{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 20.0, "eval_steps": 0, "global_step": 340, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.058823529411764705, "grad_norm": 1.234375, "learning_rate": 1e-05, "loss": 2.2766, "step": 1 }, { "epoch": 0.11764705882352941, "grad_norm": 1.125, "learning_rate": 9.99978655851684e-06, "loss": 2.2043, "step": 2 }, { "epoch": 0.17647058823529413, "grad_norm": 1.125, "learning_rate": 9.999146252290264e-06, "loss": 2.239, "step": 3 }, { "epoch": 0.23529411764705882, "grad_norm": 1.046875, "learning_rate": 9.998079135987437e-06, "loss": 2.2274, "step": 4 }, { "epoch": 0.29411764705882354, "grad_norm": 0.8984375, "learning_rate": 9.996585300715117e-06, "loss": 2.0609, "step": 5 }, { "epoch": 0.35294117647058826, "grad_norm": 0.890625, "learning_rate": 9.994664874011864e-06, "loss": 2.1648, "step": 6 }, { "epoch": 0.4117647058823529, "grad_norm": 1.0625, "learning_rate": 9.992318019837171e-06, "loss": 2.1999, "step": 7 }, { "epoch": 0.47058823529411764, "grad_norm": 0.9140625, "learning_rate": 9.989544938557453e-06, "loss": 2.2122, "step": 8 }, { "epoch": 0.5294117647058824, "grad_norm": 0.77734375, "learning_rate": 9.98634586692894e-06, "loss": 1.9256, "step": 9 }, { "epoch": 0.5882352941176471, "grad_norm": 0.85546875, "learning_rate": 9.982721078077474e-06, "loss": 2.0316, "step": 10 }, { "epoch": 0.6470588235294118, "grad_norm": 0.7734375, "learning_rate": 9.978670881475173e-06, "loss": 1.899, "step": 11 }, { "epoch": 0.7058823529411765, "grad_norm": 0.8046875, "learning_rate": 9.97419562291403e-06, "loss": 1.8132, "step": 12 }, { "epoch": 0.7647058823529411, "grad_norm": 0.8671875, "learning_rate": 9.96929568447637e-06, "loss": 1.8659, "step": 13 }, { "epoch": 0.8235294117647058, "grad_norm": 0.83203125, "learning_rate": 9.963971484502247e-06, "loss": 1.9389, "step": 14 }, { "epoch": 0.8823529411764706, "grad_norm": 0.75, "learning_rate": 9.958223477553715e-06, "loss": 1.8686, "step": 15 }, { "epoch": 0.9411764705882353, "grad_norm": 0.6796875, "learning_rate": 9.952052154376027e-06, "loss": 1.7301, "step": 16 }, { "epoch": 1.0, "grad_norm": 0.671875, "learning_rate": 9.945458041855732e-06, "loss": 1.7067, "step": 17 }, { "epoch": 1.0588235294117647, "grad_norm": 0.71484375, "learning_rate": 9.938441702975689e-06, "loss": 1.7523, "step": 18 }, { "epoch": 1.1176470588235294, "grad_norm": 0.7265625, "learning_rate": 9.931003736767013e-06, "loss": 1.615, "step": 19 }, { "epoch": 1.1764705882352942, "grad_norm": 0.77734375, "learning_rate": 9.923144778257918e-06, "loss": 1.7464, "step": 20 }, { "epoch": 1.2352941176470589, "grad_norm": 0.75390625, "learning_rate": 9.91486549841951e-06, "loss": 1.7057, "step": 21 }, { "epoch": 1.2941176470588236, "grad_norm": 0.69921875, "learning_rate": 9.906166604108494e-06, "loss": 1.4832, "step": 22 }, { "epoch": 1.3529411764705883, "grad_norm": 0.73828125, "learning_rate": 9.89704883800683e-06, "loss": 1.5449, "step": 23 }, { "epoch": 1.4117647058823528, "grad_norm": 0.8125, "learning_rate": 9.887512978558329e-06, "loss": 1.5252, "step": 24 }, { "epoch": 1.4705882352941178, "grad_norm": 0.87890625, "learning_rate": 9.877559839902185e-06, "loss": 1.4343, "step": 25 }, { "epoch": 1.5294117647058822, "grad_norm": 0.7734375, "learning_rate": 9.867190271803466e-06, "loss": 1.4411, "step": 26 }, { "epoch": 1.5882352941176472, "grad_norm": 0.7890625, "learning_rate": 9.85640515958057e-06, "loss": 1.4562, "step": 27 }, { "epoch": 1.6470588235294117, "grad_norm": 0.828125, "learning_rate": 9.845205424029639e-06, "loss": 1.3878, "step": 28 }, { "epoch": 1.7058823529411766, "grad_norm": 0.7109375, "learning_rate": 9.833592021345938e-06, "loss": 1.3791, "step": 29 }, { "epoch": 1.7647058823529411, "grad_norm": 0.8359375, "learning_rate": 9.821565943042225e-06, "loss": 1.328, "step": 30 }, { "epoch": 1.8235294117647058, "grad_norm": 0.76953125, "learning_rate": 9.809128215864096e-06, "loss": 1.3734, "step": 31 }, { "epoch": 1.8823529411764706, "grad_norm": 0.63671875, "learning_rate": 9.796279901702326e-06, "loss": 1.2231, "step": 32 }, { "epoch": 1.9411764705882353, "grad_norm": 0.6640625, "learning_rate": 9.783022097502204e-06, "loss": 1.2504, "step": 33 }, { "epoch": 2.0, "grad_norm": 0.86328125, "learning_rate": 9.76935593516989e-06, "loss": 1.2874, "step": 34 }, { "epoch": 2.0588235294117645, "grad_norm": 0.65625, "learning_rate": 9.755282581475769e-06, "loss": 1.2118, "step": 35 }, { "epoch": 2.1176470588235294, "grad_norm": 0.73046875, "learning_rate": 9.74080323795483e-06, "loss": 1.1986, "step": 36 }, { "epoch": 2.176470588235294, "grad_norm": 0.69921875, "learning_rate": 9.7259191408041e-06, "loss": 1.2208, "step": 37 }, { "epoch": 2.235294117647059, "grad_norm": 1.3671875, "learning_rate": 9.710631560777082e-06, "loss": 1.1827, "step": 38 }, { "epoch": 2.2941176470588234, "grad_norm": 0.66015625, "learning_rate": 9.694941803075285e-06, "loss": 1.1847, "step": 39 }, { "epoch": 2.3529411764705883, "grad_norm": 0.69921875, "learning_rate": 9.678851207236764e-06, "loss": 1.1876, "step": 40 }, { "epoch": 2.411764705882353, "grad_norm": 0.76953125, "learning_rate": 9.66236114702178e-06, "loss": 1.1567, "step": 41 }, { "epoch": 2.4705882352941178, "grad_norm": 0.625, "learning_rate": 9.645473030295496e-06, "loss": 1.0874, "step": 42 }, { "epoch": 2.5294117647058822, "grad_norm": 0.60546875, "learning_rate": 9.628188298907782e-06, "loss": 1.0997, "step": 43 }, { "epoch": 2.588235294117647, "grad_norm": 0.6171875, "learning_rate": 9.610508428570122e-06, "loss": 1.0843, "step": 44 }, { "epoch": 2.6470588235294117, "grad_norm": 0.64453125, "learning_rate": 9.592434928729617e-06, "loss": 1.065, "step": 45 }, { "epoch": 2.7058823529411766, "grad_norm": 0.66796875, "learning_rate": 9.573969342440107e-06, "loss": 1.0507, "step": 46 }, { "epoch": 2.764705882352941, "grad_norm": 0.6640625, "learning_rate": 9.555113246230443e-06, "loss": 1.029, "step": 47 }, { "epoch": 2.8235294117647056, "grad_norm": 0.65234375, "learning_rate": 9.535868249969882e-06, "loss": 1.0386, "step": 48 }, { "epoch": 2.8823529411764706, "grad_norm": 0.64453125, "learning_rate": 9.516235996730645e-06, "loss": 1.0454, "step": 49 }, { "epoch": 2.9411764705882355, "grad_norm": 0.578125, "learning_rate": 9.496218162647629e-06, "loss": 1.0032, "step": 50 }, { "epoch": 3.0, "grad_norm": 0.65625, "learning_rate": 9.475816456775313e-06, "loss": 1.0014, "step": 51 }, { "epoch": 3.0588235294117645, "grad_norm": 0.6953125, "learning_rate": 9.45503262094184e-06, "loss": 0.9925, "step": 52 }, { "epoch": 3.1176470588235294, "grad_norm": 0.7578125, "learning_rate": 9.43386842960031e-06, "loss": 0.9368, "step": 53 }, { "epoch": 3.176470588235294, "grad_norm": 0.58984375, "learning_rate": 9.41232568967728e-06, "loss": 0.9508, "step": 54 }, { "epoch": 3.235294117647059, "grad_norm": 0.59765625, "learning_rate": 9.39040624041849e-06, "loss": 0.9048, "step": 55 }, { "epoch": 3.2941176470588234, "grad_norm": 0.6484375, "learning_rate": 9.368111953231849e-06, "loss": 0.9199, "step": 56 }, { "epoch": 3.3529411764705883, "grad_norm": 0.77734375, "learning_rate": 9.345444731527642e-06, "loss": 0.9188, "step": 57 }, { "epoch": 3.411764705882353, "grad_norm": 0.80859375, "learning_rate": 9.32240651055604e-06, "loss": 0.9264, "step": 58 }, { "epoch": 3.4705882352941178, "grad_norm": 0.78515625, "learning_rate": 9.298999257241862e-06, "loss": 0.9272, "step": 59 }, { "epoch": 3.5294117647058822, "grad_norm": 0.56640625, "learning_rate": 9.275224970016656e-06, "loss": 0.8726, "step": 60 }, { "epoch": 3.588235294117647, "grad_norm": 0.73828125, "learning_rate": 9.251085678648072e-06, "loss": 0.8269, "step": 61 }, { "epoch": 3.6470588235294117, "grad_norm": 0.7109375, "learning_rate": 9.22658344406657e-06, "loss": 0.8624, "step": 62 }, { "epoch": 3.7058823529411766, "grad_norm": 0.66015625, "learning_rate": 9.201720358189464e-06, "loss": 0.824, "step": 63 }, { "epoch": 3.764705882352941, "grad_norm": 0.71484375, "learning_rate": 9.176498543742328e-06, "loss": 0.8, "step": 64 }, { "epoch": 3.8235294117647056, "grad_norm": 0.75390625, "learning_rate": 9.150920154077753e-06, "loss": 0.8687, "step": 65 }, { "epoch": 3.8823529411764706, "grad_norm": 0.73046875, "learning_rate": 9.124987372991512e-06, "loss": 0.8149, "step": 66 }, { "epoch": 3.9411764705882355, "grad_norm": 0.99609375, "learning_rate": 9.098702414536107e-06, "loss": 0.8413, "step": 67 }, { "epoch": 4.0, "grad_norm": 0.7265625, "learning_rate": 9.072067522831743e-06, "loss": 0.7488, "step": 68 }, { "epoch": 4.0588235294117645, "grad_norm": 0.71484375, "learning_rate": 9.045084971874738e-06, "loss": 0.7726, "step": 69 }, { "epoch": 4.117647058823529, "grad_norm": 0.7578125, "learning_rate": 9.017757065343368e-06, "loss": 0.7459, "step": 70 }, { "epoch": 4.176470588235294, "grad_norm": 0.69921875, "learning_rate": 8.990086136401199e-06, "loss": 0.7159, "step": 71 }, { "epoch": 4.235294117647059, "grad_norm": 0.765625, "learning_rate": 8.96207454749787e-06, "loss": 0.7492, "step": 72 }, { "epoch": 4.294117647058823, "grad_norm": 0.828125, "learning_rate": 8.933724690167417e-06, "loss": 0.7163, "step": 73 }, { "epoch": 4.352941176470588, "grad_norm": 0.8828125, "learning_rate": 8.905038984824079e-06, "loss": 0.7241, "step": 74 }, { "epoch": 4.411764705882353, "grad_norm": 0.69140625, "learning_rate": 8.87601988055565e-06, "loss": 0.6674, "step": 75 }, { "epoch": 4.470588235294118, "grad_norm": 0.69921875, "learning_rate": 8.846669854914395e-06, "loss": 0.644, "step": 76 }, { "epoch": 4.529411764705882, "grad_norm": 0.93359375, "learning_rate": 8.816991413705515e-06, "loss": 0.6329, "step": 77 }, { "epoch": 4.588235294117647, "grad_norm": 0.8203125, "learning_rate": 8.786987090773214e-06, "loss": 0.6597, "step": 78 }, { "epoch": 4.647058823529412, "grad_norm": 0.82421875, "learning_rate": 8.756659447784367e-06, "loss": 0.6312, "step": 79 }, { "epoch": 4.705882352941177, "grad_norm": 0.83984375, "learning_rate": 8.726011074009813e-06, "loss": 0.6067, "step": 80 }, { "epoch": 4.764705882352941, "grad_norm": 0.95703125, "learning_rate": 8.695044586103297e-06, "loss": 0.5903, "step": 81 }, { "epoch": 4.823529411764706, "grad_norm": 0.8125, "learning_rate": 8.663762627878059e-06, "loss": 0.6044, "step": 82 }, { "epoch": 4.882352941176471, "grad_norm": 0.984375, "learning_rate": 8.632167870081122e-06, "loss": 0.561, "step": 83 }, { "epoch": 4.9411764705882355, "grad_norm": 0.96484375, "learning_rate": 8.600263010165275e-06, "loss": 0.5277, "step": 84 }, { "epoch": 5.0, "grad_norm": 0.9453125, "learning_rate": 8.568050772058763e-06, "loss": 0.525, "step": 85 }, { "epoch": 5.0588235294117645, "grad_norm": 0.84375, "learning_rate": 8.535533905932739e-06, "loss": 0.5609, "step": 86 }, { "epoch": 5.117647058823529, "grad_norm": 0.8828125, "learning_rate": 8.502715187966455e-06, "loss": 0.462, "step": 87 }, { "epoch": 5.176470588235294, "grad_norm": 0.91015625, "learning_rate": 8.469597420110249e-06, "loss": 0.4672, "step": 88 }, { "epoch": 5.235294117647059, "grad_norm": 1.0390625, "learning_rate": 8.436183429846314e-06, "loss": 0.4455, "step": 89 }, { "epoch": 5.294117647058823, "grad_norm": 0.890625, "learning_rate": 8.402476069947309e-06, "loss": 0.475, "step": 90 }, { "epoch": 5.352941176470588, "grad_norm": 0.8828125, "learning_rate": 8.368478218232787e-06, "loss": 0.4027, "step": 91 }, { "epoch": 5.411764705882353, "grad_norm": 0.9609375, "learning_rate": 8.334192777323508e-06, "loss": 0.4635, "step": 92 }, { "epoch": 5.470588235294118, "grad_norm": 1.125, "learning_rate": 8.299622674393615e-06, "loss": 0.4223, "step": 93 }, { "epoch": 5.529411764705882, "grad_norm": 1.125, "learning_rate": 8.264770860920722e-06, "loss": 0.4127, "step": 94 }, { "epoch": 5.588235294117647, "grad_norm": 1.015625, "learning_rate": 8.229640312433938e-06, "loss": 0.3538, "step": 95 }, { "epoch": 5.647058823529412, "grad_norm": 1.265625, "learning_rate": 8.194234028259806e-06, "loss": 0.3591, "step": 96 }, { "epoch": 5.705882352941177, "grad_norm": 1.125, "learning_rate": 8.158555031266255e-06, "loss": 0.3207, "step": 97 }, { "epoch": 5.764705882352941, "grad_norm": 1.140625, "learning_rate": 8.122606367604497e-06, "loss": 0.3418, "step": 98 }, { "epoch": 5.823529411764706, "grad_norm": 1.21875, "learning_rate": 8.086391106448965e-06, "loss": 0.376, "step": 99 }, { "epoch": 5.882352941176471, "grad_norm": 1.15625, "learning_rate": 8.049912339735284e-06, "loss": 0.2949, "step": 100 }, { "epoch": 5.9411764705882355, "grad_norm": 0.9609375, "learning_rate": 8.013173181896283e-06, "loss": 0.2667, "step": 101 }, { "epoch": 6.0, "grad_norm": 1.03125, "learning_rate": 7.976176769596095e-06, "loss": 0.3004, "step": 102 }, { "epoch": 6.0588235294117645, "grad_norm": 1.0, "learning_rate": 7.938926261462366e-06, "loss": 0.2446, "step": 103 }, { "epoch": 6.117647058823529, "grad_norm": 1.03125, "learning_rate": 7.90142483781658e-06, "loss": 0.2519, "step": 104 }, { "epoch": 6.176470588235294, "grad_norm": 1.0703125, "learning_rate": 7.863675700402527e-06, "loss": 0.1904, "step": 105 }, { "epoch": 6.235294117647059, "grad_norm": 1.0546875, "learning_rate": 7.82568207211296e-06, "loss": 0.2112, "step": 106 }, { "epoch": 6.294117647058823, "grad_norm": 1.328125, "learning_rate": 7.787447196714428e-06, "loss": 0.2095, "step": 107 }, { "epoch": 6.352941176470588, "grad_norm": 1.203125, "learning_rate": 7.748974338570337e-06, "loss": 0.195, "step": 108 }, { "epoch": 6.411764705882353, "grad_norm": 0.9609375, "learning_rate": 7.710266782362248e-06, "loss": 0.2051, "step": 109 }, { "epoch": 6.470588235294118, "grad_norm": 1.078125, "learning_rate": 7.671327832809442e-06, "loss": 0.1555, "step": 110 }, { "epoch": 6.529411764705882, "grad_norm": 1.1171875, "learning_rate": 7.63216081438678e-06, "loss": 0.1427, "step": 111 }, { "epoch": 6.588235294117647, "grad_norm": 0.87109375, "learning_rate": 7.5927690710408606e-06, "loss": 0.1071, "step": 112 }, { "epoch": 6.647058823529412, "grad_norm": 1.0078125, "learning_rate": 7.553155965904535e-06, "loss": 0.1427, "step": 113 }, { "epoch": 6.705882352941177, "grad_norm": 0.90625, "learning_rate": 7.513324881009769e-06, "loss": 0.1262, "step": 114 }, { "epoch": 6.764705882352941, "grad_norm": 1.0, "learning_rate": 7.473279216998896e-06, "loss": 0.1334, "step": 115 }, { "epoch": 6.823529411764706, "grad_norm": 0.7265625, "learning_rate": 7.4330223928342814e-06, "loss": 0.1114, "step": 116 }, { "epoch": 6.882352941176471, "grad_norm": 0.75, "learning_rate": 7.392557845506433e-06, "loss": 0.1184, "step": 117 }, { "epoch": 6.9411764705882355, "grad_norm": 0.921875, "learning_rate": 7.351889029740548e-06, "loss": 0.0926, "step": 118 }, { "epoch": 7.0, "grad_norm": 0.76171875, "learning_rate": 7.311019417701567e-06, "loss": 0.074, "step": 119 }, { "epoch": 7.0588235294117645, "grad_norm": 0.5390625, "learning_rate": 7.269952498697734e-06, "loss": 0.051, "step": 120 }, { "epoch": 7.117647058823529, "grad_norm": 0.72265625, "learning_rate": 7.2286917788826926e-06, "loss": 0.0681, "step": 121 }, { "epoch": 7.176470588235294, "grad_norm": 0.58984375, "learning_rate": 7.187240780956133e-06, "loss": 0.0585, "step": 122 }, { "epoch": 7.235294117647059, "grad_norm": 1.09375, "learning_rate": 7.145603043863045e-06, "loss": 0.0612, "step": 123 }, { "epoch": 7.294117647058823, "grad_norm": 0.62109375, "learning_rate": 7.103782122491577e-06, "loss": 0.0371, "step": 124 }, { "epoch": 7.352941176470588, "grad_norm": 0.70703125, "learning_rate": 7.061781587369518e-06, "loss": 0.0275, "step": 125 }, { "epoch": 7.411764705882353, "grad_norm": 0.5859375, "learning_rate": 7.019605024359475e-06, "loss": 0.0351, "step": 126 }, { "epoch": 7.470588235294118, "grad_norm": 0.63671875, "learning_rate": 6.977256034352713e-06, "loss": 0.0411, "step": 127 }, { "epoch": 7.529411764705882, "grad_norm": 0.482421875, "learning_rate": 6.934738232961728e-06, "loss": 0.0326, "step": 128 }, { "epoch": 7.588235294117647, "grad_norm": 0.55078125, "learning_rate": 6.892055250211552e-06, "loss": 0.0318, "step": 129 }, { "epoch": 7.647058823529412, "grad_norm": 0.51171875, "learning_rate": 6.849210730229846e-06, "loss": 0.0302, "step": 130 }, { "epoch": 7.705882352941177, "grad_norm": 0.52734375, "learning_rate": 6.806208330935766e-06, "loss": 0.0279, "step": 131 }, { "epoch": 7.764705882352941, "grad_norm": 0.50390625, "learning_rate": 6.763051723727663e-06, "loss": 0.0227, "step": 132 }, { "epoch": 7.823529411764706, "grad_norm": 0.44921875, "learning_rate": 6.719744593169642e-06, "loss": 0.0146, "step": 133 }, { "epoch": 7.882352941176471, "grad_norm": 0.52734375, "learning_rate": 6.67629063667697e-06, "loss": 0.0204, "step": 134 }, { "epoch": 7.9411764705882355, "grad_norm": 0.482421875, "learning_rate": 6.6326935642004165e-06, "loss": 0.016, "step": 135 }, { "epoch": 8.0, "grad_norm": 0.2490234375, "learning_rate": 6.588957097909509e-06, "loss": 0.0104, "step": 136 }, { "epoch": 8.058823529411764, "grad_norm": 0.189453125, "learning_rate": 6.545084971874738e-06, "loss": 0.0086, "step": 137 }, { "epoch": 8.117647058823529, "grad_norm": 0.185546875, "learning_rate": 6.501080931748764e-06, "loss": 0.0074, "step": 138 }, { "epoch": 8.176470588235293, "grad_norm": 0.322265625, "learning_rate": 6.456948734446624e-06, "loss": 0.0112, "step": 139 }, { "epoch": 8.235294117647058, "grad_norm": 0.1982421875, "learning_rate": 6.412692147824976e-06, "loss": 0.0071, "step": 140 }, { "epoch": 8.294117647058824, "grad_norm": 0.275390625, "learning_rate": 6.368314950360416e-06, "loss": 0.0085, "step": 141 }, { "epoch": 8.352941176470589, "grad_norm": 0.6796875, "learning_rate": 6.323820930826879e-06, "loss": 0.0086, "step": 142 }, { "epoch": 8.411764705882353, "grad_norm": 0.1220703125, "learning_rate": 6.279213887972179e-06, "loss": 0.0055, "step": 143 }, { "epoch": 8.470588235294118, "grad_norm": 0.130859375, "learning_rate": 6.234497630193666e-06, "loss": 0.0055, "step": 144 }, { "epoch": 8.529411764705882, "grad_norm": 0.51171875, "learning_rate": 6.189675975213094e-06, "loss": 0.0068, "step": 145 }, { "epoch": 8.588235294117647, "grad_norm": 0.365234375, "learning_rate": 6.144752749750671e-06, "loss": 0.0053, "step": 146 }, { "epoch": 8.647058823529411, "grad_norm": 0.5390625, "learning_rate": 6.099731789198344e-06, "loss": 0.0081, "step": 147 }, { "epoch": 8.705882352941176, "grad_norm": 0.1640625, "learning_rate": 6.05461693729235e-06, "loss": 0.0057, "step": 148 }, { "epoch": 8.764705882352942, "grad_norm": 0.1376953125, "learning_rate": 6.009412045785051e-06, "loss": 0.005, "step": 149 }, { "epoch": 8.823529411764707, "grad_norm": 0.1328125, "learning_rate": 5.964120974116085e-06, "loss": 0.0043, "step": 150 }, { "epoch": 8.882352941176471, "grad_norm": 0.47265625, "learning_rate": 5.918747589082853e-06, "loss": 0.0052, "step": 151 }, { "epoch": 8.941176470588236, "grad_norm": 0.1328125, "learning_rate": 5.8732957645103946e-06, "loss": 0.0051, "step": 152 }, { "epoch": 9.0, "grad_norm": 0.07177734375, "learning_rate": 5.82776938092065e-06, "loss": 0.0032, "step": 153 }, { "epoch": 9.058823529411764, "grad_norm": 0.154296875, "learning_rate": 5.782172325201155e-06, "loss": 0.004, "step": 154 }, { "epoch": 9.117647058823529, "grad_norm": 0.248046875, "learning_rate": 5.736508490273189e-06, "loss": 0.0033, "step": 155 }, { "epoch": 9.176470588235293, "grad_norm": 0.267578125, "learning_rate": 5.690781774759412e-06, "loss": 0.0029, "step": 156 }, { "epoch": 9.235294117647058, "grad_norm": 0.059326171875, "learning_rate": 5.644996082651018e-06, "loss": 0.0029, "step": 157 }, { "epoch": 9.294117647058824, "grad_norm": 0.08154296875, "learning_rate": 5.5991553229744166e-06, "loss": 0.0031, "step": 158 }, { "epoch": 9.352941176470589, "grad_norm": 0.08447265625, "learning_rate": 5.553263409457504e-06, "loss": 0.0035, "step": 159 }, { "epoch": 9.411764705882353, "grad_norm": 0.05712890625, "learning_rate": 5.507324260195516e-06, "loss": 0.0028, "step": 160 }, { "epoch": 9.470588235294118, "grad_norm": 0.0703125, "learning_rate": 5.46134179731651e-06, "loss": 0.0025, "step": 161 }, { "epoch": 9.529411764705882, "grad_norm": 0.053955078125, "learning_rate": 5.41531994664652e-06, "loss": 0.0027, "step": 162 }, { "epoch": 9.588235294117647, "grad_norm": 0.06787109375, "learning_rate": 5.36926263737437e-06, "loss": 0.003, "step": 163 }, { "epoch": 9.647058823529411, "grad_norm": 0.0693359375, "learning_rate": 5.323173801716222e-06, "loss": 0.0029, "step": 164 }, { "epoch": 9.705882352941176, "grad_norm": 0.053466796875, "learning_rate": 5.27705737457985e-06, "loss": 0.003, "step": 165 }, { "epoch": 9.764705882352942, "grad_norm": 0.07373046875, "learning_rate": 5.230917293228699e-06, "loss": 0.0029, "step": 166 }, { "epoch": 9.823529411764707, "grad_norm": 0.06396484375, "learning_rate": 5.184757496945726e-06, "loss": 0.0029, "step": 167 }, { "epoch": 9.882352941176471, "grad_norm": 0.07080078125, "learning_rate": 5.138581926697083e-06, "loss": 0.003, "step": 168 }, { "epoch": 9.941176470588236, "grad_norm": 0.04345703125, "learning_rate": 5.09239452479565e-06, "loss": 0.0026, "step": 169 }, { "epoch": 10.0, "grad_norm": 0.048828125, "learning_rate": 5.046199234564455e-06, "loss": 0.0026, "step": 170 }, { "epoch": 10.058823529411764, "grad_norm": 0.046142578125, "learning_rate": 5e-06, "loss": 0.0024, "step": 171 }, { "epoch": 10.117647058823529, "grad_norm": 0.03955078125, "learning_rate": 4.953800765435547e-06, "loss": 0.002, "step": 172 }, { "epoch": 10.176470588235293, "grad_norm": 0.055908203125, "learning_rate": 4.907605475204352e-06, "loss": 0.0024, "step": 173 }, { "epoch": 10.235294117647058, "grad_norm": 0.046630859375, "learning_rate": 4.861418073302919e-06, "loss": 0.0025, "step": 174 }, { "epoch": 10.294117647058824, "grad_norm": 0.044677734375, "learning_rate": 4.815242503054277e-06, "loss": 0.0019, "step": 175 }, { "epoch": 10.352941176470589, "grad_norm": 0.046142578125, "learning_rate": 4.7690827067713035e-06, "loss": 0.0022, "step": 176 }, { "epoch": 10.411764705882353, "grad_norm": 0.057861328125, "learning_rate": 4.7229426254201504e-06, "loss": 0.0025, "step": 177 }, { "epoch": 10.470588235294118, "grad_norm": 0.04931640625, "learning_rate": 4.676826198283779e-06, "loss": 0.0027, "step": 178 }, { "epoch": 10.529411764705882, "grad_norm": 0.044677734375, "learning_rate": 4.630737362625631e-06, "loss": 0.0025, "step": 179 }, { "epoch": 10.588235294117647, "grad_norm": 0.0439453125, "learning_rate": 4.584680053353481e-06, "loss": 0.0021, "step": 180 }, { "epoch": 10.647058823529411, "grad_norm": 0.04443359375, "learning_rate": 4.53865820268349e-06, "loss": 0.0026, "step": 181 }, { "epoch": 10.705882352941176, "grad_norm": 0.05419921875, "learning_rate": 4.492675739804486e-06, "loss": 0.0028, "step": 182 }, { "epoch": 10.764705882352942, "grad_norm": 0.046630859375, "learning_rate": 4.446736590542497e-06, "loss": 0.0023, "step": 183 }, { "epoch": 10.823529411764707, "grad_norm": 0.044189453125, "learning_rate": 4.400844677025585e-06, "loss": 0.0027, "step": 184 }, { "epoch": 10.882352941176471, "grad_norm": 0.08544921875, "learning_rate": 4.355003917348985e-06, "loss": 0.0025, "step": 185 }, { "epoch": 10.941176470588236, "grad_norm": 0.046630859375, "learning_rate": 4.309218225240591e-06, "loss": 0.0022, "step": 186 }, { "epoch": 11.0, "grad_norm": 0.0478515625, "learning_rate": 4.263491509726812e-06, "loss": 0.0024, "step": 187 }, { "epoch": 11.058823529411764, "grad_norm": 0.04638671875, "learning_rate": 4.217827674798845e-06, "loss": 0.0023, "step": 188 }, { "epoch": 11.117647058823529, "grad_norm": 0.04296875, "learning_rate": 4.17223061907935e-06, "loss": 0.0021, "step": 189 }, { "epoch": 11.176470588235293, "grad_norm": 0.046630859375, "learning_rate": 4.126704235489606e-06, "loss": 0.002, "step": 190 }, { "epoch": 11.235294117647058, "grad_norm": 0.047607421875, "learning_rate": 4.081252410917148e-06, "loss": 0.0024, "step": 191 }, { "epoch": 11.294117647058824, "grad_norm": 0.041259765625, "learning_rate": 4.035879025883916e-06, "loss": 0.0023, "step": 192 }, { "epoch": 11.352941176470589, "grad_norm": 0.04345703125, "learning_rate": 3.99058795421495e-06, "loss": 0.0021, "step": 193 }, { "epoch": 11.411764705882353, "grad_norm": 0.047607421875, "learning_rate": 3.945383062707652e-06, "loss": 0.0025, "step": 194 }, { "epoch": 11.470588235294118, "grad_norm": 0.05224609375, "learning_rate": 3.9002682108016585e-06, "loss": 0.0026, "step": 195 }, { "epoch": 11.529411764705882, "grad_norm": 0.03564453125, "learning_rate": 3.855247250249331e-06, "loss": 0.0022, "step": 196 }, { "epoch": 11.588235294117647, "grad_norm": 0.048095703125, "learning_rate": 3.8103240247869077e-06, "loss": 0.0024, "step": 197 }, { "epoch": 11.647058823529411, "grad_norm": 0.0400390625, "learning_rate": 3.765502369806334e-06, "loss": 0.0019, "step": 198 }, { "epoch": 11.705882352941176, "grad_norm": 0.043212890625, "learning_rate": 3.720786112027822e-06, "loss": 0.0024, "step": 199 }, { "epoch": 11.764705882352942, "grad_norm": 0.03759765625, "learning_rate": 3.6761790691731207e-06, "loss": 0.0021, "step": 200 }, { "epoch": 11.823529411764707, "grad_norm": 0.04833984375, "learning_rate": 3.6316850496395863e-06, "loss": 0.0023, "step": 201 }, { "epoch": 11.882352941176471, "grad_norm": 0.035400390625, "learning_rate": 3.587307852175025e-06, "loss": 0.0018, "step": 202 }, { "epoch": 11.941176470588236, "grad_norm": 0.03662109375, "learning_rate": 3.5430512655533774e-06, "loss": 0.0021, "step": 203 }, { "epoch": 12.0, "grad_norm": 0.042724609375, "learning_rate": 3.498919068251237e-06, "loss": 0.0021, "step": 204 }, { "epoch": 12.058823529411764, "grad_norm": 0.045166015625, "learning_rate": 3.4549150281252635e-06, "loss": 0.0022, "step": 205 }, { "epoch": 12.117647058823529, "grad_norm": 0.0439453125, "learning_rate": 3.4110429020904924e-06, "loss": 0.0022, "step": 206 }, { "epoch": 12.176470588235293, "grad_norm": 0.033203125, "learning_rate": 3.3673064357995844e-06, "loss": 0.0019, "step": 207 }, { "epoch": 12.235294117647058, "grad_norm": 0.04248046875, "learning_rate": 3.3237093633230323e-06, "loss": 0.0022, "step": 208 }, { "epoch": 12.294117647058824, "grad_norm": 0.041015625, "learning_rate": 3.2802554068303595e-06, "loss": 0.0018, "step": 209 }, { "epoch": 12.352941176470589, "grad_norm": 0.041015625, "learning_rate": 3.236948276272337e-06, "loss": 0.002, "step": 210 }, { "epoch": 12.411764705882353, "grad_norm": 0.0517578125, "learning_rate": 3.1937916690642356e-06, "loss": 0.0025, "step": 211 }, { "epoch": 12.470588235294118, "grad_norm": 0.037841796875, "learning_rate": 3.150789269770155e-06, "loss": 0.0022, "step": 212 }, { "epoch": 12.529411764705882, "grad_norm": 0.037109375, "learning_rate": 3.107944749788449e-06, "loss": 0.0022, "step": 213 }, { "epoch": 12.588235294117647, "grad_norm": 0.0390625, "learning_rate": 3.0652617670382745e-06, "loss": 0.0021, "step": 214 }, { "epoch": 12.647058823529411, "grad_norm": 0.035400390625, "learning_rate": 3.0227439656472878e-06, "loss": 0.0018, "step": 215 }, { "epoch": 12.705882352941176, "grad_norm": 0.036376953125, "learning_rate": 2.980394975640526e-06, "loss": 0.002, "step": 216 }, { "epoch": 12.764705882352942, "grad_norm": 0.0439453125, "learning_rate": 2.9382184126304834e-06, "loss": 0.0023, "step": 217 }, { "epoch": 12.823529411764707, "grad_norm": 0.044921875, "learning_rate": 2.8962178775084267e-06, "loss": 0.0022, "step": 218 }, { "epoch": 12.882352941176471, "grad_norm": 0.0322265625, "learning_rate": 2.8543969561369556e-06, "loss": 0.0017, "step": 219 }, { "epoch": 12.941176470588236, "grad_norm": 0.045654296875, "learning_rate": 2.812759219043869e-06, "loss": 0.0022, "step": 220 }, { "epoch": 13.0, "grad_norm": 0.05029296875, "learning_rate": 2.771308221117309e-06, "loss": 0.0022, "step": 221 }, { "epoch": 13.058823529411764, "grad_norm": 0.041748046875, "learning_rate": 2.7300475013022666e-06, "loss": 0.0018, "step": 222 }, { "epoch": 13.117647058823529, "grad_norm": 0.0341796875, "learning_rate": 2.6889805822984348e-06, "loss": 0.0018, "step": 223 }, { "epoch": 13.176470588235293, "grad_norm": 0.0400390625, "learning_rate": 2.648110970259454e-06, "loss": 0.0023, "step": 224 }, { "epoch": 13.235294117647058, "grad_norm": 0.034423828125, "learning_rate": 2.607442154493568e-06, "loss": 0.0019, "step": 225 }, { "epoch": 13.294117647058824, "grad_norm": 0.04248046875, "learning_rate": 2.5669776071657194e-06, "loss": 0.002, "step": 226 }, { "epoch": 13.352941176470589, "grad_norm": 0.03662109375, "learning_rate": 2.526720783001107e-06, "loss": 0.0021, "step": 227 }, { "epoch": 13.411764705882353, "grad_norm": 0.0419921875, "learning_rate": 2.486675118990233e-06, "loss": 0.002, "step": 228 }, { "epoch": 13.470588235294118, "grad_norm": 0.036376953125, "learning_rate": 2.4468440340954664e-06, "loss": 0.002, "step": 229 }, { "epoch": 13.529411764705882, "grad_norm": 0.034423828125, "learning_rate": 2.4072309289591394e-06, "loss": 0.002, "step": 230 }, { "epoch": 13.588235294117647, "grad_norm": 0.039794921875, "learning_rate": 2.3678391856132203e-06, "loss": 0.0021, "step": 231 }, { "epoch": 13.647058823529411, "grad_norm": 0.049072265625, "learning_rate": 2.328672167190558e-06, "loss": 0.0023, "step": 232 }, { "epoch": 13.705882352941176, "grad_norm": 0.04248046875, "learning_rate": 2.289733217637753e-06, "loss": 0.0021, "step": 233 }, { "epoch": 13.764705882352942, "grad_norm": 0.044677734375, "learning_rate": 2.2510256614296638e-06, "loss": 0.0022, "step": 234 }, { "epoch": 13.823529411764707, "grad_norm": 0.034912109375, "learning_rate": 2.2125528032855727e-06, "loss": 0.0018, "step": 235 }, { "epoch": 13.882352941176471, "grad_norm": 0.039306640625, "learning_rate": 2.174317927887041e-06, "loss": 0.002, "step": 236 }, { "epoch": 13.941176470588236, "grad_norm": 0.049560546875, "learning_rate": 2.136324299597474e-06, "loss": 0.0024, "step": 237 }, { "epoch": 14.0, "grad_norm": 0.03759765625, "learning_rate": 2.098575162183422e-06, "loss": 0.0019, "step": 238 }, { "epoch": 14.058823529411764, "grad_norm": 0.037353515625, "learning_rate": 2.061073738537635e-06, "loss": 0.0018, "step": 239 }, { "epoch": 14.117647058823529, "grad_norm": 0.03564453125, "learning_rate": 2.023823230403907e-06, "loss": 0.0019, "step": 240 }, { "epoch": 14.176470588235293, "grad_norm": 0.031494140625, "learning_rate": 1.9868268181037186e-06, "loss": 0.0018, "step": 241 }, { "epoch": 14.235294117647058, "grad_norm": 0.043212890625, "learning_rate": 1.9500876602647167e-06, "loss": 0.0019, "step": 242 }, { "epoch": 14.294117647058824, "grad_norm": 0.045166015625, "learning_rate": 1.913608893551036e-06, "loss": 0.0023, "step": 243 }, { "epoch": 14.352941176470589, "grad_norm": 0.04541015625, "learning_rate": 1.8773936323955055e-06, "loss": 0.0022, "step": 244 }, { "epoch": 14.411764705882353, "grad_norm": 0.03955078125, "learning_rate": 1.8414449687337467e-06, "loss": 0.0021, "step": 245 }, { "epoch": 14.470588235294118, "grad_norm": 0.047119140625, "learning_rate": 1.8057659717401948e-06, "loss": 0.0023, "step": 246 }, { "epoch": 14.529411764705882, "grad_norm": 0.04541015625, "learning_rate": 1.7703596875660645e-06, "loss": 0.002, "step": 247 }, { "epoch": 14.588235294117647, "grad_norm": 0.03466796875, "learning_rate": 1.7352291390792798e-06, "loss": 0.0018, "step": 248 }, { "epoch": 14.647058823529411, "grad_norm": 0.033203125, "learning_rate": 1.7003773256063882e-06, "loss": 0.0019, "step": 249 }, { "epoch": 14.705882352941176, "grad_norm": 0.03466796875, "learning_rate": 1.6658072226764949e-06, "loss": 0.002, "step": 250 }, { "epoch": 14.764705882352942, "grad_norm": 0.03955078125, "learning_rate": 1.6315217817672142e-06, "loss": 0.002, "step": 251 }, { "epoch": 14.823529411764707, "grad_norm": 0.04296875, "learning_rate": 1.5975239300526924e-06, "loss": 0.002, "step": 252 }, { "epoch": 14.882352941176471, "grad_norm": 0.0341796875, "learning_rate": 1.5638165701536866e-06, "loss": 0.0019, "step": 253 }, { "epoch": 14.941176470588236, "grad_norm": 0.04296875, "learning_rate": 1.5304025798897521e-06, "loss": 0.0023, "step": 254 }, { "epoch": 15.0, "grad_norm": 0.0380859375, "learning_rate": 1.4972848120335453e-06, "loss": 0.0019, "step": 255 }, { "epoch": 15.058823529411764, "grad_norm": 0.03759765625, "learning_rate": 1.4644660940672628e-06, "loss": 0.002, "step": 256 }, { "epoch": 15.117647058823529, "grad_norm": 0.044677734375, "learning_rate": 1.4319492279412388e-06, "loss": 0.0022, "step": 257 }, { "epoch": 15.176470588235293, "grad_norm": 0.0458984375, "learning_rate": 1.399736989834728e-06, "loss": 0.0021, "step": 258 }, { "epoch": 15.235294117647058, "grad_norm": 0.043701171875, "learning_rate": 1.3678321299188802e-06, "loss": 0.0023, "step": 259 }, { "epoch": 15.294117647058824, "grad_norm": 0.039306640625, "learning_rate": 1.336237372121944e-06, "loss": 0.0022, "step": 260 }, { "epoch": 15.352941176470589, "grad_norm": 0.0361328125, "learning_rate": 1.3049554138967052e-06, "loss": 0.0019, "step": 261 }, { "epoch": 15.411764705882353, "grad_norm": 0.034912109375, "learning_rate": 1.2739889259901866e-06, "loss": 0.002, "step": 262 }, { "epoch": 15.470588235294118, "grad_norm": 0.044189453125, "learning_rate": 1.2433405522156334e-06, "loss": 0.0023, "step": 263 }, { "epoch": 15.529411764705882, "grad_norm": 0.04052734375, "learning_rate": 1.213012909226786e-06, "loss": 0.002, "step": 264 }, { "epoch": 15.588235294117647, "grad_norm": 0.043212890625, "learning_rate": 1.1830085862944851e-06, "loss": 0.002, "step": 265 }, { "epoch": 15.647058823529411, "grad_norm": 0.03466796875, "learning_rate": 1.1533301450856054e-06, "loss": 0.0018, "step": 266 }, { "epoch": 15.705882352941176, "grad_norm": 0.03466796875, "learning_rate": 1.1239801194443507e-06, "loss": 0.0021, "step": 267 }, { "epoch": 15.764705882352942, "grad_norm": 0.031494140625, "learning_rate": 1.0949610151759233e-06, "loss": 0.0018, "step": 268 }, { "epoch": 15.823529411764707, "grad_norm": 0.039794921875, "learning_rate": 1.066275309832584e-06, "loss": 0.0017, "step": 269 }, { "epoch": 15.882352941176471, "grad_norm": 0.02783203125, "learning_rate": 1.037925452502131e-06, "loss": 0.0018, "step": 270 }, { "epoch": 15.941176470588236, "grad_norm": 0.041259765625, "learning_rate": 1.0099138635988026e-06, "loss": 0.002, "step": 271 }, { "epoch": 16.0, "grad_norm": 0.0400390625, "learning_rate": 9.822429346566314e-07, "loss": 0.0021, "step": 272 }, { "epoch": 16.058823529411764, "grad_norm": 0.0478515625, "learning_rate": 9.549150281252633e-07, "loss": 0.0024, "step": 273 }, { "epoch": 16.11764705882353, "grad_norm": 0.0341796875, "learning_rate": 9.279324771682586e-07, "loss": 0.0017, "step": 274 }, { "epoch": 16.176470588235293, "grad_norm": 0.034423828125, "learning_rate": 9.01297585463895e-07, "loss": 0.002, "step": 275 }, { "epoch": 16.235294117647058, "grad_norm": 0.036376953125, "learning_rate": 8.750126270084891e-07, "loss": 0.0017, "step": 276 }, { "epoch": 16.294117647058822, "grad_norm": 0.041259765625, "learning_rate": 8.490798459222477e-07, "loss": 0.0021, "step": 277 }, { "epoch": 16.352941176470587, "grad_norm": 0.039794921875, "learning_rate": 8.235014562576732e-07, "loss": 0.0022, "step": 278 }, { "epoch": 16.41176470588235, "grad_norm": 0.04248046875, "learning_rate": 7.98279641810537e-07, "loss": 0.002, "step": 279 }, { "epoch": 16.470588235294116, "grad_norm": 0.036865234375, "learning_rate": 7.734165559334327e-07, "loss": 0.002, "step": 280 }, { "epoch": 16.529411764705884, "grad_norm": 0.05517578125, "learning_rate": 7.489143213519301e-07, "loss": 0.0024, "step": 281 }, { "epoch": 16.58823529411765, "grad_norm": 0.04052734375, "learning_rate": 7.24775029983345e-07, "loss": 0.0019, "step": 282 }, { "epoch": 16.647058823529413, "grad_norm": 0.03759765625, "learning_rate": 7.010007427581378e-07, "loss": 0.002, "step": 283 }, { "epoch": 16.705882352941178, "grad_norm": 0.031494140625, "learning_rate": 6.775934894439606e-07, "loss": 0.0018, "step": 284 }, { "epoch": 16.764705882352942, "grad_norm": 0.03271484375, "learning_rate": 6.545552684723583e-07, "loss": 0.0018, "step": 285 }, { "epoch": 16.823529411764707, "grad_norm": 0.043701171875, "learning_rate": 6.318880467681527e-07, "loss": 0.0022, "step": 286 }, { "epoch": 16.88235294117647, "grad_norm": 0.037841796875, "learning_rate": 6.095937595815104e-07, "loss": 0.002, "step": 287 }, { "epoch": 16.941176470588236, "grad_norm": 0.033203125, "learning_rate": 5.876743103227217e-07, "loss": 0.0017, "step": 288 }, { "epoch": 17.0, "grad_norm": 0.040771484375, "learning_rate": 5.661315703996905e-07, "loss": 0.0024, "step": 289 }, { "epoch": 17.058823529411764, "grad_norm": 0.040771484375, "learning_rate": 5.449673790581611e-07, "loss": 0.0018, "step": 290 }, { "epoch": 17.11764705882353, "grad_norm": 0.041748046875, "learning_rate": 5.241835432246888e-07, "loss": 0.002, "step": 291 }, { "epoch": 17.176470588235293, "grad_norm": 0.047119140625, "learning_rate": 5.037818373523723e-07, "loss": 0.0019, "step": 292 }, { "epoch": 17.235294117647058, "grad_norm": 0.0419921875, "learning_rate": 4.837640032693558e-07, "loss": 0.0021, "step": 293 }, { "epoch": 17.294117647058822, "grad_norm": 0.0322265625, "learning_rate": 4.641317500301173e-07, "loss": 0.0019, "step": 294 }, { "epoch": 17.352941176470587, "grad_norm": 0.042236328125, "learning_rate": 4.448867537695578e-07, "loss": 0.002, "step": 295 }, { "epoch": 17.41176470588235, "grad_norm": 0.04345703125, "learning_rate": 4.2603065755989493e-07, "loss": 0.0022, "step": 296 }, { "epoch": 17.470588235294116, "grad_norm": 0.034912109375, "learning_rate": 4.0756507127038494e-07, "loss": 0.002, "step": 297 }, { "epoch": 17.529411764705884, "grad_norm": 0.035400390625, "learning_rate": 3.894915714298775e-07, "loss": 0.0019, "step": 298 }, { "epoch": 17.58823529411765, "grad_norm": 0.0311279296875, "learning_rate": 3.71811701092219e-07, "loss": 0.0018, "step": 299 }, { "epoch": 17.647058823529413, "grad_norm": 0.039306640625, "learning_rate": 3.5452696970450674e-07, "loss": 0.0019, "step": 300 }, { "epoch": 17.705882352941178, "grad_norm": 0.037109375, "learning_rate": 3.3763885297822153e-07, "loss": 0.0019, "step": 301 }, { "epoch": 17.764705882352942, "grad_norm": 0.036865234375, "learning_rate": 3.2114879276323783e-07, "loss": 0.002, "step": 302 }, { "epoch": 17.823529411764707, "grad_norm": 0.0380859375, "learning_rate": 3.0505819692471797e-07, "loss": 0.002, "step": 303 }, { "epoch": 17.88235294117647, "grad_norm": 0.03662109375, "learning_rate": 2.893684392229185e-07, "loss": 0.002, "step": 304 }, { "epoch": 17.941176470588236, "grad_norm": 0.050537109375, "learning_rate": 2.7408085919590265e-07, "loss": 0.0023, "step": 305 }, { "epoch": 18.0, "grad_norm": 0.04296875, "learning_rate": 2.5919676204517073e-07, "loss": 0.002, "step": 306 }, { "epoch": 18.058823529411764, "grad_norm": 0.037353515625, "learning_rate": 2.447174185242324e-07, "loss": 0.0022, "step": 307 }, { "epoch": 18.11764705882353, "grad_norm": 0.039306640625, "learning_rate": 2.3064406483010947e-07, "loss": 0.0019, "step": 308 }, { "epoch": 18.176470588235293, "grad_norm": 0.048828125, "learning_rate": 2.1697790249779638e-07, "loss": 0.0024, "step": 309 }, { "epoch": 18.235294117647058, "grad_norm": 0.0390625, "learning_rate": 2.0372009829767558e-07, "loss": 0.0019, "step": 310 }, { "epoch": 18.294117647058822, "grad_norm": 0.034423828125, "learning_rate": 1.908717841359048e-07, "loss": 0.0017, "step": 311 }, { "epoch": 18.352941176470587, "grad_norm": 0.038818359375, "learning_rate": 1.7843405695777582e-07, "loss": 0.0021, "step": 312 }, { "epoch": 18.41176470588235, "grad_norm": 0.047119140625, "learning_rate": 1.664079786540629e-07, "loss": 0.0023, "step": 313 }, { "epoch": 18.470588235294116, "grad_norm": 0.041748046875, "learning_rate": 1.547945759703623e-07, "loss": 0.0019, "step": 314 }, { "epoch": 18.529411764705884, "grad_norm": 0.042724609375, "learning_rate": 1.435948404194304e-07, "loss": 0.0019, "step": 315 }, { "epoch": 18.58823529411765, "grad_norm": 0.038818359375, "learning_rate": 1.328097281965357e-07, "loss": 0.0021, "step": 316 }, { "epoch": 18.647058823529413, "grad_norm": 0.03662109375, "learning_rate": 1.22440160097817e-07, "loss": 0.0018, "step": 317 }, { "epoch": 18.705882352941178, "grad_norm": 0.0400390625, "learning_rate": 1.1248702144167123e-07, "loss": 0.002, "step": 318 }, { "epoch": 18.764705882352942, "grad_norm": 0.0361328125, "learning_rate": 1.0295116199317057e-07, "loss": 0.0019, "step": 319 }, { "epoch": 18.823529411764707, "grad_norm": 0.046875, "learning_rate": 9.383339589150776e-08, "loss": 0.002, "step": 320 }, { "epoch": 18.88235294117647, "grad_norm": 0.03857421875, "learning_rate": 8.513450158049109e-08, "loss": 0.0019, "step": 321 }, { "epoch": 18.941176470588236, "grad_norm": 0.03662109375, "learning_rate": 7.685522174208205e-08, "loss": 0.0019, "step": 322 }, { "epoch": 19.0, "grad_norm": 0.035888671875, "learning_rate": 6.899626323298714e-08, "loss": 0.0021, "step": 323 }, { "epoch": 19.058823529411764, "grad_norm": 0.032470703125, "learning_rate": 6.15582970243117e-08, "loss": 0.0019, "step": 324 }, { "epoch": 19.11764705882353, "grad_norm": 0.038330078125, "learning_rate": 5.454195814427021e-08, "loss": 0.0019, "step": 325 }, { "epoch": 19.176470588235293, "grad_norm": 0.046142578125, "learning_rate": 4.794784562397459e-08, "loss": 0.0023, "step": 326 }, { "epoch": 19.235294117647058, "grad_norm": 0.032958984375, "learning_rate": 4.177652244628627e-08, "loss": 0.0019, "step": 327 }, { "epoch": 19.294117647058822, "grad_norm": 0.037109375, "learning_rate": 3.602851549775521e-08, "loss": 0.002, "step": 328 }, { "epoch": 19.352941176470587, "grad_norm": 0.038330078125, "learning_rate": 3.0704315523631956e-08, "loss": 0.002, "step": 329 }, { "epoch": 19.41176470588235, "grad_norm": 0.03564453125, "learning_rate": 2.5804377085972278e-08, "loss": 0.002, "step": 330 }, { "epoch": 19.470588235294116, "grad_norm": 0.03369140625, "learning_rate": 2.1329118524827662e-08, "loss": 0.002, "step": 331 }, { "epoch": 19.529411764705884, "grad_norm": 0.037353515625, "learning_rate": 1.7278921922527224e-08, "loss": 0.0018, "step": 332 }, { "epoch": 19.58823529411765, "grad_norm": 0.03857421875, "learning_rate": 1.3654133071059894e-08, "loss": 0.0019, "step": 333 }, { "epoch": 19.647058823529413, "grad_norm": 0.040771484375, "learning_rate": 1.0455061442548597e-08, "loss": 0.0021, "step": 334 }, { "epoch": 19.705882352941178, "grad_norm": 0.038818359375, "learning_rate": 7.681980162830283e-09, "loss": 0.0018, "step": 335 }, { "epoch": 19.764705882352942, "grad_norm": 0.042724609375, "learning_rate": 5.3351259881379016e-09, "loss": 0.0022, "step": 336 }, { "epoch": 19.823529411764707, "grad_norm": 0.0478515625, "learning_rate": 3.41469928488547e-09, "loss": 0.0022, "step": 337 }, { "epoch": 19.88235294117647, "grad_norm": 0.032958984375, "learning_rate": 1.9208640125628618e-09, "loss": 0.0019, "step": 338 }, { "epoch": 19.941176470588236, "grad_norm": 0.039794921875, "learning_rate": 8.537477097364522e-10, "loss": 0.0019, "step": 339 }, { "epoch": 20.0, "grad_norm": 0.035400390625, "learning_rate": 2.1344148316060352e-10, "loss": 0.0021, "step": 340 } ], "logging_steps": 1.0, "max_steps": 340, "num_input_tokens_seen": 0, "num_train_epochs": 20, "save_steps": 0, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 1.155431961846612e+18, "train_batch_size": 1, "trial_name": null, "trial_params": null }