405750798722, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 11107 + }, + { + "epoch": 8.87220447284345, + "grad_norm": 0.162109375, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 11108 + }, + { + "epoch": 8.87300319488818, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 11109 + }, + { + "epoch": 8.873801916932907, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 11110 + }, + { + "epoch": 8.874600638977636, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 11111 + }, + { + "epoch": 8.875399361022364, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 11112 + }, + { + "epoch": 8.876198083067093, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 11113 + }, + { + "epoch": 8.87699680511182, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 11114 + }, + { + "epoch": 8.87779552715655, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 11115 + }, + { + "epoch": 8.878594249201278, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 11116 + }, + { + "epoch": 8.879392971246006, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.034, + "step": 11117 + }, + { + "epoch": 8.880191693290735, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 11118 + }, + { + "epoch": 8.880990415335463, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 11119 + }, + { + "epoch": 8.881789137380192, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 11120 + }, + { + "epoch": 8.88258785942492, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 11121 + }, + { + "epoch": 8.883386581469649, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 11122 + }, + { + "epoch": 8.884185303514377, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 11123 + }, + { + "epoch": 8.884984025559106, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 11124 + }, + { + "epoch": 8.885782747603834, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0398, + "step": 11125 + }, + { + "epoch": 8.886581469648561, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 11126 + }, + { + "epoch": 8.88738019169329, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 11127 + }, + { + "epoch": 8.888178913738018, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 11128 + }, + { + "epoch": 8.888977635782748, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 11129 + }, + { + "epoch": 8.889776357827476, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.015, + "step": 11130 + }, + { + "epoch": 8.890575079872205, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 11131 + }, + { + "epoch": 8.891373801916933, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 11132 + }, + { + "epoch": 8.892172523961662, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 11133 + }, + { + "epoch": 8.89297124600639, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 11134 + }, + { + "epoch": 8.893769968051117, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 11135 + }, + { + "epoch": 8.894568690095847, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 11136 + }, + { + "epoch": 8.895367412140574, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 11137 + }, + { + "epoch": 8.896166134185304, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 11138 + }, + { + "epoch": 8.896964856230031, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 11139 + }, + { + "epoch": 8.89776357827476, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 11140 + }, + { + "epoch": 8.898562300319488, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 11141 + }, + { + "epoch": 8.899361022364218, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 11142 + }, + { + "epoch": 8.900159744408946, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 11143 + }, + { + "epoch": 8.900958466453675, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 11144 + }, + { + "epoch": 8.901757188498403, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 11145 + }, + { + "epoch": 8.90255591054313, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 11146 + }, + { + "epoch": 8.90335463258786, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 11147 + }, + { + "epoch": 8.904153354632587, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 11148 + }, + { + "epoch": 8.904952076677317, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 11149 + }, + { + "epoch": 8.905750798722044, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 11150 + }, + { + "epoch": 8.906549520766774, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0145, + "step": 11151 + }, + { + "epoch": 8.907348242811501, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 11152 + }, + { + "epoch": 8.90814696485623, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 11153 + }, + { + "epoch": 8.908945686900958, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0437, + "step": 11154 + }, + { + "epoch": 8.909744408945686, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 11155 + }, + { + "epoch": 8.910543130990416, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 11156 + }, + { + "epoch": 8.911341853035143, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0378, + "step": 11157 + }, + { + "epoch": 8.912140575079873, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 11158 + }, + { + "epoch": 8.9129392971246, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 11159 + }, + { + "epoch": 8.91373801916933, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0348, + "step": 11160 + }, + { + "epoch": 8.914536741214057, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 11161 + }, + { + "epoch": 8.915335463258787, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 11162 + }, + { + "epoch": 8.916134185303514, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 11163 + }, + { + "epoch": 8.916932907348244, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 11164 + }, + { + "epoch": 8.917731629392971, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 11165 + }, + { + "epoch": 8.918530351437699, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 11166 + }, + { + "epoch": 8.919329073482428, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0153, + "step": 11167 + }, + { + "epoch": 8.920127795527156, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 11168 + }, + { + "epoch": 8.920926517571885, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 11169 + }, + { + "epoch": 8.921725239616613, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 11170 + }, + { + "epoch": 8.922523961661343, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 11171 + }, + { + "epoch": 8.92332268370607, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 11172 + }, + { + "epoch": 8.9241214057508, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 11173 + }, + { + "epoch": 8.924920127795527, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 11174 + }, + { + "epoch": 8.925718849840255, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 11175 + }, + { + "epoch": 8.926517571884984, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 11176 + }, + { + "epoch": 8.927316293929712, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 11177 + }, + { + "epoch": 8.928115015974441, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0144, + "step": 11178 + }, + { + "epoch": 8.928913738019169, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 11179 + }, + { + "epoch": 8.929712460063898, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 11180 + }, + { + "epoch": 8.930511182108626, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 11181 + }, + { + "epoch": 8.931309904153355, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 11182 + }, + { + "epoch": 8.932108626198083, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 11183 + }, + { + "epoch": 8.93290734824281, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0094, + "step": 11184 + }, + { + "epoch": 8.93370607028754, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 11185 + }, + { + "epoch": 8.934504792332268, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 11186 + }, + { + "epoch": 8.935303514376997, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 11187 + }, + { + "epoch": 8.936102236421725, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 11188 + }, + { + "epoch": 8.936900958466454, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 11189 + }, + { + "epoch": 8.937699680511182, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 11190 + }, + { + "epoch": 8.938498402555911, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 11191 + }, + { + "epoch": 8.939297124600639, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 11192 + }, + { + "epoch": 8.940095846645367, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0379, + "step": 11193 + }, + { + "epoch": 8.940894568690096, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 11194 + }, + { + "epoch": 8.941693290734824, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 11195 + }, + { + "epoch": 8.942492012779553, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 11196 + }, + { + "epoch": 8.94329073482428, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0376, + "step": 11197 + }, + { + "epoch": 8.94408945686901, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 11198 + }, + { + "epoch": 8.944888178913738, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 11199 + }, + { + "epoch": 8.945686900958467, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 11200 + }, + { + "epoch": 8.946485623003195, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 11201 + }, + { + "epoch": 8.947284345047922, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 11202 + }, + { + "epoch": 8.948083067092652, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 11203 + }, + { + "epoch": 8.94888178913738, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 11204 + }, + { + "epoch": 8.949680511182109, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 11205 + }, + { + "epoch": 8.950479233226837, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 11206 + }, + { + "epoch": 8.951277955271566, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0353, + "step": 11207 + }, + { + "epoch": 8.952076677316294, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 11208 + }, + { + "epoch": 8.952875399361023, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 11209 + }, + { + "epoch": 8.95367412140575, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 11210 + }, + { + "epoch": 8.954472843450478, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0389, + "step": 11211 + }, + { + "epoch": 8.955271565495208, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 11212 + }, + { + "epoch": 8.956070287539935, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0421, + "step": 11213 + }, + { + "epoch": 8.956869009584665, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.036, + "step": 11214 + }, + { + "epoch": 8.957667731629392, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 11215 + }, + { + "epoch": 8.958466453674122, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 11216 + }, + { + "epoch": 8.95926517571885, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 11217 + }, + { + "epoch": 8.960063897763579, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 11218 + }, + { + "epoch": 8.960862619808307, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0414, + "step": 11219 + }, + { + "epoch": 8.961661341853034, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 11220 + }, + { + "epoch": 8.962460063897764, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0362, + "step": 11221 + }, + { + "epoch": 8.963258785942491, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 11222 + }, + { + "epoch": 8.96405750798722, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 11223 + }, + { + "epoch": 8.964856230031948, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 11224 + }, + { + "epoch": 8.965654952076678, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 11225 + }, + { + "epoch": 8.966453674121405, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 11226 + }, + { + "epoch": 8.967252396166135, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 11227 + }, + { + "epoch": 8.968051118210862, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 11228 + }, + { + "epoch": 8.968849840255592, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 11229 + }, + { + "epoch": 8.96964856230032, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 11230 + }, + { + "epoch": 8.970447284345047, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 11231 + }, + { + "epoch": 8.971246006389777, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 11232 + }, + { + "epoch": 8.972044728434504, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 11233 + }, + { + "epoch": 8.972843450479234, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 11234 + }, + { + "epoch": 8.973642172523961, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 11235 + }, + { + "epoch": 8.97444089456869, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 11236 + }, + { + "epoch": 8.975239616613418, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 11237 + }, + { + "epoch": 8.976038338658148, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 11238 + }, + { + "epoch": 8.976837060702875, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 11239 + }, + { + "epoch": 8.977635782747605, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 11240 + }, + { + "epoch": 8.978434504792332, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0337, + "step": 11241 + }, + { + "epoch": 8.97923322683706, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 11242 + }, + { + "epoch": 8.98003194888179, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 11243 + }, + { + "epoch": 8.980830670926517, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 11244 + }, + { + "epoch": 8.981629392971247, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 11245 + }, + { + "epoch": 8.982428115015974, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 11246 + }, + { + "epoch": 8.983226837060704, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 11247 + }, + { + "epoch": 8.984025559105431, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 11248 + }, + { + "epoch": 8.98482428115016, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 11249 + }, + { + "epoch": 8.985623003194888, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.034, + "step": 11250 + }, + { + "epoch": 8.986421725239616, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 11251 + }, + { + "epoch": 8.987220447284345, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 11252 + }, + { + "epoch": 8.988019169329073, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 11253 + }, + { + "epoch": 8.988817891373802, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 11254 + }, + { + "epoch": 8.98961661341853, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 11255 + }, + { + "epoch": 8.99041533546326, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 11256 + }, + { + "epoch": 8.991214057507987, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 11257 + }, + { + "epoch": 8.992012779552716, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 11258 + }, + { + "epoch": 8.992811501597444, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 11259 + }, + { + "epoch": 8.993610223642172, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 11260 + }, + { + "epoch": 8.994408945686901, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 11261 + }, + { + "epoch": 8.995207667731629, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.017, + "step": 11262 + }, + { + "epoch": 8.996006389776358, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 11263 + }, + { + "epoch": 8.996805111821086, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 11264 + }, + { + "epoch": 8.997603833865815, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 11265 + }, + { + "epoch": 8.998402555910543, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 11266 + }, + { + "epoch": 8.999201277955272, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 11267 + }, + { + "epoch": 9.0, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 11268 + }, + { + "epoch": 9.000798722044728, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 11269 + }, + { + "epoch": 9.001597444089457, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 11270 + }, + { + "epoch": 9.002396166134185, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 11271 + }, + { + "epoch": 9.003194888178914, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 11272 + }, + { + "epoch": 9.003993610223642, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 11273 + }, + { + "epoch": 9.004792332268371, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0187, + "step": 11274 + }, + { + "epoch": 9.005591054313099, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 11275 + }, + { + "epoch": 9.006389776357828, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 11276 + }, + { + "epoch": 9.007188498402556, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 11277 + }, + { + "epoch": 9.007987220447284, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 11278 + }, + { + "epoch": 9.008785942492013, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 11279 + }, + { + "epoch": 9.00958466453674, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0353, + "step": 11280 + }, + { + "epoch": 9.01038338658147, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0199, + "step": 11281 + }, + { + "epoch": 9.011182108626198, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 11282 + }, + { + "epoch": 9.011980830670927, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 11283 + }, + { + "epoch": 9.012779552715655, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 11284 + }, + { + "epoch": 9.013578274760384, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 11285 + }, + { + "epoch": 9.014376996805112, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 11286 + }, + { + "epoch": 9.01517571884984, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 11287 + }, + { + "epoch": 9.015974440894569, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 11288 + }, + { + "epoch": 9.016773162939296, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 11289 + }, + { + "epoch": 9.017571884984026, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 11290 + }, + { + "epoch": 9.018370607028753, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0338, + "step": 11291 + }, + { + "epoch": 9.019169329073483, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 11292 + }, + { + "epoch": 9.01996805111821, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 11293 + }, + { + "epoch": 9.02076677316294, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 11294 + }, + { + "epoch": 9.021565495207668, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 11295 + }, + { + "epoch": 9.022364217252397, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 11296 + }, + { + "epoch": 9.023162939297125, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 11297 + }, + { + "epoch": 9.023961661341852, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 11298 + }, + { + "epoch": 9.024760383386582, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 11299 + }, + { + "epoch": 9.02555910543131, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 11300 + }, + { + "epoch": 9.026357827476039, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 11301 + }, + { + "epoch": 9.027156549520766, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0187, + "step": 11302 + }, + { + "epoch": 9.027955271565496, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 11303 + }, + { + "epoch": 9.028753993610223, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 11304 + }, + { + "epoch": 9.029552715654953, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 11305 + }, + { + "epoch": 9.03035143769968, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 11306 + }, + { + "epoch": 9.031150159744408, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 11307 + }, + { + "epoch": 9.031948881789138, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 11308 + }, + { + "epoch": 9.032747603833865, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 11309 + }, + { + "epoch": 9.033546325878595, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 11310 + }, + { + "epoch": 9.034345047923322, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 11311 + }, + { + "epoch": 9.035143769968052, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 11312 + }, + { + "epoch": 9.03594249201278, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 11313 + }, + { + "epoch": 9.036741214057509, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 11314 + }, + { + "epoch": 9.037539936102236, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 11315 + }, + { + "epoch": 9.038338658146964, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 11316 + }, + { + "epoch": 9.039137380191693, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 11317 + }, + { + "epoch": 9.039936102236421, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 11318 + }, + { + "epoch": 9.04073482428115, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 11319 + }, + { + "epoch": 9.041533546325878, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 11320 + }, + { + "epoch": 9.042332268370608, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 11321 + }, + { + "epoch": 9.043130990415335, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 11322 + }, + { + "epoch": 9.043929712460065, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0183, + "step": 11323 + }, + { + "epoch": 9.044728434504792, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 11324 + }, + { + "epoch": 9.04552715654952, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 11325 + }, + { + "epoch": 9.04632587859425, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 11326 + }, + { + "epoch": 9.047124600638977, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.034, + "step": 11327 + }, + { + "epoch": 9.047923322683706, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 11328 + }, + { + "epoch": 9.048722044728434, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 11329 + }, + { + "epoch": 9.049520766773163, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 11330 + }, + { + "epoch": 9.050319488817891, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0348, + "step": 11331 + }, + { + "epoch": 9.05111821086262, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 11332 + }, + { + "epoch": 9.051916932907348, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 11333 + }, + { + "epoch": 9.052715654952078, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 11334 + }, + { + "epoch": 9.053514376996805, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 11335 + }, + { + "epoch": 9.054313099041533, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 11336 + }, + { + "epoch": 9.055111821086262, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 11337 + }, + { + "epoch": 9.05591054313099, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 11338 + }, + { + "epoch": 9.05670926517572, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 11339 + }, + { + "epoch": 9.057507987220447, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 11340 + }, + { + "epoch": 9.058306709265176, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 11341 + }, + { + "epoch": 9.059105431309904, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 11342 + }, + { + "epoch": 9.059904153354633, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0176, + "step": 11343 + }, + { + "epoch": 9.060702875399361, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 11344 + }, + { + "epoch": 9.061501597444089, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 11345 + }, + { + "epoch": 9.062300319488818, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 11346 + }, + { + "epoch": 9.063099041533546, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0388, + "step": 11347 + }, + { + "epoch": 9.063897763578275, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 11348 + }, + { + "epoch": 9.064696485623003, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 11349 + }, + { + "epoch": 9.065495207667732, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 11350 + }, + { + "epoch": 9.06629392971246, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 11351 + }, + { + "epoch": 9.06709265175719, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 11352 + }, + { + "epoch": 9.067891373801917, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 11353 + }, + { + "epoch": 9.068690095846645, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 11354 + }, + { + "epoch": 9.069488817891374, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 11355 + }, + { + "epoch": 9.070287539936102, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 11356 + }, + { + "epoch": 9.071086261980831, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 11357 + }, + { + "epoch": 9.071884984025559, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0195, + "step": 11358 + }, + { + "epoch": 9.072683706070288, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 11359 + }, + { + "epoch": 9.073482428115016, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 11360 + }, + { + "epoch": 9.074281150159745, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 11361 + }, + { + "epoch": 9.075079872204473, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0391, + "step": 11362 + }, + { + "epoch": 9.0758785942492, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0402, + "step": 11363 + }, + { + "epoch": 9.07667731629393, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 11364 + }, + { + "epoch": 9.077476038338657, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 11365 + }, + { + "epoch": 9.078274760383387, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 11366 + }, + { + "epoch": 9.079073482428115, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 11367 + }, + { + "epoch": 9.079872204472844, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 11368 + }, + { + "epoch": 9.080670926517572, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 11369 + }, + { + "epoch": 9.081469648562301, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.037, + "step": 11370 + }, + { + "epoch": 9.082268370607029, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 11371 + }, + { + "epoch": 9.083067092651758, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0362, + "step": 11372 + }, + { + "epoch": 9.083865814696486, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 11373 + }, + { + "epoch": 9.084664536741213, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 11374 + }, + { + "epoch": 9.085463258785943, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 11375 + }, + { + "epoch": 9.08626198083067, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 11376 + }, + { + "epoch": 9.0870607028754, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 11377 + }, + { + "epoch": 9.087859424920127, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 11378 + }, + { + "epoch": 9.088658146964857, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 11379 + }, + { + "epoch": 9.089456869009584, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 11380 + }, + { + "epoch": 9.090255591054314, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 11381 + }, + { + "epoch": 9.091054313099042, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 11382 + }, + { + "epoch": 9.09185303514377, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 11383 + }, + { + "epoch": 9.092651757188499, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 11384 + }, + { + "epoch": 9.093450479233226, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 11385 + }, + { + "epoch": 9.094249201277956, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 11386 + }, + { + "epoch": 9.095047923322683, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 11387 + }, + { + "epoch": 9.095846645367413, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 11388 + }, + { + "epoch": 9.09664536741214, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 11389 + }, + { + "epoch": 9.09744408945687, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0378, + "step": 11390 + }, + { + "epoch": 9.098242811501597, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 11391 + }, + { + "epoch": 9.099041533546325, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 11392 + }, + { + "epoch": 9.099840255591054, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 11393 + }, + { + "epoch": 9.100638977635782, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 11394 + }, + { + "epoch": 9.101437699680512, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 11395 + }, + { + "epoch": 9.10223642172524, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0415, + "step": 11396 + }, + { + "epoch": 9.103035143769969, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 11397 + }, + { + "epoch": 9.103833865814696, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 11398 + }, + { + "epoch": 9.104632587859426, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 11399 + }, + { + "epoch": 9.105431309904153, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 11400 + }, + { + "epoch": 9.106230031948881, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 11401 + }, + { + "epoch": 9.10702875399361, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 11402 + }, + { + "epoch": 9.107827476038338, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 11403 + }, + { + "epoch": 9.108626198083067, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 11404 + }, + { + "epoch": 9.109424920127795, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 11405 + }, + { + "epoch": 9.110223642172524, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 11406 + }, + { + "epoch": 9.111022364217252, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 11407 + }, + { + "epoch": 9.111821086261982, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 11408 + }, + { + "epoch": 9.11261980830671, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 11409 + }, + { + "epoch": 9.113418530351439, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 11410 + }, + { + "epoch": 9.114217252396166, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 11411 + }, + { + "epoch": 9.115015974440894, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 11412 + }, + { + "epoch": 9.115814696485623, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 11413 + }, + { + "epoch": 9.116613418530351, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 11414 + }, + { + "epoch": 9.11741214057508, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 11415 + }, + { + "epoch": 9.118210862619808, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0196, + "step": 11416 + }, + { + "epoch": 9.119009584664537, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 11417 + }, + { + "epoch": 9.119808306709265, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0385, + "step": 11418 + }, + { + "epoch": 9.120607028753994, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 11419 + }, + { + "epoch": 9.121405750798722, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 11420 + }, + { + "epoch": 9.12220447284345, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 11421 + }, + { + "epoch": 9.12300319488818, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 11422 + }, + { + "epoch": 9.123801916932907, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 11423 + }, + { + "epoch": 9.124600638977636, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.036, + "step": 11424 + }, + { + "epoch": 9.125399361022364, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 11425 + }, + { + "epoch": 9.126198083067093, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 11426 + }, + { + "epoch": 9.12699680511182, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 11427 + }, + { + "epoch": 9.12779552715655, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 11428 + }, + { + "epoch": 9.128594249201278, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0361, + "step": 11429 + }, + { + "epoch": 9.129392971246006, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 11430 + }, + { + "epoch": 9.130191693290735, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.036, + "step": 11431 + }, + { + "epoch": 9.130990415335463, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 11432 + }, + { + "epoch": 9.131789137380192, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 11433 + }, + { + "epoch": 9.13258785942492, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 11434 + }, + { + "epoch": 9.133386581469649, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 11435 + }, + { + "epoch": 9.134185303514377, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 11436 + }, + { + "epoch": 9.134984025559106, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 11437 + }, + { + "epoch": 9.135782747603834, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 11438 + }, + { + "epoch": 9.136581469648561, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 11439 + }, + { + "epoch": 9.13738019169329, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 11440 + }, + { + "epoch": 9.138178913738018, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 11441 + }, + { + "epoch": 9.138977635782748, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 11442 + }, + { + "epoch": 9.139776357827476, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 11443 + }, + { + "epoch": 9.140575079872205, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 11444 + }, + { + "epoch": 9.141373801916933, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 11445 + }, + { + "epoch": 9.142172523961662, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 11446 + }, + { + "epoch": 9.14297124600639, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 11447 + }, + { + "epoch": 9.143769968051119, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 11448 + }, + { + "epoch": 9.144568690095847, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 11449 + }, + { + "epoch": 9.145367412140574, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 11450 + }, + { + "epoch": 9.146166134185304, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 11451 + }, + { + "epoch": 9.146964856230031, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0337, + "step": 11452 + }, + { + "epoch": 9.14776357827476, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 11453 + }, + { + "epoch": 9.148562300319488, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 11454 + }, + { + "epoch": 9.149361022364218, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 11455 + }, + { + "epoch": 9.150159744408946, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.038, + "step": 11456 + }, + { + "epoch": 9.150958466453675, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 11457 + }, + { + "epoch": 9.151757188498403, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 11458 + }, + { + "epoch": 9.15255591054313, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 11459 + }, + { + "epoch": 9.15335463258786, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0396, + "step": 11460 + }, + { + "epoch": 9.154153354632587, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 11461 + }, + { + "epoch": 9.154952076677317, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 11462 + }, + { + "epoch": 9.155750798722044, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 11463 + }, + { + "epoch": 9.156549520766774, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 11464 + }, + { + "epoch": 9.157348242811501, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 11465 + }, + { + "epoch": 9.15814696485623, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0388, + "step": 11466 + }, + { + "epoch": 9.158945686900958, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 11467 + }, + { + "epoch": 9.159744408945686, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 11468 + }, + { + "epoch": 9.160543130990416, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 11469 + }, + { + "epoch": 9.161341853035143, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 11470 + }, + { + "epoch": 9.162140575079873, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 11471 + }, + { + "epoch": 9.1629392971246, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0161, + "step": 11472 + }, + { + "epoch": 9.16373801916933, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 11473 + }, + { + "epoch": 9.164536741214057, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 11474 + }, + { + "epoch": 9.165335463258787, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 11475 + }, + { + "epoch": 9.166134185303514, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 11476 + }, + { + "epoch": 9.166932907348242, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 11477 + }, + { + "epoch": 9.167731629392971, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 11478 + }, + { + "epoch": 9.168530351437699, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0188, + "step": 11479 + }, + { + "epoch": 9.169329073482428, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 11480 + }, + { + "epoch": 9.170127795527156, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0415, + "step": 11481 + }, + { + "epoch": 9.170926517571885, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 11482 + }, + { + "epoch": 9.171725239616613, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 11483 + }, + { + "epoch": 9.172523961661343, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 11484 + }, + { + "epoch": 9.17332268370607, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 11485 + }, + { + "epoch": 9.1741214057508, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 11486 + }, + { + "epoch": 9.174920127795527, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 11487 + }, + { + "epoch": 9.175718849840255, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 11488 + }, + { + "epoch": 9.176517571884984, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 11489 + }, + { + "epoch": 9.177316293929712, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 11490 + }, + { + "epoch": 9.178115015974441, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 11491 + }, + { + "epoch": 9.178913738019169, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 11492 + }, + { + "epoch": 9.179712460063898, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 11493 + }, + { + "epoch": 9.180511182108626, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 11494 + }, + { + "epoch": 9.181309904153355, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0395, + "step": 11495 + }, + { + "epoch": 9.182108626198083, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0338, + "step": 11496 + }, + { + "epoch": 9.18290734824281, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 11497 + }, + { + "epoch": 9.18370607028754, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 11498 + }, + { + "epoch": 9.184504792332268, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 11499 + }, + { + "epoch": 9.185303514376997, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 11500 + }, + { + "epoch": 9.186102236421725, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 11501 + }, + { + "epoch": 9.186900958466454, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0361, + "step": 11502 + }, + { + "epoch": 9.187699680511182, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 11503 + }, + { + "epoch": 9.188498402555911, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 11504 + }, + { + "epoch": 9.189297124600639, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 11505 + }, + { + "epoch": 9.190095846645367, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 11506 + }, + { + "epoch": 9.190894568690096, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 11507 + }, + { + "epoch": 9.191693290734824, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 11508 + }, + { + "epoch": 9.192492012779553, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 11509 + }, + { + "epoch": 9.19329073482428, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 11510 + }, + { + "epoch": 9.19408945686901, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0376, + "step": 11511 + }, + { + "epoch": 9.194888178913738, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0375, + "step": 11512 + }, + { + "epoch": 9.195686900958467, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0407, + "step": 11513 + }, + { + "epoch": 9.196485623003195, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 11514 + }, + { + "epoch": 9.197284345047922, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 11515 + }, + { + "epoch": 9.198083067092652, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 11516 + }, + { + "epoch": 9.19888178913738, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 11517 + }, + { + "epoch": 9.199680511182109, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 11518 + }, + { + "epoch": 9.200479233226837, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0116, + "step": 11519 + }, + { + "epoch": 9.201277955271566, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 11520 + }, + { + "epoch": 9.202076677316294, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 11521 + }, + { + "epoch": 9.202875399361023, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 11522 + }, + { + "epoch": 9.20367412140575, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 11523 + }, + { + "epoch": 9.204472843450478, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0387, + "step": 11524 + }, + { + "epoch": 9.205271565495208, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 11525 + }, + { + "epoch": 9.206070287539935, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 11526 + }, + { + "epoch": 9.206869009584665, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 11527 + }, + { + "epoch": 9.207667731629392, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 11528 + }, + { + "epoch": 9.208466453674122, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 11529 + }, + { + "epoch": 9.20926517571885, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 11530 + }, + { + "epoch": 9.210063897763579, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0131, + "step": 11531 + }, + { + "epoch": 9.210862619808307, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 11532 + }, + { + "epoch": 9.211661341853036, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 11533 + }, + { + "epoch": 9.212460063897764, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0399, + "step": 11534 + }, + { + "epoch": 9.213258785942491, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0161, + "step": 11535 + }, + { + "epoch": 9.21405750798722, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 11536 + }, + { + "epoch": 9.214856230031948, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 11537 + }, + { + "epoch": 9.215654952076678, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 11538 + }, + { + "epoch": 9.216453674121405, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 11539 + }, + { + "epoch": 9.217252396166135, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 11540 + }, + { + "epoch": 9.218051118210862, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 11541 + }, + { + "epoch": 9.218849840255592, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 11542 + }, + { + "epoch": 9.21964856230032, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 11543 + }, + { + "epoch": 9.220447284345047, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 11544 + }, + { + "epoch": 9.221246006389777, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 11545 + }, + { + "epoch": 9.222044728434504, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 11546 + }, + { + "epoch": 9.222843450479234, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 11547 + }, + { + "epoch": 9.223642172523961, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 11548 + }, + { + "epoch": 9.22444089456869, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 11549 + }, + { + "epoch": 9.225239616613418, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 11550 + }, + { + "epoch": 9.226038338658148, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 11551 + }, + { + "epoch": 9.226837060702875, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 11552 + }, + { + "epoch": 9.227635782747603, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 11553 + }, + { + "epoch": 9.228434504792332, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0362, + "step": 11554 + }, + { + "epoch": 9.22923322683706, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0153, + "step": 11555 + }, + { + "epoch": 9.23003194888179, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 11556 + }, + { + "epoch": 9.230830670926517, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 11557 + }, + { + "epoch": 9.231629392971247, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 11558 + }, + { + "epoch": 9.232428115015974, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 11559 + }, + { + "epoch": 9.233226837060704, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 11560 + }, + { + "epoch": 9.234025559105431, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 11561 + }, + { + "epoch": 9.23482428115016, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 11562 + }, + { + "epoch": 9.235623003194888, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0369, + "step": 11563 + }, + { + "epoch": 9.236421725239616, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 11564 + }, + { + "epoch": 9.237220447284345, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 11565 + }, + { + "epoch": 9.238019169329073, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 11566 + }, + { + "epoch": 9.238817891373802, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.036, + "step": 11567 + }, + { + "epoch": 9.23961661341853, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 11568 + }, + { + "epoch": 9.24041533546326, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0353, + "step": 11569 + }, + { + "epoch": 9.241214057507987, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 11570 + }, + { + "epoch": 9.242012779552716, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 11571 + }, + { + "epoch": 9.242811501597444, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0161, + "step": 11572 + }, + { + "epoch": 9.243610223642172, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 11573 + }, + { + "epoch": 9.244408945686901, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 11574 + }, + { + "epoch": 9.245207667731629, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 11575 + }, + { + "epoch": 9.246006389776358, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 11576 + }, + { + "epoch": 9.246805111821086, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 11577 + }, + { + "epoch": 9.247603833865815, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 11578 + }, + { + "epoch": 9.248402555910543, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 11579 + }, + { + "epoch": 9.249201277955272, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 11580 + }, + { + "epoch": 9.25, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0375, + "step": 11581 + }, + { + "epoch": 9.250798722044728, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 11582 + }, + { + "epoch": 9.251597444089457, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 11583 + }, + { + "epoch": 9.252396166134185, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 11584 + }, + { + "epoch": 9.253194888178914, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 11585 + }, + { + "epoch": 9.253993610223642, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 11586 + }, + { + "epoch": 9.254792332268371, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 11587 + }, + { + "epoch": 9.255591054313099, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 11588 + }, + { + "epoch": 9.256389776357828, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 11589 + }, + { + "epoch": 9.257188498402556, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 11590 + }, + { + "epoch": 9.257987220447284, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 11591 + }, + { + "epoch": 9.258785942492013, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 11592 + }, + { + "epoch": 9.25958466453674, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 11593 + }, + { + "epoch": 9.26038338658147, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 11594 + }, + { + "epoch": 9.261182108626198, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0192, + "step": 11595 + }, + { + "epoch": 9.261980830670927, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 11596 + }, + { + "epoch": 9.262779552715655, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 11597 + }, + { + "epoch": 9.263578274760384, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 11598 + }, + { + "epoch": 9.264376996805112, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 11599 + }, + { + "epoch": 9.26517571884984, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 11600 + }, + { + "epoch": 9.265974440894569, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 11601 + }, + { + "epoch": 9.266773162939296, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 11602 + }, + { + "epoch": 9.267571884984026, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 11603 + }, + { + "epoch": 9.268370607028753, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 11604 + }, + { + "epoch": 9.269169329073483, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 11605 + }, + { + "epoch": 9.26996805111821, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 11606 + }, + { + "epoch": 9.27076677316294, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 11607 + }, + { + "epoch": 9.271565495207668, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 11608 + }, + { + "epoch": 9.272364217252397, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 11609 + }, + { + "epoch": 9.273162939297125, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 11610 + }, + { + "epoch": 9.273961661341852, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 11611 + }, + { + "epoch": 9.274760383386582, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 11612 + }, + { + "epoch": 9.27555910543131, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 11613 + }, + { + "epoch": 9.276357827476039, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 11614 + }, + { + "epoch": 9.277156549520766, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 11615 + }, + { + "epoch": 9.277955271565496, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 11616 + }, + { + "epoch": 9.278753993610223, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 11617 + }, + { + "epoch": 9.279552715654953, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 11618 + }, + { + "epoch": 9.28035143769968, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0337, + "step": 11619 + }, + { + "epoch": 9.281150159744408, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 11620 + }, + { + "epoch": 9.281948881789138, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 11621 + }, + { + "epoch": 9.282747603833865, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 11622 + }, + { + "epoch": 9.283546325878595, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 11623 + }, + { + "epoch": 9.284345047923322, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 11624 + }, + { + "epoch": 9.285143769968052, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 11625 + }, + { + "epoch": 9.28594249201278, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 11626 + }, + { + "epoch": 9.286741214057509, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 11627 + }, + { + "epoch": 9.287539936102236, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 11628 + }, + { + "epoch": 9.288338658146964, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 11629 + }, + { + "epoch": 9.289137380191693, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 11630 + }, + { + "epoch": 9.289936102236421, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 11631 + }, + { + "epoch": 9.29073482428115, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 11632 + }, + { + "epoch": 9.291533546325878, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 11633 + }, + { + "epoch": 9.292332268370608, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 11634 + }, + { + "epoch": 9.293130990415335, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 11635 + }, + { + "epoch": 9.293929712460065, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 11636 + }, + { + "epoch": 9.294728434504792, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 11637 + }, + { + "epoch": 9.295527156549522, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 11638 + }, + { + "epoch": 9.29632587859425, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 11639 + }, + { + "epoch": 9.297124600638977, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 11640 + }, + { + "epoch": 9.297923322683706, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 11641 + }, + { + "epoch": 9.298722044728434, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 11642 + }, + { + "epoch": 9.299520766773163, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0353, + "step": 11643 + }, + { + "epoch": 9.300319488817891, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 11644 + }, + { + "epoch": 9.30111821086262, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 11645 + }, + { + "epoch": 9.301916932907348, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 11646 + }, + { + "epoch": 9.302715654952078, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0199, + "step": 11647 + }, + { + "epoch": 9.303514376996805, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 11648 + }, + { + "epoch": 9.304313099041533, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.034, + "step": 11649 + }, + { + "epoch": 9.305111821086262, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0337, + "step": 11650 + }, + { + "epoch": 9.30591054313099, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 11651 + }, + { + "epoch": 9.30670926517572, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 11652 + }, + { + "epoch": 9.307507987220447, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0163, + "step": 11653 + }, + { + "epoch": 9.308306709265176, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 11654 + }, + { + "epoch": 9.309105431309904, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 11655 + }, + { + "epoch": 9.309904153354633, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 11656 + }, + { + "epoch": 9.310702875399361, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 11657 + }, + { + "epoch": 9.311501597444089, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0348, + "step": 11658 + }, + { + "epoch": 9.312300319488818, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0157, + "step": 11659 + }, + { + "epoch": 9.313099041533546, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 11660 + }, + { + "epoch": 9.313897763578275, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 11661 + }, + { + "epoch": 9.314696485623003, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 11662 + }, + { + "epoch": 9.315495207667732, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 11663 + }, + { + "epoch": 9.31629392971246, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 11664 + }, + { + "epoch": 9.31709265175719, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 11665 + }, + { + "epoch": 9.317891373801917, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0353, + "step": 11666 + }, + { + "epoch": 9.318690095846645, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 11667 + }, + { + "epoch": 9.319488817891374, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 11668 + }, + { + "epoch": 9.320287539936102, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 11669 + }, + { + "epoch": 9.321086261980831, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 11670 + }, + { + "epoch": 9.321884984025559, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 11671 + }, + { + "epoch": 9.322683706070288, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 11672 + }, + { + "epoch": 9.323482428115016, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 11673 + }, + { + "epoch": 9.324281150159745, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 11674 + }, + { + "epoch": 9.325079872204473, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 11675 + }, + { + "epoch": 9.3258785942492, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 11676 + }, + { + "epoch": 9.32667731629393, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 11677 + }, + { + "epoch": 9.327476038338657, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 11678 + }, + { + "epoch": 9.328274760383387, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 11679 + }, + { + "epoch": 9.329073482428115, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 11680 + }, + { + "epoch": 9.329872204472844, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 11681 + }, + { + "epoch": 9.330670926517572, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 11682 + }, + { + "epoch": 9.331469648562301, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 11683 + }, + { + "epoch": 9.332268370607029, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 11684 + }, + { + "epoch": 9.333067092651758, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 11685 + }, + { + "epoch": 9.333865814696486, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0173, + "step": 11686 + }, + { + "epoch": 9.334664536741213, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 11687 + }, + { + "epoch": 9.335463258785943, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 11688 + }, + { + "epoch": 9.33626198083067, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 11689 + }, + { + "epoch": 9.3370607028754, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 11690 + }, + { + "epoch": 9.337859424920127, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 11691 + }, + { + "epoch": 9.338658146964857, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 11692 + }, + { + "epoch": 9.339456869009584, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 11693 + }, + { + "epoch": 9.340255591054314, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0348, + "step": 11694 + }, + { + "epoch": 9.341054313099042, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 11695 + }, + { + "epoch": 9.34185303514377, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 11696 + }, + { + "epoch": 9.342651757188499, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 11697 + }, + { + "epoch": 9.343450479233226, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 11698 + }, + { + "epoch": 9.344249201277956, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 11699 + }, + { + "epoch": 9.345047923322683, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 11700 + }, + { + "epoch": 9.345846645367413, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 11701 + }, + { + "epoch": 9.34664536741214, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 11702 + }, + { + "epoch": 9.34744408945687, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 11703 + }, + { + "epoch": 9.348242811501597, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 11704 + }, + { + "epoch": 9.349041533546325, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 11705 + }, + { + "epoch": 9.349840255591054, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 11706 + }, + { + "epoch": 9.350638977635782, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 11707 + }, + { + "epoch": 9.351437699680512, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 11708 + }, + { + "epoch": 9.35223642172524, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 11709 + }, + { + "epoch": 9.353035143769969, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 11710 + }, + { + "epoch": 9.353833865814696, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 11711 + }, + { + "epoch": 9.354632587859426, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 11712 + }, + { + "epoch": 9.355431309904153, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 11713 + }, + { + "epoch": 9.356230031948883, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 11714 + }, + { + "epoch": 9.35702875399361, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0195, + "step": 11715 + }, + { + "epoch": 9.357827476038338, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 11716 + }, + { + "epoch": 9.358626198083067, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 11717 + }, + { + "epoch": 9.359424920127795, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 11718 + }, + { + "epoch": 9.360223642172524, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 11719 + }, + { + "epoch": 9.361022364217252, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 11720 + }, + { + "epoch": 9.361821086261982, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0396, + "step": 11721 + }, + { + "epoch": 9.36261980830671, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 11722 + }, + { + "epoch": 9.363418530351439, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 11723 + }, + { + "epoch": 9.364217252396166, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 11724 + }, + { + "epoch": 9.365015974440894, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 11725 + }, + { + "epoch": 9.365814696485623, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 11726 + }, + { + "epoch": 9.366613418530351, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 11727 + }, + { + "epoch": 9.36741214057508, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 11728 + }, + { + "epoch": 9.368210862619808, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 11729 + }, + { + "epoch": 9.369009584664537, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 11730 + }, + { + "epoch": 9.369808306709265, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 11731 + }, + { + "epoch": 9.370607028753994, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 11732 + }, + { + "epoch": 9.371405750798722, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0188, + "step": 11733 + }, + { + "epoch": 9.37220447284345, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 11734 + }, + { + "epoch": 9.37300319488818, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 11735 + }, + { + "epoch": 9.373801916932907, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 11736 + }, + { + "epoch": 9.374600638977636, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 11737 + }, + { + "epoch": 9.375399361022364, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 11738 + }, + { + "epoch": 9.376198083067093, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 11739 + }, + { + "epoch": 9.37699680511182, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 11740 + }, + { + "epoch": 9.37779552715655, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 11741 + }, + { + "epoch": 9.378594249201278, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 11742 + }, + { + "epoch": 9.379392971246006, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 11743 + }, + { + "epoch": 9.380191693290735, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 11744 + }, + { + "epoch": 9.380990415335463, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 11745 + }, + { + "epoch": 9.381789137380192, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 11746 + }, + { + "epoch": 9.38258785942492, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 11747 + }, + { + "epoch": 9.383386581469649, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 11748 + }, + { + "epoch": 9.384185303514377, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 11749 + }, + { + "epoch": 9.384984025559106, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 11750 + }, + { + "epoch": 9.385782747603834, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 11751 + }, + { + "epoch": 9.386581469648561, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 11752 + }, + { + "epoch": 9.38738019169329, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 11753 + }, + { + "epoch": 9.388178913738018, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 11754 + }, + { + "epoch": 9.388977635782748, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 11755 + }, + { + "epoch": 9.389776357827476, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 11756 + }, + { + "epoch": 9.390575079872205, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 11757 + }, + { + "epoch": 9.391373801916933, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 11758 + }, + { + "epoch": 9.392172523961662, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 11759 + }, + { + "epoch": 9.39297124600639, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 11760 + }, + { + "epoch": 9.393769968051117, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0405, + "step": 11761 + }, + { + "epoch": 9.394568690095847, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 11762 + }, + { + "epoch": 9.395367412140574, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 11763 + }, + { + "epoch": 9.396166134185304, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 11764 + }, + { + "epoch": 9.396964856230031, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 11765 + }, + { + "epoch": 9.39776357827476, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 11766 + }, + { + "epoch": 9.398562300319488, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 11767 + }, + { + "epoch": 9.399361022364218, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.017, + "step": 11768 + }, + { + "epoch": 9.400159744408946, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 11769 + }, + { + "epoch": 9.400958466453675, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 11770 + }, + { + "epoch": 9.401757188498403, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 11771 + }, + { + "epoch": 9.40255591054313, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 11772 + }, + { + "epoch": 9.40335463258786, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0338, + "step": 11773 + }, + { + "epoch": 9.404153354632587, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 11774 + }, + { + "epoch": 9.404952076677317, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 11775 + }, + { + "epoch": 9.405750798722044, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 11776 + }, + { + "epoch": 9.406549520766774, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 11777 + }, + { + "epoch": 9.407348242811501, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 11778 + }, + { + "epoch": 9.40814696485623, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 11779 + }, + { + "epoch": 9.408945686900958, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 11780 + }, + { + "epoch": 9.409744408945686, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 11781 + }, + { + "epoch": 9.410543130990416, + "grad_norm": 0.1806640625, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 11782 + }, + { + "epoch": 9.411341853035143, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 11783 + }, + { + "epoch": 9.412140575079873, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0348, + "step": 11784 + }, + { + "epoch": 9.4129392971246, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 11785 + }, + { + "epoch": 9.41373801916933, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 11786 + }, + { + "epoch": 9.414536741214057, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 11787 + }, + { + "epoch": 9.415335463258787, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 11788 + }, + { + "epoch": 9.416134185303514, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 11789 + }, + { + "epoch": 9.416932907348242, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 11790 + }, + { + "epoch": 9.417731629392971, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 11791 + }, + { + "epoch": 9.418530351437699, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 11792 + }, + { + "epoch": 9.419329073482428, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 11793 + }, + { + "epoch": 9.420127795527156, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 11794 + }, + { + "epoch": 9.420926517571885, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 11795 + }, + { + "epoch": 9.421725239616613, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 11796 + }, + { + "epoch": 9.422523961661343, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 11797 + }, + { + "epoch": 9.42332268370607, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 11798 + }, + { + "epoch": 9.4241214057508, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 11799 + }, + { + "epoch": 9.424920127795527, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0379, + "step": 11800 + }, + { + "epoch": 9.425718849840255, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 11801 + }, + { + "epoch": 9.426517571884984, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 11802 + }, + { + "epoch": 9.427316293929712, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 11803 + }, + { + "epoch": 9.428115015974441, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 11804 + }, + { + "epoch": 9.428913738019169, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 11805 + }, + { + "epoch": 9.429712460063898, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 11806 + }, + { + "epoch": 9.430511182108626, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 11807 + }, + { + "epoch": 9.431309904153355, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 11808 + }, + { + "epoch": 9.432108626198083, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 11809 + }, + { + "epoch": 9.43290734824281, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0146, + "step": 11810 + }, + { + "epoch": 9.43370607028754, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 11811 + }, + { + "epoch": 9.434504792332268, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 11812 + }, + { + "epoch": 9.435303514376997, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 11813 + }, + { + "epoch": 9.436102236421725, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0337, + "step": 11814 + }, + { + "epoch": 9.436900958466454, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 11815 + }, + { + "epoch": 9.437699680511182, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0462, + "step": 11816 + }, + { + "epoch": 9.438498402555911, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 11817 + }, + { + "epoch": 9.439297124600639, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0188, + "step": 11818 + }, + { + "epoch": 9.440095846645367, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 11819 + }, + { + "epoch": 9.440894568690096, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0348, + "step": 11820 + }, + { + "epoch": 9.441693290734824, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 11821 + }, + { + "epoch": 9.442492012779553, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 11822 + }, + { + "epoch": 9.44329073482428, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 11823 + }, + { + "epoch": 9.44408945686901, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 11824 + }, + { + "epoch": 9.444888178913738, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 11825 + }, + { + "epoch": 9.445686900958467, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0388, + "step": 11826 + }, + { + "epoch": 9.446485623003195, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 11827 + }, + { + "epoch": 9.447284345047922, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 11828 + }, + { + "epoch": 9.448083067092652, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 11829 + }, + { + "epoch": 9.44888178913738, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 11830 + }, + { + "epoch": 9.449680511182109, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 11831 + }, + { + "epoch": 9.450479233226837, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.034, + "step": 11832 + }, + { + "epoch": 9.451277955271566, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 11833 + }, + { + "epoch": 9.452076677316294, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 11834 + }, + { + "epoch": 9.452875399361023, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 11835 + }, + { + "epoch": 9.45367412140575, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 11836 + }, + { + "epoch": 9.454472843450478, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0181, + "step": 11837 + }, + { + "epoch": 9.455271565495208, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0383, + "step": 11838 + }, + { + "epoch": 9.456070287539935, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 11839 + }, + { + "epoch": 9.456869009584665, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 11840 + }, + { + "epoch": 9.457667731629392, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 11841 + }, + { + "epoch": 9.458466453674122, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 11842 + }, + { + "epoch": 9.45926517571885, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 11843 + }, + { + "epoch": 9.460063897763579, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 11844 + }, + { + "epoch": 9.460862619808307, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 11845 + }, + { + "epoch": 9.461661341853036, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 11846 + }, + { + "epoch": 9.462460063897764, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 11847 + }, + { + "epoch": 9.463258785942491, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 11848 + }, + { + "epoch": 9.46405750798722, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 11849 + }, + { + "epoch": 9.464856230031948, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 11850 + }, + { + "epoch": 9.465654952076678, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 11851 + }, + { + "epoch": 9.466453674121405, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 11852 + }, + { + "epoch": 9.467252396166135, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 11853 + }, + { + "epoch": 9.468051118210862, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 11854 + }, + { + "epoch": 9.468849840255592, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 11855 + }, + { + "epoch": 9.46964856230032, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 11856 + }, + { + "epoch": 9.470447284345047, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 11857 + }, + { + "epoch": 9.471246006389777, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 11858 + }, + { + "epoch": 9.472044728434504, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 11859 + }, + { + "epoch": 9.472843450479234, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 11860 + }, + { + "epoch": 9.473642172523961, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 11861 + }, + { + "epoch": 9.47444089456869, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 11862 + }, + { + "epoch": 9.475239616613418, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 11863 + }, + { + "epoch": 9.476038338658148, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 11864 + }, + { + "epoch": 9.476837060702875, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 11865 + }, + { + "epoch": 9.477635782747603, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 11866 + }, + { + "epoch": 9.478434504792332, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 11867 + }, + { + "epoch": 9.47923322683706, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 11868 + }, + { + "epoch": 9.48003194888179, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 11869 + }, + { + "epoch": 9.480830670926517, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 11870 + }, + { + "epoch": 9.481629392971247, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0361, + "step": 11871 + }, + { + "epoch": 9.482428115015974, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 11872 + }, + { + "epoch": 9.483226837060704, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 11873 + }, + { + "epoch": 9.484025559105431, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0395, + "step": 11874 + }, + { + "epoch": 9.48482428115016, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 11875 + }, + { + "epoch": 9.485623003194888, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 11876 + }, + { + "epoch": 9.486421725239616, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 11877 + }, + { + "epoch": 9.487220447284345, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 11878 + }, + { + "epoch": 9.488019169329073, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 11879 + }, + { + "epoch": 9.488817891373802, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 11880 + }, + { + "epoch": 9.48961661341853, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 11881 + }, + { + "epoch": 9.49041533546326, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 11882 + }, + { + "epoch": 9.491214057507987, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 11883 + }, + { + "epoch": 9.492012779552716, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 11884 + }, + { + "epoch": 9.492811501597444, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 11885 + }, + { + "epoch": 9.493610223642172, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 11886 + }, + { + "epoch": 9.494408945686901, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 11887 + }, + { + "epoch": 9.495207667731629, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 11888 + }, + { + "epoch": 9.496006389776358, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 11889 + }, + { + "epoch": 9.496805111821086, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0179, + "step": 11890 + }, + { + "epoch": 9.497603833865815, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 11891 + }, + { + "epoch": 9.498402555910543, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0369, + "step": 11892 + }, + { + "epoch": 9.499201277955272, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 11893 + }, + { + "epoch": 9.5, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 11894 + }, + { + "epoch": 9.500798722044728, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 11895 + }, + { + "epoch": 9.501597444089457, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 11896 + }, + { + "epoch": 9.502396166134185, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 11897 + }, + { + "epoch": 9.503194888178914, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 11898 + }, + { + "epoch": 9.503993610223642, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 11899 + }, + { + "epoch": 9.504792332268371, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 11900 + }, + { + "epoch": 9.505591054313099, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 11901 + }, + { + "epoch": 9.506389776357828, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 11902 + }, + { + "epoch": 9.507188498402556, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 11903 + }, + { + "epoch": 9.507987220447284, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 11904 + }, + { + "epoch": 9.508785942492013, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0362, + "step": 11905 + }, + { + "epoch": 9.50958466453674, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 11906 + }, + { + "epoch": 9.51038338658147, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 11907 + }, + { + "epoch": 9.511182108626198, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0338, + "step": 11908 + }, + { + "epoch": 9.511980830670927, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 11909 + }, + { + "epoch": 9.512779552715655, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0182, + "step": 11910 + }, + { + "epoch": 9.513578274760384, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 11911 + }, + { + "epoch": 9.514376996805112, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 11912 + }, + { + "epoch": 9.51517571884984, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 11913 + }, + { + "epoch": 9.515974440894569, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 11914 + }, + { + "epoch": 9.516773162939296, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 11915 + }, + { + "epoch": 9.517571884984026, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.034, + "step": 11916 + }, + { + "epoch": 9.518370607028753, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 11917 + }, + { + "epoch": 9.519169329073483, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 11918 + }, + { + "epoch": 9.51996805111821, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 11919 + }, + { + "epoch": 9.52076677316294, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 11920 + }, + { + "epoch": 9.521565495207668, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 11921 + }, + { + "epoch": 9.522364217252395, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 11922 + }, + { + "epoch": 9.523162939297125, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 11923 + }, + { + "epoch": 9.523961661341852, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 11924 + }, + { + "epoch": 9.524760383386582, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 11925 + }, + { + "epoch": 9.52555910543131, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 11926 + }, + { + "epoch": 9.526357827476039, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 11927 + }, + { + "epoch": 9.527156549520766, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 11928 + }, + { + "epoch": 9.527955271565496, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0394, + "step": 11929 + }, + { + "epoch": 9.528753993610223, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 11930 + }, + { + "epoch": 9.529552715654953, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 11931 + }, + { + "epoch": 9.53035143769968, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 11932 + }, + { + "epoch": 9.531150159744408, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 11933 + }, + { + "epoch": 9.531948881789138, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0375, + "step": 11934 + }, + { + "epoch": 9.532747603833865, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0354, + "step": 11935 + }, + { + "epoch": 9.533546325878595, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.036, + "step": 11936 + }, + { + "epoch": 9.534345047923322, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 11937 + }, + { + "epoch": 9.535143769968052, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 11938 + }, + { + "epoch": 9.53594249201278, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0338, + "step": 11939 + }, + { + "epoch": 9.536741214057509, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 11940 + }, + { + "epoch": 9.537539936102236, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 11941 + }, + { + "epoch": 9.538338658146966, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 11942 + }, + { + "epoch": 9.539137380191693, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 11943 + }, + { + "epoch": 9.539936102236421, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 11944 + }, + { + "epoch": 9.54073482428115, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 11945 + }, + { + "epoch": 9.541533546325878, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 11946 + }, + { + "epoch": 9.542332268370608, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0378, + "step": 11947 + }, + { + "epoch": 9.543130990415335, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 11948 + }, + { + "epoch": 9.543929712460065, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 11949 + }, + { + "epoch": 9.544728434504792, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 11950 + }, + { + "epoch": 9.545527156549522, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 11951 + }, + { + "epoch": 9.54632587859425, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 11952 + }, + { + "epoch": 9.547124600638977, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 11953 + }, + { + "epoch": 9.547923322683706, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 11954 + }, + { + "epoch": 9.548722044728434, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 11955 + }, + { + "epoch": 9.549520766773163, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0378, + "step": 11956 + }, + { + "epoch": 9.550319488817891, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 11957 + }, + { + "epoch": 9.55111821086262, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.038, + "step": 11958 + }, + { + "epoch": 9.551916932907348, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 11959 + }, + { + "epoch": 9.552715654952078, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 11960 + }, + { + "epoch": 9.553514376996805, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 11961 + }, + { + "epoch": 9.554313099041533, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 11962 + }, + { + "epoch": 9.555111821086262, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 11963 + }, + { + "epoch": 9.55591054313099, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0393, + "step": 11964 + }, + { + "epoch": 9.55670926517572, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 11965 + }, + { + "epoch": 9.557507987220447, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 11966 + }, + { + "epoch": 9.558306709265176, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 11967 + }, + { + "epoch": 9.559105431309904, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 11968 + }, + { + "epoch": 9.559904153354633, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 11969 + }, + { + "epoch": 9.560702875399361, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 11970 + }, + { + "epoch": 9.561501597444089, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 11971 + }, + { + "epoch": 9.562300319488818, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 11972 + }, + { + "epoch": 9.563099041533546, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 11973 + }, + { + "epoch": 9.563897763578275, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 11974 + }, + { + "epoch": 9.564696485623003, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 11975 + }, + { + "epoch": 9.565495207667732, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 11976 + }, + { + "epoch": 9.56629392971246, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 11977 + }, + { + "epoch": 9.56709265175719, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 11978 + }, + { + "epoch": 9.567891373801917, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0176, + "step": 11979 + }, + { + "epoch": 9.568690095846645, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 11980 + }, + { + "epoch": 9.569488817891374, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 11981 + }, + { + "epoch": 9.570287539936102, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 11982 + }, + { + "epoch": 9.571086261980831, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 11983 + }, + { + "epoch": 9.571884984025559, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 11984 + }, + { + "epoch": 9.572683706070288, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 11985 + }, + { + "epoch": 9.573482428115016, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 11986 + }, + { + "epoch": 9.574281150159745, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 11987 + }, + { + "epoch": 9.575079872204473, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 11988 + }, + { + "epoch": 9.5758785942492, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 11989 + }, + { + "epoch": 9.57667731629393, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 11990 + }, + { + "epoch": 9.577476038338657, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 11991 + }, + { + "epoch": 9.578274760383387, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 11992 + }, + { + "epoch": 9.579073482428115, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 11993 + }, + { + "epoch": 9.579872204472844, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 11994 + }, + { + "epoch": 9.580670926517572, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 11995 + }, + { + "epoch": 9.581469648562301, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.037, + "step": 11996 + }, + { + "epoch": 9.582268370607029, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 11997 + }, + { + "epoch": 9.583067092651756, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 11998 + }, + { + "epoch": 9.583865814696486, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.017, + "step": 11999 + }, + { + "epoch": 9.584664536741213, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 12000 + }, + { + "epoch": 9.585463258785943, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 12001 + }, + { + "epoch": 9.58626198083067, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 12002 + }, + { + "epoch": 9.5870607028754, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 12003 + }, + { + "epoch": 9.587859424920127, + "grad_norm": 0.17578125, + "learning_rate": 0.0005, + "loss": 1.0361, + "step": 12004 + }, + { + "epoch": 9.588658146964857, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0377, + "step": 12005 + }, + { + "epoch": 9.589456869009584, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 12006 + }, + { + "epoch": 9.590255591054314, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 12007 + }, + { + "epoch": 9.591054313099042, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 12008 + }, + { + "epoch": 9.59185303514377, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 12009 + }, + { + "epoch": 9.592651757188499, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 12010 + }, + { + "epoch": 9.593450479233226, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 12011 + }, + { + "epoch": 9.594249201277956, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 12012 + }, + { + "epoch": 9.595047923322683, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 12013 + }, + { + "epoch": 9.595846645367413, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 12014 + }, + { + "epoch": 9.59664536741214, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 12015 + }, + { + "epoch": 9.59744408945687, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 12016 + }, + { + "epoch": 9.598242811501597, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 12017 + }, + { + "epoch": 9.599041533546325, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 12018 + }, + { + "epoch": 9.599840255591054, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 12019 + }, + { + "epoch": 9.600638977635782, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 12020 + }, + { + "epoch": 9.601437699680512, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 12021 + }, + { + "epoch": 9.60223642172524, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 12022 + }, + { + "epoch": 9.603035143769969, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 12023 + }, + { + "epoch": 9.603833865814696, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 12024 + }, + { + "epoch": 9.604632587859426, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 12025 + }, + { + "epoch": 9.605431309904153, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 12026 + }, + { + "epoch": 9.606230031948883, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 12027 + }, + { + "epoch": 9.60702875399361, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 12028 + }, + { + "epoch": 9.607827476038338, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 12029 + }, + { + "epoch": 9.608626198083067, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 12030 + }, + { + "epoch": 9.609424920127795, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 12031 + }, + { + "epoch": 9.610223642172524, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 12032 + }, + { + "epoch": 9.611022364217252, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0419, + "step": 12033 + }, + { + "epoch": 9.611821086261982, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 12034 + }, + { + "epoch": 9.61261980830671, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 12035 + }, + { + "epoch": 9.613418530351439, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 12036 + }, + { + "epoch": 9.614217252396166, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 12037 + }, + { + "epoch": 9.615015974440894, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 12038 + }, + { + "epoch": 9.615814696485623, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 12039 + }, + { + "epoch": 9.616613418530351, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 12040 + }, + { + "epoch": 9.61741214057508, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 12041 + }, + { + "epoch": 9.618210862619808, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 12042 + }, + { + "epoch": 9.619009584664537, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 12043 + }, + { + "epoch": 9.619808306709265, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.037, + "step": 12044 + }, + { + "epoch": 9.620607028753994, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 12045 + }, + { + "epoch": 9.621405750798722, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 12046 + }, + { + "epoch": 9.62220447284345, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 12047 + }, + { + "epoch": 9.62300319488818, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 12048 + }, + { + "epoch": 9.623801916932907, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 12049 + }, + { + "epoch": 9.624600638977636, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 12050 + }, + { + "epoch": 9.625399361022364, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 12051 + }, + { + "epoch": 9.626198083067093, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 12052 + }, + { + "epoch": 9.62699680511182, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 12053 + }, + { + "epoch": 9.62779552715655, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 12054 + }, + { + "epoch": 9.628594249201278, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 12055 + }, + { + "epoch": 9.629392971246006, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 12056 + }, + { + "epoch": 9.630191693290735, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 12057 + }, + { + "epoch": 9.630990415335463, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 12058 + }, + { + "epoch": 9.631789137380192, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 12059 + }, + { + "epoch": 9.63258785942492, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 12060 + }, + { + "epoch": 9.633386581469649, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0373, + "step": 12061 + }, + { + "epoch": 9.634185303514377, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 12062 + }, + { + "epoch": 9.634984025559106, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 12063 + }, + { + "epoch": 9.635782747603834, + "grad_norm": 0.1728515625, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 12064 + }, + { + "epoch": 9.636581469648561, + "grad_norm": 0.169921875, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 12065 + }, + { + "epoch": 9.63738019169329, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 12066 + }, + { + "epoch": 9.638178913738018, + "grad_norm": 0.173828125, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 12067 + }, + { + "epoch": 9.638977635782748, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 12068 + }, + { + "epoch": 9.639776357827476, + "grad_norm": 0.1630859375, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 12069 + }, + { + "epoch": 9.640575079872205, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 12070 + }, + { + "epoch": 9.641373801916933, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 12071 + }, + { + "epoch": 9.642172523961662, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0396, + "step": 12072 + }, + { + "epoch": 9.64297124600639, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 12073 + }, + { + "epoch": 9.643769968051117, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 12074 + }, + { + "epoch": 9.644568690095847, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 12075 + }, + { + "epoch": 9.645367412140574, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 12076 + }, + { + "epoch": 9.646166134185304, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 12077 + }, + { + "epoch": 9.646964856230031, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 12078 + }, + { + "epoch": 9.64776357827476, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0389, + "step": 12079 + }, + { + "epoch": 9.648562300319488, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 12080 + }, + { + "epoch": 9.649361022364218, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 12081 + }, + { + "epoch": 9.650159744408946, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 12082 + }, + { + "epoch": 9.650958466453675, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 12083 + }, + { + "epoch": 9.651757188498403, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 12084 + }, + { + "epoch": 9.65255591054313, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 12085 + }, + { + "epoch": 9.65335463258786, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 12086 + }, + { + "epoch": 9.654153354632587, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 12087 + }, + { + "epoch": 9.654952076677317, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 12088 + }, + { + "epoch": 9.655750798722044, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0409, + "step": 12089 + }, + { + "epoch": 9.656549520766774, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 12090 + }, + { + "epoch": 9.657348242811501, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 12091 + }, + { + "epoch": 9.65814696485623, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 12092 + }, + { + "epoch": 9.658945686900958, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 12093 + }, + { + "epoch": 9.659744408945686, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 12094 + }, + { + "epoch": 9.660543130990416, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 12095 + }, + { + "epoch": 9.661341853035143, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 12096 + }, + { + "epoch": 9.662140575079873, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 12097 + }, + { + "epoch": 9.6629392971246, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 12098 + }, + { + "epoch": 9.66373801916933, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 12099 + }, + { + "epoch": 9.664536741214057, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 12100 + }, + { + "epoch": 9.665335463258787, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 12101 + }, + { + "epoch": 9.666134185303514, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 12102 + }, + { + "epoch": 9.666932907348244, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 12103 + }, + { + "epoch": 9.667731629392971, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 12104 + }, + { + "epoch": 9.668530351437699, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 12105 + }, + { + "epoch": 9.669329073482428, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 12106 + }, + { + "epoch": 9.670127795527156, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 12107 + }, + { + "epoch": 9.670926517571885, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 12108 + }, + { + "epoch": 9.671725239616613, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0337, + "step": 12109 + }, + { + "epoch": 9.672523961661343, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 12110 + }, + { + "epoch": 9.67332268370607, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 12111 + }, + { + "epoch": 9.6741214057508, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 12112 + }, + { + "epoch": 9.674920127795527, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 12113 + }, + { + "epoch": 9.675718849840255, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 12114 + }, + { + "epoch": 9.676517571884984, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 12115 + }, + { + "epoch": 9.677316293929712, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 12116 + }, + { + "epoch": 9.678115015974441, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 12117 + }, + { + "epoch": 9.678913738019169, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 12118 + }, + { + "epoch": 9.679712460063898, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 12119 + }, + { + "epoch": 9.680511182108626, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 12120 + }, + { + "epoch": 9.681309904153355, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0169, + "step": 12121 + }, + { + "epoch": 9.682108626198083, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 12122 + }, + { + "epoch": 9.68290734824281, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0385, + "step": 12123 + }, + { + "epoch": 9.68370607028754, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 12124 + }, + { + "epoch": 9.684504792332268, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 12125 + }, + { + "epoch": 9.685303514376997, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0338, + "step": 12126 + }, + { + "epoch": 9.686102236421725, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 12127 + }, + { + "epoch": 9.686900958466454, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 12128 + }, + { + "epoch": 9.687699680511182, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 12129 + }, + { + "epoch": 9.688498402555911, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 12130 + }, + { + "epoch": 9.689297124600639, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 12131 + }, + { + "epoch": 9.690095846645367, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 12132 + }, + { + "epoch": 9.690894568690096, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0145, + "step": 12133 + }, + { + "epoch": 9.691693290734824, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 12134 + }, + { + "epoch": 9.692492012779553, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0384, + "step": 12135 + }, + { + "epoch": 9.69329073482428, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 12136 + }, + { + "epoch": 9.69408945686901, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 12137 + }, + { + "epoch": 9.694888178913738, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 12138 + }, + { + "epoch": 9.695686900958467, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 12139 + }, + { + "epoch": 9.696485623003195, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 12140 + }, + { + "epoch": 9.697284345047922, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 12141 + }, + { + "epoch": 9.698083067092652, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 12142 + }, + { + "epoch": 9.69888178913738, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 12143 + }, + { + "epoch": 9.699680511182109, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 12144 + }, + { + "epoch": 9.700479233226837, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 12145 + }, + { + "epoch": 9.701277955271566, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 12146 + }, + { + "epoch": 9.702076677316294, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 12147 + }, + { + "epoch": 9.702875399361023, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 12148 + }, + { + "epoch": 9.70367412140575, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 12149 + }, + { + "epoch": 9.704472843450478, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 12150 + }, + { + "epoch": 9.705271565495208, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 12151 + }, + { + "epoch": 9.706070287539935, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 12152 + }, + { + "epoch": 9.706869009584665, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 12153 + }, + { + "epoch": 9.707667731629392, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 12154 + }, + { + "epoch": 9.708466453674122, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 12155 + }, + { + "epoch": 9.70926517571885, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 12156 + }, + { + "epoch": 9.710063897763579, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 12157 + }, + { + "epoch": 9.710862619808307, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0172, + "step": 12158 + }, + { + "epoch": 9.711661341853034, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 12159 + }, + { + "epoch": 9.712460063897764, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 12160 + }, + { + "epoch": 9.713258785942491, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 12161 + }, + { + "epoch": 9.71405750798722, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0158, + "step": 12162 + }, + { + "epoch": 9.714856230031948, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 12163 + }, + { + "epoch": 9.715654952076678, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 12164 + }, + { + "epoch": 9.716453674121405, + "grad_norm": 0.2197265625, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 12165 + }, + { + "epoch": 9.717252396166135, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 12166 + }, + { + "epoch": 9.718051118210862, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 12167 + }, + { + "epoch": 9.718849840255592, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0337, + "step": 12168 + }, + { + "epoch": 9.71964856230032, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 12169 + }, + { + "epoch": 9.720447284345047, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 12170 + }, + { + "epoch": 9.721246006389777, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 12171 + }, + { + "epoch": 9.722044728434504, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 12172 + }, + { + "epoch": 9.722843450479234, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 12173 + }, + { + "epoch": 9.723642172523961, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 12174 + }, + { + "epoch": 9.72444089456869, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 12175 + }, + { + "epoch": 9.725239616613418, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 12176 + }, + { + "epoch": 9.726038338658148, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 12177 + }, + { + "epoch": 9.726837060702875, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 12178 + }, + { + "epoch": 9.727635782747605, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 12179 + }, + { + "epoch": 9.728434504792332, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 12180 + }, + { + "epoch": 9.72923322683706, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 12181 + }, + { + "epoch": 9.73003194888179, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 12182 + }, + { + "epoch": 9.730830670926517, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 12183 + }, + { + "epoch": 9.731629392971247, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 12184 + }, + { + "epoch": 9.732428115015974, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 12185 + }, + { + "epoch": 9.733226837060704, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 12186 + }, + { + "epoch": 9.734025559105431, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 12187 + }, + { + "epoch": 9.73482428115016, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0348, + "step": 12188 + }, + { + "epoch": 9.735623003194888, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 12189 + }, + { + "epoch": 9.736421725239616, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 12190 + }, + { + "epoch": 9.737220447284345, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 12191 + }, + { + "epoch": 9.738019169329073, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 12192 + }, + { + "epoch": 9.738817891373802, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 12193 + }, + { + "epoch": 9.73961661341853, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 12194 + }, + { + "epoch": 9.74041533546326, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 12195 + }, + { + "epoch": 9.741214057507987, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0337, + "step": 12196 + }, + { + "epoch": 9.742012779552716, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 12197 + }, + { + "epoch": 9.742811501597444, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0361, + "step": 12198 + }, + { + "epoch": 9.743610223642172, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0402, + "step": 12199 + }, + { + "epoch": 9.744408945686901, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 12200 + }, + { + "epoch": 9.745207667731629, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 12201 + }, + { + "epoch": 9.746006389776358, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 12202 + }, + { + "epoch": 9.746805111821086, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 12203 + }, + { + "epoch": 9.747603833865815, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 12204 + }, + { + "epoch": 9.748402555910543, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 12205 + }, + { + "epoch": 9.749201277955272, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 12206 + }, + { + "epoch": 9.75, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 12207 + }, + { + "epoch": 9.750798722044728, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 12208 + }, + { + "epoch": 9.751597444089457, + "grad_norm": 0.169921875, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 12209 + }, + { + "epoch": 9.752396166134185, + "grad_norm": 0.1962890625, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 12210 + }, + { + "epoch": 9.753194888178914, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 12211 + }, + { + "epoch": 9.753993610223642, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 12212 + }, + { + "epoch": 9.754792332268371, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 12213 + }, + { + "epoch": 9.755591054313099, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 12214 + }, + { + "epoch": 9.756389776357828, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 12215 + }, + { + "epoch": 9.757188498402556, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 12216 + }, + { + "epoch": 9.757987220447284, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 12217 + }, + { + "epoch": 9.758785942492013, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 12218 + }, + { + "epoch": 9.75958466453674, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 12219 + }, + { + "epoch": 9.76038338658147, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 12220 + }, + { + "epoch": 9.761182108626198, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0373, + "step": 12221 + }, + { + "epoch": 9.761980830670927, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 12222 + }, + { + "epoch": 9.762779552715655, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 12223 + }, + { + "epoch": 9.763578274760384, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 12224 + }, + { + "epoch": 9.764376996805112, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 12225 + }, + { + "epoch": 9.76517571884984, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 12226 + }, + { + "epoch": 9.765974440894569, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 12227 + }, + { + "epoch": 9.766773162939296, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 12228 + }, + { + "epoch": 9.767571884984026, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 12229 + }, + { + "epoch": 9.768370607028753, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 12230 + }, + { + "epoch": 9.769169329073483, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 12231 + }, + { + "epoch": 9.76996805111821, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 12232 + }, + { + "epoch": 9.77076677316294, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 12233 + }, + { + "epoch": 9.771565495207668, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 12234 + }, + { + "epoch": 9.772364217252395, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 12235 + }, + { + "epoch": 9.773162939297125, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 12236 + }, + { + "epoch": 9.773961661341852, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.015, + "step": 12237 + }, + { + "epoch": 9.774760383386582, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 12238 + }, + { + "epoch": 9.77555910543131, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 12239 + }, + { + "epoch": 9.776357827476039, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 12240 + }, + { + "epoch": 9.777156549520766, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 12241 + }, + { + "epoch": 9.777955271565496, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 12242 + }, + { + "epoch": 9.778753993610223, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 12243 + }, + { + "epoch": 9.779552715654953, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 12244 + }, + { + "epoch": 9.78035143769968, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 12245 + }, + { + "epoch": 9.781150159744408, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 12246 + }, + { + "epoch": 9.781948881789138, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 12247 + }, + { + "epoch": 9.782747603833865, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 12248 + }, + { + "epoch": 9.783546325878595, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 12249 + }, + { + "epoch": 9.784345047923322, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 12250 + }, + { + "epoch": 9.785143769968052, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 12251 + }, + { + "epoch": 9.78594249201278, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 12252 + }, + { + "epoch": 9.786741214057509, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 12253 + }, + { + "epoch": 9.787539936102236, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0375, + "step": 12254 + }, + { + "epoch": 9.788338658146966, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 12255 + }, + { + "epoch": 9.789137380191693, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 12256 + }, + { + "epoch": 9.789936102236421, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 12257 + }, + { + "epoch": 9.79073482428115, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0369, + "step": 12258 + }, + { + "epoch": 9.791533546325878, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 12259 + }, + { + "epoch": 9.792332268370608, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 12260 + }, + { + "epoch": 9.793130990415335, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 12261 + }, + { + "epoch": 9.793929712460065, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 12262 + }, + { + "epoch": 9.794728434504792, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 12263 + }, + { + "epoch": 9.795527156549522, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 12264 + }, + { + "epoch": 9.79632587859425, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 12265 + }, + { + "epoch": 9.797124600638977, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0384, + "step": 12266 + }, + { + "epoch": 9.797923322683706, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0416, + "step": 12267 + }, + { + "epoch": 9.798722044728434, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 12268 + }, + { + "epoch": 9.799520766773163, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 12269 + }, + { + "epoch": 9.800319488817891, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 12270 + }, + { + "epoch": 9.80111821086262, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 12271 + }, + { + "epoch": 9.801916932907348, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 12272 + }, + { + "epoch": 9.802715654952078, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 12273 + }, + { + "epoch": 9.803514376996805, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 12274 + }, + { + "epoch": 9.804313099041533, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 12275 + }, + { + "epoch": 9.805111821086262, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 12276 + }, + { + "epoch": 9.80591054313099, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 12277 + }, + { + "epoch": 9.80670926517572, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 12278 + }, + { + "epoch": 9.807507987220447, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 12279 + }, + { + "epoch": 9.808306709265176, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 12280 + }, + { + "epoch": 9.809105431309904, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 12281 + }, + { + "epoch": 9.809904153354633, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 12282 + }, + { + "epoch": 9.810702875399361, + "grad_norm": 0.181640625, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 12283 + }, + { + "epoch": 9.811501597444089, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 12284 + }, + { + "epoch": 9.812300319488818, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 12285 + }, + { + "epoch": 9.813099041533546, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 12286 + }, + { + "epoch": 9.813897763578275, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 12287 + }, + { + "epoch": 9.814696485623003, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0148, + "step": 12288 + }, + { + "epoch": 9.815495207667732, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 12289 + }, + { + "epoch": 9.81629392971246, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 12290 + }, + { + "epoch": 9.81709265175719, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 12291 + }, + { + "epoch": 9.817891373801917, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 12292 + }, + { + "epoch": 9.818690095846645, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 12293 + }, + { + "epoch": 9.819488817891374, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 12294 + }, + { + "epoch": 9.820287539936102, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 12295 + }, + { + "epoch": 9.821086261980831, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 12296 + }, + { + "epoch": 9.821884984025559, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 12297 + }, + { + "epoch": 9.822683706070288, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 12298 + }, + { + "epoch": 9.823482428115016, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 12299 + }, + { + "epoch": 9.824281150159745, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 12300 + }, + { + "epoch": 9.825079872204473, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 12301 + }, + { + "epoch": 9.8258785942492, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 12302 + }, + { + "epoch": 9.82667731629393, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 12303 + }, + { + "epoch": 9.827476038338657, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 12304 + }, + { + "epoch": 9.828274760383387, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 12305 + }, + { + "epoch": 9.829073482428115, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 12306 + }, + { + "epoch": 9.829872204472844, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 12307 + }, + { + "epoch": 9.830670926517572, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 12308 + }, + { + "epoch": 9.831469648562301, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 12309 + }, + { + "epoch": 9.832268370607029, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 12310 + }, + { + "epoch": 9.833067092651756, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 12311 + }, + { + "epoch": 9.833865814696486, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 12312 + }, + { + "epoch": 9.834664536741213, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 12313 + }, + { + "epoch": 9.835463258785943, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 12314 + }, + { + "epoch": 9.83626198083067, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 12315 + }, + { + "epoch": 9.8370607028754, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 12316 + }, + { + "epoch": 9.837859424920127, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 12317 + }, + { + "epoch": 9.838658146964857, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0369, + "step": 12318 + }, + { + "epoch": 9.839456869009584, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0337, + "step": 12319 + }, + { + "epoch": 9.840255591054314, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 12320 + }, + { + "epoch": 9.841054313099042, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 12321 + }, + { + "epoch": 9.84185303514377, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 12322 + }, + { + "epoch": 9.842651757188499, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 12323 + }, + { + "epoch": 9.843450479233226, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 12324 + }, + { + "epoch": 9.844249201277956, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 12325 + }, + { + "epoch": 9.845047923322683, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 12326 + }, + { + "epoch": 9.845846645367413, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 12327 + }, + { + "epoch": 9.84664536741214, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 12328 + }, + { + "epoch": 9.84744408945687, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0337, + "step": 12329 + }, + { + "epoch": 9.848242811501597, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.034, + "step": 12330 + }, + { + "epoch": 9.849041533546325, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0439, + "step": 12331 + }, + { + "epoch": 9.849840255591054, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 12332 + }, + { + "epoch": 9.850638977635782, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 12333 + }, + { + "epoch": 9.851437699680512, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 12334 + }, + { + "epoch": 9.85223642172524, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 12335 + }, + { + "epoch": 9.853035143769969, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 12336 + }, + { + "epoch": 9.853833865814696, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 12337 + }, + { + "epoch": 9.854632587859426, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 12338 + }, + { + "epoch": 9.855431309904153, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 12339 + }, + { + "epoch": 9.856230031948883, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0169, + "step": 12340 + }, + { + "epoch": 9.85702875399361, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 12341 + }, + { + "epoch": 9.857827476038338, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 12342 + }, + { + "epoch": 9.858626198083067, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 12343 + }, + { + "epoch": 9.859424920127795, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 12344 + }, + { + "epoch": 9.860223642172524, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 12345 + }, + { + "epoch": 9.861022364217252, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 12346 + }, + { + "epoch": 9.861821086261982, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 12347 + }, + { + "epoch": 9.86261980830671, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 12348 + }, + { + "epoch": 9.863418530351439, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 12349 + }, + { + "epoch": 9.864217252396166, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 12350 + }, + { + "epoch": 9.865015974440894, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 12351 + }, + { + "epoch": 9.865814696485623, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 12352 + }, + { + "epoch": 9.866613418530351, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 12353 + }, + { + "epoch": 9.86741214057508, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 12354 + }, + { + "epoch": 9.868210862619808, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 12355 + }, + { + "epoch": 9.869009584664537, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 12356 + }, + { + "epoch": 9.869808306709265, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 12357 + }, + { + "epoch": 9.870607028753994, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 12358 + }, + { + "epoch": 9.871405750798722, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 12359 + }, + { + "epoch": 9.87220447284345, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0401, + "step": 12360 + }, + { + "epoch": 9.87300319488818, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 12361 + }, + { + "epoch": 9.873801916932907, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 12362 + }, + { + "epoch": 9.874600638977636, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 12363 + }, + { + "epoch": 9.875399361022364, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 12364 + }, + { + "epoch": 9.876198083067093, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 12365 + }, + { + "epoch": 9.87699680511182, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 12366 + }, + { + "epoch": 9.87779552715655, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 12367 + }, + { + "epoch": 9.878594249201278, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 12368 + }, + { + "epoch": 9.879392971246006, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 12369 + }, + { + "epoch": 9.880191693290735, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 12370 + }, + { + "epoch": 9.880990415335463, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 12371 + }, + { + "epoch": 9.881789137380192, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 12372 + }, + { + "epoch": 9.88258785942492, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 12373 + }, + { + "epoch": 9.883386581469649, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 12374 + }, + { + "epoch": 9.884185303514377, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 12375 + }, + { + "epoch": 9.884984025559106, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 12376 + }, + { + "epoch": 9.885782747603834, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 12377 + }, + { + "epoch": 9.886581469648561, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 12378 + }, + { + "epoch": 9.88738019169329, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 12379 + }, + { + "epoch": 9.888178913738018, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 12380 + }, + { + "epoch": 9.888977635782748, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 12381 + }, + { + "epoch": 9.889776357827476, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 12382 + }, + { + "epoch": 9.890575079872205, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 12383 + }, + { + "epoch": 9.891373801916933, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 12384 + }, + { + "epoch": 9.892172523961662, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 12385 + }, + { + "epoch": 9.89297124600639, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 12386 + }, + { + "epoch": 9.893769968051117, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 12387 + }, + { + "epoch": 9.894568690095847, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0143, + "step": 12388 + }, + { + "epoch": 9.895367412140574, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 12389 + }, + { + "epoch": 9.896166134185304, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 12390 + }, + { + "epoch": 9.896964856230031, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 12391 + }, + { + "epoch": 9.89776357827476, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 12392 + }, + { + "epoch": 9.898562300319488, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 12393 + }, + { + "epoch": 9.899361022364218, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 12394 + }, + { + "epoch": 9.900159744408946, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 12395 + }, + { + "epoch": 9.900958466453675, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 12396 + }, + { + "epoch": 9.901757188498403, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0353, + "step": 12397 + }, + { + "epoch": 9.90255591054313, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 12398 + }, + { + "epoch": 9.90335463258786, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 12399 + }, + { + "epoch": 9.904153354632587, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 12400 + }, + { + "epoch": 9.904952076677317, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 12401 + }, + { + "epoch": 9.905750798722044, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 12402 + }, + { + "epoch": 9.906549520766774, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 12403 + }, + { + "epoch": 9.907348242811501, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 12404 + }, + { + "epoch": 9.90814696485623, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 12405 + }, + { + "epoch": 9.908945686900958, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 12406 + }, + { + "epoch": 9.909744408945686, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 12407 + }, + { + "epoch": 9.910543130990416, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 12408 + }, + { + "epoch": 9.911341853035143, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 12409 + }, + { + "epoch": 9.912140575079873, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 12410 + }, + { + "epoch": 9.9129392971246, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0157, + "step": 12411 + }, + { + "epoch": 9.91373801916933, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 12412 + }, + { + "epoch": 9.914536741214057, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 12413 + }, + { + "epoch": 9.915335463258787, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 12414 + }, + { + "epoch": 9.916134185303514, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0181, + "step": 12415 + }, + { + "epoch": 9.916932907348244, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 12416 + }, + { + "epoch": 9.917731629392971, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0168, + "step": 12417 + }, + { + "epoch": 9.918530351437699, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0188, + "step": 12418 + }, + { + "epoch": 9.919329073482428, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 12419 + }, + { + "epoch": 9.920127795527156, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 12420 + }, + { + "epoch": 9.920926517571885, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 12421 + }, + { + "epoch": 9.921725239616613, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 12422 + }, + { + "epoch": 9.922523961661343, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 12423 + }, + { + "epoch": 9.92332268370607, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 12424 + }, + { + "epoch": 9.9241214057508, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 12425 + }, + { + "epoch": 9.924920127795527, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 12426 + }, + { + "epoch": 9.925718849840255, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0337, + "step": 12427 + }, + { + "epoch": 9.926517571884984, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 12428 + }, + { + "epoch": 9.927316293929712, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 12429 + }, + { + "epoch": 9.928115015974441, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 12430 + }, + { + "epoch": 9.928913738019169, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 12431 + }, + { + "epoch": 9.929712460063898, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 12432 + }, + { + "epoch": 9.930511182108626, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0174, + "step": 12433 + }, + { + "epoch": 9.931309904153355, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 12434 + }, + { + "epoch": 9.932108626198083, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 12435 + }, + { + "epoch": 9.93290734824281, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 12436 + }, + { + "epoch": 9.93370607028754, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 12437 + }, + { + "epoch": 9.934504792332268, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 12438 + }, + { + "epoch": 9.935303514376997, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 12439 + }, + { + "epoch": 9.936102236421725, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 12440 + }, + { + "epoch": 9.936900958466454, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 12441 + }, + { + "epoch": 9.937699680511182, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 12442 + }, + { + "epoch": 9.938498402555911, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 12443 + }, + { + "epoch": 9.939297124600639, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 12444 + }, + { + "epoch": 9.940095846645367, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 12445 + }, + { + "epoch": 9.940894568690096, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 12446 + }, + { + "epoch": 9.941693290734824, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 12447 + }, + { + "epoch": 9.942492012779553, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 12448 + }, + { + "epoch": 9.94329073482428, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 12449 + }, + { + "epoch": 9.94408945686901, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 12450 + }, + { + "epoch": 9.944888178913738, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 12451 + }, + { + "epoch": 9.945686900958467, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 12452 + }, + { + "epoch": 9.946485623003195, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 12453 + }, + { + "epoch": 9.947284345047922, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 12454 + }, + { + "epoch": 9.948083067092652, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0164, + "step": 12455 + }, + { + "epoch": 9.94888178913738, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 12456 + }, + { + "epoch": 9.949680511182109, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 12457 + }, + { + "epoch": 9.950479233226837, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 12458 + }, + { + "epoch": 9.951277955271566, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 12459 + }, + { + "epoch": 9.952076677316294, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 12460 + }, + { + "epoch": 9.952875399361023, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 12461 + }, + { + "epoch": 9.95367412140575, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 12462 + }, + { + "epoch": 9.954472843450478, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 12463 + }, + { + "epoch": 9.955271565495208, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 12464 + }, + { + "epoch": 9.956070287539935, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 12465 + }, + { + "epoch": 9.956869009584665, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 12466 + }, + { + "epoch": 9.957667731629392, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 12467 + }, + { + "epoch": 9.958466453674122, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 12468 + }, + { + "epoch": 9.95926517571885, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 12469 + }, + { + "epoch": 9.960063897763579, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 12470 + }, + { + "epoch": 9.960862619808307, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 12471 + }, + { + "epoch": 9.961661341853034, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 12472 + }, + { + "epoch": 9.962460063897764, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 12473 + }, + { + "epoch": 9.963258785942491, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 12474 + }, + { + "epoch": 9.96405750798722, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 12475 + }, + { + "epoch": 9.964856230031948, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0161, + "step": 12476 + }, + { + "epoch": 9.965654952076678, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 12477 + }, + { + "epoch": 9.966453674121405, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 12478 + }, + { + "epoch": 9.967252396166135, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 12479 + }, + { + "epoch": 9.968051118210862, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 12480 + }, + { + "epoch": 9.968849840255592, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 12481 + }, + { + "epoch": 9.96964856230032, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 12482 + }, + { + "epoch": 9.970447284345047, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 12483 + }, + { + "epoch": 9.971246006389777, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 12484 + }, + { + "epoch": 9.972044728434504, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 12485 + }, + { + "epoch": 9.972843450479234, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 12486 + }, + { + "epoch": 9.973642172523961, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 12487 + }, + { + "epoch": 9.97444089456869, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 12488 + }, + { + "epoch": 9.975239616613418, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 12489 + }, + { + "epoch": 9.976038338658148, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 12490 + }, + { + "epoch": 9.976837060702875, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 12491 + }, + { + "epoch": 9.977635782747605, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 12492 + }, + { + "epoch": 9.978434504792332, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0159, + "step": 12493 + }, + { + "epoch": 9.97923322683706, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 12494 + }, + { + "epoch": 9.98003194888179, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 12495 + }, + { + "epoch": 9.980830670926517, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0369, + "step": 12496 + }, + { + "epoch": 9.981629392971247, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 12497 + }, + { + "epoch": 9.982428115015974, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 12498 + }, + { + "epoch": 9.983226837060704, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 12499 + }, + { + "epoch": 9.984025559105431, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 12500 + }, + { + "epoch": 9.98482428115016, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0177, + "step": 12501 + }, + { + "epoch": 9.985623003194888, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 12502 + }, + { + "epoch": 9.986421725239616, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 12503 + }, + { + "epoch": 9.987220447284345, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 12504 + }, + { + "epoch": 9.988019169329073, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 12505 + }, + { + "epoch": 9.988817891373802, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 12506 + }, + { + "epoch": 9.98961661341853, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 12507 + }, + { + "epoch": 9.99041533546326, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 12508 + }, + { + "epoch": 9.991214057507987, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 12509 + }, + { + "epoch": 9.992012779552716, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 12510 + }, + { + "epoch": 9.992811501597444, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 12511 + }, + { + "epoch": 9.993610223642172, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 12512 + }, + { + "epoch": 9.994408945686901, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 12513 + }, + { + "epoch": 9.995207667731629, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.018, + "step": 12514 + }, + { + "epoch": 9.996006389776358, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 12515 + }, + { + "epoch": 9.996805111821086, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 12516 + }, + { + "epoch": 9.997603833865815, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 12517 + }, + { + "epoch": 9.998402555910543, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 12518 + }, + { + "epoch": 9.999201277955272, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 12519 + }, + { + "epoch": 10.0, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 12520 + }, + { + "epoch": 10.000798722044728, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 12521 + }, + { + "epoch": 10.001597444089457, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 12522 + }, + { + "epoch": 10.002396166134185, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0192, + "step": 12523 + }, + { + "epoch": 10.003194888178914, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 12524 + }, + { + "epoch": 10.003993610223642, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 12525 + }, + { + "epoch": 10.004792332268371, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 12526 + }, + { + "epoch": 10.005591054313099, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 12527 + }, + { + "epoch": 10.006389776357828, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 12528 + }, + { + "epoch": 10.007188498402556, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 12529 + }, + { + "epoch": 10.007987220447284, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 12530 + }, + { + "epoch": 10.008785942492013, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 12531 + }, + { + "epoch": 10.00958466453674, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 12532 + }, + { + "epoch": 10.01038338658147, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 12533 + }, + { + "epoch": 10.011182108626198, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0393, + "step": 12534 + }, + { + "epoch": 10.011980830670927, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 12535 + }, + { + "epoch": 10.012779552715655, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 12536 + }, + { + "epoch": 10.013578274760384, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 12537 + }, + { + "epoch": 10.014376996805112, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 12538 + }, + { + "epoch": 10.01517571884984, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 12539 + }, + { + "epoch": 10.015974440894569, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 12540 + }, + { + "epoch": 10.016773162939296, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 12541 + }, + { + "epoch": 10.017571884984026, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 12542 + }, + { + "epoch": 10.018370607028753, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 12543 + }, + { + "epoch": 10.019169329073483, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0134, + "step": 12544 + }, + { + "epoch": 10.01996805111821, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 12545 + }, + { + "epoch": 10.02076677316294, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 12546 + }, + { + "epoch": 10.021565495207668, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 12547 + }, + { + "epoch": 10.022364217252397, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 12548 + }, + { + "epoch": 10.023162939297125, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 12549 + }, + { + "epoch": 10.023961661341852, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 12550 + }, + { + "epoch": 10.024760383386582, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 12551 + }, + { + "epoch": 10.02555910543131, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 12552 + }, + { + "epoch": 10.026357827476039, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 12553 + }, + { + "epoch": 10.027156549520766, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 12554 + }, + { + "epoch": 10.027955271565496, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 12555 + }, + { + "epoch": 10.028753993610223, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 12556 + }, + { + "epoch": 10.029552715654953, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0175, + "step": 12557 + }, + { + "epoch": 10.03035143769968, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 12558 + }, + { + "epoch": 10.031150159744408, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0379, + "step": 12559 + }, + { + "epoch": 10.031948881789138, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 12560 + }, + { + "epoch": 10.032747603833865, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0361, + "step": 12561 + }, + { + "epoch": 10.033546325878595, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 12562 + }, + { + "epoch": 10.034345047923322, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 12563 + }, + { + "epoch": 10.035143769968052, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 12564 + }, + { + "epoch": 10.03594249201278, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 12565 + }, + { + "epoch": 10.036741214057509, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 12566 + }, + { + "epoch": 10.037539936102236, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 12567 + }, + { + "epoch": 10.038338658146964, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 12568 + }, + { + "epoch": 10.039137380191693, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 12569 + }, + { + "epoch": 10.039936102236421, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 12570 + }, + { + "epoch": 10.04073482428115, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 12571 + }, + { + "epoch": 10.041533546325878, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 12572 + }, + { + "epoch": 10.042332268370608, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 12573 + }, + { + "epoch": 10.043130990415335, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 12574 + }, + { + "epoch": 10.043929712460065, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 12575 + }, + { + "epoch": 10.044728434504792, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 12576 + }, + { + "epoch": 10.04552715654952, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 12577 + }, + { + "epoch": 10.04632587859425, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 12578 + }, + { + "epoch": 10.047124600638977, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0406, + "step": 12579 + }, + { + "epoch": 10.047923322683706, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 12580 + }, + { + "epoch": 10.048722044728434, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 12581 + }, + { + "epoch": 10.049520766773163, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 12582 + }, + { + "epoch": 10.050319488817891, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0174, + "step": 12583 + }, + { + "epoch": 10.05111821086262, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 12584 + }, + { + "epoch": 10.051916932907348, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 12585 + }, + { + "epoch": 10.052715654952078, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0173, + "step": 12586 + }, + { + "epoch": 10.053514376996805, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 12587 + }, + { + "epoch": 10.054313099041533, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 12588 + }, + { + "epoch": 10.055111821086262, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 12589 + }, + { + "epoch": 10.05591054313099, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0442, + "step": 12590 + }, + { + "epoch": 10.05670926517572, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 12591 + }, + { + "epoch": 10.057507987220447, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 12592 + }, + { + "epoch": 10.058306709265176, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 12593 + }, + { + "epoch": 10.059105431309904, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 12594 + }, + { + "epoch": 10.059904153354633, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 12595 + }, + { + "epoch": 10.060702875399361, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 12596 + }, + { + "epoch": 10.061501597444089, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 12597 + }, + { + "epoch": 10.062300319488818, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 12598 + }, + { + "epoch": 10.063099041533546, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 12599 + }, + { + "epoch": 10.063897763578275, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 12600 + }, + { + "epoch": 10.064696485623003, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 12601 + }, + { + "epoch": 10.065495207667732, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 12602 + }, + { + "epoch": 10.06629392971246, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 12603 + }, + { + "epoch": 10.06709265175719, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 12604 + }, + { + "epoch": 10.067891373801917, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 12605 + }, + { + "epoch": 10.068690095846645, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 12606 + }, + { + "epoch": 10.069488817891374, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 12607 + }, + { + "epoch": 10.070287539936102, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 12608 + }, + { + "epoch": 10.071086261980831, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0362, + "step": 12609 + }, + { + "epoch": 10.071884984025559, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 12610 + }, + { + "epoch": 10.072683706070288, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 12611 + }, + { + "epoch": 10.073482428115016, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 12612 + }, + { + "epoch": 10.074281150159745, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0413, + "step": 12613 + }, + { + "epoch": 10.075079872204473, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 12614 + }, + { + "epoch": 10.0758785942492, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 12615 + }, + { + "epoch": 10.07667731629393, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 12616 + }, + { + "epoch": 10.077476038338657, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 12617 + }, + { + "epoch": 10.078274760383387, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 12618 + }, + { + "epoch": 10.079073482428115, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 12619 + }, + { + "epoch": 10.079872204472844, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 12620 + }, + { + "epoch": 10.080670926517572, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 12621 + }, + { + "epoch": 10.081469648562301, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 12622 + }, + { + "epoch": 10.082268370607029, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 12623 + }, + { + "epoch": 10.083067092651758, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 12624 + }, + { + "epoch": 10.083865814696486, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 12625 + }, + { + "epoch": 10.084664536741213, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 12626 + }, + { + "epoch": 10.085463258785943, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 12627 + }, + { + "epoch": 10.08626198083067, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 12628 + }, + { + "epoch": 10.0870607028754, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 12629 + }, + { + "epoch": 10.087859424920127, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 12630 + }, + { + "epoch": 10.088658146964857, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 12631 + }, + { + "epoch": 10.089456869009584, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 12632 + }, + { + "epoch": 10.090255591054314, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 12633 + }, + { + "epoch": 10.091054313099042, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 12634 + }, + { + "epoch": 10.09185303514377, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 12635 + }, + { + "epoch": 10.092651757188499, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0176, + "step": 12636 + }, + { + "epoch": 10.093450479233226, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0166, + "step": 12637 + }, + { + "epoch": 10.094249201277956, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0192, + "step": 12638 + }, + { + "epoch": 10.095047923322683, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 12639 + }, + { + "epoch": 10.095846645367413, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 12640 + }, + { + "epoch": 10.09664536741214, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 12641 + }, + { + "epoch": 10.09744408945687, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 12642 + }, + { + "epoch": 10.098242811501597, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 12643 + }, + { + "epoch": 10.099041533546325, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.017, + "step": 12644 + }, + { + "epoch": 10.099840255591054, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 12645 + }, + { + "epoch": 10.100638977635782, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 12646 + }, + { + "epoch": 10.101437699680512, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 12647 + }, + { + "epoch": 10.10223642172524, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 12648 + }, + { + "epoch": 10.103035143769969, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 12649 + }, + { + "epoch": 10.103833865814696, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.042, + "step": 12650 + }, + { + "epoch": 10.104632587859426, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 12651 + }, + { + "epoch": 10.105431309904153, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 12652 + }, + { + "epoch": 10.106230031948881, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 12653 + }, + { + "epoch": 10.10702875399361, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 12654 + }, + { + "epoch": 10.107827476038338, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 12655 + }, + { + "epoch": 10.108626198083067, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 12656 + }, + { + "epoch": 10.109424920127795, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 12657 + }, + { + "epoch": 10.110223642172524, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 12658 + }, + { + "epoch": 10.111022364217252, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 12659 + }, + { + "epoch": 10.111821086261982, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0183, + "step": 12660 + }, + { + "epoch": 10.11261980830671, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 12661 + }, + { + "epoch": 10.113418530351439, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 12662 + }, + { + "epoch": 10.114217252396166, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 12663 + }, + { + "epoch": 10.115015974440894, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 12664 + }, + { + "epoch": 10.115814696485623, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 12665 + }, + { + "epoch": 10.116613418530351, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 12666 + }, + { + "epoch": 10.11741214057508, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 12667 + }, + { + "epoch": 10.118210862619808, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 12668 + }, + { + "epoch": 10.119009584664537, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 12669 + }, + { + "epoch": 10.119808306709265, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 12670 + }, + { + "epoch": 10.120607028753994, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 12671 + }, + { + "epoch": 10.121405750798722, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 12672 + }, + { + "epoch": 10.12220447284345, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 12673 + }, + { + "epoch": 10.12300319488818, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 12674 + }, + { + "epoch": 10.123801916932907, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 12675 + }, + { + "epoch": 10.124600638977636, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 12676 + }, + { + "epoch": 10.125399361022364, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 12677 + }, + { + "epoch": 10.126198083067093, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 12678 + }, + { + "epoch": 10.12699680511182, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 12679 + }, + { + "epoch": 10.12779552715655, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 12680 + }, + { + "epoch": 10.128594249201278, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 12681 + }, + { + "epoch": 10.129392971246006, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 12682 + }, + { + "epoch": 10.130191693290735, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 12683 + }, + { + "epoch": 10.130990415335463, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 12684 + }, + { + "epoch": 10.131789137380192, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 12685 + }, + { + "epoch": 10.13258785942492, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 12686 + }, + { + "epoch": 10.133386581469649, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 12687 + }, + { + "epoch": 10.134185303514377, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 12688 + }, + { + "epoch": 10.134984025559106, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 12689 + }, + { + "epoch": 10.135782747603834, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 12690 + }, + { + "epoch": 10.136581469648561, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 12691 + }, + { + "epoch": 10.13738019169329, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 12692 + }, + { + "epoch": 10.138178913738018, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 12693 + }, + { + "epoch": 10.138977635782748, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 12694 + }, + { + "epoch": 10.139776357827476, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 12695 + }, + { + "epoch": 10.140575079872205, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 12696 + }, + { + "epoch": 10.141373801916933, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 12697 + }, + { + "epoch": 10.142172523961662, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 12698 + }, + { + "epoch": 10.14297124600639, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 12699 + }, + { + "epoch": 10.143769968051119, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 12700 + }, + { + "epoch": 10.144568690095847, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 12701 + }, + { + "epoch": 10.145367412140574, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 12702 + }, + { + "epoch": 10.146166134185304, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 12703 + }, + { + "epoch": 10.146964856230031, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 12704 + }, + { + "epoch": 10.14776357827476, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 12705 + }, + { + "epoch": 10.148562300319488, + "grad_norm": 0.054931640625, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 12706 + }, + { + "epoch": 10.149361022364218, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 12707 + }, + { + "epoch": 10.150159744408946, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 12708 + }, + { + "epoch": 10.150958466453675, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 12709 + }, + { + "epoch": 10.151757188498403, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 12710 + }, + { + "epoch": 10.15255591054313, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0125, + "step": 12711 + }, + { + "epoch": 10.15335463258786, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0337, + "step": 12712 + }, + { + "epoch": 10.154153354632587, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0338, + "step": 12713 + }, + { + "epoch": 10.154952076677317, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0337, + "step": 12714 + }, + { + "epoch": 10.155750798722044, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 12715 + }, + { + "epoch": 10.156549520766774, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 12716 + }, + { + "epoch": 10.157348242811501, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 12717 + }, + { + "epoch": 10.15814696485623, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 12718 + }, + { + "epoch": 10.158945686900958, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0362, + "step": 12719 + }, + { + "epoch": 10.159744408945686, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 12720 + }, + { + "epoch": 10.160543130990416, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 12721 + }, + { + "epoch": 10.161341853035143, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 12722 + }, + { + "epoch": 10.162140575079873, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 12723 + }, + { + "epoch": 10.1629392971246, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 12724 + }, + { + "epoch": 10.16373801916933, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 12725 + }, + { + "epoch": 10.164536741214057, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 12726 + }, + { + "epoch": 10.165335463258787, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 12727 + }, + { + "epoch": 10.166134185303514, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 12728 + }, + { + "epoch": 10.166932907348242, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 12729 + }, + { + "epoch": 10.167731629392971, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 12730 + }, + { + "epoch": 10.168530351437699, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 12731 + }, + { + "epoch": 10.169329073482428, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 12732 + }, + { + "epoch": 10.170127795527156, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 12733 + }, + { + "epoch": 10.170926517571885, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 12734 + }, + { + "epoch": 10.171725239616613, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 12735 + }, + { + "epoch": 10.172523961661343, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0151, + "step": 12736 + }, + { + "epoch": 10.17332268370607, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 12737 + }, + { + "epoch": 10.1741214057508, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 12738 + }, + { + "epoch": 10.174920127795527, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 12739 + }, + { + "epoch": 10.175718849840255, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 12740 + }, + { + "epoch": 10.176517571884984, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 12741 + }, + { + "epoch": 10.177316293929712, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 12742 + }, + { + "epoch": 10.178115015974441, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 12743 + }, + { + "epoch": 10.178913738019169, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 12744 + }, + { + "epoch": 10.179712460063898, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 12745 + }, + { + "epoch": 10.180511182108626, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 12746 + }, + { + "epoch": 10.181309904153355, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 12747 + }, + { + "epoch": 10.182108626198083, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 12748 + }, + { + "epoch": 10.18290734824281, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 12749 + }, + { + "epoch": 10.18370607028754, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0443, + "step": 12750 + }, + { + "epoch": 10.184504792332268, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 12751 + }, + { + "epoch": 10.185303514376997, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 12752 + }, + { + "epoch": 10.186102236421725, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 12753 + }, + { + "epoch": 10.186900958466454, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 12754 + }, + { + "epoch": 10.187699680511182, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 12755 + }, + { + "epoch": 10.188498402555911, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 12756 + }, + { + "epoch": 10.189297124600639, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 12757 + }, + { + "epoch": 10.190095846645367, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 12758 + }, + { + "epoch": 10.190894568690096, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 12759 + }, + { + "epoch": 10.191693290734824, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 12760 + }, + { + "epoch": 10.192492012779553, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 12761 + }, + { + "epoch": 10.19329073482428, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 12762 + }, + { + "epoch": 10.19408945686901, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 12763 + }, + { + "epoch": 10.194888178913738, + "grad_norm": 0.1640625, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 12764 + }, + { + "epoch": 10.195686900958467, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 12765 + }, + { + "epoch": 10.196485623003195, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 12766 + }, + { + "epoch": 10.197284345047922, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 12767 + }, + { + "epoch": 10.198083067092652, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 12768 + }, + { + "epoch": 10.19888178913738, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 12769 + }, + { + "epoch": 10.199680511182109, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 12770 + }, + { + "epoch": 10.200479233226837, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 12771 + }, + { + "epoch": 10.201277955271566, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 12772 + }, + { + "epoch": 10.202076677316294, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 12773 + }, + { + "epoch": 10.202875399361023, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 12774 + }, + { + "epoch": 10.20367412140575, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 12775 + }, + { + "epoch": 10.204472843450478, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 12776 + }, + { + "epoch": 10.205271565495208, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 12777 + }, + { + "epoch": 10.206070287539935, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 12778 + }, + { + "epoch": 10.206869009584665, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 12779 + }, + { + "epoch": 10.207667731629392, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 12780 + }, + { + "epoch": 10.208466453674122, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 12781 + }, + { + "epoch": 10.20926517571885, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 12782 + }, + { + "epoch": 10.210063897763579, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 12783 + }, + { + "epoch": 10.210862619808307, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 12784 + }, + { + "epoch": 10.211661341853036, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 12785 + }, + { + "epoch": 10.212460063897764, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 12786 + }, + { + "epoch": 10.213258785942491, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 12787 + }, + { + "epoch": 10.21405750798722, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 12788 + }, + { + "epoch": 10.214856230031948, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 12789 + }, + { + "epoch": 10.215654952076678, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 12790 + }, + { + "epoch": 10.216453674121405, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 12791 + }, + { + "epoch": 10.217252396166135, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 12792 + }, + { + "epoch": 10.218051118210862, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 12793 + }, + { + "epoch": 10.218849840255592, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 12794 + }, + { + "epoch": 10.21964856230032, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 12795 + }, + { + "epoch": 10.220447284345047, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 12796 + }, + { + "epoch": 10.221246006389777, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 12797 + }, + { + "epoch": 10.222044728434504, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 12798 + }, + { + "epoch": 10.222843450479234, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 12799 + }, + { + "epoch": 10.223642172523961, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0192, + "step": 12800 + }, + { + "epoch": 10.22444089456869, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 12801 + }, + { + "epoch": 10.225239616613418, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 12802 + }, + { + "epoch": 10.226038338658148, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 12803 + }, + { + "epoch": 10.226837060702875, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 12804 + }, + { + "epoch": 10.227635782747603, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 12805 + }, + { + "epoch": 10.228434504792332, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 12806 + }, + { + "epoch": 10.22923322683706, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0161, + "step": 12807 + }, + { + "epoch": 10.23003194888179, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 12808 + }, + { + "epoch": 10.230830670926517, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 12809 + }, + { + "epoch": 10.231629392971247, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 12810 + }, + { + "epoch": 10.232428115015974, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 12811 + }, + { + "epoch": 10.233226837060704, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 12812 + }, + { + "epoch": 10.234025559105431, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 12813 + }, + { + "epoch": 10.23482428115016, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0173, + "step": 12814 + }, + { + "epoch": 10.235623003194888, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 12815 + }, + { + "epoch": 10.236421725239616, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 12816 + }, + { + "epoch": 10.237220447284345, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 12817 + }, + { + "epoch": 10.238019169329073, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 12818 + }, + { + "epoch": 10.238817891373802, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0407, + "step": 12819 + }, + { + "epoch": 10.23961661341853, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 12820 + }, + { + "epoch": 10.24041533546326, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 12821 + }, + { + "epoch": 10.241214057507987, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 12822 + }, + { + "epoch": 10.242012779552716, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 12823 + }, + { + "epoch": 10.242811501597444, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0173, + "step": 12824 + }, + { + "epoch": 10.243610223642172, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 12825 + }, + { + "epoch": 10.244408945686901, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 12826 + }, + { + "epoch": 10.245207667731629, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 12827 + }, + { + "epoch": 10.246006389776358, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 12828 + }, + { + "epoch": 10.246805111821086, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 12829 + }, + { + "epoch": 10.247603833865815, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 12830 + }, + { + "epoch": 10.248402555910543, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 12831 + }, + { + "epoch": 10.249201277955272, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 12832 + }, + { + "epoch": 10.25, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 12833 + }, + { + "epoch": 10.250798722044728, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 12834 + }, + { + "epoch": 10.251597444089457, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 12835 + }, + { + "epoch": 10.252396166134185, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 12836 + }, + { + "epoch": 10.253194888178914, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 12837 + }, + { + "epoch": 10.253993610223642, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0152, + "step": 12838 + }, + { + "epoch": 10.254792332268371, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 12839 + }, + { + "epoch": 10.255591054313099, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0174, + "step": 12840 + }, + { + "epoch": 10.256389776357828, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 12841 + }, + { + "epoch": 10.257188498402556, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 12842 + }, + { + "epoch": 10.257987220447284, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 12843 + }, + { + "epoch": 10.258785942492013, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 12844 + }, + { + "epoch": 10.25958466453674, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 12845 + }, + { + "epoch": 10.26038338658147, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 12846 + }, + { + "epoch": 10.261182108626198, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 12847 + }, + { + "epoch": 10.261980830670927, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.034, + "step": 12848 + }, + { + "epoch": 10.262779552715655, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0158, + "step": 12849 + }, + { + "epoch": 10.263578274760384, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 12850 + }, + { + "epoch": 10.264376996805112, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 12851 + }, + { + "epoch": 10.26517571884984, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 12852 + }, + { + "epoch": 10.265974440894569, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 12853 + }, + { + "epoch": 10.266773162939296, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 12854 + }, + { + "epoch": 10.267571884984026, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 12855 + }, + { + "epoch": 10.268370607028753, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 12856 + }, + { + "epoch": 10.269169329073483, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 12857 + }, + { + "epoch": 10.26996805111821, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 12858 + }, + { + "epoch": 10.27076677316294, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 12859 + }, + { + "epoch": 10.271565495207668, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 12860 + }, + { + "epoch": 10.272364217252397, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 12861 + }, + { + "epoch": 10.273162939297125, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 12862 + }, + { + "epoch": 10.273961661341852, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0138, + "step": 12863 + }, + { + "epoch": 10.274760383386582, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 12864 + }, + { + "epoch": 10.27555910543131, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 12865 + }, + { + "epoch": 10.276357827476039, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 12866 + }, + { + "epoch": 10.277156549520766, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 12867 + }, + { + "epoch": 10.277955271565496, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 12868 + }, + { + "epoch": 10.278753993610223, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 12869 + }, + { + "epoch": 10.279552715654953, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 12870 + }, + { + "epoch": 10.28035143769968, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 12871 + }, + { + "epoch": 10.281150159744408, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 12872 + }, + { + "epoch": 10.281948881789138, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0168, + "step": 12873 + }, + { + "epoch": 10.282747603833865, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0192, + "step": 12874 + }, + { + "epoch": 10.283546325878595, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 12875 + }, + { + "epoch": 10.284345047923322, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 12876 + }, + { + "epoch": 10.285143769968052, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 12877 + }, + { + "epoch": 10.28594249201278, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 12878 + }, + { + "epoch": 10.286741214057509, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 12879 + }, + { + "epoch": 10.287539936102236, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 12880 + }, + { + "epoch": 10.288338658146964, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 12881 + }, + { + "epoch": 10.289137380191693, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 12882 + }, + { + "epoch": 10.289936102236421, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 12883 + }, + { + "epoch": 10.29073482428115, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 12884 + }, + { + "epoch": 10.291533546325878, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 12885 + }, + { + "epoch": 10.292332268370608, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 12886 + }, + { + "epoch": 10.293130990415335, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 12887 + }, + { + "epoch": 10.293929712460065, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 12888 + }, + { + "epoch": 10.294728434504792, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 12889 + }, + { + "epoch": 10.295527156549522, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 12890 + }, + { + "epoch": 10.29632587859425, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 12891 + }, + { + "epoch": 10.297124600638977, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 12892 + }, + { + "epoch": 10.297923322683706, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 12893 + }, + { + "epoch": 10.298722044728434, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0404, + "step": 12894 + }, + { + "epoch": 10.299520766773163, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 12895 + }, + { + "epoch": 10.300319488817891, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 12896 + }, + { + "epoch": 10.30111821086262, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 12897 + }, + { + "epoch": 10.301916932907348, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.037, + "step": 12898 + }, + { + "epoch": 10.302715654952078, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 12899 + }, + { + "epoch": 10.303514376996805, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 12900 + }, + { + "epoch": 10.304313099041533, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 12901 + }, + { + "epoch": 10.305111821086262, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 12902 + }, + { + "epoch": 10.30591054313099, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 12903 + }, + { + "epoch": 10.30670926517572, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 12904 + }, + { + "epoch": 10.307507987220447, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 12905 + }, + { + "epoch": 10.308306709265176, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 12906 + }, + { + "epoch": 10.309105431309904, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 12907 + }, + { + "epoch": 10.309904153354633, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 12908 + }, + { + "epoch": 10.310702875399361, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 12909 + }, + { + "epoch": 10.311501597444089, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 12910 + }, + { + "epoch": 10.312300319488818, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 12911 + }, + { + "epoch": 10.313099041533546, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 12912 + }, + { + "epoch": 10.313897763578275, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 12913 + }, + { + "epoch": 10.314696485623003, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 12914 + }, + { + "epoch": 10.315495207667732, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 12915 + }, + { + "epoch": 10.31629392971246, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 12916 + }, + { + "epoch": 10.31709265175719, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 12917 + }, + { + "epoch": 10.317891373801917, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 12918 + }, + { + "epoch": 10.318690095846645, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 12919 + }, + { + "epoch": 10.319488817891374, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 12920 + }, + { + "epoch": 10.320287539936102, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 12921 + }, + { + "epoch": 10.321086261980831, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 12922 + }, + { + "epoch": 10.321884984025559, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.037, + "step": 12923 + }, + { + "epoch": 10.322683706070288, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 12924 + }, + { + "epoch": 10.323482428115016, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 12925 + }, + { + "epoch": 10.324281150159745, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 12926 + }, + { + "epoch": 10.325079872204473, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 12927 + }, + { + "epoch": 10.3258785942492, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 12928 + }, + { + "epoch": 10.32667731629393, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 12929 + }, + { + "epoch": 10.327476038338657, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 12930 + }, + { + "epoch": 10.328274760383387, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 12931 + }, + { + "epoch": 10.329073482428115, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 12932 + }, + { + "epoch": 10.329872204472844, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 12933 + }, + { + "epoch": 10.330670926517572, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 12934 + }, + { + "epoch": 10.331469648562301, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 12935 + }, + { + "epoch": 10.332268370607029, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 12936 + }, + { + "epoch": 10.333067092651758, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 12937 + }, + { + "epoch": 10.333865814696486, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 12938 + }, + { + "epoch": 10.334664536741213, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0183, + "step": 12939 + }, + { + "epoch": 10.335463258785943, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0205, + "step": 12940 + }, + { + "epoch": 10.33626198083067, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 12941 + }, + { + "epoch": 10.3370607028754, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 12942 + }, + { + "epoch": 10.337859424920127, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 12943 + }, + { + "epoch": 10.338658146964857, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 12944 + }, + { + "epoch": 10.339456869009584, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0353, + "step": 12945 + }, + { + "epoch": 10.340255591054314, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 12946 + }, + { + "epoch": 10.341054313099042, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 12947 + }, + { + "epoch": 10.34185303514377, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 12948 + }, + { + "epoch": 10.342651757188499, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 12949 + }, + { + "epoch": 10.343450479233226, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 12950 + }, + { + "epoch": 10.344249201277956, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 12951 + }, + { + "epoch": 10.345047923322683, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 12952 + }, + { + "epoch": 10.345846645367413, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 12953 + }, + { + "epoch": 10.34664536741214, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 12954 + }, + { + "epoch": 10.34744408945687, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 12955 + }, + { + "epoch": 10.348242811501597, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 12956 + }, + { + "epoch": 10.349041533546325, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 12957 + }, + { + "epoch": 10.349840255591054, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 12958 + }, + { + "epoch": 10.350638977635782, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 12959 + }, + { + "epoch": 10.351437699680512, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 12960 + }, + { + "epoch": 10.35223642172524, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 12961 + }, + { + "epoch": 10.353035143769969, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 12962 + }, + { + "epoch": 10.353833865814696, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 12963 + }, + { + "epoch": 10.354632587859426, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 12964 + }, + { + "epoch": 10.355431309904153, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 12965 + }, + { + "epoch": 10.356230031948883, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 12966 + }, + { + "epoch": 10.35702875399361, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 12967 + }, + { + "epoch": 10.357827476038338, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 12968 + }, + { + "epoch": 10.358626198083067, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 12969 + }, + { + "epoch": 10.359424920127795, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0372, + "step": 12970 + }, + { + "epoch": 10.360223642172524, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 12971 + }, + { + "epoch": 10.361022364217252, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 12972 + }, + { + "epoch": 10.361821086261982, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 12973 + }, + { + "epoch": 10.36261980830671, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 12974 + }, + { + "epoch": 10.363418530351439, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 12975 + }, + { + "epoch": 10.364217252396166, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 12976 + }, + { + "epoch": 10.365015974440894, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0175, + "step": 12977 + }, + { + "epoch": 10.365814696485623, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 12978 + }, + { + "epoch": 10.366613418530351, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 12979 + }, + { + "epoch": 10.36741214057508, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 12980 + }, + { + "epoch": 10.368210862619808, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0178, + "step": 12981 + }, + { + "epoch": 10.369009584664537, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 12982 + }, + { + "epoch": 10.369808306709265, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 12983 + }, + { + "epoch": 10.370607028753994, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 12984 + }, + { + "epoch": 10.371405750798722, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 12985 + }, + { + "epoch": 10.37220447284345, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 12986 + }, + { + "epoch": 10.37300319488818, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 12987 + }, + { + "epoch": 10.373801916932907, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0098, + "step": 12988 + }, + { + "epoch": 10.374600638977636, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 12989 + }, + { + "epoch": 10.375399361022364, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 12990 + }, + { + "epoch": 10.376198083067093, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 12991 + }, + { + "epoch": 10.37699680511182, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.034, + "step": 12992 + }, + { + "epoch": 10.37779552715655, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0396, + "step": 12993 + }, + { + "epoch": 10.378594249201278, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 12994 + }, + { + "epoch": 10.379392971246006, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 12995 + }, + { + "epoch": 10.380191693290735, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 12996 + }, + { + "epoch": 10.380990415335463, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 12997 + }, + { + "epoch": 10.381789137380192, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 12998 + }, + { + "epoch": 10.38258785942492, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 12999 + }, + { + "epoch": 10.383386581469649, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 13000 + }, + { + "epoch": 10.384185303514377, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 13001 + }, + { + "epoch": 10.384984025559106, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 13002 + }, + { + "epoch": 10.385782747603834, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 13003 + }, + { + "epoch": 10.386581469648561, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 13004 + }, + { + "epoch": 10.38738019169329, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 13005 + }, + { + "epoch": 10.388178913738018, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 13006 + }, + { + "epoch": 10.388977635782748, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 13007 + }, + { + "epoch": 10.389776357827476, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 13008 + }, + { + "epoch": 10.390575079872205, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 13009 + }, + { + "epoch": 10.391373801916933, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 13010 + }, + { + "epoch": 10.392172523961662, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 13011 + }, + { + "epoch": 10.39297124600639, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 13012 + }, + { + "epoch": 10.393769968051117, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 13013 + }, + { + "epoch": 10.394568690095847, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 13014 + }, + { + "epoch": 10.395367412140574, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 13015 + }, + { + "epoch": 10.396166134185304, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 13016 + }, + { + "epoch": 10.396964856230031, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 13017 + }, + { + "epoch": 10.39776357827476, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 13018 + }, + { + "epoch": 10.398562300319488, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0405, + "step": 13019 + }, + { + "epoch": 10.399361022364218, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 13020 + }, + { + "epoch": 10.400159744408946, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0143, + "step": 13021 + }, + { + "epoch": 10.400958466453675, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 13022 + }, + { + "epoch": 10.401757188498403, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 13023 + }, + { + "epoch": 10.40255591054313, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 13024 + }, + { + "epoch": 10.40335463258786, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 13025 + }, + { + "epoch": 10.404153354632587, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 13026 + }, + { + "epoch": 10.404952076677317, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 13027 + }, + { + "epoch": 10.405750798722044, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 13028 + }, + { + "epoch": 10.406549520766774, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 13029 + }, + { + "epoch": 10.407348242811501, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 13030 + }, + { + "epoch": 10.40814696485623, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 13031 + }, + { + "epoch": 10.408945686900958, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0389, + "step": 13032 + }, + { + "epoch": 10.409744408945686, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0376, + "step": 13033 + }, + { + "epoch": 10.410543130990416, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 13034 + }, + { + "epoch": 10.411341853035143, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 13035 + }, + { + "epoch": 10.412140575079873, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 13036 + }, + { + "epoch": 10.4129392971246, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 13037 + }, + { + "epoch": 10.41373801916933, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 13038 + }, + { + "epoch": 10.414536741214057, + "grad_norm": 0.1611328125, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 13039 + }, + { + "epoch": 10.415335463258787, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 13040 + }, + { + "epoch": 10.416134185303514, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 13041 + }, + { + "epoch": 10.416932907348242, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 13042 + }, + { + "epoch": 10.417731629392971, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 13043 + }, + { + "epoch": 10.418530351437699, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 13044 + }, + { + "epoch": 10.419329073482428, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 13045 + }, + { + "epoch": 10.420127795527156, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 13046 + }, + { + "epoch": 10.420926517571885, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 13047 + }, + { + "epoch": 10.421725239616613, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 13048 + }, + { + "epoch": 10.422523961661343, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 13049 + }, + { + "epoch": 10.42332268370607, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 13050 + }, + { + "epoch": 10.4241214057508, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 13051 + }, + { + "epoch": 10.424920127795527, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0196, + "step": 13052 + }, + { + "epoch": 10.425718849840255, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 13053 + }, + { + "epoch": 10.426517571884984, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 13054 + }, + { + "epoch": 10.427316293929712, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 13055 + }, + { + "epoch": 10.428115015974441, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 13056 + }, + { + "epoch": 10.428913738019169, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 13057 + }, + { + "epoch": 10.429712460063898, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0199, + "step": 13058 + }, + { + "epoch": 10.430511182108626, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 13059 + }, + { + "epoch": 10.431309904153355, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 13060 + }, + { + "epoch": 10.432108626198083, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 13061 + }, + { + "epoch": 10.43290734824281, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 13062 + }, + { + "epoch": 10.43370607028754, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 13063 + }, + { + "epoch": 10.434504792332268, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 13064 + }, + { + "epoch": 10.435303514376997, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 13065 + }, + { + "epoch": 10.436102236421725, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 13066 + }, + { + "epoch": 10.436900958466454, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 13067 + }, + { + "epoch": 10.437699680511182, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 13068 + }, + { + "epoch": 10.438498402555911, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 13069 + }, + { + "epoch": 10.439297124600639, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 13070 + }, + { + "epoch": 10.440095846645367, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 13071 + }, + { + "epoch": 10.440894568690096, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 13072 + }, + { + "epoch": 10.441693290734824, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 13073 + }, + { + "epoch": 10.442492012779553, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 13074 + }, + { + "epoch": 10.44329073482428, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 13075 + }, + { + "epoch": 10.44408945686901, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 13076 + }, + { + "epoch": 10.444888178913738, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0182, + "step": 13077 + }, + { + "epoch": 10.445686900958467, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 13078 + }, + { + "epoch": 10.446485623003195, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 13079 + }, + { + "epoch": 10.447284345047922, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 13080 + }, + { + "epoch": 10.448083067092652, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 13081 + }, + { + "epoch": 10.44888178913738, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 13082 + }, + { + "epoch": 10.449680511182109, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0169, + "step": 13083 + }, + { + "epoch": 10.450479233226837, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 13084 + }, + { + "epoch": 10.451277955271566, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 13085 + }, + { + "epoch": 10.452076677316294, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 13086 + }, + { + "epoch": 10.452875399361023, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 13087 + }, + { + "epoch": 10.45367412140575, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 13088 + }, + { + "epoch": 10.454472843450478, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 13089 + }, + { + "epoch": 10.455271565495208, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0395, + "step": 13090 + }, + { + "epoch": 10.456070287539935, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 13091 + }, + { + "epoch": 10.456869009584665, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0185, + "step": 13092 + }, + { + "epoch": 10.457667731629392, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0147, + "step": 13093 + }, + { + "epoch": 10.458466453674122, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 13094 + }, + { + "epoch": 10.45926517571885, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 13095 + }, + { + "epoch": 10.460063897763579, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 13096 + }, + { + "epoch": 10.460862619808307, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 13097 + }, + { + "epoch": 10.461661341853036, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 13098 + }, + { + "epoch": 10.462460063897764, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 13099 + }, + { + "epoch": 10.463258785942491, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 13100 + }, + { + "epoch": 10.46405750798722, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 13101 + }, + { + "epoch": 10.464856230031948, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 13102 + }, + { + "epoch": 10.465654952076678, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 13103 + }, + { + "epoch": 10.466453674121405, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 13104 + }, + { + "epoch": 10.467252396166135, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 13105 + }, + { + "epoch": 10.468051118210862, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 13106 + }, + { + "epoch": 10.468849840255592, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 13107 + }, + { + "epoch": 10.46964856230032, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 13108 + }, + { + "epoch": 10.470447284345047, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 13109 + }, + { + "epoch": 10.471246006389777, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 13110 + }, + { + "epoch": 10.472044728434504, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 13111 + }, + { + "epoch": 10.472843450479234, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 13112 + }, + { + "epoch": 10.473642172523961, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 13113 + }, + { + "epoch": 10.47444089456869, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 13114 + }, + { + "epoch": 10.475239616613418, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 13115 + }, + { + "epoch": 10.476038338658148, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 13116 + }, + { + "epoch": 10.476837060702875, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 13117 + }, + { + "epoch": 10.477635782747603, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 13118 + }, + { + "epoch": 10.478434504792332, + "grad_norm": 0.051513671875, + "learning_rate": 0.0005, + "loss": 1.0205, + "step": 13119 + }, + { + "epoch": 10.47923322683706, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0195, + "step": 13120 + }, + { + "epoch": 10.48003194888179, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 13121 + }, + { + "epoch": 10.480830670926517, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 13122 + }, + { + "epoch": 10.481629392971247, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 13123 + }, + { + "epoch": 10.482428115015974, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 13124 + }, + { + "epoch": 10.483226837060704, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 13125 + }, + { + "epoch": 10.484025559105431, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 13126 + }, + { + "epoch": 10.48482428115016, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 13127 + }, + { + "epoch": 10.485623003194888, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 13128 + }, + { + "epoch": 10.486421725239616, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 13129 + }, + { + "epoch": 10.487220447284345, + "grad_norm": 0.055908203125, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 13130 + }, + { + "epoch": 10.488019169329073, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 13131 + }, + { + "epoch": 10.488817891373802, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 13132 + }, + { + "epoch": 10.48961661341853, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 13133 + }, + { + "epoch": 10.49041533546326, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 13134 + }, + { + "epoch": 10.491214057507987, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 13135 + }, + { + "epoch": 10.492012779552716, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 13136 + }, + { + "epoch": 10.492811501597444, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 13137 + }, + { + "epoch": 10.493610223642172, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 13138 + }, + { + "epoch": 10.494408945686901, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 13139 + }, + { + "epoch": 10.495207667731629, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 13140 + }, + { + "epoch": 10.496006389776358, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 13141 + }, + { + "epoch": 10.496805111821086, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 13142 + }, + { + "epoch": 10.497603833865815, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 13143 + }, + { + "epoch": 10.498402555910543, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 13144 + }, + { + "epoch": 10.499201277955272, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 13145 + }, + { + "epoch": 10.5, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 13146 + }, + { + "epoch": 10.500798722044728, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 13147 + }, + { + "epoch": 10.501597444089457, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0181, + "step": 13148 + }, + { + "epoch": 10.502396166134185, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 13149 + }, + { + "epoch": 10.503194888178914, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 13150 + }, + { + "epoch": 10.503993610223642, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 13151 + }, + { + "epoch": 10.504792332268371, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 13152 + }, + { + "epoch": 10.505591054313099, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0163, + "step": 13153 + }, + { + "epoch": 10.506389776357828, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 13154 + }, + { + "epoch": 10.507188498402556, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 13155 + }, + { + "epoch": 10.507987220447284, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 13156 + }, + { + "epoch": 10.508785942492013, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 13157 + }, + { + "epoch": 10.50958466453674, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 13158 + }, + { + "epoch": 10.51038338658147, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 13159 + }, + { + "epoch": 10.511182108626198, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0338, + "step": 13160 + }, + { + "epoch": 10.511980830670927, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 13161 + }, + { + "epoch": 10.512779552715655, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 13162 + }, + { + "epoch": 10.513578274760384, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0199, + "step": 13163 + }, + { + "epoch": 10.514376996805112, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 13164 + }, + { + "epoch": 10.51517571884984, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 13165 + }, + { + "epoch": 10.515974440894569, + "grad_norm": 0.193359375, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 13166 + }, + { + "epoch": 10.516773162939296, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 13167 + }, + { + "epoch": 10.517571884984026, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 13168 + }, + { + "epoch": 10.518370607028753, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 13169 + }, + { + "epoch": 10.519169329073483, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 13170 + }, + { + "epoch": 10.51996805111821, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 13171 + }, + { + "epoch": 10.52076677316294, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 13172 + }, + { + "epoch": 10.521565495207668, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 13173 + }, + { + "epoch": 10.522364217252395, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 13174 + }, + { + "epoch": 10.523162939297125, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 13175 + }, + { + "epoch": 10.523961661341852, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 13176 + }, + { + "epoch": 10.524760383386582, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 13177 + }, + { + "epoch": 10.52555910543131, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 13178 + }, + { + "epoch": 10.526357827476039, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 13179 + }, + { + "epoch": 10.527156549520766, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 13180 + }, + { + "epoch": 10.527955271565496, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 13181 + }, + { + "epoch": 10.528753993610223, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 13182 + }, + { + "epoch": 10.529552715654953, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 13183 + }, + { + "epoch": 10.53035143769968, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 13184 + }, + { + "epoch": 10.531150159744408, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 13185 + }, + { + "epoch": 10.531948881789138, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0182, + "step": 13186 + }, + { + "epoch": 10.532747603833865, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 13187 + }, + { + "epoch": 10.533546325878595, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 13188 + }, + { + "epoch": 10.534345047923322, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 13189 + }, + { + "epoch": 10.535143769968052, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 13190 + }, + { + "epoch": 10.53594249201278, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 13191 + }, + { + "epoch": 10.536741214057509, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 13192 + }, + { + "epoch": 10.537539936102236, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 13193 + }, + { + "epoch": 10.538338658146966, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 13194 + }, + { + "epoch": 10.539137380191693, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 13195 + }, + { + "epoch": 10.539936102236421, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 13196 + }, + { + "epoch": 10.54073482428115, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 13197 + }, + { + "epoch": 10.541533546325878, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 13198 + }, + { + "epoch": 10.542332268370608, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 13199 + }, + { + "epoch": 10.543130990415335, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 13200 + }, + { + "epoch": 10.543929712460065, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 13201 + }, + { + "epoch": 10.544728434504792, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0192, + "step": 13202 + }, + { + "epoch": 10.545527156549522, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0185, + "step": 13203 + }, + { + "epoch": 10.54632587859425, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 13204 + }, + { + "epoch": 10.547124600638977, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 13205 + }, + { + "epoch": 10.547923322683706, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 13206 + }, + { + "epoch": 10.548722044728434, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 13207 + }, + { + "epoch": 10.549520766773163, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.01, + "step": 13208 + }, + { + "epoch": 10.550319488817891, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 13209 + }, + { + "epoch": 10.55111821086262, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 13210 + }, + { + "epoch": 10.551916932907348, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 13211 + }, + { + "epoch": 10.552715654952078, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 13212 + }, + { + "epoch": 10.553514376996805, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 13213 + }, + { + "epoch": 10.554313099041533, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 13214 + }, + { + "epoch": 10.555111821086262, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 13215 + }, + { + "epoch": 10.55591054313099, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 13216 + }, + { + "epoch": 10.55670926517572, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0157, + "step": 13217 + }, + { + "epoch": 10.557507987220447, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0205, + "step": 13218 + }, + { + "epoch": 10.558306709265176, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 13219 + }, + { + "epoch": 10.559105431309904, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 13220 + }, + { + "epoch": 10.559904153354633, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 13221 + }, + { + "epoch": 10.560702875399361, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 13222 + }, + { + "epoch": 10.561501597444089, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 13223 + }, + { + "epoch": 10.562300319488818, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 13224 + }, + { + "epoch": 10.563099041533546, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 13225 + }, + { + "epoch": 10.563897763578275, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0199, + "step": 13226 + }, + { + "epoch": 10.564696485623003, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 13227 + }, + { + "epoch": 10.565495207667732, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.036, + "step": 13228 + }, + { + "epoch": 10.56629392971246, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 13229 + }, + { + "epoch": 10.56709265175719, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 13230 + }, + { + "epoch": 10.567891373801917, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 13231 + }, + { + "epoch": 10.568690095846645, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 13232 + }, + { + "epoch": 10.569488817891374, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 13233 + }, + { + "epoch": 10.570287539936102, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 13234 + }, + { + "epoch": 10.571086261980831, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 13235 + }, + { + "epoch": 10.571884984025559, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 13236 + }, + { + "epoch": 10.572683706070288, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 13237 + }, + { + "epoch": 10.573482428115016, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 13238 + }, + { + "epoch": 10.574281150159745, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 13239 + }, + { + "epoch": 10.575079872204473, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 13240 + }, + { + "epoch": 10.5758785942492, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0401, + "step": 13241 + }, + { + "epoch": 10.57667731629393, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 13242 + }, + { + "epoch": 10.577476038338657, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 13243 + }, + { + "epoch": 10.578274760383387, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 13244 + }, + { + "epoch": 10.579073482428115, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 13245 + }, + { + "epoch": 10.579872204472844, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 13246 + }, + { + "epoch": 10.580670926517572, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 13247 + }, + { + "epoch": 10.581469648562301, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 13248 + }, + { + "epoch": 10.582268370607029, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 13249 + }, + { + "epoch": 10.583067092651756, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 13250 + }, + { + "epoch": 10.583865814696486, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 13251 + }, + { + "epoch": 10.584664536741213, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 13252 + }, + { + "epoch": 10.585463258785943, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0159, + "step": 13253 + }, + { + "epoch": 10.58626198083067, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 13254 + }, + { + "epoch": 10.5870607028754, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 13255 + }, + { + "epoch": 10.587859424920127, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 13256 + }, + { + "epoch": 10.588658146964857, + "grad_norm": 0.2041015625, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 13257 + }, + { + "epoch": 10.589456869009584, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 13258 + }, + { + "epoch": 10.590255591054314, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 13259 + }, + { + "epoch": 10.591054313099042, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 13260 + }, + { + "epoch": 10.59185303514377, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 13261 + }, + { + "epoch": 10.592651757188499, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 13262 + }, + { + "epoch": 10.593450479233226, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0176, + "step": 13263 + }, + { + "epoch": 10.594249201277956, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 13264 + }, + { + "epoch": 10.595047923322683, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 13265 + }, + { + "epoch": 10.595846645367413, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 13266 + }, + { + "epoch": 10.59664536741214, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 13267 + }, + { + "epoch": 10.59744408945687, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 13268 + }, + { + "epoch": 10.598242811501597, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 13269 + }, + { + "epoch": 10.599041533546325, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 13270 + }, + { + "epoch": 10.599840255591054, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 13271 + }, + { + "epoch": 10.600638977635782, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 13272 + }, + { + "epoch": 10.601437699680512, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 13273 + }, + { + "epoch": 10.60223642172524, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 13274 + }, + { + "epoch": 10.603035143769969, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 13275 + }, + { + "epoch": 10.603833865814696, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 13276 + }, + { + "epoch": 10.604632587859426, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 13277 + }, + { + "epoch": 10.605431309904153, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 13278 + }, + { + "epoch": 10.606230031948883, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 13279 + }, + { + "epoch": 10.60702875399361, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 13280 + }, + { + "epoch": 10.607827476038338, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 13281 + }, + { + "epoch": 10.608626198083067, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 13282 + }, + { + "epoch": 10.609424920127795, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 13283 + }, + { + "epoch": 10.610223642172524, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 13284 + }, + { + "epoch": 10.611022364217252, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 13285 + }, + { + "epoch": 10.611821086261982, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 13286 + }, + { + "epoch": 10.61261980830671, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 13287 + }, + { + "epoch": 10.613418530351439, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 13288 + }, + { + "epoch": 10.614217252396166, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 13289 + }, + { + "epoch": 10.615015974440894, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 13290 + }, + { + "epoch": 10.615814696485623, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 13291 + }, + { + "epoch": 10.616613418530351, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 13292 + }, + { + "epoch": 10.61741214057508, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 13293 + }, + { + "epoch": 10.618210862619808, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 13294 + }, + { + "epoch": 10.619009584664537, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 13295 + }, + { + "epoch": 10.619808306709265, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 13296 + }, + { + "epoch": 10.620607028753994, + "grad_norm": 0.208984375, + "learning_rate": 0.0005, + "loss": 1.0127, + "step": 13297 + }, + { + "epoch": 10.621405750798722, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 13298 + }, + { + "epoch": 10.62220447284345, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 13299 + }, + { + "epoch": 10.62300319488818, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 13300 + }, + { + "epoch": 10.623801916932907, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0373, + "step": 13301 + }, + { + "epoch": 10.624600638977636, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 13302 + }, + { + "epoch": 10.625399361022364, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 13303 + }, + { + "epoch": 10.626198083067093, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 13304 + }, + { + "epoch": 10.62699680511182, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 13305 + }, + { + "epoch": 10.62779552715655, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 13306 + }, + { + "epoch": 10.628594249201278, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 13307 + }, + { + "epoch": 10.629392971246006, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 13308 + }, + { + "epoch": 10.630191693290735, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 13309 + }, + { + "epoch": 10.630990415335463, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0123, + "step": 13310 + }, + { + "epoch": 10.631789137380192, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 13311 + }, + { + "epoch": 10.63258785942492, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 13312 + }, + { + "epoch": 10.633386581469649, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 13313 + }, + { + "epoch": 10.634185303514377, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 13314 + }, + { + "epoch": 10.634984025559106, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0171, + "step": 13315 + }, + { + "epoch": 10.635782747603834, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 13316 + }, + { + "epoch": 10.636581469648561, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 13317 + }, + { + "epoch": 10.63738019169329, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 13318 + }, + { + "epoch": 10.638178913738018, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 13319 + }, + { + "epoch": 10.638977635782748, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 13320 + }, + { + "epoch": 10.639776357827476, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 13321 + }, + { + "epoch": 10.640575079872205, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0185, + "step": 13322 + }, + { + "epoch": 10.641373801916933, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 13323 + }, + { + "epoch": 10.642172523961662, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 13324 + }, + { + "epoch": 10.64297124600639, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 13325 + }, + { + "epoch": 10.643769968051117, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 13326 + }, + { + "epoch": 10.644568690095847, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 13327 + }, + { + "epoch": 10.645367412140574, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 13328 + }, + { + "epoch": 10.646166134185304, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0176, + "step": 13329 + }, + { + "epoch": 10.646964856230031, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 13330 + }, + { + "epoch": 10.64776357827476, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 13331 + }, + { + "epoch": 10.648562300319488, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 13332 + }, + { + "epoch": 10.649361022364218, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 13333 + }, + { + "epoch": 10.650159744408946, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 13334 + }, + { + "epoch": 10.650958466453675, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 13335 + }, + { + "epoch": 10.651757188498403, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 13336 + }, + { + "epoch": 10.65255591054313, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 13337 + }, + { + "epoch": 10.65335463258786, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 13338 + }, + { + "epoch": 10.654153354632587, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 13339 + }, + { + "epoch": 10.654952076677317, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 13340 + }, + { + "epoch": 10.655750798722044, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 13341 + }, + { + "epoch": 10.656549520766774, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 13342 + }, + { + "epoch": 10.657348242811501, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 13343 + }, + { + "epoch": 10.65814696485623, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 13344 + }, + { + "epoch": 10.658945686900958, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0176, + "step": 13345 + }, + { + "epoch": 10.659744408945686, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 13346 + }, + { + "epoch": 10.660543130990416, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 13347 + }, + { + "epoch": 10.661341853035143, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 13348 + }, + { + "epoch": 10.662140575079873, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 13349 + }, + { + "epoch": 10.6629392971246, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 13350 + }, + { + "epoch": 10.66373801916933, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 13351 + }, + { + "epoch": 10.664536741214057, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 13352 + }, + { + "epoch": 10.665335463258787, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 13353 + }, + { + "epoch": 10.666134185303514, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 13354 + }, + { + "epoch": 10.666932907348244, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 13355 + }, + { + "epoch": 10.667731629392971, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 13356 + }, + { + "epoch": 10.668530351437699, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 13357 + }, + { + "epoch": 10.669329073482428, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 13358 + }, + { + "epoch": 10.670127795527156, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 13359 + }, + { + "epoch": 10.670926517571885, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 13360 + }, + { + "epoch": 10.671725239616613, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0166, + "step": 13361 + }, + { + "epoch": 10.672523961661343, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 13362 + }, + { + "epoch": 10.67332268370607, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 13363 + }, + { + "epoch": 10.6741214057508, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0385, + "step": 13364 + }, + { + "epoch": 10.674920127795527, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 13365 + }, + { + "epoch": 10.675718849840255, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0161, + "step": 13366 + }, + { + "epoch": 10.676517571884984, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 13367 + }, + { + "epoch": 10.677316293929712, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 13368 + }, + { + "epoch": 10.678115015974441, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 13369 + }, + { + "epoch": 10.678913738019169, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 13370 + }, + { + "epoch": 10.679712460063898, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 13371 + }, + { + "epoch": 10.680511182108626, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 13372 + }, + { + "epoch": 10.681309904153355, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 13373 + }, + { + "epoch": 10.682108626198083, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 13374 + }, + { + "epoch": 10.68290734824281, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 13375 + }, + { + "epoch": 10.68370607028754, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0199, + "step": 13376 + }, + { + "epoch": 10.684504792332268, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 13377 + }, + { + "epoch": 10.685303514376997, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 13378 + }, + { + "epoch": 10.686102236421725, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 13379 + }, + { + "epoch": 10.686900958466454, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 13380 + }, + { + "epoch": 10.687699680511182, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 13381 + }, + { + "epoch": 10.688498402555911, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 13382 + }, + { + "epoch": 10.689297124600639, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 13383 + }, + { + "epoch": 10.690095846645367, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 13384 + }, + { + "epoch": 10.690894568690096, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 13385 + }, + { + "epoch": 10.691693290734824, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 13386 + }, + { + "epoch": 10.692492012779553, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 13387 + }, + { + "epoch": 10.69329073482428, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 13388 + }, + { + "epoch": 10.69408945686901, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 13389 + }, + { + "epoch": 10.694888178913738, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 13390 + }, + { + "epoch": 10.695686900958467, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 13391 + }, + { + "epoch": 10.696485623003195, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 13392 + }, + { + "epoch": 10.697284345047922, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 13393 + }, + { + "epoch": 10.698083067092652, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 13394 + }, + { + "epoch": 10.69888178913738, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 13395 + }, + { + "epoch": 10.699680511182109, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 13396 + }, + { + "epoch": 10.700479233226837, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0168, + "step": 13397 + }, + { + "epoch": 10.701277955271566, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 13398 + }, + { + "epoch": 10.702076677316294, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 13399 + }, + { + "epoch": 10.702875399361023, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 13400 + }, + { + "epoch": 10.70367412140575, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 13401 + }, + { + "epoch": 10.704472843450478, + "grad_norm": 0.16796875, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 13402 + }, + { + "epoch": 10.705271565495208, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 13403 + }, + { + "epoch": 10.706070287539935, + "grad_norm": 0.181640625, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 13404 + }, + { + "epoch": 10.706869009584665, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 13405 + }, + { + "epoch": 10.707667731629392, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 13406 + }, + { + "epoch": 10.708466453674122, + "grad_norm": 0.1728515625, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 13407 + }, + { + "epoch": 10.70926517571885, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 13408 + }, + { + "epoch": 10.710063897763579, + "grad_norm": 0.193359375, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 13409 + }, + { + "epoch": 10.710862619808307, + "grad_norm": 0.166015625, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 13410 + }, + { + "epoch": 10.711661341853034, + "grad_norm": 0.1533203125, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 13411 + }, + { + "epoch": 10.712460063897764, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 13412 + }, + { + "epoch": 10.713258785942491, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 13413 + }, + { + "epoch": 10.71405750798722, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 13414 + }, + { + "epoch": 10.714856230031948, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0138, + "step": 13415 + }, + { + "epoch": 10.715654952076678, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 13416 + }, + { + "epoch": 10.716453674121405, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 13417 + }, + { + "epoch": 10.717252396166135, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 13418 + }, + { + "epoch": 10.718051118210862, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 13419 + }, + { + "epoch": 10.718849840255592, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 13420 + }, + { + "epoch": 10.71964856230032, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 13421 + }, + { + "epoch": 10.720447284345047, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 13422 + }, + { + "epoch": 10.721246006389777, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 13423 + }, + { + "epoch": 10.722044728434504, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 13424 + }, + { + "epoch": 10.722843450479234, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 13425 + }, + { + "epoch": 10.723642172523961, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 13426 + }, + { + "epoch": 10.72444089456869, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 13427 + }, + { + "epoch": 10.725239616613418, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 13428 + }, + { + "epoch": 10.726038338658148, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0158, + "step": 13429 + }, + { + "epoch": 10.726837060702875, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 13430 + }, + { + "epoch": 10.727635782747605, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 13431 + }, + { + "epoch": 10.728434504792332, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 13432 + }, + { + "epoch": 10.72923322683706, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0175, + "step": 13433 + }, + { + "epoch": 10.73003194888179, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 13434 + }, + { + "epoch": 10.730830670926517, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 13435 + }, + { + "epoch": 10.731629392971247, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0196, + "step": 13436 + }, + { + "epoch": 10.732428115015974, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 13437 + }, + { + "epoch": 10.733226837060704, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 13438 + }, + { + "epoch": 10.734025559105431, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0411, + "step": 13439 + }, + { + "epoch": 10.73482428115016, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 13440 + }, + { + "epoch": 10.735623003194888, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 13441 + }, + { + "epoch": 10.736421725239616, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 13442 + }, + { + "epoch": 10.737220447284345, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0337, + "step": 13443 + }, + { + "epoch": 10.738019169329073, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 13444 + }, + { + "epoch": 10.738817891373802, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 13445 + }, + { + "epoch": 10.73961661341853, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 13446 + }, + { + "epoch": 10.74041533546326, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 13447 + }, + { + "epoch": 10.741214057507987, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 13448 + }, + { + "epoch": 10.742012779552716, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 13449 + }, + { + "epoch": 10.742811501597444, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 13450 + }, + { + "epoch": 10.743610223642172, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 13451 + }, + { + "epoch": 10.744408945686901, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 13452 + }, + { + "epoch": 10.745207667731629, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 13453 + }, + { + "epoch": 10.746006389776358, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 13454 + }, + { + "epoch": 10.746805111821086, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 13455 + }, + { + "epoch": 10.747603833865815, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 13456 + }, + { + "epoch": 10.748402555910543, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 13457 + }, + { + "epoch": 10.749201277955272, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 13458 + }, + { + "epoch": 10.75, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 13459 + }, + { + "epoch": 10.750798722044728, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 13460 + }, + { + "epoch": 10.751597444089457, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 13461 + }, + { + "epoch": 10.752396166134185, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 13462 + }, + { + "epoch": 10.753194888178914, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 13463 + }, + { + "epoch": 10.753993610223642, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 13464 + }, + { + "epoch": 10.754792332268371, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 13465 + }, + { + "epoch": 10.755591054313099, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 13466 + }, + { + "epoch": 10.756389776357828, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 13467 + }, + { + "epoch": 10.757188498402556, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 13468 + }, + { + "epoch": 10.757987220447284, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 13469 + }, + { + "epoch": 10.758785942492013, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 13470 + }, + { + "epoch": 10.75958466453674, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 13471 + }, + { + "epoch": 10.76038338658147, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 13472 + }, + { + "epoch": 10.761182108626198, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 13473 + }, + { + "epoch": 10.761980830670927, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 13474 + }, + { + "epoch": 10.762779552715655, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 13475 + }, + { + "epoch": 10.763578274760384, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 13476 + }, + { + "epoch": 10.764376996805112, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 13477 + }, + { + "epoch": 10.76517571884984, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 13478 + }, + { + "epoch": 10.765974440894569, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 13479 + }, + { + "epoch": 10.766773162939296, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0177, + "step": 13480 + }, + { + "epoch": 10.767571884984026, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0348, + "step": 13481 + }, + { + "epoch": 10.768370607028753, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 13482 + }, + { + "epoch": 10.769169329073483, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 13483 + }, + { + "epoch": 10.76996805111821, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 13484 + }, + { + "epoch": 10.77076677316294, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 13485 + }, + { + "epoch": 10.771565495207668, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 13486 + }, + { + "epoch": 10.772364217252395, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 13487 + }, + { + "epoch": 10.773162939297125, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0173, + "step": 13488 + }, + { + "epoch": 10.773961661341852, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 13489 + }, + { + "epoch": 10.774760383386582, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0372, + "step": 13490 + }, + { + "epoch": 10.77555910543131, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 13491 + }, + { + "epoch": 10.776357827476039, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 13492 + }, + { + "epoch": 10.777156549520766, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 13493 + }, + { + "epoch": 10.777955271565496, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 13494 + }, + { + "epoch": 10.778753993610223, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 13495 + }, + { + "epoch": 10.779552715654953, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 13496 + }, + { + "epoch": 10.78035143769968, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 13497 + }, + { + "epoch": 10.781150159744408, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 13498 + }, + { + "epoch": 10.781948881789138, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 13499 + }, + { + "epoch": 10.782747603833865, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 13500 + }, + { + "epoch": 10.783546325878595, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 13501 + }, + { + "epoch": 10.784345047923322, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 13502 + }, + { + "epoch": 10.785143769968052, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 13503 + }, + { + "epoch": 10.78594249201278, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 13504 + }, + { + "epoch": 10.786741214057509, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 13505 + }, + { + "epoch": 10.787539936102236, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 13506 + }, + { + "epoch": 10.788338658146966, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 13507 + }, + { + "epoch": 10.789137380191693, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0181, + "step": 13508 + }, + { + "epoch": 10.789936102236421, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.038, + "step": 13509 + }, + { + "epoch": 10.79073482428115, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 13510 + }, + { + "epoch": 10.791533546325878, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 13511 + }, + { + "epoch": 10.792332268370608, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 13512 + }, + { + "epoch": 10.793130990415335, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 13513 + }, + { + "epoch": 10.793929712460065, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 13514 + }, + { + "epoch": 10.794728434504792, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 13515 + }, + { + "epoch": 10.795527156549522, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 13516 + }, + { + "epoch": 10.79632587859425, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 13517 + }, + { + "epoch": 10.797124600638977, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0141, + "step": 13518 + }, + { + "epoch": 10.797923322683706, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 13519 + }, + { + "epoch": 10.798722044728434, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 13520 + }, + { + "epoch": 10.799520766773163, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 13521 + }, + { + "epoch": 10.800319488817891, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 13522 + }, + { + "epoch": 10.80111821086262, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 13523 + }, + { + "epoch": 10.801916932907348, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 13524 + }, + { + "epoch": 10.802715654952078, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 13525 + }, + { + "epoch": 10.803514376996805, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 13526 + }, + { + "epoch": 10.804313099041533, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 13527 + }, + { + "epoch": 10.805111821086262, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 13528 + }, + { + "epoch": 10.80591054313099, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 13529 + }, + { + "epoch": 10.80670926517572, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 13530 + }, + { + "epoch": 10.807507987220447, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 13531 + }, + { + "epoch": 10.808306709265176, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 13532 + }, + { + "epoch": 10.809105431309904, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 13533 + }, + { + "epoch": 10.809904153354633, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.037, + "step": 13534 + }, + { + "epoch": 10.810702875399361, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 13535 + }, + { + "epoch": 10.811501597444089, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 13536 + }, + { + "epoch": 10.812300319488818, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 13537 + }, + { + "epoch": 10.813099041533546, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 13538 + }, + { + "epoch": 10.813897763578275, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 13539 + }, + { + "epoch": 10.814696485623003, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.038, + "step": 13540 + }, + { + "epoch": 10.815495207667732, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0182, + "step": 13541 + }, + { + "epoch": 10.81629392971246, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.018, + "step": 13542 + }, + { + "epoch": 10.81709265175719, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 13543 + }, + { + "epoch": 10.817891373801917, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 13544 + }, + { + "epoch": 10.818690095846645, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 13545 + }, + { + "epoch": 10.819488817891374, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 13546 + }, + { + "epoch": 10.820287539936102, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 13547 + }, + { + "epoch": 10.821086261980831, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 13548 + }, + { + "epoch": 10.821884984025559, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 13549 + }, + { + "epoch": 10.822683706070288, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 13550 + }, + { + "epoch": 10.823482428115016, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 13551 + }, + { + "epoch": 10.824281150159745, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 13552 + }, + { + "epoch": 10.825079872204473, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 13553 + }, + { + "epoch": 10.8258785942492, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 13554 + }, + { + "epoch": 10.82667731629393, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 13555 + }, + { + "epoch": 10.827476038338657, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 13556 + }, + { + "epoch": 10.828274760383387, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 13557 + }, + { + "epoch": 10.829073482428115, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0142, + "step": 13558 + }, + { + "epoch": 10.829872204472844, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 13559 + }, + { + "epoch": 10.830670926517572, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 13560 + }, + { + "epoch": 10.831469648562301, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 13561 + }, + { + "epoch": 10.832268370607029, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0181, + "step": 13562 + }, + { + "epoch": 10.833067092651756, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 13563 + }, + { + "epoch": 10.833865814696486, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 13564 + }, + { + "epoch": 10.834664536741213, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 13565 + }, + { + "epoch": 10.835463258785943, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 13566 + }, + { + "epoch": 10.83626198083067, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 13567 + }, + { + "epoch": 10.8370607028754, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0354, + "step": 13568 + }, + { + "epoch": 10.837859424920127, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 13569 + }, + { + "epoch": 10.838658146964857, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 13570 + }, + { + "epoch": 10.839456869009584, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0123, + "step": 13571 + }, + { + "epoch": 10.840255591054314, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 13572 + }, + { + "epoch": 10.841054313099042, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 13573 + }, + { + "epoch": 10.84185303514377, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 13574 + }, + { + "epoch": 10.842651757188499, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 13575 + }, + { + "epoch": 10.843450479233226, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0384, + "step": 13576 + }, + { + "epoch": 10.844249201277956, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 13577 + }, + { + "epoch": 10.845047923322683, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 13578 + }, + { + "epoch": 10.845846645367413, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 13579 + }, + { + "epoch": 10.84664536741214, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 13580 + }, + { + "epoch": 10.84744408945687, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 13581 + }, + { + "epoch": 10.848242811501597, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 13582 + }, + { + "epoch": 10.849041533546325, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 13583 + }, + { + "epoch": 10.849840255591054, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.012, + "step": 13584 + }, + { + "epoch": 10.850638977635782, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 13585 + }, + { + "epoch": 10.851437699680512, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 13586 + }, + { + "epoch": 10.85223642172524, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 13587 + }, + { + "epoch": 10.853035143769969, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 13588 + }, + { + "epoch": 10.853833865814696, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 13589 + }, + { + "epoch": 10.854632587859426, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 13590 + }, + { + "epoch": 10.855431309904153, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 13591 + }, + { + "epoch": 10.856230031948883, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 13592 + }, + { + "epoch": 10.85702875399361, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 13593 + }, + { + "epoch": 10.857827476038338, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 13594 + }, + { + "epoch": 10.858626198083067, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 13595 + }, + { + "epoch": 10.859424920127795, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 13596 + }, + { + "epoch": 10.860223642172524, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 13597 + }, + { + "epoch": 10.861022364217252, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 13598 + }, + { + "epoch": 10.861821086261982, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 13599 + }, + { + "epoch": 10.86261980830671, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0145, + "step": 13600 + }, + { + "epoch": 10.863418530351439, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 13601 + }, + { + "epoch": 10.864217252396166, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0174, + "step": 13602 + }, + { + "epoch": 10.865015974440894, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 13603 + }, + { + "epoch": 10.865814696485623, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 13604 + }, + { + "epoch": 10.866613418530351, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0171, + "step": 13605 + }, + { + "epoch": 10.86741214057508, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 13606 + }, + { + "epoch": 10.868210862619808, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 13607 + }, + { + "epoch": 10.869009584664537, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 13608 + }, + { + "epoch": 10.869808306709265, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 13609 + }, + { + "epoch": 10.870607028753994, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.018, + "step": 13610 + }, + { + "epoch": 10.871405750798722, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 13611 + }, + { + "epoch": 10.87220447284345, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 13612 + }, + { + "epoch": 10.87300319488818, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0354, + "step": 13613 + }, + { + "epoch": 10.873801916932907, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 13614 + }, + { + "epoch": 10.874600638977636, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 13615 + }, + { + "epoch": 10.875399361022364, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 13616 + }, + { + "epoch": 10.876198083067093, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0124, + "step": 13617 + }, + { + "epoch": 10.87699680511182, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 13618 + }, + { + "epoch": 10.87779552715655, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 13619 + }, + { + "epoch": 10.878594249201278, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 13620 + }, + { + "epoch": 10.879392971246006, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 13621 + }, + { + "epoch": 10.880191693290735, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 13622 + }, + { + "epoch": 10.880990415335463, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 13623 + }, + { + "epoch": 10.881789137380192, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 13624 + }, + { + "epoch": 10.88258785942492, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 13625 + }, + { + "epoch": 10.883386581469649, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 13626 + }, + { + "epoch": 10.884185303514377, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 13627 + }, + { + "epoch": 10.884984025559106, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 13628 + }, + { + "epoch": 10.885782747603834, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 13629 + }, + { + "epoch": 10.886581469648561, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 13630 + }, + { + "epoch": 10.88738019169329, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 13631 + }, + { + "epoch": 10.888178913738018, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 13632 + }, + { + "epoch": 10.888977635782748, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 13633 + }, + { + "epoch": 10.889776357827476, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 13634 + }, + { + "epoch": 10.890575079872205, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 13635 + }, + { + "epoch": 10.891373801916933, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 13636 + }, + { + "epoch": 10.892172523961662, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0181, + "step": 13637 + }, + { + "epoch": 10.89297124600639, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 13638 + }, + { + "epoch": 10.893769968051117, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 13639 + }, + { + "epoch": 10.894568690095847, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 13640 + }, + { + "epoch": 10.895367412140574, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 13641 + }, + { + "epoch": 10.896166134185304, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 13642 + }, + { + "epoch": 10.896964856230031, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 13643 + }, + { + "epoch": 10.89776357827476, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 13644 + }, + { + "epoch": 10.898562300319488, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 13645 + }, + { + "epoch": 10.899361022364218, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0357, + "step": 13646 + }, + { + "epoch": 10.900159744408946, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 13647 + }, + { + "epoch": 10.900958466453675, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 13648 + }, + { + "epoch": 10.901757188498403, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 13649 + }, + { + "epoch": 10.90255591054313, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 13650 + }, + { + "epoch": 10.90335463258786, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 13651 + }, + { + "epoch": 10.904153354632587, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 13652 + }, + { + "epoch": 10.904952076677317, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 13653 + }, + { + "epoch": 10.905750798722044, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 13654 + }, + { + "epoch": 10.906549520766774, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 13655 + }, + { + "epoch": 10.907348242811501, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 13656 + }, + { + "epoch": 10.90814696485623, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 13657 + }, + { + "epoch": 10.908945686900958, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 13658 + }, + { + "epoch": 10.909744408945686, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 13659 + }, + { + "epoch": 10.910543130990416, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 13660 + }, + { + "epoch": 10.911341853035143, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 13661 + }, + { + "epoch": 10.912140575079873, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 13662 + }, + { + "epoch": 10.9129392971246, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 13663 + }, + { + "epoch": 10.91373801916933, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 13664 + }, + { + "epoch": 10.914536741214057, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 13665 + }, + { + "epoch": 10.915335463258787, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 13666 + }, + { + "epoch": 10.916134185303514, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 13667 + }, + { + "epoch": 10.916932907348244, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 13668 + }, + { + "epoch": 10.917731629392971, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 13669 + }, + { + "epoch": 10.918530351437699, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 13670 + }, + { + "epoch": 10.919329073482428, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 13671 + }, + { + "epoch": 10.920127795527156, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 13672 + }, + { + "epoch": 10.920926517571885, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 13673 + }, + { + "epoch": 10.921725239616613, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 13674 + }, + { + "epoch": 10.922523961661343, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 13675 + }, + { + "epoch": 10.92332268370607, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 13676 + }, + { + "epoch": 10.9241214057508, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 13677 + }, + { + "epoch": 10.924920127795527, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 13678 + }, + { + "epoch": 10.925718849840255, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0094, + "step": 13679 + }, + { + "epoch": 10.926517571884984, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 13680 + }, + { + "epoch": 10.927316293929712, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 13681 + }, + { + "epoch": 10.928115015974441, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 13682 + }, + { + "epoch": 10.928913738019169, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 13683 + }, + { + "epoch": 10.929712460063898, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0354, + "step": 13684 + }, + { + "epoch": 10.930511182108626, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 13685 + }, + { + "epoch": 10.931309904153355, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 13686 + }, + { + "epoch": 10.932108626198083, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 13687 + }, + { + "epoch": 10.93290734824281, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 13688 + }, + { + "epoch": 10.93370607028754, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 13689 + }, + { + "epoch": 10.934504792332268, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 13690 + }, + { + "epoch": 10.935303514376997, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0372, + "step": 13691 + }, + { + "epoch": 10.936102236421725, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 13692 + }, + { + "epoch": 10.936900958466454, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 13693 + }, + { + "epoch": 10.937699680511182, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 13694 + }, + { + "epoch": 10.938498402555911, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 13695 + }, + { + "epoch": 10.939297124600639, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0196, + "step": 13696 + }, + { + "epoch": 10.940095846645367, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 13697 + }, + { + "epoch": 10.940894568690096, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 13698 + }, + { + "epoch": 10.941693290734824, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 13699 + }, + { + "epoch": 10.942492012779553, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 13700 + }, + { + "epoch": 10.94329073482428, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 13701 + }, + { + "epoch": 10.94408945686901, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 13702 + }, + { + "epoch": 10.944888178913738, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 13703 + }, + { + "epoch": 10.945686900958467, + "grad_norm": 0.1904296875, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 13704 + }, + { + "epoch": 10.946485623003195, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 13705 + }, + { + "epoch": 10.947284345047922, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 13706 + }, + { + "epoch": 10.948083067092652, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 13707 + }, + { + "epoch": 10.94888178913738, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 13708 + }, + { + "epoch": 10.949680511182109, + "grad_norm": 0.169921875, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 13709 + }, + { + "epoch": 10.950479233226837, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0384, + "step": 13710 + }, + { + "epoch": 10.951277955271566, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 13711 + }, + { + "epoch": 10.952076677316294, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 13712 + }, + { + "epoch": 10.952875399361023, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 13713 + }, + { + "epoch": 10.95367412140575, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 13714 + }, + { + "epoch": 10.954472843450478, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 13715 + }, + { + "epoch": 10.955271565495208, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 13716 + }, + { + "epoch": 10.956070287539935, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 13717 + }, + { + "epoch": 10.956869009584665, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 13718 + }, + { + "epoch": 10.957667731629392, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 13719 + }, + { + "epoch": 10.958466453674122, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 13720 + }, + { + "epoch": 10.95926517571885, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 13721 + }, + { + "epoch": 10.960063897763579, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 13722 + }, + { + "epoch": 10.960862619808307, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 13723 + }, + { + "epoch": 10.961661341853034, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 13724 + }, + { + "epoch": 10.962460063897764, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 13725 + }, + { + "epoch": 10.963258785942491, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 13726 + }, + { + "epoch": 10.96405750798722, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 13727 + }, + { + "epoch": 10.964856230031948, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 13728 + }, + { + "epoch": 10.965654952076678, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 13729 + }, + { + "epoch": 10.966453674121405, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 13730 + }, + { + "epoch": 10.967252396166135, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 13731 + }, + { + "epoch": 10.968051118210862, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 13732 + }, + { + "epoch": 10.968849840255592, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 13733 + }, + { + "epoch": 10.96964856230032, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 13734 + }, + { + "epoch": 10.970447284345047, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0175, + "step": 13735 + }, + { + "epoch": 10.971246006389777, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 13736 + }, + { + "epoch": 10.972044728434504, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 13737 + }, + { + "epoch": 10.972843450479234, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.0095, + "step": 13738 + }, + { + "epoch": 10.973642172523961, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 13739 + }, + { + "epoch": 10.97444089456869, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 13740 + }, + { + "epoch": 10.975239616613418, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.018, + "step": 13741 + }, + { + "epoch": 10.976038338658148, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 13742 + }, + { + "epoch": 10.976837060702875, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0377, + "step": 13743 + }, + { + "epoch": 10.977635782747605, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 13744 + }, + { + "epoch": 10.978434504792332, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 13745 + }, + { + "epoch": 10.97923322683706, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 13746 + }, + { + "epoch": 10.98003194888179, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 13747 + }, + { + "epoch": 10.980830670926517, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 13748 + }, + { + "epoch": 10.981629392971247, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 13749 + }, + { + "epoch": 10.982428115015974, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 13750 + }, + { + "epoch": 10.983226837060704, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 13751 + }, + { + "epoch": 10.984025559105431, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 13752 + }, + { + "epoch": 10.98482428115016, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 13753 + }, + { + "epoch": 10.985623003194888, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 13754 + }, + { + "epoch": 10.986421725239616, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 13755 + }, + { + "epoch": 10.987220447284345, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.039, + "step": 13756 + }, + { + "epoch": 10.988019169329073, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 13757 + }, + { + "epoch": 10.988817891373802, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 13758 + }, + { + "epoch": 10.98961661341853, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 13759 + }, + { + "epoch": 10.99041533546326, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 13760 + }, + { + "epoch": 10.991214057507987, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 13761 + }, + { + "epoch": 10.992012779552716, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 13762 + }, + { + "epoch": 10.992811501597444, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 13763 + }, + { + "epoch": 10.993610223642172, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 13764 + }, + { + "epoch": 10.994408945686901, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 13765 + }, + { + "epoch": 10.995207667731629, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 13766 + }, + { + "epoch": 10.996006389776358, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 13767 + }, + { + "epoch": 10.996805111821086, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 13768 + }, + { + "epoch": 10.997603833865815, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 13769 + }, + { + "epoch": 10.998402555910543, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 13770 + }, + { + "epoch": 10.999201277955272, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0163, + "step": 13771 + }, + { + "epoch": 11.0, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0162, + "step": 13772 + }, + { + "epoch": 11.000798722044728, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0182, + "step": 13773 + }, + { + "epoch": 11.001597444089457, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 13774 + }, + { + "epoch": 11.002396166134185, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 13775 + }, + { + "epoch": 11.003194888178914, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 13776 + }, + { + "epoch": 11.003993610223642, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0177, + "step": 13777 + }, + { + "epoch": 11.004792332268371, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 13778 + }, + { + "epoch": 11.005591054313099, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 13779 + }, + { + "epoch": 11.006389776357828, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 13780 + }, + { + "epoch": 11.007188498402556, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 13781 + }, + { + "epoch": 11.007987220447284, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 13782 + }, + { + "epoch": 11.008785942492013, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 13783 + }, + { + "epoch": 11.00958466453674, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 13784 + }, + { + "epoch": 11.01038338658147, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 13785 + }, + { + "epoch": 11.011182108626198, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 13786 + }, + { + "epoch": 11.011980830670927, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 13787 + }, + { + "epoch": 11.012779552715655, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 13788 + }, + { + "epoch": 11.013578274760384, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 13789 + }, + { + "epoch": 11.014376996805112, + "grad_norm": 0.16796875, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 13790 + }, + { + "epoch": 11.01517571884984, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 13791 + }, + { + "epoch": 11.015974440894569, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.0173, + "step": 13792 + }, + { + "epoch": 11.016773162939296, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 13793 + }, + { + "epoch": 11.017571884984026, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 13794 + }, + { + "epoch": 11.018370607028753, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 13795 + }, + { + "epoch": 11.019169329073483, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 13796 + }, + { + "epoch": 11.01996805111821, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 13797 + }, + { + "epoch": 11.02076677316294, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 13798 + }, + { + "epoch": 11.021565495207668, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 13799 + }, + { + "epoch": 11.022364217252397, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 13800 + }, + { + "epoch": 11.023162939297125, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 13801 + }, + { + "epoch": 11.023961661341852, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0196, + "step": 13802 + }, + { + "epoch": 11.024760383386582, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 13803 + }, + { + "epoch": 11.02555910543131, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 13804 + }, + { + "epoch": 11.026357827476039, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 13805 + }, + { + "epoch": 11.027156549520766, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 13806 + }, + { + "epoch": 11.027955271565496, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 13807 + }, + { + "epoch": 11.028753993610223, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 13808 + }, + { + "epoch": 11.029552715654953, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 13809 + }, + { + "epoch": 11.03035143769968, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0182, + "step": 13810 + }, + { + "epoch": 11.031150159744408, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 13811 + }, + { + "epoch": 11.031948881789138, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 13812 + }, + { + "epoch": 11.032747603833865, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.034, + "step": 13813 + }, + { + "epoch": 11.033546325878595, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 13814 + }, + { + "epoch": 11.034345047923322, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 13815 + }, + { + "epoch": 11.035143769968052, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 13816 + }, + { + "epoch": 11.03594249201278, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 13817 + }, + { + "epoch": 11.036741214057509, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 13818 + }, + { + "epoch": 11.037539936102236, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 13819 + }, + { + "epoch": 11.038338658146964, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 13820 + }, + { + "epoch": 11.039137380191693, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 13821 + }, + { + "epoch": 11.039936102236421, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0178, + "step": 13822 + }, + { + "epoch": 11.04073482428115, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 13823 + }, + { + "epoch": 11.041533546325878, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 13824 + }, + { + "epoch": 11.042332268370608, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 13825 + }, + { + "epoch": 11.043130990415335, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 13826 + }, + { + "epoch": 11.043929712460065, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 13827 + }, + { + "epoch": 11.044728434504792, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 13828 + }, + { + "epoch": 11.04552715654952, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0121, + "step": 13829 + }, + { + "epoch": 11.04632587859425, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 13830 + }, + { + "epoch": 11.047124600638977, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 13831 + }, + { + "epoch": 11.047923322683706, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0361, + "step": 13832 + }, + { + "epoch": 11.048722044728434, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 13833 + }, + { + "epoch": 11.049520766773163, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 13834 + }, + { + "epoch": 11.050319488817891, + "grad_norm": 0.205078125, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 13835 + }, + { + "epoch": 11.05111821086262, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 13836 + }, + { + "epoch": 11.051916932907348, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0166, + "step": 13837 + }, + { + "epoch": 11.052715654952078, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 13838 + }, + { + "epoch": 11.053514376996805, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 13839 + }, + { + "epoch": 11.054313099041533, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 13840 + }, + { + "epoch": 11.055111821086262, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 13841 + }, + { + "epoch": 11.05591054313099, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 13842 + }, + { + "epoch": 11.05670926517572, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 13843 + }, + { + "epoch": 11.057507987220447, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 13844 + }, + { + "epoch": 11.058306709265176, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 13845 + }, + { + "epoch": 11.059105431309904, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 13846 + }, + { + "epoch": 11.059904153354633, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 13847 + }, + { + "epoch": 11.060702875399361, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 13848 + }, + { + "epoch": 11.061501597444089, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 13849 + }, + { + "epoch": 11.062300319488818, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 13850 + }, + { + "epoch": 11.063099041533546, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 13851 + }, + { + "epoch": 11.063897763578275, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0338, + "step": 13852 + }, + { + "epoch": 11.064696485623003, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 13853 + }, + { + "epoch": 11.065495207667732, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 13854 + }, + { + "epoch": 11.06629392971246, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 13855 + }, + { + "epoch": 11.06709265175719, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 13856 + }, + { + "epoch": 11.067891373801917, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 13857 + }, + { + "epoch": 11.068690095846645, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 13858 + }, + { + "epoch": 11.069488817891374, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 13859 + }, + { + "epoch": 11.070287539936102, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 13860 + }, + { + "epoch": 11.071086261980831, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 13861 + }, + { + "epoch": 11.071884984025559, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 13862 + }, + { + "epoch": 11.072683706070288, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0182, + "step": 13863 + }, + { + "epoch": 11.073482428115016, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0165, + "step": 13864 + }, + { + "epoch": 11.074281150159745, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 13865 + }, + { + "epoch": 11.075079872204473, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 13866 + }, + { + "epoch": 11.0758785942492, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0188, + "step": 13867 + }, + { + "epoch": 11.07667731629393, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 13868 + }, + { + "epoch": 11.077476038338657, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 13869 + }, + { + "epoch": 11.078274760383387, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0155, + "step": 13870 + }, + { + "epoch": 11.079073482428115, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 13871 + }, + { + "epoch": 11.079872204472844, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 13872 + }, + { + "epoch": 11.080670926517572, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 13873 + }, + { + "epoch": 11.081469648562301, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 13874 + }, + { + "epoch": 11.082268370607029, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0365, + "step": 13875 + }, + { + "epoch": 11.083067092651758, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 13876 + }, + { + "epoch": 11.083865814696486, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 13877 + }, + { + "epoch": 11.084664536741213, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 13878 + }, + { + "epoch": 11.085463258785943, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 13879 + }, + { + "epoch": 11.08626198083067, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0205, + "step": 13880 + }, + { + "epoch": 11.0870607028754, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 13881 + }, + { + "epoch": 11.087859424920127, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 13882 + }, + { + "epoch": 11.088658146964857, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 13883 + }, + { + "epoch": 11.089456869009584, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 13884 + }, + { + "epoch": 11.090255591054314, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0173, + "step": 13885 + }, + { + "epoch": 11.091054313099042, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 13886 + }, + { + "epoch": 11.09185303514377, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 13887 + }, + { + "epoch": 11.092651757188499, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 13888 + }, + { + "epoch": 11.093450479233226, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 13889 + }, + { + "epoch": 11.094249201277956, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 13890 + }, + { + "epoch": 11.095047923322683, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 13891 + }, + { + "epoch": 11.095846645367413, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 13892 + }, + { + "epoch": 11.09664536741214, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 13893 + }, + { + "epoch": 11.09744408945687, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 13894 + }, + { + "epoch": 11.098242811501597, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0382, + "step": 13895 + }, + { + "epoch": 11.099041533546325, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 13896 + }, + { + "epoch": 11.099840255591054, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 13897 + }, + { + "epoch": 11.100638977635782, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 13898 + }, + { + "epoch": 11.101437699680512, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 13899 + }, + { + "epoch": 11.10223642172524, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0115, + "step": 13900 + }, + { + "epoch": 11.103035143769969, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 13901 + }, + { + "epoch": 11.103833865814696, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 13902 + }, + { + "epoch": 11.104632587859426, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 13903 + }, + { + "epoch": 11.105431309904153, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 13904 + }, + { + "epoch": 11.106230031948881, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 13905 + }, + { + "epoch": 11.10702875399361, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 13906 + }, + { + "epoch": 11.107827476038338, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 13907 + }, + { + "epoch": 11.108626198083067, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 13908 + }, + { + "epoch": 11.109424920127795, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0122, + "step": 13909 + }, + { + "epoch": 11.110223642172524, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 13910 + }, + { + "epoch": 11.111022364217252, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 13911 + }, + { + "epoch": 11.111821086261982, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 13912 + }, + { + "epoch": 11.11261980830671, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 13913 + }, + { + "epoch": 11.113418530351439, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 13914 + }, + { + "epoch": 11.114217252396166, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 13915 + }, + { + "epoch": 11.115015974440894, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 13916 + }, + { + "epoch": 11.115814696485623, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 13917 + }, + { + "epoch": 11.116613418530351, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 13918 + }, + { + "epoch": 11.11741214057508, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 13919 + }, + { + "epoch": 11.118210862619808, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 13920 + }, + { + "epoch": 11.119009584664537, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 13921 + }, + { + "epoch": 11.119808306709265, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0163, + "step": 13922 + }, + { + "epoch": 11.120607028753994, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0156, + "step": 13923 + }, + { + "epoch": 11.121405750798722, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 13924 + }, + { + "epoch": 11.12220447284345, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0119, + "step": 13925 + }, + { + "epoch": 11.12300319488818, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 13926 + }, + { + "epoch": 11.123801916932907, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 13927 + }, + { + "epoch": 11.124600638977636, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0192, + "step": 13928 + }, + { + "epoch": 11.125399361022364, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 13929 + }, + { + "epoch": 11.126198083067093, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 13930 + }, + { + "epoch": 11.12699680511182, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 13931 + }, + { + "epoch": 11.12779552715655, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 13932 + }, + { + "epoch": 11.128594249201278, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 13933 + }, + { + "epoch": 11.129392971246006, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 13934 + }, + { + "epoch": 11.130191693290735, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 13935 + }, + { + "epoch": 11.130990415335463, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 13936 + }, + { + "epoch": 11.131789137380192, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0168, + "step": 13937 + }, + { + "epoch": 11.13258785942492, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 13938 + }, + { + "epoch": 11.133386581469649, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 13939 + }, + { + "epoch": 11.134185303514377, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 13940 + }, + { + "epoch": 11.134984025559106, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0337, + "step": 13941 + }, + { + "epoch": 11.135782747603834, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0354, + "step": 13942 + }, + { + "epoch": 11.136581469648561, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 13943 + }, + { + "epoch": 11.13738019169329, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 13944 + }, + { + "epoch": 11.138178913738018, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 13945 + }, + { + "epoch": 11.138977635782748, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 13946 + }, + { + "epoch": 11.139776357827476, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 13947 + }, + { + "epoch": 11.140575079872205, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 13948 + }, + { + "epoch": 11.141373801916933, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 13949 + }, + { + "epoch": 11.142172523961662, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 13950 + }, + { + "epoch": 11.14297124600639, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 13951 + }, + { + "epoch": 11.143769968051119, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 13952 + }, + { + "epoch": 11.144568690095847, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 13953 + }, + { + "epoch": 11.145367412140574, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.038, + "step": 13954 + }, + { + "epoch": 11.146166134185304, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 13955 + }, + { + "epoch": 11.146964856230031, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0174, + "step": 13956 + }, + { + "epoch": 11.14776357827476, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 13957 + }, + { + "epoch": 11.148562300319488, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 13958 + }, + { + "epoch": 11.149361022364218, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 13959 + }, + { + "epoch": 11.150159744408946, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 13960 + }, + { + "epoch": 11.150958466453675, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0196, + "step": 13961 + }, + { + "epoch": 11.151757188498403, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 13962 + }, + { + "epoch": 11.15255591054313, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 13963 + }, + { + "epoch": 11.15335463258786, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 13964 + }, + { + "epoch": 11.154153354632587, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 13965 + }, + { + "epoch": 11.154952076677317, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 13966 + }, + { + "epoch": 11.155750798722044, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 13967 + }, + { + "epoch": 11.156549520766774, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 13968 + }, + { + "epoch": 11.157348242811501, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 13969 + }, + { + "epoch": 11.15814696485623, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 13970 + }, + { + "epoch": 11.158945686900958, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 13971 + }, + { + "epoch": 11.159744408945686, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0338, + "step": 13972 + }, + { + "epoch": 11.160543130990416, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 13973 + }, + { + "epoch": 11.161341853035143, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 13974 + }, + { + "epoch": 11.162140575079873, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 13975 + }, + { + "epoch": 11.1629392971246, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 13976 + }, + { + "epoch": 11.16373801916933, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 13977 + }, + { + "epoch": 11.164536741214057, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 13978 + }, + { + "epoch": 11.165335463258787, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 13979 + }, + { + "epoch": 11.166134185303514, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0423, + "step": 13980 + }, + { + "epoch": 11.166932907348242, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 13981 + }, + { + "epoch": 11.167731629392971, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 13982 + }, + { + "epoch": 11.168530351437699, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 13983 + }, + { + "epoch": 11.169329073482428, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 13984 + }, + { + "epoch": 11.170127795527156, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 13985 + }, + { + "epoch": 11.170926517571885, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 13986 + }, + { + "epoch": 11.171725239616613, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 13987 + }, + { + "epoch": 11.172523961661343, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 13988 + }, + { + "epoch": 11.17332268370607, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 13989 + }, + { + "epoch": 11.1741214057508, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 13990 + }, + { + "epoch": 11.174920127795527, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 13991 + }, + { + "epoch": 11.175718849840255, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 13992 + }, + { + "epoch": 11.176517571884984, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 13993 + }, + { + "epoch": 11.177316293929712, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 13994 + }, + { + "epoch": 11.178115015974441, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 13995 + }, + { + "epoch": 11.178913738019169, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 13996 + }, + { + "epoch": 11.179712460063898, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 13997 + }, + { + "epoch": 11.180511182108626, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 13998 + }, + { + "epoch": 11.181309904153355, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 13999 + }, + { + "epoch": 11.182108626198083, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 14000 + }, + { + "epoch": 11.18290734824281, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 14001 + }, + { + "epoch": 11.18370607028754, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 14002 + }, + { + "epoch": 11.184504792332268, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 14003 + }, + { + "epoch": 11.185303514376997, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 14004 + }, + { + "epoch": 11.186102236421725, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 14005 + }, + { + "epoch": 11.186900958466454, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 14006 + }, + { + "epoch": 11.187699680511182, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0389, + "step": 14007 + }, + { + "epoch": 11.188498402555911, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 14008 + }, + { + "epoch": 11.189297124600639, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 14009 + }, + { + "epoch": 11.190095846645367, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 14010 + }, + { + "epoch": 11.190894568690096, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 14011 + }, + { + "epoch": 11.191693290734824, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 14012 + }, + { + "epoch": 11.192492012779553, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 14013 + }, + { + "epoch": 11.19329073482428, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 14014 + }, + { + "epoch": 11.19408945686901, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 14015 + }, + { + "epoch": 11.194888178913738, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 14016 + }, + { + "epoch": 11.195686900958467, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0183, + "step": 14017 + }, + { + "epoch": 11.196485623003195, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0185, + "step": 14018 + }, + { + "epoch": 11.197284345047922, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 14019 + }, + { + "epoch": 11.198083067092652, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 14020 + }, + { + "epoch": 11.19888178913738, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 14021 + }, + { + "epoch": 11.199680511182109, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 14022 + }, + { + "epoch": 11.200479233226837, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 14023 + }, + { + "epoch": 11.201277955271566, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 14024 + }, + { + "epoch": 11.202076677316294, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 14025 + }, + { + "epoch": 11.202875399361023, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 14026 + }, + { + "epoch": 11.20367412140575, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 14027 + }, + { + "epoch": 11.204472843450478, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 14028 + }, + { + "epoch": 11.205271565495208, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 14029 + }, + { + "epoch": 11.206070287539935, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 14030 + }, + { + "epoch": 11.206869009584665, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 14031 + }, + { + "epoch": 11.207667731629392, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 14032 + }, + { + "epoch": 11.208466453674122, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 14033 + }, + { + "epoch": 11.20926517571885, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 14034 + }, + { + "epoch": 11.210063897763579, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 14035 + }, + { + "epoch": 11.210862619808307, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 14036 + }, + { + "epoch": 11.211661341853036, + "grad_norm": 0.1640625, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 14037 + }, + { + "epoch": 11.212460063897764, + "grad_norm": 0.16015625, + "learning_rate": 0.0005, + "loss": 1.036, + "step": 14038 + }, + { + "epoch": 11.213258785942491, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 14039 + }, + { + "epoch": 11.21405750798722, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 14040 + }, + { + "epoch": 11.214856230031948, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0155, + "step": 14041 + }, + { + "epoch": 11.215654952076678, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 14042 + }, + { + "epoch": 11.216453674121405, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 14043 + }, + { + "epoch": 11.217252396166135, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 14044 + }, + { + "epoch": 11.218051118210862, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0205, + "step": 14045 + }, + { + "epoch": 11.218849840255592, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 14046 + }, + { + "epoch": 11.21964856230032, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 14047 + }, + { + "epoch": 11.220447284345047, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 14048 + }, + { + "epoch": 11.221246006389777, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 14049 + }, + { + "epoch": 11.222044728434504, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 14050 + }, + { + "epoch": 11.222843450479234, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 14051 + }, + { + "epoch": 11.223642172523961, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 14052 + }, + { + "epoch": 11.22444089456869, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 14053 + }, + { + "epoch": 11.225239616613418, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 14054 + }, + { + "epoch": 11.226038338658148, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 14055 + }, + { + "epoch": 11.226837060702875, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 14056 + }, + { + "epoch": 11.227635782747603, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 14057 + }, + { + "epoch": 11.228434504792332, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 14058 + }, + { + "epoch": 11.22923322683706, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 14059 + }, + { + "epoch": 11.23003194888179, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 14060 + }, + { + "epoch": 11.230830670926517, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 14061 + }, + { + "epoch": 11.231629392971247, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 14062 + }, + { + "epoch": 11.232428115015974, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 14063 + }, + { + "epoch": 11.233226837060704, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 14064 + }, + { + "epoch": 11.234025559105431, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 14065 + }, + { + "epoch": 11.23482428115016, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 14066 + }, + { + "epoch": 11.235623003194888, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 14067 + }, + { + "epoch": 11.236421725239616, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 14068 + }, + { + "epoch": 11.237220447284345, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 14069 + }, + { + "epoch": 11.238019169329073, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 14070 + }, + { + "epoch": 11.238817891373802, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0427, + "step": 14071 + }, + { + "epoch": 11.23961661341853, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 14072 + }, + { + "epoch": 11.24041533546326, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 14073 + }, + { + "epoch": 11.241214057507987, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0192, + "step": 14074 + }, + { + "epoch": 11.242012779552716, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 14075 + }, + { + "epoch": 11.242811501597444, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 14076 + }, + { + "epoch": 11.243610223642172, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 14077 + }, + { + "epoch": 11.244408945686901, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0196, + "step": 14078 + }, + { + "epoch": 11.245207667731629, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 14079 + }, + { + "epoch": 11.246006389776358, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 14080 + }, + { + "epoch": 11.246805111821086, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 14081 + }, + { + "epoch": 11.247603833865815, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 14082 + }, + { + "epoch": 11.248402555910543, + "grad_norm": 0.177734375, + "learning_rate": 0.0005, + "loss": 1.0427, + "step": 14083 + }, + { + "epoch": 11.249201277955272, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 14084 + }, + { + "epoch": 11.25, + "grad_norm": 0.1806640625, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 14085 + }, + { + "epoch": 11.250798722044728, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 14086 + }, + { + "epoch": 11.251597444089457, + "grad_norm": 0.1923828125, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 14087 + }, + { + "epoch": 11.252396166134185, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0353, + "step": 14088 + }, + { + "epoch": 11.253194888178914, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0177, + "step": 14089 + }, + { + "epoch": 11.253993610223642, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 14090 + }, + { + "epoch": 11.254792332268371, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 14091 + }, + { + "epoch": 11.255591054313099, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0114, + "step": 14092 + }, + { + "epoch": 11.256389776357828, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 14093 + }, + { + "epoch": 11.257188498402556, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 14094 + }, + { + "epoch": 11.257987220447284, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 14095 + }, + { + "epoch": 11.258785942492013, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 14096 + }, + { + "epoch": 11.25958466453674, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0166, + "step": 14097 + }, + { + "epoch": 11.26038338658147, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 14098 + }, + { + "epoch": 11.261182108626198, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 14099 + }, + { + "epoch": 11.261980830670927, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 14100 + }, + { + "epoch": 11.262779552715655, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 14101 + }, + { + "epoch": 11.263578274760384, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 14102 + }, + { + "epoch": 11.264376996805112, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0188, + "step": 14103 + }, + { + "epoch": 11.26517571884984, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 14104 + }, + { + "epoch": 11.265974440894569, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 14105 + }, + { + "epoch": 11.266773162939296, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 14106 + }, + { + "epoch": 11.267571884984026, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 14107 + }, + { + "epoch": 11.268370607028753, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0128, + "step": 14108 + }, + { + "epoch": 11.269169329073483, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 14109 + }, + { + "epoch": 11.26996805111821, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0187, + "step": 14110 + }, + { + "epoch": 11.27076677316294, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 14111 + }, + { + "epoch": 11.271565495207668, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 14112 + }, + { + "epoch": 11.272364217252397, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 14113 + }, + { + "epoch": 11.273162939297125, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 14114 + }, + { + "epoch": 11.273961661341852, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 14115 + }, + { + "epoch": 11.274760383386582, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0403, + "step": 14116 + }, + { + "epoch": 11.27555910543131, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 14117 + }, + { + "epoch": 11.276357827476039, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 14118 + }, + { + "epoch": 11.277156549520766, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 14119 + }, + { + "epoch": 11.277955271565496, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 14120 + }, + { + "epoch": 11.278753993610223, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 14121 + }, + { + "epoch": 11.279552715654953, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 14122 + }, + { + "epoch": 11.28035143769968, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 14123 + }, + { + "epoch": 11.281150159744408, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0148, + "step": 14124 + }, + { + "epoch": 11.281948881789138, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 14125 + }, + { + "epoch": 11.282747603833865, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 14126 + }, + { + "epoch": 11.283546325878595, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0185, + "step": 14127 + }, + { + "epoch": 11.284345047923322, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 14128 + }, + { + "epoch": 11.285143769968052, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 14129 + }, + { + "epoch": 11.28594249201278, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 14130 + }, + { + "epoch": 11.286741214057509, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0159, + "step": 14131 + }, + { + "epoch": 11.287539936102236, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 14132 + }, + { + "epoch": 11.288338658146964, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 14133 + }, + { + "epoch": 11.289137380191693, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0183, + "step": 14134 + }, + { + "epoch": 11.289936102236421, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0173, + "step": 14135 + }, + { + "epoch": 11.29073482428115, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 14136 + }, + { + "epoch": 11.291533546325878, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 14137 + }, + { + "epoch": 11.292332268370608, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 14138 + }, + { + "epoch": 11.293130990415335, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 14139 + }, + { + "epoch": 11.293929712460065, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0338, + "step": 14140 + }, + { + "epoch": 11.294728434504792, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 14141 + }, + { + "epoch": 11.295527156549522, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 14142 + }, + { + "epoch": 11.29632587859425, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 14143 + }, + { + "epoch": 11.297124600638977, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.012, + "step": 14144 + }, + { + "epoch": 11.297923322683706, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 14145 + }, + { + "epoch": 11.298722044728434, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 14146 + }, + { + "epoch": 11.299520766773163, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 14147 + }, + { + "epoch": 11.300319488817891, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 14148 + }, + { + "epoch": 11.30111821086262, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 14149 + }, + { + "epoch": 11.301916932907348, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 14150 + }, + { + "epoch": 11.302715654952078, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 14151 + }, + { + "epoch": 11.303514376996805, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 14152 + }, + { + "epoch": 11.304313099041533, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 14153 + }, + { + "epoch": 11.305111821086262, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 14154 + }, + { + "epoch": 11.30591054313099, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 14155 + }, + { + "epoch": 11.30670926517572, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0172, + "step": 14156 + }, + { + "epoch": 11.307507987220447, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 14157 + }, + { + "epoch": 11.308306709265176, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 14158 + }, + { + "epoch": 11.309105431309904, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 14159 + }, + { + "epoch": 11.309904153354633, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 14160 + }, + { + "epoch": 11.310702875399361, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 14161 + }, + { + "epoch": 11.311501597444089, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0164, + "step": 14162 + }, + { + "epoch": 11.312300319488818, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 14163 + }, + { + "epoch": 11.313099041533546, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 14164 + }, + { + "epoch": 11.313897763578275, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 14165 + }, + { + "epoch": 11.314696485623003, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 14166 + }, + { + "epoch": 11.315495207667732, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 14167 + }, + { + "epoch": 11.31629392971246, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 14168 + }, + { + "epoch": 11.31709265175719, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 14169 + }, + { + "epoch": 11.317891373801917, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 14170 + }, + { + "epoch": 11.318690095846645, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 14171 + }, + { + "epoch": 11.319488817891374, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 14172 + }, + { + "epoch": 11.320287539936102, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 14173 + }, + { + "epoch": 11.321086261980831, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 14174 + }, + { + "epoch": 11.321884984025559, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 14175 + }, + { + "epoch": 11.322683706070288, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 14176 + }, + { + "epoch": 11.323482428115016, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0179, + "step": 14177 + }, + { + "epoch": 11.324281150159745, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 14178 + }, + { + "epoch": 11.325079872204473, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 14179 + }, + { + "epoch": 11.3258785942492, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 14180 + }, + { + "epoch": 11.32667731629393, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 14181 + }, + { + "epoch": 11.327476038338657, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 14182 + }, + { + "epoch": 11.328274760383387, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0118, + "step": 14183 + }, + { + "epoch": 11.329073482428115, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 14184 + }, + { + "epoch": 11.329872204472844, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 14185 + }, + { + "epoch": 11.330670926517572, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 14186 + }, + { + "epoch": 11.331469648562301, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 14187 + }, + { + "epoch": 11.332268370607029, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 14188 + }, + { + "epoch": 11.333067092651758, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 14189 + }, + { + "epoch": 11.333865814696486, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0177, + "step": 14190 + }, + { + "epoch": 11.334664536741213, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 14191 + }, + { + "epoch": 11.335463258785943, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 14192 + }, + { + "epoch": 11.33626198083067, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 14193 + }, + { + "epoch": 11.3370607028754, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 14194 + }, + { + "epoch": 11.337859424920127, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 14195 + }, + { + "epoch": 11.338658146964857, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 14196 + }, + { + "epoch": 11.339456869009584, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.038, + "step": 14197 + }, + { + "epoch": 11.340255591054314, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 14198 + }, + { + "epoch": 11.341054313099042, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 14199 + }, + { + "epoch": 11.34185303514377, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 14200 + }, + { + "epoch": 11.342651757188499, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 14201 + }, + { + "epoch": 11.343450479233226, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 14202 + }, + { + "epoch": 11.344249201277956, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 14203 + }, + { + "epoch": 11.345047923322683, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 14204 + }, + { + "epoch": 11.345846645367413, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 14205 + }, + { + "epoch": 11.34664536741214, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 14206 + }, + { + "epoch": 11.34744408945687, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 14207 + }, + { + "epoch": 11.348242811501597, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 14208 + }, + { + "epoch": 11.349041533546325, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 14209 + }, + { + "epoch": 11.349840255591054, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 14210 + }, + { + "epoch": 11.350638977635782, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 14211 + }, + { + "epoch": 11.351437699680512, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 14212 + }, + { + "epoch": 11.35223642172524, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 14213 + }, + { + "epoch": 11.353035143769969, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 14214 + }, + { + "epoch": 11.353833865814696, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 14215 + }, + { + "epoch": 11.354632587859426, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 14216 + }, + { + "epoch": 11.355431309904153, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 14217 + }, + { + "epoch": 11.356230031948883, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 14218 + }, + { + "epoch": 11.35702875399361, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 14219 + }, + { + "epoch": 11.357827476038338, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 14220 + }, + { + "epoch": 11.358626198083067, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 14221 + }, + { + "epoch": 11.359424920127795, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 14222 + }, + { + "epoch": 11.360223642172524, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 14223 + }, + { + "epoch": 11.361022364217252, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 14224 + }, + { + "epoch": 11.361821086261982, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 14225 + }, + { + "epoch": 11.36261980830671, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 14226 + }, + { + "epoch": 11.363418530351439, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 14227 + }, + { + "epoch": 11.364217252396166, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 14228 + }, + { + "epoch": 11.365015974440894, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 14229 + }, + { + "epoch": 11.365814696485623, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 14230 + }, + { + "epoch": 11.366613418530351, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 14231 + }, + { + "epoch": 11.36741214057508, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 14232 + }, + { + "epoch": 11.368210862619808, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 14233 + }, + { + "epoch": 11.369009584664537, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 14234 + }, + { + "epoch": 11.369808306709265, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 14235 + }, + { + "epoch": 11.370607028753994, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 14236 + }, + { + "epoch": 11.371405750798722, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0098, + "step": 14237 + }, + { + "epoch": 11.37220447284345, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 14238 + }, + { + "epoch": 11.37300319488818, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 14239 + }, + { + "epoch": 11.373801916932907, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 14240 + }, + { + "epoch": 11.374600638977636, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 14241 + }, + { + "epoch": 11.375399361022364, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0385, + "step": 14242 + }, + { + "epoch": 11.376198083067093, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 14243 + }, + { + "epoch": 11.37699680511182, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 14244 + }, + { + "epoch": 11.37779552715655, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 14245 + }, + { + "epoch": 11.378594249201278, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 14246 + }, + { + "epoch": 11.379392971246006, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 14247 + }, + { + "epoch": 11.380191693290735, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 14248 + }, + { + "epoch": 11.380990415335463, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 14249 + }, + { + "epoch": 11.381789137380192, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 14250 + }, + { + "epoch": 11.38258785942492, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 14251 + }, + { + "epoch": 11.383386581469649, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0157, + "step": 14252 + }, + { + "epoch": 11.384185303514377, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 14253 + }, + { + "epoch": 11.384984025559106, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0132, + "step": 14254 + }, + { + "epoch": 11.385782747603834, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 14255 + }, + { + "epoch": 11.386581469648561, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0127, + "step": 14256 + }, + { + "epoch": 11.38738019169329, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 14257 + }, + { + "epoch": 11.388178913738018, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 14258 + }, + { + "epoch": 11.388977635782748, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 14259 + }, + { + "epoch": 11.389776357827476, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 14260 + }, + { + "epoch": 11.390575079872205, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 14261 + }, + { + "epoch": 11.391373801916933, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 14262 + }, + { + "epoch": 11.392172523961662, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 14263 + }, + { + "epoch": 11.39297124600639, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 14264 + }, + { + "epoch": 11.393769968051117, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 14265 + }, + { + "epoch": 11.394568690095847, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 14266 + }, + { + "epoch": 11.395367412140574, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0161, + "step": 14267 + }, + { + "epoch": 11.396166134185304, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 14268 + }, + { + "epoch": 11.396964856230031, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0125, + "step": 14269 + }, + { + "epoch": 11.39776357827476, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 14270 + }, + { + "epoch": 11.398562300319488, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 14271 + }, + { + "epoch": 11.399361022364218, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 14272 + }, + { + "epoch": 11.400159744408946, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 14273 + }, + { + "epoch": 11.400958466453675, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 14274 + }, + { + "epoch": 11.401757188498403, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 14275 + }, + { + "epoch": 11.40255591054313, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 14276 + }, + { + "epoch": 11.40335463258786, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 14277 + }, + { + "epoch": 11.404153354632587, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 14278 + }, + { + "epoch": 11.404952076677317, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 14279 + }, + { + "epoch": 11.405750798722044, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 14280 + }, + { + "epoch": 11.406549520766774, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 14281 + }, + { + "epoch": 11.407348242811501, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 14282 + }, + { + "epoch": 11.40814696485623, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0177, + "step": 14283 + }, + { + "epoch": 11.408945686900958, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 14284 + }, + { + "epoch": 11.409744408945686, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 14285 + }, + { + "epoch": 11.410543130990416, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 14286 + }, + { + "epoch": 11.411341853035143, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 14287 + }, + { + "epoch": 11.412140575079873, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0163, + "step": 14288 + }, + { + "epoch": 11.4129392971246, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 14289 + }, + { + "epoch": 11.41373801916933, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 14290 + }, + { + "epoch": 11.414536741214057, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0138, + "step": 14291 + }, + { + "epoch": 11.415335463258787, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 14292 + }, + { + "epoch": 11.416134185303514, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 14293 + }, + { + "epoch": 11.416932907348242, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0167, + "step": 14294 + }, + { + "epoch": 11.417731629392971, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0166, + "step": 14295 + }, + { + "epoch": 11.418530351437699, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0153, + "step": 14296 + }, + { + "epoch": 11.419329073482428, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 14297 + }, + { + "epoch": 11.420127795527156, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 14298 + }, + { + "epoch": 11.420926517571885, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 14299 + }, + { + "epoch": 11.421725239616613, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 14300 + }, + { + "epoch": 11.422523961661343, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 14301 + }, + { + "epoch": 11.42332268370607, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0148, + "step": 14302 + }, + { + "epoch": 11.4241214057508, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 14303 + }, + { + "epoch": 11.424920127795527, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0116, + "step": 14304 + }, + { + "epoch": 11.425718849840255, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 14305 + }, + { + "epoch": 11.426517571884984, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 14306 + }, + { + "epoch": 11.427316293929712, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 14307 + }, + { + "epoch": 11.428115015974441, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 14308 + }, + { + "epoch": 11.428913738019169, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 14309 + }, + { + "epoch": 11.429712460063898, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 14310 + }, + { + "epoch": 11.430511182108626, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 14311 + }, + { + "epoch": 11.431309904153355, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 14312 + }, + { + "epoch": 11.432108626198083, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 14313 + }, + { + "epoch": 11.43290734824281, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 14314 + }, + { + "epoch": 11.43370607028754, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 14315 + }, + { + "epoch": 11.434504792332268, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0192, + "step": 14316 + }, + { + "epoch": 11.435303514376997, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 14317 + }, + { + "epoch": 11.436102236421725, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0205, + "step": 14318 + }, + { + "epoch": 11.436900958466454, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 14319 + }, + { + "epoch": 11.437699680511182, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 14320 + }, + { + "epoch": 11.438498402555911, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 14321 + }, + { + "epoch": 11.439297124600639, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 14322 + }, + { + "epoch": 11.440095846645367, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 14323 + }, + { + "epoch": 11.440894568690096, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 14324 + }, + { + "epoch": 11.441693290734824, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 14325 + }, + { + "epoch": 11.442492012779553, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 14326 + }, + { + "epoch": 11.44329073482428, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0182, + "step": 14327 + }, + { + "epoch": 11.44408945686901, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 14328 + }, + { + "epoch": 11.444888178913738, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 14329 + }, + { + "epoch": 11.445686900958467, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 14330 + }, + { + "epoch": 11.446485623003195, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0133, + "step": 14331 + }, + { + "epoch": 11.447284345047922, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 14332 + }, + { + "epoch": 11.448083067092652, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 14333 + }, + { + "epoch": 11.44888178913738, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 14334 + }, + { + "epoch": 11.449680511182109, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 14335 + }, + { + "epoch": 11.450479233226837, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 14336 + }, + { + "epoch": 11.451277955271566, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 14337 + }, + { + "epoch": 11.452076677316294, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0192, + "step": 14338 + }, + { + "epoch": 11.452875399361023, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 14339 + }, + { + "epoch": 11.45367412140575, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 14340 + }, + { + "epoch": 11.454472843450478, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 14341 + }, + { + "epoch": 11.455271565495208, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 14342 + }, + { + "epoch": 11.456070287539935, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 14343 + }, + { + "epoch": 11.456869009584665, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 14344 + }, + { + "epoch": 11.457667731629392, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 14345 + }, + { + "epoch": 11.458466453674122, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 14346 + }, + { + "epoch": 11.45926517571885, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 14347 + }, + { + "epoch": 11.460063897763579, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 14348 + }, + { + "epoch": 11.460862619808307, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 14349 + }, + { + "epoch": 11.461661341853036, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 14350 + }, + { + "epoch": 11.462460063897764, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0348, + "step": 14351 + }, + { + "epoch": 11.463258785942491, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 14352 + }, + { + "epoch": 11.46405750798722, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 14353 + }, + { + "epoch": 11.464856230031948, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 14354 + }, + { + "epoch": 11.465654952076678, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 14355 + }, + { + "epoch": 11.466453674121405, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 14356 + }, + { + "epoch": 11.467252396166135, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 14357 + }, + { + "epoch": 11.468051118210862, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 14358 + }, + { + "epoch": 11.468849840255592, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 14359 + }, + { + "epoch": 11.46964856230032, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0164, + "step": 14360 + }, + { + "epoch": 11.470447284345047, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 14361 + }, + { + "epoch": 11.471246006389777, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 14362 + }, + { + "epoch": 11.472044728434504, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0144, + "step": 14363 + }, + { + "epoch": 11.472843450479234, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 14364 + }, + { + "epoch": 11.473642172523961, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 14365 + }, + { + "epoch": 11.47444089456869, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 14366 + }, + { + "epoch": 11.475239616613418, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 14367 + }, + { + "epoch": 11.476038338658148, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 14368 + }, + { + "epoch": 11.476837060702875, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 14369 + }, + { + "epoch": 11.477635782747603, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0199, + "step": 14370 + }, + { + "epoch": 11.478434504792332, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 14371 + }, + { + "epoch": 11.47923322683706, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 14372 + }, + { + "epoch": 11.48003194888179, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 14373 + }, + { + "epoch": 11.480830670926517, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 14374 + }, + { + "epoch": 11.481629392971247, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 14375 + }, + { + "epoch": 11.482428115015974, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 14376 + }, + { + "epoch": 11.483226837060704, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 14377 + }, + { + "epoch": 11.484025559105431, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0145, + "step": 14378 + }, + { + "epoch": 11.48482428115016, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 14379 + }, + { + "epoch": 11.485623003194888, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 14380 + }, + { + "epoch": 11.486421725239616, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 14381 + }, + { + "epoch": 11.487220447284345, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 14382 + }, + { + "epoch": 11.488019169329073, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 14383 + }, + { + "epoch": 11.488817891373802, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 14384 + }, + { + "epoch": 11.48961661341853, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 14385 + }, + { + "epoch": 11.49041533546326, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 14386 + }, + { + "epoch": 11.491214057507987, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 14387 + }, + { + "epoch": 11.492012779552716, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 14388 + }, + { + "epoch": 11.492811501597444, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 14389 + }, + { + "epoch": 11.493610223642172, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 14390 + }, + { + "epoch": 11.494408945686901, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0185, + "step": 14391 + }, + { + "epoch": 11.495207667731629, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 14392 + }, + { + "epoch": 11.496006389776358, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 14393 + }, + { + "epoch": 11.496805111821086, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 14394 + }, + { + "epoch": 11.497603833865815, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 14395 + }, + { + "epoch": 11.498402555910543, + "grad_norm": 0.1904296875, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 14396 + }, + { + "epoch": 11.499201277955272, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 14397 + }, + { + "epoch": 11.5, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 14398 + }, + { + "epoch": 11.500798722044728, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.015, + "step": 14399 + }, + { + "epoch": 11.501597444089457, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 14400 + }, + { + "epoch": 11.502396166134185, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 14401 + }, + { + "epoch": 11.503194888178914, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 14402 + }, + { + "epoch": 11.503993610223642, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 14403 + }, + { + "epoch": 11.504792332268371, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0171, + "step": 14404 + }, + { + "epoch": 11.505591054313099, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 14405 + }, + { + "epoch": 11.506389776357828, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 14406 + }, + { + "epoch": 11.507188498402556, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0183, + "step": 14407 + }, + { + "epoch": 11.507987220447284, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0174, + "step": 14408 + }, + { + "epoch": 11.508785942492013, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 14409 + }, + { + "epoch": 11.50958466453674, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 14410 + }, + { + "epoch": 11.51038338658147, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 14411 + }, + { + "epoch": 11.511182108626198, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 14412 + }, + { + "epoch": 11.511980830670927, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 14413 + }, + { + "epoch": 11.512779552715655, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 14414 + }, + { + "epoch": 11.513578274760384, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 14415 + }, + { + "epoch": 11.514376996805112, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 14416 + }, + { + "epoch": 11.51517571884984, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 14417 + }, + { + "epoch": 11.515974440894569, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 14418 + }, + { + "epoch": 11.516773162939296, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 14419 + }, + { + "epoch": 11.517571884984026, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0143, + "step": 14420 + }, + { + "epoch": 11.518370607028753, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 14421 + }, + { + "epoch": 11.519169329073483, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 14422 + }, + { + "epoch": 11.51996805111821, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 14423 + }, + { + "epoch": 11.52076677316294, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 14424 + }, + { + "epoch": 11.521565495207668, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 14425 + }, + { + "epoch": 11.522364217252395, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.013, + "step": 14426 + }, + { + "epoch": 11.523162939297125, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 14427 + }, + { + "epoch": 11.523961661341852, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 14428 + }, + { + "epoch": 11.524760383386582, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 14429 + }, + { + "epoch": 11.52555910543131, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 14430 + }, + { + "epoch": 11.526357827476039, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 14431 + }, + { + "epoch": 11.527156549520766, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 14432 + }, + { + "epoch": 11.527955271565496, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 14433 + }, + { + "epoch": 11.528753993610223, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 14434 + }, + { + "epoch": 11.529552715654953, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0178, + "step": 14435 + }, + { + "epoch": 11.53035143769968, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 14436 + }, + { + "epoch": 11.531150159744408, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 14437 + }, + { + "epoch": 11.531948881789138, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0372, + "step": 14438 + }, + { + "epoch": 11.532747603833865, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 14439 + }, + { + "epoch": 11.533546325878595, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 14440 + }, + { + "epoch": 11.534345047923322, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 14441 + }, + { + "epoch": 11.535143769968052, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0179, + "step": 14442 + }, + { + "epoch": 11.53594249201278, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 14443 + }, + { + "epoch": 11.536741214057509, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 14444 + }, + { + "epoch": 11.537539936102236, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 14445 + }, + { + "epoch": 11.538338658146966, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 14446 + }, + { + "epoch": 11.539137380191693, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 14447 + }, + { + "epoch": 11.539936102236421, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0147, + "step": 14448 + }, + { + "epoch": 11.54073482428115, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 14449 + }, + { + "epoch": 11.541533546325878, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 14450 + }, + { + "epoch": 11.542332268370608, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 14451 + }, + { + "epoch": 11.543130990415335, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 14452 + }, + { + "epoch": 11.543929712460065, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 14453 + }, + { + "epoch": 11.544728434504792, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 14454 + }, + { + "epoch": 11.545527156549522, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 14455 + }, + { + "epoch": 11.54632587859425, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 14456 + }, + { + "epoch": 11.547124600638977, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 14457 + }, + { + "epoch": 11.547923322683706, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 14458 + }, + { + "epoch": 11.548722044728434, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 14459 + }, + { + "epoch": 11.549520766773163, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 14460 + }, + { + "epoch": 11.550319488817891, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 14461 + }, + { + "epoch": 11.55111821086262, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 14462 + }, + { + "epoch": 11.551916932907348, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 14463 + }, + { + "epoch": 11.552715654952078, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 14464 + }, + { + "epoch": 11.553514376996805, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 14465 + }, + { + "epoch": 11.554313099041533, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 14466 + }, + { + "epoch": 11.555111821086262, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 14467 + }, + { + "epoch": 11.55591054313099, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 14468 + }, + { + "epoch": 11.55670926517572, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 14469 + }, + { + "epoch": 11.557507987220447, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 14470 + }, + { + "epoch": 11.558306709265176, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 14471 + }, + { + "epoch": 11.559105431309904, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 14472 + }, + { + "epoch": 11.559904153354633, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 14473 + }, + { + "epoch": 11.560702875399361, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0181, + "step": 14474 + }, + { + "epoch": 11.561501597444089, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 14475 + }, + { + "epoch": 11.562300319488818, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 14476 + }, + { + "epoch": 11.563099041533546, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 14477 + }, + { + "epoch": 11.563897763578275, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 14478 + }, + { + "epoch": 11.564696485623003, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 14479 + }, + { + "epoch": 11.565495207667732, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 14480 + }, + { + "epoch": 11.56629392971246, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0427, + "step": 14481 + }, + { + "epoch": 11.56709265175719, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 14482 + }, + { + "epoch": 11.567891373801917, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 14483 + }, + { + "epoch": 11.568690095846645, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 14484 + }, + { + "epoch": 11.569488817891374, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 14485 + }, + { + "epoch": 11.570287539936102, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 14486 + }, + { + "epoch": 11.571086261980831, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 14487 + }, + { + "epoch": 11.571884984025559, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0169, + "step": 14488 + }, + { + "epoch": 11.572683706070288, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 14489 + }, + { + "epoch": 11.573482428115016, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 14490 + }, + { + "epoch": 11.574281150159745, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 14491 + }, + { + "epoch": 11.575079872204473, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 14492 + }, + { + "epoch": 11.5758785942492, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 14493 + }, + { + "epoch": 11.57667731629393, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 14494 + }, + { + "epoch": 11.577476038338657, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 14495 + }, + { + "epoch": 11.578274760383387, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 14496 + }, + { + "epoch": 11.579073482428115, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 14497 + }, + { + "epoch": 11.579872204472844, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 14498 + }, + { + "epoch": 11.580670926517572, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0177, + "step": 14499 + }, + { + "epoch": 11.581469648562301, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 14500 + }, + { + "epoch": 11.582268370607029, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 14501 + }, + { + "epoch": 11.583067092651756, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 14502 + }, + { + "epoch": 11.583865814696486, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 14503 + }, + { + "epoch": 11.584664536741213, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 14504 + }, + { + "epoch": 11.585463258785943, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 14505 + }, + { + "epoch": 11.58626198083067, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0128, + "step": 14506 + }, + { + "epoch": 11.5870607028754, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 14507 + }, + { + "epoch": 11.587859424920127, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 14508 + }, + { + "epoch": 11.588658146964857, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 14509 + }, + { + "epoch": 11.589456869009584, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.014, + "step": 14510 + }, + { + "epoch": 11.590255591054314, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 14511 + }, + { + "epoch": 11.591054313099042, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 14512 + }, + { + "epoch": 11.59185303514377, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 14513 + }, + { + "epoch": 11.592651757188499, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 14514 + }, + { + "epoch": 11.593450479233226, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 14515 + }, + { + "epoch": 11.594249201277956, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 14516 + }, + { + "epoch": 11.595047923322683, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0205, + "step": 14517 + }, + { + "epoch": 11.595846645367413, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 14518 + }, + { + "epoch": 11.59664536741214, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0183, + "step": 14519 + }, + { + "epoch": 11.59744408945687, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 14520 + }, + { + "epoch": 11.598242811501597, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 14521 + }, + { + "epoch": 11.599041533546325, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 14522 + }, + { + "epoch": 11.599840255591054, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0195, + "step": 14523 + }, + { + "epoch": 11.600638977635782, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0169, + "step": 14524 + }, + { + "epoch": 11.601437699680512, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 14525 + }, + { + "epoch": 11.60223642172524, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 14526 + }, + { + "epoch": 11.603035143769969, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 14527 + }, + { + "epoch": 11.603833865814696, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0163, + "step": 14528 + }, + { + "epoch": 11.604632587859426, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 14529 + }, + { + "epoch": 11.605431309904153, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 14530 + }, + { + "epoch": 11.606230031948883, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 14531 + }, + { + "epoch": 11.60702875399361, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 14532 + }, + { + "epoch": 11.607827476038338, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.016, + "step": 14533 + }, + { + "epoch": 11.608626198083067, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 14534 + }, + { + "epoch": 11.609424920127795, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 14535 + }, + { + "epoch": 11.610223642172524, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 14536 + }, + { + "epoch": 11.611022364217252, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 14537 + }, + { + "epoch": 11.611821086261982, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 14538 + }, + { + "epoch": 11.61261980830671, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 14539 + }, + { + "epoch": 11.613418530351439, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 14540 + }, + { + "epoch": 11.614217252396166, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 14541 + }, + { + "epoch": 11.615015974440894, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0162, + "step": 14542 + }, + { + "epoch": 11.615814696485623, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 14543 + }, + { + "epoch": 11.616613418530351, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 14544 + }, + { + "epoch": 11.61741214057508, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.034, + "step": 14545 + }, + { + "epoch": 11.618210862619808, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 14546 + }, + { + "epoch": 11.619009584664537, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 14547 + }, + { + "epoch": 11.619808306709265, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 14548 + }, + { + "epoch": 11.620607028753994, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 14549 + }, + { + "epoch": 11.621405750798722, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 14550 + }, + { + "epoch": 11.62220447284345, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 14551 + }, + { + "epoch": 11.62300319488818, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 14552 + }, + { + "epoch": 11.623801916932907, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 14553 + }, + { + "epoch": 11.624600638977636, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 14554 + }, + { + "epoch": 11.625399361022364, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0418, + "step": 14555 + }, + { + "epoch": 11.626198083067093, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 14556 + }, + { + "epoch": 11.62699680511182, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 14557 + }, + { + "epoch": 11.62779552715655, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 14558 + }, + { + "epoch": 11.628594249201278, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 14559 + }, + { + "epoch": 11.629392971246006, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 14560 + }, + { + "epoch": 11.630191693290735, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 14561 + }, + { + "epoch": 11.630990415335463, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 14562 + }, + { + "epoch": 11.631789137380192, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 14563 + }, + { + "epoch": 11.63258785942492, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 14564 + }, + { + "epoch": 11.633386581469649, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 14565 + }, + { + "epoch": 11.634185303514377, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 14566 + }, + { + "epoch": 11.634984025559106, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 14567 + }, + { + "epoch": 11.635782747603834, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0205, + "step": 14568 + }, + { + "epoch": 11.636581469648561, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 14569 + }, + { + "epoch": 11.63738019169329, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 14570 + }, + { + "epoch": 11.638178913738018, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 14571 + }, + { + "epoch": 11.638977635782748, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 14572 + }, + { + "epoch": 11.639776357827476, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 14573 + }, + { + "epoch": 11.640575079872205, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 14574 + }, + { + "epoch": 11.641373801916933, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 14575 + }, + { + "epoch": 11.642172523961662, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 14576 + }, + { + "epoch": 11.64297124600639, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0157, + "step": 14577 + }, + { + "epoch": 11.643769968051117, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 14578 + }, + { + "epoch": 11.644568690095847, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 14579 + }, + { + "epoch": 11.645367412140574, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 14580 + }, + { + "epoch": 11.646166134185304, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 14581 + }, + { + "epoch": 11.646964856230031, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0181, + "step": 14582 + }, + { + "epoch": 11.64776357827476, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 14583 + }, + { + "epoch": 11.648562300319488, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 14584 + }, + { + "epoch": 11.649361022364218, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 14585 + }, + { + "epoch": 11.650159744408946, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 14586 + }, + { + "epoch": 11.650958466453675, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 14587 + }, + { + "epoch": 11.651757188498403, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 14588 + }, + { + "epoch": 11.65255591054313, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 14589 + }, + { + "epoch": 11.65335463258786, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 14590 + }, + { + "epoch": 11.654153354632587, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 14591 + }, + { + "epoch": 11.654952076677317, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 14592 + }, + { + "epoch": 11.655750798722044, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 14593 + }, + { + "epoch": 11.656549520766774, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 14594 + }, + { + "epoch": 11.657348242811501, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 14595 + }, + { + "epoch": 11.65814696485623, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 14596 + }, + { + "epoch": 11.658945686900958, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 14597 + }, + { + "epoch": 11.659744408945686, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 14598 + }, + { + "epoch": 11.660543130990416, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 14599 + }, + { + "epoch": 11.661341853035143, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.034, + "step": 14600 + }, + { + "epoch": 11.662140575079873, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 14601 + }, + { + "epoch": 11.6629392971246, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 14602 + }, + { + "epoch": 11.66373801916933, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 14603 + }, + { + "epoch": 11.664536741214057, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 14604 + }, + { + "epoch": 11.665335463258787, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 14605 + }, + { + "epoch": 11.666134185303514, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 14606 + }, + { + "epoch": 11.666932907348244, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0173, + "step": 14607 + }, + { + "epoch": 11.667731629392971, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 14608 + }, + { + "epoch": 11.668530351437699, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 14609 + }, + { + "epoch": 11.669329073482428, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 14610 + }, + { + "epoch": 11.670127795527156, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 14611 + }, + { + "epoch": 11.670926517571885, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 14612 + }, + { + "epoch": 11.671725239616613, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 14613 + }, + { + "epoch": 11.672523961661343, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.016, + "step": 14614 + }, + { + "epoch": 11.67332268370607, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 14615 + }, + { + "epoch": 11.6741214057508, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 14616 + }, + { + "epoch": 11.674920127795527, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 14617 + }, + { + "epoch": 11.675718849840255, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 14618 + }, + { + "epoch": 11.676517571884984, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 14619 + }, + { + "epoch": 11.677316293929712, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 14620 + }, + { + "epoch": 11.678115015974441, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 14621 + }, + { + "epoch": 11.678913738019169, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 14622 + }, + { + "epoch": 11.679712460063898, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 14623 + }, + { + "epoch": 11.680511182108626, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 14624 + }, + { + "epoch": 11.681309904153355, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0165, + "step": 14625 + }, + { + "epoch": 11.682108626198083, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 14626 + }, + { + "epoch": 11.68290734824281, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0362, + "step": 14627 + }, + { + "epoch": 11.68370607028754, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 14628 + }, + { + "epoch": 11.684504792332268, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 14629 + }, + { + "epoch": 11.685303514376997, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 14630 + }, + { + "epoch": 11.686102236421725, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 14631 + }, + { + "epoch": 11.686900958466454, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0143, + "step": 14632 + }, + { + "epoch": 11.687699680511182, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 14633 + }, + { + "epoch": 11.688498402555911, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 14634 + }, + { + "epoch": 11.689297124600639, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0165, + "step": 14635 + }, + { + "epoch": 11.690095846645367, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 14636 + }, + { + "epoch": 11.690894568690096, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 14637 + }, + { + "epoch": 11.691693290734824, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 14638 + }, + { + "epoch": 11.692492012779553, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0185, + "step": 14639 + }, + { + "epoch": 11.69329073482428, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0192, + "step": 14640 + }, + { + "epoch": 11.69408945686901, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0373, + "step": 14641 + }, + { + "epoch": 11.694888178913738, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 14642 + }, + { + "epoch": 11.695686900958467, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 14643 + }, + { + "epoch": 11.696485623003195, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 14644 + }, + { + "epoch": 11.697284345047922, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0169, + "step": 14645 + }, + { + "epoch": 11.698083067092652, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 14646 + }, + { + "epoch": 11.69888178913738, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.018, + "step": 14647 + }, + { + "epoch": 11.699680511182109, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 14648 + }, + { + "epoch": 11.700479233226837, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 14649 + }, + { + "epoch": 11.701277955271566, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0205, + "step": 14650 + }, + { + "epoch": 11.702076677316294, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 14651 + }, + { + "epoch": 11.702875399361023, + "grad_norm": 0.169921875, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 14652 + }, + { + "epoch": 11.70367412140575, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 14653 + }, + { + "epoch": 11.704472843450478, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 14654 + }, + { + "epoch": 11.705271565495208, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 14655 + }, + { + "epoch": 11.706070287539935, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 14656 + }, + { + "epoch": 11.706869009584665, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 14657 + }, + { + "epoch": 11.707667731629392, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 14658 + }, + { + "epoch": 11.708466453674122, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 14659 + }, + { + "epoch": 11.70926517571885, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0174, + "step": 14660 + }, + { + "epoch": 11.710063897763579, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 14661 + }, + { + "epoch": 11.710862619808307, + "grad_norm": 0.1787109375, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 14662 + }, + { + "epoch": 11.711661341853034, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 14663 + }, + { + "epoch": 11.712460063897764, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 14664 + }, + { + "epoch": 11.713258785942491, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 14665 + }, + { + "epoch": 11.71405750798722, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 14666 + }, + { + "epoch": 11.714856230031948, + "grad_norm": 0.16796875, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 14667 + }, + { + "epoch": 11.715654952076678, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0151, + "step": 14668 + }, + { + "epoch": 11.716453674121405, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 14669 + }, + { + "epoch": 11.717252396166135, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 14670 + }, + { + "epoch": 11.718051118210862, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 14671 + }, + { + "epoch": 11.718849840255592, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 14672 + }, + { + "epoch": 11.71964856230032, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 14673 + }, + { + "epoch": 11.720447284345047, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 14674 + }, + { + "epoch": 11.721246006389777, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 14675 + }, + { + "epoch": 11.722044728434504, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 14676 + }, + { + "epoch": 11.722843450479234, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 14677 + }, + { + "epoch": 11.723642172523961, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 14678 + }, + { + "epoch": 11.72444089456869, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 14679 + }, + { + "epoch": 11.725239616613418, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 14680 + }, + { + "epoch": 11.726038338658148, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 14681 + }, + { + "epoch": 11.726837060702875, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 14682 + }, + { + "epoch": 11.727635782747605, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 14683 + }, + { + "epoch": 11.728434504792332, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 14684 + }, + { + "epoch": 11.72923322683706, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 14685 + }, + { + "epoch": 11.73003194888179, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 14686 + }, + { + "epoch": 11.730830670926517, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 14687 + }, + { + "epoch": 11.731629392971247, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0163, + "step": 14688 + }, + { + "epoch": 11.732428115015974, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 14689 + }, + { + "epoch": 11.733226837060704, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 14690 + }, + { + "epoch": 11.734025559105431, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 14691 + }, + { + "epoch": 11.73482428115016, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 14692 + }, + { + "epoch": 11.735623003194888, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 14693 + }, + { + "epoch": 11.736421725239616, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 14694 + }, + { + "epoch": 11.737220447284345, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 14695 + }, + { + "epoch": 11.738019169329073, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0126, + "step": 14696 + }, + { + "epoch": 11.738817891373802, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 14697 + }, + { + "epoch": 11.73961661341853, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 14698 + }, + { + "epoch": 11.74041533546326, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 14699 + }, + { + "epoch": 11.741214057507987, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 14700 + }, + { + "epoch": 11.742012779552716, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 14701 + }, + { + "epoch": 11.742811501597444, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 14702 + }, + { + "epoch": 11.743610223642172, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 14703 + }, + { + "epoch": 11.744408945686901, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 14704 + }, + { + "epoch": 11.745207667731629, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 14705 + }, + { + "epoch": 11.746006389776358, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 14706 + }, + { + "epoch": 11.746805111821086, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 14707 + }, + { + "epoch": 11.747603833865815, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 14708 + }, + { + "epoch": 11.748402555910543, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 14709 + }, + { + "epoch": 11.749201277955272, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 14710 + }, + { + "epoch": 11.75, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 14711 + }, + { + "epoch": 11.750798722044728, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 14712 + }, + { + "epoch": 11.751597444089457, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 14713 + }, + { + "epoch": 11.752396166134185, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 14714 + }, + { + "epoch": 11.753194888178914, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 14715 + }, + { + "epoch": 11.753993610223642, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0167, + "step": 14716 + }, + { + "epoch": 11.754792332268371, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 14717 + }, + { + "epoch": 11.755591054313099, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 14718 + }, + { + "epoch": 11.756389776357828, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 14719 + }, + { + "epoch": 11.757188498402556, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 14720 + }, + { + "epoch": 11.757987220447284, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0156, + "step": 14721 + }, + { + "epoch": 11.758785942492013, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 14722 + }, + { + "epoch": 11.75958466453674, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 14723 + }, + { + "epoch": 11.76038338658147, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 14724 + }, + { + "epoch": 11.761182108626198, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 14725 + }, + { + "epoch": 11.761980830670927, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 14726 + }, + { + "epoch": 11.762779552715655, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 14727 + }, + { + "epoch": 11.763578274760384, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 14728 + }, + { + "epoch": 11.764376996805112, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 14729 + }, + { + "epoch": 11.76517571884984, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 14730 + }, + { + "epoch": 11.765974440894569, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 14731 + }, + { + "epoch": 11.766773162939296, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 14732 + }, + { + "epoch": 11.767571884984026, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 14733 + }, + { + "epoch": 11.768370607028753, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 14734 + }, + { + "epoch": 11.769169329073483, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 14735 + }, + { + "epoch": 11.76996805111821, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 14736 + }, + { + "epoch": 11.77076677316294, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 14737 + }, + { + "epoch": 11.771565495207668, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 14738 + }, + { + "epoch": 11.772364217252395, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 14739 + }, + { + "epoch": 11.773162939297125, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 14740 + }, + { + "epoch": 11.773961661341852, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 14741 + }, + { + "epoch": 11.774760383386582, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 14742 + }, + { + "epoch": 11.77555910543131, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 14743 + }, + { + "epoch": 11.776357827476039, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 14744 + }, + { + "epoch": 11.777156549520766, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 14745 + }, + { + "epoch": 11.777955271565496, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 14746 + }, + { + "epoch": 11.778753993610223, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 14747 + }, + { + "epoch": 11.779552715654953, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 14748 + }, + { + "epoch": 11.78035143769968, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 14749 + }, + { + "epoch": 11.781150159744408, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 14750 + }, + { + "epoch": 11.781948881789138, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 14751 + }, + { + "epoch": 11.782747603833865, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 14752 + }, + { + "epoch": 11.783546325878595, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 14753 + }, + { + "epoch": 11.784345047923322, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 14754 + }, + { + "epoch": 11.785143769968052, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0174, + "step": 14755 + }, + { + "epoch": 11.78594249201278, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 14756 + }, + { + "epoch": 11.786741214057509, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 14757 + }, + { + "epoch": 11.787539936102236, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 14758 + }, + { + "epoch": 11.788338658146966, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 14759 + }, + { + "epoch": 11.789137380191693, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 14760 + }, + { + "epoch": 11.789936102236421, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 14761 + }, + { + "epoch": 11.79073482428115, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 14762 + }, + { + "epoch": 11.791533546325878, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 14763 + }, + { + "epoch": 11.792332268370608, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 14764 + }, + { + "epoch": 11.793130990415335, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 14765 + }, + { + "epoch": 11.793929712460065, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0196, + "step": 14766 + }, + { + "epoch": 11.794728434504792, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 14767 + }, + { + "epoch": 11.795527156549522, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 14768 + }, + { + "epoch": 11.79632587859425, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 14769 + }, + { + "epoch": 11.797124600638977, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 14770 + }, + { + "epoch": 11.797923322683706, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 14771 + }, + { + "epoch": 11.798722044728434, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 14772 + }, + { + "epoch": 11.799520766773163, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 14773 + }, + { + "epoch": 11.800319488817891, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 14774 + }, + { + "epoch": 11.80111821086262, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 14775 + }, + { + "epoch": 11.801916932907348, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 14776 + }, + { + "epoch": 11.802715654952078, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0403, + "step": 14777 + }, + { + "epoch": 11.803514376996805, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 14778 + }, + { + "epoch": 11.804313099041533, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 14779 + }, + { + "epoch": 11.805111821086262, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 14780 + }, + { + "epoch": 11.80591054313099, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 14781 + }, + { + "epoch": 11.80670926517572, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 14782 + }, + { + "epoch": 11.807507987220447, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 14783 + }, + { + "epoch": 11.808306709265176, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 14784 + }, + { + "epoch": 11.809105431309904, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 14785 + }, + { + "epoch": 11.809904153354633, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0162, + "step": 14786 + }, + { + "epoch": 11.810702875399361, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 14787 + }, + { + "epoch": 11.811501597444089, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 14788 + }, + { + "epoch": 11.812300319488818, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 14789 + }, + { + "epoch": 11.813099041533546, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 14790 + }, + { + "epoch": 11.813897763578275, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 14791 + }, + { + "epoch": 11.814696485623003, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 14792 + }, + { + "epoch": 11.815495207667732, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 14793 + }, + { + "epoch": 11.81629392971246, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 14794 + }, + { + "epoch": 11.81709265175719, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0176, + "step": 14795 + }, + { + "epoch": 11.817891373801917, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 14796 + }, + { + "epoch": 11.818690095846645, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 14797 + }, + { + "epoch": 11.819488817891374, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 14798 + }, + { + "epoch": 11.820287539936102, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 14799 + }, + { + "epoch": 11.821086261980831, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0164, + "step": 14800 + }, + { + "epoch": 11.821884984025559, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 14801 + }, + { + "epoch": 11.822683706070288, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 14802 + }, + { + "epoch": 11.823482428115016, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 14803 + }, + { + "epoch": 11.824281150159745, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 14804 + }, + { + "epoch": 11.825079872204473, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 14805 + }, + { + "epoch": 11.8258785942492, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 14806 + }, + { + "epoch": 11.82667731629393, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 14807 + }, + { + "epoch": 11.827476038338657, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 14808 + }, + { + "epoch": 11.828274760383387, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 14809 + }, + { + "epoch": 11.829073482428115, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 14810 + }, + { + "epoch": 11.829872204472844, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 14811 + }, + { + "epoch": 11.830670926517572, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 14812 + }, + { + "epoch": 11.831469648562301, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 14813 + }, + { + "epoch": 11.832268370607029, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 14814 + }, + { + "epoch": 11.833067092651756, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 14815 + }, + { + "epoch": 11.833865814696486, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 14816 + }, + { + "epoch": 11.834664536741213, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 14817 + }, + { + "epoch": 11.835463258785943, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0177, + "step": 14818 + }, + { + "epoch": 11.83626198083067, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 14819 + }, + { + "epoch": 11.8370607028754, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 14820 + }, + { + "epoch": 11.837859424920127, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0356, + "step": 14821 + }, + { + "epoch": 11.838658146964857, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 14822 + }, + { + "epoch": 11.839456869009584, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0181, + "step": 14823 + }, + { + "epoch": 11.840255591054314, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 14824 + }, + { + "epoch": 11.841054313099042, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0146, + "step": 14825 + }, + { + "epoch": 11.84185303514377, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 14826 + }, + { + "epoch": 11.842651757188499, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 14827 + }, + { + "epoch": 11.843450479233226, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 14828 + }, + { + "epoch": 11.844249201277956, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 14829 + }, + { + "epoch": 11.845047923322683, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 14830 + }, + { + "epoch": 11.845846645367413, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 14831 + }, + { + "epoch": 11.84664536741214, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 14832 + }, + { + "epoch": 11.84744408945687, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 14833 + }, + { + "epoch": 11.848242811501597, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 14834 + }, + { + "epoch": 11.849041533546325, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 14835 + }, + { + "epoch": 11.849840255591054, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 14836 + }, + { + "epoch": 11.850638977635782, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 14837 + }, + { + "epoch": 11.851437699680512, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 14838 + }, + { + "epoch": 11.85223642172524, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 14839 + }, + { + "epoch": 11.853035143769969, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 14840 + }, + { + "epoch": 11.853833865814696, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 14841 + }, + { + "epoch": 11.854632587859426, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 14842 + }, + { + "epoch": 11.855431309904153, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 14843 + }, + { + "epoch": 11.856230031948883, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 14844 + }, + { + "epoch": 11.85702875399361, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 14845 + }, + { + "epoch": 11.857827476038338, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 14846 + }, + { + "epoch": 11.858626198083067, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 14847 + }, + { + "epoch": 11.859424920127795, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 14848 + }, + { + "epoch": 11.860223642172524, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 14849 + }, + { + "epoch": 11.861022364217252, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 14850 + }, + { + "epoch": 11.861821086261982, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 14851 + }, + { + "epoch": 11.86261980830671, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0188, + "step": 14852 + }, + { + "epoch": 11.863418530351439, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 14853 + }, + { + "epoch": 11.864217252396166, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 14854 + }, + { + "epoch": 11.865015974440894, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 14855 + }, + { + "epoch": 11.865814696485623, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 14856 + }, + { + "epoch": 11.866613418530351, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 14857 + }, + { + "epoch": 11.86741214057508, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 14858 + }, + { + "epoch": 11.868210862619808, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 14859 + }, + { + "epoch": 11.869009584664537, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 14860 + }, + { + "epoch": 11.869808306709265, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 14861 + }, + { + "epoch": 11.870607028753994, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 14862 + }, + { + "epoch": 11.871405750798722, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 14863 + }, + { + "epoch": 11.87220447284345, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 14864 + }, + { + "epoch": 11.87300319488818, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 14865 + }, + { + "epoch": 11.873801916932907, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0364, + "step": 14866 + }, + { + "epoch": 11.874600638977636, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 14867 + }, + { + "epoch": 11.875399361022364, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 14868 + }, + { + "epoch": 11.876198083067093, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 14869 + }, + { + "epoch": 11.87699680511182, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 14870 + }, + { + "epoch": 11.87779552715655, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 14871 + }, + { + "epoch": 11.878594249201278, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 14872 + }, + { + "epoch": 11.879392971246006, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 14873 + }, + { + "epoch": 11.880191693290735, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0156, + "step": 14874 + }, + { + "epoch": 11.880990415335463, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 14875 + }, + { + "epoch": 11.881789137380192, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0195, + "step": 14876 + }, + { + "epoch": 11.88258785942492, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 14877 + }, + { + "epoch": 11.883386581469649, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 14878 + }, + { + "epoch": 11.884185303514377, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 14879 + }, + { + "epoch": 11.884984025559106, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 14880 + }, + { + "epoch": 11.885782747603834, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 14881 + }, + { + "epoch": 11.886581469648561, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 14882 + }, + { + "epoch": 11.88738019169329, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 14883 + }, + { + "epoch": 11.888178913738018, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 14884 + }, + { + "epoch": 11.888977635782748, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 14885 + }, + { + "epoch": 11.889776357827476, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 14886 + }, + { + "epoch": 11.890575079872205, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 14887 + }, + { + "epoch": 11.891373801916933, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 14888 + }, + { + "epoch": 11.892172523961662, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 14889 + }, + { + "epoch": 11.89297124600639, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0168, + "step": 14890 + }, + { + "epoch": 11.893769968051117, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 14891 + }, + { + "epoch": 11.894568690095847, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 14892 + }, + { + "epoch": 11.895367412140574, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 14893 + }, + { + "epoch": 11.896166134185304, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 14894 + }, + { + "epoch": 11.896964856230031, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 14895 + }, + { + "epoch": 11.89776357827476, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 14896 + }, + { + "epoch": 11.898562300319488, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 14897 + }, + { + "epoch": 11.899361022364218, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 14898 + }, + { + "epoch": 11.900159744408946, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 14899 + }, + { + "epoch": 11.900958466453675, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 14900 + }, + { + "epoch": 11.901757188498403, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 14901 + }, + { + "epoch": 11.90255591054313, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 14902 + }, + { + "epoch": 11.90335463258786, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 14903 + }, + { + "epoch": 11.904153354632587, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 14904 + }, + { + "epoch": 11.904952076677317, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 14905 + }, + { + "epoch": 11.905750798722044, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 14906 + }, + { + "epoch": 11.906549520766774, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 14907 + }, + { + "epoch": 11.907348242811501, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 14908 + }, + { + "epoch": 11.90814696485623, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 14909 + }, + { + "epoch": 11.908945686900958, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0192, + "step": 14910 + }, + { + "epoch": 11.909744408945686, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 14911 + }, + { + "epoch": 11.910543130990416, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 14912 + }, + { + "epoch": 11.911341853035143, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 14913 + }, + { + "epoch": 11.912140575079873, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0205, + "step": 14914 + }, + { + "epoch": 11.9129392971246, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 14915 + }, + { + "epoch": 11.91373801916933, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0196, + "step": 14916 + }, + { + "epoch": 11.914536741214057, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 14917 + }, + { + "epoch": 11.915335463258787, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 14918 + }, + { + "epoch": 11.916134185303514, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 14919 + }, + { + "epoch": 11.916932907348244, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 14920 + }, + { + "epoch": 11.917731629392971, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 14921 + }, + { + "epoch": 11.918530351437699, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0185, + "step": 14922 + }, + { + "epoch": 11.919329073482428, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 14923 + }, + { + "epoch": 11.920127795527156, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 14924 + }, + { + "epoch": 11.920926517571885, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0166, + "step": 14925 + }, + { + "epoch": 11.921725239616613, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 14926 + }, + { + "epoch": 11.922523961661343, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 14927 + }, + { + "epoch": 11.92332268370607, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 14928 + }, + { + "epoch": 11.9241214057508, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 14929 + }, + { + "epoch": 11.924920127795527, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 14930 + }, + { + "epoch": 11.925718849840255, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 14931 + }, + { + "epoch": 11.926517571884984, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 14932 + }, + { + "epoch": 11.927316293929712, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 14933 + }, + { + "epoch": 11.928115015974441, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 14934 + }, + { + "epoch": 11.928913738019169, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 14935 + }, + { + "epoch": 11.929712460063898, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 14936 + }, + { + "epoch": 11.930511182108626, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 14937 + }, + { + "epoch": 11.931309904153355, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 14938 + }, + { + "epoch": 11.932108626198083, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 14939 + }, + { + "epoch": 11.93290734824281, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 14940 + }, + { + "epoch": 11.93370607028754, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 14941 + }, + { + "epoch": 11.934504792332268, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 14942 + }, + { + "epoch": 11.935303514376997, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 14943 + }, + { + "epoch": 11.936102236421725, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0183, + "step": 14944 + }, + { + "epoch": 11.936900958466454, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 14945 + }, + { + "epoch": 11.937699680511182, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 14946 + }, + { + "epoch": 11.938498402555911, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 14947 + }, + { + "epoch": 11.939297124600639, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 14948 + }, + { + "epoch": 11.940095846645367, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 14949 + }, + { + "epoch": 11.940894568690096, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 14950 + }, + { + "epoch": 11.941693290734824, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 14951 + }, + { + "epoch": 11.942492012779553, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 14952 + }, + { + "epoch": 11.94329073482428, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 14953 + }, + { + "epoch": 11.94408945686901, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 14954 + }, + { + "epoch": 11.944888178913738, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 14955 + }, + { + "epoch": 11.945686900958467, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 14956 + }, + { + "epoch": 11.946485623003195, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 14957 + }, + { + "epoch": 11.947284345047922, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0353, + "step": 14958 + }, + { + "epoch": 11.948083067092652, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 14959 + }, + { + "epoch": 11.94888178913738, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 14960 + }, + { + "epoch": 11.949680511182109, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 14961 + }, + { + "epoch": 11.950479233226837, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 14962 + }, + { + "epoch": 11.951277955271566, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 14963 + }, + { + "epoch": 11.952076677316294, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 14964 + }, + { + "epoch": 11.952875399361023, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 14965 + }, + { + "epoch": 11.95367412140575, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 14966 + }, + { + "epoch": 11.954472843450478, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 14967 + }, + { + "epoch": 11.955271565495208, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 14968 + }, + { + "epoch": 11.956070287539935, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 14969 + }, + { + "epoch": 11.956869009584665, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 14970 + }, + { + "epoch": 11.957667731629392, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 14971 + }, + { + "epoch": 11.958466453674122, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 14972 + }, + { + "epoch": 11.95926517571885, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 14973 + }, + { + "epoch": 11.960063897763579, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 14974 + }, + { + "epoch": 11.960862619808307, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 14975 + }, + { + "epoch": 11.961661341853034, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 14976 + }, + { + "epoch": 11.962460063897764, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 14977 + }, + { + "epoch": 11.963258785942491, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0151, + "step": 14978 + }, + { + "epoch": 11.96405750798722, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 14979 + }, + { + "epoch": 11.964856230031948, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 14980 + }, + { + "epoch": 11.965654952076678, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 14981 + }, + { + "epoch": 11.966453674121405, + "grad_norm": 0.2041015625, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 14982 + }, + { + "epoch": 11.967252396166135, + "grad_norm": 0.05615234375, + "learning_rate": 0.0005, + "loss": 1.0129, + "step": 14983 + }, + { + "epoch": 11.968051118210862, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 14984 + }, + { + "epoch": 11.968849840255592, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 14985 + }, + { + "epoch": 11.96964856230032, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 14986 + }, + { + "epoch": 11.970447284345047, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 14987 + }, + { + "epoch": 11.971246006389777, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0374, + "step": 14988 + }, + { + "epoch": 11.972044728434504, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 14989 + }, + { + "epoch": 11.972843450479234, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0182, + "step": 14990 + }, + { + "epoch": 11.973642172523961, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 14991 + }, + { + "epoch": 11.97444089456869, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 14992 + }, + { + "epoch": 11.975239616613418, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 14993 + }, + { + "epoch": 11.976038338658148, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 14994 + }, + { + "epoch": 11.976837060702875, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 14995 + }, + { + "epoch": 11.977635782747605, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 14996 + }, + { + "epoch": 11.978434504792332, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 14997 + }, + { + "epoch": 11.97923322683706, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 14998 + }, + { + "epoch": 11.98003194888179, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 14999 + }, + { + "epoch": 11.980830670926517, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 15000 + }, + { + "epoch": 11.981629392971247, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0149, + "step": 15001 + }, + { + "epoch": 11.982428115015974, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 15002 + }, + { + "epoch": 11.983226837060704, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 15003 + }, + { + "epoch": 11.984025559105431, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 15004 + }, + { + "epoch": 11.98482428115016, + "grad_norm": 0.1787109375, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 15005 + }, + { + "epoch": 11.985623003194888, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 15006 + }, + { + "epoch": 11.986421725239616, + "grad_norm": 0.1982421875, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 15007 + }, + { + "epoch": 11.987220447284345, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 15008 + }, + { + "epoch": 11.988019169329073, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 15009 + }, + { + "epoch": 11.988817891373802, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 15010 + }, + { + "epoch": 11.98961661341853, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 15011 + }, + { + "epoch": 11.99041533546326, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 15012 + }, + { + "epoch": 11.991214057507987, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 15013 + }, + { + "epoch": 11.992012779552716, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 15014 + }, + { + "epoch": 11.992811501597444, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 15015 + }, + { + "epoch": 11.993610223642172, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 15016 + }, + { + "epoch": 11.994408945686901, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 15017 + }, + { + "epoch": 11.995207667731629, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 15018 + }, + { + "epoch": 11.996006389776358, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 15019 + }, + { + "epoch": 11.996805111821086, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 15020 + }, + { + "epoch": 11.997603833865815, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0199, + "step": 15021 + }, + { + "epoch": 11.998402555910543, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 15022 + }, + { + "epoch": 11.999201277955272, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 15023 + }, + { + "epoch": 12.0, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0115, + "step": 15024 + }, + { + "epoch": 12.000798722044728, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0199, + "step": 15025 + }, + { + "epoch": 12.001597444089457, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 15026 + }, + { + "epoch": 12.002396166134185, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 15027 + }, + { + "epoch": 12.003194888178914, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 15028 + }, + { + "epoch": 12.003993610223642, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 15029 + }, + { + "epoch": 12.004792332268371, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 15030 + }, + { + "epoch": 12.005591054313099, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 15031 + }, + { + "epoch": 12.006389776357828, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 15032 + }, + { + "epoch": 12.007188498402556, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0135, + "step": 15033 + }, + { + "epoch": 12.007987220447284, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 15034 + }, + { + "epoch": 12.008785942492013, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 15035 + }, + { + "epoch": 12.00958466453674, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 15036 + }, + { + "epoch": 12.01038338658147, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 15037 + }, + { + "epoch": 12.011182108626198, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 15038 + }, + { + "epoch": 12.011980830670927, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0134, + "step": 15039 + }, + { + "epoch": 12.012779552715655, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 15040 + }, + { + "epoch": 12.013578274760384, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0353, + "step": 15041 + }, + { + "epoch": 12.014376996805112, + "grad_norm": 0.057373046875, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 15042 + }, + { + "epoch": 12.01517571884984, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 15043 + }, + { + "epoch": 12.015974440894569, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0168, + "step": 15044 + }, + { + "epoch": 12.016773162939296, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 15045 + }, + { + "epoch": 12.017571884984026, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 15046 + }, + { + "epoch": 12.018370607028753, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 15047 + }, + { + "epoch": 12.019169329073483, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 15048 + }, + { + "epoch": 12.01996805111821, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 15049 + }, + { + "epoch": 12.02076677316294, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 15050 + }, + { + "epoch": 12.021565495207668, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0205, + "step": 15051 + }, + { + "epoch": 12.022364217252397, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 15052 + }, + { + "epoch": 12.023162939297125, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 15053 + }, + { + "epoch": 12.023961661341852, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 15054 + }, + { + "epoch": 12.024760383386582, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0171, + "step": 15055 + }, + { + "epoch": 12.02555910543131, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 15056 + }, + { + "epoch": 12.026357827476039, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 15057 + }, + { + "epoch": 12.027156549520766, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 15058 + }, + { + "epoch": 12.027955271565496, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 15059 + }, + { + "epoch": 12.028753993610223, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 15060 + }, + { + "epoch": 12.029552715654953, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 15061 + }, + { + "epoch": 12.03035143769968, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 15062 + }, + { + "epoch": 12.031150159744408, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 15063 + }, + { + "epoch": 12.031948881789138, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0163, + "step": 15064 + }, + { + "epoch": 12.032747603833865, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 15065 + }, + { + "epoch": 12.033546325878595, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 15066 + }, + { + "epoch": 12.034345047923322, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0153, + "step": 15067 + }, + { + "epoch": 12.035143769968052, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 15068 + }, + { + "epoch": 12.03594249201278, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 15069 + }, + { + "epoch": 12.036741214057509, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 15070 + }, + { + "epoch": 12.037539936102236, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 15071 + }, + { + "epoch": 12.038338658146964, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 15072 + }, + { + "epoch": 12.039137380191693, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0178, + "step": 15073 + }, + { + "epoch": 12.039936102236421, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 15074 + }, + { + "epoch": 12.04073482428115, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 15075 + }, + { + "epoch": 12.041533546325878, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 15076 + }, + { + "epoch": 12.042332268370608, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 15077 + }, + { + "epoch": 12.043130990415335, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 15078 + }, + { + "epoch": 12.043929712460065, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0181, + "step": 15079 + }, + { + "epoch": 12.044728434504792, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0174, + "step": 15080 + }, + { + "epoch": 12.04552715654952, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 15081 + }, + { + "epoch": 12.04632587859425, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 15082 + }, + { + "epoch": 12.047124600638977, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 15083 + }, + { + "epoch": 12.047923322683706, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 15084 + }, + { + "epoch": 12.048722044728434, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0119, + "step": 15085 + }, + { + "epoch": 12.049520766773163, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 15086 + }, + { + "epoch": 12.050319488817891, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 15087 + }, + { + "epoch": 12.05111821086262, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 15088 + }, + { + "epoch": 12.051916932907348, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 15089 + }, + { + "epoch": 12.052715654952078, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 15090 + }, + { + "epoch": 12.053514376996805, + "grad_norm": 0.1630859375, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 15091 + }, + { + "epoch": 12.054313099041533, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 15092 + }, + { + "epoch": 12.055111821086262, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0187, + "step": 15093 + }, + { + "epoch": 12.05591054313099, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 15094 + }, + { + "epoch": 12.05670926517572, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 15095 + }, + { + "epoch": 12.057507987220447, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 15096 + }, + { + "epoch": 12.058306709265176, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 15097 + }, + { + "epoch": 12.059105431309904, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 15098 + }, + { + "epoch": 12.059904153354633, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 15099 + }, + { + "epoch": 12.060702875399361, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 15100 + }, + { + "epoch": 12.061501597444089, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0165, + "step": 15101 + }, + { + "epoch": 12.062300319488818, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 15102 + }, + { + "epoch": 12.063099041533546, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 15103 + }, + { + "epoch": 12.063897763578275, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 15104 + }, + { + "epoch": 12.064696485623003, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 15105 + }, + { + "epoch": 12.065495207667732, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 15106 + }, + { + "epoch": 12.06629392971246, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 15107 + }, + { + "epoch": 12.06709265175719, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 15108 + }, + { + "epoch": 12.067891373801917, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 15109 + }, + { + "epoch": 12.068690095846645, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 15110 + }, + { + "epoch": 12.069488817891374, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 15111 + }, + { + "epoch": 12.070287539936102, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 15112 + }, + { + "epoch": 12.071086261980831, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 15113 + }, + { + "epoch": 12.071884984025559, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 15114 + }, + { + "epoch": 12.072683706070288, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 15115 + }, + { + "epoch": 12.073482428115016, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 15116 + }, + { + "epoch": 12.074281150159745, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0199, + "step": 15117 + }, + { + "epoch": 12.075079872204473, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0158, + "step": 15118 + }, + { + "epoch": 12.0758785942492, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0167, + "step": 15119 + }, + { + "epoch": 12.07667731629393, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 15120 + }, + { + "epoch": 12.077476038338657, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 15121 + }, + { + "epoch": 12.078274760383387, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 15122 + }, + { + "epoch": 12.079073482428115, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0183, + "step": 15123 + }, + { + "epoch": 12.079872204472844, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 15124 + }, + { + "epoch": 12.080670926517572, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 15125 + }, + { + "epoch": 12.081469648562301, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 15126 + }, + { + "epoch": 12.082268370607029, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0181, + "step": 15127 + }, + { + "epoch": 12.083067092651758, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 15128 + }, + { + "epoch": 12.083865814696486, + "grad_norm": 0.1640625, + "learning_rate": 0.0005, + "loss": 1.0182, + "step": 15129 + }, + { + "epoch": 12.084664536741213, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 15130 + }, + { + "epoch": 12.085463258785943, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 15131 + }, + { + "epoch": 12.08626198083067, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 15132 + }, + { + "epoch": 12.0870607028754, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0157, + "step": 15133 + }, + { + "epoch": 12.087859424920127, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 15134 + }, + { + "epoch": 12.088658146964857, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 15135 + }, + { + "epoch": 12.089456869009584, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 15136 + }, + { + "epoch": 12.090255591054314, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 15137 + }, + { + "epoch": 12.091054313099042, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 15138 + }, + { + "epoch": 12.09185303514377, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 15139 + }, + { + "epoch": 12.092651757188499, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 15140 + }, + { + "epoch": 12.093450479233226, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 15141 + }, + { + "epoch": 12.094249201277956, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 15142 + }, + { + "epoch": 12.095047923322683, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 15143 + }, + { + "epoch": 12.095846645367413, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 15144 + }, + { + "epoch": 12.09664536741214, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 15145 + }, + { + "epoch": 12.09744408945687, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 15146 + }, + { + "epoch": 12.098242811501597, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 15147 + }, + { + "epoch": 12.099041533546325, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 15148 + }, + { + "epoch": 12.099840255591054, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0196, + "step": 15149 + }, + { + "epoch": 12.100638977635782, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 15150 + }, + { + "epoch": 12.101437699680512, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 15151 + }, + { + "epoch": 12.10223642172524, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 15152 + }, + { + "epoch": 12.103035143769969, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 15153 + }, + { + "epoch": 12.103833865814696, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 15154 + }, + { + "epoch": 12.104632587859426, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 15155 + }, + { + "epoch": 12.105431309904153, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 15156 + }, + { + "epoch": 12.106230031948881, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 15157 + }, + { + "epoch": 12.10702875399361, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 15158 + }, + { + "epoch": 12.107827476038338, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0164, + "step": 15159 + }, + { + "epoch": 12.108626198083067, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 15160 + }, + { + "epoch": 12.109424920127795, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 15161 + }, + { + "epoch": 12.110223642172524, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 15162 + }, + { + "epoch": 12.111022364217252, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 15163 + }, + { + "epoch": 12.111821086261982, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 15164 + }, + { + "epoch": 12.11261980830671, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 15165 + }, + { + "epoch": 12.113418530351439, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 15166 + }, + { + "epoch": 12.114217252396166, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0182, + "step": 15167 + }, + { + "epoch": 12.115015974440894, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 15168 + }, + { + "epoch": 12.115814696485623, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 15169 + }, + { + "epoch": 12.116613418530351, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0172, + "step": 15170 + }, + { + "epoch": 12.11741214057508, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 15171 + }, + { + "epoch": 12.118210862619808, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0177, + "step": 15172 + }, + { + "epoch": 12.119009584664537, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0138, + "step": 15173 + }, + { + "epoch": 12.119808306709265, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 15174 + }, + { + "epoch": 12.120607028753994, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 15175 + }, + { + "epoch": 12.121405750798722, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 15176 + }, + { + "epoch": 12.12220447284345, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 15177 + }, + { + "epoch": 12.12300319488818, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 15178 + }, + { + "epoch": 12.123801916932907, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 15179 + }, + { + "epoch": 12.124600638977636, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 15180 + }, + { + "epoch": 12.125399361022364, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 15181 + }, + { + "epoch": 12.126198083067093, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 15182 + }, + { + "epoch": 12.12699680511182, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 15183 + }, + { + "epoch": 12.12779552715655, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 15184 + }, + { + "epoch": 12.128594249201278, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 15185 + }, + { + "epoch": 12.129392971246006, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 15186 + }, + { + "epoch": 12.130191693290735, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 15187 + }, + { + "epoch": 12.130990415335463, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 15188 + }, + { + "epoch": 12.131789137380192, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 15189 + }, + { + "epoch": 12.13258785942492, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 15190 + }, + { + "epoch": 12.133386581469649, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 15191 + }, + { + "epoch": 12.134185303514377, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0337, + "step": 15192 + }, + { + "epoch": 12.134984025559106, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 15193 + }, + { + "epoch": 12.135782747603834, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 15194 + }, + { + "epoch": 12.136581469648561, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 15195 + }, + { + "epoch": 12.13738019169329, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 15196 + }, + { + "epoch": 12.138178913738018, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 15197 + }, + { + "epoch": 12.138977635782748, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.016, + "step": 15198 + }, + { + "epoch": 12.139776357827476, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 15199 + }, + { + "epoch": 12.140575079872205, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0143, + "step": 15200 + }, + { + "epoch": 12.141373801916933, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 15201 + }, + { + "epoch": 12.142172523961662, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 15202 + }, + { + "epoch": 12.14297124600639, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 15203 + }, + { + "epoch": 12.143769968051119, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 15204 + }, + { + "epoch": 12.144568690095847, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 15205 + }, + { + "epoch": 12.145367412140574, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 15206 + }, + { + "epoch": 12.146166134185304, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 15207 + }, + { + "epoch": 12.146964856230031, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 15208 + }, + { + "epoch": 12.14776357827476, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 15209 + }, + { + "epoch": 12.148562300319488, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0176, + "step": 15210 + }, + { + "epoch": 12.149361022364218, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0123, + "step": 15211 + }, + { + "epoch": 12.150159744408946, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 15212 + }, + { + "epoch": 12.150958466453675, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 15213 + }, + { + "epoch": 12.151757188498403, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 15214 + }, + { + "epoch": 12.15255591054313, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0187, + "step": 15215 + }, + { + "epoch": 12.15335463258786, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 15216 + }, + { + "epoch": 12.154153354632587, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 15217 + }, + { + "epoch": 12.154952076677317, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 15218 + }, + { + "epoch": 12.155750798722044, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 15219 + }, + { + "epoch": 12.156549520766774, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 15220 + }, + { + "epoch": 12.157348242811501, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 15221 + }, + { + "epoch": 12.15814696485623, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 15222 + }, + { + "epoch": 12.158945686900958, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0205, + "step": 15223 + }, + { + "epoch": 12.159744408945686, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 15224 + }, + { + "epoch": 12.160543130990416, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0192, + "step": 15225 + }, + { + "epoch": 12.161341853035143, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 15226 + }, + { + "epoch": 12.162140575079873, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0179, + "step": 15227 + }, + { + "epoch": 12.1629392971246, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 15228 + }, + { + "epoch": 12.16373801916933, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 15229 + }, + { + "epoch": 12.164536741214057, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 15230 + }, + { + "epoch": 12.165335463258787, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 15231 + }, + { + "epoch": 12.166134185303514, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0148, + "step": 15232 + }, + { + "epoch": 12.166932907348242, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 15233 + }, + { + "epoch": 12.167731629392971, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 15234 + }, + { + "epoch": 12.168530351437699, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 15235 + }, + { + "epoch": 12.169329073482428, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 15236 + }, + { + "epoch": 12.170127795527156, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 15237 + }, + { + "epoch": 12.170926517571885, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 15238 + }, + { + "epoch": 12.171725239616613, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0185, + "step": 15239 + }, + { + "epoch": 12.172523961661343, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 15240 + }, + { + "epoch": 12.17332268370607, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 15241 + }, + { + "epoch": 12.1741214057508, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 15242 + }, + { + "epoch": 12.174920127795527, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 15243 + }, + { + "epoch": 12.175718849840255, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 15244 + }, + { + "epoch": 12.176517571884984, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 15245 + }, + { + "epoch": 12.177316293929712, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 15246 + }, + { + "epoch": 12.178115015974441, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 15247 + }, + { + "epoch": 12.178913738019169, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0132, + "step": 15248 + }, + { + "epoch": 12.179712460063898, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 15249 + }, + { + "epoch": 12.180511182108626, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 15250 + }, + { + "epoch": 12.181309904153355, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 15251 + }, + { + "epoch": 12.182108626198083, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 15252 + }, + { + "epoch": 12.18290734824281, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 15253 + }, + { + "epoch": 12.18370607028754, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 15254 + }, + { + "epoch": 12.184504792332268, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0146, + "step": 15255 + }, + { + "epoch": 12.185303514376997, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 15256 + }, + { + "epoch": 12.186102236421725, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 15257 + }, + { + "epoch": 12.186900958466454, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 15258 + }, + { + "epoch": 12.187699680511182, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 15259 + }, + { + "epoch": 12.188498402555911, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0185, + "step": 15260 + }, + { + "epoch": 12.189297124600639, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.013, + "step": 15261 + }, + { + "epoch": 12.190095846645367, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 15262 + }, + { + "epoch": 12.190894568690096, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 15263 + }, + { + "epoch": 12.191693290734824, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 15264 + }, + { + "epoch": 12.192492012779553, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 15265 + }, + { + "epoch": 12.19329073482428, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 15266 + }, + { + "epoch": 12.19408945686901, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 15267 + }, + { + "epoch": 12.194888178913738, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 15268 + }, + { + "epoch": 12.195686900958467, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0166, + "step": 15269 + }, + { + "epoch": 12.196485623003195, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 15270 + }, + { + "epoch": 12.197284345047922, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0166, + "step": 15271 + }, + { + "epoch": 12.198083067092652, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0119, + "step": 15272 + }, + { + "epoch": 12.19888178913738, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 15273 + }, + { + "epoch": 12.199680511182109, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 15274 + }, + { + "epoch": 12.200479233226837, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.016, + "step": 15275 + }, + { + "epoch": 12.201277955271566, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 15276 + }, + { + "epoch": 12.202076677316294, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 15277 + }, + { + "epoch": 12.202875399361023, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 15278 + }, + { + "epoch": 12.20367412140575, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 15279 + }, + { + "epoch": 12.204472843450478, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 15280 + }, + { + "epoch": 12.205271565495208, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 15281 + }, + { + "epoch": 12.206070287539935, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 15282 + }, + { + "epoch": 12.206869009584665, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 15283 + }, + { + "epoch": 12.207667731629392, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 15284 + }, + { + "epoch": 12.208466453674122, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.016, + "step": 15285 + }, + { + "epoch": 12.20926517571885, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 15286 + }, + { + "epoch": 12.210063897763579, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 15287 + }, + { + "epoch": 12.210862619808307, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 15288 + }, + { + "epoch": 12.211661341853036, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0114, + "step": 15289 + }, + { + "epoch": 12.212460063897764, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 15290 + }, + { + "epoch": 12.213258785942491, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 15291 + }, + { + "epoch": 12.21405750798722, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 15292 + }, + { + "epoch": 12.214856230031948, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 15293 + }, + { + "epoch": 12.215654952076678, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 15294 + }, + { + "epoch": 12.216453674121405, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 15295 + }, + { + "epoch": 12.217252396166135, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 15296 + }, + { + "epoch": 12.218051118210862, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0134, + "step": 15297 + }, + { + "epoch": 12.218849840255592, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 15298 + }, + { + "epoch": 12.21964856230032, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0147, + "step": 15299 + }, + { + "epoch": 12.220447284345047, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 15300 + }, + { + "epoch": 12.221246006389777, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 15301 + }, + { + "epoch": 12.222044728434504, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 15302 + }, + { + "epoch": 12.222843450479234, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 15303 + }, + { + "epoch": 12.223642172523961, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 15304 + }, + { + "epoch": 12.22444089456869, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 15305 + }, + { + "epoch": 12.225239616613418, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 15306 + }, + { + "epoch": 12.226038338658148, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 15307 + }, + { + "epoch": 12.226837060702875, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 15308 + }, + { + "epoch": 12.227635782747603, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0171, + "step": 15309 + }, + { + "epoch": 12.228434504792332, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 15310 + }, + { + "epoch": 12.22923322683706, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 15311 + }, + { + "epoch": 12.23003194888179, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 15312 + }, + { + "epoch": 12.230830670926517, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 15313 + }, + { + "epoch": 12.231629392971247, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 15314 + }, + { + "epoch": 12.232428115015974, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0176, + "step": 15315 + }, + { + "epoch": 12.233226837060704, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0187, + "step": 15316 + }, + { + "epoch": 12.234025559105431, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 15317 + }, + { + "epoch": 12.23482428115016, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 15318 + }, + { + "epoch": 12.235623003194888, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0146, + "step": 15319 + }, + { + "epoch": 12.236421725239616, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 15320 + }, + { + "epoch": 12.237220447284345, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 15321 + }, + { + "epoch": 12.238019169329073, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 15322 + }, + { + "epoch": 12.238817891373802, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 15323 + }, + { + "epoch": 12.23961661341853, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 15324 + }, + { + "epoch": 12.24041533546326, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 15325 + }, + { + "epoch": 12.241214057507987, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0171, + "step": 15326 + }, + { + "epoch": 12.242012779552716, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 15327 + }, + { + "epoch": 12.242811501597444, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 15328 + }, + { + "epoch": 12.243610223642172, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 15329 + }, + { + "epoch": 12.244408945686901, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0129, + "step": 15330 + }, + { + "epoch": 12.245207667731629, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 15331 + }, + { + "epoch": 12.246006389776358, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 15332 + }, + { + "epoch": 12.246805111821086, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 15333 + }, + { + "epoch": 12.247603833865815, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 15334 + }, + { + "epoch": 12.248402555910543, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 15335 + }, + { + "epoch": 12.249201277955272, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 15336 + }, + { + "epoch": 12.25, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 15337 + }, + { + "epoch": 12.250798722044728, + "grad_norm": 0.1806640625, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 15338 + }, + { + "epoch": 12.251597444089457, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 15339 + }, + { + "epoch": 12.252396166134185, + "grad_norm": 0.1904296875, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 15340 + }, + { + "epoch": 12.253194888178914, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 15341 + }, + { + "epoch": 12.253993610223642, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0392, + "step": 15342 + }, + { + "epoch": 12.254792332268371, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 15343 + }, + { + "epoch": 12.255591054313099, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 15344 + }, + { + "epoch": 12.256389776357828, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 15345 + }, + { + "epoch": 12.257188498402556, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0349, + "step": 15346 + }, + { + "epoch": 12.257987220447284, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 15347 + }, + { + "epoch": 12.258785942492013, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 15348 + }, + { + "epoch": 12.25958466453674, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.018, + "step": 15349 + }, + { + "epoch": 12.26038338658147, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 15350 + }, + { + "epoch": 12.261182108626198, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 15351 + }, + { + "epoch": 12.261980830670927, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 15352 + }, + { + "epoch": 12.262779552715655, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0173, + "step": 15353 + }, + { + "epoch": 12.263578274760384, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.016, + "step": 15354 + }, + { + "epoch": 12.264376996805112, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 15355 + }, + { + "epoch": 12.26517571884984, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 15356 + }, + { + "epoch": 12.265974440894569, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 15357 + }, + { + "epoch": 12.266773162939296, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 15358 + }, + { + "epoch": 12.267571884984026, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 15359 + }, + { + "epoch": 12.268370607028753, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 15360 + }, + { + "epoch": 12.269169329073483, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 15361 + }, + { + "epoch": 12.26996805111821, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.016, + "step": 15362 + }, + { + "epoch": 12.27076677316294, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 15363 + }, + { + "epoch": 12.271565495207668, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.034, + "step": 15364 + }, + { + "epoch": 12.272364217252397, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 15365 + }, + { + "epoch": 12.273162939297125, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 15366 + }, + { + "epoch": 12.273961661341852, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 15367 + }, + { + "epoch": 12.274760383386582, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 15368 + }, + { + "epoch": 12.27555910543131, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 15369 + }, + { + "epoch": 12.276357827476039, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0165, + "step": 15370 + }, + { + "epoch": 12.277156549520766, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 15371 + }, + { + "epoch": 12.277955271565496, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 15372 + }, + { + "epoch": 12.278753993610223, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0181, + "step": 15373 + }, + { + "epoch": 12.279552715654953, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 15374 + }, + { + "epoch": 12.28035143769968, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 15375 + }, + { + "epoch": 12.281150159744408, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 15376 + }, + { + "epoch": 12.281948881789138, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 15377 + }, + { + "epoch": 12.282747603833865, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 15378 + }, + { + "epoch": 12.283546325878595, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 15379 + }, + { + "epoch": 12.284345047923322, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0152, + "step": 15380 + }, + { + "epoch": 12.285143769968052, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 15381 + }, + { + "epoch": 12.28594249201278, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0174, + "step": 15382 + }, + { + "epoch": 12.286741214057509, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 15383 + }, + { + "epoch": 12.287539936102236, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 15384 + }, + { + "epoch": 12.288338658146964, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 15385 + }, + { + "epoch": 12.289137380191693, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 15386 + }, + { + "epoch": 12.289936102236421, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 15387 + }, + { + "epoch": 12.29073482428115, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.016, + "step": 15388 + }, + { + "epoch": 12.291533546325878, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 15389 + }, + { + "epoch": 12.292332268370608, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 15390 + }, + { + "epoch": 12.293130990415335, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 15391 + }, + { + "epoch": 12.293929712460065, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 15392 + }, + { + "epoch": 12.294728434504792, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 15393 + }, + { + "epoch": 12.295527156549522, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 15394 + }, + { + "epoch": 12.29632587859425, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 15395 + }, + { + "epoch": 12.297124600638977, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0188, + "step": 15396 + }, + { + "epoch": 12.297923322683706, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 15397 + }, + { + "epoch": 12.298722044728434, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0368, + "step": 15398 + }, + { + "epoch": 12.299520766773163, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 15399 + }, + { + "epoch": 12.300319488817891, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 15400 + }, + { + "epoch": 12.30111821086262, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 15401 + }, + { + "epoch": 12.301916932907348, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 15402 + }, + { + "epoch": 12.302715654952078, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 15403 + }, + { + "epoch": 12.303514376996805, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 15404 + }, + { + "epoch": 12.304313099041533, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 15405 + }, + { + "epoch": 12.305111821086262, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0152, + "step": 15406 + }, + { + "epoch": 12.30591054313099, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 15407 + }, + { + "epoch": 12.30670926517572, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0103, + "step": 15408 + }, + { + "epoch": 12.307507987220447, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 15409 + }, + { + "epoch": 12.308306709265176, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 15410 + }, + { + "epoch": 12.309105431309904, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 15411 + }, + { + "epoch": 12.309904153354633, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 15412 + }, + { + "epoch": 12.310702875399361, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 15413 + }, + { + "epoch": 12.311501597444089, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 15414 + }, + { + "epoch": 12.312300319488818, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 15415 + }, + { + "epoch": 12.313099041533546, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 15416 + }, + { + "epoch": 12.313897763578275, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 15417 + }, + { + "epoch": 12.314696485623003, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 15418 + }, + { + "epoch": 12.315495207667732, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 15419 + }, + { + "epoch": 12.31629392971246, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 15420 + }, + { + "epoch": 12.31709265175719, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 15421 + }, + { + "epoch": 12.317891373801917, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 15422 + }, + { + "epoch": 12.318690095846645, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 15423 + }, + { + "epoch": 12.319488817891374, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 15424 + }, + { + "epoch": 12.320287539936102, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0167, + "step": 15425 + }, + { + "epoch": 12.321086261980831, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 15426 + }, + { + "epoch": 12.321884984025559, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 15427 + }, + { + "epoch": 12.322683706070288, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 15428 + }, + { + "epoch": 12.323482428115016, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.018, + "step": 15429 + }, + { + "epoch": 12.324281150159745, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 15430 + }, + { + "epoch": 12.325079872204473, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0185, + "step": 15431 + }, + { + "epoch": 12.3258785942492, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 15432 + }, + { + "epoch": 12.32667731629393, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0172, + "step": 15433 + }, + { + "epoch": 12.327476038338657, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 15434 + }, + { + "epoch": 12.328274760383387, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 15435 + }, + { + "epoch": 12.329073482428115, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 15436 + }, + { + "epoch": 12.329872204472844, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0179, + "step": 15437 + }, + { + "epoch": 12.330670926517572, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 15438 + }, + { + "epoch": 12.331469648562301, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 15439 + }, + { + "epoch": 12.332268370607029, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 15440 + }, + { + "epoch": 12.333067092651758, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 15441 + }, + { + "epoch": 12.333865814696486, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 15442 + }, + { + "epoch": 12.334664536741213, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 15443 + }, + { + "epoch": 12.335463258785943, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 15444 + }, + { + "epoch": 12.33626198083067, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0185, + "step": 15445 + }, + { + "epoch": 12.3370607028754, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 15446 + }, + { + "epoch": 12.337859424920127, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 15447 + }, + { + "epoch": 12.338658146964857, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 15448 + }, + { + "epoch": 12.339456869009584, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 15449 + }, + { + "epoch": 12.340255591054314, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 15450 + }, + { + "epoch": 12.341054313099042, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 15451 + }, + { + "epoch": 12.34185303514377, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.013, + "step": 15452 + }, + { + "epoch": 12.342651757188499, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 15453 + }, + { + "epoch": 12.343450479233226, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 15454 + }, + { + "epoch": 12.344249201277956, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0187, + "step": 15455 + }, + { + "epoch": 12.345047923322683, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 15456 + }, + { + "epoch": 12.345846645367413, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 15457 + }, + { + "epoch": 12.34664536741214, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 15458 + }, + { + "epoch": 12.34744408945687, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 15459 + }, + { + "epoch": 12.348242811501597, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 15460 + }, + { + "epoch": 12.349041533546325, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 15461 + }, + { + "epoch": 12.349840255591054, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 15462 + }, + { + "epoch": 12.350638977635782, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0117, + "step": 15463 + }, + { + "epoch": 12.351437699680512, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 15464 + }, + { + "epoch": 12.35223642172524, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.039, + "step": 15465 + }, + { + "epoch": 12.353035143769969, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 15466 + }, + { + "epoch": 12.353833865814696, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 15467 + }, + { + "epoch": 12.354632587859426, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 15468 + }, + { + "epoch": 12.355431309904153, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 15469 + }, + { + "epoch": 12.356230031948883, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 15470 + }, + { + "epoch": 12.35702875399361, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 15471 + }, + { + "epoch": 12.357827476038338, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 15472 + }, + { + "epoch": 12.358626198083067, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0144, + "step": 15473 + }, + { + "epoch": 12.359424920127795, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 15474 + }, + { + "epoch": 12.360223642172524, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 15475 + }, + { + "epoch": 12.361022364217252, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 15476 + }, + { + "epoch": 12.361821086261982, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 15477 + }, + { + "epoch": 12.36261980830671, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 15478 + }, + { + "epoch": 12.363418530351439, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 15479 + }, + { + "epoch": 12.364217252396166, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 15480 + }, + { + "epoch": 12.365015974440894, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 15481 + }, + { + "epoch": 12.365814696485623, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 15482 + }, + { + "epoch": 12.366613418530351, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 15483 + }, + { + "epoch": 12.36741214057508, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 15484 + }, + { + "epoch": 12.368210862619808, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0126, + "step": 15485 + }, + { + "epoch": 12.369009584664537, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 15486 + }, + { + "epoch": 12.369808306709265, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0196, + "step": 15487 + }, + { + "epoch": 12.370607028753994, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 15488 + }, + { + "epoch": 12.371405750798722, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 15489 + }, + { + "epoch": 12.37220447284345, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.016, + "step": 15490 + }, + { + "epoch": 12.37300319488818, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 15491 + }, + { + "epoch": 12.373801916932907, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 15492 + }, + { + "epoch": 12.374600638977636, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0076, + "step": 15493 + }, + { + "epoch": 12.375399361022364, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 15494 + }, + { + "epoch": 12.376198083067093, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 15495 + }, + { + "epoch": 12.37699680511182, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 15496 + }, + { + "epoch": 12.37779552715655, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 15497 + }, + { + "epoch": 12.378594249201278, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 15498 + }, + { + "epoch": 12.379392971246006, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0174, + "step": 15499 + }, + { + "epoch": 12.380191693290735, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 15500 + }, + { + "epoch": 12.380990415335463, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0388, + "step": 15501 + }, + { + "epoch": 12.381789137380192, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 15502 + }, + { + "epoch": 12.38258785942492, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 15503 + }, + { + "epoch": 12.383386581469649, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 15504 + }, + { + "epoch": 12.384185303514377, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 15505 + }, + { + "epoch": 12.384984025559106, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0176, + "step": 15506 + }, + { + "epoch": 12.385782747603834, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 15507 + }, + { + "epoch": 12.386581469648561, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 15508 + }, + { + "epoch": 12.38738019169329, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 15509 + }, + { + "epoch": 12.388178913738018, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 15510 + }, + { + "epoch": 12.388977635782748, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 15511 + }, + { + "epoch": 12.389776357827476, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 15512 + }, + { + "epoch": 12.390575079872205, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 15513 + }, + { + "epoch": 12.391373801916933, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 15514 + }, + { + "epoch": 12.392172523961662, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 15515 + }, + { + "epoch": 12.39297124600639, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 15516 + }, + { + "epoch": 12.393769968051117, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 15517 + }, + { + "epoch": 12.394568690095847, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 15518 + }, + { + "epoch": 12.395367412140574, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 15519 + }, + { + "epoch": 12.396166134185304, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 15520 + }, + { + "epoch": 12.396964856230031, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 15521 + }, + { + "epoch": 12.39776357827476, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0169, + "step": 15522 + }, + { + "epoch": 12.398562300319488, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 15523 + }, + { + "epoch": 12.399361022364218, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0338, + "step": 15524 + }, + { + "epoch": 12.400159744408946, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 15525 + }, + { + "epoch": 12.400958466453675, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 15526 + }, + { + "epoch": 12.401757188498403, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 15527 + }, + { + "epoch": 12.40255591054313, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 15528 + }, + { + "epoch": 12.40335463258786, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 15529 + }, + { + "epoch": 12.404153354632587, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 15530 + }, + { + "epoch": 12.404952076677317, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 15531 + }, + { + "epoch": 12.405750798722044, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 15532 + }, + { + "epoch": 12.406549520766774, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 15533 + }, + { + "epoch": 12.407348242811501, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 15534 + }, + { + "epoch": 12.40814696485623, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 15535 + }, + { + "epoch": 12.408945686900958, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 15536 + }, + { + "epoch": 12.409744408945686, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 15537 + }, + { + "epoch": 12.410543130990416, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.038, + "step": 15538 + }, + { + "epoch": 12.411341853035143, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 15539 + }, + { + "epoch": 12.412140575079873, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0162, + "step": 15540 + }, + { + "epoch": 12.4129392971246, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 15541 + }, + { + "epoch": 12.41373801916933, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 15542 + }, + { + "epoch": 12.414536741214057, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0199, + "step": 15543 + }, + { + "epoch": 12.415335463258787, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 15544 + }, + { + "epoch": 12.416134185303514, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 15545 + }, + { + "epoch": 12.416932907348242, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 15546 + }, + { + "epoch": 12.417731629392971, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 15547 + }, + { + "epoch": 12.418530351437699, + "grad_norm": 0.1806640625, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 15548 + }, + { + "epoch": 12.419329073482428, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 15549 + }, + { + "epoch": 12.420127795527156, + "grad_norm": 0.255859375, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 15550 + }, + { + "epoch": 12.420926517571885, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 15551 + }, + { + "epoch": 12.421725239616613, + "grad_norm": 0.27734375, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 15552 + }, + { + "epoch": 12.422523961661343, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 15553 + }, + { + "epoch": 12.42332268370607, + "grad_norm": 0.1982421875, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 15554 + }, + { + "epoch": 12.4241214057508, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 15555 + }, + { + "epoch": 12.424920127795527, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 15556 + }, + { + "epoch": 12.425718849840255, + "grad_norm": 0.1572265625, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 15557 + }, + { + "epoch": 12.426517571884984, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0187, + "step": 15558 + }, + { + "epoch": 12.427316293929712, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 15559 + }, + { + "epoch": 12.428115015974441, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 15560 + }, + { + "epoch": 12.428913738019169, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 15561 + }, + { + "epoch": 12.429712460063898, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0149, + "step": 15562 + }, + { + "epoch": 12.430511182108626, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 15563 + }, + { + "epoch": 12.431309904153355, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 15564 + }, + { + "epoch": 12.432108626198083, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 15565 + }, + { + "epoch": 12.43290734824281, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 15566 + }, + { + "epoch": 12.43370607028754, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 15567 + }, + { + "epoch": 12.434504792332268, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0181, + "step": 15568 + }, + { + "epoch": 12.435303514376997, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 15569 + }, + { + "epoch": 12.436102236421725, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 15570 + }, + { + "epoch": 12.436900958466454, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0199, + "step": 15571 + }, + { + "epoch": 12.437699680511182, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0176, + "step": 15572 + }, + { + "epoch": 12.438498402555911, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 15573 + }, + { + "epoch": 12.439297124600639, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 15574 + }, + { + "epoch": 12.440095846645367, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 15575 + }, + { + "epoch": 12.440894568690096, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 15576 + }, + { + "epoch": 12.441693290734824, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 15577 + }, + { + "epoch": 12.442492012779553, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0187, + "step": 15578 + }, + { + "epoch": 12.44329073482428, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 15579 + }, + { + "epoch": 12.44408945686901, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 15580 + }, + { + "epoch": 12.444888178913738, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 15581 + }, + { + "epoch": 12.445686900958467, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 15582 + }, + { + "epoch": 12.446485623003195, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 15583 + }, + { + "epoch": 12.447284345047922, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0187, + "step": 15584 + }, + { + "epoch": 12.448083067092652, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 15585 + }, + { + "epoch": 12.44888178913738, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 15586 + }, + { + "epoch": 12.449680511182109, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 15587 + }, + { + "epoch": 12.450479233226837, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 15588 + }, + { + "epoch": 12.451277955271566, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 15589 + }, + { + "epoch": 12.452076677316294, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 15590 + }, + { + "epoch": 12.452875399361023, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 15591 + }, + { + "epoch": 12.45367412140575, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 15592 + }, + { + "epoch": 12.454472843450478, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 15593 + }, + { + "epoch": 12.455271565495208, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 15594 + }, + { + "epoch": 12.456070287539935, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 15595 + }, + { + "epoch": 12.456869009584665, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0196, + "step": 15596 + }, + { + "epoch": 12.457667731629392, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0161, + "step": 15597 + }, + { + "epoch": 12.458466453674122, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0138, + "step": 15598 + }, + { + "epoch": 12.45926517571885, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0126, + "step": 15599 + }, + { + "epoch": 12.460063897763579, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 15600 + }, + { + "epoch": 12.460862619808307, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 15601 + }, + { + "epoch": 12.461661341853036, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 15602 + }, + { + "epoch": 12.462460063897764, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 15603 + }, + { + "epoch": 12.463258785942491, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 15604 + }, + { + "epoch": 12.46405750798722, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 15605 + }, + { + "epoch": 12.464856230031948, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 15606 + }, + { + "epoch": 12.465654952076678, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0176, + "step": 15607 + }, + { + "epoch": 12.466453674121405, + "grad_norm": 0.0546875, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 15608 + }, + { + "epoch": 12.467252396166135, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 15609 + }, + { + "epoch": 12.468051118210862, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 15610 + }, + { + "epoch": 12.468849840255592, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 15611 + }, + { + "epoch": 12.46964856230032, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 15612 + }, + { + "epoch": 12.470447284345047, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 15613 + }, + { + "epoch": 12.471246006389777, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 15614 + }, + { + "epoch": 12.472044728434504, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 15615 + }, + { + "epoch": 12.472843450479234, + "grad_norm": 0.16796875, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 15616 + }, + { + "epoch": 12.473642172523961, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 15617 + }, + { + "epoch": 12.47444089456869, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 15618 + }, + { + "epoch": 12.475239616613418, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 15619 + }, + { + "epoch": 12.476038338658148, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 15620 + }, + { + "epoch": 12.476837060702875, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 15621 + }, + { + "epoch": 12.477635782747603, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 15622 + }, + { + "epoch": 12.478434504792332, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 15623 + }, + { + "epoch": 12.47923322683706, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0171, + "step": 15624 + }, + { + "epoch": 12.48003194888179, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 15625 + }, + { + "epoch": 12.480830670926517, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0149, + "step": 15626 + }, + { + "epoch": 12.481629392971247, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 15627 + }, + { + "epoch": 12.482428115015974, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 15628 + }, + { + "epoch": 12.483226837060704, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 15629 + }, + { + "epoch": 12.484025559105431, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 15630 + }, + { + "epoch": 12.48482428115016, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 15631 + }, + { + "epoch": 12.485623003194888, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 15632 + }, + { + "epoch": 12.486421725239616, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 15633 + }, + { + "epoch": 12.487220447284345, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0161, + "step": 15634 + }, + { + "epoch": 12.488019169329073, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 15635 + }, + { + "epoch": 12.488817891373802, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 15636 + }, + { + "epoch": 12.48961661341853, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 15637 + }, + { + "epoch": 12.49041533546326, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 15638 + }, + { + "epoch": 12.491214057507987, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 15639 + }, + { + "epoch": 12.492012779552716, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 15640 + }, + { + "epoch": 12.492811501597444, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 15641 + }, + { + "epoch": 12.493610223642172, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 15642 + }, + { + "epoch": 12.494408945686901, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 15643 + }, + { + "epoch": 12.495207667731629, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 15644 + }, + { + "epoch": 12.496006389776358, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0105, + "step": 15645 + }, + { + "epoch": 12.496805111821086, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 15646 + }, + { + "epoch": 12.497603833865815, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0177, + "step": 15647 + }, + { + "epoch": 12.498402555910543, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 15648 + }, + { + "epoch": 12.499201277955272, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 15649 + }, + { + "epoch": 12.5, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 15650 + }, + { + "epoch": 12.500798722044728, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 15651 + }, + { + "epoch": 12.501597444089457, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 15652 + }, + { + "epoch": 12.502396166134185, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 15653 + }, + { + "epoch": 12.503194888178914, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0188, + "step": 15654 + }, + { + "epoch": 12.503993610223642, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 15655 + }, + { + "epoch": 12.504792332268371, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 15656 + }, + { + "epoch": 12.505591054313099, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 15657 + }, + { + "epoch": 12.506389776357828, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 15658 + }, + { + "epoch": 12.507188498402556, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 15659 + }, + { + "epoch": 12.507987220447284, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 15660 + }, + { + "epoch": 12.508785942492013, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 15661 + }, + { + "epoch": 12.50958466453674, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0196, + "step": 15662 + }, + { + "epoch": 12.51038338658147, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 15663 + }, + { + "epoch": 12.511182108626198, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 15664 + }, + { + "epoch": 12.511980830670927, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0132, + "step": 15665 + }, + { + "epoch": 12.512779552715655, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 15666 + }, + { + "epoch": 12.513578274760384, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0172, + "step": 15667 + }, + { + "epoch": 12.514376996805112, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 15668 + }, + { + "epoch": 12.51517571884984, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 15669 + }, + { + "epoch": 12.515974440894569, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 15670 + }, + { + "epoch": 12.516773162939296, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 15671 + }, + { + "epoch": 12.517571884984026, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 15672 + }, + { + "epoch": 12.518370607028753, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 15673 + }, + { + "epoch": 12.519169329073483, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0173, + "step": 15674 + }, + { + "epoch": 12.51996805111821, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 15675 + }, + { + "epoch": 12.52076677316294, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 15676 + }, + { + "epoch": 12.521565495207668, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 15677 + }, + { + "epoch": 12.522364217252395, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 15678 + }, + { + "epoch": 12.523162939297125, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 15679 + }, + { + "epoch": 12.523961661341852, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 15680 + }, + { + "epoch": 12.524760383386582, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 15681 + }, + { + "epoch": 12.52555910543131, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 15682 + }, + { + "epoch": 12.526357827476039, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0147, + "step": 15683 + }, + { + "epoch": 12.527156549520766, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0179, + "step": 15684 + }, + { + "epoch": 12.527955271565496, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 15685 + }, + { + "epoch": 12.528753993610223, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 15686 + }, + { + "epoch": 12.529552715654953, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 15687 + }, + { + "epoch": 12.53035143769968, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 15688 + }, + { + "epoch": 12.531150159744408, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 15689 + }, + { + "epoch": 12.531948881789138, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 15690 + }, + { + "epoch": 12.532747603833865, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0187, + "step": 15691 + }, + { + "epoch": 12.533546325878595, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 15692 + }, + { + "epoch": 12.534345047923322, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 15693 + }, + { + "epoch": 12.535143769968052, + "grad_norm": 0.18359375, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 15694 + }, + { + "epoch": 12.53594249201278, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 15695 + }, + { + "epoch": 12.536741214057509, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 15696 + }, + { + "epoch": 12.537539936102236, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 15697 + }, + { + "epoch": 12.538338658146966, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.015, + "step": 15698 + }, + { + "epoch": 12.539137380191693, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0177, + "step": 15699 + }, + { + "epoch": 12.539936102236421, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 15700 + }, + { + "epoch": 12.54073482428115, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 15701 + }, + { + "epoch": 12.541533546325878, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 15702 + }, + { + "epoch": 12.542332268370608, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 15703 + }, + { + "epoch": 12.543130990415335, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 15704 + }, + { + "epoch": 12.543929712460065, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 15705 + }, + { + "epoch": 12.544728434504792, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 15706 + }, + { + "epoch": 12.545527156549522, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 15707 + }, + { + "epoch": 12.54632587859425, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 15708 + }, + { + "epoch": 12.547124600638977, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 15709 + }, + { + "epoch": 12.547923322683706, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 15710 + }, + { + "epoch": 12.548722044728434, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 15711 + }, + { + "epoch": 12.549520766773163, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 15712 + }, + { + "epoch": 12.550319488817891, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.018, + "step": 15713 + }, + { + "epoch": 12.55111821086262, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 15714 + }, + { + "epoch": 12.551916932907348, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 15715 + }, + { + "epoch": 12.552715654952078, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0174, + "step": 15716 + }, + { + "epoch": 12.553514376996805, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0187, + "step": 15717 + }, + { + "epoch": 12.554313099041533, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0199, + "step": 15718 + }, + { + "epoch": 12.555111821086262, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 15719 + }, + { + "epoch": 12.55591054313099, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 15720 + }, + { + "epoch": 12.55670926517572, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 15721 + }, + { + "epoch": 12.557507987220447, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0148, + "step": 15722 + }, + { + "epoch": 12.558306709265176, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0126, + "step": 15723 + }, + { + "epoch": 12.559105431309904, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 15724 + }, + { + "epoch": 12.559904153354633, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 15725 + }, + { + "epoch": 12.560702875399361, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 15726 + }, + { + "epoch": 12.561501597444089, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 15727 + }, + { + "epoch": 12.562300319488818, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 15728 + }, + { + "epoch": 12.563099041533546, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 15729 + }, + { + "epoch": 12.563897763578275, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 15730 + }, + { + "epoch": 12.564696485623003, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0159, + "step": 15731 + }, + { + "epoch": 12.565495207667732, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 15732 + }, + { + "epoch": 12.56629392971246, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0179, + "step": 15733 + }, + { + "epoch": 12.56709265175719, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 15734 + }, + { + "epoch": 12.567891373801917, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 15735 + }, + { + "epoch": 12.568690095846645, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 15736 + }, + { + "epoch": 12.569488817891374, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 15737 + }, + { + "epoch": 12.570287539936102, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 15738 + }, + { + "epoch": 12.571086261980831, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 15739 + }, + { + "epoch": 12.571884984025559, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 15740 + }, + { + "epoch": 12.572683706070288, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 15741 + }, + { + "epoch": 12.573482428115016, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 15742 + }, + { + "epoch": 12.574281150159745, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0333, + "step": 15743 + }, + { + "epoch": 12.575079872204473, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 15744 + }, + { + "epoch": 12.5758785942492, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 15745 + }, + { + "epoch": 12.57667731629393, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 15746 + }, + { + "epoch": 12.577476038338657, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 15747 + }, + { + "epoch": 12.578274760383387, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 15748 + }, + { + "epoch": 12.579073482428115, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 15749 + }, + { + "epoch": 12.579872204472844, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 15750 + }, + { + "epoch": 12.580670926517572, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.018, + "step": 15751 + }, + { + "epoch": 12.581469648562301, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0185, + "step": 15752 + }, + { + "epoch": 12.582268370607029, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 15753 + }, + { + "epoch": 12.583067092651756, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 15754 + }, + { + "epoch": 12.583865814696486, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 15755 + }, + { + "epoch": 12.584664536741213, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 15756 + }, + { + "epoch": 12.585463258785943, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 15757 + }, + { + "epoch": 12.58626198083067, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 15758 + }, + { + "epoch": 12.5870607028754, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.036, + "step": 15759 + }, + { + "epoch": 12.587859424920127, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0155, + "step": 15760 + }, + { + "epoch": 12.588658146964857, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 15761 + }, + { + "epoch": 12.589456869009584, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 15762 + }, + { + "epoch": 12.590255591054314, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 15763 + }, + { + "epoch": 12.591054313099042, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0158, + "step": 15764 + }, + { + "epoch": 12.59185303514377, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 15765 + }, + { + "epoch": 12.592651757188499, + "grad_norm": 0.1611328125, + "learning_rate": 0.0005, + "loss": 1.0158, + "step": 15766 + }, + { + "epoch": 12.593450479233226, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 15767 + }, + { + "epoch": 12.594249201277956, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 15768 + }, + { + "epoch": 12.595047923322683, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 15769 + }, + { + "epoch": 12.595846645367413, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 15770 + }, + { + "epoch": 12.59664536741214, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 15771 + }, + { + "epoch": 12.59744408945687, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 15772 + }, + { + "epoch": 12.598242811501597, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.017, + "step": 15773 + }, + { + "epoch": 12.599041533546325, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0199, + "step": 15774 + }, + { + "epoch": 12.599840255591054, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 15775 + }, + { + "epoch": 12.600638977635782, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 15776 + }, + { + "epoch": 12.601437699680512, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 15777 + }, + { + "epoch": 12.60223642172524, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 15778 + }, + { + "epoch": 12.603035143769969, + "grad_norm": 0.1630859375, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 15779 + }, + { + "epoch": 12.603833865814696, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 15780 + }, + { + "epoch": 12.604632587859426, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 15781 + }, + { + "epoch": 12.605431309904153, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 15782 + }, + { + "epoch": 12.606230031948883, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 15783 + }, + { + "epoch": 12.60702875399361, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 15784 + }, + { + "epoch": 12.607827476038338, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 15785 + }, + { + "epoch": 12.608626198083067, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 15786 + }, + { + "epoch": 12.609424920127795, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 15787 + }, + { + "epoch": 12.610223642172524, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0371, + "step": 15788 + }, + { + "epoch": 12.611022364217252, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 15789 + }, + { + "epoch": 12.611821086261982, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 15790 + }, + { + "epoch": 12.61261980830671, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 15791 + }, + { + "epoch": 12.613418530351439, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 15792 + }, + { + "epoch": 12.614217252396166, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 15793 + }, + { + "epoch": 12.615015974440894, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 15794 + }, + { + "epoch": 12.615814696485623, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 15795 + }, + { + "epoch": 12.616613418530351, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 15796 + }, + { + "epoch": 12.61741214057508, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 15797 + }, + { + "epoch": 12.618210862619808, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 15798 + }, + { + "epoch": 12.619009584664537, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 15799 + }, + { + "epoch": 12.619808306709265, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 15800 + }, + { + "epoch": 12.620607028753994, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 15801 + }, + { + "epoch": 12.621405750798722, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 15802 + }, + { + "epoch": 12.62220447284345, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 15803 + }, + { + "epoch": 12.62300319488818, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 15804 + }, + { + "epoch": 12.623801916932907, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0112, + "step": 15805 + }, + { + "epoch": 12.624600638977636, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 15806 + }, + { + "epoch": 12.625399361022364, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 15807 + }, + { + "epoch": 12.626198083067093, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 15808 + }, + { + "epoch": 12.62699680511182, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 15809 + }, + { + "epoch": 12.62779552715655, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 15810 + }, + { + "epoch": 12.628594249201278, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0181, + "step": 15811 + }, + { + "epoch": 12.629392971246006, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0161, + "step": 15812 + }, + { + "epoch": 12.630191693290735, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0145, + "step": 15813 + }, + { + "epoch": 12.630990415335463, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 15814 + }, + { + "epoch": 12.631789137380192, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 15815 + }, + { + "epoch": 12.63258785942492, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 15816 + }, + { + "epoch": 12.633386581469649, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 15817 + }, + { + "epoch": 12.634185303514377, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 15818 + }, + { + "epoch": 12.634984025559106, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0177, + "step": 15819 + }, + { + "epoch": 12.635782747603834, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0205, + "step": 15820 + }, + { + "epoch": 12.636581469648561, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0179, + "step": 15821 + }, + { + "epoch": 12.63738019169329, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 15822 + }, + { + "epoch": 12.638178913738018, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 15823 + }, + { + "epoch": 12.638977635782748, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 15824 + }, + { + "epoch": 12.639776357827476, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 15825 + }, + { + "epoch": 12.640575079872205, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 15826 + }, + { + "epoch": 12.641373801916933, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0188, + "step": 15827 + }, + { + "epoch": 12.642172523961662, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 15828 + }, + { + "epoch": 12.64297124600639, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 15829 + }, + { + "epoch": 12.643769968051117, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 15830 + }, + { + "epoch": 12.644568690095847, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 15831 + }, + { + "epoch": 12.645367412140574, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 15832 + }, + { + "epoch": 12.646166134185304, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 15833 + }, + { + "epoch": 12.646964856230031, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 15834 + }, + { + "epoch": 12.64776357827476, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.0155, + "step": 15835 + }, + { + "epoch": 12.648562300319488, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 15836 + }, + { + "epoch": 12.649361022364218, + "grad_norm": 0.1923828125, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 15837 + }, + { + "epoch": 12.650159744408946, + "grad_norm": 0.201171875, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 15838 + }, + { + "epoch": 12.650958466453675, + "grad_norm": 0.1953125, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 15839 + }, + { + "epoch": 12.651757188498403, + "grad_norm": 0.2373046875, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 15840 + }, + { + "epoch": 12.65255591054313, + "grad_norm": 0.212890625, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 15841 + }, + { + "epoch": 12.65335463258786, + "grad_norm": 0.166015625, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 15842 + }, + { + "epoch": 12.654153354632587, + "grad_norm": 0.19140625, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 15843 + }, + { + "epoch": 12.654952076677317, + "grad_norm": 0.1767578125, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 15844 + }, + { + "epoch": 12.655750798722044, + "grad_norm": 0.2333984375, + "learning_rate": 0.0005, + "loss": 1.0176, + "step": 15845 + }, + { + "epoch": 12.656549520766774, + "grad_norm": 0.185546875, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 15846 + }, + { + "epoch": 12.657348242811501, + "grad_norm": 0.279296875, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 15847 + }, + { + "epoch": 12.65814696485623, + "grad_norm": 0.1982421875, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 15848 + }, + { + "epoch": 12.658945686900958, + "grad_norm": 0.365234375, + "learning_rate": 0.0005, + "loss": 1.0195, + "step": 15849 + }, + { + "epoch": 12.659744408945686, + "grad_norm": 0.1806640625, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 15850 + }, + { + "epoch": 12.660543130990416, + "grad_norm": 0.35546875, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 15851 + }, + { + "epoch": 12.661341853035143, + "grad_norm": 0.2333984375, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 15852 + }, + { + "epoch": 12.662140575079873, + "grad_norm": 0.443359375, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 15853 + }, + { + "epoch": 12.6629392971246, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 15854 + }, + { + "epoch": 12.66373801916933, + "grad_norm": 0.427734375, + "learning_rate": 0.0005, + "loss": 1.0153, + "step": 15855 + }, + { + "epoch": 12.664536741214057, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 15856 + }, + { + "epoch": 12.665335463258787, + "grad_norm": 0.4765625, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 15857 + }, + { + "epoch": 12.666134185303514, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 15858 + }, + { + "epoch": 12.666932907348244, + "grad_norm": 0.3984375, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 15859 + }, + { + "epoch": 12.667731629392971, + "grad_norm": 0.2392578125, + "learning_rate": 0.0005, + "loss": 1.0205, + "step": 15860 + }, + { + "epoch": 12.668530351437699, + "grad_norm": 0.51953125, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 15861 + }, + { + "epoch": 12.669329073482428, + "grad_norm": 0.1787109375, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 15862 + }, + { + "epoch": 12.670127795527156, + "grad_norm": 0.59375, + "learning_rate": 0.0005, + "loss": 1.0154, + "step": 15863 + }, + { + "epoch": 12.670926517571885, + "grad_norm": 0.171875, + "learning_rate": 0.0005, + "loss": 1.0196, + "step": 15864 + }, + { + "epoch": 12.671725239616613, + "grad_norm": 0.734375, + "learning_rate": 0.0005, + "loss": 1.0354, + "step": 15865 + }, + { + "epoch": 12.672523961661343, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 15866 + }, + { + "epoch": 12.67332268370607, + "grad_norm": 0.45703125, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 15867 + }, + { + "epoch": 12.6741214057508, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 15868 + }, + { + "epoch": 12.674920127795527, + "grad_norm": 0.388671875, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 15869 + }, + { + "epoch": 12.675718849840255, + "grad_norm": 0.23046875, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 15870 + }, + { + "epoch": 12.676517571884984, + "grad_norm": 0.43359375, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 15871 + }, + { + "epoch": 12.677316293929712, + "grad_norm": 0.310546875, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 15872 + }, + { + "epoch": 12.678115015974441, + "grad_norm": 0.57421875, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 15873 + }, + { + "epoch": 12.678913738019169, + "grad_norm": 0.248046875, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 15874 + }, + { + "epoch": 12.679712460063898, + "grad_norm": 0.59375, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 15875 + }, + { + "epoch": 12.680511182108626, + "grad_norm": 0.287109375, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 15876 + }, + { + "epoch": 12.681309904153355, + "grad_norm": 0.69921875, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 15877 + }, + { + "epoch": 12.682108626198083, + "grad_norm": 0.314453125, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 15878 + }, + { + "epoch": 12.68290734824281, + "grad_norm": 0.7421875, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 15879 + }, + { + "epoch": 12.68370607028754, + "grad_norm": 0.4140625, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 15880 + }, + { + "epoch": 12.684504792332268, + "grad_norm": 0.8984375, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 15881 + }, + { + "epoch": 12.685303514376997, + "grad_norm": 0.341796875, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 15882 + }, + { + "epoch": 12.686102236421725, + "grad_norm": 0.70703125, + "learning_rate": 0.0005, + "loss": 1.0453, + "step": 15883 + }, + { + "epoch": 12.686900958466454, + "grad_norm": 0.58984375, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 15884 + }, + { + "epoch": 12.687699680511182, + "grad_norm": 0.6953125, + "learning_rate": 0.0005, + "loss": 1.0407, + "step": 15885 + }, + { + "epoch": 12.688498402555911, + "grad_norm": 0.59765625, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 15886 + }, + { + "epoch": 12.689297124600639, + "grad_norm": 0.423828125, + "learning_rate": 0.0005, + "loss": 1.0408, + "step": 15887 + }, + { + "epoch": 12.690095846645367, + "grad_norm": 0.447265625, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 15888 + }, + { + "epoch": 12.690894568690096, + "grad_norm": 0.1572265625, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 15889 + }, + { + "epoch": 12.691693290734824, + "grad_norm": 0.470703125, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 15890 + }, + { + "epoch": 12.692492012779553, + "grad_norm": 0.380859375, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 15891 + }, + { + "epoch": 12.69329073482428, + "grad_norm": 0.38671875, + "learning_rate": 0.0005, + "loss": 1.0122, + "step": 15892 + }, + { + "epoch": 12.69408945686901, + "grad_norm": 0.423828125, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 15893 + }, + { + "epoch": 12.694888178913738, + "grad_norm": 0.265625, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 15894 + }, + { + "epoch": 12.695686900958467, + "grad_norm": 0.306640625, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 15895 + }, + { + "epoch": 12.696485623003195, + "grad_norm": 0.16796875, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 15896 + }, + { + "epoch": 12.697284345047922, + "grad_norm": 0.234375, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 15897 + }, + { + "epoch": 12.698083067092652, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0205, + "step": 15898 + }, + { + "epoch": 12.69888178913738, + "grad_norm": 0.1630859375, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 15899 + }, + { + "epoch": 12.699680511182109, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 15900 + }, + { + "epoch": 12.700479233226837, + "grad_norm": 0.166015625, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 15901 + }, + { + "epoch": 12.701277955271566, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 15902 + }, + { + "epoch": 12.702076677316294, + "grad_norm": 0.173828125, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 15903 + }, + { + "epoch": 12.702875399361023, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 15904 + }, + { + "epoch": 12.70367412140575, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 15905 + }, + { + "epoch": 12.704472843450478, + "grad_norm": 0.1640625, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 15906 + }, + { + "epoch": 12.705271565495208, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 15907 + }, + { + "epoch": 12.706070287539935, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.038, + "step": 15908 + }, + { + "epoch": 12.706869009584665, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 15909 + }, + { + "epoch": 12.707667731629392, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 15910 + }, + { + "epoch": 12.708466453674122, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.0378, + "step": 15911 + }, + { + "epoch": 12.70926517571885, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 15912 + }, + { + "epoch": 12.710063897763579, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 15913 + }, + { + "epoch": 12.710862619808307, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 15914 + }, + { + "epoch": 12.711661341853034, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 15915 + }, + { + "epoch": 12.712460063897764, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0138, + "step": 15916 + }, + { + "epoch": 12.713258785942491, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 15917 + }, + { + "epoch": 12.71405750798722, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 15918 + }, + { + "epoch": 12.714856230031948, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 15919 + }, + { + "epoch": 12.715654952076678, + "grad_norm": 0.166015625, + "learning_rate": 0.0005, + "loss": 1.0174, + "step": 15920 + }, + { + "epoch": 12.716453674121405, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 15921 + }, + { + "epoch": 12.717252396166135, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 15922 + }, + { + "epoch": 12.718051118210862, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0183, + "step": 15923 + }, + { + "epoch": 12.718849840255592, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 15924 + }, + { + "epoch": 12.71964856230032, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 15925 + }, + { + "epoch": 12.720447284345047, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 15926 + }, + { + "epoch": 12.721246006389777, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 15927 + }, + { + "epoch": 12.722044728434504, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 15928 + }, + { + "epoch": 12.722843450479234, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 15929 + }, + { + "epoch": 12.723642172523961, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 15930 + }, + { + "epoch": 12.72444089456869, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 15931 + }, + { + "epoch": 12.725239616613418, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 15932 + }, + { + "epoch": 12.726038338658148, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 15933 + }, + { + "epoch": 12.726837060702875, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 15934 + }, + { + "epoch": 12.727635782747605, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 15935 + }, + { + "epoch": 12.728434504792332, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 15936 + }, + { + "epoch": 12.72923322683706, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 15937 + }, + { + "epoch": 12.73003194888179, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0176, + "step": 15938 + }, + { + "epoch": 12.730830670926517, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 15939 + }, + { + "epoch": 12.731629392971247, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 15940 + }, + { + "epoch": 12.732428115015974, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 15941 + }, + { + "epoch": 12.733226837060704, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 15942 + }, + { + "epoch": 12.734025559105431, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 15943 + }, + { + "epoch": 12.73482428115016, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 15944 + }, + { + "epoch": 12.735623003194888, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 15945 + }, + { + "epoch": 12.736421725239616, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 15946 + }, + { + "epoch": 12.737220447284345, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 15947 + }, + { + "epoch": 12.738019169329073, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 15948 + }, + { + "epoch": 12.738817891373802, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 15949 + }, + { + "epoch": 12.73961661341853, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0163, + "step": 15950 + }, + { + "epoch": 12.74041533546326, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 15951 + }, + { + "epoch": 12.741214057507987, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 15952 + }, + { + "epoch": 12.742012779552716, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0165, + "step": 15953 + }, + { + "epoch": 12.742811501597444, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 15954 + }, + { + "epoch": 12.743610223642172, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 15955 + }, + { + "epoch": 12.744408945686901, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 15956 + }, + { + "epoch": 12.745207667731629, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.013, + "step": 15957 + }, + { + "epoch": 12.746006389776358, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 15958 + }, + { + "epoch": 12.746805111821086, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 15959 + }, + { + "epoch": 12.747603833865815, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 15960 + }, + { + "epoch": 12.748402555910543, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0157, + "step": 15961 + }, + { + "epoch": 12.749201277955272, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 15962 + }, + { + "epoch": 12.75, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 15963 + }, + { + "epoch": 12.750798722044728, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 15964 + }, + { + "epoch": 12.751597444089457, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 15965 + }, + { + "epoch": 12.752396166134185, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 15966 + }, + { + "epoch": 12.753194888178914, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 15967 + }, + { + "epoch": 12.753993610223642, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0142, + "step": 15968 + }, + { + "epoch": 12.754792332268371, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 15969 + }, + { + "epoch": 12.755591054313099, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 15970 + }, + { + "epoch": 12.756389776357828, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 15971 + }, + { + "epoch": 12.757188498402556, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 15972 + }, + { + "epoch": 12.757987220447284, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 15973 + }, + { + "epoch": 12.758785942492013, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 15974 + }, + { + "epoch": 12.75958466453674, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 15975 + }, + { + "epoch": 12.76038338658147, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 15976 + }, + { + "epoch": 12.761182108626198, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0152, + "step": 15977 + }, + { + "epoch": 12.761980830670927, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0178, + "step": 15978 + }, + { + "epoch": 12.762779552715655, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0175, + "step": 15979 + }, + { + "epoch": 12.763578274760384, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 15980 + }, + { + "epoch": 12.764376996805112, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 15981 + }, + { + "epoch": 12.76517571884984, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 15982 + }, + { + "epoch": 12.765974440894569, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 15983 + }, + { + "epoch": 12.766773162939296, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 15984 + }, + { + "epoch": 12.767571884984026, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 15985 + }, + { + "epoch": 12.768370607028753, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0173, + "step": 15986 + }, + { + "epoch": 12.769169329073483, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 15987 + }, + { + "epoch": 12.76996805111821, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 15988 + }, + { + "epoch": 12.77076677316294, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 15989 + }, + { + "epoch": 12.771565495207668, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 15990 + }, + { + "epoch": 12.772364217252395, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 15991 + }, + { + "epoch": 12.773162939297125, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 15992 + }, + { + "epoch": 12.773961661341852, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0174, + "step": 15993 + }, + { + "epoch": 12.774760383386582, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 15994 + }, + { + "epoch": 12.77555910543131, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0196, + "step": 15995 + }, + { + "epoch": 12.776357827476039, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 15996 + }, + { + "epoch": 12.777156549520766, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0385, + "step": 15997 + }, + { + "epoch": 12.777955271565496, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0359, + "step": 15998 + }, + { + "epoch": 12.778753993610223, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 15999 + }, + { + "epoch": 12.779552715654953, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 16000 + }, + { + "epoch": 12.78035143769968, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 16001 + }, + { + "epoch": 12.781150159744408, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 16002 + }, + { + "epoch": 12.781948881789138, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 16003 + }, + { + "epoch": 12.782747603833865, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0127, + "step": 16004 + }, + { + "epoch": 12.783546325878595, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 16005 + }, + { + "epoch": 12.784345047923322, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 16006 + }, + { + "epoch": 12.785143769968052, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 16007 + }, + { + "epoch": 12.78594249201278, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0162, + "step": 16008 + }, + { + "epoch": 12.786741214057509, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 16009 + }, + { + "epoch": 12.787539936102236, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 16010 + }, + { + "epoch": 12.788338658146966, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 16011 + }, + { + "epoch": 12.789137380191693, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0362, + "step": 16012 + }, + { + "epoch": 12.789936102236421, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 16013 + }, + { + "epoch": 12.79073482428115, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 16014 + }, + { + "epoch": 12.791533546325878, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 16015 + }, + { + "epoch": 12.792332268370608, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 16016 + }, + { + "epoch": 12.793130990415335, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0195, + "step": 16017 + }, + { + "epoch": 12.793929712460065, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 16018 + }, + { + "epoch": 12.794728434504792, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 16019 + }, + { + "epoch": 12.795527156549522, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0341, + "step": 16020 + }, + { + "epoch": 12.79632587859425, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 16021 + }, + { + "epoch": 12.797124600638977, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0152, + "step": 16022 + }, + { + "epoch": 12.797923322683706, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 16023 + }, + { + "epoch": 12.798722044728434, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 16024 + }, + { + "epoch": 12.799520766773163, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 16025 + }, + { + "epoch": 12.800319488817891, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 16026 + }, + { + "epoch": 12.80111821086262, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 16027 + }, + { + "epoch": 12.801916932907348, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 16028 + }, + { + "epoch": 12.802715654952078, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 16029 + }, + { + "epoch": 12.803514376996805, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 16030 + }, + { + "epoch": 12.804313099041533, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 16031 + }, + { + "epoch": 12.805111821086262, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 16032 + }, + { + "epoch": 12.80591054313099, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 16033 + }, + { + "epoch": 12.80670926517572, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 16034 + }, + { + "epoch": 12.807507987220447, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0146, + "step": 16035 + }, + { + "epoch": 12.808306709265176, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 16036 + }, + { + "epoch": 12.809105431309904, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 16037 + }, + { + "epoch": 12.809904153354633, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 16038 + }, + { + "epoch": 12.810702875399361, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0173, + "step": 16039 + }, + { + "epoch": 12.811501597444089, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 16040 + }, + { + "epoch": 12.812300319488818, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 16041 + }, + { + "epoch": 12.813099041533546, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0334, + "step": 16042 + }, + { + "epoch": 12.813897763578275, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 16043 + }, + { + "epoch": 12.814696485623003, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 16044 + }, + { + "epoch": 12.815495207667732, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 16045 + }, + { + "epoch": 12.81629392971246, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 16046 + }, + { + "epoch": 12.81709265175719, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 16047 + }, + { + "epoch": 12.817891373801917, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 16048 + }, + { + "epoch": 12.818690095846645, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 16049 + }, + { + "epoch": 12.819488817891374, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 16050 + }, + { + "epoch": 12.820287539936102, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 16051 + }, + { + "epoch": 12.821086261980831, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0157, + "step": 16052 + }, + { + "epoch": 12.821884984025559, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 16053 + }, + { + "epoch": 12.822683706070288, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 16054 + }, + { + "epoch": 12.823482428115016, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 16055 + }, + { + "epoch": 12.824281150159745, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0122, + "step": 16056 + }, + { + "epoch": 12.825079872204473, + "grad_norm": 0.05810546875, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 16057 + }, + { + "epoch": 12.8258785942492, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 16058 + }, + { + "epoch": 12.82667731629393, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 16059 + }, + { + "epoch": 12.827476038338657, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 16060 + }, + { + "epoch": 12.828274760383387, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 16061 + }, + { + "epoch": 12.829073482428115, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 16062 + }, + { + "epoch": 12.829872204472844, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.018, + "step": 16063 + }, + { + "epoch": 12.830670926517572, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0185, + "step": 16064 + }, + { + "epoch": 12.831469648562301, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 16065 + }, + { + "epoch": 12.832268370607029, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 16066 + }, + { + "epoch": 12.833067092651756, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0162, + "step": 16067 + }, + { + "epoch": 12.833865814696486, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0124, + "step": 16068 + }, + { + "epoch": 12.834664536741213, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 16069 + }, + { + "epoch": 12.835463258785943, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.034, + "step": 16070 + }, + { + "epoch": 12.83626198083067, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 16071 + }, + { + "epoch": 12.8370607028754, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 16072 + }, + { + "epoch": 12.837859424920127, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0205, + "step": 16073 + }, + { + "epoch": 12.838658146964857, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 16074 + }, + { + "epoch": 12.839456869009584, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0174, + "step": 16075 + }, + { + "epoch": 12.840255591054314, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 16076 + }, + { + "epoch": 12.841054313099042, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 16077 + }, + { + "epoch": 12.84185303514377, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 16078 + }, + { + "epoch": 12.842651757188499, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 16079 + }, + { + "epoch": 12.843450479233226, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 16080 + }, + { + "epoch": 12.844249201277956, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0179, + "step": 16081 + }, + { + "epoch": 12.845047923322683, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 16082 + }, + { + "epoch": 12.845846645367413, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 16083 + }, + { + "epoch": 12.84664536741214, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 16084 + }, + { + "epoch": 12.84744408945687, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0363, + "step": 16085 + }, + { + "epoch": 12.848242811501597, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 16086 + }, + { + "epoch": 12.849041533546325, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 16087 + }, + { + "epoch": 12.849840255591054, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0337, + "step": 16088 + }, + { + "epoch": 12.850638977635782, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0117, + "step": 16089 + }, + { + "epoch": 12.851437699680512, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.017, + "step": 16090 + }, + { + "epoch": 12.85223642172524, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 16091 + }, + { + "epoch": 12.853035143769969, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0156, + "step": 16092 + }, + { + "epoch": 12.853833865814696, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 16093 + }, + { + "epoch": 12.854632587859426, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 16094 + }, + { + "epoch": 12.855431309904153, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 16095 + }, + { + "epoch": 12.856230031948883, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 16096 + }, + { + "epoch": 12.85702875399361, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 16097 + }, + { + "epoch": 12.857827476038338, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 16098 + }, + { + "epoch": 12.858626198083067, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 16099 + }, + { + "epoch": 12.859424920127795, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 16100 + }, + { + "epoch": 12.860223642172524, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0168, + "step": 16101 + }, + { + "epoch": 12.861022364217252, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0179, + "step": 16102 + }, + { + "epoch": 12.861821086261982, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 16103 + }, + { + "epoch": 12.86261980830671, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 16104 + }, + { + "epoch": 12.863418530351439, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 16105 + }, + { + "epoch": 12.864217252396166, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 16106 + }, + { + "epoch": 12.865015974440894, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 16107 + }, + { + "epoch": 12.865814696485623, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 16108 + }, + { + "epoch": 12.866613418530351, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 16109 + }, + { + "epoch": 12.86741214057508, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 16110 + }, + { + "epoch": 12.868210862619808, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 16111 + }, + { + "epoch": 12.869009584664537, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 16112 + }, + { + "epoch": 12.869808306709265, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0136, + "step": 16113 + }, + { + "epoch": 12.870607028753994, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 16114 + }, + { + "epoch": 12.871405750798722, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0148, + "step": 16115 + }, + { + "epoch": 12.87220447284345, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 16116 + }, + { + "epoch": 12.87300319488818, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 16117 + }, + { + "epoch": 12.873801916932907, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0165, + "step": 16118 + }, + { + "epoch": 12.874600638977636, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 16119 + }, + { + "epoch": 12.875399361022364, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 16120 + }, + { + "epoch": 12.876198083067093, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 16121 + }, + { + "epoch": 12.87699680511182, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.015, + "step": 16122 + }, + { + "epoch": 12.87779552715655, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0138, + "step": 16123 + }, + { + "epoch": 12.878594249201278, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 16124 + }, + { + "epoch": 12.879392971246006, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 16125 + }, + { + "epoch": 12.880191693290735, + "grad_norm": 0.169921875, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 16126 + }, + { + "epoch": 12.880990415335463, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 16127 + }, + { + "epoch": 12.881789137380192, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 16128 + }, + { + "epoch": 12.88258785942492, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 16129 + }, + { + "epoch": 12.883386581469649, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 16130 + }, + { + "epoch": 12.884185303514377, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 16131 + }, + { + "epoch": 12.884984025559106, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 16132 + }, + { + "epoch": 12.885782747603834, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 16133 + }, + { + "epoch": 12.886581469648561, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 16134 + }, + { + "epoch": 12.88738019169329, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 16135 + }, + { + "epoch": 12.888178913738018, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 16136 + }, + { + "epoch": 12.888977635782748, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0155, + "step": 16137 + }, + { + "epoch": 12.889776357827476, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.01, + "step": 16138 + }, + { + "epoch": 12.890575079872205, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 16139 + }, + { + "epoch": 12.891373801916933, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 16140 + }, + { + "epoch": 12.892172523961662, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 16141 + }, + { + "epoch": 12.89297124600639, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0181, + "step": 16142 + }, + { + "epoch": 12.893769968051117, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0182, + "step": 16143 + }, + { + "epoch": 12.894568690095847, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0172, + "step": 16144 + }, + { + "epoch": 12.895367412140574, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 16145 + }, + { + "epoch": 12.896166134185304, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 16146 + }, + { + "epoch": 12.896964856230031, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0195, + "step": 16147 + }, + { + "epoch": 12.89776357827476, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 16148 + }, + { + "epoch": 12.898562300319488, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 16149 + }, + { + "epoch": 12.899361022364218, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 16150 + }, + { + "epoch": 12.900159744408946, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 16151 + }, + { + "epoch": 12.900958466453675, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0167, + "step": 16152 + }, + { + "epoch": 12.901757188498403, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 16153 + }, + { + "epoch": 12.90255591054313, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0136, + "step": 16154 + }, + { + "epoch": 12.90335463258786, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0129, + "step": 16155 + }, + { + "epoch": 12.904153354632587, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0119, + "step": 16156 + }, + { + "epoch": 12.904952076677317, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 16157 + }, + { + "epoch": 12.905750798722044, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 16158 + }, + { + "epoch": 12.906549520766774, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 16159 + }, + { + "epoch": 12.907348242811501, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0183, + "step": 16160 + }, + { + "epoch": 12.90814696485623, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 16161 + }, + { + "epoch": 12.908945686900958, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 16162 + }, + { + "epoch": 12.909744408945686, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 16163 + }, + { + "epoch": 12.910543130990416, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 16164 + }, + { + "epoch": 12.911341853035143, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 16165 + }, + { + "epoch": 12.912140575079873, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 16166 + }, + { + "epoch": 12.9129392971246, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 16167 + }, + { + "epoch": 12.91373801916933, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0181, + "step": 16168 + }, + { + "epoch": 12.914536741214057, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 16169 + }, + { + "epoch": 12.915335463258787, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0092, + "step": 16170 + }, + { + "epoch": 12.916134185303514, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 16171 + }, + { + "epoch": 12.916932907348244, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 16172 + }, + { + "epoch": 12.917731629392971, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 16173 + }, + { + "epoch": 12.918530351437699, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 16174 + }, + { + "epoch": 12.919329073482428, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 16175 + }, + { + "epoch": 12.920127795527156, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 16176 + }, + { + "epoch": 12.920926517571885, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 16177 + }, + { + "epoch": 12.921725239616613, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 16178 + }, + { + "epoch": 12.922523961661343, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 16179 + }, + { + "epoch": 12.92332268370607, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0332, + "step": 16180 + }, + { + "epoch": 12.9241214057508, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 16181 + }, + { + "epoch": 12.924920127795527, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 16182 + }, + { + "epoch": 12.925718849840255, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 16183 + }, + { + "epoch": 12.926517571884984, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 16184 + }, + { + "epoch": 12.927316293929712, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 16185 + }, + { + "epoch": 12.928115015974441, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0119, + "step": 16186 + }, + { + "epoch": 12.928913738019169, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 16187 + }, + { + "epoch": 12.929712460063898, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0172, + "step": 16188 + }, + { + "epoch": 12.930511182108626, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 16189 + }, + { + "epoch": 12.931309904153355, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0195, + "step": 16190 + }, + { + "epoch": 12.932108626198083, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 16191 + }, + { + "epoch": 12.93290734824281, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 16192 + }, + { + "epoch": 12.93370607028754, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0175, + "step": 16193 + }, + { + "epoch": 12.934504792332268, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0119, + "step": 16194 + }, + { + "epoch": 12.935303514376997, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 16195 + }, + { + "epoch": 12.936102236421725, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 16196 + }, + { + "epoch": 12.936900958466454, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 16197 + }, + { + "epoch": 12.937699680511182, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 16198 + }, + { + "epoch": 12.938498402555911, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.018, + "step": 16199 + }, + { + "epoch": 12.939297124600639, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 16200 + }, + { + "epoch": 12.940095846645367, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 16201 + }, + { + "epoch": 12.940894568690096, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 16202 + }, + { + "epoch": 12.941693290734824, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 16203 + }, + { + "epoch": 12.942492012779553, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0162, + "step": 16204 + }, + { + "epoch": 12.94329073482428, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 16205 + }, + { + "epoch": 12.94408945686901, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 16206 + }, + { + "epoch": 12.944888178913738, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 16207 + }, + { + "epoch": 12.945686900958467, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 16208 + }, + { + "epoch": 12.946485623003195, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0132, + "step": 16209 + }, + { + "epoch": 12.947284345047922, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 16210 + }, + { + "epoch": 12.948083067092652, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 16211 + }, + { + "epoch": 12.94888178913738, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0159, + "step": 16212 + }, + { + "epoch": 12.949680511182109, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 16213 + }, + { + "epoch": 12.950479233226837, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 16214 + }, + { + "epoch": 12.951277955271566, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 16215 + }, + { + "epoch": 12.952076677316294, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 16216 + }, + { + "epoch": 12.952875399361023, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 16217 + }, + { + "epoch": 12.95367412140575, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 16218 + }, + { + "epoch": 12.954472843450478, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 16219 + }, + { + "epoch": 12.955271565495208, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 16220 + }, + { + "epoch": 12.956070287539935, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 16221 + }, + { + "epoch": 12.956869009584665, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 16222 + }, + { + "epoch": 12.957667731629392, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 16223 + }, + { + "epoch": 12.958466453674122, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 16224 + }, + { + "epoch": 12.95926517571885, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 16225 + }, + { + "epoch": 12.960063897763579, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 16226 + }, + { + "epoch": 12.960862619808307, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 16227 + }, + { + "epoch": 12.961661341853034, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 16228 + }, + { + "epoch": 12.962460063897764, + "grad_norm": 0.169921875, + "learning_rate": 0.0005, + "loss": 1.0172, + "step": 16229 + }, + { + "epoch": 12.963258785942491, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 16230 + }, + { + "epoch": 12.96405750798722, + "grad_norm": 0.1484375, + "learning_rate": 0.0005, + "loss": 1.018, + "step": 16231 + }, + { + "epoch": 12.964856230031948, + "grad_norm": 0.1767578125, + "learning_rate": 0.0005, + "loss": 1.0125, + "step": 16232 + }, + { + "epoch": 12.965654952076678, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 16233 + }, + { + "epoch": 12.966453674121405, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0185, + "step": 16234 + }, + { + "epoch": 12.967252396166135, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 16235 + }, + { + "epoch": 12.968051118210862, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 16236 + }, + { + "epoch": 12.968849840255592, + "grad_norm": 0.166015625, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 16237 + }, + { + "epoch": 12.96964856230032, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 16238 + }, + { + "epoch": 12.970447284345047, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 16239 + }, + { + "epoch": 12.971246006389777, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 16240 + }, + { + "epoch": 12.972044728434504, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 16241 + }, + { + "epoch": 12.972843450479234, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 16242 + }, + { + "epoch": 12.973642172523961, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0174, + "step": 16243 + }, + { + "epoch": 12.97444089456869, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0156, + "step": 16244 + }, + { + "epoch": 12.975239616613418, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 16245 + }, + { + "epoch": 12.976038338658148, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0182, + "step": 16246 + }, + { + "epoch": 12.976837060702875, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0196, + "step": 16247 + }, + { + "epoch": 12.977635782747605, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0179, + "step": 16248 + }, + { + "epoch": 12.978434504792332, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 16249 + }, + { + "epoch": 12.97923322683706, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0188, + "step": 16250 + }, + { + "epoch": 12.98003194888179, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 16251 + }, + { + "epoch": 12.980830670926517, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 16252 + }, + { + "epoch": 12.981629392971247, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0192, + "step": 16253 + }, + { + "epoch": 12.982428115015974, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 16254 + }, + { + "epoch": 12.983226837060704, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 16255 + }, + { + "epoch": 12.984025559105431, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0174, + "step": 16256 + }, + { + "epoch": 12.98482428115016, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 16257 + }, + { + "epoch": 12.985623003194888, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0157, + "step": 16258 + }, + { + "epoch": 12.986421725239616, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0205, + "step": 16259 + }, + { + "epoch": 12.987220447284345, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 16260 + }, + { + "epoch": 12.988019169329073, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 16261 + }, + { + "epoch": 12.988817891373802, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0157, + "step": 16262 + }, + { + "epoch": 12.98961661341853, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.018, + "step": 16263 + }, + { + "epoch": 12.99041533546326, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 16264 + }, + { + "epoch": 12.991214057507987, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 16265 + }, + { + "epoch": 12.992012779552716, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0131, + "step": 16266 + }, + { + "epoch": 12.992811501597444, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 16267 + }, + { + "epoch": 12.993610223642172, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 16268 + }, + { + "epoch": 12.994408945686901, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 16269 + }, + { + "epoch": 12.995207667731629, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 16270 + }, + { + "epoch": 12.996006389776358, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 16271 + }, + { + "epoch": 12.996805111821086, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 16272 + }, + { + "epoch": 12.997603833865815, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 16273 + }, + { + "epoch": 12.998402555910543, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 16274 + }, + { + "epoch": 12.999201277955272, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 16275 + }, + { + "epoch": 13.0, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 16276 + }, + { + "epoch": 13.000798722044728, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 16277 + }, + { + "epoch": 13.001597444089457, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 16278 + }, + { + "epoch": 13.002396166134185, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0187, + "step": 16279 + }, + { + "epoch": 13.003194888178914, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 16280 + }, + { + "epoch": 13.003993610223642, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 16281 + }, + { + "epoch": 13.004792332268371, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 16282 + }, + { + "epoch": 13.005591054313099, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0175, + "step": 16283 + }, + { + "epoch": 13.006389776357828, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0113, + "step": 16284 + }, + { + "epoch": 13.007188498402556, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 16285 + }, + { + "epoch": 13.007987220447284, + "grad_norm": 0.16796875, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 16286 + }, + { + "epoch": 13.008785942492013, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 16287 + }, + { + "epoch": 13.00958466453674, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0146, + "step": 16288 + }, + { + "epoch": 13.01038338658147, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 16289 + }, + { + "epoch": 13.011182108626198, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0181, + "step": 16290 + }, + { + "epoch": 13.011980830670927, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 16291 + }, + { + "epoch": 13.012779552715655, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 16292 + }, + { + "epoch": 13.013578274760384, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 16293 + }, + { + "epoch": 13.014376996805112, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 16294 + }, + { + "epoch": 13.01517571884984, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 16295 + }, + { + "epoch": 13.015974440894569, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 16296 + }, + { + "epoch": 13.016773162939296, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 16297 + }, + { + "epoch": 13.017571884984026, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 16298 + }, + { + "epoch": 13.018370607028753, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 16299 + }, + { + "epoch": 13.019169329073483, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 16300 + }, + { + "epoch": 13.01996805111821, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0104, + "step": 16301 + }, + { + "epoch": 13.02076677316294, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 16302 + }, + { + "epoch": 13.021565495207668, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 16303 + }, + { + "epoch": 13.022364217252397, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0158, + "step": 16304 + }, + { + "epoch": 13.023162939297125, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 16305 + }, + { + "epoch": 13.023961661341852, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 16306 + }, + { + "epoch": 13.024760383386582, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0159, + "step": 16307 + }, + { + "epoch": 13.02555910543131, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 16308 + }, + { + "epoch": 13.026357827476039, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0147, + "step": 16309 + }, + { + "epoch": 13.027156549520766, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 16310 + }, + { + "epoch": 13.027955271565496, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0181, + "step": 16311 + }, + { + "epoch": 13.028753993610223, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 16312 + }, + { + "epoch": 13.029552715654953, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0182, + "step": 16313 + }, + { + "epoch": 13.03035143769968, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 16314 + }, + { + "epoch": 13.031150159744408, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 16315 + }, + { + "epoch": 13.031948881789138, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 16316 + }, + { + "epoch": 13.032747603833865, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 16317 + }, + { + "epoch": 13.033546325878595, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 16318 + }, + { + "epoch": 13.034345047923322, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 16319 + }, + { + "epoch": 13.035143769968052, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0155, + "step": 16320 + }, + { + "epoch": 13.03594249201278, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 16321 + }, + { + "epoch": 13.036741214057509, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 16322 + }, + { + "epoch": 13.037539936102236, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 16323 + }, + { + "epoch": 13.038338658146964, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 16324 + }, + { + "epoch": 13.039137380191693, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0167, + "step": 16325 + }, + { + "epoch": 13.039936102236421, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0173, + "step": 16326 + }, + { + "epoch": 13.04073482428115, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.015, + "step": 16327 + }, + { + "epoch": 13.041533546325878, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 16328 + }, + { + "epoch": 13.042332268370608, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 16329 + }, + { + "epoch": 13.043130990415335, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0124, + "step": 16330 + }, + { + "epoch": 13.043929712460065, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 16331 + }, + { + "epoch": 13.044728434504792, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 16332 + }, + { + "epoch": 13.04552715654952, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0156, + "step": 16333 + }, + { + "epoch": 13.04632587859425, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 16334 + }, + { + "epoch": 13.047124600638977, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 16335 + }, + { + "epoch": 13.047923322683706, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 16336 + }, + { + "epoch": 13.048722044728434, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 16337 + }, + { + "epoch": 13.049520766773163, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 16338 + }, + { + "epoch": 13.050319488817891, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 16339 + }, + { + "epoch": 13.05111821086262, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 16340 + }, + { + "epoch": 13.051916932907348, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0187, + "step": 16341 + }, + { + "epoch": 13.052715654952078, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0205, + "step": 16342 + }, + { + "epoch": 13.053514376996805, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 16343 + }, + { + "epoch": 13.054313099041533, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 16344 + }, + { + "epoch": 13.055111821086262, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 16345 + }, + { + "epoch": 13.05591054313099, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0124, + "step": 16346 + }, + { + "epoch": 13.05670926517572, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.0136, + "step": 16347 + }, + { + "epoch": 13.057507987220447, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 16348 + }, + { + "epoch": 13.058306709265176, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 16349 + }, + { + "epoch": 13.059105431309904, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0126, + "step": 16350 + }, + { + "epoch": 13.059904153354633, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 16351 + }, + { + "epoch": 13.060702875399361, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 16352 + }, + { + "epoch": 13.061501597444089, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0195, + "step": 16353 + }, + { + "epoch": 13.062300319488818, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 16354 + }, + { + "epoch": 13.063099041533546, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 16355 + }, + { + "epoch": 13.063897763578275, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 16356 + }, + { + "epoch": 13.064696485623003, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0149, + "step": 16357 + }, + { + "epoch": 13.065495207667732, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0128, + "step": 16358 + }, + { + "epoch": 13.06629392971246, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 16359 + }, + { + "epoch": 13.06709265175719, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 16360 + }, + { + "epoch": 13.067891373801917, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 16361 + }, + { + "epoch": 13.068690095846645, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0171, + "step": 16362 + }, + { + "epoch": 13.069488817891374, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 16363 + }, + { + "epoch": 13.070287539936102, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0162, + "step": 16364 + }, + { + "epoch": 13.071086261980831, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 16365 + }, + { + "epoch": 13.071884984025559, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0196, + "step": 16366 + }, + { + "epoch": 13.072683706070288, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 16367 + }, + { + "epoch": 13.073482428115016, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 16368 + }, + { + "epoch": 13.074281150159745, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 16369 + }, + { + "epoch": 13.075079872204473, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 16370 + }, + { + "epoch": 13.0758785942492, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 16371 + }, + { + "epoch": 13.07667731629393, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0175, + "step": 16372 + }, + { + "epoch": 13.077476038338657, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 16373 + }, + { + "epoch": 13.078274760383387, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 16374 + }, + { + "epoch": 13.079073482428115, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0148, + "step": 16375 + }, + { + "epoch": 13.079872204472844, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 16376 + }, + { + "epoch": 13.080670926517572, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 16377 + }, + { + "epoch": 13.081469648562301, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 16378 + }, + { + "epoch": 13.082268370607029, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 16379 + }, + { + "epoch": 13.083067092651758, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 16380 + }, + { + "epoch": 13.083865814696486, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0182, + "step": 16381 + }, + { + "epoch": 13.084664536741213, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 16382 + }, + { + "epoch": 13.085463258785943, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 16383 + }, + { + "epoch": 13.08626198083067, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 16384 + }, + { + "epoch": 13.0870607028754, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0118, + "step": 16385 + }, + { + "epoch": 13.087859424920127, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 16386 + }, + { + "epoch": 13.088658146964857, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 16387 + }, + { + "epoch": 13.089456869009584, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 16388 + }, + { + "epoch": 13.090255591054314, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 16389 + }, + { + "epoch": 13.091054313099042, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0159, + "step": 16390 + }, + { + "epoch": 13.09185303514377, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 16391 + }, + { + "epoch": 13.092651757188499, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 16392 + }, + { + "epoch": 13.093450479233226, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 16393 + }, + { + "epoch": 13.094249201277956, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0154, + "step": 16394 + }, + { + "epoch": 13.095047923322683, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0129, + "step": 16395 + }, + { + "epoch": 13.095846645367413, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0187, + "step": 16396 + }, + { + "epoch": 13.09664536741214, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 16397 + }, + { + "epoch": 13.09744408945687, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 16398 + }, + { + "epoch": 13.098242811501597, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0187, + "step": 16399 + }, + { + "epoch": 13.099041533546325, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0188, + "step": 16400 + }, + { + "epoch": 13.099840255591054, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 16401 + }, + { + "epoch": 13.100638977635782, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 16402 + }, + { + "epoch": 13.101437699680512, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 16403 + }, + { + "epoch": 13.10223642172524, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0182, + "step": 16404 + }, + { + "epoch": 13.103035143769969, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 16405 + }, + { + "epoch": 13.103833865814696, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 16406 + }, + { + "epoch": 13.104632587859426, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.018, + "step": 16407 + }, + { + "epoch": 13.105431309904153, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 16408 + }, + { + "epoch": 13.106230031948881, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 16409 + }, + { + "epoch": 13.10702875399361, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.016, + "step": 16410 + }, + { + "epoch": 13.107827476038338, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 16411 + }, + { + "epoch": 13.108626198083067, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 16412 + }, + { + "epoch": 13.109424920127795, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0165, + "step": 16413 + }, + { + "epoch": 13.110223642172524, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 16414 + }, + { + "epoch": 13.111022364217252, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 16415 + }, + { + "epoch": 13.111821086261982, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0362, + "step": 16416 + }, + { + "epoch": 13.11261980830671, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 16417 + }, + { + "epoch": 13.113418530351439, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 16418 + }, + { + "epoch": 13.114217252396166, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 16419 + }, + { + "epoch": 13.115015974440894, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 16420 + }, + { + "epoch": 13.115814696485623, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 16421 + }, + { + "epoch": 13.116613418530351, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 16422 + }, + { + "epoch": 13.11741214057508, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 16423 + }, + { + "epoch": 13.118210862619808, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 16424 + }, + { + "epoch": 13.119009584664537, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 16425 + }, + { + "epoch": 13.119808306709265, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 16426 + }, + { + "epoch": 13.120607028753994, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 16427 + }, + { + "epoch": 13.121405750798722, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 16428 + }, + { + "epoch": 13.12220447284345, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 16429 + }, + { + "epoch": 13.12300319488818, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 16430 + }, + { + "epoch": 13.123801916932907, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 16431 + }, + { + "epoch": 13.124600638977636, + "grad_norm": 0.14453125, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 16432 + }, + { + "epoch": 13.125399361022364, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 16433 + }, + { + "epoch": 13.126198083067093, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 16434 + }, + { + "epoch": 13.12699680511182, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 16435 + }, + { + "epoch": 13.12779552715655, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0178, + "step": 16436 + }, + { + "epoch": 13.128594249201278, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.018, + "step": 16437 + }, + { + "epoch": 13.129392971246006, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 16438 + }, + { + "epoch": 13.130191693290735, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 16439 + }, + { + "epoch": 13.130990415335463, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 16440 + }, + { + "epoch": 13.131789137380192, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0144, + "step": 16441 + }, + { + "epoch": 13.13258785942492, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 16442 + }, + { + "epoch": 13.133386581469649, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 16443 + }, + { + "epoch": 13.134185303514377, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 16444 + }, + { + "epoch": 13.134984025559106, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 16445 + }, + { + "epoch": 13.135782747603834, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 16446 + }, + { + "epoch": 13.136581469648561, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 16447 + }, + { + "epoch": 13.13738019169329, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 16448 + }, + { + "epoch": 13.138178913738018, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0166, + "step": 16449 + }, + { + "epoch": 13.138977635782748, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 16450 + }, + { + "epoch": 13.139776357827476, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 16451 + }, + { + "epoch": 13.140575079872205, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0168, + "step": 16452 + }, + { + "epoch": 13.141373801916933, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 16453 + }, + { + "epoch": 13.142172523961662, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 16454 + }, + { + "epoch": 13.14297124600639, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0154, + "step": 16455 + }, + { + "epoch": 13.143769968051119, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 16456 + }, + { + "epoch": 13.144568690095847, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0166, + "step": 16457 + }, + { + "epoch": 13.145367412140574, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 16458 + }, + { + "epoch": 13.146166134185304, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 16459 + }, + { + "epoch": 13.146964856230031, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 16460 + }, + { + "epoch": 13.14776357827476, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 16461 + }, + { + "epoch": 13.148562300319488, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 16462 + }, + { + "epoch": 13.149361022364218, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 16463 + }, + { + "epoch": 13.150159744408946, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 16464 + }, + { + "epoch": 13.150958466453675, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 16465 + }, + { + "epoch": 13.151757188498403, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 16466 + }, + { + "epoch": 13.15255591054313, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 16467 + }, + { + "epoch": 13.15335463258786, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 16468 + }, + { + "epoch": 13.154153354632587, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0177, + "step": 16469 + }, + { + "epoch": 13.154952076677317, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0161, + "step": 16470 + }, + { + "epoch": 13.155750798722044, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 16471 + }, + { + "epoch": 13.156549520766774, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 16472 + }, + { + "epoch": 13.157348242811501, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 16473 + }, + { + "epoch": 13.15814696485623, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0159, + "step": 16474 + }, + { + "epoch": 13.158945686900958, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 16475 + }, + { + "epoch": 13.159744408945686, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0143, + "step": 16476 + }, + { + "epoch": 13.160543130990416, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0169, + "step": 16477 + }, + { + "epoch": 13.161341853035143, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0196, + "step": 16478 + }, + { + "epoch": 13.162140575079873, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 16479 + }, + { + "epoch": 13.1629392971246, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 16480 + }, + { + "epoch": 13.16373801916933, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0158, + "step": 16481 + }, + { + "epoch": 13.164536741214057, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 16482 + }, + { + "epoch": 13.165335463258787, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0171, + "step": 16483 + }, + { + "epoch": 13.166134185303514, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 16484 + }, + { + "epoch": 13.166932907348242, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0187, + "step": 16485 + }, + { + "epoch": 13.167731629392971, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 16486 + }, + { + "epoch": 13.168530351437699, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 16487 + }, + { + "epoch": 13.169329073482428, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 16488 + }, + { + "epoch": 13.170127795527156, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0171, + "step": 16489 + }, + { + "epoch": 13.170926517571885, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 16490 + }, + { + "epoch": 13.171725239616613, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0181, + "step": 16491 + }, + { + "epoch": 13.172523961661343, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 16492 + }, + { + "epoch": 13.17332268370607, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 16493 + }, + { + "epoch": 13.1741214057508, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 16494 + }, + { + "epoch": 13.174920127795527, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 16495 + }, + { + "epoch": 13.175718849840255, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 16496 + }, + { + "epoch": 13.176517571884984, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0163, + "step": 16497 + }, + { + "epoch": 13.177316293929712, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 16498 + }, + { + "epoch": 13.178115015974441, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 16499 + }, + { + "epoch": 13.178913738019169, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 16500 + }, + { + "epoch": 13.179712460063898, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 16501 + }, + { + "epoch": 13.180511182108626, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 16502 + }, + { + "epoch": 13.181309904153355, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 16503 + }, + { + "epoch": 13.182108626198083, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 16504 + }, + { + "epoch": 13.18290734824281, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 16505 + }, + { + "epoch": 13.18370607028754, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 16506 + }, + { + "epoch": 13.184504792332268, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 16507 + }, + { + "epoch": 13.185303514376997, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0179, + "step": 16508 + }, + { + "epoch": 13.186102236421725, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 16509 + }, + { + "epoch": 13.186900958466454, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.013, + "step": 16510 + }, + { + "epoch": 13.187699680511182, + "grad_norm": 0.171875, + "learning_rate": 0.0005, + "loss": 1.0177, + "step": 16511 + }, + { + "epoch": 13.188498402555911, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 16512 + }, + { + "epoch": 13.189297124600639, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 16513 + }, + { + "epoch": 13.190095846645367, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 16514 + }, + { + "epoch": 13.190894568690096, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0205, + "step": 16515 + }, + { + "epoch": 13.191693290734824, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0167, + "step": 16516 + }, + { + "epoch": 13.192492012779553, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0109, + "step": 16517 + }, + { + "epoch": 13.19329073482428, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 16518 + }, + { + "epoch": 13.19408945686901, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0195, + "step": 16519 + }, + { + "epoch": 13.194888178913738, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 16520 + }, + { + "epoch": 13.195686900958467, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 16521 + }, + { + "epoch": 13.196485623003195, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0159, + "step": 16522 + }, + { + "epoch": 13.197284345047922, + "grad_norm": 0.16015625, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 16523 + }, + { + "epoch": 13.198083067092652, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 16524 + }, + { + "epoch": 13.19888178913738, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 16525 + }, + { + "epoch": 13.199680511182109, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 16526 + }, + { + "epoch": 13.200479233226837, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 16527 + }, + { + "epoch": 13.201277955271566, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 16528 + }, + { + "epoch": 13.202076677316294, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 16529 + }, + { + "epoch": 13.202875399361023, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0182, + "step": 16530 + }, + { + "epoch": 13.20367412140575, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 16531 + }, + { + "epoch": 13.204472843450478, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 16532 + }, + { + "epoch": 13.205271565495208, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 16533 + }, + { + "epoch": 13.206070287539935, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 16534 + }, + { + "epoch": 13.206869009584665, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 16535 + }, + { + "epoch": 13.207667731629392, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 16536 + }, + { + "epoch": 13.208466453674122, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0407, + "step": 16537 + }, + { + "epoch": 13.20926517571885, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 16538 + }, + { + "epoch": 13.210063897763579, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 16539 + }, + { + "epoch": 13.210862619808307, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 16540 + }, + { + "epoch": 13.211661341853036, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 16541 + }, + { + "epoch": 13.212460063897764, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 16542 + }, + { + "epoch": 13.213258785942491, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 16543 + }, + { + "epoch": 13.21405750798722, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 16544 + }, + { + "epoch": 13.214856230031948, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0163, + "step": 16545 + }, + { + "epoch": 13.215654952076678, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0092, + "step": 16546 + }, + { + "epoch": 13.216453674121405, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 16547 + }, + { + "epoch": 13.217252396166135, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0388, + "step": 16548 + }, + { + "epoch": 13.218051118210862, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 16549 + }, + { + "epoch": 13.218849840255592, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 16550 + }, + { + "epoch": 13.21964856230032, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0112, + "step": 16551 + }, + { + "epoch": 13.220447284345047, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 16552 + }, + { + "epoch": 13.221246006389777, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 16553 + }, + { + "epoch": 13.222044728434504, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 16554 + }, + { + "epoch": 13.222843450479234, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 16555 + }, + { + "epoch": 13.223642172523961, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0118, + "step": 16556 + }, + { + "epoch": 13.22444089456869, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 16557 + }, + { + "epoch": 13.225239616613418, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 16558 + }, + { + "epoch": 13.226038338658148, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 16559 + }, + { + "epoch": 13.226837060702875, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 16560 + }, + { + "epoch": 13.227635782747603, + "grad_norm": 0.1728515625, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 16561 + }, + { + "epoch": 13.228434504792332, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 16562 + }, + { + "epoch": 13.22923322683706, + "grad_norm": 0.1689453125, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 16563 + }, + { + "epoch": 13.23003194888179, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.0175, + "step": 16564 + }, + { + "epoch": 13.230830670926517, + "grad_norm": 0.162109375, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 16565 + }, + { + "epoch": 13.231629392971247, + "grad_norm": 0.169921875, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 16566 + }, + { + "epoch": 13.232428115015974, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 16567 + }, + { + "epoch": 13.233226837060704, + "grad_norm": 0.17578125, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 16568 + }, + { + "epoch": 13.234025559105431, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 16569 + }, + { + "epoch": 13.23482428115016, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 16570 + }, + { + "epoch": 13.235623003194888, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0199, + "step": 16571 + }, + { + "epoch": 13.236421725239616, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 16572 + }, + { + "epoch": 13.237220447284345, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 16573 + }, + { + "epoch": 13.238019169329073, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 16574 + }, + { + "epoch": 13.238817891373802, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 16575 + }, + { + "epoch": 13.23961661341853, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0162, + "step": 16576 + }, + { + "epoch": 13.24041533546326, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0163, + "step": 16577 + }, + { + "epoch": 13.241214057507987, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 16578 + }, + { + "epoch": 13.242012779552716, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 16579 + }, + { + "epoch": 13.242811501597444, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 16580 + }, + { + "epoch": 13.243610223642172, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 16581 + }, + { + "epoch": 13.244408945686901, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 16582 + }, + { + "epoch": 13.245207667731629, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 16583 + }, + { + "epoch": 13.246006389776358, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 16584 + }, + { + "epoch": 13.246805111821086, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 16585 + }, + { + "epoch": 13.247603833865815, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 16586 + }, + { + "epoch": 13.248402555910543, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 16587 + }, + { + "epoch": 13.249201277955272, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0168, + "step": 16588 + }, + { + "epoch": 13.25, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 16589 + }, + { + "epoch": 13.250798722044728, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0168, + "step": 16590 + }, + { + "epoch": 13.251597444089457, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 16591 + }, + { + "epoch": 13.252396166134185, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0161, + "step": 16592 + }, + { + "epoch": 13.253194888178914, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 16593 + }, + { + "epoch": 13.253993610223642, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 16594 + }, + { + "epoch": 13.254792332268371, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0192, + "step": 16595 + }, + { + "epoch": 13.255591054313099, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 16596 + }, + { + "epoch": 13.256389776357828, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 16597 + }, + { + "epoch": 13.257188498402556, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.015, + "step": 16598 + }, + { + "epoch": 13.257987220447284, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 16599 + }, + { + "epoch": 13.258785942492013, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 16600 + }, + { + "epoch": 13.25958466453674, + "grad_norm": 0.173828125, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 16601 + }, + { + "epoch": 13.26038338658147, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0132, + "step": 16602 + }, + { + "epoch": 13.261182108626198, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 16603 + }, + { + "epoch": 13.261980830670927, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 16604 + }, + { + "epoch": 13.262779552715655, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 16605 + }, + { + "epoch": 13.263578274760384, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 16606 + }, + { + "epoch": 13.264376996805112, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 16607 + }, + { + "epoch": 13.26517571884984, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0182, + "step": 16608 + }, + { + "epoch": 13.265974440894569, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 16609 + }, + { + "epoch": 13.266773162939296, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 16610 + }, + { + "epoch": 13.267571884984026, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 16611 + }, + { + "epoch": 13.268370607028753, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 16612 + }, + { + "epoch": 13.269169329073483, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 16613 + }, + { + "epoch": 13.26996805111821, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 16614 + }, + { + "epoch": 13.27076677316294, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 16615 + }, + { + "epoch": 13.271565495207668, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 16616 + }, + { + "epoch": 13.272364217252397, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 16617 + }, + { + "epoch": 13.273162939297125, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 16618 + }, + { + "epoch": 13.273961661341852, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0199, + "step": 16619 + }, + { + "epoch": 13.274760383386582, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 16620 + }, + { + "epoch": 13.27555910543131, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 16621 + }, + { + "epoch": 13.276357827476039, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 16622 + }, + { + "epoch": 13.277156549520766, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0185, + "step": 16623 + }, + { + "epoch": 13.277955271565496, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 16624 + }, + { + "epoch": 13.278753993610223, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 16625 + }, + { + "epoch": 13.279552715654953, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 16626 + }, + { + "epoch": 13.28035143769968, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 16627 + }, + { + "epoch": 13.281150159744408, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0161, + "step": 16628 + }, + { + "epoch": 13.281948881789138, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 16629 + }, + { + "epoch": 13.282747603833865, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0092, + "step": 16630 + }, + { + "epoch": 13.283546325878595, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0195, + "step": 16631 + }, + { + "epoch": 13.284345047923322, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 16632 + }, + { + "epoch": 13.285143769968052, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0126, + "step": 16633 + }, + { + "epoch": 13.28594249201278, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0152, + "step": 16634 + }, + { + "epoch": 13.286741214057509, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.016, + "step": 16635 + }, + { + "epoch": 13.287539936102236, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 16636 + }, + { + "epoch": 13.288338658146964, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0181, + "step": 16637 + }, + { + "epoch": 13.289137380191693, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 16638 + }, + { + "epoch": 13.289936102236421, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 16639 + }, + { + "epoch": 13.29073482428115, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 16640 + }, + { + "epoch": 13.291533546325878, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0196, + "step": 16641 + }, + { + "epoch": 13.292332268370608, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 16642 + }, + { + "epoch": 13.293130990415335, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 16643 + }, + { + "epoch": 13.293929712460065, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 16644 + }, + { + "epoch": 13.294728434504792, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 16645 + }, + { + "epoch": 13.295527156549522, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 16646 + }, + { + "epoch": 13.29632587859425, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 16647 + }, + { + "epoch": 13.297124600638977, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 16648 + }, + { + "epoch": 13.297923322683706, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0199, + "step": 16649 + }, + { + "epoch": 13.298722044728434, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0339, + "step": 16650 + }, + { + "epoch": 13.299520766773163, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 16651 + }, + { + "epoch": 13.300319488817891, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0141, + "step": 16652 + }, + { + "epoch": 13.30111821086262, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 16653 + }, + { + "epoch": 13.301916932907348, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 16654 + }, + { + "epoch": 13.302715654952078, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 16655 + }, + { + "epoch": 13.303514376996805, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 16656 + }, + { + "epoch": 13.304313099041533, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 16657 + }, + { + "epoch": 13.305111821086262, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 16658 + }, + { + "epoch": 13.30591054313099, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 16659 + }, + { + "epoch": 13.30670926517572, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0129, + "step": 16660 + }, + { + "epoch": 13.307507987220447, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0171, + "step": 16661 + }, + { + "epoch": 13.308306709265176, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 16662 + }, + { + "epoch": 13.309105431309904, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 16663 + }, + { + "epoch": 13.309904153354633, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 16664 + }, + { + "epoch": 13.310702875399361, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 16665 + }, + { + "epoch": 13.311501597444089, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 16666 + }, + { + "epoch": 13.312300319488818, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.034, + "step": 16667 + }, + { + "epoch": 13.313099041533546, + "grad_norm": 0.173828125, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 16668 + }, + { + "epoch": 13.313897763578275, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0177, + "step": 16669 + }, + { + "epoch": 13.314696485623003, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 16670 + }, + { + "epoch": 13.315495207667732, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0099, + "step": 16671 + }, + { + "epoch": 13.31629392971246, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0187, + "step": 16672 + }, + { + "epoch": 13.31709265175719, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 16673 + }, + { + "epoch": 13.317891373801917, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 16674 + }, + { + "epoch": 13.318690095846645, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0187, + "step": 16675 + }, + { + "epoch": 13.319488817891374, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 16676 + }, + { + "epoch": 13.320287539936102, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 16677 + }, + { + "epoch": 13.321086261980831, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 16678 + }, + { + "epoch": 13.321884984025559, + "grad_norm": 0.053955078125, + "learning_rate": 0.0005, + "loss": 1.0182, + "step": 16679 + }, + { + "epoch": 13.322683706070288, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0156, + "step": 16680 + }, + { + "epoch": 13.323482428115016, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 16681 + }, + { + "epoch": 13.324281150159745, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0165, + "step": 16682 + }, + { + "epoch": 13.325079872204473, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 16683 + }, + { + "epoch": 13.3258785942492, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0172, + "step": 16684 + }, + { + "epoch": 13.32667731629393, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 16685 + }, + { + "epoch": 13.327476038338657, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 16686 + }, + { + "epoch": 13.328274760383387, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 16687 + }, + { + "epoch": 13.329073482428115, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 16688 + }, + { + "epoch": 13.329872204472844, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0157, + "step": 16689 + }, + { + "epoch": 13.330670926517572, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 16690 + }, + { + "epoch": 13.331469648562301, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 16691 + }, + { + "epoch": 13.332268370607029, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 16692 + }, + { + "epoch": 13.333067092651758, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0165, + "step": 16693 + }, + { + "epoch": 13.333865814696486, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 16694 + }, + { + "epoch": 13.334664536741213, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 16695 + }, + { + "epoch": 13.335463258785943, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 16696 + }, + { + "epoch": 13.33626198083067, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 16697 + }, + { + "epoch": 13.3370607028754, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 16698 + }, + { + "epoch": 13.337859424920127, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 16699 + }, + { + "epoch": 13.338658146964857, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 16700 + }, + { + "epoch": 13.339456869009584, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 16701 + }, + { + "epoch": 13.340255591054314, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 16702 + }, + { + "epoch": 13.341054313099042, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 16703 + }, + { + "epoch": 13.34185303514377, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0168, + "step": 16704 + }, + { + "epoch": 13.342651757188499, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.018, + "step": 16705 + }, + { + "epoch": 13.343450479233226, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 16706 + }, + { + "epoch": 13.344249201277956, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 16707 + }, + { + "epoch": 13.345047923322683, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 16708 + }, + { + "epoch": 13.345846645367413, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0187, + "step": 16709 + }, + { + "epoch": 13.34664536741214, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 16710 + }, + { + "epoch": 13.34744408945687, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0109, + "step": 16711 + }, + { + "epoch": 13.348242811501597, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 16712 + }, + { + "epoch": 13.349041533546325, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 16713 + }, + { + "epoch": 13.349840255591054, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 16714 + }, + { + "epoch": 13.350638977635782, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 16715 + }, + { + "epoch": 13.351437699680512, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0161, + "step": 16716 + }, + { + "epoch": 13.35223642172524, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 16717 + }, + { + "epoch": 13.353035143769969, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 16718 + }, + { + "epoch": 13.353833865814696, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 16719 + }, + { + "epoch": 13.354632587859426, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 16720 + }, + { + "epoch": 13.355431309904153, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 16721 + }, + { + "epoch": 13.356230031948883, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 16722 + }, + { + "epoch": 13.35702875399361, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 16723 + }, + { + "epoch": 13.357827476038338, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.017, + "step": 16724 + }, + { + "epoch": 13.358626198083067, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 16725 + }, + { + "epoch": 13.359424920127795, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0176, + "step": 16726 + }, + { + "epoch": 13.360223642172524, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0181, + "step": 16727 + }, + { + "epoch": 13.361022364217252, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 16728 + }, + { + "epoch": 13.361821086261982, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 16729 + }, + { + "epoch": 13.36261980830671, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 16730 + }, + { + "epoch": 13.363418530351439, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0152, + "step": 16731 + }, + { + "epoch": 13.364217252396166, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 16732 + }, + { + "epoch": 13.365015974440894, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 16733 + }, + { + "epoch": 13.365814696485623, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 16734 + }, + { + "epoch": 13.366613418530351, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 16735 + }, + { + "epoch": 13.36741214057508, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 16736 + }, + { + "epoch": 13.368210862619808, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 16737 + }, + { + "epoch": 13.369009584664537, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 16738 + }, + { + "epoch": 13.369808306709265, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 16739 + }, + { + "epoch": 13.370607028753994, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 16740 + }, + { + "epoch": 13.371405750798722, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 16741 + }, + { + "epoch": 13.37220447284345, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 16742 + }, + { + "epoch": 13.37300319488818, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0205, + "step": 16743 + }, + { + "epoch": 13.373801916932907, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0165, + "step": 16744 + }, + { + "epoch": 13.374600638977636, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 16745 + }, + { + "epoch": 13.375399361022364, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0177, + "step": 16746 + }, + { + "epoch": 13.376198083067093, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 16747 + }, + { + "epoch": 13.37699680511182, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 16748 + }, + { + "epoch": 13.37779552715655, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 16749 + }, + { + "epoch": 13.378594249201278, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 16750 + }, + { + "epoch": 13.379392971246006, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 16751 + }, + { + "epoch": 13.380191693290735, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 16752 + }, + { + "epoch": 13.380990415335463, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 16753 + }, + { + "epoch": 13.381789137380192, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 16754 + }, + { + "epoch": 13.38258785942492, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 16755 + }, + { + "epoch": 13.383386581469649, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 16756 + }, + { + "epoch": 13.384185303514377, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 16757 + }, + { + "epoch": 13.384984025559106, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0181, + "step": 16758 + }, + { + "epoch": 13.385782747603834, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 16759 + }, + { + "epoch": 13.386581469648561, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 16760 + }, + { + "epoch": 13.38738019169329, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 16761 + }, + { + "epoch": 13.388178913738018, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0156, + "step": 16762 + }, + { + "epoch": 13.388977635782748, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0161, + "step": 16763 + }, + { + "epoch": 13.389776357827476, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 16764 + }, + { + "epoch": 13.390575079872205, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 16765 + }, + { + "epoch": 13.391373801916933, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 16766 + }, + { + "epoch": 13.392172523961662, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 16767 + }, + { + "epoch": 13.39297124600639, + "grad_norm": 0.205078125, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 16768 + }, + { + "epoch": 13.393769968051117, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 16769 + }, + { + "epoch": 13.394568690095847, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 16770 + }, + { + "epoch": 13.395367412140574, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 16771 + }, + { + "epoch": 13.396166134185304, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 16772 + }, + { + "epoch": 13.396964856230031, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0165, + "step": 16773 + }, + { + "epoch": 13.39776357827476, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 16774 + }, + { + "epoch": 13.398562300319488, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 16775 + }, + { + "epoch": 13.399361022364218, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 16776 + }, + { + "epoch": 13.400159744408946, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 16777 + }, + { + "epoch": 13.400958466453675, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 16778 + }, + { + "epoch": 13.401757188498403, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 16779 + }, + { + "epoch": 13.40255591054313, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 16780 + }, + { + "epoch": 13.40335463258786, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 16781 + }, + { + "epoch": 13.404153354632587, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0351, + "step": 16782 + }, + { + "epoch": 13.404952076677317, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 16783 + }, + { + "epoch": 13.405750798722044, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 16784 + }, + { + "epoch": 13.406549520766774, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 16785 + }, + { + "epoch": 13.407348242811501, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 16786 + }, + { + "epoch": 13.40814696485623, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0155, + "step": 16787 + }, + { + "epoch": 13.408945686900958, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 16788 + }, + { + "epoch": 13.409744408945686, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 16789 + }, + { + "epoch": 13.410543130990416, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 16790 + }, + { + "epoch": 13.411341853035143, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 16791 + }, + { + "epoch": 13.412140575079873, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 16792 + }, + { + "epoch": 13.4129392971246, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 16793 + }, + { + "epoch": 13.41373801916933, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 16794 + }, + { + "epoch": 13.414536741214057, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0181, + "step": 16795 + }, + { + "epoch": 13.415335463258787, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 16796 + }, + { + "epoch": 13.416134185303514, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 16797 + }, + { + "epoch": 13.416932907348242, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0161, + "step": 16798 + }, + { + "epoch": 13.417731629392971, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 16799 + }, + { + "epoch": 13.418530351437699, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 16800 + }, + { + "epoch": 13.419329073482428, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 16801 + }, + { + "epoch": 13.420127795527156, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 16802 + }, + { + "epoch": 13.420926517571885, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 16803 + }, + { + "epoch": 13.421725239616613, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 16804 + }, + { + "epoch": 13.422523961661343, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 16805 + }, + { + "epoch": 13.42332268370607, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 16806 + }, + { + "epoch": 13.4241214057508, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 16807 + }, + { + "epoch": 13.424920127795527, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 16808 + }, + { + "epoch": 13.425718849840255, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 16809 + }, + { + "epoch": 13.426517571884984, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0167, + "step": 16810 + }, + { + "epoch": 13.427316293929712, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 16811 + }, + { + "epoch": 13.428115015974441, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 16812 + }, + { + "epoch": 13.428913738019169, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 16813 + }, + { + "epoch": 13.429712460063898, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0173, + "step": 16814 + }, + { + "epoch": 13.430511182108626, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0168, + "step": 16815 + }, + { + "epoch": 13.431309904153355, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.018, + "step": 16816 + }, + { + "epoch": 13.432108626198083, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 16817 + }, + { + "epoch": 13.43290734824281, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 16818 + }, + { + "epoch": 13.43370607028754, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0145, + "step": 16819 + }, + { + "epoch": 13.434504792332268, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 16820 + }, + { + "epoch": 13.435303514376997, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 16821 + }, + { + "epoch": 13.436102236421725, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 16822 + }, + { + "epoch": 13.436900958466454, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0087, + "step": 16823 + }, + { + "epoch": 13.437699680511182, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 16824 + }, + { + "epoch": 13.438498402555911, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 16825 + }, + { + "epoch": 13.439297124600639, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0175, + "step": 16826 + }, + { + "epoch": 13.440095846645367, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 16827 + }, + { + "epoch": 13.440894568690096, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 16828 + }, + { + "epoch": 13.441693290734824, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 16829 + }, + { + "epoch": 13.442492012779553, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 16830 + }, + { + "epoch": 13.44329073482428, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0196, + "step": 16831 + }, + { + "epoch": 13.44408945686901, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0158, + "step": 16832 + }, + { + "epoch": 13.444888178913738, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 16833 + }, + { + "epoch": 13.445686900958467, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 16834 + }, + { + "epoch": 13.446485623003195, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 16835 + }, + { + "epoch": 13.447284345047922, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 16836 + }, + { + "epoch": 13.448083067092652, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0148, + "step": 16837 + }, + { + "epoch": 13.44888178913738, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.017, + "step": 16838 + }, + { + "epoch": 13.449680511182109, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 16839 + }, + { + "epoch": 13.450479233226837, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 16840 + }, + { + "epoch": 13.451277955271566, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 16841 + }, + { + "epoch": 13.452076677316294, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 16842 + }, + { + "epoch": 13.452875399361023, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 16843 + }, + { + "epoch": 13.45367412140575, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 16844 + }, + { + "epoch": 13.454472843450478, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 16845 + }, + { + "epoch": 13.455271565495208, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 16846 + }, + { + "epoch": 13.456070287539935, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 16847 + }, + { + "epoch": 13.456869009584665, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 16848 + }, + { + "epoch": 13.457667731629392, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 16849 + }, + { + "epoch": 13.458466453674122, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 16850 + }, + { + "epoch": 13.45926517571885, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 16851 + }, + { + "epoch": 13.460063897763579, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0125, + "step": 16852 + }, + { + "epoch": 13.460862619808307, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0181, + "step": 16853 + }, + { + "epoch": 13.461661341853036, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 16854 + }, + { + "epoch": 13.462460063897764, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 16855 + }, + { + "epoch": 13.463258785942491, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 16856 + }, + { + "epoch": 13.46405750798722, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 16857 + }, + { + "epoch": 13.464856230031948, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 16858 + }, + { + "epoch": 13.465654952076678, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 16859 + }, + { + "epoch": 13.466453674121405, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 16860 + }, + { + "epoch": 13.467252396166135, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0125, + "step": 16861 + }, + { + "epoch": 13.468051118210862, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 16862 + }, + { + "epoch": 13.468849840255592, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 16863 + }, + { + "epoch": 13.46964856230032, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0171, + "step": 16864 + }, + { + "epoch": 13.470447284345047, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 16865 + }, + { + "epoch": 13.471246006389777, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 16866 + }, + { + "epoch": 13.472044728434504, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 16867 + }, + { + "epoch": 13.472843450479234, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 16868 + }, + { + "epoch": 13.473642172523961, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 16869 + }, + { + "epoch": 13.47444089456869, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0205, + "step": 16870 + }, + { + "epoch": 13.475239616613418, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 16871 + }, + { + "epoch": 13.476038338658148, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0175, + "step": 16872 + }, + { + "epoch": 13.476837060702875, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 16873 + }, + { + "epoch": 13.477635782747603, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 16874 + }, + { + "epoch": 13.478434504792332, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 16875 + }, + { + "epoch": 13.47923322683706, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0148, + "step": 16876 + }, + { + "epoch": 13.48003194888179, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0187, + "step": 16877 + }, + { + "epoch": 13.480830670926517, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 16878 + }, + { + "epoch": 13.481629392971247, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 16879 + }, + { + "epoch": 13.482428115015974, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 16880 + }, + { + "epoch": 13.483226837060704, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 16881 + }, + { + "epoch": 13.484025559105431, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 16882 + }, + { + "epoch": 13.48482428115016, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.018, + "step": 16883 + }, + { + "epoch": 13.485623003194888, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 16884 + }, + { + "epoch": 13.486421725239616, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0127, + "step": 16885 + }, + { + "epoch": 13.487220447284345, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.013, + "step": 16886 + }, + { + "epoch": 13.488019169329073, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 16887 + }, + { + "epoch": 13.488817891373802, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 16888 + }, + { + "epoch": 13.48961661341853, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 16889 + }, + { + "epoch": 13.49041533546326, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 16890 + }, + { + "epoch": 13.491214057507987, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 16891 + }, + { + "epoch": 13.492012779552716, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 16892 + }, + { + "epoch": 13.492811501597444, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 16893 + }, + { + "epoch": 13.493610223642172, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0121, + "step": 16894 + }, + { + "epoch": 13.494408945686901, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 16895 + }, + { + "epoch": 13.495207667731629, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 16896 + }, + { + "epoch": 13.496006389776358, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.009, + "step": 16897 + }, + { + "epoch": 13.496805111821086, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 16898 + }, + { + "epoch": 13.497603833865815, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 16899 + }, + { + "epoch": 13.498402555910543, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0129, + "step": 16900 + }, + { + "epoch": 13.499201277955272, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 16901 + }, + { + "epoch": 13.5, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 16902 + }, + { + "epoch": 13.500798722044728, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0174, + "step": 16903 + }, + { + "epoch": 13.501597444089457, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0179, + "step": 16904 + }, + { + "epoch": 13.502396166134185, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0188, + "step": 16905 + }, + { + "epoch": 13.503194888178914, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0153, + "step": 16906 + }, + { + "epoch": 13.503993610223642, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 16907 + }, + { + "epoch": 13.504792332268371, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 16908 + }, + { + "epoch": 13.505591054313099, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0205, + "step": 16909 + }, + { + "epoch": 13.506389776357828, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 16910 + }, + { + "epoch": 13.507188498402556, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0181, + "step": 16911 + }, + { + "epoch": 13.507987220447284, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0171, + "step": 16912 + }, + { + "epoch": 13.508785942492013, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0195, + "step": 16913 + }, + { + "epoch": 13.50958466453674, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0196, + "step": 16914 + }, + { + "epoch": 13.51038338658147, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0154, + "step": 16915 + }, + { + "epoch": 13.511182108626198, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 16916 + }, + { + "epoch": 13.511980830670927, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 16917 + }, + { + "epoch": 13.512779552715655, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 16918 + }, + { + "epoch": 13.513578274760384, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 16919 + }, + { + "epoch": 13.514376996805112, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 16920 + }, + { + "epoch": 13.51517571884984, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0355, + "step": 16921 + }, + { + "epoch": 13.515974440894569, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0195, + "step": 16922 + }, + { + "epoch": 13.516773162939296, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 16923 + }, + { + "epoch": 13.517571884984026, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 16924 + }, + { + "epoch": 13.518370607028753, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 16925 + }, + { + "epoch": 13.519169329073483, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 16926 + }, + { + "epoch": 13.51996805111821, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0345, + "step": 16927 + }, + { + "epoch": 13.52076677316294, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 16928 + }, + { + "epoch": 13.521565495207668, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 16929 + }, + { + "epoch": 13.522364217252395, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 16930 + }, + { + "epoch": 13.523162939297125, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 16931 + }, + { + "epoch": 13.523961661341852, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 16932 + }, + { + "epoch": 13.524760383386582, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 16933 + }, + { + "epoch": 13.52555910543131, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 16934 + }, + { + "epoch": 13.526357827476039, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 16935 + }, + { + "epoch": 13.527156549520766, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 16936 + }, + { + "epoch": 13.527955271565496, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 16937 + }, + { + "epoch": 13.528753993610223, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0112, + "step": 16938 + }, + { + "epoch": 13.529552715654953, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 16939 + }, + { + "epoch": 13.53035143769968, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 16940 + }, + { + "epoch": 13.531150159744408, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 16941 + }, + { + "epoch": 13.531948881789138, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 16942 + }, + { + "epoch": 13.532747603833865, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 16943 + }, + { + "epoch": 13.533546325878595, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 16944 + }, + { + "epoch": 13.534345047923322, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0127, + "step": 16945 + }, + { + "epoch": 13.535143769968052, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 16946 + }, + { + "epoch": 13.53594249201278, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 16947 + }, + { + "epoch": 13.536741214057509, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 16948 + }, + { + "epoch": 13.537539936102236, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 16949 + }, + { + "epoch": 13.538338658146966, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 16950 + }, + { + "epoch": 13.539137380191693, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.014, + "step": 16951 + }, + { + "epoch": 13.539936102236421, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 16952 + }, + { + "epoch": 13.54073482428115, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 16953 + }, + { + "epoch": 13.541533546325878, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 16954 + }, + { + "epoch": 13.542332268370608, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0168, + "step": 16955 + }, + { + "epoch": 13.543130990415335, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0195, + "step": 16956 + }, + { + "epoch": 13.543929712460065, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 16957 + }, + { + "epoch": 13.544728434504792, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 16958 + }, + { + "epoch": 13.545527156549522, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 16959 + }, + { + "epoch": 13.54632587859425, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0138, + "step": 16960 + }, + { + "epoch": 13.547124600638977, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 16961 + }, + { + "epoch": 13.547923322683706, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 16962 + }, + { + "epoch": 13.548722044728434, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0162, + "step": 16963 + }, + { + "epoch": 13.549520766773163, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 16964 + }, + { + "epoch": 13.550319488817891, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 16965 + }, + { + "epoch": 13.55111821086262, + "grad_norm": 0.177734375, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 16966 + }, + { + "epoch": 13.551916932907348, + "grad_norm": 0.1767578125, + "learning_rate": 0.0005, + "loss": 1.0164, + "step": 16967 + }, + { + "epoch": 13.552715654952078, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0088, + "step": 16968 + }, + { + "epoch": 13.553514376996805, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 16969 + }, + { + "epoch": 13.554313099041533, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 16970 + }, + { + "epoch": 13.555111821086262, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 16971 + }, + { + "epoch": 13.55591054313099, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 16972 + }, + { + "epoch": 13.55670926517572, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 16973 + }, + { + "epoch": 13.557507987220447, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 16974 + }, + { + "epoch": 13.558306709265176, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 16975 + }, + { + "epoch": 13.559105431309904, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 16976 + }, + { + "epoch": 13.559904153354633, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0173, + "step": 16977 + }, + { + "epoch": 13.560702875399361, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 16978 + }, + { + "epoch": 13.561501597444089, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 16979 + }, + { + "epoch": 13.562300319488818, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0154, + "step": 16980 + }, + { + "epoch": 13.563099041533546, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0175, + "step": 16981 + }, + { + "epoch": 13.563897763578275, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 16982 + }, + { + "epoch": 13.564696485623003, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 16983 + }, + { + "epoch": 13.565495207667732, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 16984 + }, + { + "epoch": 13.56629392971246, + "grad_norm": 0.169921875, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 16985 + }, + { + "epoch": 13.56709265175719, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 16986 + }, + { + "epoch": 13.567891373801917, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.0177, + "step": 16987 + }, + { + "epoch": 13.568690095846645, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 16988 + }, + { + "epoch": 13.569488817891374, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 16989 + }, + { + "epoch": 13.570287539936102, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 16990 + }, + { + "epoch": 13.571086261980831, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 16991 + }, + { + "epoch": 13.571884984025559, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 16992 + }, + { + "epoch": 13.572683706070288, + "grad_norm": 0.11767578125, + "learning_rate": 0.0005, + "loss": 1.0157, + "step": 16993 + }, + { + "epoch": 13.573482428115016, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 16994 + }, + { + "epoch": 13.574281150159745, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0135, + "step": 16995 + }, + { + "epoch": 13.575079872204473, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 16996 + }, + { + "epoch": 13.5758785942492, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0145, + "step": 16997 + }, + { + "epoch": 13.57667731629393, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 16998 + }, + { + "epoch": 13.577476038338657, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 16999 + }, + { + "epoch": 13.578274760383387, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 17000 + }, + { + "epoch": 13.579073482428115, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0195, + "step": 17001 + }, + { + "epoch": 13.579872204472844, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0168, + "step": 17002 + }, + { + "epoch": 13.580670926517572, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0132, + "step": 17003 + }, + { + "epoch": 13.581469648562301, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0187, + "step": 17004 + }, + { + "epoch": 13.582268370607029, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 17005 + }, + { + "epoch": 13.583067092651756, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 17006 + }, + { + "epoch": 13.583865814696486, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0169, + "step": 17007 + }, + { + "epoch": 13.584664536741213, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 17008 + }, + { + "epoch": 13.585463258785943, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 17009 + }, + { + "epoch": 13.58626198083067, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 17010 + }, + { + "epoch": 13.5870607028754, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0205, + "step": 17011 + }, + { + "epoch": 13.587859424920127, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.018, + "step": 17012 + }, + { + "epoch": 13.588658146964857, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0177, + "step": 17013 + }, + { + "epoch": 13.589456869009584, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 17014 + }, + { + "epoch": 13.590255591054314, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 17015 + }, + { + "epoch": 13.591054313099042, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 17016 + }, + { + "epoch": 13.59185303514377, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 17017 + }, + { + "epoch": 13.592651757188499, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 17018 + }, + { + "epoch": 13.593450479233226, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 17019 + }, + { + "epoch": 13.594249201277956, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 17020 + }, + { + "epoch": 13.595047923322683, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 17021 + }, + { + "epoch": 13.595846645367413, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0199, + "step": 17022 + }, + { + "epoch": 13.59664536741214, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 17023 + }, + { + "epoch": 13.59744408945687, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 17024 + }, + { + "epoch": 13.598242811501597, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 17025 + }, + { + "epoch": 13.599041533546325, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 17026 + }, + { + "epoch": 13.599840255591054, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 17027 + }, + { + "epoch": 13.600638977635782, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 17028 + }, + { + "epoch": 13.601437699680512, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0154, + "step": 17029 + }, + { + "epoch": 13.60223642172524, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 17030 + }, + { + "epoch": 13.603035143769969, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 17031 + }, + { + "epoch": 13.603833865814696, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 17032 + }, + { + "epoch": 13.604632587859426, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 17033 + }, + { + "epoch": 13.605431309904153, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 17034 + }, + { + "epoch": 13.606230031948883, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 17035 + }, + { + "epoch": 13.60702875399361, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.018, + "step": 17036 + }, + { + "epoch": 13.607827476038338, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0143, + "step": 17037 + }, + { + "epoch": 13.608626198083067, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 17038 + }, + { + "epoch": 13.609424920127795, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 17039 + }, + { + "epoch": 13.610223642172524, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 17040 + }, + { + "epoch": 13.611022364217252, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 17041 + }, + { + "epoch": 13.611821086261982, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0139, + "step": 17042 + }, + { + "epoch": 13.61261980830671, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 17043 + }, + { + "epoch": 13.613418530351439, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0195, + "step": 17044 + }, + { + "epoch": 13.614217252396166, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 17045 + }, + { + "epoch": 13.615015974440894, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 17046 + }, + { + "epoch": 13.615814696485623, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0199, + "step": 17047 + }, + { + "epoch": 13.616613418530351, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0105, + "step": 17048 + }, + { + "epoch": 13.61741214057508, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0141, + "step": 17049 + }, + { + "epoch": 13.618210862619808, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0176, + "step": 17050 + }, + { + "epoch": 13.619009584664537, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 17051 + }, + { + "epoch": 13.619808306709265, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 17052 + }, + { + "epoch": 13.620607028753994, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 17053 + }, + { + "epoch": 13.621405750798722, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 17054 + }, + { + "epoch": 13.62220447284345, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 17055 + }, + { + "epoch": 13.62300319488818, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 17056 + }, + { + "epoch": 13.623801916932907, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 17057 + }, + { + "epoch": 13.624600638977636, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 17058 + }, + { + "epoch": 13.625399361022364, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0182, + "step": 17059 + }, + { + "epoch": 13.626198083067093, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 17060 + }, + { + "epoch": 13.62699680511182, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 17061 + }, + { + "epoch": 13.62779552715655, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 17062 + }, + { + "epoch": 13.628594249201278, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0174, + "step": 17063 + }, + { + "epoch": 13.629392971246006, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 17064 + }, + { + "epoch": 13.630191693290735, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 17065 + }, + { + "epoch": 13.630990415335463, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 17066 + }, + { + "epoch": 13.631789137380192, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 17067 + }, + { + "epoch": 13.63258785942492, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0335, + "step": 17068 + }, + { + "epoch": 13.633386581469649, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0182, + "step": 17069 + }, + { + "epoch": 13.634185303514377, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 17070 + }, + { + "epoch": 13.634984025559106, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 17071 + }, + { + "epoch": 13.635782747603834, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0195, + "step": 17072 + }, + { + "epoch": 13.636581469648561, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 17073 + }, + { + "epoch": 13.63738019169329, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 17074 + }, + { + "epoch": 13.638178913738018, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0171, + "step": 17075 + }, + { + "epoch": 13.638977635782748, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 17076 + }, + { + "epoch": 13.639776357827476, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0149, + "step": 17077 + }, + { + "epoch": 13.640575079872205, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 17078 + }, + { + "epoch": 13.641373801916933, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0179, + "step": 17079 + }, + { + "epoch": 13.642172523961662, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0146, + "step": 17080 + }, + { + "epoch": 13.64297124600639, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 17081 + }, + { + "epoch": 13.643769968051117, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 17082 + }, + { + "epoch": 13.644568690095847, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 17083 + }, + { + "epoch": 13.645367412140574, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 17084 + }, + { + "epoch": 13.646166134185304, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 17085 + }, + { + "epoch": 13.646964856230031, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 17086 + }, + { + "epoch": 13.64776357827476, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.018, + "step": 17087 + }, + { + "epoch": 13.648562300319488, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0169, + "step": 17088 + }, + { + "epoch": 13.649361022364218, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 17089 + }, + { + "epoch": 13.650159744408946, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 17090 + }, + { + "epoch": 13.650958466453675, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 17091 + }, + { + "epoch": 13.651757188498403, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0136, + "step": 17092 + }, + { + "epoch": 13.65255591054313, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 17093 + }, + { + "epoch": 13.65335463258786, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 17094 + }, + { + "epoch": 13.654153354632587, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 17095 + }, + { + "epoch": 13.654952076677317, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 17096 + }, + { + "epoch": 13.655750798722044, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0105, + "step": 17097 + }, + { + "epoch": 13.656549520766774, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0143, + "step": 17098 + }, + { + "epoch": 13.657348242811501, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 17099 + }, + { + "epoch": 13.65814696485623, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 17100 + }, + { + "epoch": 13.658945686900958, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0147, + "step": 17101 + }, + { + "epoch": 13.659744408945686, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0168, + "step": 17102 + }, + { + "epoch": 13.660543130990416, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0171, + "step": 17103 + }, + { + "epoch": 13.661341853035143, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 17104 + }, + { + "epoch": 13.662140575079873, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0175, + "step": 17105 + }, + { + "epoch": 13.6629392971246, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 17106 + }, + { + "epoch": 13.66373801916933, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0171, + "step": 17107 + }, + { + "epoch": 13.664536741214057, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 17108 + }, + { + "epoch": 13.665335463258787, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0182, + "step": 17109 + }, + { + "epoch": 13.666134185303514, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 17110 + }, + { + "epoch": 13.666932907348244, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0343, + "step": 17111 + }, + { + "epoch": 13.667731629392971, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 17112 + }, + { + "epoch": 13.668530351437699, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 17113 + }, + { + "epoch": 13.669329073482428, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0199, + "step": 17114 + }, + { + "epoch": 13.670127795527156, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 17115 + }, + { + "epoch": 13.670926517571885, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0192, + "step": 17116 + }, + { + "epoch": 13.671725239616613, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 17117 + }, + { + "epoch": 13.672523961661343, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 17118 + }, + { + "epoch": 13.67332268370607, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 17119 + }, + { + "epoch": 13.6741214057508, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0195, + "step": 17120 + }, + { + "epoch": 13.674920127795527, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 17121 + }, + { + "epoch": 13.675718849840255, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0174, + "step": 17122 + }, + { + "epoch": 13.676517571884984, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0117, + "step": 17123 + }, + { + "epoch": 13.677316293929712, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 17124 + }, + { + "epoch": 13.678115015974441, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 17125 + }, + { + "epoch": 13.678913738019169, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 17126 + }, + { + "epoch": 13.679712460063898, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 17127 + }, + { + "epoch": 13.680511182108626, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 17128 + }, + { + "epoch": 13.681309904153355, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 17129 + }, + { + "epoch": 13.682108626198083, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 17130 + }, + { + "epoch": 13.68290734824281, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 17131 + }, + { + "epoch": 13.68370607028754, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 17132 + }, + { + "epoch": 13.684504792332268, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 17133 + }, + { + "epoch": 13.685303514376997, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 17134 + }, + { + "epoch": 13.686102236421725, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0156, + "step": 17135 + }, + { + "epoch": 13.686900958466454, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 17136 + }, + { + "epoch": 13.687699680511182, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 17137 + }, + { + "epoch": 13.688498402555911, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 17138 + }, + { + "epoch": 13.689297124600639, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 17139 + }, + { + "epoch": 13.690095846645367, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 17140 + }, + { + "epoch": 13.690894568690096, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0176, + "step": 17141 + }, + { + "epoch": 13.691693290734824, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 17142 + }, + { + "epoch": 13.692492012779553, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 17143 + }, + { + "epoch": 13.69329073482428, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 17144 + }, + { + "epoch": 13.69408945686901, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.0175, + "step": 17145 + }, + { + "epoch": 13.694888178913738, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 17146 + }, + { + "epoch": 13.695686900958467, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 17147 + }, + { + "epoch": 13.696485623003195, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.01, + "step": 17148 + }, + { + "epoch": 13.697284345047922, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 17149 + }, + { + "epoch": 13.698083067092652, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0176, + "step": 17150 + }, + { + "epoch": 13.69888178913738, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 17151 + }, + { + "epoch": 13.699680511182109, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 17152 + }, + { + "epoch": 13.700479233226837, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 17153 + }, + { + "epoch": 13.701277955271566, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 17154 + }, + { + "epoch": 13.702076677316294, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 17155 + }, + { + "epoch": 13.702875399361023, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 17156 + }, + { + "epoch": 13.70367412140575, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0196, + "step": 17157 + }, + { + "epoch": 13.704472843450478, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0098, + "step": 17158 + }, + { + "epoch": 13.705271565495208, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 17159 + }, + { + "epoch": 13.706070287539935, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.018, + "step": 17160 + }, + { + "epoch": 13.706869009584665, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 17161 + }, + { + "epoch": 13.707667731629392, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 17162 + }, + { + "epoch": 13.708466453674122, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 17163 + }, + { + "epoch": 13.70926517571885, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0316, + "step": 17164 + }, + { + "epoch": 13.710063897763579, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.016, + "step": 17165 + }, + { + "epoch": 13.710862619808307, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 17166 + }, + { + "epoch": 13.711661341853034, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 17167 + }, + { + "epoch": 13.712460063897764, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 17168 + }, + { + "epoch": 13.713258785942491, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 17169 + }, + { + "epoch": 13.71405750798722, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0168, + "step": 17170 + }, + { + "epoch": 13.714856230031948, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0148, + "step": 17171 + }, + { + "epoch": 13.715654952076678, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 17172 + }, + { + "epoch": 13.716453674121405, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 17173 + }, + { + "epoch": 13.717252396166135, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 17174 + }, + { + "epoch": 13.718051118210862, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0139, + "step": 17175 + }, + { + "epoch": 13.718849840255592, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 17176 + }, + { + "epoch": 13.71964856230032, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 17177 + }, + { + "epoch": 13.720447284345047, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 17178 + }, + { + "epoch": 13.721246006389777, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0205, + "step": 17179 + }, + { + "epoch": 13.722044728434504, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 17180 + }, + { + "epoch": 13.722843450479234, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 17181 + }, + { + "epoch": 13.723642172523961, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0188, + "step": 17182 + }, + { + "epoch": 13.72444089456869, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 17183 + }, + { + "epoch": 13.725239616613418, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 17184 + }, + { + "epoch": 13.726038338658148, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0162, + "step": 17185 + }, + { + "epoch": 13.726837060702875, + "grad_norm": 0.171875, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 17186 + }, + { + "epoch": 13.727635782747605, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 17187 + }, + { + "epoch": 13.728434504792332, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 17188 + }, + { + "epoch": 13.72923322683706, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 17189 + }, + { + "epoch": 13.73003194888179, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0172, + "step": 17190 + }, + { + "epoch": 13.730830670926517, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 17191 + }, + { + "epoch": 13.731629392971247, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 17192 + }, + { + "epoch": 13.732428115015974, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0163, + "step": 17193 + }, + { + "epoch": 13.733226837060704, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 17194 + }, + { + "epoch": 13.734025559105431, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 17195 + }, + { + "epoch": 13.73482428115016, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 17196 + }, + { + "epoch": 13.735623003194888, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 17197 + }, + { + "epoch": 13.736421725239616, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 17198 + }, + { + "epoch": 13.737220447284345, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 17199 + }, + { + "epoch": 13.738019169329073, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 17200 + }, + { + "epoch": 13.738817891373802, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 17201 + }, + { + "epoch": 13.73961661341853, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 17202 + }, + { + "epoch": 13.74041533546326, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 17203 + }, + { + "epoch": 13.741214057507987, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0156, + "step": 17204 + }, + { + "epoch": 13.742012779552716, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0139, + "step": 17205 + }, + { + "epoch": 13.742811501597444, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 17206 + }, + { + "epoch": 13.743610223642172, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 17207 + }, + { + "epoch": 13.744408945686901, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 17208 + }, + { + "epoch": 13.745207667731629, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0157, + "step": 17209 + }, + { + "epoch": 13.746006389776358, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 17210 + }, + { + "epoch": 13.746805111821086, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0179, + "step": 17211 + }, + { + "epoch": 13.747603833865815, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0159, + "step": 17212 + }, + { + "epoch": 13.748402555910543, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 17213 + }, + { + "epoch": 13.749201277955272, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 17214 + }, + { + "epoch": 13.75, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 17215 + }, + { + "epoch": 13.750798722044728, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0132, + "step": 17216 + }, + { + "epoch": 13.751597444089457, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 17217 + }, + { + "epoch": 13.752396166134185, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 17218 + }, + { + "epoch": 13.753194888178914, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0129, + "step": 17219 + }, + { + "epoch": 13.753993610223642, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 17220 + }, + { + "epoch": 13.754792332268371, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 17221 + }, + { + "epoch": 13.755591054313099, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 17222 + }, + { + "epoch": 13.756389776357828, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 17223 + }, + { + "epoch": 13.757188498402556, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 17224 + }, + { + "epoch": 13.757987220447284, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0188, + "step": 17225 + }, + { + "epoch": 13.758785942492013, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 17226 + }, + { + "epoch": 13.75958466453674, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 17227 + }, + { + "epoch": 13.76038338658147, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 17228 + }, + { + "epoch": 13.761182108626198, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0172, + "step": 17229 + }, + { + "epoch": 13.761980830670927, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 17230 + }, + { + "epoch": 13.762779552715655, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 17231 + }, + { + "epoch": 13.763578274760384, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.012, + "step": 17232 + }, + { + "epoch": 13.764376996805112, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 17233 + }, + { + "epoch": 13.76517571884984, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.016, + "step": 17234 + }, + { + "epoch": 13.765974440894569, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 17235 + }, + { + "epoch": 13.766773162939296, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 17236 + }, + { + "epoch": 13.767571884984026, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 17237 + }, + { + "epoch": 13.768370607028753, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 17238 + }, + { + "epoch": 13.769169329073483, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 17239 + }, + { + "epoch": 13.76996805111821, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.014, + "step": 17240 + }, + { + "epoch": 13.77076677316294, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 17241 + }, + { + "epoch": 13.771565495207668, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 17242 + }, + { + "epoch": 13.772364217252395, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 17243 + }, + { + "epoch": 13.773162939297125, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 17244 + }, + { + "epoch": 13.773961661341852, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 17245 + }, + { + "epoch": 13.774760383386582, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 17246 + }, + { + "epoch": 13.77555910543131, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 17247 + }, + { + "epoch": 13.776357827476039, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 17248 + }, + { + "epoch": 13.777156549520766, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0205, + "step": 17249 + }, + { + "epoch": 13.777955271565496, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 17250 + }, + { + "epoch": 13.778753993610223, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0308, + "step": 17251 + }, + { + "epoch": 13.779552715654953, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 17252 + }, + { + "epoch": 13.78035143769968, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 17253 + }, + { + "epoch": 13.781150159744408, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0328, + "step": 17254 + }, + { + "epoch": 13.781948881789138, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 17255 + }, + { + "epoch": 13.782747603833865, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 17256 + }, + { + "epoch": 13.783546325878595, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0169, + "step": 17257 + }, + { + "epoch": 13.784345047923322, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 17258 + }, + { + "epoch": 13.785143769968052, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 17259 + }, + { + "epoch": 13.78594249201278, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 17260 + }, + { + "epoch": 13.786741214057509, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 17261 + }, + { + "epoch": 13.787539936102236, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 17262 + }, + { + "epoch": 13.788338658146966, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0168, + "step": 17263 + }, + { + "epoch": 13.789137380191693, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 17264 + }, + { + "epoch": 13.789936102236421, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 17265 + }, + { + "epoch": 13.79073482428115, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 17266 + }, + { + "epoch": 13.791533546325878, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 17267 + }, + { + "epoch": 13.792332268370608, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 17268 + }, + { + "epoch": 13.793130990415335, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 17269 + }, + { + "epoch": 13.793929712460065, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 17270 + }, + { + "epoch": 13.794728434504792, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 17271 + }, + { + "epoch": 13.795527156549522, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0169, + "step": 17272 + }, + { + "epoch": 13.79632587859425, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 17273 + }, + { + "epoch": 13.797124600638977, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0187, + "step": 17274 + }, + { + "epoch": 13.797923322683706, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 17275 + }, + { + "epoch": 13.798722044728434, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 17276 + }, + { + "epoch": 13.799520766773163, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 17277 + }, + { + "epoch": 13.800319488817891, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 17278 + }, + { + "epoch": 13.80111821086262, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 17279 + }, + { + "epoch": 13.801916932907348, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0134, + "step": 17280 + }, + { + "epoch": 13.802715654952078, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 17281 + }, + { + "epoch": 13.803514376996805, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0195, + "step": 17282 + }, + { + "epoch": 13.804313099041533, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 17283 + }, + { + "epoch": 13.805111821086262, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0142, + "step": 17284 + }, + { + "epoch": 13.80591054313099, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0182, + "step": 17285 + }, + { + "epoch": 13.80670926517572, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 17286 + }, + { + "epoch": 13.807507987220447, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 17287 + }, + { + "epoch": 13.808306709265176, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0167, + "step": 17288 + }, + { + "epoch": 13.809105431309904, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 17289 + }, + { + "epoch": 13.809904153354633, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 17290 + }, + { + "epoch": 13.810702875399361, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 17291 + }, + { + "epoch": 13.811501597444089, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 17292 + }, + { + "epoch": 13.812300319488818, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0176, + "step": 17293 + }, + { + "epoch": 13.813099041533546, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0171, + "step": 17294 + }, + { + "epoch": 13.813897763578275, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 17295 + }, + { + "epoch": 13.814696485623003, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 17296 + }, + { + "epoch": 13.815495207667732, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0178, + "step": 17297 + }, + { + "epoch": 13.81629392971246, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0153, + "step": 17298 + }, + { + "epoch": 13.81709265175719, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 17299 + }, + { + "epoch": 13.817891373801917, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 17300 + }, + { + "epoch": 13.818690095846645, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0174, + "step": 17301 + }, + { + "epoch": 13.819488817891374, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 17302 + }, + { + "epoch": 13.820287539936102, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0165, + "step": 17303 + }, + { + "epoch": 13.821086261980831, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0172, + "step": 17304 + }, + { + "epoch": 13.821884984025559, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 17305 + }, + { + "epoch": 13.822683706070288, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 17306 + }, + { + "epoch": 13.823482428115016, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 17307 + }, + { + "epoch": 13.824281150159745, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 17308 + }, + { + "epoch": 13.825079872204473, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 17309 + }, + { + "epoch": 13.8258785942492, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 17310 + }, + { + "epoch": 13.82667731629393, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 17311 + }, + { + "epoch": 13.827476038338657, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.018, + "step": 17312 + }, + { + "epoch": 13.828274760383387, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0381, + "step": 17313 + }, + { + "epoch": 13.829073482428115, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 17314 + }, + { + "epoch": 13.829872204472844, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 17315 + }, + { + "epoch": 13.830670926517572, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 17316 + }, + { + "epoch": 13.831469648562301, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 17317 + }, + { + "epoch": 13.832268370607029, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 17318 + }, + { + "epoch": 13.833067092651756, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 17319 + }, + { + "epoch": 13.833865814696486, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 17320 + }, + { + "epoch": 13.834664536741213, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 17321 + }, + { + "epoch": 13.835463258785943, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 17322 + }, + { + "epoch": 13.83626198083067, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 17323 + }, + { + "epoch": 13.8370607028754, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 17324 + }, + { + "epoch": 13.837859424920127, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 17325 + }, + { + "epoch": 13.838658146964857, + "grad_norm": 0.1240234375, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 17326 + }, + { + "epoch": 13.839456869009584, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0165, + "step": 17327 + }, + { + "epoch": 13.840255591054314, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 17328 + }, + { + "epoch": 13.841054313099042, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0178, + "step": 17329 + }, + { + "epoch": 13.84185303514377, + "grad_norm": 0.185546875, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 17330 + }, + { + "epoch": 13.842651757188499, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 17331 + }, + { + "epoch": 13.843450479233226, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 17332 + }, + { + "epoch": 13.844249201277956, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 17333 + }, + { + "epoch": 13.845047923322683, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0205, + "step": 17334 + }, + { + "epoch": 13.845846645367413, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 17335 + }, + { + "epoch": 13.84664536741214, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 17336 + }, + { + "epoch": 13.84744408945687, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0158, + "step": 17337 + }, + { + "epoch": 13.848242811501597, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0188, + "step": 17338 + }, + { + "epoch": 13.849041533546325, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 17339 + }, + { + "epoch": 13.849840255591054, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 17340 + }, + { + "epoch": 13.850638977635782, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0166, + "step": 17341 + }, + { + "epoch": 13.851437699680512, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 17342 + }, + { + "epoch": 13.85223642172524, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 17343 + }, + { + "epoch": 13.853035143769969, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0152, + "step": 17344 + }, + { + "epoch": 13.853833865814696, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0142, + "step": 17345 + }, + { + "epoch": 13.854632587859426, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 17346 + }, + { + "epoch": 13.855431309904153, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0205, + "step": 17347 + }, + { + "epoch": 13.856230031948883, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0159, + "step": 17348 + }, + { + "epoch": 13.85702875399361, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 17349 + }, + { + "epoch": 13.857827476038338, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 17350 + }, + { + "epoch": 13.858626198083067, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 17351 + }, + { + "epoch": 13.859424920127795, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 17352 + }, + { + "epoch": 13.860223642172524, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 17353 + }, + { + "epoch": 13.861022364217252, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 17354 + }, + { + "epoch": 13.861821086261982, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 17355 + }, + { + "epoch": 13.86261980830671, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0161, + "step": 17356 + }, + { + "epoch": 13.863418530351439, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 17357 + }, + { + "epoch": 13.864217252396166, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0163, + "step": 17358 + }, + { + "epoch": 13.865015974440894, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 17359 + }, + { + "epoch": 13.865814696485623, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 17360 + }, + { + "epoch": 13.866613418530351, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0159, + "step": 17361 + }, + { + "epoch": 13.86741214057508, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0175, + "step": 17362 + }, + { + "epoch": 13.868210862619808, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 17363 + }, + { + "epoch": 13.869009584664537, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 17364 + }, + { + "epoch": 13.869808306709265, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0179, + "step": 17365 + }, + { + "epoch": 13.870607028753994, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0155, + "step": 17366 + }, + { + "epoch": 13.871405750798722, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 17367 + }, + { + "epoch": 13.87220447284345, + "grad_norm": 0.1728515625, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 17368 + }, + { + "epoch": 13.87300319488818, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0119, + "step": 17369 + }, + { + "epoch": 13.873801916932907, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0163, + "step": 17370 + }, + { + "epoch": 13.874600638977636, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0182, + "step": 17371 + }, + { + "epoch": 13.875399361022364, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0196, + "step": 17372 + }, + { + "epoch": 13.876198083067093, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 17373 + }, + { + "epoch": 13.87699680511182, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 17374 + }, + { + "epoch": 13.87779552715655, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 17375 + }, + { + "epoch": 13.878594249201278, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.016, + "step": 17376 + }, + { + "epoch": 13.879392971246006, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 17377 + }, + { + "epoch": 13.880191693290735, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 17378 + }, + { + "epoch": 13.880990415335463, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0161, + "step": 17379 + }, + { + "epoch": 13.881789137380192, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0117, + "step": 17380 + }, + { + "epoch": 13.88258785942492, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0182, + "step": 17381 + }, + { + "epoch": 13.883386581469649, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 17382 + }, + { + "epoch": 13.884185303514377, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0145, + "step": 17383 + }, + { + "epoch": 13.884984025559106, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 17384 + }, + { + "epoch": 13.885782747603834, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 17385 + }, + { + "epoch": 13.886581469648561, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 17386 + }, + { + "epoch": 13.88738019169329, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0155, + "step": 17387 + }, + { + "epoch": 13.888178913738018, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0167, + "step": 17388 + }, + { + "epoch": 13.888977635782748, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 17389 + }, + { + "epoch": 13.889776357827476, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 17390 + }, + { + "epoch": 13.890575079872205, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 17391 + }, + { + "epoch": 13.891373801916933, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 17392 + }, + { + "epoch": 13.892172523961662, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 17393 + }, + { + "epoch": 13.89297124600639, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0199, + "step": 17394 + }, + { + "epoch": 13.893769968051117, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 17395 + }, + { + "epoch": 13.894568690095847, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0192, + "step": 17396 + }, + { + "epoch": 13.895367412140574, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 17397 + }, + { + "epoch": 13.896166134185304, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 17398 + }, + { + "epoch": 13.896964856230031, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 17399 + }, + { + "epoch": 13.89776357827476, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 17400 + }, + { + "epoch": 13.898562300319488, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 17401 + }, + { + "epoch": 13.899361022364218, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 17402 + }, + { + "epoch": 13.900159744408946, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 17403 + }, + { + "epoch": 13.900958466453675, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 17404 + }, + { + "epoch": 13.901757188498403, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.012, + "step": 17405 + }, + { + "epoch": 13.90255591054313, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 17406 + }, + { + "epoch": 13.90335463258786, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0192, + "step": 17407 + }, + { + "epoch": 13.904153354632587, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 17408 + }, + { + "epoch": 13.904952076677317, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 17409 + }, + { + "epoch": 13.905750798722044, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 17410 + }, + { + "epoch": 13.906549520766774, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 17411 + }, + { + "epoch": 13.907348242811501, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0116, + "step": 17412 + }, + { + "epoch": 13.90814696485623, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 17413 + }, + { + "epoch": 13.908945686900958, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0168, + "step": 17414 + }, + { + "epoch": 13.909744408945686, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 17415 + }, + { + "epoch": 13.910543130990416, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 17416 + }, + { + "epoch": 13.911341853035143, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0178, + "step": 17417 + }, + { + "epoch": 13.912140575079873, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 17418 + }, + { + "epoch": 13.9129392971246, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 17419 + }, + { + "epoch": 13.91373801916933, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 17420 + }, + { + "epoch": 13.914536741214057, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0148, + "step": 17421 + }, + { + "epoch": 13.915335463258787, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 17422 + }, + { + "epoch": 13.916134185303514, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 17423 + }, + { + "epoch": 13.916932907348244, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 17424 + }, + { + "epoch": 13.917731629392971, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 17425 + }, + { + "epoch": 13.918530351437699, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 17426 + }, + { + "epoch": 13.919329073482428, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0143, + "step": 17427 + }, + { + "epoch": 13.920127795527156, + "grad_norm": 0.1591796875, + "learning_rate": 0.0005, + "loss": 1.0183, + "step": 17428 + }, + { + "epoch": 13.920926517571885, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 17429 + }, + { + "epoch": 13.921725239616613, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 17430 + }, + { + "epoch": 13.922523961661343, + "grad_norm": 0.16015625, + "learning_rate": 0.0005, + "loss": 1.0175, + "step": 17431 + }, + { + "epoch": 13.92332268370607, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 17432 + }, + { + "epoch": 13.9241214057508, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0172, + "step": 17433 + }, + { + "epoch": 13.924920127795527, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 17434 + }, + { + "epoch": 13.925718849840255, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0101, + "step": 17435 + }, + { + "epoch": 13.926517571884984, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 17436 + }, + { + "epoch": 13.927316293929712, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0147, + "step": 17437 + }, + { + "epoch": 13.928115015974441, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 17438 + }, + { + "epoch": 13.928913738019169, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0161, + "step": 17439 + }, + { + "epoch": 13.929712460063898, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 17440 + }, + { + "epoch": 13.930511182108626, + "grad_norm": 0.1337890625, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 17441 + }, + { + "epoch": 13.931309904153355, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 17442 + }, + { + "epoch": 13.932108626198083, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 17443 + }, + { + "epoch": 13.93290734824281, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 17444 + }, + { + "epoch": 13.93370607028754, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 17445 + }, + { + "epoch": 13.934504792332268, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 17446 + }, + { + "epoch": 13.935303514376997, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 17447 + }, + { + "epoch": 13.936102236421725, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 17448 + }, + { + "epoch": 13.936900958466454, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 17449 + }, + { + "epoch": 13.937699680511182, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 17450 + }, + { + "epoch": 13.938498402555911, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 17451 + }, + { + "epoch": 13.939297124600639, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 17452 + }, + { + "epoch": 13.940095846645367, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 17453 + }, + { + "epoch": 13.940894568690096, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0171, + "step": 17454 + }, + { + "epoch": 13.941693290734824, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0164, + "step": 17455 + }, + { + "epoch": 13.942492012779553, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 17456 + }, + { + "epoch": 13.94329073482428, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 17457 + }, + { + "epoch": 13.94408945686901, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 17458 + }, + { + "epoch": 13.944888178913738, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 17459 + }, + { + "epoch": 13.945686900958467, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 17460 + }, + { + "epoch": 13.946485623003195, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 17461 + }, + { + "epoch": 13.947284345047922, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 17462 + }, + { + "epoch": 13.948083067092652, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 17463 + }, + { + "epoch": 13.94888178913738, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 17464 + }, + { + "epoch": 13.949680511182109, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 17465 + }, + { + "epoch": 13.950479233226837, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0162, + "step": 17466 + }, + { + "epoch": 13.951277955271566, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 17467 + }, + { + "epoch": 13.952076677316294, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 17468 + }, + { + "epoch": 13.952875399361023, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 17469 + }, + { + "epoch": 13.95367412140575, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0121, + "step": 17470 + }, + { + "epoch": 13.954472843450478, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0158, + "step": 17471 + }, + { + "epoch": 13.955271565495208, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 17472 + }, + { + "epoch": 13.956070287539935, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 17473 + }, + { + "epoch": 13.956869009584665, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 17474 + }, + { + "epoch": 13.957667731629392, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 17475 + }, + { + "epoch": 13.958466453674122, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 17476 + }, + { + "epoch": 13.95926517571885, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 17477 + }, + { + "epoch": 13.960063897763579, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 17478 + }, + { + "epoch": 13.960862619808307, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 17479 + }, + { + "epoch": 13.961661341853034, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 17480 + }, + { + "epoch": 13.962460063897764, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0146, + "step": 17481 + }, + { + "epoch": 13.963258785942491, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 17482 + }, + { + "epoch": 13.96405750798722, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 17483 + }, + { + "epoch": 13.964856230031948, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 17484 + }, + { + "epoch": 13.965654952076678, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 17485 + }, + { + "epoch": 13.966453674121405, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0137, + "step": 17486 + }, + { + "epoch": 13.967252396166135, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 17487 + }, + { + "epoch": 13.968051118210862, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 17488 + }, + { + "epoch": 13.968849840255592, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 17489 + }, + { + "epoch": 13.96964856230032, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 17490 + }, + { + "epoch": 13.970447284345047, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 17491 + }, + { + "epoch": 13.971246006389777, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.011, + "step": 17492 + }, + { + "epoch": 13.972044728434504, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0321, + "step": 17493 + }, + { + "epoch": 13.972843450479234, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 17494 + }, + { + "epoch": 13.973642172523961, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0366, + "step": 17495 + }, + { + "epoch": 13.97444089456869, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0155, + "step": 17496 + }, + { + "epoch": 13.975239616613418, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 17497 + }, + { + "epoch": 13.976038338658148, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 17498 + }, + { + "epoch": 13.976837060702875, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 17499 + }, + { + "epoch": 13.977635782747605, + "grad_norm": 0.056396484375, + "learning_rate": 0.0005, + "loss": 1.0172, + "step": 17500 + }, + { + "epoch": 13.978434504792332, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 17501 + }, + { + "epoch": 13.97923322683706, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0099, + "step": 17502 + }, + { + "epoch": 13.98003194888179, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 17503 + }, + { + "epoch": 13.980830670926517, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 17504 + }, + { + "epoch": 13.981629392971247, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.016, + "step": 17505 + }, + { + "epoch": 13.982428115015974, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 17506 + }, + { + "epoch": 13.983226837060704, + "grad_norm": 0.1396484375, + "learning_rate": 0.0005, + "loss": 1.0139, + "step": 17507 + }, + { + "epoch": 13.984025559105431, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 17508 + }, + { + "epoch": 13.98482428115016, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0347, + "step": 17509 + }, + { + "epoch": 13.985623003194888, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 17510 + }, + { + "epoch": 13.986421725239616, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 17511 + }, + { + "epoch": 13.987220447284345, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 17512 + }, + { + "epoch": 13.988019169329073, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 17513 + }, + { + "epoch": 13.988817891373802, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 17514 + }, + { + "epoch": 13.98961661341853, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0152, + "step": 17515 + }, + { + "epoch": 13.99041533546326, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 17516 + }, + { + "epoch": 13.991214057507987, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 17517 + }, + { + "epoch": 13.992012779552716, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 17518 + }, + { + "epoch": 13.992811501597444, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 17519 + }, + { + "epoch": 13.993610223642172, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 17520 + }, + { + "epoch": 13.994408945686901, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 17521 + }, + { + "epoch": 13.995207667731629, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 17522 + }, + { + "epoch": 13.996006389776358, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 17523 + }, + { + "epoch": 13.996805111821086, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 17524 + }, + { + "epoch": 13.997603833865815, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.0192, + "step": 17525 + }, + { + "epoch": 13.998402555910543, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 17526 + }, + { + "epoch": 13.999201277955272, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 17527 + }, + { + "epoch": 14.0, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 17528 + }, + { + "epoch": 14.000798722044728, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 17529 + }, + { + "epoch": 14.001597444089457, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 17530 + }, + { + "epoch": 14.002396166134185, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0149, + "step": 17531 + }, + { + "epoch": 14.003194888178914, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 17532 + }, + { + "epoch": 14.003993610223642, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 17533 + }, + { + "epoch": 14.004792332268371, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 17534 + }, + { + "epoch": 14.005591054313099, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 17535 + }, + { + "epoch": 14.006389776357828, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0096, + "step": 17536 + }, + { + "epoch": 14.007188498402556, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 17537 + }, + { + "epoch": 14.007987220447284, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 17538 + }, + { + "epoch": 14.008785942492013, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0163, + "step": 17539 + }, + { + "epoch": 14.00958466453674, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0143, + "step": 17540 + }, + { + "epoch": 14.01038338658147, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 17541 + }, + { + "epoch": 14.011182108626198, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.015, + "step": 17542 + }, + { + "epoch": 14.011980830670927, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 17543 + }, + { + "epoch": 14.012779552715655, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 17544 + }, + { + "epoch": 14.013578274760384, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 17545 + }, + { + "epoch": 14.014376996805112, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 17546 + }, + { + "epoch": 14.01517571884984, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0135, + "step": 17547 + }, + { + "epoch": 14.015974440894569, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0195, + "step": 17548 + }, + { + "epoch": 14.016773162939296, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 17549 + }, + { + "epoch": 14.017571884984026, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 17550 + }, + { + "epoch": 14.018370607028753, + "grad_norm": 0.1474609375, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 17551 + }, + { + "epoch": 14.019169329073483, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 17552 + }, + { + "epoch": 14.01996805111821, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.015, + "step": 17553 + }, + { + "epoch": 14.02076677316294, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 17554 + }, + { + "epoch": 14.021565495207668, + "grad_norm": 0.12060546875, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 17555 + }, + { + "epoch": 14.022364217252397, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0164, + "step": 17556 + }, + { + "epoch": 14.023162939297125, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 17557 + }, + { + "epoch": 14.023961661341852, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0165, + "step": 17558 + }, + { + "epoch": 14.024760383386582, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 17559 + }, + { + "epoch": 14.02555910543131, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 17560 + }, + { + "epoch": 14.026357827476039, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0164, + "step": 17561 + }, + { + "epoch": 14.027156549520766, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 17562 + }, + { + "epoch": 14.027955271565496, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 17563 + }, + { + "epoch": 14.028753993610223, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 17564 + }, + { + "epoch": 14.029552715654953, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 17565 + }, + { + "epoch": 14.03035143769968, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 17566 + }, + { + "epoch": 14.031150159744408, + "grad_norm": 0.138671875, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 17567 + }, + { + "epoch": 14.031948881789138, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0196, + "step": 17568 + }, + { + "epoch": 14.032747603833865, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 17569 + }, + { + "epoch": 14.033546325878595, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 17570 + }, + { + "epoch": 14.034345047923322, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0178, + "step": 17571 + }, + { + "epoch": 14.035143769968052, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 17572 + }, + { + "epoch": 14.03594249201278, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 17573 + }, + { + "epoch": 14.036741214057509, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 17574 + }, + { + "epoch": 14.037539936102236, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 17575 + }, + { + "epoch": 14.038338658146964, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0344, + "step": 17576 + }, + { + "epoch": 14.039137380191693, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0319, + "step": 17577 + }, + { + "epoch": 14.039936102236421, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 17578 + }, + { + "epoch": 14.04073482428115, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 17579 + }, + { + "epoch": 14.041533546325878, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 17580 + }, + { + "epoch": 14.042332268370608, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 17581 + }, + { + "epoch": 14.043130990415335, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0196, + "step": 17582 + }, + { + "epoch": 14.043929712460065, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0178, + "step": 17583 + }, + { + "epoch": 14.044728434504792, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 17584 + }, + { + "epoch": 14.04552715654952, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 17585 + }, + { + "epoch": 14.04632587859425, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 17586 + }, + { + "epoch": 14.047124600638977, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 17587 + }, + { + "epoch": 14.047923322683706, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 17588 + }, + { + "epoch": 14.048722044728434, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 17589 + }, + { + "epoch": 14.049520766773163, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0171, + "step": 17590 + }, + { + "epoch": 14.050319488817891, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 17591 + }, + { + "epoch": 14.05111821086262, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 17592 + }, + { + "epoch": 14.051916932907348, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 17593 + }, + { + "epoch": 14.052715654952078, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0178, + "step": 17594 + }, + { + "epoch": 14.053514376996805, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 17595 + }, + { + "epoch": 14.054313099041533, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 17596 + }, + { + "epoch": 14.055111821086262, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0171, + "step": 17597 + }, + { + "epoch": 14.05591054313099, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 17598 + }, + { + "epoch": 14.05670926517572, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 17599 + }, + { + "epoch": 14.057507987220447, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 17600 + }, + { + "epoch": 14.058306709265176, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0168, + "step": 17601 + }, + { + "epoch": 14.059105431309904, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 17602 + }, + { + "epoch": 14.059904153354633, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 17603 + }, + { + "epoch": 14.060702875399361, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 17604 + }, + { + "epoch": 14.061501597444089, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0165, + "step": 17605 + }, + { + "epoch": 14.062300319488818, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 17606 + }, + { + "epoch": 14.063099041533546, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 17607 + }, + { + "epoch": 14.063897763578275, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0116, + "step": 17608 + }, + { + "epoch": 14.064696485623003, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0128, + "step": 17609 + }, + { + "epoch": 14.065495207667732, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 17610 + }, + { + "epoch": 14.06629392971246, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0157, + "step": 17611 + }, + { + "epoch": 14.06709265175719, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 17612 + }, + { + "epoch": 14.067891373801917, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 17613 + }, + { + "epoch": 14.068690095846645, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 17614 + }, + { + "epoch": 14.069488817891374, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0192, + "step": 17615 + }, + { + "epoch": 14.070287539936102, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0134, + "step": 17616 + }, + { + "epoch": 14.071086261980831, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 17617 + }, + { + "epoch": 14.071884984025559, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 17618 + }, + { + "epoch": 14.072683706070288, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0124, + "step": 17619 + }, + { + "epoch": 14.073482428115016, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 17620 + }, + { + "epoch": 14.074281150159745, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 17621 + }, + { + "epoch": 14.075079872204473, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 17622 + }, + { + "epoch": 14.0758785942492, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 17623 + }, + { + "epoch": 14.07667731629393, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 17624 + }, + { + "epoch": 14.077476038338657, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 17625 + }, + { + "epoch": 14.078274760383387, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 17626 + }, + { + "epoch": 14.079073482428115, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0187, + "step": 17627 + }, + { + "epoch": 14.079872204472844, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 17628 + }, + { + "epoch": 14.080670926517572, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0144, + "step": 17629 + }, + { + "epoch": 14.081469648562301, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 17630 + }, + { + "epoch": 14.082268370607029, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 17631 + }, + { + "epoch": 14.083067092651758, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 17632 + }, + { + "epoch": 14.083865814696486, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0117, + "step": 17633 + }, + { + "epoch": 14.084664536741213, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 17634 + }, + { + "epoch": 14.085463258785943, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 17635 + }, + { + "epoch": 14.08626198083067, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.015, + "step": 17636 + }, + { + "epoch": 14.0870607028754, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0185, + "step": 17637 + }, + { + "epoch": 14.087859424920127, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 17638 + }, + { + "epoch": 14.088658146964857, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0153, + "step": 17639 + }, + { + "epoch": 14.089456869009584, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 17640 + }, + { + "epoch": 14.090255591054314, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 17641 + }, + { + "epoch": 14.091054313099042, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 17642 + }, + { + "epoch": 14.09185303514377, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 17643 + }, + { + "epoch": 14.092651757188499, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0133, + "step": 17644 + }, + { + "epoch": 14.093450479233226, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 17645 + }, + { + "epoch": 14.094249201277956, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 17646 + }, + { + "epoch": 14.095047923322683, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 17647 + }, + { + "epoch": 14.095846645367413, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 17648 + }, + { + "epoch": 14.09664536741214, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0085, + "step": 17649 + }, + { + "epoch": 14.09744408945687, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 17650 + }, + { + "epoch": 14.098242811501597, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 17651 + }, + { + "epoch": 14.099041533546325, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 17652 + }, + { + "epoch": 14.099840255591054, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 17653 + }, + { + "epoch": 14.100638977635782, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 17654 + }, + { + "epoch": 14.101437699680512, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0138, + "step": 17655 + }, + { + "epoch": 14.10223642172524, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 17656 + }, + { + "epoch": 14.103035143769969, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 17657 + }, + { + "epoch": 14.103833865814696, + "grad_norm": 0.0556640625, + "learning_rate": 0.0005, + "loss": 1.018, + "step": 17658 + }, + { + "epoch": 14.104632587859426, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 17659 + }, + { + "epoch": 14.105431309904153, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 17660 + }, + { + "epoch": 14.106230031948881, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 17661 + }, + { + "epoch": 14.10702875399361, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 17662 + }, + { + "epoch": 14.107827476038338, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 17663 + }, + { + "epoch": 14.108626198083067, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 17664 + }, + { + "epoch": 14.109424920127795, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 17665 + }, + { + "epoch": 14.110223642172524, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 17666 + }, + { + "epoch": 14.111022364217252, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0168, + "step": 17667 + }, + { + "epoch": 14.111821086261982, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0185, + "step": 17668 + }, + { + "epoch": 14.11261980830671, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 17669 + }, + { + "epoch": 14.113418530351439, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 17670 + }, + { + "epoch": 14.114217252396166, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 17671 + }, + { + "epoch": 14.115015974440894, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 17672 + }, + { + "epoch": 14.115814696485623, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 17673 + }, + { + "epoch": 14.116613418530351, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0141, + "step": 17674 + }, + { + "epoch": 14.11741214057508, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 17675 + }, + { + "epoch": 14.118210862619808, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0129, + "step": 17676 + }, + { + "epoch": 14.119009584664537, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 17677 + }, + { + "epoch": 14.119808306709265, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 17678 + }, + { + "epoch": 14.120607028753994, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 17679 + }, + { + "epoch": 14.121405750798722, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 17680 + }, + { + "epoch": 14.12220447284345, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 17681 + }, + { + "epoch": 14.12300319488818, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0071, + "step": 17682 + }, + { + "epoch": 14.123801916932907, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 17683 + }, + { + "epoch": 14.124600638977636, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 17684 + }, + { + "epoch": 14.125399361022364, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 17685 + }, + { + "epoch": 14.126198083067093, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 17686 + }, + { + "epoch": 14.12699680511182, + "grad_norm": 0.11962890625, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 17687 + }, + { + "epoch": 14.12779552715655, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 17688 + }, + { + "epoch": 14.128594249201278, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0159, + "step": 17689 + }, + { + "epoch": 14.129392971246006, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0175, + "step": 17690 + }, + { + "epoch": 14.130191693290735, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.035, + "step": 17691 + }, + { + "epoch": 14.130990415335463, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 17692 + }, + { + "epoch": 14.131789137380192, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0183, + "step": 17693 + }, + { + "epoch": 14.13258785942492, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 17694 + }, + { + "epoch": 14.133386581469649, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 17695 + }, + { + "epoch": 14.134185303514377, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 17696 + }, + { + "epoch": 14.134984025559106, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 17697 + }, + { + "epoch": 14.135782747603834, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 17698 + }, + { + "epoch": 14.136581469648561, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 17699 + }, + { + "epoch": 14.13738019169329, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0174, + "step": 17700 + }, + { + "epoch": 14.138178913738018, + "grad_norm": 0.1416015625, + "learning_rate": 0.0005, + "loss": 1.0176, + "step": 17701 + }, + { + "epoch": 14.138977635782748, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 17702 + }, + { + "epoch": 14.139776357827476, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 17703 + }, + { + "epoch": 14.140575079872205, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 17704 + }, + { + "epoch": 14.141373801916933, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 17705 + }, + { + "epoch": 14.142172523961662, + "grad_norm": 0.158203125, + "learning_rate": 0.0005, + "loss": 1.0126, + "step": 17706 + }, + { + "epoch": 14.14297124600639, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 17707 + }, + { + "epoch": 14.143769968051119, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 17708 + }, + { + "epoch": 14.144568690095847, + "grad_norm": 0.1455078125, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 17709 + }, + { + "epoch": 14.145367412140574, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 17710 + }, + { + "epoch": 14.146166134185304, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 17711 + }, + { + "epoch": 14.146964856230031, + "grad_norm": 0.10693359375, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 17712 + }, + { + "epoch": 14.14776357827476, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 17713 + }, + { + "epoch": 14.148562300319488, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 17714 + }, + { + "epoch": 14.149361022364218, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 17715 + }, + { + "epoch": 14.150159744408946, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 17716 + }, + { + "epoch": 14.150958466453675, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 17717 + }, + { + "epoch": 14.151757188498403, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 17718 + }, + { + "epoch": 14.15255591054313, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0177, + "step": 17719 + }, + { + "epoch": 14.15335463258786, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 17720 + }, + { + "epoch": 14.154153354632587, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 17721 + }, + { + "epoch": 14.154952076677317, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 17722 + }, + { + "epoch": 14.155750798722044, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 17723 + }, + { + "epoch": 14.156549520766774, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 17724 + }, + { + "epoch": 14.157348242811501, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 17725 + }, + { + "epoch": 14.15814696485623, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 17726 + }, + { + "epoch": 14.158945686900958, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0138, + "step": 17727 + }, + { + "epoch": 14.159744408945686, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 17728 + }, + { + "epoch": 14.160543130990416, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 17729 + }, + { + "epoch": 14.161341853035143, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0182, + "step": 17730 + }, + { + "epoch": 14.162140575079873, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0179, + "step": 17731 + }, + { + "epoch": 14.1629392971246, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0169, + "step": 17732 + }, + { + "epoch": 14.16373801916933, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 17733 + }, + { + "epoch": 14.164536741214057, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 17734 + }, + { + "epoch": 14.165335463258787, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 17735 + }, + { + "epoch": 14.166134185303514, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 17736 + }, + { + "epoch": 14.166932907348242, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 17737 + }, + { + "epoch": 14.167731629392971, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 17738 + }, + { + "epoch": 14.168530351437699, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 17739 + }, + { + "epoch": 14.169329073482428, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 17740 + }, + { + "epoch": 14.170127795527156, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 17741 + }, + { + "epoch": 14.170926517571885, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 17742 + }, + { + "epoch": 14.171725239616613, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 17743 + }, + { + "epoch": 14.172523961661343, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0195, + "step": 17744 + }, + { + "epoch": 14.17332268370607, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0178, + "step": 17745 + }, + { + "epoch": 14.1741214057508, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 17746 + }, + { + "epoch": 14.174920127795527, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 17747 + }, + { + "epoch": 14.175718849840255, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 17748 + }, + { + "epoch": 14.176517571884984, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0137, + "step": 17749 + }, + { + "epoch": 14.177316293929712, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0205, + "step": 17750 + }, + { + "epoch": 14.178115015974441, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 17751 + }, + { + "epoch": 14.178913738019169, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 17752 + }, + { + "epoch": 14.179712460063898, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 17753 + }, + { + "epoch": 14.180511182108626, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0187, + "step": 17754 + }, + { + "epoch": 14.181309904153355, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 17755 + }, + { + "epoch": 14.182108626198083, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 17756 + }, + { + "epoch": 14.18290734824281, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 17757 + }, + { + "epoch": 14.18370607028754, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 17758 + }, + { + "epoch": 14.184504792332268, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 17759 + }, + { + "epoch": 14.185303514376997, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 17760 + }, + { + "epoch": 14.186102236421725, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 17761 + }, + { + "epoch": 14.186900958466454, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 17762 + }, + { + "epoch": 14.187699680511182, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 17763 + }, + { + "epoch": 14.188498402555911, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 17764 + }, + { + "epoch": 14.189297124600639, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0195, + "step": 17765 + }, + { + "epoch": 14.190095846645367, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.0196, + "step": 17766 + }, + { + "epoch": 14.190894568690096, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 17767 + }, + { + "epoch": 14.191693290734824, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0179, + "step": 17768 + }, + { + "epoch": 14.192492012779553, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 17769 + }, + { + "epoch": 14.19329073482428, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0172, + "step": 17770 + }, + { + "epoch": 14.19408945686901, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 17771 + }, + { + "epoch": 14.194888178913738, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 17772 + }, + { + "epoch": 14.195686900958467, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0179, + "step": 17773 + }, + { + "epoch": 14.196485623003195, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 17774 + }, + { + "epoch": 14.197284345047922, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0166, + "step": 17775 + }, + { + "epoch": 14.198083067092652, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.016, + "step": 17776 + }, + { + "epoch": 14.19888178913738, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0101, + "step": 17777 + }, + { + "epoch": 14.199680511182109, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 17778 + }, + { + "epoch": 14.200479233226837, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 17779 + }, + { + "epoch": 14.201277955271566, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 17780 + }, + { + "epoch": 14.202076677316294, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0195, + "step": 17781 + }, + { + "epoch": 14.202875399361023, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 17782 + }, + { + "epoch": 14.20367412140575, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 17783 + }, + { + "epoch": 14.204472843450478, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 17784 + }, + { + "epoch": 14.205271565495208, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0049, + "step": 17785 + }, + { + "epoch": 14.206070287539935, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 17786 + }, + { + "epoch": 14.206869009584665, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 17787 + }, + { + "epoch": 14.207667731629392, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 17788 + }, + { + "epoch": 14.208466453674122, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 17789 + }, + { + "epoch": 14.20926517571885, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 17790 + }, + { + "epoch": 14.210063897763579, + "grad_norm": 0.1552734375, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 17791 + }, + { + "epoch": 14.210862619808307, + "grad_norm": 0.140625, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 17792 + }, + { + "epoch": 14.211661341853036, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 17793 + }, + { + "epoch": 14.212460063897764, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0145, + "step": 17794 + }, + { + "epoch": 14.213258785942491, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 17795 + }, + { + "epoch": 14.21405750798722, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 17796 + }, + { + "epoch": 14.214856230031948, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 17797 + }, + { + "epoch": 14.215654952076678, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 17798 + }, + { + "epoch": 14.216453674121405, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 17799 + }, + { + "epoch": 14.217252396166135, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0169, + "step": 17800 + }, + { + "epoch": 14.218051118210862, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0181, + "step": 17801 + }, + { + "epoch": 14.218849840255592, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 17802 + }, + { + "epoch": 14.21964856230032, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0309, + "step": 17803 + }, + { + "epoch": 14.220447284345047, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0169, + "step": 17804 + }, + { + "epoch": 14.221246006389777, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0096, + "step": 17805 + }, + { + "epoch": 14.222044728434504, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 17806 + }, + { + "epoch": 14.222843450479234, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 17807 + }, + { + "epoch": 14.223642172523961, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 17808 + }, + { + "epoch": 14.22444089456869, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 17809 + }, + { + "epoch": 14.225239616613418, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 17810 + }, + { + "epoch": 14.226038338658148, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 17811 + }, + { + "epoch": 14.226837060702875, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 17812 + }, + { + "epoch": 14.227635782747603, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 17813 + }, + { + "epoch": 14.228434504792332, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0304, + "step": 17814 + }, + { + "epoch": 14.22923322683706, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 17815 + }, + { + "epoch": 14.23003194888179, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 17816 + }, + { + "epoch": 14.230830670926517, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 17817 + }, + { + "epoch": 14.231629392971247, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 17818 + }, + { + "epoch": 14.232428115015974, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 17819 + }, + { + "epoch": 14.233226837060704, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 17820 + }, + { + "epoch": 14.234025559105431, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 17821 + }, + { + "epoch": 14.23482428115016, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0187, + "step": 17822 + }, + { + "epoch": 14.235623003194888, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0188, + "step": 17823 + }, + { + "epoch": 14.236421725239616, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0185, + "step": 17824 + }, + { + "epoch": 14.237220447284345, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 17825 + }, + { + "epoch": 14.238019169329073, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0199, + "step": 17826 + }, + { + "epoch": 14.238817891373802, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 17827 + }, + { + "epoch": 14.23961661341853, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0114, + "step": 17828 + }, + { + "epoch": 14.24041533546326, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0162, + "step": 17829 + }, + { + "epoch": 14.241214057507987, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0141, + "step": 17830 + }, + { + "epoch": 14.242012779552716, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0163, + "step": 17831 + }, + { + "epoch": 14.242811501597444, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 17832 + }, + { + "epoch": 14.243610223642172, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0113, + "step": 17833 + }, + { + "epoch": 14.244408945686901, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0187, + "step": 17834 + }, + { + "epoch": 14.245207667731629, + "grad_norm": 0.11474609375, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 17835 + }, + { + "epoch": 14.246006389776358, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 17836 + }, + { + "epoch": 14.246805111821086, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 17837 + }, + { + "epoch": 14.247603833865815, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0117, + "step": 17838 + }, + { + "epoch": 14.248402555910543, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 17839 + }, + { + "epoch": 14.249201277955272, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0148, + "step": 17840 + }, + { + "epoch": 14.25, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 17841 + }, + { + "epoch": 14.250798722044728, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 17842 + }, + { + "epoch": 14.251597444089457, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0142, + "step": 17843 + }, + { + "epoch": 14.252396166134185, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 17844 + }, + { + "epoch": 14.253194888178914, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0187, + "step": 17845 + }, + { + "epoch": 14.253993610223642, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0138, + "step": 17846 + }, + { + "epoch": 14.254792332268371, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 17847 + }, + { + "epoch": 14.255591054313099, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 17848 + }, + { + "epoch": 14.256389776357828, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0192, + "step": 17849 + }, + { + "epoch": 14.257188498402556, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 17850 + }, + { + "epoch": 14.257987220447284, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.016, + "step": 17851 + }, + { + "epoch": 14.258785942492013, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 17852 + }, + { + "epoch": 14.25958466453674, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0151, + "step": 17853 + }, + { + "epoch": 14.26038338658147, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0144, + "step": 17854 + }, + { + "epoch": 14.261182108626198, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 17855 + }, + { + "epoch": 14.261980830670927, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 17856 + }, + { + "epoch": 14.262779552715655, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 17857 + }, + { + "epoch": 14.263578274760384, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0205, + "step": 17858 + }, + { + "epoch": 14.264376996805112, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0169, + "step": 17859 + }, + { + "epoch": 14.26517571884984, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 17860 + }, + { + "epoch": 14.265974440894569, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 17861 + }, + { + "epoch": 14.266773162939296, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0181, + "step": 17862 + }, + { + "epoch": 14.267571884984026, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0156, + "step": 17863 + }, + { + "epoch": 14.268370607028753, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0162, + "step": 17864 + }, + { + "epoch": 14.269169329073483, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 17865 + }, + { + "epoch": 14.26996805111821, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 17866 + }, + { + "epoch": 14.27076677316294, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 17867 + }, + { + "epoch": 14.271565495207668, + "grad_norm": 0.162109375, + "learning_rate": 0.0005, + "loss": 1.0195, + "step": 17868 + }, + { + "epoch": 14.272364217252397, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.018, + "step": 17869 + }, + { + "epoch": 14.273162939297125, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0172, + "step": 17870 + }, + { + "epoch": 14.273961661341852, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 17871 + }, + { + "epoch": 14.274760383386582, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 17872 + }, + { + "epoch": 14.27555910543131, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 17873 + }, + { + "epoch": 14.276357827476039, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 17874 + }, + { + "epoch": 14.277156549520766, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 17875 + }, + { + "epoch": 14.277955271565496, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 17876 + }, + { + "epoch": 14.278753993610223, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0171, + "step": 17877 + }, + { + "epoch": 14.279552715654953, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0167, + "step": 17878 + }, + { + "epoch": 14.28035143769968, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0136, + "step": 17879 + }, + { + "epoch": 14.281150159744408, + "grad_norm": 0.15234375, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 17880 + }, + { + "epoch": 14.281948881789138, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 17881 + }, + { + "epoch": 14.282747603833865, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 17882 + }, + { + "epoch": 14.283546325878595, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 17883 + }, + { + "epoch": 14.284345047923322, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 17884 + }, + { + "epoch": 14.285143769968052, + "grad_norm": 0.201171875, + "learning_rate": 0.0005, + "loss": 1.016, + "step": 17885 + }, + { + "epoch": 14.28594249201278, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 17886 + }, + { + "epoch": 14.286741214057509, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 17887 + }, + { + "epoch": 14.287539936102236, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 17888 + }, + { + "epoch": 14.288338658146964, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 17889 + }, + { + "epoch": 14.289137380191693, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 17890 + }, + { + "epoch": 14.289936102236421, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0159, + "step": 17891 + }, + { + "epoch": 14.29073482428115, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 17892 + }, + { + "epoch": 14.291533546325878, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0176, + "step": 17893 + }, + { + "epoch": 14.292332268370608, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 17894 + }, + { + "epoch": 14.293130990415335, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 17895 + }, + { + "epoch": 14.293929712460065, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 17896 + }, + { + "epoch": 14.294728434504792, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 17897 + }, + { + "epoch": 14.295527156549522, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 17898 + }, + { + "epoch": 14.29632587859425, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 17899 + }, + { + "epoch": 14.297124600638977, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 17900 + }, + { + "epoch": 14.297923322683706, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0119, + "step": 17901 + }, + { + "epoch": 14.298722044728434, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 17902 + }, + { + "epoch": 14.299520766773163, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 17903 + }, + { + "epoch": 14.300319488817891, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 17904 + }, + { + "epoch": 14.30111821086262, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 17905 + }, + { + "epoch": 14.301916932907348, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 17906 + }, + { + "epoch": 14.302715654952078, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 17907 + }, + { + "epoch": 14.303514376996805, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 17908 + }, + { + "epoch": 14.304313099041533, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 17909 + }, + { + "epoch": 14.305111821086262, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 17910 + }, + { + "epoch": 14.30591054313099, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 17911 + }, + { + "epoch": 14.30670926517572, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 17912 + }, + { + "epoch": 14.307507987220447, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0153, + "step": 17913 + }, + { + "epoch": 14.308306709265176, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0183, + "step": 17914 + }, + { + "epoch": 14.309105431309904, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 17915 + }, + { + "epoch": 14.309904153354633, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 17916 + }, + { + "epoch": 14.310702875399361, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0325, + "step": 17917 + }, + { + "epoch": 14.311501597444089, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 17918 + }, + { + "epoch": 14.312300319488818, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0168, + "step": 17919 + }, + { + "epoch": 14.313099041533546, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 17920 + }, + { + "epoch": 14.313897763578275, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0315, + "step": 17921 + }, + { + "epoch": 14.314696485623003, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 17922 + }, + { + "epoch": 14.315495207667732, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0152, + "step": 17923 + }, + { + "epoch": 14.31629392971246, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 17924 + }, + { + "epoch": 14.31709265175719, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0097, + "step": 17925 + }, + { + "epoch": 14.317891373801917, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 17926 + }, + { + "epoch": 14.318690095846645, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0145, + "step": 17927 + }, + { + "epoch": 14.319488817891374, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 17928 + }, + { + "epoch": 14.320287539936102, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 17929 + }, + { + "epoch": 14.321086261980831, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 17930 + }, + { + "epoch": 14.321884984025559, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0172, + "step": 17931 + }, + { + "epoch": 14.322683706070288, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 17932 + }, + { + "epoch": 14.323482428115016, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0151, + "step": 17933 + }, + { + "epoch": 14.324281150159745, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0102, + "step": 17934 + }, + { + "epoch": 14.325079872204473, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0156, + "step": 17935 + }, + { + "epoch": 14.3258785942492, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 17936 + }, + { + "epoch": 14.32667731629393, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0156, + "step": 17937 + }, + { + "epoch": 14.327476038338657, + "grad_norm": 0.05859375, + "learning_rate": 0.0005, + "loss": 1.0112, + "step": 17938 + }, + { + "epoch": 14.328274760383387, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 17939 + }, + { + "epoch": 14.329073482428115, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0199, + "step": 17940 + }, + { + "epoch": 14.329872204472844, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 17941 + }, + { + "epoch": 14.330670926517572, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.016, + "step": 17942 + }, + { + "epoch": 14.331469648562301, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 17943 + }, + { + "epoch": 14.332268370607029, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 17944 + }, + { + "epoch": 14.333067092651758, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 17945 + }, + { + "epoch": 14.333865814696486, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0199, + "step": 17946 + }, + { + "epoch": 14.334664536741213, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0136, + "step": 17947 + }, + { + "epoch": 14.335463258785943, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0161, + "step": 17948 + }, + { + "epoch": 14.33626198083067, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 17949 + }, + { + "epoch": 14.3370607028754, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.013, + "step": 17950 + }, + { + "epoch": 14.337859424920127, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 17951 + }, + { + "epoch": 14.338658146964857, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0181, + "step": 17952 + }, + { + "epoch": 14.339456869009584, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 17953 + }, + { + "epoch": 14.340255591054314, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 17954 + }, + { + "epoch": 14.341054313099042, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 17955 + }, + { + "epoch": 14.34185303514377, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0098, + "step": 17956 + }, + { + "epoch": 14.342651757188499, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 17957 + }, + { + "epoch": 14.343450479233226, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.033, + "step": 17958 + }, + { + "epoch": 14.344249201277956, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 17959 + }, + { + "epoch": 14.345047923322683, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 17960 + }, + { + "epoch": 14.345846645367413, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 17961 + }, + { + "epoch": 14.34664536741214, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 17962 + }, + { + "epoch": 14.34744408945687, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 17963 + }, + { + "epoch": 14.348242811501597, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0152, + "step": 17964 + }, + { + "epoch": 14.349041533546325, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0117, + "step": 17965 + }, + { + "epoch": 14.349840255591054, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 17966 + }, + { + "epoch": 14.350638977635782, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0159, + "step": 17967 + }, + { + "epoch": 14.351437699680512, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 17968 + }, + { + "epoch": 14.35223642172524, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 17969 + }, + { + "epoch": 14.353035143769969, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0101, + "step": 17970 + }, + { + "epoch": 14.353833865814696, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 17971 + }, + { + "epoch": 14.354632587859426, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0188, + "step": 17972 + }, + { + "epoch": 14.355431309904153, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 17973 + }, + { + "epoch": 14.356230031948883, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0111, + "step": 17974 + }, + { + "epoch": 14.35702875399361, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 17975 + }, + { + "epoch": 14.357827476038338, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 17976 + }, + { + "epoch": 14.358626198083067, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0174, + "step": 17977 + }, + { + "epoch": 14.359424920127795, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 17978 + }, + { + "epoch": 14.360223642172524, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 17979 + }, + { + "epoch": 14.361022364217252, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 17980 + }, + { + "epoch": 14.361821086261982, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0077, + "step": 17981 + }, + { + "epoch": 14.36261980830671, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0195, + "step": 17982 + }, + { + "epoch": 14.363418530351439, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 17983 + }, + { + "epoch": 14.364217252396166, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 17984 + }, + { + "epoch": 14.365015974440894, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 17985 + }, + { + "epoch": 14.365814696485623, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0166, + "step": 17986 + }, + { + "epoch": 14.366613418530351, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0192, + "step": 17987 + }, + { + "epoch": 14.36741214057508, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 17988 + }, + { + "epoch": 14.368210862619808, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 17989 + }, + { + "epoch": 14.369009584664537, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 17990 + }, + { + "epoch": 14.369808306709265, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 17991 + }, + { + "epoch": 14.370607028753994, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0125, + "step": 17992 + }, + { + "epoch": 14.371405750798722, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0185, + "step": 17993 + }, + { + "epoch": 14.37220447284345, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 17994 + }, + { + "epoch": 14.37300319488818, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 17995 + }, + { + "epoch": 14.373801916932907, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0157, + "step": 17996 + }, + { + "epoch": 14.374600638977636, + "grad_norm": 0.1708984375, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 17997 + }, + { + "epoch": 14.375399361022364, + "grad_norm": 0.1787109375, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 17998 + }, + { + "epoch": 14.376198083067093, + "grad_norm": 0.2138671875, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 17999 + }, + { + "epoch": 14.37699680511182, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 18000 + }, + { + "epoch": 14.37779552715655, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 18001 + }, + { + "epoch": 14.378594249201278, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 18002 + }, + { + "epoch": 14.379392971246006, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 18003 + }, + { + "epoch": 14.380191693290735, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 18004 + }, + { + "epoch": 14.380990415335463, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0295, + "step": 18005 + }, + { + "epoch": 14.381789137380192, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0088, + "step": 18006 + }, + { + "epoch": 14.38258785942492, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0172, + "step": 18007 + }, + { + "epoch": 14.383386581469649, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 18008 + }, + { + "epoch": 14.384185303514377, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0149, + "step": 18009 + }, + { + "epoch": 14.384984025559106, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0156, + "step": 18010 + }, + { + "epoch": 14.385782747603834, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 18011 + }, + { + "epoch": 14.386581469648561, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0185, + "step": 18012 + }, + { + "epoch": 14.38738019169329, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 18013 + }, + { + "epoch": 14.388178913738018, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 18014 + }, + { + "epoch": 14.388977635782748, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 18015 + }, + { + "epoch": 14.389776357827476, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0175, + "step": 18016 + }, + { + "epoch": 14.390575079872205, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 18017 + }, + { + "epoch": 14.391373801916933, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0156, + "step": 18018 + }, + { + "epoch": 14.392172523961662, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 18019 + }, + { + "epoch": 14.39297124600639, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 18020 + }, + { + "epoch": 14.393769968051117, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 18021 + }, + { + "epoch": 14.394568690095847, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 18022 + }, + { + "epoch": 14.395367412140574, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 18023 + }, + { + "epoch": 14.396166134185304, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0152, + "step": 18024 + }, + { + "epoch": 14.396964856230031, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0178, + "step": 18025 + }, + { + "epoch": 14.39776357827476, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 18026 + }, + { + "epoch": 14.398562300319488, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0174, + "step": 18027 + }, + { + "epoch": 14.399361022364218, + "grad_norm": 0.189453125, + "learning_rate": 0.0005, + "loss": 1.0188, + "step": 18028 + }, + { + "epoch": 14.400159744408946, + "grad_norm": 0.1435546875, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 18029 + }, + { + "epoch": 14.400958466453675, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 18030 + }, + { + "epoch": 14.401757188498403, + "grad_norm": 0.166015625, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 18031 + }, + { + "epoch": 14.40255591054313, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.017, + "step": 18032 + }, + { + "epoch": 14.40335463258786, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 18033 + }, + { + "epoch": 14.404153354632587, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 18034 + }, + { + "epoch": 14.404952076677317, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 18035 + }, + { + "epoch": 14.405750798722044, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 18036 + }, + { + "epoch": 14.406549520766774, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 18037 + }, + { + "epoch": 14.407348242811501, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 18038 + }, + { + "epoch": 14.40814696485623, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0168, + "step": 18039 + }, + { + "epoch": 14.408945686900958, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0183, + "step": 18040 + }, + { + "epoch": 14.409744408945686, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0183, + "step": 18041 + }, + { + "epoch": 14.410543130990416, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0192, + "step": 18042 + }, + { + "epoch": 14.411341853035143, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 18043 + }, + { + "epoch": 14.412140575079873, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 18044 + }, + { + "epoch": 14.4129392971246, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 18045 + }, + { + "epoch": 14.41373801916933, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 18046 + }, + { + "epoch": 14.414536741214057, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0161, + "step": 18047 + }, + { + "epoch": 14.415335463258787, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0171, + "step": 18048 + }, + { + "epoch": 14.416134185303514, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 18049 + }, + { + "epoch": 14.416932907348242, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0133, + "step": 18050 + }, + { + "epoch": 14.417731629392971, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 18051 + }, + { + "epoch": 14.418530351437699, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 18052 + }, + { + "epoch": 14.419329073482428, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0146, + "step": 18053 + }, + { + "epoch": 14.420127795527156, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0169, + "step": 18054 + }, + { + "epoch": 14.420926517571885, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0164, + "step": 18055 + }, + { + "epoch": 14.421725239616613, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 18056 + }, + { + "epoch": 14.422523961661343, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 18057 + }, + { + "epoch": 14.42332268370607, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 18058 + }, + { + "epoch": 14.4241214057508, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0153, + "step": 18059 + }, + { + "epoch": 14.424920127795527, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 18060 + }, + { + "epoch": 14.425718849840255, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0165, + "step": 18061 + }, + { + "epoch": 14.426517571884984, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0183, + "step": 18062 + }, + { + "epoch": 14.427316293929712, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 18063 + }, + { + "epoch": 14.428115015974441, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0183, + "step": 18064 + }, + { + "epoch": 14.428913738019169, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 18065 + }, + { + "epoch": 14.429712460063898, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0172, + "step": 18066 + }, + { + "epoch": 14.430511182108626, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 18067 + }, + { + "epoch": 14.431309904153355, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0173, + "step": 18068 + }, + { + "epoch": 14.432108626198083, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 18069 + }, + { + "epoch": 14.43290734824281, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0073, + "step": 18070 + }, + { + "epoch": 14.43370607028754, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 18071 + }, + { + "epoch": 14.434504792332268, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0128, + "step": 18072 + }, + { + "epoch": 14.435303514376997, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0161, + "step": 18073 + }, + { + "epoch": 14.436102236421725, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 18074 + }, + { + "epoch": 14.436900958466454, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 18075 + }, + { + "epoch": 14.437699680511182, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 18076 + }, + { + "epoch": 14.438498402555911, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 18077 + }, + { + "epoch": 14.439297124600639, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 18078 + }, + { + "epoch": 14.440095846645367, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 18079 + }, + { + "epoch": 14.440894568690096, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 18080 + }, + { + "epoch": 14.441693290734824, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0188, + "step": 18081 + }, + { + "epoch": 14.442492012779553, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 18082 + }, + { + "epoch": 14.44329073482428, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0139, + "step": 18083 + }, + { + "epoch": 14.44408945686901, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0294, + "step": 18084 + }, + { + "epoch": 14.444888178913738, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 18085 + }, + { + "epoch": 14.445686900958467, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0173, + "step": 18086 + }, + { + "epoch": 14.446485623003195, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0121, + "step": 18087 + }, + { + "epoch": 14.447284345047922, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0081, + "step": 18088 + }, + { + "epoch": 14.448083067092652, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 18089 + }, + { + "epoch": 14.44888178913738, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 18090 + }, + { + "epoch": 14.449680511182109, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 18091 + }, + { + "epoch": 14.450479233226837, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0113, + "step": 18092 + }, + { + "epoch": 14.451277955271566, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.011, + "step": 18093 + }, + { + "epoch": 14.452076677316294, + "grad_norm": 0.125, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 18094 + }, + { + "epoch": 14.452875399361023, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 18095 + }, + { + "epoch": 14.45367412140575, + "grad_norm": 0.1611328125, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 18096 + }, + { + "epoch": 14.454472843450478, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 18097 + }, + { + "epoch": 14.455271565495208, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 18098 + }, + { + "epoch": 14.456070287539935, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 18099 + }, + { + "epoch": 14.456869009584665, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 18100 + }, + { + "epoch": 14.457667731629392, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0165, + "step": 18101 + }, + { + "epoch": 14.458466453674122, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0176, + "step": 18102 + }, + { + "epoch": 14.45926517571885, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0177, + "step": 18103 + }, + { + "epoch": 14.460063897763579, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 18104 + }, + { + "epoch": 14.460862619808307, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 18105 + }, + { + "epoch": 14.461661341853036, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 18106 + }, + { + "epoch": 14.462460063897764, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0174, + "step": 18107 + }, + { + "epoch": 14.463258785942491, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0153, + "step": 18108 + }, + { + "epoch": 14.46405750798722, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 18109 + }, + { + "epoch": 14.464856230031948, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0196, + "step": 18110 + }, + { + "epoch": 14.465654952076678, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 18111 + }, + { + "epoch": 14.466453674121405, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 18112 + }, + { + "epoch": 14.467252396166135, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0183, + "step": 18113 + }, + { + "epoch": 14.468051118210862, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 18114 + }, + { + "epoch": 14.468849840255592, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 18115 + }, + { + "epoch": 14.46964856230032, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 18116 + }, + { + "epoch": 14.470447284345047, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0162, + "step": 18117 + }, + { + "epoch": 14.471246006389777, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 18118 + }, + { + "epoch": 14.472044728434504, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 18119 + }, + { + "epoch": 14.472843450479234, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 18120 + }, + { + "epoch": 14.473642172523961, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 18121 + }, + { + "epoch": 14.47444089456869, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 18122 + }, + { + "epoch": 14.475239616613418, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 18123 + }, + { + "epoch": 14.476038338658148, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 18124 + }, + { + "epoch": 14.476837060702875, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 18125 + }, + { + "epoch": 14.477635782747603, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 18126 + }, + { + "epoch": 14.478434504792332, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 18127 + }, + { + "epoch": 14.47923322683706, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 18128 + }, + { + "epoch": 14.48003194888179, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 18129 + }, + { + "epoch": 14.480830670926517, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 18130 + }, + { + "epoch": 14.481629392971247, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 18131 + }, + { + "epoch": 14.482428115015974, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 18132 + }, + { + "epoch": 14.483226837060704, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 18133 + }, + { + "epoch": 14.484025559105431, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 18134 + }, + { + "epoch": 14.48482428115016, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 18135 + }, + { + "epoch": 14.485623003194888, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0151, + "step": 18136 + }, + { + "epoch": 14.486421725239616, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 18137 + }, + { + "epoch": 14.487220447284345, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0177, + "step": 18138 + }, + { + "epoch": 14.488019169329073, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 18139 + }, + { + "epoch": 14.488817891373802, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 18140 + }, + { + "epoch": 14.48961661341853, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 18141 + }, + { + "epoch": 14.49041533546326, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 18142 + }, + { + "epoch": 14.491214057507987, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 18143 + }, + { + "epoch": 14.492012779552716, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0106, + "step": 18144 + }, + { + "epoch": 14.492811501597444, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 18145 + }, + { + "epoch": 14.493610223642172, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 18146 + }, + { + "epoch": 14.494408945686901, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 18147 + }, + { + "epoch": 14.495207667731629, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 18148 + }, + { + "epoch": 14.496006389776358, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 18149 + }, + { + "epoch": 14.496805111821086, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 18150 + }, + { + "epoch": 14.497603833865815, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0127, + "step": 18151 + }, + { + "epoch": 14.498402555910543, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 18152 + }, + { + "epoch": 14.499201277955272, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0181, + "step": 18153 + }, + { + "epoch": 14.5, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 18154 + }, + { + "epoch": 14.500798722044728, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0152, + "step": 18155 + }, + { + "epoch": 14.501597444089457, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 18156 + }, + { + "epoch": 14.502396166134185, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 18157 + }, + { + "epoch": 14.503194888178914, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0151, + "step": 18158 + }, + { + "epoch": 14.503993610223642, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 18159 + }, + { + "epoch": 14.504792332268371, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 18160 + }, + { + "epoch": 14.505591054313099, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0199, + "step": 18161 + }, + { + "epoch": 14.506389776357828, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 18162 + }, + { + "epoch": 14.507188498402556, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 18163 + }, + { + "epoch": 14.507987220447284, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 18164 + }, + { + "epoch": 14.508785942492013, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 18165 + }, + { + "epoch": 14.50958466453674, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0114, + "step": 18166 + }, + { + "epoch": 14.51038338658147, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.014, + "step": 18167 + }, + { + "epoch": 14.511182108626198, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 18168 + }, + { + "epoch": 14.511980830670927, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0132, + "step": 18169 + }, + { + "epoch": 14.512779552715655, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 18170 + }, + { + "epoch": 14.513578274760384, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0132, + "step": 18171 + }, + { + "epoch": 14.514376996805112, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 18172 + }, + { + "epoch": 14.51517571884984, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 18173 + }, + { + "epoch": 14.515974440894569, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 18174 + }, + { + "epoch": 14.516773162939296, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0192, + "step": 18175 + }, + { + "epoch": 14.517571884984026, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 18176 + }, + { + "epoch": 14.518370607028753, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0175, + "step": 18177 + }, + { + "epoch": 14.519169329073483, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 18178 + }, + { + "epoch": 14.51996805111821, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 18179 + }, + { + "epoch": 14.52076677316294, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0169, + "step": 18180 + }, + { + "epoch": 14.521565495207668, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 18181 + }, + { + "epoch": 14.522364217252395, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0155, + "step": 18182 + }, + { + "epoch": 14.523162939297125, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 18183 + }, + { + "epoch": 14.523961661341852, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 18184 + }, + { + "epoch": 14.524760383386582, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 18185 + }, + { + "epoch": 14.52555910543131, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0153, + "step": 18186 + }, + { + "epoch": 14.526357827476039, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 18187 + }, + { + "epoch": 14.527156549520766, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0178, + "step": 18188 + }, + { + "epoch": 14.527955271565496, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0134, + "step": 18189 + }, + { + "epoch": 14.528753993610223, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0196, + "step": 18190 + }, + { + "epoch": 14.529552715654953, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 18191 + }, + { + "epoch": 14.53035143769968, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 18192 + }, + { + "epoch": 14.531150159744408, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0161, + "step": 18193 + }, + { + "epoch": 14.531948881789138, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0199, + "step": 18194 + }, + { + "epoch": 14.532747603833865, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 18195 + }, + { + "epoch": 14.533546325878595, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 18196 + }, + { + "epoch": 14.534345047923322, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 18197 + }, + { + "epoch": 14.535143769968052, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 18198 + }, + { + "epoch": 14.53594249201278, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 18199 + }, + { + "epoch": 14.536741214057509, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 18200 + }, + { + "epoch": 14.537539936102236, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0175, + "step": 18201 + }, + { + "epoch": 14.538338658146966, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 18202 + }, + { + "epoch": 14.539137380191693, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0155, + "step": 18203 + }, + { + "epoch": 14.539936102236421, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 18204 + }, + { + "epoch": 14.54073482428115, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 18205 + }, + { + "epoch": 14.541533546325878, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 18206 + }, + { + "epoch": 14.542332268370608, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 18207 + }, + { + "epoch": 14.543130990415335, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0154, + "step": 18208 + }, + { + "epoch": 14.543929712460065, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 18209 + }, + { + "epoch": 14.544728434504792, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0185, + "step": 18210 + }, + { + "epoch": 14.545527156549522, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 18211 + }, + { + "epoch": 14.54632587859425, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0147, + "step": 18212 + }, + { + "epoch": 14.547124600638977, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 18213 + }, + { + "epoch": 14.547923322683706, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0167, + "step": 18214 + }, + { + "epoch": 14.548722044728434, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 18215 + }, + { + "epoch": 14.549520766773163, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 18216 + }, + { + "epoch": 14.550319488817891, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 18217 + }, + { + "epoch": 14.55111821086262, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 18218 + }, + { + "epoch": 14.551916932907348, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0159, + "step": 18219 + }, + { + "epoch": 14.552715654952078, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 18220 + }, + { + "epoch": 14.553514376996805, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 18221 + }, + { + "epoch": 14.554313099041533, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 18222 + }, + { + "epoch": 14.555111821086262, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 18223 + }, + { + "epoch": 14.55591054313099, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 18224 + }, + { + "epoch": 14.55670926517572, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 18225 + }, + { + "epoch": 14.557507987220447, + "grad_norm": 0.1171875, + "learning_rate": 0.0005, + "loss": 1.0135, + "step": 18226 + }, + { + "epoch": 14.558306709265176, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 18227 + }, + { + "epoch": 14.559105431309904, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0367, + "step": 18228 + }, + { + "epoch": 14.559904153354633, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 18229 + }, + { + "epoch": 14.560702875399361, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 18230 + }, + { + "epoch": 14.561501597444089, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 18231 + }, + { + "epoch": 14.562300319488818, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 18232 + }, + { + "epoch": 14.563099041533546, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0174, + "step": 18233 + }, + { + "epoch": 14.563897763578275, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 18234 + }, + { + "epoch": 14.564696485623003, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 18235 + }, + { + "epoch": 14.565495207667732, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 18236 + }, + { + "epoch": 14.56629392971246, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 18237 + }, + { + "epoch": 14.56709265175719, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 18238 + }, + { + "epoch": 14.567891373801917, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 18239 + }, + { + "epoch": 14.568690095846645, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.018, + "step": 18240 + }, + { + "epoch": 14.569488817891374, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0187, + "step": 18241 + }, + { + "epoch": 14.570287539936102, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0167, + "step": 18242 + }, + { + "epoch": 14.571086261980831, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 18243 + }, + { + "epoch": 14.571884984025559, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 18244 + }, + { + "epoch": 14.572683706070288, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 18245 + }, + { + "epoch": 14.573482428115016, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0185, + "step": 18246 + }, + { + "epoch": 14.574281150159745, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 18247 + }, + { + "epoch": 14.575079872204473, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 18248 + }, + { + "epoch": 14.5758785942492, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0172, + "step": 18249 + }, + { + "epoch": 14.57667731629393, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 18250 + }, + { + "epoch": 14.577476038338657, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0195, + "step": 18251 + }, + { + "epoch": 14.578274760383387, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 18252 + }, + { + "epoch": 14.579073482428115, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 18253 + }, + { + "epoch": 14.579872204472844, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 18254 + }, + { + "epoch": 14.580670926517572, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 18255 + }, + { + "epoch": 14.581469648562301, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 18256 + }, + { + "epoch": 14.582268370607029, + "grad_norm": 0.1650390625, + "learning_rate": 0.0005, + "loss": 1.0342, + "step": 18257 + }, + { + "epoch": 14.583067092651756, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 18258 + }, + { + "epoch": 14.583865814696486, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0148, + "step": 18259 + }, + { + "epoch": 14.584664536741213, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0177, + "step": 18260 + }, + { + "epoch": 14.585463258785943, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 18261 + }, + { + "epoch": 14.58626198083067, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0183, + "step": 18262 + }, + { + "epoch": 14.5870607028754, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0103, + "step": 18263 + }, + { + "epoch": 14.587859424920127, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 18264 + }, + { + "epoch": 14.588658146964857, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.009, + "step": 18265 + }, + { + "epoch": 14.589456869009584, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 18266 + }, + { + "epoch": 14.590255591054314, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 18267 + }, + { + "epoch": 14.591054313099042, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 18268 + }, + { + "epoch": 14.59185303514377, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 18269 + }, + { + "epoch": 14.592651757188499, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0143, + "step": 18270 + }, + { + "epoch": 14.593450479233226, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 18271 + }, + { + "epoch": 14.594249201277956, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 18272 + }, + { + "epoch": 14.595047923322683, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0141, + "step": 18273 + }, + { + "epoch": 14.595846645367413, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 18274 + }, + { + "epoch": 14.59664536741214, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 18275 + }, + { + "epoch": 14.59744408945687, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0183, + "step": 18276 + }, + { + "epoch": 14.598242811501597, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 18277 + }, + { + "epoch": 14.599041533546325, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.016, + "step": 18278 + }, + { + "epoch": 14.599840255591054, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 18279 + }, + { + "epoch": 14.600638977635782, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.018, + "step": 18280 + }, + { + "epoch": 14.601437699680512, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 18281 + }, + { + "epoch": 14.60223642172524, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 18282 + }, + { + "epoch": 14.603035143769969, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0329, + "step": 18283 + }, + { + "epoch": 14.603833865814696, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 18284 + }, + { + "epoch": 14.604632587859426, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.013, + "step": 18285 + }, + { + "epoch": 14.605431309904153, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 18286 + }, + { + "epoch": 14.606230031948883, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0139, + "step": 18287 + }, + { + "epoch": 14.60702875399361, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 18288 + }, + { + "epoch": 14.607827476038338, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 18289 + }, + { + "epoch": 14.608626198083067, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 18290 + }, + { + "epoch": 14.609424920127795, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 18291 + }, + { + "epoch": 14.610223642172524, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 18292 + }, + { + "epoch": 14.611022364217252, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 18293 + }, + { + "epoch": 14.611821086261982, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 18294 + }, + { + "epoch": 14.61261980830671, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0144, + "step": 18295 + }, + { + "epoch": 14.613418530351439, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0148, + "step": 18296 + }, + { + "epoch": 14.614217252396166, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0153, + "step": 18297 + }, + { + "epoch": 14.615015974440894, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0114, + "step": 18298 + }, + { + "epoch": 14.615814696485623, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0176, + "step": 18299 + }, + { + "epoch": 14.616613418530351, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 18300 + }, + { + "epoch": 14.61741214057508, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0114, + "step": 18301 + }, + { + "epoch": 14.618210862619808, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 18302 + }, + { + "epoch": 14.619009584664537, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 18303 + }, + { + "epoch": 14.619808306709265, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 18304 + }, + { + "epoch": 14.620607028753994, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0159, + "step": 18305 + }, + { + "epoch": 14.621405750798722, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 18306 + }, + { + "epoch": 14.62220447284345, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 18307 + }, + { + "epoch": 14.62300319488818, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 18308 + }, + { + "epoch": 14.623801916932907, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 18309 + }, + { + "epoch": 14.624600638977636, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 18310 + }, + { + "epoch": 14.625399361022364, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 18311 + }, + { + "epoch": 14.626198083067093, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0168, + "step": 18312 + }, + { + "epoch": 14.62699680511182, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0182, + "step": 18313 + }, + { + "epoch": 14.62779552715655, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 18314 + }, + { + "epoch": 14.628594249201278, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 18315 + }, + { + "epoch": 14.629392971246006, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 18316 + }, + { + "epoch": 14.630191693290735, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0196, + "step": 18317 + }, + { + "epoch": 14.630990415335463, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 18318 + }, + { + "epoch": 14.631789137380192, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 18319 + }, + { + "epoch": 14.63258785942492, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0179, + "step": 18320 + }, + { + "epoch": 14.633386581469649, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 18321 + }, + { + "epoch": 14.634185303514377, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 18322 + }, + { + "epoch": 14.634984025559106, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0182, + "step": 18323 + }, + { + "epoch": 14.635782747603834, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 18324 + }, + { + "epoch": 14.636581469648561, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.016, + "step": 18325 + }, + { + "epoch": 14.63738019169329, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 18326 + }, + { + "epoch": 14.638178913738018, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 18327 + }, + { + "epoch": 14.638977635782748, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 18328 + }, + { + "epoch": 14.639776357827476, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0324, + "step": 18329 + }, + { + "epoch": 14.640575079872205, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 18330 + }, + { + "epoch": 14.641373801916933, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0122, + "step": 18331 + }, + { + "epoch": 14.642172523961662, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 18332 + }, + { + "epoch": 14.64297124600639, + "grad_norm": 0.1328125, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 18333 + }, + { + "epoch": 14.643769968051117, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 18334 + }, + { + "epoch": 14.644568690095847, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 18335 + }, + { + "epoch": 14.645367412140574, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0168, + "step": 18336 + }, + { + "epoch": 14.646166134185304, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 18337 + }, + { + "epoch": 14.646964856230031, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 18338 + }, + { + "epoch": 14.64776357827476, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 18339 + }, + { + "epoch": 14.648562300319488, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 18340 + }, + { + "epoch": 14.649361022364218, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 18341 + }, + { + "epoch": 14.650159744408946, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 18342 + }, + { + "epoch": 14.650958466453675, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 18343 + }, + { + "epoch": 14.651757188498403, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 18344 + }, + { + "epoch": 14.65255591054313, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 18345 + }, + { + "epoch": 14.65335463258786, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 18346 + }, + { + "epoch": 14.654153354632587, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 18347 + }, + { + "epoch": 14.654952076677317, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 18348 + }, + { + "epoch": 14.655750798722044, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0185, + "step": 18349 + }, + { + "epoch": 14.656549520766774, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 18350 + }, + { + "epoch": 14.657348242811501, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 18351 + }, + { + "epoch": 14.65814696485623, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 18352 + }, + { + "epoch": 14.658945686900958, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 18353 + }, + { + "epoch": 14.659744408945686, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.017, + "step": 18354 + }, + { + "epoch": 14.660543130990416, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 18355 + }, + { + "epoch": 14.661341853035143, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0173, + "step": 18356 + }, + { + "epoch": 14.662140575079873, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0179, + "step": 18357 + }, + { + "epoch": 14.6629392971246, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0153, + "step": 18358 + }, + { + "epoch": 14.66373801916933, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0205, + "step": 18359 + }, + { + "epoch": 14.664536741214057, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 18360 + }, + { + "epoch": 14.665335463258787, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 18361 + }, + { + "epoch": 14.666134185303514, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 18362 + }, + { + "epoch": 14.666932907348244, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 18363 + }, + { + "epoch": 14.667731629392971, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0127, + "step": 18364 + }, + { + "epoch": 14.668530351437699, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 18365 + }, + { + "epoch": 14.669329073482428, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 18366 + }, + { + "epoch": 14.670127795527156, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0178, + "step": 18367 + }, + { + "epoch": 14.670926517571885, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 18368 + }, + { + "epoch": 14.671725239616613, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0176, + "step": 18369 + }, + { + "epoch": 14.672523961661343, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 18370 + }, + { + "epoch": 14.67332268370607, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 18371 + }, + { + "epoch": 14.6741214057508, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0135, + "step": 18372 + }, + { + "epoch": 14.674920127795527, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0153, + "step": 18373 + }, + { + "epoch": 14.675718849840255, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 18374 + }, + { + "epoch": 14.676517571884984, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 18375 + }, + { + "epoch": 14.677316293929712, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 18376 + }, + { + "epoch": 14.678115015974441, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.014, + "step": 18377 + }, + { + "epoch": 14.678913738019169, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 18378 + }, + { + "epoch": 14.679712460063898, + "grad_norm": 0.1318359375, + "learning_rate": 0.0005, + "loss": 1.0188, + "step": 18379 + }, + { + "epoch": 14.680511182108626, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 18380 + }, + { + "epoch": 14.681309904153355, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 18381 + }, + { + "epoch": 14.682108626198083, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 18382 + }, + { + "epoch": 14.68290734824281, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 18383 + }, + { + "epoch": 14.68370607028754, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 18384 + }, + { + "epoch": 14.684504792332268, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 18385 + }, + { + "epoch": 14.685303514376997, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0154, + "step": 18386 + }, + { + "epoch": 14.686102236421725, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0192, + "step": 18387 + }, + { + "epoch": 14.686900958466454, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0113, + "step": 18388 + }, + { + "epoch": 14.687699680511182, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 18389 + }, + { + "epoch": 14.688498402555911, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 18390 + }, + { + "epoch": 14.689297124600639, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 18391 + }, + { + "epoch": 14.690095846645367, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0182, + "step": 18392 + }, + { + "epoch": 14.690894568690096, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0136, + "step": 18393 + }, + { + "epoch": 14.691693290734824, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 18394 + }, + { + "epoch": 14.692492012779553, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 18395 + }, + { + "epoch": 14.69329073482428, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0131, + "step": 18396 + }, + { + "epoch": 14.69408945686901, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0188, + "step": 18397 + }, + { + "epoch": 14.694888178913738, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 18398 + }, + { + "epoch": 14.695686900958467, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 18399 + }, + { + "epoch": 14.696485623003195, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 18400 + }, + { + "epoch": 14.697284345047922, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0151, + "step": 18401 + }, + { + "epoch": 14.698083067092652, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 18402 + }, + { + "epoch": 14.69888178913738, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 18403 + }, + { + "epoch": 14.699680511182109, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0293, + "step": 18404 + }, + { + "epoch": 14.700479233226837, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0178, + "step": 18405 + }, + { + "epoch": 14.701277955271566, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0158, + "step": 18406 + }, + { + "epoch": 14.702076677316294, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0121, + "step": 18407 + }, + { + "epoch": 14.702875399361023, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0164, + "step": 18408 + }, + { + "epoch": 14.70367412140575, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 18409 + }, + { + "epoch": 14.704472843450478, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 18410 + }, + { + "epoch": 14.705271565495208, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0192, + "step": 18411 + }, + { + "epoch": 14.706070287539935, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0154, + "step": 18412 + }, + { + "epoch": 14.706869009584665, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0178, + "step": 18413 + }, + { + "epoch": 14.707667731629392, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 18414 + }, + { + "epoch": 14.708466453674122, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 18415 + }, + { + "epoch": 14.70926517571885, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 18416 + }, + { + "epoch": 14.710063897763579, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 18417 + }, + { + "epoch": 14.710862619808307, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 18418 + }, + { + "epoch": 14.711661341853034, + "grad_norm": 0.103515625, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 18419 + }, + { + "epoch": 14.712460063897764, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 18420 + }, + { + "epoch": 14.713258785942491, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 18421 + }, + { + "epoch": 14.71405750798722, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 18422 + }, + { + "epoch": 14.714856230031948, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 18423 + }, + { + "epoch": 14.715654952076678, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 18424 + }, + { + "epoch": 14.716453674121405, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0305, + "step": 18425 + }, + { + "epoch": 14.717252396166135, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 18426 + }, + { + "epoch": 14.718051118210862, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0313, + "step": 18427 + }, + { + "epoch": 14.718849840255592, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 18428 + }, + { + "epoch": 14.71964856230032, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 18429 + }, + { + "epoch": 14.720447284345047, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0165, + "step": 18430 + }, + { + "epoch": 14.721246006389777, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 18431 + }, + { + "epoch": 14.722044728434504, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 18432 + }, + { + "epoch": 14.722843450479234, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 18433 + }, + { + "epoch": 14.723642172523961, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0192, + "step": 18434 + }, + { + "epoch": 14.72444089456869, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0188, + "step": 18435 + }, + { + "epoch": 14.725239616613418, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0352, + "step": 18436 + }, + { + "epoch": 14.726038338658148, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 18437 + }, + { + "epoch": 14.726837060702875, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0161, + "step": 18438 + }, + { + "epoch": 14.727635782747605, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 18439 + }, + { + "epoch": 14.728434504792332, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 18440 + }, + { + "epoch": 14.72923322683706, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0153, + "step": 18441 + }, + { + "epoch": 14.73003194888179, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 18442 + }, + { + "epoch": 14.730830670926517, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 18443 + }, + { + "epoch": 14.731629392971247, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 18444 + }, + { + "epoch": 14.732428115015974, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 18445 + }, + { + "epoch": 14.733226837060704, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.015, + "step": 18446 + }, + { + "epoch": 14.734025559105431, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 18447 + }, + { + "epoch": 14.73482428115016, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 18448 + }, + { + "epoch": 14.735623003194888, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 18449 + }, + { + "epoch": 14.736421725239616, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0346, + "step": 18450 + }, + { + "epoch": 14.737220447284345, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0187, + "step": 18451 + }, + { + "epoch": 14.738019169329073, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 18452 + }, + { + "epoch": 14.738817891373802, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0117, + "step": 18453 + }, + { + "epoch": 14.73961661341853, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0152, + "step": 18454 + }, + { + "epoch": 14.74041533546326, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 18455 + }, + { + "epoch": 14.741214057507987, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 18456 + }, + { + "epoch": 14.742012779552716, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 18457 + }, + { + "epoch": 14.742811501597444, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0118, + "step": 18458 + }, + { + "epoch": 14.743610223642172, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0176, + "step": 18459 + }, + { + "epoch": 14.744408945686901, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 18460 + }, + { + "epoch": 14.745207667731629, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0182, + "step": 18461 + }, + { + "epoch": 14.746006389776358, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0178, + "step": 18462 + }, + { + "epoch": 14.746805111821086, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 18463 + }, + { + "epoch": 14.747603833865815, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 18464 + }, + { + "epoch": 14.748402555910543, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 18465 + }, + { + "epoch": 14.749201277955272, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 18466 + }, + { + "epoch": 14.75, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 18467 + }, + { + "epoch": 14.750798722044728, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0052, + "step": 18468 + }, + { + "epoch": 14.751597444089457, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0196, + "step": 18469 + }, + { + "epoch": 14.752396166134185, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 18470 + }, + { + "epoch": 14.753194888178914, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 18471 + }, + { + "epoch": 14.753993610223642, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0199, + "step": 18472 + }, + { + "epoch": 14.754792332268371, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 18473 + }, + { + "epoch": 14.755591054313099, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 18474 + }, + { + "epoch": 14.756389776357828, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 18475 + }, + { + "epoch": 14.757188498402556, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 18476 + }, + { + "epoch": 14.757987220447284, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 18477 + }, + { + "epoch": 14.758785942492013, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0109, + "step": 18478 + }, + { + "epoch": 14.75958466453674, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0105, + "step": 18479 + }, + { + "epoch": 14.76038338658147, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 18480 + }, + { + "epoch": 14.761182108626198, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 18481 + }, + { + "epoch": 14.761980830670927, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0195, + "step": 18482 + }, + { + "epoch": 14.762779552715655, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 18483 + }, + { + "epoch": 14.763578274760384, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0172, + "step": 18484 + }, + { + "epoch": 14.764376996805112, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 18485 + }, + { + "epoch": 14.76517571884984, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 18486 + }, + { + "epoch": 14.765974440894569, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 18487 + }, + { + "epoch": 14.766773162939296, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 18488 + }, + { + "epoch": 14.767571884984026, + "grad_norm": 0.05419921875, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 18489 + }, + { + "epoch": 14.768370607028753, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0185, + "step": 18490 + }, + { + "epoch": 14.769169329073483, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 18491 + }, + { + "epoch": 14.76996805111821, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 18492 + }, + { + "epoch": 14.77076677316294, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 18493 + }, + { + "epoch": 14.771565495207668, + "grad_norm": 0.055419921875, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 18494 + }, + { + "epoch": 14.772364217252395, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 18495 + }, + { + "epoch": 14.773162939297125, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 18496 + }, + { + "epoch": 14.773961661341852, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0171, + "step": 18497 + }, + { + "epoch": 14.774760383386582, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 18498 + }, + { + "epoch": 14.77555910543131, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 18499 + }, + { + "epoch": 14.776357827476039, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 18500 + }, + { + "epoch": 14.777156549520766, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0185, + "step": 18501 + }, + { + "epoch": 14.777955271565496, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0369, + "step": 18502 + }, + { + "epoch": 14.778753993610223, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 18503 + }, + { + "epoch": 14.779552715654953, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 18504 + }, + { + "epoch": 14.78035143769968, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0183, + "step": 18505 + }, + { + "epoch": 14.781150159744408, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 18506 + }, + { + "epoch": 14.781948881789138, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0129, + "step": 18507 + }, + { + "epoch": 14.782747603833865, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 18508 + }, + { + "epoch": 14.783546325878595, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 18509 + }, + { + "epoch": 14.784345047923322, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0157, + "step": 18510 + }, + { + "epoch": 14.785143769968052, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 18511 + }, + { + "epoch": 14.78594249201278, + "grad_norm": 0.05908203125, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 18512 + }, + { + "epoch": 14.786741214057509, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 18513 + }, + { + "epoch": 14.787539936102236, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 18514 + }, + { + "epoch": 14.788338658146966, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 18515 + }, + { + "epoch": 14.789137380191693, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0085, + "step": 18516 + }, + { + "epoch": 14.789936102236421, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 18517 + }, + { + "epoch": 14.79073482428115, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 18518 + }, + { + "epoch": 14.791533546325878, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0181, + "step": 18519 + }, + { + "epoch": 14.792332268370608, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0157, + "step": 18520 + }, + { + "epoch": 14.793130990415335, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0168, + "step": 18521 + }, + { + "epoch": 14.793929712460065, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 18522 + }, + { + "epoch": 14.794728434504792, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0168, + "step": 18523 + }, + { + "epoch": 14.795527156549522, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 18524 + }, + { + "epoch": 14.79632587859425, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 18525 + }, + { + "epoch": 14.797124600638977, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 18526 + }, + { + "epoch": 14.797923322683706, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 18527 + }, + { + "epoch": 14.798722044728434, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 18528 + }, + { + "epoch": 14.799520766773163, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.018, + "step": 18529 + }, + { + "epoch": 14.800319488817891, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 18530 + }, + { + "epoch": 14.80111821086262, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0183, + "step": 18531 + }, + { + "epoch": 14.801916932907348, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 18532 + }, + { + "epoch": 14.802715654952078, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0238, + "step": 18533 + }, + { + "epoch": 14.803514376996805, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 18534 + }, + { + "epoch": 14.804313099041533, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 18535 + }, + { + "epoch": 14.805111821086262, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 18536 + }, + { + "epoch": 14.80591054313099, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0124, + "step": 18537 + }, + { + "epoch": 14.80670926517572, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0173, + "step": 18538 + }, + { + "epoch": 14.807507987220447, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 18539 + }, + { + "epoch": 14.808306709265176, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 18540 + }, + { + "epoch": 14.809105431309904, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0199, + "step": 18541 + }, + { + "epoch": 14.809904153354633, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0199, + "step": 18542 + }, + { + "epoch": 14.810702875399361, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 18543 + }, + { + "epoch": 14.811501597444089, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 18544 + }, + { + "epoch": 14.812300319488818, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 18545 + }, + { + "epoch": 14.813099041533546, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 18546 + }, + { + "epoch": 14.813897763578275, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 18547 + }, + { + "epoch": 14.814696485623003, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 18548 + }, + { + "epoch": 14.815495207667732, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 18549 + }, + { + "epoch": 14.81629392971246, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 18550 + }, + { + "epoch": 14.81709265175719, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0155, + "step": 18551 + }, + { + "epoch": 14.817891373801917, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0168, + "step": 18552 + }, + { + "epoch": 14.818690095846645, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0131, + "step": 18553 + }, + { + "epoch": 14.819488817891374, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 18554 + }, + { + "epoch": 14.820287539936102, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0153, + "step": 18555 + }, + { + "epoch": 14.821086261980831, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 18556 + }, + { + "epoch": 14.821884984025559, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 18557 + }, + { + "epoch": 14.822683706070288, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.017, + "step": 18558 + }, + { + "epoch": 14.823482428115016, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 18559 + }, + { + "epoch": 14.824281150159745, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 18560 + }, + { + "epoch": 14.825079872204473, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0128, + "step": 18561 + }, + { + "epoch": 14.8258785942492, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 18562 + }, + { + "epoch": 14.82667731629393, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0159, + "step": 18563 + }, + { + "epoch": 14.827476038338657, + "grad_norm": 0.1123046875, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 18564 + }, + { + "epoch": 14.828274760383387, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0317, + "step": 18565 + }, + { + "epoch": 14.829073482428115, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0132, + "step": 18566 + }, + { + "epoch": 14.829872204472844, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 18567 + }, + { + "epoch": 14.830670926517572, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 18568 + }, + { + "epoch": 14.831469648562301, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.015, + "step": 18569 + }, + { + "epoch": 14.832268370607029, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0158, + "step": 18570 + }, + { + "epoch": 14.833067092651756, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 18571 + }, + { + "epoch": 14.833865814696486, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 18572 + }, + { + "epoch": 14.834664536741213, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0145, + "step": 18573 + }, + { + "epoch": 14.835463258785943, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0285, + "step": 18574 + }, + { + "epoch": 14.83626198083067, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.032, + "step": 18575 + }, + { + "epoch": 14.8370607028754, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0322, + "step": 18576 + }, + { + "epoch": 14.837859424920127, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 18577 + }, + { + "epoch": 14.838658146964857, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0192, + "step": 18578 + }, + { + "epoch": 14.839456869009584, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0137, + "step": 18579 + }, + { + "epoch": 14.840255591054314, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 18580 + }, + { + "epoch": 14.841054313099042, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 18581 + }, + { + "epoch": 14.84185303514377, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 18582 + }, + { + "epoch": 14.842651757188499, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 18583 + }, + { + "epoch": 14.843450479233226, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 18584 + }, + { + "epoch": 14.844249201277956, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 18585 + }, + { + "epoch": 14.845047923322683, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0177, + "step": 18586 + }, + { + "epoch": 14.845846645367413, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 18587 + }, + { + "epoch": 14.84664536741214, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0185, + "step": 18588 + }, + { + "epoch": 14.84744408945687, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.018, + "step": 18589 + }, + { + "epoch": 14.848242811501597, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 18590 + }, + { + "epoch": 14.849041533546325, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 18591 + }, + { + "epoch": 14.849840255591054, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 18592 + }, + { + "epoch": 14.850638977635782, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0185, + "step": 18593 + }, + { + "epoch": 14.851437699680512, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 18594 + }, + { + "epoch": 14.85223642172524, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 18595 + }, + { + "epoch": 14.853035143769969, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 18596 + }, + { + "epoch": 14.853833865814696, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 18597 + }, + { + "epoch": 14.854632587859426, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 18598 + }, + { + "epoch": 14.855431309904153, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 18599 + }, + { + "epoch": 14.856230031948883, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0172, + "step": 18600 + }, + { + "epoch": 14.85702875399361, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 18601 + }, + { + "epoch": 14.857827476038338, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 18602 + }, + { + "epoch": 14.858626198083067, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 18603 + }, + { + "epoch": 14.859424920127795, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0273, + "step": 18604 + }, + { + "epoch": 14.860223642172524, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 18605 + }, + { + "epoch": 14.861022364217252, + "grad_norm": 0.1279296875, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 18606 + }, + { + "epoch": 14.861821086261982, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 18607 + }, + { + "epoch": 14.86261980830671, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0116, + "step": 18608 + }, + { + "epoch": 14.863418530351439, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 18609 + }, + { + "epoch": 14.864217252396166, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 18610 + }, + { + "epoch": 14.865015974440894, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 18611 + }, + { + "epoch": 14.865814696485623, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 18612 + }, + { + "epoch": 14.866613418530351, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0146, + "step": 18613 + }, + { + "epoch": 14.86741214057508, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0155, + "step": 18614 + }, + { + "epoch": 14.868210862619808, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 18615 + }, + { + "epoch": 14.869009584664537, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 18616 + }, + { + "epoch": 14.869808306709265, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 18617 + }, + { + "epoch": 14.870607028753994, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 18618 + }, + { + "epoch": 14.871405750798722, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 18619 + }, + { + "epoch": 14.87220447284345, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 18620 + }, + { + "epoch": 14.87300319488818, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 18621 + }, + { + "epoch": 14.873801916932907, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 18622 + }, + { + "epoch": 14.874600638977636, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 18623 + }, + { + "epoch": 14.875399361022364, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 18624 + }, + { + "epoch": 14.876198083067093, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 18625 + }, + { + "epoch": 14.87699680511182, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 18626 + }, + { + "epoch": 14.87779552715655, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 18627 + }, + { + "epoch": 14.878594249201278, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0195, + "step": 18628 + }, + { + "epoch": 14.879392971246006, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0195, + "step": 18629 + }, + { + "epoch": 14.880191693290735, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 18630 + }, + { + "epoch": 14.880990415335463, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 18631 + }, + { + "epoch": 14.881789137380192, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 18632 + }, + { + "epoch": 14.88258785942492, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 18633 + }, + { + "epoch": 14.883386581469649, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 18634 + }, + { + "epoch": 14.884185303514377, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0134, + "step": 18635 + }, + { + "epoch": 14.884984025559106, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 18636 + }, + { + "epoch": 14.885782747603834, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 18637 + }, + { + "epoch": 14.886581469648561, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 18638 + }, + { + "epoch": 14.88738019169329, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0172, + "step": 18639 + }, + { + "epoch": 14.888178913738018, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 18640 + }, + { + "epoch": 14.888977635782748, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0169, + "step": 18641 + }, + { + "epoch": 14.889776357827476, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 18642 + }, + { + "epoch": 14.890575079872205, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0192, + "step": 18643 + }, + { + "epoch": 14.891373801916933, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 18644 + }, + { + "epoch": 14.892172523961662, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 18645 + }, + { + "epoch": 14.89297124600639, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 18646 + }, + { + "epoch": 14.893769968051117, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 18647 + }, + { + "epoch": 14.894568690095847, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 18648 + }, + { + "epoch": 14.895367412140574, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0196, + "step": 18649 + }, + { + "epoch": 14.896166134185304, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 18650 + }, + { + "epoch": 14.896964856230031, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 18651 + }, + { + "epoch": 14.89776357827476, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0196, + "step": 18652 + }, + { + "epoch": 14.898562300319488, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 18653 + }, + { + "epoch": 14.899361022364218, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 18654 + }, + { + "epoch": 14.900159744408946, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0205, + "step": 18655 + }, + { + "epoch": 14.900958466453675, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 18656 + }, + { + "epoch": 14.901757188498403, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 18657 + }, + { + "epoch": 14.90255591054313, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.011, + "step": 18658 + }, + { + "epoch": 14.90335463258786, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 18659 + }, + { + "epoch": 14.904153354632587, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0188, + "step": 18660 + }, + { + "epoch": 14.904952076677317, + "grad_norm": 0.052001953125, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 18661 + }, + { + "epoch": 14.905750798722044, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 18662 + }, + { + "epoch": 14.906549520766774, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 18663 + }, + { + "epoch": 14.907348242811501, + "grad_norm": 0.056884765625, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 18664 + }, + { + "epoch": 14.90814696485623, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 18665 + }, + { + "epoch": 14.908945686900958, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0175, + "step": 18666 + }, + { + "epoch": 14.909744408945686, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0188, + "step": 18667 + }, + { + "epoch": 14.910543130990416, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.0139, + "step": 18668 + }, + { + "epoch": 14.911341853035143, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0173, + "step": 18669 + }, + { + "epoch": 14.912140575079873, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 18670 + }, + { + "epoch": 14.9129392971246, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0131, + "step": 18671 + }, + { + "epoch": 14.91373801916933, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 18672 + }, + { + "epoch": 14.914536741214057, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 18673 + }, + { + "epoch": 14.915335463258787, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 18674 + }, + { + "epoch": 14.916134185303514, + "grad_norm": 0.11181640625, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 18675 + }, + { + "epoch": 14.916932907348244, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 18676 + }, + { + "epoch": 14.917731629392971, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0145, + "step": 18677 + }, + { + "epoch": 14.918530351437699, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 18678 + }, + { + "epoch": 14.919329073482428, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 18679 + }, + { + "epoch": 14.920127795527156, + "grad_norm": 0.1689453125, + "learning_rate": 0.0005, + "loss": 1.0095, + "step": 18680 + }, + { + "epoch": 14.920926517571885, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 18681 + }, + { + "epoch": 14.921725239616613, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 18682 + }, + { + "epoch": 14.922523961661343, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 18683 + }, + { + "epoch": 14.92332268370607, + "grad_norm": 0.162109375, + "learning_rate": 0.0005, + "loss": 1.0143, + "step": 18684 + }, + { + "epoch": 14.9241214057508, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 18685 + }, + { + "epoch": 14.924920127795527, + "grad_norm": 0.1611328125, + "learning_rate": 0.0005, + "loss": 1.0182, + "step": 18686 + }, + { + "epoch": 14.925718849840255, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0188, + "step": 18687 + }, + { + "epoch": 14.926517571884984, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.016, + "step": 18688 + }, + { + "epoch": 14.927316293929712, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 18689 + }, + { + "epoch": 14.928115015974441, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 18690 + }, + { + "epoch": 14.928913738019169, + "grad_norm": 0.1826171875, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 18691 + }, + { + "epoch": 14.929712460063898, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 18692 + }, + { + "epoch": 14.930511182108626, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 18693 + }, + { + "epoch": 14.931309904153355, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0137, + "step": 18694 + }, + { + "epoch": 14.932108626198083, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 18695 + }, + { + "epoch": 14.93290734824281, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0301, + "step": 18696 + }, + { + "epoch": 14.93370607028754, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0135, + "step": 18697 + }, + { + "epoch": 14.934504792332268, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 18698 + }, + { + "epoch": 14.935303514376997, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0162, + "step": 18699 + }, + { + "epoch": 14.936102236421725, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0144, + "step": 18700 + }, + { + "epoch": 14.936900958466454, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0166, + "step": 18701 + }, + { + "epoch": 14.937699680511182, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0155, + "step": 18702 + }, + { + "epoch": 14.938498402555911, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.013, + "step": 18703 + }, + { + "epoch": 14.939297124600639, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 18704 + }, + { + "epoch": 14.940095846645367, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0161, + "step": 18705 + }, + { + "epoch": 14.940894568690096, + "grad_norm": 0.1640625, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 18706 + }, + { + "epoch": 14.941693290734824, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.011, + "step": 18707 + }, + { + "epoch": 14.942492012779553, + "grad_norm": 0.1357421875, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 18708 + }, + { + "epoch": 14.94329073482428, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 18709 + }, + { + "epoch": 14.94408945686901, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0161, + "step": 18710 + }, + { + "epoch": 14.944888178913738, + "grad_norm": 0.17578125, + "learning_rate": 0.0005, + "loss": 1.0155, + "step": 18711 + }, + { + "epoch": 14.945686900958467, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 18712 + }, + { + "epoch": 14.946485623003195, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.0182, + "step": 18713 + }, + { + "epoch": 14.947284345047922, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 18714 + }, + { + "epoch": 14.948083067092652, + "grad_norm": 0.1220703125, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 18715 + }, + { + "epoch": 14.94888178913738, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 18716 + }, + { + "epoch": 14.949680511182109, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0097, + "step": 18717 + }, + { + "epoch": 14.950479233226837, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 18718 + }, + { + "epoch": 14.951277955271566, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0157, + "step": 18719 + }, + { + "epoch": 14.952076677316294, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 18720 + }, + { + "epoch": 14.952875399361023, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0177, + "step": 18721 + }, + { + "epoch": 14.95367412140575, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 18722 + }, + { + "epoch": 14.954472843450478, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 18723 + }, + { + "epoch": 14.955271565495208, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 18724 + }, + { + "epoch": 14.956070287539935, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 18725 + }, + { + "epoch": 14.956869009584665, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0171, + "step": 18726 + }, + { + "epoch": 14.957667731629392, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 18727 + }, + { + "epoch": 14.958466453674122, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 18728 + }, + { + "epoch": 14.95926517571885, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0123, + "step": 18729 + }, + { + "epoch": 14.960063897763579, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 18730 + }, + { + "epoch": 14.960862619808307, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0134, + "step": 18731 + }, + { + "epoch": 14.961661341853034, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 18732 + }, + { + "epoch": 14.962460063897764, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0113, + "step": 18733 + }, + { + "epoch": 14.963258785942491, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 18734 + }, + { + "epoch": 14.96405750798722, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 18735 + }, + { + "epoch": 14.964856230031948, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 18736 + }, + { + "epoch": 14.965654952076678, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 18737 + }, + { + "epoch": 14.966453674121405, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 18738 + }, + { + "epoch": 14.967252396166135, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0358, + "step": 18739 + }, + { + "epoch": 14.968051118210862, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0122, + "step": 18740 + }, + { + "epoch": 14.968849840255592, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 18741 + }, + { + "epoch": 14.96964856230032, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 18742 + }, + { + "epoch": 14.970447284345047, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0185, + "step": 18743 + }, + { + "epoch": 14.971246006389777, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0169, + "step": 18744 + }, + { + "epoch": 14.972044728434504, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 18745 + }, + { + "epoch": 14.972843450479234, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 18746 + }, + { + "epoch": 14.973642172523961, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 18747 + }, + { + "epoch": 14.97444089456869, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 18748 + }, + { + "epoch": 14.975239616613418, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 18749 + }, + { + "epoch": 14.976038338658148, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 18750 + }, + { + "epoch": 14.976837060702875, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 18751 + }, + { + "epoch": 14.977635782747605, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 18752 + }, + { + "epoch": 14.978434504792332, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 18753 + }, + { + "epoch": 14.97923322683706, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0149, + "step": 18754 + }, + { + "epoch": 14.98003194888179, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 18755 + }, + { + "epoch": 14.980830670926517, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0314, + "step": 18756 + }, + { + "epoch": 14.981629392971247, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 18757 + }, + { + "epoch": 14.982428115015974, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 18758 + }, + { + "epoch": 14.983226837060704, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 18759 + }, + { + "epoch": 14.984025559105431, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 18760 + }, + { + "epoch": 14.98482428115016, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 18761 + }, + { + "epoch": 14.985623003194888, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 18762 + }, + { + "epoch": 14.986421725239616, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 18763 + }, + { + "epoch": 14.987220447284345, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 18764 + }, + { + "epoch": 14.988019169329073, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0156, + "step": 18765 + }, + { + "epoch": 14.988817891373802, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0282, + "step": 18766 + }, + { + "epoch": 14.98961661341853, + "grad_norm": 0.119140625, + "learning_rate": 0.0005, + "loss": 1.0284, + "step": 18767 + }, + { + "epoch": 14.99041533546326, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0165, + "step": 18768 + }, + { + "epoch": 14.991214057507987, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 18769 + }, + { + "epoch": 14.992012779552716, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 18770 + }, + { + "epoch": 14.992811501597444, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 18771 + }, + { + "epoch": 14.993610223642172, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 18772 + }, + { + "epoch": 14.994408945686901, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0147, + "step": 18773 + }, + { + "epoch": 14.995207667731629, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 18774 + }, + { + "epoch": 14.996006389776358, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 18775 + }, + { + "epoch": 14.996805111821086, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0141, + "step": 18776 + }, + { + "epoch": 14.997603833865815, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0178, + "step": 18777 + }, + { + "epoch": 14.998402555910543, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0177, + "step": 18778 + }, + { + "epoch": 14.999201277955272, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0183, + "step": 18779 + }, + { + "epoch": 15.0, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0199, + "step": 18780 + }, + { + "epoch": 15.000798722044728, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0063, + "step": 18781 + }, + { + "epoch": 15.001597444089457, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0134, + "step": 18782 + }, + { + "epoch": 15.002396166134185, + "grad_norm": 0.13671875, + "learning_rate": 0.0005, + "loss": 1.0138, + "step": 18783 + }, + { + "epoch": 15.003194888178914, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 18784 + }, + { + "epoch": 15.003993610223642, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 18785 + }, + { + "epoch": 15.004792332268371, + "grad_norm": 0.12353515625, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 18786 + }, + { + "epoch": 15.005591054313099, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 18787 + }, + { + "epoch": 15.006389776357828, + "grad_norm": 0.1572265625, + "learning_rate": 0.0005, + "loss": 1.0137, + "step": 18788 + }, + { + "epoch": 15.007188498402556, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0168, + "step": 18789 + }, + { + "epoch": 15.007987220447284, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0183, + "step": 18790 + }, + { + "epoch": 15.008785942492013, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0179, + "step": 18791 + }, + { + "epoch": 15.00958466453674, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 18792 + }, + { + "epoch": 15.01038338658147, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 18793 + }, + { + "epoch": 15.011182108626198, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0178, + "step": 18794 + }, + { + "epoch": 15.011980830670927, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 18795 + }, + { + "epoch": 15.012779552715655, + "grad_norm": 0.1630859375, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 18796 + }, + { + "epoch": 15.013578274760384, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 18797 + }, + { + "epoch": 15.014376996805112, + "grad_norm": 0.134765625, + "learning_rate": 0.0005, + "loss": 1.0187, + "step": 18798 + }, + { + "epoch": 15.01517571884984, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0173, + "step": 18799 + }, + { + "epoch": 15.015974440894569, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0199, + "step": 18800 + }, + { + "epoch": 15.016773162939296, + "grad_norm": 0.10205078125, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 18801 + }, + { + "epoch": 15.017571884984026, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0138, + "step": 18802 + }, + { + "epoch": 15.018370607028753, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.011, + "step": 18803 + }, + { + "epoch": 15.019169329073483, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0169, + "step": 18804 + }, + { + "epoch": 15.01996805111821, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0177, + "step": 18805 + }, + { + "epoch": 15.02076677316294, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0159, + "step": 18806 + }, + { + "epoch": 15.021565495207668, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 18807 + }, + { + "epoch": 15.022364217252397, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 18808 + }, + { + "epoch": 15.023162939297125, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 18809 + }, + { + "epoch": 15.023961661341852, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 18810 + }, + { + "epoch": 15.024760383386582, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0118, + "step": 18811 + }, + { + "epoch": 15.02555910543131, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0035, + "step": 18812 + }, + { + "epoch": 15.026357827476039, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 18813 + }, + { + "epoch": 15.027156549520766, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0145, + "step": 18814 + }, + { + "epoch": 15.027955271565496, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 18815 + }, + { + "epoch": 15.028753993610223, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0192, + "step": 18816 + }, + { + "epoch": 15.029552715654953, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 18817 + }, + { + "epoch": 15.03035143769968, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0176, + "step": 18818 + }, + { + "epoch": 15.031150159744408, + "grad_norm": 0.15625, + "learning_rate": 0.0005, + "loss": 1.0158, + "step": 18819 + }, + { + "epoch": 15.031948881789138, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0196, + "step": 18820 + }, + { + "epoch": 15.032747603833865, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 18821 + }, + { + "epoch": 15.033546325878595, + "grad_norm": 0.12451171875, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 18822 + }, + { + "epoch": 15.034345047923322, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 18823 + }, + { + "epoch": 15.035143769968052, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0112, + "step": 18824 + }, + { + "epoch": 15.03594249201278, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0173, + "step": 18825 + }, + { + "epoch": 15.036741214057509, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 18826 + }, + { + "epoch": 15.037539936102236, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 18827 + }, + { + "epoch": 15.038338658146964, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 18828 + }, + { + "epoch": 15.039137380191693, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 18829 + }, + { + "epoch": 15.039936102236421, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 18830 + }, + { + "epoch": 15.04073482428115, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 18831 + }, + { + "epoch": 15.041533546325878, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0111, + "step": 18832 + }, + { + "epoch": 15.042332268370608, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0118, + "step": 18833 + }, + { + "epoch": 15.043130990415335, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 18834 + }, + { + "epoch": 15.043929712460065, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0327, + "step": 18835 + }, + { + "epoch": 15.044728434504792, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0156, + "step": 18836 + }, + { + "epoch": 15.04552715654952, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 18837 + }, + { + "epoch": 15.04632587859425, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 18838 + }, + { + "epoch": 15.047124600638977, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 18839 + }, + { + "epoch": 15.047923322683706, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 18840 + }, + { + "epoch": 15.048722044728434, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 18841 + }, + { + "epoch": 15.049520766773163, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 18842 + }, + { + "epoch": 15.050319488817891, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0131, + "step": 18843 + }, + { + "epoch": 15.05111821086262, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0177, + "step": 18844 + }, + { + "epoch": 15.051916932907348, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0181, + "step": 18845 + }, + { + "epoch": 15.052715654952078, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 18846 + }, + { + "epoch": 15.053514376996805, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 18847 + }, + { + "epoch": 15.054313099041533, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 18848 + }, + { + "epoch": 15.055111821086262, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 18849 + }, + { + "epoch": 15.05591054313099, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 18850 + }, + { + "epoch": 15.05670926517572, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0205, + "step": 18851 + }, + { + "epoch": 15.057507987220447, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0131, + "step": 18852 + }, + { + "epoch": 15.058306709265176, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 18853 + }, + { + "epoch": 15.059105431309904, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 18854 + }, + { + "epoch": 15.059904153354633, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 18855 + }, + { + "epoch": 15.060702875399361, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 18856 + }, + { + "epoch": 15.061501597444089, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0118, + "step": 18857 + }, + { + "epoch": 15.062300319488818, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 18858 + }, + { + "epoch": 15.063099041533546, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 18859 + }, + { + "epoch": 15.063897763578275, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 18860 + }, + { + "epoch": 15.064696485623003, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 18861 + }, + { + "epoch": 15.065495207667732, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0165, + "step": 18862 + }, + { + "epoch": 15.06629392971246, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0244, + "step": 18863 + }, + { + "epoch": 15.06709265175719, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.018, + "step": 18864 + }, + { + "epoch": 15.067891373801917, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 18865 + }, + { + "epoch": 15.068690095846645, + "grad_norm": 0.1103515625, + "learning_rate": 0.0005, + "loss": 1.0124, + "step": 18866 + }, + { + "epoch": 15.069488817891374, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 18867 + }, + { + "epoch": 15.070287539936102, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 18868 + }, + { + "epoch": 15.071086261980831, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 18869 + }, + { + "epoch": 15.071884984025559, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0165, + "step": 18870 + }, + { + "epoch": 15.072683706070288, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 18871 + }, + { + "epoch": 15.073482428115016, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 18872 + }, + { + "epoch": 15.074281150159745, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 18873 + }, + { + "epoch": 15.075079872204473, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 18874 + }, + { + "epoch": 15.0758785942492, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 18875 + }, + { + "epoch": 15.07667731629393, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 18876 + }, + { + "epoch": 15.077476038338657, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0089, + "step": 18877 + }, + { + "epoch": 15.078274760383387, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 18878 + }, + { + "epoch": 15.079073482428115, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 18879 + }, + { + "epoch": 15.079872204472844, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 18880 + }, + { + "epoch": 15.080670926517572, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0148, + "step": 18881 + }, + { + "epoch": 15.081469648562301, + "grad_norm": 0.057861328125, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 18882 + }, + { + "epoch": 15.082268370607029, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 18883 + }, + { + "epoch": 15.083067092651758, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0277, + "step": 18884 + }, + { + "epoch": 15.083865814696486, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0155, + "step": 18885 + }, + { + "epoch": 15.084664536741213, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.015, + "step": 18886 + }, + { + "epoch": 15.085463258785943, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 18887 + }, + { + "epoch": 15.08626198083067, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 18888 + }, + { + "epoch": 15.0870607028754, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 18889 + }, + { + "epoch": 15.087859424920127, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 18890 + }, + { + "epoch": 15.088658146964857, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0138, + "step": 18891 + }, + { + "epoch": 15.089456869009584, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0144, + "step": 18892 + }, + { + "epoch": 15.090255591054314, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.0336, + "step": 18893 + }, + { + "epoch": 15.091054313099042, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 18894 + }, + { + "epoch": 15.09185303514377, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 18895 + }, + { + "epoch": 15.092651757188499, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.018, + "step": 18896 + }, + { + "epoch": 15.093450479233226, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 18897 + }, + { + "epoch": 15.094249201277956, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 18898 + }, + { + "epoch": 15.095047923322683, + "grad_norm": 0.11572265625, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 18899 + }, + { + "epoch": 15.095846645367413, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0126, + "step": 18900 + }, + { + "epoch": 15.09664536741214, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0126, + "step": 18901 + }, + { + "epoch": 15.09744408945687, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0188, + "step": 18902 + }, + { + "epoch": 15.098242811501597, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0097, + "step": 18903 + }, + { + "epoch": 15.099041533546325, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 18904 + }, + { + "epoch": 15.099840255591054, + "grad_norm": 0.123046875, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 18905 + }, + { + "epoch": 15.100638977635782, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 18906 + }, + { + "epoch": 15.101437699680512, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0179, + "step": 18907 + }, + { + "epoch": 15.10223642172524, + "grad_norm": 0.1162109375, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 18908 + }, + { + "epoch": 15.103035143769969, + "grad_norm": 0.1181640625, + "learning_rate": 0.0005, + "loss": 1.0278, + "step": 18909 + }, + { + "epoch": 15.103833865814696, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 18910 + }, + { + "epoch": 15.104632587859426, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0187, + "step": 18911 + }, + { + "epoch": 15.105431309904153, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 18912 + }, + { + "epoch": 15.106230031948881, + "grad_norm": 0.1015625, + "learning_rate": 0.0005, + "loss": 1.0205, + "step": 18913 + }, + { + "epoch": 15.10702875399361, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 18914 + }, + { + "epoch": 15.107827476038338, + "grad_norm": 0.09521484375, + "learning_rate": 0.0005, + "loss": 1.0178, + "step": 18915 + }, + { + "epoch": 15.108626198083067, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 18916 + }, + { + "epoch": 15.109424920127795, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 18917 + }, + { + "epoch": 15.110223642172524, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 18918 + }, + { + "epoch": 15.111022364217252, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0164, + "step": 18919 + }, + { + "epoch": 15.111821086261982, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 18920 + }, + { + "epoch": 15.11261980830671, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 18921 + }, + { + "epoch": 15.113418530351439, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0113, + "step": 18922 + }, + { + "epoch": 15.114217252396166, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0167, + "step": 18923 + }, + { + "epoch": 15.115015974440894, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 18924 + }, + { + "epoch": 15.115814696485623, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.018, + "step": 18925 + }, + { + "epoch": 15.116613418530351, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 18926 + }, + { + "epoch": 15.11741214057508, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 18927 + }, + { + "epoch": 15.118210862619808, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0195, + "step": 18928 + }, + { + "epoch": 15.119009584664537, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0303, + "step": 18929 + }, + { + "epoch": 15.119808306709265, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 18930 + }, + { + "epoch": 15.120607028753994, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0162, + "step": 18931 + }, + { + "epoch": 15.121405750798722, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 18932 + }, + { + "epoch": 15.12220447284345, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0097, + "step": 18933 + }, + { + "epoch": 15.12300319488818, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 18934 + }, + { + "epoch": 15.123801916932907, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.013, + "step": 18935 + }, + { + "epoch": 15.124600638977636, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0177, + "step": 18936 + }, + { + "epoch": 15.125399361022364, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0185, + "step": 18937 + }, + { + "epoch": 15.126198083067093, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 18938 + }, + { + "epoch": 15.12699680511182, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0085, + "step": 18939 + }, + { + "epoch": 15.12779552715655, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0137, + "step": 18940 + }, + { + "epoch": 15.128594249201278, + "grad_norm": 0.12255859375, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 18941 + }, + { + "epoch": 15.129392971246006, + "grad_norm": 0.10302734375, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 18942 + }, + { + "epoch": 15.130191693290735, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0183, + "step": 18943 + }, + { + "epoch": 15.130990415335463, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 18944 + }, + { + "epoch": 15.131789137380192, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 18945 + }, + { + "epoch": 15.13258785942492, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0318, + "step": 18946 + }, + { + "epoch": 15.133386581469649, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0156, + "step": 18947 + }, + { + "epoch": 15.134185303514377, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 18948 + }, + { + "epoch": 15.134984025559106, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 18949 + }, + { + "epoch": 15.135782747603834, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 18950 + }, + { + "epoch": 15.136581469648561, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0165, + "step": 18951 + }, + { + "epoch": 15.13738019169329, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0185, + "step": 18952 + }, + { + "epoch": 15.138178913738018, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 18953 + }, + { + "epoch": 15.138977635782748, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 18954 + }, + { + "epoch": 15.139776357827476, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 18955 + }, + { + "epoch": 15.140575079872205, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 18956 + }, + { + "epoch": 15.141373801916933, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0152, + "step": 18957 + }, + { + "epoch": 15.142172523961662, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0165, + "step": 18958 + }, + { + "epoch": 15.14297124600639, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 18959 + }, + { + "epoch": 15.143769968051119, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0169, + "step": 18960 + }, + { + "epoch": 15.144568690095847, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.017, + "step": 18961 + }, + { + "epoch": 15.145367412140574, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0207, + "step": 18962 + }, + { + "epoch": 15.146166134185304, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 18963 + }, + { + "epoch": 15.146964856230031, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0171, + "step": 18964 + }, + { + "epoch": 15.14776357827476, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 18965 + }, + { + "epoch": 15.148562300319488, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0153, + "step": 18966 + }, + { + "epoch": 15.149361022364218, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 18967 + }, + { + "epoch": 15.150159744408946, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 18968 + }, + { + "epoch": 15.150958466453675, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 18969 + }, + { + "epoch": 15.151757188498403, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0199, + "step": 18970 + }, + { + "epoch": 15.15255591054313, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 18971 + }, + { + "epoch": 15.15335463258786, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 18972 + }, + { + "epoch": 15.154153354632587, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 18973 + }, + { + "epoch": 15.154952076677317, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 18974 + }, + { + "epoch": 15.155750798722044, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0127, + "step": 18975 + }, + { + "epoch": 15.156549520766774, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 18976 + }, + { + "epoch": 15.157348242811501, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0178, + "step": 18977 + }, + { + "epoch": 15.15814696485623, + "grad_norm": 0.058349609375, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 18978 + }, + { + "epoch": 15.158945686900958, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0098, + "step": 18979 + }, + { + "epoch": 15.159744408945686, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0141, + "step": 18980 + }, + { + "epoch": 15.160543130990416, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0182, + "step": 18981 + }, + { + "epoch": 15.161341853035143, + "grad_norm": 0.11279296875, + "learning_rate": 0.0005, + "loss": 1.0119, + "step": 18982 + }, + { + "epoch": 15.162140575079873, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0163, + "step": 18983 + }, + { + "epoch": 15.1629392971246, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0161, + "step": 18984 + }, + { + "epoch": 15.16373801916933, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0166, + "step": 18985 + }, + { + "epoch": 15.164536741214057, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 18986 + }, + { + "epoch": 15.165335463258787, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0115, + "step": 18987 + }, + { + "epoch": 15.166134185303514, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0147, + "step": 18988 + }, + { + "epoch": 15.166932907348242, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 18989 + }, + { + "epoch": 15.167731629392971, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0111, + "step": 18990 + }, + { + "epoch": 15.168530351437699, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0157, + "step": 18991 + }, + { + "epoch": 15.169329073482428, + "grad_norm": 0.0986328125, + "learning_rate": 0.0005, + "loss": 1.0175, + "step": 18992 + }, + { + "epoch": 15.170127795527156, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 18993 + }, + { + "epoch": 15.170926517571885, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 18994 + }, + { + "epoch": 15.171725239616613, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 18995 + }, + { + "epoch": 15.172523961661343, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0174, + "step": 18996 + }, + { + "epoch": 15.17332268370607, + "grad_norm": 0.05712890625, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 18997 + }, + { + "epoch": 15.1741214057508, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 18998 + }, + { + "epoch": 15.174920127795527, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 18999 + }, + { + "epoch": 15.175718849840255, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 19000 + }, + { + "epoch": 15.176517571884984, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0181, + "step": 19001 + }, + { + "epoch": 15.177316293929712, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0171, + "step": 19002 + }, + { + "epoch": 15.178115015974441, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0187, + "step": 19003 + }, + { + "epoch": 15.178913738019169, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0286, + "step": 19004 + }, + { + "epoch": 15.179712460063898, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 19005 + }, + { + "epoch": 15.180511182108626, + "grad_norm": 0.06201171875, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 19006 + }, + { + "epoch": 15.181309904153355, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 19007 + }, + { + "epoch": 15.182108626198083, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 19008 + }, + { + "epoch": 15.18290734824281, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0142, + "step": 19009 + }, + { + "epoch": 15.18370607028754, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0157, + "step": 19010 + }, + { + "epoch": 15.184504792332268, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 19011 + }, + { + "epoch": 15.185303514376997, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0133, + "step": 19012 + }, + { + "epoch": 15.186102236421725, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 19013 + }, + { + "epoch": 15.186900958466454, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0192, + "step": 19014 + }, + { + "epoch": 15.187699680511182, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0165, + "step": 19015 + }, + { + "epoch": 15.188498402555911, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0152, + "step": 19016 + }, + { + "epoch": 15.189297124600639, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 19017 + }, + { + "epoch": 15.190095846645367, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 19018 + }, + { + "epoch": 15.190894568690096, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 19019 + }, + { + "epoch": 15.191693290734824, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0188, + "step": 19020 + }, + { + "epoch": 15.192492012779553, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0168, + "step": 19021 + }, + { + "epoch": 15.19329073482428, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0171, + "step": 19022 + }, + { + "epoch": 15.19408945686901, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 19023 + }, + { + "epoch": 15.194888178913738, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0128, + "step": 19024 + }, + { + "epoch": 15.195686900958467, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.018, + "step": 19025 + }, + { + "epoch": 15.196485623003195, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 19026 + }, + { + "epoch": 15.197284345047922, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0179, + "step": 19027 + }, + { + "epoch": 15.198083067092652, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0296, + "step": 19028 + }, + { + "epoch": 15.19888178913738, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 19029 + }, + { + "epoch": 15.199680511182109, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0134, + "step": 19030 + }, + { + "epoch": 15.200479233226837, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 19031 + }, + { + "epoch": 15.201277955271566, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 19032 + }, + { + "epoch": 15.202076677316294, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 19033 + }, + { + "epoch": 15.202875399361023, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0149, + "step": 19034 + }, + { + "epoch": 15.20367412140575, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0145, + "step": 19035 + }, + { + "epoch": 15.204472843450478, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0196, + "step": 19036 + }, + { + "epoch": 15.205271565495208, + "grad_norm": 0.099609375, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 19037 + }, + { + "epoch": 15.206070287539935, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0173, + "step": 19038 + }, + { + "epoch": 15.206869009584665, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.023, + "step": 19039 + }, + { + "epoch": 15.207667731629392, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 19040 + }, + { + "epoch": 15.208466453674122, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0081, + "step": 19041 + }, + { + "epoch": 15.20926517571885, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0132, + "step": 19042 + }, + { + "epoch": 15.210063897763579, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 19043 + }, + { + "epoch": 15.210862619808307, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0178, + "step": 19044 + }, + { + "epoch": 15.211661341853036, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 19045 + }, + { + "epoch": 15.212460063897764, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 19046 + }, + { + "epoch": 15.213258785942491, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0196, + "step": 19047 + }, + { + "epoch": 15.21405750798722, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 19048 + }, + { + "epoch": 15.214856230031948, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 19049 + }, + { + "epoch": 15.215654952076678, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 19050 + }, + { + "epoch": 15.216453674121405, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0147, + "step": 19051 + }, + { + "epoch": 15.217252396166135, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 19052 + }, + { + "epoch": 15.218051118210862, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 19053 + }, + { + "epoch": 15.218849840255592, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 19054 + }, + { + "epoch": 15.21964856230032, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0179, + "step": 19055 + }, + { + "epoch": 15.220447284345047, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 19056 + }, + { + "epoch": 15.221246006389777, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 19057 + }, + { + "epoch": 15.222044728434504, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0163, + "step": 19058 + }, + { + "epoch": 15.222843450479234, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0169, + "step": 19059 + }, + { + "epoch": 15.223642172523961, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 19060 + }, + { + "epoch": 15.22444089456869, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0222, + "step": 19061 + }, + { + "epoch": 15.225239616613418, + "grad_norm": 0.12109375, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 19062 + }, + { + "epoch": 15.226038338658148, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0192, + "step": 19063 + }, + { + "epoch": 15.226837060702875, + "grad_norm": 0.154296875, + "learning_rate": 0.0005, + "loss": 1.0157, + "step": 19064 + }, + { + "epoch": 15.227635782747603, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0143, + "step": 19065 + }, + { + "epoch": 15.228434504792332, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 19066 + }, + { + "epoch": 15.22923322683706, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0157, + "step": 19067 + }, + { + "epoch": 15.23003194888179, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0173, + "step": 19068 + }, + { + "epoch": 15.230830670926517, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0172, + "step": 19069 + }, + { + "epoch": 15.231629392971247, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 19070 + }, + { + "epoch": 15.232428115015974, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 19071 + }, + { + "epoch": 15.233226837060704, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 19072 + }, + { + "epoch": 15.234025559105431, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 19073 + }, + { + "epoch": 15.23482428115016, + "grad_norm": 0.126953125, + "learning_rate": 0.0005, + "loss": 1.0139, + "step": 19074 + }, + { + "epoch": 15.235623003194888, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0135, + "step": 19075 + }, + { + "epoch": 15.236421725239616, + "grad_norm": 0.11083984375, + "learning_rate": 0.0005, + "loss": 1.0171, + "step": 19076 + }, + { + "epoch": 15.237220447284345, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 19077 + }, + { + "epoch": 15.238019169329073, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 19078 + }, + { + "epoch": 15.238817891373802, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0246, + "step": 19079 + }, + { + "epoch": 15.23961661341853, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 19080 + }, + { + "epoch": 15.24041533546326, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 19081 + }, + { + "epoch": 15.241214057507987, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 19082 + }, + { + "epoch": 15.242012779552716, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0114, + "step": 19083 + }, + { + "epoch": 15.242811501597444, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 19084 + }, + { + "epoch": 15.243610223642172, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 19085 + }, + { + "epoch": 15.244408945686901, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.018, + "step": 19086 + }, + { + "epoch": 15.245207667731629, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.01, + "step": 19087 + }, + { + "epoch": 15.246006389776358, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 19088 + }, + { + "epoch": 15.246805111821086, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 19089 + }, + { + "epoch": 15.247603833865815, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 19090 + }, + { + "epoch": 15.248402555910543, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 19091 + }, + { + "epoch": 15.249201277955272, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0175, + "step": 19092 + }, + { + "epoch": 15.25, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.014, + "step": 19093 + }, + { + "epoch": 15.250798722044728, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.024, + "step": 19094 + }, + { + "epoch": 15.251597444089457, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0177, + "step": 19095 + }, + { + "epoch": 15.252396166134185, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 19096 + }, + { + "epoch": 15.253194888178914, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0202, + "step": 19097 + }, + { + "epoch": 15.253993610223642, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.014, + "step": 19098 + }, + { + "epoch": 15.254792332268371, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0192, + "step": 19099 + }, + { + "epoch": 15.255591054313099, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 19100 + }, + { + "epoch": 15.256389776357828, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.011, + "step": 19101 + }, + { + "epoch": 15.257188498402556, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 19102 + }, + { + "epoch": 15.257987220447284, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 19103 + }, + { + "epoch": 15.258785942492013, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 19104 + }, + { + "epoch": 15.25958466453674, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.0159, + "step": 19105 + }, + { + "epoch": 15.26038338658147, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 19106 + }, + { + "epoch": 15.261182108626198, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0192, + "step": 19107 + }, + { + "epoch": 15.261980830670927, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0195, + "step": 19108 + }, + { + "epoch": 15.262779552715655, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0093, + "step": 19109 + }, + { + "epoch": 15.263578274760384, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 19110 + }, + { + "epoch": 15.264376996805112, + "grad_norm": 0.11376953125, + "learning_rate": 0.0005, + "loss": 1.017, + "step": 19111 + }, + { + "epoch": 15.26517571884984, + "grad_norm": 0.146484375, + "learning_rate": 0.0005, + "loss": 1.0155, + "step": 19112 + }, + { + "epoch": 15.265974440894569, + "grad_norm": 0.0908203125, + "learning_rate": 0.0005, + "loss": 1.0181, + "step": 19113 + }, + { + "epoch": 15.266773162939296, + "grad_norm": 0.1494140625, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 19114 + }, + { + "epoch": 15.267571884984026, + "grad_norm": 0.0966796875, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 19115 + }, + { + "epoch": 15.268370607028753, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 19116 + }, + { + "epoch": 15.269169329073483, + "grad_norm": 0.1259765625, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 19117 + }, + { + "epoch": 15.26996805111821, + "grad_norm": 0.10009765625, + "learning_rate": 0.0005, + "loss": 1.0166, + "step": 19118 + }, + { + "epoch": 15.27076677316294, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.013, + "step": 19119 + }, + { + "epoch": 15.271565495207668, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 19120 + }, + { + "epoch": 15.272364217252397, + "grad_norm": 0.11328125, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 19121 + }, + { + "epoch": 15.273162939297125, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0109, + "step": 19122 + }, + { + "epoch": 15.273961661341852, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0337, + "step": 19123 + }, + { + "epoch": 15.274760383386582, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 19124 + }, + { + "epoch": 15.27555910543131, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 19125 + }, + { + "epoch": 15.276357827476039, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 19126 + }, + { + "epoch": 15.277156549520766, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 19127 + }, + { + "epoch": 15.277955271565496, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0167, + "step": 19128 + }, + { + "epoch": 15.278753993610223, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 19129 + }, + { + "epoch": 15.279552715654953, + "grad_norm": 0.0849609375, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 19130 + }, + { + "epoch": 15.28035143769968, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0223, + "step": 19131 + }, + { + "epoch": 15.281150159744408, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 19132 + }, + { + "epoch": 15.281948881789138, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0188, + "step": 19133 + }, + { + "epoch": 15.282747603833865, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 19134 + }, + { + "epoch": 15.283546325878595, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0259, + "step": 19135 + }, + { + "epoch": 15.284345047923322, + "grad_norm": 0.107421875, + "learning_rate": 0.0005, + "loss": 1.0196, + "step": 19136 + }, + { + "epoch": 15.285143769968052, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0187, + "step": 19137 + }, + { + "epoch": 15.28594249201278, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 19138 + }, + { + "epoch": 15.286741214057509, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 19139 + }, + { + "epoch": 15.287539936102236, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0199, + "step": 19140 + }, + { + "epoch": 15.288338658146964, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0171, + "step": 19141 + }, + { + "epoch": 15.289137380191693, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 19142 + }, + { + "epoch": 15.289936102236421, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0139, + "step": 19143 + }, + { + "epoch": 15.29073482428115, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0135, + "step": 19144 + }, + { + "epoch": 15.291533546325878, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0173, + "step": 19145 + }, + { + "epoch": 15.292332268370608, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 19146 + }, + { + "epoch": 15.293130990415335, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0139, + "step": 19147 + }, + { + "epoch": 15.293929712460065, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0154, + "step": 19148 + }, + { + "epoch": 15.294728434504792, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0108, + "step": 19149 + }, + { + "epoch": 15.295527156549522, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 19150 + }, + { + "epoch": 15.29632587859425, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 19151 + }, + { + "epoch": 15.297124600638977, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 19152 + }, + { + "epoch": 15.297923322683706, + "grad_norm": 0.1513671875, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 19153 + }, + { + "epoch": 15.298722044728434, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 19154 + }, + { + "epoch": 15.299520766773163, + "grad_norm": 0.142578125, + "learning_rate": 0.0005, + "loss": 1.0153, + "step": 19155 + }, + { + "epoch": 15.300319488817891, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0185, + "step": 19156 + }, + { + "epoch": 15.30111821086262, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0154, + "step": 19157 + }, + { + "epoch": 15.301916932907348, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 19158 + }, + { + "epoch": 15.302715654952078, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0142, + "step": 19159 + }, + { + "epoch": 15.303514376996805, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 19160 + }, + { + "epoch": 15.304313099041533, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 19161 + }, + { + "epoch": 15.305111821086262, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0166, + "step": 19162 + }, + { + "epoch": 15.30591054313099, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 19163 + }, + { + "epoch": 15.30670926517572, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 19164 + }, + { + "epoch": 15.307507987220447, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 19165 + }, + { + "epoch": 15.308306709265176, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.011, + "step": 19166 + }, + { + "epoch": 15.309105431309904, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.018, + "step": 19167 + }, + { + "epoch": 15.309904153354633, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0167, + "step": 19168 + }, + { + "epoch": 15.310702875399361, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0137, + "step": 19169 + }, + { + "epoch": 15.311501597444089, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.016, + "step": 19170 + }, + { + "epoch": 15.312300319488818, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0126, + "step": 19171 + }, + { + "epoch": 15.313099041533546, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 19172 + }, + { + "epoch": 15.313897763578275, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 19173 + }, + { + "epoch": 15.314696485623003, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0116, + "step": 19174 + }, + { + "epoch": 15.315495207667732, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0185, + "step": 19175 + }, + { + "epoch": 15.31629392971246, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0183, + "step": 19176 + }, + { + "epoch": 15.31709265175719, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0171, + "step": 19177 + }, + { + "epoch": 15.317891373801917, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 19178 + }, + { + "epoch": 15.318690095846645, + "grad_norm": 0.130859375, + "learning_rate": 0.0005, + "loss": 1.0102, + "step": 19179 + }, + { + "epoch": 15.319488817891374, + "grad_norm": 0.109375, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 19180 + }, + { + "epoch": 15.320287539936102, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 19181 + }, + { + "epoch": 15.321086261980831, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 19182 + }, + { + "epoch": 15.321884984025559, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.011, + "step": 19183 + }, + { + "epoch": 15.322683706070288, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.0204, + "step": 19184 + }, + { + "epoch": 15.323482428115016, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0169, + "step": 19185 + }, + { + "epoch": 15.324281150159745, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0127, + "step": 19186 + }, + { + "epoch": 15.325079872204473, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0307, + "step": 19187 + }, + { + "epoch": 15.3258785942492, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0167, + "step": 19188 + }, + { + "epoch": 15.32667731629393, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 19189 + }, + { + "epoch": 15.327476038338657, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0218, + "step": 19190 + }, + { + "epoch": 15.328274760383387, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0221, + "step": 19191 + }, + { + "epoch": 15.329073482428115, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 19192 + }, + { + "epoch": 15.329872204472844, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 19193 + }, + { + "epoch": 15.330670926517572, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0241, + "step": 19194 + }, + { + "epoch": 15.331469648562301, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 19195 + }, + { + "epoch": 15.332268370607029, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0227, + "step": 19196 + }, + { + "epoch": 15.333067092651758, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0326, + "step": 19197 + }, + { + "epoch": 15.333865814696486, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0163, + "step": 19198 + }, + { + "epoch": 15.334664536741213, + "grad_norm": 0.09765625, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 19199 + }, + { + "epoch": 15.335463258785943, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 19200 + }, + { + "epoch": 15.33626198083067, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0188, + "step": 19201 + }, + { + "epoch": 15.3370607028754, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 19202 + }, + { + "epoch": 15.337859424920127, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 19203 + }, + { + "epoch": 15.338658146964857, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 19204 + }, + { + "epoch": 15.339456869009584, + "grad_norm": 0.12158203125, + "learning_rate": 0.0005, + "loss": 1.0271, + "step": 19205 + }, + { + "epoch": 15.340255591054314, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 19206 + }, + { + "epoch": 15.341054313099042, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 19207 + }, + { + "epoch": 15.34185303514377, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 19208 + }, + { + "epoch": 15.342651757188499, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0199, + "step": 19209 + }, + { + "epoch": 15.343450479233226, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0145, + "step": 19210 + }, + { + "epoch": 15.344249201277956, + "grad_norm": 0.0595703125, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 19211 + }, + { + "epoch": 15.345047923322683, + "grad_norm": 0.0859375, + "learning_rate": 0.0005, + "loss": 1.0137, + "step": 19212 + }, + { + "epoch": 15.345846645367413, + "grad_norm": 0.091796875, + "learning_rate": 0.0005, + "loss": 1.0108, + "step": 19213 + }, + { + "epoch": 15.34664536741214, + "grad_norm": 0.10107421875, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 19214 + }, + { + "epoch": 15.34744408945687, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0224, + "step": 19215 + }, + { + "epoch": 15.348242811501597, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 19216 + }, + { + "epoch": 15.349041533546325, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0128, + "step": 19217 + }, + { + "epoch": 15.349840255591054, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0162, + "step": 19218 + }, + { + "epoch": 15.350638977635782, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0253, + "step": 19219 + }, + { + "epoch": 15.351437699680512, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0299, + "step": 19220 + }, + { + "epoch": 15.35223642172524, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 19221 + }, + { + "epoch": 15.353035143769969, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0169, + "step": 19222 + }, + { + "epoch": 15.353833865814696, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0177, + "step": 19223 + }, + { + "epoch": 15.354632587859426, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0197, + "step": 19224 + }, + { + "epoch": 15.355431309904153, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 19225 + }, + { + "epoch": 15.356230031948883, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 19226 + }, + { + "epoch": 15.35702875399361, + "grad_norm": 0.10888671875, + "learning_rate": 0.0005, + "loss": 1.0157, + "step": 19227 + }, + { + "epoch": 15.357827476038338, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 19228 + }, + { + "epoch": 15.358626198083067, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0269, + "step": 19229 + }, + { + "epoch": 15.359424920127795, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 19230 + }, + { + "epoch": 15.360223642172524, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 19231 + }, + { + "epoch": 15.361022364217252, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0174, + "step": 19232 + }, + { + "epoch": 15.361821086261982, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0239, + "step": 19233 + }, + { + "epoch": 15.36261980830671, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0121, + "step": 19234 + }, + { + "epoch": 15.363418530351439, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0306, + "step": 19235 + }, + { + "epoch": 15.364217252396166, + "grad_norm": 0.056640625, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 19236 + }, + { + "epoch": 15.365015974440894, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0161, + "step": 19237 + }, + { + "epoch": 15.365814696485623, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0302, + "step": 19238 + }, + { + "epoch": 15.366613418530351, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 19239 + }, + { + "epoch": 15.36741214057508, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0165, + "step": 19240 + }, + { + "epoch": 15.368210862619808, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 19241 + }, + { + "epoch": 15.369009584664537, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0168, + "step": 19242 + }, + { + "epoch": 15.369808306709265, + "grad_norm": 0.09814453125, + "learning_rate": 0.0005, + "loss": 1.0179, + "step": 19243 + }, + { + "epoch": 15.370607028753994, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 19244 + }, + { + "epoch": 15.371405750798722, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 19245 + }, + { + "epoch": 15.37220447284345, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0192, + "step": 19246 + }, + { + "epoch": 15.37300319488818, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0175, + "step": 19247 + }, + { + "epoch": 15.373801916932907, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 19248 + }, + { + "epoch": 15.374600638977636, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 19249 + }, + { + "epoch": 15.375399361022364, + "grad_norm": 0.1083984375, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 19250 + }, + { + "epoch": 15.376198083067093, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 19251 + }, + { + "epoch": 15.37699680511182, + "grad_norm": 0.171875, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 19252 + }, + { + "epoch": 15.37779552715655, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 19253 + }, + { + "epoch": 15.378594249201278, + "grad_norm": 0.16015625, + "learning_rate": 0.0005, + "loss": 1.018, + "step": 19254 + }, + { + "epoch": 15.379392971246006, + "grad_norm": 0.1201171875, + "learning_rate": 0.0005, + "loss": 1.0263, + "step": 19255 + }, + { + "epoch": 15.380191693290735, + "grad_norm": 0.111328125, + "learning_rate": 0.0005, + "loss": 1.0163, + "step": 19256 + }, + { + "epoch": 15.380990415335463, + "grad_norm": 0.11865234375, + "learning_rate": 0.0005, + "loss": 1.0147, + "step": 19257 + }, + { + "epoch": 15.381789137380192, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0183, + "step": 19258 + }, + { + "epoch": 15.38258785942492, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0171, + "step": 19259 + }, + { + "epoch": 15.383386581469649, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0203, + "step": 19260 + }, + { + "epoch": 15.384185303514377, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.031, + "step": 19261 + }, + { + "epoch": 15.384984025559106, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 19262 + }, + { + "epoch": 15.385782747603834, + "grad_norm": 0.07958984375, + "learning_rate": 0.0005, + "loss": 1.0242, + "step": 19263 + }, + { + "epoch": 15.386581469648561, + "grad_norm": 0.115234375, + "learning_rate": 0.0005, + "loss": 1.0188, + "step": 19264 + }, + { + "epoch": 15.38738019169329, + "grad_norm": 0.09912109375, + "learning_rate": 0.0005, + "loss": 1.0177, + "step": 19265 + }, + { + "epoch": 15.388178913738018, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0181, + "step": 19266 + }, + { + "epoch": 15.388977635782748, + "grad_norm": 0.09423828125, + "learning_rate": 0.0005, + "loss": 1.0179, + "step": 19267 + }, + { + "epoch": 15.389776357827476, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 19268 + }, + { + "epoch": 15.390575079872205, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 19269 + }, + { + "epoch": 15.391373801916933, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 19270 + }, + { + "epoch": 15.392172523961662, + "grad_norm": 0.087890625, + "learning_rate": 0.0005, + "loss": 1.0195, + "step": 19271 + }, + { + "epoch": 15.39297124600639, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0195, + "step": 19272 + }, + { + "epoch": 15.393769968051117, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.022, + "step": 19273 + }, + { + "epoch": 15.394568690095847, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0124, + "step": 19274 + }, + { + "epoch": 15.395367412140574, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 19275 + }, + { + "epoch": 15.396166134185304, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 19276 + }, + { + "epoch": 15.396964856230031, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0199, + "step": 19277 + }, + { + "epoch": 15.39776357827476, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0161, + "step": 19278 + }, + { + "epoch": 15.398562300319488, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 19279 + }, + { + "epoch": 15.399361022364218, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.017, + "step": 19280 + }, + { + "epoch": 15.400159744408946, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 19281 + }, + { + "epoch": 15.400958466453675, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 19282 + }, + { + "epoch": 15.401757188498403, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 19283 + }, + { + "epoch": 15.40255591054313, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 19284 + }, + { + "epoch": 15.40335463258786, + "grad_norm": 0.1005859375, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 19285 + }, + { + "epoch": 15.404153354632587, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0219, + "step": 19286 + }, + { + "epoch": 15.404952076677317, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0267, + "step": 19287 + }, + { + "epoch": 15.405750798722044, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0274, + "step": 19288 + }, + { + "epoch": 15.406549520766774, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 19289 + }, + { + "epoch": 15.407348242811501, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0165, + "step": 19290 + }, + { + "epoch": 15.40814696485623, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0133, + "step": 19291 + }, + { + "epoch": 15.408945686900958, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 19292 + }, + { + "epoch": 15.409744408945686, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0159, + "step": 19293 + }, + { + "epoch": 15.410543130990416, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.019, + "step": 19294 + }, + { + "epoch": 15.411341853035143, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 19295 + }, + { + "epoch": 15.412140575079873, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0283, + "step": 19296 + }, + { + "epoch": 15.4129392971246, + "grad_norm": 0.09716796875, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 19297 + }, + { + "epoch": 15.41373801916933, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 19298 + }, + { + "epoch": 15.414536741214057, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0171, + "step": 19299 + }, + { + "epoch": 15.415335463258787, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 19300 + }, + { + "epoch": 15.416134185303514, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 19301 + }, + { + "epoch": 15.416932907348242, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 19302 + }, + { + "epoch": 15.417731629392971, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 19303 + }, + { + "epoch": 15.418530351437699, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0161, + "step": 19304 + }, + { + "epoch": 15.419329073482428, + "grad_norm": 0.06982421875, + "learning_rate": 0.0005, + "loss": 1.0245, + "step": 19305 + }, + { + "epoch": 15.420127795527156, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 19306 + }, + { + "epoch": 15.420926517571885, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0147, + "step": 19307 + }, + { + "epoch": 15.421725239616613, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 19308 + }, + { + "epoch": 15.422523961661343, + "grad_norm": 0.095703125, + "learning_rate": 0.0005, + "loss": 1.027, + "step": 19309 + }, + { + "epoch": 15.42332268370607, + "grad_norm": 0.1142578125, + "learning_rate": 0.0005, + "loss": 1.0289, + "step": 19310 + }, + { + "epoch": 15.4241214057508, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 19311 + }, + { + "epoch": 15.424920127795527, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 19312 + }, + { + "epoch": 15.425718849840255, + "grad_norm": 0.08935546875, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 19313 + }, + { + "epoch": 15.426517571884984, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0177, + "step": 19314 + }, + { + "epoch": 15.427316293929712, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0118, + "step": 19315 + }, + { + "epoch": 15.428115015974441, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0139, + "step": 19316 + }, + { + "epoch": 15.428913738019169, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0287, + "step": 19317 + }, + { + "epoch": 15.429712460063898, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 19318 + }, + { + "epoch": 15.430511182108626, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 19319 + }, + { + "epoch": 15.431309904153355, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 19320 + }, + { + "epoch": 15.432108626198083, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 19321 + }, + { + "epoch": 15.43290734824281, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0164, + "step": 19322 + }, + { + "epoch": 15.43370607028754, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0177, + "step": 19323 + }, + { + "epoch": 15.434504792332268, + "grad_norm": 0.062255859375, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 19324 + }, + { + "epoch": 15.435303514376997, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 19325 + }, + { + "epoch": 15.436102236421725, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0297, + "step": 19326 + }, + { + "epoch": 15.436900958466454, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 19327 + }, + { + "epoch": 15.437699680511182, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0179, + "step": 19328 + }, + { + "epoch": 15.438498402555911, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.016, + "step": 19329 + }, + { + "epoch": 15.439297124600639, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0157, + "step": 19330 + }, + { + "epoch": 15.440095846645367, + "grad_norm": 0.058837890625, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 19331 + }, + { + "epoch": 15.440894568690096, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 19332 + }, + { + "epoch": 15.441693290734824, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0235, + "step": 19333 + }, + { + "epoch": 15.442492012779553, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.018, + "step": 19334 + }, + { + "epoch": 15.44329073482428, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.014, + "step": 19335 + }, + { + "epoch": 15.44408945686901, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.028, + "step": 19336 + }, + { + "epoch": 15.444888178913738, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0237, + "step": 19337 + }, + { + "epoch": 15.445686900958467, + "grad_norm": 0.09619140625, + "learning_rate": 0.0005, + "loss": 1.0154, + "step": 19338 + }, + { + "epoch": 15.446485623003195, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0188, + "step": 19339 + }, + { + "epoch": 15.447284345047922, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.011, + "step": 19340 + }, + { + "epoch": 15.448083067092652, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.0255, + "step": 19341 + }, + { + "epoch": 15.44888178913738, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0201, + "step": 19342 + }, + { + "epoch": 15.449680511182109, + "grad_norm": 0.08544921875, + "learning_rate": 0.0005, + "loss": 1.0167, + "step": 19343 + }, + { + "epoch": 15.450479233226837, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0142, + "step": 19344 + }, + { + "epoch": 15.451277955271566, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0214, + "step": 19345 + }, + { + "epoch": 15.452076677316294, + "grad_norm": 0.0947265625, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 19346 + }, + { + "epoch": 15.452875399361023, + "grad_norm": 0.09130859375, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 19347 + }, + { + "epoch": 15.45367412140575, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 19348 + }, + { + "epoch": 15.454472843450478, + "grad_norm": 0.10546875, + "learning_rate": 0.0005, + "loss": 1.0144, + "step": 19349 + }, + { + "epoch": 15.455271565495208, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0208, + "step": 19350 + }, + { + "epoch": 15.456070287539935, + "grad_norm": 0.1298828125, + "learning_rate": 0.0005, + "loss": 1.0248, + "step": 19351 + }, + { + "epoch": 15.456869009584665, + "grad_norm": 0.11669921875, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 19352 + }, + { + "epoch": 15.457667731629392, + "grad_norm": 0.08837890625, + "learning_rate": 0.0005, + "loss": 1.0226, + "step": 19353 + }, + { + "epoch": 15.458466453674122, + "grad_norm": 0.10986328125, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 19354 + }, + { + "epoch": 15.45926517571885, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0178, + "step": 19355 + }, + { + "epoch": 15.460063897763579, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0186, + "step": 19356 + }, + { + "epoch": 15.460862619808307, + "grad_norm": 0.0830078125, + "learning_rate": 0.0005, + "loss": 1.029, + "step": 19357 + }, + { + "epoch": 15.461661341853036, + "grad_norm": 0.1064453125, + "learning_rate": 0.0005, + "loss": 1.0233, + "step": 19358 + }, + { + "epoch": 15.462460063897764, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0191, + "step": 19359 + }, + { + "epoch": 15.463258785942491, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0261, + "step": 19360 + }, + { + "epoch": 15.46405750798722, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0276, + "step": 19361 + }, + { + "epoch": 15.464856230031948, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.017, + "step": 19362 + }, + { + "epoch": 15.465654952076678, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0164, + "step": 19363 + }, + { + "epoch": 15.466453674121405, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 19364 + }, + { + "epoch": 15.467252396166135, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0176, + "step": 19365 + }, + { + "epoch": 15.468051118210862, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0188, + "step": 19366 + }, + { + "epoch": 15.468849840255592, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 19367 + }, + { + "epoch": 15.46964856230032, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0331, + "step": 19368 + }, + { + "epoch": 15.470447284345047, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0323, + "step": 19369 + }, + { + "epoch": 15.471246006389777, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 19370 + }, + { + "epoch": 15.472044728434504, + "grad_norm": 0.0869140625, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 19371 + }, + { + "epoch": 15.472843450479234, + "grad_norm": 0.1044921875, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 19372 + }, + { + "epoch": 15.473642172523961, + "grad_norm": 0.0888671875, + "learning_rate": 0.0005, + "loss": 1.0193, + "step": 19373 + }, + { + "epoch": 15.47444089456869, + "grad_norm": 0.150390625, + "learning_rate": 0.0005, + "loss": 1.0091, + "step": 19374 + }, + { + "epoch": 15.475239616613418, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0179, + "step": 19375 + }, + { + "epoch": 15.476038338658148, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0188, + "step": 19376 + }, + { + "epoch": 15.476837060702875, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 19377 + }, + { + "epoch": 15.477635782747603, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0185, + "step": 19378 + }, + { + "epoch": 15.478434504792332, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0187, + "step": 19379 + }, + { + "epoch": 15.47923322683706, + "grad_norm": 0.09228515625, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 19380 + }, + { + "epoch": 15.48003194888179, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0158, + "step": 19381 + }, + { + "epoch": 15.480830670926517, + "grad_norm": 0.10498046875, + "learning_rate": 0.0005, + "loss": 1.0162, + "step": 19382 + }, + { + "epoch": 15.481629392971247, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0231, + "step": 19383 + }, + { + "epoch": 15.482428115015974, + "grad_norm": 0.083984375, + "learning_rate": 0.0005, + "loss": 1.03, + "step": 19384 + }, + { + "epoch": 15.483226837060704, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0291, + "step": 19385 + }, + { + "epoch": 15.484025559105431, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.013, + "step": 19386 + }, + { + "epoch": 15.48482428115016, + "grad_norm": 0.1025390625, + "learning_rate": 0.0005, + "loss": 1.0183, + "step": 19387 + }, + { + "epoch": 15.485623003194888, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0298, + "step": 19388 + }, + { + "epoch": 15.486421725239616, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0173, + "step": 19389 + }, + { + "epoch": 15.487220447284345, + "grad_norm": 0.0791015625, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 19390 + }, + { + "epoch": 15.488019169329073, + "grad_norm": 0.059326171875, + "learning_rate": 0.0005, + "loss": 1.0195, + "step": 19391 + }, + { + "epoch": 15.488817891373802, + "grad_norm": 0.12890625, + "learning_rate": 0.0005, + "loss": 1.0112, + "step": 19392 + }, + { + "epoch": 15.48961661341853, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0168, + "step": 19393 + }, + { + "epoch": 15.49041533546326, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0158, + "step": 19394 + }, + { + "epoch": 15.491214057507987, + "grad_norm": 0.08984375, + "learning_rate": 0.0005, + "loss": 1.0158, + "step": 19395 + }, + { + "epoch": 15.492012779552716, + "grad_norm": 0.07275390625, + "learning_rate": 0.0005, + "loss": 1.0175, + "step": 19396 + }, + { + "epoch": 15.492811501597444, + "grad_norm": 0.0625, + "learning_rate": 0.0005, + "loss": 1.0236, + "step": 19397 + }, + { + "epoch": 15.493610223642172, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 19398 + }, + { + "epoch": 15.494408945686901, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 19399 + }, + { + "epoch": 15.495207667731629, + "grad_norm": 0.0634765625, + "learning_rate": 0.0005, + "loss": 1.0209, + "step": 19400 + }, + { + "epoch": 15.496006389776358, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 19401 + }, + { + "epoch": 15.496805111821086, + "grad_norm": 0.07666015625, + "learning_rate": 0.0005, + "loss": 1.02, + "step": 19402 + }, + { + "epoch": 15.497603833865815, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0252, + "step": 19403 + }, + { + "epoch": 15.498402555910543, + "grad_norm": 0.060546875, + "learning_rate": 0.0005, + "loss": 1.0111, + "step": 19404 + }, + { + "epoch": 15.499201277955272, + "grad_norm": 0.060302734375, + "learning_rate": 0.0005, + "loss": 1.018, + "step": 19405 + }, + { + "epoch": 15.5, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0275, + "step": 19406 + }, + { + "epoch": 15.500798722044728, + "grad_norm": 0.06884765625, + "learning_rate": 0.0005, + "loss": 1.0178, + "step": 19407 + }, + { + "epoch": 15.501597444089457, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.017, + "step": 19408 + }, + { + "epoch": 15.502396166134185, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0256, + "step": 19409 + }, + { + "epoch": 15.503194888178914, + "grad_norm": 0.06298828125, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 19410 + }, + { + "epoch": 15.503993610223642, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0169, + "step": 19411 + }, + { + "epoch": 15.504792332268371, + "grad_norm": 0.10595703125, + "learning_rate": 0.0005, + "loss": 1.0115, + "step": 19412 + }, + { + "epoch": 15.505591054313099, + "grad_norm": 0.054443359375, + "learning_rate": 0.0005, + "loss": 1.0143, + "step": 19413 + }, + { + "epoch": 15.506389776357828, + "grad_norm": 0.0927734375, + "learning_rate": 0.0005, + "loss": 1.0162, + "step": 19414 + }, + { + "epoch": 15.507188498402556, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0311, + "step": 19415 + }, + { + "epoch": 15.507987220447284, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.026, + "step": 19416 + }, + { + "epoch": 15.508785942492013, + "grad_norm": 0.10400390625, + "learning_rate": 0.0005, + "loss": 1.0288, + "step": 19417 + }, + { + "epoch": 15.50958466453674, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0132, + "step": 19418 + }, + { + "epoch": 15.51038338658147, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0254, + "step": 19419 + }, + { + "epoch": 15.511182108626198, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0312, + "step": 19420 + }, + { + "epoch": 15.511980830670927, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0205, + "step": 19421 + }, + { + "epoch": 15.512779552715655, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 19422 + }, + { + "epoch": 15.513578274760384, + "grad_norm": 0.1376953125, + "learning_rate": 0.0005, + "loss": 1.0281, + "step": 19423 + }, + { + "epoch": 15.514376996805112, + "grad_norm": 0.07421875, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 19424 + }, + { + "epoch": 15.51517571884984, + "grad_norm": 0.09375, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 19425 + }, + { + "epoch": 15.515974440894569, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 19426 + }, + { + "epoch": 15.516773162939296, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0152, + "step": 19427 + }, + { + "epoch": 15.517571884984026, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0258, + "step": 19428 + }, + { + "epoch": 15.518370607028753, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 19429 + }, + { + "epoch": 15.519169329073483, + "grad_norm": 0.09033203125, + "learning_rate": 0.0005, + "loss": 1.0185, + "step": 19430 + }, + { + "epoch": 15.51996805111821, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 19431 + }, + { + "epoch": 15.52076677316294, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0181, + "step": 19432 + }, + { + "epoch": 15.521565495207668, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0234, + "step": 19433 + }, + { + "epoch": 15.522364217252395, + "grad_norm": 0.08447265625, + "learning_rate": 0.0005, + "loss": 1.0292, + "step": 19434 + }, + { + "epoch": 15.523162939297125, + "grad_norm": 0.06494140625, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 19435 + }, + { + "epoch": 15.523961661341852, + "grad_norm": 0.08251953125, + "learning_rate": 0.0005, + "loss": 1.0228, + "step": 19436 + }, + { + "epoch": 15.524760383386582, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.004, + "step": 19437 + }, + { + "epoch": 15.52555910543131, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0199, + "step": 19438 + }, + { + "epoch": 15.526357827476039, + "grad_norm": 0.080078125, + "learning_rate": 0.0005, + "loss": 1.0213, + "step": 19439 + }, + { + "epoch": 15.527156549520766, + "grad_norm": 0.0576171875, + "learning_rate": 0.0005, + "loss": 1.0189, + "step": 19440 + }, + { + "epoch": 15.527955271565496, + "grad_norm": 0.09326171875, + "learning_rate": 0.0005, + "loss": 1.0211, + "step": 19441 + }, + { + "epoch": 15.528753993610223, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 19442 + }, + { + "epoch": 15.529552715654953, + "grad_norm": 0.06396484375, + "learning_rate": 0.0005, + "loss": 1.0268, + "step": 19443 + }, + { + "epoch": 15.53035143769968, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0176, + "step": 19444 + }, + { + "epoch": 15.531150159744408, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.0156, + "step": 19445 + }, + { + "epoch": 15.531948881789138, + "grad_norm": 0.0615234375, + "learning_rate": 0.0005, + "loss": 1.0172, + "step": 19446 + }, + { + "epoch": 15.532747603833865, + "grad_norm": 0.08740234375, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 19447 + }, + { + "epoch": 15.533546325878595, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.025, + "step": 19448 + }, + { + "epoch": 15.534345047923322, + "grad_norm": 0.068359375, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 19449 + }, + { + "epoch": 15.535143769968052, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0174, + "step": 19450 + }, + { + "epoch": 15.53594249201278, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0147, + "step": 19451 + }, + { + "epoch": 15.536741214057509, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 19452 + }, + { + "epoch": 15.537539936102236, + "grad_norm": 0.07177734375, + "learning_rate": 0.0005, + "loss": 1.0152, + "step": 19453 + }, + { + "epoch": 15.538338658146966, + "grad_norm": 0.06103515625, + "learning_rate": 0.0005, + "loss": 1.0164, + "step": 19454 + }, + { + "epoch": 15.539137380191693, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0215, + "step": 19455 + }, + { + "epoch": 15.539936102236421, + "grad_norm": 0.06787109375, + "learning_rate": 0.0005, + "loss": 1.0251, + "step": 19456 + }, + { + "epoch": 15.54073482428115, + "grad_norm": 0.07763671875, + "learning_rate": 0.0005, + "loss": 1.0196, + "step": 19457 + }, + { + "epoch": 15.541533546325878, + "grad_norm": 0.07373046875, + "learning_rate": 0.0005, + "loss": 1.0194, + "step": 19458 + }, + { + "epoch": 15.542332268370608, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0172, + "step": 19459 + }, + { + "epoch": 15.543130990415335, + "grad_norm": 0.076171875, + "learning_rate": 0.0005, + "loss": 1.0184, + "step": 19460 + }, + { + "epoch": 15.543929712460065, + "grad_norm": 0.08056640625, + "learning_rate": 0.0005, + "loss": 1.0157, + "step": 19461 + }, + { + "epoch": 15.544728434504792, + "grad_norm": 0.0703125, + "learning_rate": 0.0005, + "loss": 1.0249, + "step": 19462 + }, + { + "epoch": 15.545527156549522, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0161, + "step": 19463 + }, + { + "epoch": 15.54632587859425, + "grad_norm": 0.0732421875, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 19464 + }, + { + "epoch": 15.547124600638977, + "grad_norm": 0.06591796875, + "learning_rate": 0.0005, + "loss": 1.0176, + "step": 19465 + }, + { + "epoch": 15.547923322683706, + "grad_norm": 0.08203125, + "learning_rate": 0.0005, + "loss": 1.0091, + "step": 19466 + }, + { + "epoch": 15.548722044728434, + "grad_norm": 0.06640625, + "learning_rate": 0.0005, + "loss": 1.0229, + "step": 19467 + }, + { + "epoch": 15.549520766773163, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0266, + "step": 19468 + }, + { + "epoch": 15.550319488817891, + "grad_norm": 0.06689453125, + "learning_rate": 0.0005, + "loss": 1.0182, + "step": 19469 + }, + { + "epoch": 15.55111821086262, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.017, + "step": 19470 + }, + { + "epoch": 15.551916932907348, + "grad_norm": 0.059814453125, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 19471 + }, + { + "epoch": 15.552715654952078, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0232, + "step": 19472 + }, + { + "epoch": 15.553514376996805, + "grad_norm": 0.07470703125, + "learning_rate": 0.0005, + "loss": 1.0216, + "step": 19473 + }, + { + "epoch": 15.554313099041533, + "grad_norm": 0.0771484375, + "learning_rate": 0.0005, + "loss": 1.0272, + "step": 19474 + }, + { + "epoch": 15.555111821086262, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 19475 + }, + { + "epoch": 15.55591054313099, + "grad_norm": 0.10791015625, + "learning_rate": 0.0005, + "loss": 1.0178, + "step": 19476 + }, + { + "epoch": 15.55670926517572, + "grad_norm": 0.06005859375, + "learning_rate": 0.0005, + "loss": 1.0225, + "step": 19477 + }, + { + "epoch": 15.557507987220447, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 19478 + }, + { + "epoch": 15.558306709265176, + "grad_norm": 0.0712890625, + "learning_rate": 0.0005, + "loss": 1.0107, + "step": 19479 + }, + { + "epoch": 15.559105431309904, + "grad_norm": 0.05517578125, + "learning_rate": 0.0005, + "loss": 1.0279, + "step": 19480 + }, + { + "epoch": 15.559904153354633, + "grad_norm": 0.0673828125, + "learning_rate": 0.0005, + "loss": 1.0212, + "step": 19481 + }, + { + "epoch": 15.560702875399361, + "grad_norm": 0.060791015625, + "learning_rate": 0.0005, + "loss": 1.0152, + "step": 19482 + }, + { + "epoch": 15.561501597444089, + "grad_norm": 0.0751953125, + "learning_rate": 0.0005, + "loss": 1.021, + "step": 19483 + }, + { + "epoch": 15.562300319488818, + "grad_norm": 0.0810546875, + "learning_rate": 0.0005, + "loss": 1.0265, + "step": 19484 + }, + { + "epoch": 15.563099041533546, + "grad_norm": 0.064453125, + "learning_rate": 0.0005, + "loss": 1.0206, + "step": 19485 + }, + { + "epoch": 15.563897763578275, + "grad_norm": 0.07861328125, + "learning_rate": 0.0005, + "loss": 1.0262, + "step": 19486 + }, + { + "epoch": 15.564696485623003, + "grad_norm": 0.07080078125, + "learning_rate": 0.0005, + "loss": 1.0198, + "step": 19487 + }, + { + "epoch": 15.565495207667732, + "grad_norm": 0.0654296875, + "learning_rate": 0.0005, + "loss": 1.0257, + "step": 19488 + }, + { + "epoch": 15.56629392971246, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0182, + "step": 19489 + }, + { + "epoch": 15.56709265175719, + "grad_norm": 0.08349609375, + "learning_rate": 0.0005, + "loss": 1.0126, + "step": 19490 + }, + { + "epoch": 15.567891373801917, + "grad_norm": 0.08642578125, + "learning_rate": 0.0005, + "loss": 1.0243, + "step": 19491 + }, + { + "epoch": 15.568690095846645, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0177, + "step": 19492 + }, + { + "epoch": 15.569488817891374, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0181, + "step": 19493 + }, + { + "epoch": 15.570287539936102, + "grad_norm": 0.061767578125, + "learning_rate": 0.0005, + "loss": 1.0161, + "step": 19494 + }, + { + "epoch": 15.571086261980831, + "grad_norm": 0.07568359375, + "learning_rate": 0.0005, + "loss": 1.0264, + "step": 19495 + }, + { + "epoch": 15.571884984025559, + "grad_norm": 0.061279296875, + "learning_rate": 0.0005, + "loss": 1.0131, + "step": 19496 + }, + { + "epoch": 15.572683706070288, + "grad_norm": 0.078125, + "learning_rate": 0.0005, + "loss": 1.0247, + "step": 19497 + }, + { + "epoch": 15.573482428115016, + "grad_norm": 0.0693359375, + "learning_rate": 0.0005, + "loss": 1.0217, + "step": 19498 + }, + { + "epoch": 15.574281150159745, + "grad_norm": 0.072265625, + "learning_rate": 0.0005, + "loss": 1.0176, + "step": 19499 + }, + { + "epoch": 15.575079872204473, + "grad_norm": 0.08154296875, + "learning_rate": 0.0005, + "loss": 1.0183, + "step": 19500 + } + ], + "logging_steps": 1.0, + "max_steps": 80128, + "num_input_tokens_seen": 0, + "num_train_epochs": 64, + "save_steps": 500, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 9.802032495179334e+18, + "train_batch_size": 128, + "trial_name": null, + "trial_params": null +}