diff --git "a/checkpoint-13000/trainer_state.json" "b/checkpoint-13000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-13000/trainer_state.json"
@@ -0,0 +1,91033 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 19.461077844311376,
+  "eval_steps": 500,
+  "global_step": 13000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0014970059880239522,
+      "grad_norm": 12.25,
+      "learning_rate": 0.0008,
+      "loss": 3.8621,
+      "step": 1
+    },
+    {
+      "epoch": 0.0029940119760479044,
+      "grad_norm": 147.0,
+      "learning_rate": 0.0008,
+      "loss": 17.8731,
+      "step": 2
+    },
+    {
+      "epoch": 0.004491017964071856,
+      "grad_norm": 19.375,
+      "learning_rate": 0.0008,
+      "loss": 3.9524,
+      "step": 3
+    },
+    {
+      "epoch": 0.005988023952095809,
+      "grad_norm": 39.25,
+      "learning_rate": 0.0008,
+      "loss": 9.6708,
+      "step": 4
+    },
+    {
+      "epoch": 0.0074850299401197605,
+      "grad_norm": 8.5,
+      "learning_rate": 0.0008,
+      "loss": 2.3872,
+      "step": 5
+    },
+    {
+      "epoch": 0.008982035928143712,
+      "grad_norm": 25.0,
+      "learning_rate": 0.0008,
+      "loss": 3.4757,
+      "step": 6
+    },
+    {
+      "epoch": 0.010479041916167664,
+      "grad_norm": 6.6875,
+      "learning_rate": 0.0008,
+      "loss": 2.1171,
+      "step": 7
+    },
+    {
+      "epoch": 0.011976047904191617,
+      "grad_norm": 29.0,
+      "learning_rate": 0.0008,
+      "loss": 3.6517,
+      "step": 8
+    },
+    {
+      "epoch": 0.01347305389221557,
+      "grad_norm": 33.5,
+      "learning_rate": 0.0008,
+      "loss": 4.0679,
+      "step": 9
+    },
+    {
+      "epoch": 0.014970059880239521,
+      "grad_norm": 15.1875,
+      "learning_rate": 0.0008,
+      "loss": 2.6297,
+      "step": 10
+    },
+    {
+      "epoch": 0.016467065868263474,
+      "grad_norm": 5.4375,
+      "learning_rate": 0.0008,
+      "loss": 2.1564,
+      "step": 11
+    },
+    {
+      "epoch": 0.017964071856287425,
+      "grad_norm": 6.5,
+      "learning_rate": 0.0008,
+      "loss": 2.2034,
+      "step": 12
+    },
+    {
+      "epoch": 0.019461077844311378,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.0008,
+      "loss": 2.0095,
+      "step": 13
+    },
+    {
+      "epoch": 0.020958083832335328,
+      "grad_norm": 5.5,
+      "learning_rate": 0.0008,
+      "loss": 2.2219,
+      "step": 14
+    },
+    {
+      "epoch": 0.02245508982035928,
+      "grad_norm": 3.359375,
+      "learning_rate": 0.0008,
+      "loss": 2.1824,
+      "step": 15
+    },
+    {
+      "epoch": 0.023952095808383235,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.0008,
+      "loss": 2.0574,
+      "step": 16
+    },
+    {
+      "epoch": 0.025449101796407185,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.0008,
+      "loss": 2.0102,
+      "step": 17
+    },
+    {
+      "epoch": 0.02694610778443114,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0008,
+      "loss": 2.0261,
+      "step": 18
+    },
+    {
+      "epoch": 0.02844311377245509,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.0008,
+      "loss": 1.9611,
+      "step": 19
+    },
+    {
+      "epoch": 0.029940119760479042,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.9503,
+      "step": 20
+    },
+    {
+      "epoch": 0.03143712574850299,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.0008,
+      "loss": 1.9254,
+      "step": 21
+    },
+    {
+      "epoch": 0.03293413173652695,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.0008,
+      "loss": 1.9061,
+      "step": 22
+    },
+    {
+      "epoch": 0.0344311377245509,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.0008,
+      "loss": 1.875,
+      "step": 23
+    },
+    {
+      "epoch": 0.03592814371257485,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0008,
+      "loss": 1.8203,
+      "step": 24
+    },
+    {
+      "epoch": 0.0374251497005988,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.0008,
+      "loss": 1.7867,
+      "step": 25
+    },
+    {
+      "epoch": 0.038922155688622756,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.0008,
+      "loss": 1.7805,
+      "step": 26
+    },
+    {
+      "epoch": 0.040419161676646706,
+      "grad_norm": 1.25,
+      "learning_rate": 0.0008,
+      "loss": 1.7473,
+      "step": 27
+    },
+    {
+      "epoch": 0.041916167664670656,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0008,
+      "loss": 1.7456,
+      "step": 28
+    },
+    {
+      "epoch": 0.04341317365269461,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0008,
+      "loss": 1.776,
+      "step": 29
+    },
+    {
+      "epoch": 0.04491017964071856,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.0008,
+      "loss": 1.722,
+      "step": 30
+    },
+    {
+      "epoch": 0.04640718562874251,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.0008,
+      "loss": 1.6853,
+      "step": 31
+    },
+    {
+      "epoch": 0.04790419161676647,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0008,
+      "loss": 1.649,
+      "step": 32
+    },
+    {
+      "epoch": 0.04940119760479042,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0008,
+      "loss": 1.6499,
+      "step": 33
+    },
+    {
+      "epoch": 0.05089820359281437,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0008,
+      "loss": 1.6225,
+      "step": 34
+    },
+    {
+      "epoch": 0.05239520958083832,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0008,
+      "loss": 1.6077,
+      "step": 35
+    },
+    {
+      "epoch": 0.05389221556886228,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0008,
+      "loss": 1.5821,
+      "step": 36
+    },
+    {
+      "epoch": 0.05538922155688623,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0008,
+      "loss": 1.5864,
+      "step": 37
+    },
+    {
+      "epoch": 0.05688622754491018,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0008,
+      "loss": 1.5684,
+      "step": 38
+    },
+    {
+      "epoch": 0.058383233532934134,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.5676,
+      "step": 39
+    },
+    {
+      "epoch": 0.059880239520958084,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0008,
+      "loss": 1.554,
+      "step": 40
+    },
+    {
+      "epoch": 0.061377245508982034,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.5503,
+      "step": 41
+    },
+    {
+      "epoch": 0.06287425149700598,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0008,
+      "loss": 1.5388,
+      "step": 42
+    },
+    {
+      "epoch": 0.06437125748502993,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0008,
+      "loss": 1.547,
+      "step": 43
+    },
+    {
+      "epoch": 0.0658682634730539,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0008,
+      "loss": 1.5204,
+      "step": 44
+    },
+    {
+      "epoch": 0.06736526946107785,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.511,
+      "step": 45
+    },
+    {
+      "epoch": 0.0688622754491018,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.5279,
+      "step": 46
+    },
+    {
+      "epoch": 0.07035928143712575,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0008,
+      "loss": 1.5328,
+      "step": 47
+    },
+    {
+      "epoch": 0.0718562874251497,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0008,
+      "loss": 1.5189,
+      "step": 48
+    },
+    {
+      "epoch": 0.07335329341317365,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 0.0008,
+      "loss": 1.5183,
+      "step": 49
+    },
+    {
+      "epoch": 0.0748502994011976,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.5107,
+      "step": 50
+    },
+    {
+      "epoch": 0.07634730538922156,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0008,
+      "loss": 1.507,
+      "step": 51
+    },
+    {
+      "epoch": 0.07784431137724551,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.483,
+      "step": 52
+    },
+    {
+      "epoch": 0.07934131736526946,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.4858,
+      "step": 53
+    },
+    {
+      "epoch": 0.08083832335329341,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.4907,
+      "step": 54
+    },
+    {
+      "epoch": 0.08233532934131736,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0008,
+      "loss": 1.4804,
+      "step": 55
+    },
+    {
+      "epoch": 0.08383233532934131,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0008,
+      "loss": 1.4833,
+      "step": 56
+    },
+    {
+      "epoch": 0.08532934131736528,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0008,
+      "loss": 1.4743,
+      "step": 57
+    },
+    {
+      "epoch": 0.08682634730538923,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0008,
+      "loss": 1.4808,
+      "step": 58
+    },
+    {
+      "epoch": 0.08832335329341318,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0008,
+      "loss": 1.4926,
+      "step": 59
+    },
+    {
+      "epoch": 0.08982035928143713,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0008,
+      "loss": 1.4869,
+      "step": 60
+    },
+    {
+      "epoch": 0.09131736526946108,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.4629,
+      "step": 61
+    },
+    {
+      "epoch": 0.09281437125748503,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0008,
+      "loss": 1.4634,
+      "step": 62
+    },
+    {
+      "epoch": 0.09431137724550898,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0008,
+      "loss": 1.4597,
+      "step": 63
+    },
+    {
+      "epoch": 0.09580838323353294,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.4483,
+      "step": 64
+    },
+    {
+      "epoch": 0.09730538922155689,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0008,
+      "loss": 1.452,
+      "step": 65
+    },
+    {
+      "epoch": 0.09880239520958084,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0008,
+      "loss": 1.4557,
+      "step": 66
+    },
+    {
+      "epoch": 0.10029940119760479,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0008,
+      "loss": 1.4346,
+      "step": 67
+    },
+    {
+      "epoch": 0.10179640718562874,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.4549,
+      "step": 68
+    },
+    {
+      "epoch": 0.10329341317365269,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0008,
+      "loss": 1.4367,
+      "step": 69
+    },
+    {
+      "epoch": 0.10479041916167664,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0008,
+      "loss": 1.4298,
+      "step": 70
+    },
+    {
+      "epoch": 0.1062874251497006,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0008,
+      "loss": 1.4434,
+      "step": 71
+    },
+    {
+      "epoch": 0.10778443113772455,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0008,
+      "loss": 1.4333,
+      "step": 72
+    },
+    {
+      "epoch": 0.1092814371257485,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0008,
+      "loss": 1.4331,
+      "step": 73
+    },
+    {
+      "epoch": 0.11077844311377245,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0008,
+      "loss": 1.4263,
+      "step": 74
+    },
+    {
+      "epoch": 0.1122754491017964,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0008,
+      "loss": 1.4307,
+      "step": 75
+    },
+    {
+      "epoch": 0.11377245508982035,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0008,
+      "loss": 1.4165,
+      "step": 76
+    },
+    {
+      "epoch": 0.11526946107784432,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0008,
+      "loss": 1.4138,
+      "step": 77
+    },
+    {
+      "epoch": 0.11676646706586827,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0008,
+      "loss": 1.4204,
+      "step": 78
+    },
+    {
+      "epoch": 0.11826347305389222,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0008,
+      "loss": 1.4122,
+      "step": 79
+    },
+    {
+      "epoch": 0.11976047904191617,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0008,
+      "loss": 1.4143,
+      "step": 80
+    },
+    {
+      "epoch": 0.12125748502994012,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0008,
+      "loss": 1.4032,
+      "step": 81
+    },
+    {
+      "epoch": 0.12275449101796407,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0008,
+      "loss": 1.4069,
+      "step": 82
+    },
+    {
+      "epoch": 0.12425149700598802,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0008,
+      "loss": 1.4059,
+      "step": 83
+    },
+    {
+      "epoch": 0.12574850299401197,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0008,
+      "loss": 1.4065,
+      "step": 84
+    },
+    {
+      "epoch": 0.12724550898203593,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0008,
+      "loss": 1.3974,
+      "step": 85
+    },
+    {
+      "epoch": 0.12874251497005987,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0008,
+      "loss": 1.3896,
+      "step": 86
+    },
+    {
+      "epoch": 0.13023952095808383,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0008,
+      "loss": 1.4011,
+      "step": 87
+    },
+    {
+      "epoch": 0.1317365269461078,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0008,
+      "loss": 1.391,
+      "step": 88
+    },
+    {
+      "epoch": 0.13323353293413173,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0008,
+      "loss": 1.3908,
+      "step": 89
+    },
+    {
+      "epoch": 0.1347305389221557,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.0008,
+      "loss": 1.4263,
+      "step": 90
+    },
+    {
+      "epoch": 0.13622754491017963,
+      "grad_norm": 5.34375,
+      "learning_rate": 0.0008,
+      "loss": 1.5911,
+      "step": 91
+    },
+    {
+      "epoch": 0.1377245508982036,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.0008,
+      "loss": 1.4479,
+      "step": 92
+    },
+    {
+      "epoch": 0.13922155688622753,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.0008,
+      "loss": 1.5098,
+      "step": 93
+    },
+    {
+      "epoch": 0.1407185628742515,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0008,
+      "loss": 1.4637,
+      "step": 94
+    },
+    {
+      "epoch": 0.14221556886227546,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0008,
+      "loss": 1.4494,
+      "step": 95
+    },
+    {
+      "epoch": 0.1437125748502994,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0008,
+      "loss": 1.4602,
+      "step": 96
+    },
+    {
+      "epoch": 0.14520958083832336,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.0008,
+      "loss": 1.4749,
+      "step": 97
+    },
+    {
+      "epoch": 0.1467065868263473,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.0008,
+      "loss": 1.4621,
+      "step": 98
+    },
+    {
+      "epoch": 0.14820359281437126,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0008,
+      "loss": 1.4306,
+      "step": 99
+    },
+    {
+      "epoch": 0.1497005988023952,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0008,
+      "loss": 1.4282,
+      "step": 100
+    },
+    {
+      "epoch": 0.15119760479041916,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0008,
+      "loss": 1.4251,
+      "step": 101
+    },
+    {
+      "epoch": 0.15269461077844312,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0008,
+      "loss": 1.4319,
+      "step": 102
+    },
+    {
+      "epoch": 0.15419161676646706,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0008,
+      "loss": 1.4123,
+      "step": 103
+    },
+    {
+      "epoch": 0.15568862275449102,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0008,
+      "loss": 1.4124,
+      "step": 104
+    },
+    {
+      "epoch": 0.15718562874251496,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0008,
+      "loss": 1.4222,
+      "step": 105
+    },
+    {
+      "epoch": 0.15868263473053892,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0008,
+      "loss": 1.4084,
+      "step": 106
+    },
+    {
+      "epoch": 0.1601796407185629,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0008,
+      "loss": 1.4113,
+      "step": 107
+    },
+    {
+      "epoch": 0.16167664670658682,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0008,
+      "loss": 1.403,
+      "step": 108
+    },
+    {
+      "epoch": 0.1631736526946108,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0008,
+      "loss": 1.3906,
+      "step": 109
+    },
+    {
+      "epoch": 0.16467065868263472,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.3873,
+      "step": 110
+    },
+    {
+      "epoch": 0.1661676646706587,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0008,
+      "loss": 1.3932,
+      "step": 111
+    },
+    {
+      "epoch": 0.16766467065868262,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0008,
+      "loss": 1.397,
+      "step": 112
+    },
+    {
+      "epoch": 0.1691616766467066,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0008,
+      "loss": 1.388,
+      "step": 113
+    },
+    {
+      "epoch": 0.17065868263473055,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0008,
+      "loss": 1.3821,
+      "step": 114
+    },
+    {
+      "epoch": 0.1721556886227545,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0008,
+      "loss": 1.3754,
+      "step": 115
+    },
+    {
+      "epoch": 0.17365269461077845,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0008,
+      "loss": 1.3932,
+      "step": 116
+    },
+    {
+      "epoch": 0.1751497005988024,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.0008,
+      "loss": 1.3986,
+      "step": 117
+    },
+    {
+      "epoch": 0.17664670658682635,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0008,
+      "loss": 1.4088,
+      "step": 118
+    },
+    {
+      "epoch": 0.1781437125748503,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.0008,
+      "loss": 1.4281,
+      "step": 119
+    },
+    {
+      "epoch": 0.17964071856287425,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0008,
+      "loss": 1.3933,
+      "step": 120
+    },
+    {
+      "epoch": 0.18113772455089822,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.0008,
+      "loss": 1.4178,
+      "step": 121
+    },
+    {
+      "epoch": 0.18263473053892215,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0008,
+      "loss": 1.3884,
+      "step": 122
+    },
+    {
+      "epoch": 0.18413173652694612,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0008,
+      "loss": 1.3867,
+      "step": 123
+    },
+    {
+      "epoch": 0.18562874251497005,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0008,
+      "loss": 1.3795,
+      "step": 124
+    },
+    {
+      "epoch": 0.18712574850299402,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0008,
+      "loss": 1.3872,
+      "step": 125
+    },
+    {
+      "epoch": 0.18862275449101795,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0008,
+      "loss": 1.3764,
+      "step": 126
+    },
+    {
+      "epoch": 0.19011976047904192,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0008,
+      "loss": 1.3773,
+      "step": 127
+    },
+    {
+      "epoch": 0.19161676646706588,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0008,
+      "loss": 1.377,
+      "step": 128
+    },
+    {
+      "epoch": 0.19311377245508982,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0008,
+      "loss": 1.3746,
+      "step": 129
+    },
+    {
+      "epoch": 0.19461077844311378,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0008,
+      "loss": 1.3691,
+      "step": 130
+    },
+    {
+      "epoch": 0.19610778443113772,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 0.0008,
+      "loss": 1.373,
+      "step": 131
+    },
+    {
+      "epoch": 0.19760479041916168,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0008,
+      "loss": 1.3726,
+      "step": 132
+    },
+    {
+      "epoch": 0.19910179640718562,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0008,
+      "loss": 1.3653,
+      "step": 133
+    },
+    {
+      "epoch": 0.20059880239520958,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0008,
+      "loss": 1.3572,
+      "step": 134
+    },
+    {
+      "epoch": 0.20209580838323354,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0008,
+      "loss": 1.3518,
+      "step": 135
+    },
+    {
+      "epoch": 0.20359281437125748,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0008,
+      "loss": 1.3573,
+      "step": 136
+    },
+    {
+      "epoch": 0.20508982035928144,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0008,
+      "loss": 1.3523,
+      "step": 137
+    },
+    {
+      "epoch": 0.20658682634730538,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.3393,
+      "step": 138
+    },
+    {
+      "epoch": 0.20808383233532934,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0008,
+      "loss": 1.3482,
+      "step": 139
+    },
+    {
+      "epoch": 0.20958083832335328,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.3397,
+      "step": 140
+    },
+    {
+      "epoch": 0.21107784431137724,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.3446,
+      "step": 141
+    },
+    {
+      "epoch": 0.2125748502994012,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0008,
+      "loss": 1.3432,
+      "step": 142
+    },
+    {
+      "epoch": 0.21407185628742514,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0008,
+      "loss": 1.3407,
+      "step": 143
+    },
+    {
+      "epoch": 0.2155688622754491,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0008,
+      "loss": 1.334,
+      "step": 144
+    },
+    {
+      "epoch": 0.21706586826347304,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0008,
+      "loss": 1.3358,
+      "step": 145
+    },
+    {
+      "epoch": 0.218562874251497,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0008,
+      "loss": 1.3459,
+      "step": 146
+    },
+    {
+      "epoch": 0.22005988023952097,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0008,
+      "loss": 1.3315,
+      "step": 147
+    },
+    {
+      "epoch": 0.2215568862275449,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0008,
+      "loss": 1.3374,
+      "step": 148
+    },
+    {
+      "epoch": 0.22305389221556887,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0008,
+      "loss": 1.3344,
+      "step": 149
+    },
+    {
+      "epoch": 0.2245508982035928,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0008,
+      "loss": 1.345,
+      "step": 150
+    },
+    {
+      "epoch": 0.22604790419161677,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0008,
+      "loss": 1.3301,
+      "step": 151
+    },
+    {
+      "epoch": 0.2275449101796407,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0008,
+      "loss": 1.331,
+      "step": 152
+    },
+    {
+      "epoch": 0.22904191616766467,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.322,
+      "step": 153
+    },
+    {
+      "epoch": 0.23053892215568864,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0008,
+      "loss": 1.3297,
+      "step": 154
+    },
+    {
+      "epoch": 0.23203592814371257,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0008,
+      "loss": 1.3389,
+      "step": 155
+    },
+    {
+      "epoch": 0.23353293413173654,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0008,
+      "loss": 1.3189,
+      "step": 156
+    },
+    {
+      "epoch": 0.23502994011976047,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0008,
+      "loss": 1.3325,
+      "step": 157
+    },
+    {
+      "epoch": 0.23652694610778444,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0008,
+      "loss": 1.3364,
+      "step": 158
+    },
+    {
+      "epoch": 0.23802395209580837,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0008,
+      "loss": 1.3401,
+      "step": 159
+    },
+    {
+      "epoch": 0.23952095808383234,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0008,
+      "loss": 1.3334,
+      "step": 160
+    },
+    {
+      "epoch": 0.2410179640718563,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0008,
+      "loss": 1.3468,
+      "step": 161
+    },
+    {
+      "epoch": 0.24251497005988024,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0008,
+      "loss": 1.407,
+      "step": 162
+    },
+    {
+      "epoch": 0.2440119760479042,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0008,
+      "loss": 1.3459,
+      "step": 163
+    },
+    {
+      "epoch": 0.24550898203592814,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.0008,
+      "loss": 1.3732,
+      "step": 164
+    },
+    {
+      "epoch": 0.2470059880239521,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0008,
+      "loss": 1.3869,
+      "step": 165
+    },
+    {
+      "epoch": 0.24850299401197604,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0008,
+      "loss": 1.3686,
+      "step": 166
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0008,
+      "loss": 1.3486,
+      "step": 167
+    },
+    {
+      "epoch": 0.25149700598802394,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0008,
+      "loss": 1.357,
+      "step": 168
+    },
+    {
+      "epoch": 0.25299401197604793,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0008,
+      "loss": 1.3466,
+      "step": 169
+    },
+    {
+      "epoch": 0.25449101796407186,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0008,
+      "loss": 1.3512,
+      "step": 170
+    },
+    {
+      "epoch": 0.2559880239520958,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0008,
+      "loss": 1.3607,
+      "step": 171
+    },
+    {
+      "epoch": 0.25748502994011974,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0008,
+      "loss": 1.3398,
+      "step": 172
+    },
+    {
+      "epoch": 0.25898203592814373,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.3372,
+      "step": 173
+    },
+    {
+      "epoch": 0.26047904191616766,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0008,
+      "loss": 1.3342,
+      "step": 174
+    },
+    {
+      "epoch": 0.2619760479041916,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0008,
+      "loss": 1.3536,
+      "step": 175
+    },
+    {
+      "epoch": 0.2634730538922156,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0008,
+      "loss": 1.3416,
+      "step": 176
+    },
+    {
+      "epoch": 0.26497005988023953,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0008,
+      "loss": 1.3383,
+      "step": 177
+    },
+    {
+      "epoch": 0.26646706586826346,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0008,
+      "loss": 1.3285,
+      "step": 178
+    },
+    {
+      "epoch": 0.2679640718562874,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0008,
+      "loss": 1.3252,
+      "step": 179
+    },
+    {
+      "epoch": 0.2694610778443114,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0008,
+      "loss": 1.3281,
+      "step": 180
+    },
+    {
+      "epoch": 0.27095808383233533,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0008,
+      "loss": 1.3373,
+      "step": 181
+    },
+    {
+      "epoch": 0.27245508982035926,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0008,
+      "loss": 1.3247,
+      "step": 182
+    },
+    {
+      "epoch": 0.27395209580838326,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0008,
+      "loss": 1.3321,
+      "step": 183
+    },
+    {
+      "epoch": 0.2754491017964072,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0008,
+      "loss": 1.3243,
+      "step": 184
+    },
+    {
+      "epoch": 0.27694610778443113,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0008,
+      "loss": 1.323,
+      "step": 185
+    },
+    {
+      "epoch": 0.27844311377245506,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.3149,
+      "step": 186
+    },
+    {
+      "epoch": 0.27994011976047906,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.3184,
+      "step": 187
+    },
+    {
+      "epoch": 0.281437125748503,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.3214,
+      "step": 188
+    },
+    {
+      "epoch": 0.28293413173652693,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0008,
+      "loss": 1.3092,
+      "step": 189
+    },
+    {
+      "epoch": 0.2844311377245509,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.3186,
+      "step": 190
+    },
+    {
+      "epoch": 0.28592814371257486,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.3194,
+      "step": 191
+    },
+    {
+      "epoch": 0.2874251497005988,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.3081,
+      "step": 192
+    },
+    {
+      "epoch": 0.28892215568862273,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.3161,
+      "step": 193
+    },
+    {
+      "epoch": 0.2904191616766467,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.3115,
+      "step": 194
+    },
+    {
+      "epoch": 0.29191616766467066,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.3152,
+      "step": 195
+    },
+    {
+      "epoch": 0.2934131736526946,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.3126,
+      "step": 196
+    },
+    {
+      "epoch": 0.2949101796407186,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.3046,
+      "step": 197
+    },
+    {
+      "epoch": 0.2964071856287425,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.3053,
+      "step": 198
+    },
+    {
+      "epoch": 0.29790419161676646,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.3166,
+      "step": 199
+    },
+    {
+      "epoch": 0.2994011976047904,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.3115,
+      "step": 200
+    },
+    {
+      "epoch": 0.3008982035928144,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.3097,
+      "step": 201
+    },
+    {
+      "epoch": 0.3023952095808383,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.3051,
+      "step": 202
+    },
+    {
+      "epoch": 0.30389221556886226,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.3034,
+      "step": 203
+    },
+    {
+      "epoch": 0.30538922155688625,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0008,
+      "loss": 1.305,
+      "step": 204
+    },
+    {
+      "epoch": 0.3068862275449102,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.2913,
+      "step": 205
+    },
+    {
+      "epoch": 0.3083832335329341,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0008,
+      "loss": 1.3031,
+      "step": 206
+    },
+    {
+      "epoch": 0.30988023952095806,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.3113,
+      "step": 207
+    },
+    {
+      "epoch": 0.31137724550898205,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.3012,
+      "step": 208
+    },
+    {
+      "epoch": 0.312874251497006,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.2968,
+      "step": 209
+    },
+    {
+      "epoch": 0.3143712574850299,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.2994,
+      "step": 210
+    },
+    {
+      "epoch": 0.3158682634730539,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.3021,
+      "step": 211
+    },
+    {
+      "epoch": 0.31736526946107785,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0008,
+      "loss": 1.3061,
+      "step": 212
+    },
+    {
+      "epoch": 0.3188622754491018,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0008,
+      "loss": 1.3072,
+      "step": 213
+    },
+    {
+      "epoch": 0.3203592814371258,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0008,
+      "loss": 1.3489,
+      "step": 214
+    },
+    {
+      "epoch": 0.3218562874251497,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0008,
+      "loss": 1.3112,
+      "step": 215
+    },
+    {
+      "epoch": 0.32335329341317365,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.0008,
+      "loss": 1.3582,
+      "step": 216
+    },
+    {
+      "epoch": 0.3248502994011976,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.0008,
+      "loss": 1.3626,
+      "step": 217
+    },
+    {
+      "epoch": 0.3263473053892216,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0008,
+      "loss": 1.3273,
+      "step": 218
+    },
+    {
+      "epoch": 0.3278443113772455,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0008,
+      "loss": 1.3443,
+      "step": 219
+    },
+    {
+      "epoch": 0.32934131736526945,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0008,
+      "loss": 1.3579,
+      "step": 220
+    },
+    {
+      "epoch": 0.33083832335329344,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0008,
+      "loss": 1.3535,
+      "step": 221
+    },
+    {
+      "epoch": 0.3323353293413174,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0008,
+      "loss": 1.3344,
+      "step": 222
+    },
+    {
+      "epoch": 0.3338323353293413,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0008,
+      "loss": 1.3299,
+      "step": 223
+    },
+    {
+      "epoch": 0.33532934131736525,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.3354,
+      "step": 224
+    },
+    {
+      "epoch": 0.33682634730538924,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0008,
+      "loss": 1.3313,
+      "step": 225
+    },
+    {
+      "epoch": 0.3383233532934132,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0008,
+      "loss": 1.3231,
+      "step": 226
+    },
+    {
+      "epoch": 0.3398203592814371,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0008,
+      "loss": 1.3302,
+      "step": 227
+    },
+    {
+      "epoch": 0.3413173652694611,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0008,
+      "loss": 1.3175,
+      "step": 228
+    },
+    {
+      "epoch": 0.34281437125748504,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0008,
+      "loss": 1.3205,
+      "step": 229
+    },
+    {
+      "epoch": 0.344311377245509,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0008,
+      "loss": 1.3124,
+      "step": 230
+    },
+    {
+      "epoch": 0.3458083832335329,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0008,
+      "loss": 1.3172,
+      "step": 231
+    },
+    {
+      "epoch": 0.3473053892215569,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.3148,
+      "step": 232
+    },
+    {
+      "epoch": 0.34880239520958084,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0008,
+      "loss": 1.3087,
+      "step": 233
+    },
+    {
+      "epoch": 0.3502994011976048,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.3094,
+      "step": 234
+    },
+    {
+      "epoch": 0.35179640718562877,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0008,
+      "loss": 1.3113,
+      "step": 235
+    },
+    {
+      "epoch": 0.3532934131736527,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.3054,
+      "step": 236
+    },
+    {
+      "epoch": 0.35479041916167664,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0008,
+      "loss": 1.3145,
+      "step": 237
+    },
+    {
+      "epoch": 0.3562874251497006,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0008,
+      "loss": 1.304,
+      "step": 238
+    },
+    {
+      "epoch": 0.35778443113772457,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.3063,
+      "step": 239
+    },
+    {
+      "epoch": 0.3592814371257485,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.3018,
+      "step": 240
+    },
+    {
+      "epoch": 0.36077844311377244,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.3051,
+      "step": 241
+    },
+    {
+      "epoch": 0.36227544910179643,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.2974,
+      "step": 242
+    },
+    {
+      "epoch": 0.36377245508982037,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.303,
+      "step": 243
+    },
+    {
+      "epoch": 0.3652694610778443,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.3004,
+      "step": 244
+    },
+    {
+      "epoch": 0.36676646706586824,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.2908,
+      "step": 245
+    },
+    {
+      "epoch": 0.36826347305389223,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.2944,
+      "step": 246
+    },
+    {
+      "epoch": 0.36976047904191617,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.3007,
+      "step": 247
+    },
+    {
+      "epoch": 0.3712574850299401,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.2976,
+      "step": 248
+    },
+    {
+      "epoch": 0.3727544910179641,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.293,
+      "step": 249
+    },
+    {
+      "epoch": 0.37425149700598803,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.2994,
+      "step": 250
+    },
+    {
+      "epoch": 0.37574850299401197,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.2936,
+      "step": 251
+    },
+    {
+      "epoch": 0.3772455089820359,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.2927,
+      "step": 252
+    },
+    {
+      "epoch": 0.3787425149700599,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.2853,
+      "step": 253
+    },
+    {
+      "epoch": 0.38023952095808383,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0008,
+      "loss": 1.285,
+      "step": 254
+    },
+    {
+      "epoch": 0.38173652694610777,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.2995,
+      "step": 255
+    },
+    {
+      "epoch": 0.38323353293413176,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0008,
+      "loss": 1.2804,
+      "step": 256
+    },
+    {
+      "epoch": 0.3847305389221557,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.2919,
+      "step": 257
+    },
+    {
+      "epoch": 0.38622754491017963,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.2924,
+      "step": 258
+    },
+    {
+      "epoch": 0.38772455089820357,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.2885,
+      "step": 259
+    },
+    {
+      "epoch": 0.38922155688622756,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.2904,
+      "step": 260
+    },
+    {
+      "epoch": 0.3907185628742515,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.2872,
+      "step": 261
+    },
+    {
+      "epoch": 0.39221556886227543,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.2919,
+      "step": 262
+    },
+    {
+      "epoch": 0.3937125748502994,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.2957,
+      "step": 263
+    },
+    {
+      "epoch": 0.39520958083832336,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.2855,
+      "step": 264
+    },
+    {
+      "epoch": 0.3967065868263473,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0008,
+      "loss": 1.2825,
+      "step": 265
+    },
+    {
+      "epoch": 0.39820359281437123,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.2848,
+      "step": 266
+    },
+    {
+      "epoch": 0.3997005988023952,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0008,
+      "loss": 1.2872,
+      "step": 267
+    },
+    {
+      "epoch": 0.40119760479041916,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0008,
+      "loss": 1.2917,
+      "step": 268
+    },
+    {
+      "epoch": 0.4026946107784431,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.2817,
+      "step": 269
+    },
+    {
+      "epoch": 0.4041916167664671,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0008,
+      "loss": 1.2888,
+      "step": 270
+    },
+    {
+      "epoch": 0.405688622754491,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0008,
+      "loss": 1.287,
+      "step": 271
+    },
+    {
+      "epoch": 0.40718562874251496,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.2871,
+      "step": 272
+    },
+    {
+      "epoch": 0.4086826347305389,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.2878,
+      "step": 273
+    },
+    {
+      "epoch": 0.4101796407185629,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.2845,
+      "step": 274
+    },
+    {
+      "epoch": 0.4116766467065868,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.274,
+      "step": 275
+    },
+    {
+      "epoch": 0.41317365269461076,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0008,
+      "loss": 1.2769,
+      "step": 276
+    },
+    {
+      "epoch": 0.41467065868263475,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.284,
+      "step": 277
+    },
+    {
+      "epoch": 0.4161676646706587,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.2755,
+      "step": 278
+    },
+    {
+      "epoch": 0.4176646706586826,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0008,
+      "loss": 1.278,
+      "step": 279
+    },
+    {
+      "epoch": 0.41916167664670656,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0008,
+      "loss": 1.2831,
+      "step": 280
+    },
+    {
+      "epoch": 0.42065868263473055,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.2768,
+      "step": 281
+    },
+    {
+      "epoch": 0.4221556886227545,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.2748,
+      "step": 282
+    },
+    {
+      "epoch": 0.4236526946107784,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.2761,
+      "step": 283
+    },
+    {
+      "epoch": 0.4251497005988024,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.2679,
+      "step": 284
+    },
+    {
+      "epoch": 0.42664670658682635,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.276,
+      "step": 285
+    },
+    {
+      "epoch": 0.4281437125748503,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.2812,
+      "step": 286
+    },
+    {
+      "epoch": 0.4296407185628742,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.2695,
+      "step": 287
+    },
+    {
+      "epoch": 0.4311377245508982,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.2785,
+      "step": 288
+    },
+    {
+      "epoch": 0.43263473053892215,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0008,
+      "loss": 1.2744,
+      "step": 289
+    },
+    {
+      "epoch": 0.4341317365269461,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0008,
+      "loss": 1.2805,
+      "step": 290
+    },
+    {
+      "epoch": 0.4356287425149701,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0008,
+      "loss": 1.2771,
+      "step": 291
+    },
+    {
+      "epoch": 0.437125748502994,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0008,
+      "loss": 1.2746,
+      "step": 292
+    },
+    {
+      "epoch": 0.43862275449101795,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0008,
+      "loss": 1.284,
+      "step": 293
+    },
+    {
+      "epoch": 0.44011976047904194,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.2814,
+      "step": 294
+    },
+    {
+      "epoch": 0.4416167664670659,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.2742,
+      "step": 295
+    },
+    {
+      "epoch": 0.4431137724550898,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.2815,
+      "step": 296
+    },
+    {
+      "epoch": 0.44461077844311375,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0008,
+      "loss": 1.3037,
+      "step": 297
+    },
+    {
+      "epoch": 0.44610778443113774,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.0008,
+      "loss": 1.3257,
+      "step": 298
+    },
+    {
+      "epoch": 0.4476047904191617,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.0008,
+      "loss": 1.3587,
+      "step": 299
+    },
+    {
+      "epoch": 0.4491017964071856,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0008,
+      "loss": 1.3257,
+      "step": 300
+    },
+    {
+      "epoch": 0.4505988023952096,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0008,
+      "loss": 1.3205,
+      "step": 301
+    },
+    {
+      "epoch": 0.45209580838323354,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0008,
+      "loss": 1.3186,
+      "step": 302
+    },
+    {
+      "epoch": 0.4535928143712575,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0008,
+      "loss": 1.3042,
+      "step": 303
+    },
+    {
+      "epoch": 0.4550898203592814,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0008,
+      "loss": 1.299,
+      "step": 304
+    },
+    {
+      "epoch": 0.4565868263473054,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0008,
+      "loss": 1.2962,
+      "step": 305
+    },
+    {
+      "epoch": 0.45808383233532934,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0008,
+      "loss": 1.2882,
+      "step": 306
+    },
+    {
+      "epoch": 0.4595808383233533,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0008,
+      "loss": 1.2876,
+      "step": 307
+    },
+    {
+      "epoch": 0.46107784431137727,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0008,
+      "loss": 1.2961,
+      "step": 308
+    },
+    {
+      "epoch": 0.4625748502994012,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0008,
+      "loss": 1.295,
+      "step": 309
+    },
+    {
+      "epoch": 0.46407185628742514,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0008,
+      "loss": 1.2868,
+      "step": 310
+    },
+    {
+      "epoch": 0.4655688622754491,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0008,
+      "loss": 1.2954,
+      "step": 311
+    },
+    {
+      "epoch": 0.46706586826347307,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0008,
+      "loss": 1.2825,
+      "step": 312
+    },
+    {
+      "epoch": 0.468562874251497,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0008,
+      "loss": 1.2827,
+      "step": 313
+    },
+    {
+      "epoch": 0.47005988023952094,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0008,
+      "loss": 1.2897,
+      "step": 314
+    },
+    {
+      "epoch": 0.47155688622754494,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0008,
+      "loss": 1.2906,
+      "step": 315
+    },
+    {
+      "epoch": 0.47305389221556887,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.2799,
+      "step": 316
+    },
+    {
+      "epoch": 0.4745508982035928,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.2748,
+      "step": 317
+    },
+    {
+      "epoch": 0.47604790419161674,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.2831,
+      "step": 318
+    },
+    {
+      "epoch": 0.47754491017964074,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0008,
+      "loss": 1.2808,
+      "step": 319
+    },
+    {
+      "epoch": 0.47904191616766467,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.2816,
+      "step": 320
+    },
+    {
+      "epoch": 0.4805389221556886,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.2835,
+      "step": 321
+    },
+    {
+      "epoch": 0.4820359281437126,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0008,
+      "loss": 1.2818,
+      "step": 322
+    },
+    {
+      "epoch": 0.48353293413173654,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0008,
+      "loss": 1.2763,
+      "step": 323
+    },
+    {
+      "epoch": 0.48502994011976047,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.2805,
+      "step": 324
+    },
+    {
+      "epoch": 0.4865269461077844,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.2805,
+      "step": 325
+    },
+    {
+      "epoch": 0.4880239520958084,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0008,
+      "loss": 1.276,
+      "step": 326
+    },
+    {
+      "epoch": 0.48952095808383234,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.2827,
+      "step": 327
+    },
+    {
+      "epoch": 0.49101796407185627,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.2658,
+      "step": 328
+    },
+    {
+      "epoch": 0.49251497005988026,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.2749,
+      "step": 329
+    },
+    {
+      "epoch": 0.4940119760479042,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.2715,
+      "step": 330
+    },
+    {
+      "epoch": 0.49550898203592814,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.2661,
+      "step": 331
+    },
+    {
+      "epoch": 0.49700598802395207,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.2714,
+      "step": 332
+    },
+    {
+      "epoch": 0.49850299401197606,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0008,
+      "loss": 1.2666,
+      "step": 333
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.2837,
+      "step": 334
+    },
+    {
+      "epoch": 0.5014970059880239,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.2692,
+      "step": 335
+    },
+    {
+      "epoch": 0.5029940119760479,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.268,
+      "step": 336
+    },
+    {
+      "epoch": 0.5044910179640718,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.2758,
+      "step": 337
+    },
+    {
+      "epoch": 0.5059880239520959,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.2652,
+      "step": 338
+    },
+    {
+      "epoch": 0.5074850299401198,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.2729,
+      "step": 339
+    },
+    {
+      "epoch": 0.5089820359281437,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.2636,
+      "step": 340
+    },
+    {
+      "epoch": 0.5104790419161677,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.2692,
+      "step": 341
+    },
+    {
+      "epoch": 0.5119760479041916,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.2694,
+      "step": 342
+    },
+    {
+      "epoch": 0.5134730538922155,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.2742,
+      "step": 343
+    },
+    {
+      "epoch": 0.5149700598802395,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.267,
+      "step": 344
+    },
+    {
+      "epoch": 0.5164670658682635,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.2677,
+      "step": 345
+    },
+    {
+      "epoch": 0.5179640718562875,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.2671,
+      "step": 346
+    },
+    {
+      "epoch": 0.5194610778443114,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.2737,
+      "step": 347
+    },
+    {
+      "epoch": 0.5209580838323353,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0008,
+      "loss": 1.2797,
+      "step": 348
+    },
+    {
+      "epoch": 0.5224550898203593,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.2638,
+      "step": 349
+    },
+    {
+      "epoch": 0.5239520958083832,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.2663,
+      "step": 350
+    },
+    {
+      "epoch": 0.5254491017964071,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.2715,
+      "step": 351
+    },
+    {
+      "epoch": 0.5269461077844312,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.2662,
+      "step": 352
+    },
+    {
+      "epoch": 0.5284431137724551,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0008,
+      "loss": 1.2748,
+      "step": 353
+    },
+    {
+      "epoch": 0.5299401197604791,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.2716,
+      "step": 354
+    },
+    {
+      "epoch": 0.531437125748503,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.2581,
+      "step": 355
+    },
+    {
+      "epoch": 0.5329341317365269,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.2679,
+      "step": 356
+    },
+    {
+      "epoch": 0.5344311377245509,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.2615,
+      "step": 357
+    },
+    {
+      "epoch": 0.5359281437125748,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0008,
+      "loss": 1.2642,
+      "step": 358
+    },
+    {
+      "epoch": 0.5374251497005988,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0008,
+      "loss": 1.267,
+      "step": 359
+    },
+    {
+      "epoch": 0.5389221556886228,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0008,
+      "loss": 1.2599,
+      "step": 360
+    },
+    {
+      "epoch": 0.5404191616766467,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0008,
+      "loss": 1.2684,
+      "step": 361
+    },
+    {
+      "epoch": 0.5419161676646707,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.2617,
+      "step": 362
+    },
+    {
+      "epoch": 0.5434131736526946,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0008,
+      "loss": 1.2601,
+      "step": 363
+    },
+    {
+      "epoch": 0.5449101796407185,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.2675,
+      "step": 364
+    },
+    {
+      "epoch": 0.5464071856287425,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0008,
+      "loss": 1.269,
+      "step": 365
+    },
+    {
+      "epoch": 0.5479041916167665,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.264,
+      "step": 366
+    },
+    {
+      "epoch": 0.5494011976047904,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.2626,
+      "step": 367
+    },
+    {
+      "epoch": 0.5508982035928144,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.2651,
+      "step": 368
+    },
+    {
+      "epoch": 0.5523952095808383,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.2659,
+      "step": 369
+    },
+    {
+      "epoch": 0.5538922155688623,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0008,
+      "loss": 1.2623,
+      "step": 370
+    },
+    {
+      "epoch": 0.5553892215568862,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0008,
+      "loss": 1.2614,
+      "step": 371
+    },
+    {
+      "epoch": 0.5568862275449101,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.2689,
+      "step": 372
+    },
+    {
+      "epoch": 0.5583832335329342,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.2632,
+      "step": 373
+    },
+    {
+      "epoch": 0.5598802395209581,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.2591,
+      "step": 374
+    },
+    {
+      "epoch": 0.561377245508982,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0008,
+      "loss": 1.2601,
+      "step": 375
+    },
+    {
+      "epoch": 0.562874251497006,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.2628,
+      "step": 376
+    },
+    {
+      "epoch": 0.5643712574850299,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.2577,
+      "step": 377
+    },
+    {
+      "epoch": 0.5658682634730539,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.2571,
+      "step": 378
+    },
+    {
+      "epoch": 0.5673652694610778,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.2581,
+      "step": 379
+    },
+    {
+      "epoch": 0.5688622754491018,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0008,
+      "loss": 1.2596,
+      "step": 380
+    },
+    {
+      "epoch": 0.5703592814371258,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.2629,
+      "step": 381
+    },
+    {
+      "epoch": 0.5718562874251497,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.2597,
+      "step": 382
+    },
+    {
+      "epoch": 0.5733532934131736,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.2626,
+      "step": 383
+    },
+    {
+      "epoch": 0.5748502994011976,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.2557,
+      "step": 384
+    },
+    {
+      "epoch": 0.5763473053892215,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0008,
+      "loss": 1.2703,
+      "step": 385
+    },
+    {
+      "epoch": 0.5778443113772455,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0008,
+      "loss": 1.2595,
+      "step": 386
+    },
+    {
+      "epoch": 0.5793413173652695,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.2534,
+      "step": 387
+    },
+    {
+      "epoch": 0.5808383233532934,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0008,
+      "loss": 1.2662,
+      "step": 388
+    },
+    {
+      "epoch": 0.5823353293413174,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0008,
+      "loss": 1.2494,
+      "step": 389
+    },
+    {
+      "epoch": 0.5838323353293413,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0008,
+      "loss": 1.2616,
+      "step": 390
+    },
+    {
+      "epoch": 0.5853293413173652,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0008,
+      "loss": 1.2516,
+      "step": 391
+    },
+    {
+      "epoch": 0.5868263473053892,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.2502,
+      "step": 392
+    },
+    {
+      "epoch": 0.5883233532934131,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0008,
+      "loss": 1.2538,
+      "step": 393
+    },
+    {
+      "epoch": 0.5898203592814372,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0008,
+      "loss": 1.2571,
+      "step": 394
+    },
+    {
+      "epoch": 0.5913173652694611,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0008,
+      "loss": 1.2501,
+      "step": 395
+    },
+    {
+      "epoch": 0.592814371257485,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.2572,
+      "step": 396
+    },
+    {
+      "epoch": 0.594311377245509,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0008,
+      "loss": 1.2573,
+      "step": 397
+    },
+    {
+      "epoch": 0.5958083832335329,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.2487,
+      "step": 398
+    },
+    {
+      "epoch": 0.5973053892215568,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.2464,
+      "step": 399
+    },
+    {
+      "epoch": 0.5988023952095808,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.251,
+      "step": 400
+    },
+    {
+      "epoch": 0.6002994011976048,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0008,
+      "loss": 1.2577,
+      "step": 401
+    },
+    {
+      "epoch": 0.6017964071856288,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.2532,
+      "step": 402
+    },
+    {
+      "epoch": 0.6032934131736527,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0008,
+      "loss": 1.2586,
+      "step": 403
+    },
+    {
+      "epoch": 0.6047904191616766,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0008,
+      "loss": 1.2574,
+      "step": 404
+    },
+    {
+      "epoch": 0.6062874251497006,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0008,
+      "loss": 1.2516,
+      "step": 405
+    },
+    {
+      "epoch": 0.6077844311377245,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0008,
+      "loss": 1.2707,
+      "step": 406
+    },
+    {
+      "epoch": 0.6092814371257484,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0008,
+      "loss": 1.3835,
+      "step": 407
+    },
+    {
+      "epoch": 0.6107784431137725,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.0008,
+      "loss": 1.3279,
+      "step": 408
+    },
+    {
+      "epoch": 0.6122754491017964,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.0008,
+      "loss": 1.4153,
+      "step": 409
+    },
+    {
+      "epoch": 0.6137724550898204,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0008,
+      "loss": 1.3757,
+      "step": 410
+    },
+    {
+      "epoch": 0.6152694610778443,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.0008,
+      "loss": 1.4161,
+      "step": 411
+    },
+    {
+      "epoch": 0.6167664670658682,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.0008,
+      "loss": 1.3939,
+      "step": 412
+    },
+    {
+      "epoch": 0.6182634730538922,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0008,
+      "loss": 1.3756,
+      "step": 413
+    },
+    {
+      "epoch": 0.6197604790419161,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0008,
+      "loss": 1.373,
+      "step": 414
+    },
+    {
+      "epoch": 0.6212574850299402,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.3487,
+      "step": 415
+    },
+    {
+      "epoch": 0.6227544910179641,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.3556,
+      "step": 416
+    },
+    {
+      "epoch": 0.624251497005988,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0008,
+      "loss": 1.3364,
+      "step": 417
+    },
+    {
+      "epoch": 0.625748502994012,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0008,
+      "loss": 1.3407,
+      "step": 418
+    },
+    {
+      "epoch": 0.6272455089820359,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0008,
+      "loss": 1.3387,
+      "step": 419
+    },
+    {
+      "epoch": 0.6287425149700598,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0008,
+      "loss": 1.3285,
+      "step": 420
+    },
+    {
+      "epoch": 0.6302395209580839,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0008,
+      "loss": 1.3225,
+      "step": 421
+    },
+    {
+      "epoch": 0.6317365269461078,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.3073,
+      "step": 422
+    },
+    {
+      "epoch": 0.6332335329341318,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0008,
+      "loss": 1.3177,
+      "step": 423
+    },
+    {
+      "epoch": 0.6347305389221557,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 0.0008,
+      "loss": 1.3074,
+      "step": 424
+    },
+    {
+      "epoch": 0.6362275449101796,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0008,
+      "loss": 1.2996,
+      "step": 425
+    },
+    {
+      "epoch": 0.6377245508982036,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0008,
+      "loss": 1.3035,
+      "step": 426
+    },
+    {
+      "epoch": 0.6392215568862275,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0008,
+      "loss": 1.3063,
+      "step": 427
+    },
+    {
+      "epoch": 0.6407185628742516,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0008,
+      "loss": 1.3045,
+      "step": 428
+    },
+    {
+      "epoch": 0.6422155688622755,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0008,
+      "loss": 1.2989,
+      "step": 429
+    },
+    {
+      "epoch": 0.6437125748502994,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0008,
+      "loss": 1.293,
+      "step": 430
+    },
+    {
+      "epoch": 0.6452095808383234,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0008,
+      "loss": 1.2885,
+      "step": 431
+    },
+    {
+      "epoch": 0.6467065868263473,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0008,
+      "loss": 1.2915,
+      "step": 432
+    },
+    {
+      "epoch": 0.6482035928143712,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0008,
+      "loss": 1.2892,
+      "step": 433
+    },
+    {
+      "epoch": 0.6497005988023952,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0008,
+      "loss": 1.2859,
+      "step": 434
+    },
+    {
+      "epoch": 0.6511976047904192,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0008,
+      "loss": 1.2829,
+      "step": 435
+    },
+    {
+      "epoch": 0.6526946107784432,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0008,
+      "loss": 1.2888,
+      "step": 436
+    },
+    {
+      "epoch": 0.6541916167664671,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.278,
+      "step": 437
+    },
+    {
+      "epoch": 0.655688622754491,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0008,
+      "loss": 1.2785,
+      "step": 438
+    },
+    {
+      "epoch": 0.657185628742515,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0008,
+      "loss": 1.2866,
+      "step": 439
+    },
+    {
+      "epoch": 0.6586826347305389,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.2802,
+      "step": 440
+    },
+    {
+      "epoch": 0.6601796407185628,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.2791,
+      "step": 441
+    },
+    {
+      "epoch": 0.6616766467065869,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.2747,
+      "step": 442
+    },
+    {
+      "epoch": 0.6631736526946108,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.2801,
+      "step": 443
+    },
+    {
+      "epoch": 0.6646706586826348,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0008,
+      "loss": 1.2771,
+      "step": 444
+    },
+    {
+      "epoch": 0.6661676646706587,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.2773,
+      "step": 445
+    },
+    {
+      "epoch": 0.6676646706586826,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.2719,
+      "step": 446
+    },
+    {
+      "epoch": 0.6691616766467066,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0008,
+      "loss": 1.2724,
+      "step": 447
+    },
+    {
+      "epoch": 0.6706586826347305,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.2643,
+      "step": 448
+    },
+    {
+      "epoch": 0.6721556886227545,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.2621,
+      "step": 449
+    },
+    {
+      "epoch": 0.6736526946107785,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0008,
+      "loss": 1.272,
+      "step": 450
+    },
+    {
+      "epoch": 0.6751497005988024,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.2579,
+      "step": 451
+    },
+    {
+      "epoch": 0.6766467065868264,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.2697,
+      "step": 452
+    },
+    {
+      "epoch": 0.6781437125748503,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0008,
+      "loss": 1.2567,
+      "step": 453
+    },
+    {
+      "epoch": 0.6796407185628742,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.2649,
+      "step": 454
+    },
+    {
+      "epoch": 0.6811377245508982,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.2707,
+      "step": 455
+    },
+    {
+      "epoch": 0.6826347305389222,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.2685,
+      "step": 456
+    },
+    {
+      "epoch": 0.6841317365269461,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.2598,
+      "step": 457
+    },
+    {
+      "epoch": 0.6856287425149701,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.2675,
+      "step": 458
+    },
+    {
+      "epoch": 0.687125748502994,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.2612,
+      "step": 459
+    },
+    {
+      "epoch": 0.688622754491018,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.2551,
+      "step": 460
+    },
+    {
+      "epoch": 0.6901197604790419,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.2577,
+      "step": 461
+    },
+    {
+      "epoch": 0.6916167664670658,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.2606,
+      "step": 462
+    },
+    {
+      "epoch": 0.6931137724550899,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.2602,
+      "step": 463
+    },
+    {
+      "epoch": 0.6946107784431138,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0008,
+      "loss": 1.2544,
+      "step": 464
+    },
+    {
+      "epoch": 0.6961077844311377,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.2547,
+      "step": 465
+    },
+    {
+      "epoch": 0.6976047904191617,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0008,
+      "loss": 1.2575,
+      "step": 466
+    },
+    {
+      "epoch": 0.6991017964071856,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0008,
+      "loss": 1.2726,
+      "step": 467
+    },
+    {
+      "epoch": 0.7005988023952096,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.0008,
+      "loss": 1.3157,
+      "step": 468
+    },
+    {
+      "epoch": 0.7020958083832335,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0008,
+      "loss": 1.2837,
+      "step": 469
+    },
+    {
+      "epoch": 0.7035928143712575,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0008,
+      "loss": 1.3059,
+      "step": 470
+    },
+    {
+      "epoch": 0.7050898203592815,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0008,
+      "loss": 1.3061,
+      "step": 471
+    },
+    {
+      "epoch": 0.7065868263473054,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0008,
+      "loss": 1.2931,
+      "step": 472
+    },
+    {
+      "epoch": 0.7080838323353293,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0008,
+      "loss": 1.2869,
+      "step": 473
+    },
+    {
+      "epoch": 0.7095808383233533,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0008,
+      "loss": 1.2761,
+      "step": 474
+    },
+    {
+      "epoch": 0.7110778443113772,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0008,
+      "loss": 1.2764,
+      "step": 475
+    },
+    {
+      "epoch": 0.7125748502994012,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0008,
+      "loss": 1.2857,
+      "step": 476
+    },
+    {
+      "epoch": 0.7140718562874252,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0008,
+      "loss": 1.2777,
+      "step": 477
+    },
+    {
+      "epoch": 0.7155688622754491,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.2747,
+      "step": 478
+    },
+    {
+      "epoch": 0.7170658682634731,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.2732,
+      "step": 479
+    },
+    {
+      "epoch": 0.718562874251497,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0008,
+      "loss": 1.2695,
+      "step": 480
+    },
+    {
+      "epoch": 0.7200598802395209,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0008,
+      "loss": 1.2706,
+      "step": 481
+    },
+    {
+      "epoch": 0.7215568862275449,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.0008,
+      "loss": 1.2783,
+      "step": 482
+    },
+    {
+      "epoch": 0.7230538922155688,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0008,
+      "loss": 1.2682,
+      "step": 483
+    },
+    {
+      "epoch": 0.7245508982035929,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0008,
+      "loss": 1.2582,
+      "step": 484
+    },
+    {
+      "epoch": 0.7260479041916168,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0008,
+      "loss": 1.275,
+      "step": 485
+    },
+    {
+      "epoch": 0.7275449101796407,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0008,
+      "loss": 1.2671,
+      "step": 486
+    },
+    {
+      "epoch": 0.7290419161676647,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.2655,
+      "step": 487
+    },
+    {
+      "epoch": 0.7305389221556886,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0008,
+      "loss": 1.2631,
+      "step": 488
+    },
+    {
+      "epoch": 0.7320359281437125,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0008,
+      "loss": 1.2566,
+      "step": 489
+    },
+    {
+      "epoch": 0.7335329341317365,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.2566,
+      "step": 490
+    },
+    {
+      "epoch": 0.7350299401197605,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.2554,
+      "step": 491
+    },
+    {
+      "epoch": 0.7365269461077845,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.2556,
+      "step": 492
+    },
+    {
+      "epoch": 0.7380239520958084,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0008,
+      "loss": 1.2574,
+      "step": 493
+    },
+    {
+      "epoch": 0.7395209580838323,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0008,
+      "loss": 1.257,
+      "step": 494
+    },
+    {
+      "epoch": 0.7410179640718563,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.2622,
+      "step": 495
+    },
+    {
+      "epoch": 0.7425149700598802,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.2635,
+      "step": 496
+    },
+    {
+      "epoch": 0.7440119760479041,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.2536,
+      "step": 497
+    },
+    {
+      "epoch": 0.7455089820359282,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.2551,
+      "step": 498
+    },
+    {
+      "epoch": 0.7470059880239521,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.2509,
+      "step": 499
+    },
+    {
+      "epoch": 0.7485029940119761,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.2617,
+      "step": 500
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.2542,
+      "step": 501
+    },
+    {
+      "epoch": 0.7514970059880239,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.254,
+      "step": 502
+    },
+    {
+      "epoch": 0.7529940119760479,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.2516,
+      "step": 503
+    },
+    {
+      "epoch": 0.7544910179640718,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.2534,
+      "step": 504
+    },
+    {
+      "epoch": 0.7559880239520959,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.2536,
+      "step": 505
+    },
+    {
+      "epoch": 0.7574850299401198,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.2457,
+      "step": 506
+    },
+    {
+      "epoch": 0.7589820359281437,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.2576,
+      "step": 507
+    },
+    {
+      "epoch": 0.7604790419161677,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.2521,
+      "step": 508
+    },
+    {
+      "epoch": 0.7619760479041916,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.2629,
+      "step": 509
+    },
+    {
+      "epoch": 0.7634730538922155,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.2444,
+      "step": 510
+    },
+    {
+      "epoch": 0.7649700598802395,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.2592,
+      "step": 511
+    },
+    {
+      "epoch": 0.7664670658682635,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.2509,
+      "step": 512
+    },
+    {
+      "epoch": 0.7679640718562875,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.2494,
+      "step": 513
+    },
+    {
+      "epoch": 0.7694610778443114,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.2442,
+      "step": 514
+    },
+    {
+      "epoch": 0.7709580838323353,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.2492,
+      "step": 515
+    },
+    {
+      "epoch": 0.7724550898203593,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.2487,
+      "step": 516
+    },
+    {
+      "epoch": 0.7739520958083832,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.2504,
+      "step": 517
+    },
+    {
+      "epoch": 0.7754491017964071,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0008,
+      "loss": 1.2428,
+      "step": 518
+    },
+    {
+      "epoch": 0.7769461077844312,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.2555,
+      "step": 519
+    },
+    {
+      "epoch": 0.7784431137724551,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.2431,
+      "step": 520
+    },
+    {
+      "epoch": 0.7799401197604791,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.2415,
+      "step": 521
+    },
+    {
+      "epoch": 0.781437125748503,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.2507,
+      "step": 522
+    },
+    {
+      "epoch": 0.7829341317365269,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.2531,
+      "step": 523
+    },
+    {
+      "epoch": 0.7844311377245509,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.2447,
+      "step": 524
+    },
+    {
+      "epoch": 0.7859281437125748,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0008,
+      "loss": 1.2391,
+      "step": 525
+    },
+    {
+      "epoch": 0.7874251497005988,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.2424,
+      "step": 526
+    },
+    {
+      "epoch": 0.7889221556886228,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.2469,
+      "step": 527
+    },
+    {
+      "epoch": 0.7904191616766467,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.2487,
+      "step": 528
+    },
+    {
+      "epoch": 0.7919161676646707,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0008,
+      "loss": 1.2405,
+      "step": 529
+    },
+    {
+      "epoch": 0.7934131736526946,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.2493,
+      "step": 530
+    },
+    {
+      "epoch": 0.7949101796407185,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0008,
+      "loss": 1.2407,
+      "step": 531
+    },
+    {
+      "epoch": 0.7964071856287425,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0008,
+      "loss": 1.2494,
+      "step": 532
+    },
+    {
+      "epoch": 0.7979041916167665,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.2468,
+      "step": 533
+    },
+    {
+      "epoch": 0.7994011976047904,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.2456,
+      "step": 534
+    },
+    {
+      "epoch": 0.8008982035928144,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.2483,
+      "step": 535
+    },
+    {
+      "epoch": 0.8023952095808383,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.2478,
+      "step": 536
+    },
+    {
+      "epoch": 0.8038922155688623,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.2463,
+      "step": 537
+    },
+    {
+      "epoch": 0.8053892215568862,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.2434,
+      "step": 538
+    },
+    {
+      "epoch": 0.8068862275449101,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.2472,
+      "step": 539
+    },
+    {
+      "epoch": 0.8083832335329342,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.2514,
+      "step": 540
+    },
+    {
+      "epoch": 0.8098802395209581,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.2483,
+      "step": 541
+    },
+    {
+      "epoch": 0.811377245508982,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.2396,
+      "step": 542
+    },
+    {
+      "epoch": 0.812874251497006,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.2467,
+      "step": 543
+    },
+    {
+      "epoch": 0.8143712574850299,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0008,
+      "loss": 1.2467,
+      "step": 544
+    },
+    {
+      "epoch": 0.8158682634730539,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0008,
+      "loss": 1.2465,
+      "step": 545
+    },
+    {
+      "epoch": 0.8173652694610778,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.2495,
+      "step": 546
+    },
+    {
+      "epoch": 0.8188622754491018,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0008,
+      "loss": 1.2516,
+      "step": 547
+    },
+    {
+      "epoch": 0.8203592814371258,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0008,
+      "loss": 1.2468,
+      "step": 548
+    },
+    {
+      "epoch": 0.8218562874251497,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0008,
+      "loss": 1.2523,
+      "step": 549
+    },
+    {
+      "epoch": 0.8233532934131736,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0008,
+      "loss": 1.2377,
+      "step": 550
+    },
+    {
+      "epoch": 0.8248502994011976,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0008,
+      "loss": 1.2487,
+      "step": 551
+    },
+    {
+      "epoch": 0.8263473053892215,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.2451,
+      "step": 552
+    },
+    {
+      "epoch": 0.8278443113772455,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0008,
+      "loss": 1.2477,
+      "step": 553
+    },
+    {
+      "epoch": 0.8293413173652695,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0008,
+      "loss": 1.2495,
+      "step": 554
+    },
+    {
+      "epoch": 0.8308383233532934,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0008,
+      "loss": 1.2438,
+      "step": 555
+    },
+    {
+      "epoch": 0.8323353293413174,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0008,
+      "loss": 1.2469,
+      "step": 556
+    },
+    {
+      "epoch": 0.8338323353293413,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.2444,
+      "step": 557
+    },
+    {
+      "epoch": 0.8353293413173652,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0008,
+      "loss": 1.2592,
+      "step": 558
+    },
+    {
+      "epoch": 0.8368263473053892,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.2508,
+      "step": 559
+    },
+    {
+      "epoch": 0.8383233532934131,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0008,
+      "loss": 1.2474,
+      "step": 560
+    },
+    {
+      "epoch": 0.8398203592814372,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0008,
+      "loss": 1.2452,
+      "step": 561
+    },
+    {
+      "epoch": 0.8413173652694611,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.2421,
+      "step": 562
+    },
+    {
+      "epoch": 0.842814371257485,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0008,
+      "loss": 1.2452,
+      "step": 563
+    },
+    {
+      "epoch": 0.844311377245509,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.2469,
+      "step": 564
+    },
+    {
+      "epoch": 0.8458083832335329,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.2491,
+      "step": 565
+    },
+    {
+      "epoch": 0.8473053892215568,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.2406,
+      "step": 566
+    },
+    {
+      "epoch": 0.8488023952095808,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.2382,
+      "step": 567
+    },
+    {
+      "epoch": 0.8502994011976048,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.2385,
+      "step": 568
+    },
+    {
+      "epoch": 0.8517964071856288,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0008,
+      "loss": 1.2477,
+      "step": 569
+    },
+    {
+      "epoch": 0.8532934131736527,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0008,
+      "loss": 1.2428,
+      "step": 570
+    },
+    {
+      "epoch": 0.8547904191616766,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.2465,
+      "step": 571
+    },
+    {
+      "epoch": 0.8562874251497006,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.2548,
+      "step": 572
+    },
+    {
+      "epoch": 0.8577844311377245,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.2468,
+      "step": 573
+    },
+    {
+      "epoch": 0.8592814371257484,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.2416,
+      "step": 574
+    },
+    {
+      "epoch": 0.8607784431137725,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.2437,
+      "step": 575
+    },
+    {
+      "epoch": 0.8622754491017964,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.2291,
+      "step": 576
+    },
+    {
+      "epoch": 0.8637724550898204,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.2391,
+      "step": 577
+    },
+    {
+      "epoch": 0.8652694610778443,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.2451,
+      "step": 578
+    },
+    {
+      "epoch": 0.8667664670658682,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.2325,
+      "step": 579
+    },
+    {
+      "epoch": 0.8682634730538922,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.2394,
+      "step": 580
+    },
+    {
+      "epoch": 0.8697604790419161,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.2332,
+      "step": 581
+    },
+    {
+      "epoch": 0.8712574850299402,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.239,
+      "step": 582
+    },
+    {
+      "epoch": 0.8727544910179641,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.2363,
+      "step": 583
+    },
+    {
+      "epoch": 0.874251497005988,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.2463,
+      "step": 584
+    },
+    {
+      "epoch": 0.875748502994012,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.2459,
+      "step": 585
+    },
+    {
+      "epoch": 0.8772455089820359,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.2403,
+      "step": 586
+    },
+    {
+      "epoch": 0.8787425149700598,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0008,
+      "loss": 1.249,
+      "step": 587
+    },
+    {
+      "epoch": 0.8802395209580839,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.2359,
+      "step": 588
+    },
+    {
+      "epoch": 0.8817365269461078,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0008,
+      "loss": 1.2448,
+      "step": 589
+    },
+    {
+      "epoch": 0.8832335329341318,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0008,
+      "loss": 1.2473,
+      "step": 590
+    },
+    {
+      "epoch": 0.8847305389221557,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0008,
+      "loss": 1.2437,
+      "step": 591
+    },
+    {
+      "epoch": 0.8862275449101796,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0008,
+      "loss": 1.243,
+      "step": 592
+    },
+    {
+      "epoch": 0.8877245508982036,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.2377,
+      "step": 593
+    },
+    {
+      "epoch": 0.8892215568862275,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0008,
+      "loss": 1.2365,
+      "step": 594
+    },
+    {
+      "epoch": 0.8907185628742516,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0008,
+      "loss": 1.2386,
+      "step": 595
+    },
+    {
+      "epoch": 0.8922155688622755,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0008,
+      "loss": 1.2347,
+      "step": 596
+    },
+    {
+      "epoch": 0.8937125748502994,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0008,
+      "loss": 1.2459,
+      "step": 597
+    },
+    {
+      "epoch": 0.8952095808383234,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0008,
+      "loss": 1.2417,
+      "step": 598
+    },
+    {
+      "epoch": 0.8967065868263473,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0008,
+      "loss": 1.2378,
+      "step": 599
+    },
+    {
+      "epoch": 0.8982035928143712,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0008,
+      "loss": 1.2402,
+      "step": 600
+    },
+    {
+      "epoch": 0.8997005988023952,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.2324,
+      "step": 601
+    },
+    {
+      "epoch": 0.9011976047904192,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.2395,
+      "step": 602
+    },
+    {
+      "epoch": 0.9026946107784432,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0008,
+      "loss": 1.2374,
+      "step": 603
+    },
+    {
+      "epoch": 0.9041916167664671,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0008,
+      "loss": 1.2418,
+      "step": 604
+    },
+    {
+      "epoch": 0.905688622754491,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0008,
+      "loss": 1.2431,
+      "step": 605
+    },
+    {
+      "epoch": 0.907185628742515,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.2416,
+      "step": 606
+    },
+    {
+      "epoch": 0.9086826347305389,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0008,
+      "loss": 1.2369,
+      "step": 607
+    },
+    {
+      "epoch": 0.9101796407185628,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0008,
+      "loss": 1.2399,
+      "step": 608
+    },
+    {
+      "epoch": 0.9116766467065869,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0008,
+      "loss": 1.2445,
+      "step": 609
+    },
+    {
+      "epoch": 0.9131736526946108,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0008,
+      "loss": 1.2343,
+      "step": 610
+    },
+    {
+      "epoch": 0.9146706586826348,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0008,
+      "loss": 1.2317,
+      "step": 611
+    },
+    {
+      "epoch": 0.9161676646706587,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0008,
+      "loss": 1.2381,
+      "step": 612
+    },
+    {
+      "epoch": 0.9176646706586826,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0008,
+      "loss": 1.2429,
+      "step": 613
+    },
+    {
+      "epoch": 0.9191616766467066,
+      "grad_norm": 1.625,
+      "learning_rate": 0.0008,
+      "loss": 1.3109,
+      "step": 614
+    },
+    {
+      "epoch": 0.9206586826347305,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.0008,
+      "loss": 1.2647,
+      "step": 615
+    },
+    {
+      "epoch": 0.9221556886227545,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0008,
+      "loss": 1.2848,
+      "step": 616
+    },
+    {
+      "epoch": 0.9236526946107785,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0008,
+      "loss": 1.2868,
+      "step": 617
+    },
+    {
+      "epoch": 0.9251497005988024,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0008,
+      "loss": 1.2641,
+      "step": 618
+    },
+    {
+      "epoch": 0.9266467065868264,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.261,
+      "step": 619
+    },
+    {
+      "epoch": 0.9281437125748503,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.2685,
+      "step": 620
+    },
+    {
+      "epoch": 0.9296407185628742,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 0.0008,
+      "loss": 1.2581,
+      "step": 621
+    },
+    {
+      "epoch": 0.9311377245508982,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0008,
+      "loss": 1.2602,
+      "step": 622
+    },
+    {
+      "epoch": 0.9326347305389222,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0008,
+      "loss": 1.2504,
+      "step": 623
+    },
+    {
+      "epoch": 0.9341317365269461,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0008,
+      "loss": 1.2536,
+      "step": 624
+    },
+    {
+      "epoch": 0.9356287425149701,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0008,
+      "loss": 1.2571,
+      "step": 625
+    },
+    {
+      "epoch": 0.937125748502994,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0008,
+      "loss": 1.2459,
+      "step": 626
+    },
+    {
+      "epoch": 0.938622754491018,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0008,
+      "loss": 1.2506,
+      "step": 627
+    },
+    {
+      "epoch": 0.9401197604790419,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0008,
+      "loss": 1.2524,
+      "step": 628
+    },
+    {
+      "epoch": 0.9416167664670658,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0008,
+      "loss": 1.2505,
+      "step": 629
+    },
+    {
+      "epoch": 0.9431137724550899,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0008,
+      "loss": 1.2487,
+      "step": 630
+    },
+    {
+      "epoch": 0.9446107784431138,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0008,
+      "loss": 1.2488,
+      "step": 631
+    },
+    {
+      "epoch": 0.9461077844311377,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0008,
+      "loss": 1.2459,
+      "step": 632
+    },
+    {
+      "epoch": 0.9476047904191617,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.2413,
+      "step": 633
+    },
+    {
+      "epoch": 0.9491017964071856,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.2407,
+      "step": 634
+    },
+    {
+      "epoch": 0.9505988023952096,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.2363,
+      "step": 635
+    },
+    {
+      "epoch": 0.9520958083832335,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0008,
+      "loss": 1.2352,
+      "step": 636
+    },
+    {
+      "epoch": 0.9535928143712575,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.2365,
+      "step": 637
+    },
+    {
+      "epoch": 0.9550898203592815,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.2447,
+      "step": 638
+    },
+    {
+      "epoch": 0.9565868263473054,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.245,
+      "step": 639
+    },
+    {
+      "epoch": 0.9580838323353293,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.2441,
+      "step": 640
+    },
+    {
+      "epoch": 0.9595808383233533,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.2406,
+      "step": 641
+    },
+    {
+      "epoch": 0.9610778443113772,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.2385,
+      "step": 642
+    },
+    {
+      "epoch": 0.9625748502994012,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.2329,
+      "step": 643
+    },
+    {
+      "epoch": 0.9640718562874252,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.2412,
+      "step": 644
+    },
+    {
+      "epoch": 0.9655688622754491,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.2366,
+      "step": 645
+    },
+    {
+      "epoch": 0.9670658682634731,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0008,
+      "loss": 1.2435,
+      "step": 646
+    },
+    {
+      "epoch": 0.968562874251497,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.2438,
+      "step": 647
+    },
+    {
+      "epoch": 0.9700598802395209,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.2343,
+      "step": 648
+    },
+    {
+      "epoch": 0.9715568862275449,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0008,
+      "loss": 1.2347,
+      "step": 649
+    },
+    {
+      "epoch": 0.9730538922155688,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.2401,
+      "step": 650
+    },
+    {
+      "epoch": 0.9745508982035929,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.235,
+      "step": 651
+    },
+    {
+      "epoch": 0.9760479041916168,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.2344,
+      "step": 652
+    },
+    {
+      "epoch": 0.9775449101796407,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.2315,
+      "step": 653
+    },
+    {
+      "epoch": 0.9790419161676647,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.2327,
+      "step": 654
+    },
+    {
+      "epoch": 0.9805389221556886,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.236,
+      "step": 655
+    },
+    {
+      "epoch": 0.9820359281437125,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.2296,
+      "step": 656
+    },
+    {
+      "epoch": 0.9835329341317365,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.2338,
+      "step": 657
+    },
+    {
+      "epoch": 0.9850299401197605,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.2435,
+      "step": 658
+    },
+    {
+      "epoch": 0.9865269461077845,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0008,
+      "loss": 1.2294,
+      "step": 659
+    },
+    {
+      "epoch": 0.9880239520958084,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.232,
+      "step": 660
+    },
+    {
+      "epoch": 0.9895209580838323,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.2238,
+      "step": 661
+    },
+    {
+      "epoch": 0.9910179640718563,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.2328,
+      "step": 662
+    },
+    {
+      "epoch": 0.9925149700598802,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.2334,
+      "step": 663
+    },
+    {
+      "epoch": 0.9940119760479041,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.2371,
+      "step": 664
+    },
+    {
+      "epoch": 0.9955089820359282,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.2422,
+      "step": 665
+    },
+    {
+      "epoch": 0.9970059880239521,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.2314,
+      "step": 666
+    },
+    {
+      "epoch": 0.9985029940119761,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.2306,
+      "step": 667
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.2275,
+      "step": 668
+    },
+    {
+      "epoch": 1.001497005988024,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.2372,
+      "step": 669
+    },
+    {
+      "epoch": 1.0029940119760479,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.2366,
+      "step": 670
+    },
+    {
+      "epoch": 1.0044910179640718,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.2293,
+      "step": 671
+    },
+    {
+      "epoch": 1.0059880239520957,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.2365,
+      "step": 672
+    },
+    {
+      "epoch": 1.0074850299401197,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.2288,
+      "step": 673
+    },
+    {
+      "epoch": 1.0089820359281436,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.2309,
+      "step": 674
+    },
+    {
+      "epoch": 1.0104790419161676,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.2281,
+      "step": 675
+    },
+    {
+      "epoch": 1.0119760479041917,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.2286,
+      "step": 676
+    },
+    {
+      "epoch": 1.0134730538922156,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.2302,
+      "step": 677
+    },
+    {
+      "epoch": 1.0149700598802396,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.2238,
+      "step": 678
+    },
+    {
+      "epoch": 1.0164670658682635,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.2237,
+      "step": 679
+    },
+    {
+      "epoch": 1.0179640718562875,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.222,
+      "step": 680
+    },
+    {
+      "epoch": 1.0194610778443114,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.2323,
+      "step": 681
+    },
+    {
+      "epoch": 1.0209580838323353,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.2246,
+      "step": 682
+    },
+    {
+      "epoch": 1.0224550898203593,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.2249,
+      "step": 683
+    },
+    {
+      "epoch": 1.0239520958083832,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.2348,
+      "step": 684
+    },
+    {
+      "epoch": 1.0254491017964071,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.2309,
+      "step": 685
+    },
+    {
+      "epoch": 1.026946107784431,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.2295,
+      "step": 686
+    },
+    {
+      "epoch": 1.028443113772455,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.2291,
+      "step": 687
+    },
+    {
+      "epoch": 1.029940119760479,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0008,
+      "loss": 1.2245,
+      "step": 688
+    },
+    {
+      "epoch": 1.031437125748503,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.2247,
+      "step": 689
+    },
+    {
+      "epoch": 1.032934131736527,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.2322,
+      "step": 690
+    },
+    {
+      "epoch": 1.034431137724551,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.2273,
+      "step": 691
+    },
+    {
+      "epoch": 1.035928143712575,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.223,
+      "step": 692
+    },
+    {
+      "epoch": 1.0374251497005988,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.2276,
+      "step": 693
+    },
+    {
+      "epoch": 1.0389221556886228,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.221,
+      "step": 694
+    },
+    {
+      "epoch": 1.0404191616766467,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.2212,
+      "step": 695
+    },
+    {
+      "epoch": 1.0419161676646707,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.222,
+      "step": 696
+    },
+    {
+      "epoch": 1.0434131736526946,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.2294,
+      "step": 697
+    },
+    {
+      "epoch": 1.0449101796407185,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.2231,
+      "step": 698
+    },
+    {
+      "epoch": 1.0464071856287425,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.2274,
+      "step": 699
+    },
+    {
+      "epoch": 1.0479041916167664,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.2232,
+      "step": 700
+    },
+    {
+      "epoch": 1.0494011976047903,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.2189,
+      "step": 701
+    },
+    {
+      "epoch": 1.0508982035928143,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.2336,
+      "step": 702
+    },
+    {
+      "epoch": 1.0523952095808382,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.23,
+      "step": 703
+    },
+    {
+      "epoch": 1.0538922155688624,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0008,
+      "loss": 1.2317,
+      "step": 704
+    },
+    {
+      "epoch": 1.0553892215568863,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.2261,
+      "step": 705
+    },
+    {
+      "epoch": 1.0568862275449102,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0008,
+      "loss": 1.2241,
+      "step": 706
+    },
+    {
+      "epoch": 1.0583832335329342,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0008,
+      "loss": 1.2223,
+      "step": 707
+    },
+    {
+      "epoch": 1.0598802395209581,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0008,
+      "loss": 1.2178,
+      "step": 708
+    },
+    {
+      "epoch": 1.061377245508982,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.2348,
+      "step": 709
+    },
+    {
+      "epoch": 1.062874251497006,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0008,
+      "loss": 1.2284,
+      "step": 710
+    },
+    {
+      "epoch": 1.06437125748503,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0008,
+      "loss": 1.2272,
+      "step": 711
+    },
+    {
+      "epoch": 1.0658682634730539,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0008,
+      "loss": 1.2345,
+      "step": 712
+    },
+    {
+      "epoch": 1.0673652694610778,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0008,
+      "loss": 1.2282,
+      "step": 713
+    },
+    {
+      "epoch": 1.0688622754491017,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0008,
+      "loss": 1.2278,
+      "step": 714
+    },
+    {
+      "epoch": 1.0703592814371257,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0008,
+      "loss": 1.2277,
+      "step": 715
+    },
+    {
+      "epoch": 1.0718562874251496,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0008,
+      "loss": 1.2264,
+      "step": 716
+    },
+    {
+      "epoch": 1.0733532934131738,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0008,
+      "loss": 1.2238,
+      "step": 717
+    },
+    {
+      "epoch": 1.0748502994011977,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0008,
+      "loss": 1.218,
+      "step": 718
+    },
+    {
+      "epoch": 1.0763473053892216,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.0008,
+      "loss": 1.2397,
+      "step": 719
+    },
+    {
+      "epoch": 1.0778443113772456,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.0008,
+      "loss": 1.3052,
+      "step": 720
+    },
+    {
+      "epoch": 1.0793413173652695,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.2408,
+      "step": 721
+    },
+    {
+      "epoch": 1.0808383233532934,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0008,
+      "loss": 1.2891,
+      "step": 722
+    },
+    {
+      "epoch": 1.0823353293413174,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0008,
+      "loss": 1.2595,
+      "step": 723
+    },
+    {
+      "epoch": 1.0838323353293413,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0008,
+      "loss": 1.2605,
+      "step": 724
+    },
+    {
+      "epoch": 1.0853293413173652,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0008,
+      "loss": 1.2537,
+      "step": 725
+    },
+    {
+      "epoch": 1.0868263473053892,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0008,
+      "loss": 1.2526,
+      "step": 726
+    },
+    {
+      "epoch": 1.0883233532934131,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.2418,
+      "step": 727
+    },
+    {
+      "epoch": 1.089820359281437,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.2439,
+      "step": 728
+    },
+    {
+      "epoch": 1.091317365269461,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0008,
+      "loss": 1.2389,
+      "step": 729
+    },
+    {
+      "epoch": 1.092814371257485,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0008,
+      "loss": 1.2388,
+      "step": 730
+    },
+    {
+      "epoch": 1.0943113772455089,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0008,
+      "loss": 1.2394,
+      "step": 731
+    },
+    {
+      "epoch": 1.095808383233533,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.2443,
+      "step": 732
+    },
+    {
+      "epoch": 1.097305389221557,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0008,
+      "loss": 1.2308,
+      "step": 733
+    },
+    {
+      "epoch": 1.098802395209581,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0008,
+      "loss": 1.23,
+      "step": 734
+    },
+    {
+      "epoch": 1.1002994011976048,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.2342,
+      "step": 735
+    },
+    {
+      "epoch": 1.1017964071856288,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.2372,
+      "step": 736
+    },
+    {
+      "epoch": 1.1032934131736527,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0008,
+      "loss": 1.2339,
+      "step": 737
+    },
+    {
+      "epoch": 1.1047904191616766,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0008,
+      "loss": 1.2252,
+      "step": 738
+    },
+    {
+      "epoch": 1.1062874251497006,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0008,
+      "loss": 1.2298,
+      "step": 739
+    },
+    {
+      "epoch": 1.1077844311377245,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.2372,
+      "step": 740
+    },
+    {
+      "epoch": 1.1092814371257484,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0008,
+      "loss": 1.2223,
+      "step": 741
+    },
+    {
+      "epoch": 1.1107784431137724,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.2322,
+      "step": 742
+    },
+    {
+      "epoch": 1.1122754491017963,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.2261,
+      "step": 743
+    },
+    {
+      "epoch": 1.1137724550898203,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.2329,
+      "step": 744
+    },
+    {
+      "epoch": 1.1152694610778444,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.2268,
+      "step": 745
+    },
+    {
+      "epoch": 1.1167664670658684,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.2296,
+      "step": 746
+    },
+    {
+      "epoch": 1.1182634730538923,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.2252,
+      "step": 747
+    },
+    {
+      "epoch": 1.1197604790419162,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.2342,
+      "step": 748
+    },
+    {
+      "epoch": 1.1212574850299402,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.2307,
+      "step": 749
+    },
+    {
+      "epoch": 1.122754491017964,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0008,
+      "loss": 1.224,
+      "step": 750
+    },
+    {
+      "epoch": 1.124251497005988,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.2255,
+      "step": 751
+    },
+    {
+      "epoch": 1.125748502994012,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.2305,
+      "step": 752
+    },
+    {
+      "epoch": 1.127245508982036,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.228,
+      "step": 753
+    },
+    {
+      "epoch": 1.1287425149700598,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.2216,
+      "step": 754
+    },
+    {
+      "epoch": 1.1302395209580838,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.2284,
+      "step": 755
+    },
+    {
+      "epoch": 1.1317365269461077,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.2271,
+      "step": 756
+    },
+    {
+      "epoch": 1.1332335329341316,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.2271,
+      "step": 757
+    },
+    {
+      "epoch": 1.1347305389221556,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.2278,
+      "step": 758
+    },
+    {
+      "epoch": 1.1362275449101795,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.2244,
+      "step": 759
+    },
+    {
+      "epoch": 1.1377245508982037,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.2322,
+      "step": 760
+    },
+    {
+      "epoch": 1.1392215568862276,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.2234,
+      "step": 761
+    },
+    {
+      "epoch": 1.1407185628742516,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.2225,
+      "step": 762
+    },
+    {
+      "epoch": 1.1422155688622755,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.2199,
+      "step": 763
+    },
+    {
+      "epoch": 1.1437125748502994,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.224,
+      "step": 764
+    },
+    {
+      "epoch": 1.1452095808383234,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.2217,
+      "step": 765
+    },
+    {
+      "epoch": 1.1467065868263473,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.2292,
+      "step": 766
+    },
+    {
+      "epoch": 1.1482035928143712,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.2188,
+      "step": 767
+    },
+    {
+      "epoch": 1.1497005988023952,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.2298,
+      "step": 768
+    },
+    {
+      "epoch": 1.151197604790419,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.2207,
+      "step": 769
+    },
+    {
+      "epoch": 1.152694610778443,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.2295,
+      "step": 770
+    },
+    {
+      "epoch": 1.154191616766467,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.2203,
+      "step": 771
+    },
+    {
+      "epoch": 1.1556886227544911,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.2198,
+      "step": 772
+    },
+    {
+      "epoch": 1.157185628742515,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.2228,
+      "step": 773
+    },
+    {
+      "epoch": 1.158682634730539,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.2141,
+      "step": 774
+    },
+    {
+      "epoch": 1.160179640718563,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.22,
+      "step": 775
+    },
+    {
+      "epoch": 1.1616766467065869,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.2225,
+      "step": 776
+    },
+    {
+      "epoch": 1.1631736526946108,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.2187,
+      "step": 777
+    },
+    {
+      "epoch": 1.1646706586826348,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.2214,
+      "step": 778
+    },
+    {
+      "epoch": 1.1661676646706587,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.2203,
+      "step": 779
+    },
+    {
+      "epoch": 1.1676646706586826,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.2223,
+      "step": 780
+    },
+    {
+      "epoch": 1.1691616766467066,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.219,
+      "step": 781
+    },
+    {
+      "epoch": 1.1706586826347305,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.2217,
+      "step": 782
+    },
+    {
+      "epoch": 1.1721556886227544,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.2189,
+      "step": 783
+    },
+    {
+      "epoch": 1.1736526946107784,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.2181,
+      "step": 784
+    },
+    {
+      "epoch": 1.1751497005988023,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.2276,
+      "step": 785
+    },
+    {
+      "epoch": 1.1766467065868262,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.2178,
+      "step": 786
+    },
+    {
+      "epoch": 1.1781437125748502,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.2254,
+      "step": 787
+    },
+    {
+      "epoch": 1.1796407185628743,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.2179,
+      "step": 788
+    },
+    {
+      "epoch": 1.1811377245508983,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.219,
+      "step": 789
+    },
+    {
+      "epoch": 1.1826347305389222,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.2253,
+      "step": 790
+    },
+    {
+      "epoch": 1.1841317365269461,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.2195,
+      "step": 791
+    },
+    {
+      "epoch": 1.18562874251497,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.2154,
+      "step": 792
+    },
+    {
+      "epoch": 1.187125748502994,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.2129,
+      "step": 793
+    },
+    {
+      "epoch": 1.188622754491018,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.2196,
+      "step": 794
+    },
+    {
+      "epoch": 1.1901197604790419,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.2216,
+      "step": 795
+    },
+    {
+      "epoch": 1.1916167664670658,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.2218,
+      "step": 796
+    },
+    {
+      "epoch": 1.1931137724550898,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.2223,
+      "step": 797
+    },
+    {
+      "epoch": 1.1946107784431137,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.2176,
+      "step": 798
+    },
+    {
+      "epoch": 1.1961077844311376,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.2178,
+      "step": 799
+    },
+    {
+      "epoch": 1.1976047904191618,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.2196,
+      "step": 800
+    },
+    {
+      "epoch": 1.1991017964071857,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.222,
+      "step": 801
+    },
+    {
+      "epoch": 1.2005988023952097,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.2207,
+      "step": 802
+    },
+    {
+      "epoch": 1.2020958083832336,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.2174,
+      "step": 803
+    },
+    {
+      "epoch": 1.2035928143712575,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.2221,
+      "step": 804
+    },
+    {
+      "epoch": 1.2050898203592815,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.2173,
+      "step": 805
+    },
+    {
+      "epoch": 1.2065868263473054,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.2134,
+      "step": 806
+    },
+    {
+      "epoch": 1.2080838323353293,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.2161,
+      "step": 807
+    },
+    {
+      "epoch": 1.2095808383233533,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0008,
+      "loss": 1.2128,
+      "step": 808
+    },
+    {
+      "epoch": 1.2110778443113772,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0008,
+      "loss": 1.2226,
+      "step": 809
+    },
+    {
+      "epoch": 1.2125748502994012,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.0008,
+      "loss": 1.2246,
+      "step": 810
+    },
+    {
+      "epoch": 1.214071856287425,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.0008,
+      "loss": 1.3315,
+      "step": 811
+    },
+    {
+      "epoch": 1.215568862275449,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0008,
+      "loss": 1.2585,
+      "step": 812
+    },
+    {
+      "epoch": 1.217065868263473,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.0008,
+      "loss": 1.2749,
+      "step": 813
+    },
+    {
+      "epoch": 1.218562874251497,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.0008,
+      "loss": 1.2944,
+      "step": 814
+    },
+    {
+      "epoch": 1.220059880239521,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0008,
+      "loss": 1.2678,
+      "step": 815
+    },
+    {
+      "epoch": 1.221556886227545,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0008,
+      "loss": 1.2712,
+      "step": 816
+    },
+    {
+      "epoch": 1.223053892215569,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0008,
+      "loss": 1.2548,
+      "step": 817
+    },
+    {
+      "epoch": 1.2245508982035929,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0008,
+      "loss": 1.2598,
+      "step": 818
+    },
+    {
+      "epoch": 1.2260479041916168,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0008,
+      "loss": 1.2494,
+      "step": 819
+    },
+    {
+      "epoch": 1.2275449101796407,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0008,
+      "loss": 1.258,
+      "step": 820
+    },
+    {
+      "epoch": 1.2290419161676647,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0008,
+      "loss": 1.2467,
+      "step": 821
+    },
+    {
+      "epoch": 1.2305389221556886,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0008,
+      "loss": 1.238,
+      "step": 822
+    },
+    {
+      "epoch": 1.2320359281437125,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.2271,
+      "step": 823
+    },
+    {
+      "epoch": 1.2335329341317365,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0008,
+      "loss": 1.2371,
+      "step": 824
+    },
+    {
+      "epoch": 1.2350299401197604,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0008,
+      "loss": 1.2355,
+      "step": 825
+    },
+    {
+      "epoch": 1.2365269461077844,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.2355,
+      "step": 826
+    },
+    {
+      "epoch": 1.2380239520958083,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.2356,
+      "step": 827
+    },
+    {
+      "epoch": 1.2395209580838324,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.2343,
+      "step": 828
+    },
+    {
+      "epoch": 1.2410179640718564,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.2363,
+      "step": 829
+    },
+    {
+      "epoch": 1.2425149700598803,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0008,
+      "loss": 1.2363,
+      "step": 830
+    },
+    {
+      "epoch": 1.2440119760479043,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0008,
+      "loss": 1.22,
+      "step": 831
+    },
+    {
+      "epoch": 1.2455089820359282,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.2313,
+      "step": 832
+    },
+    {
+      "epoch": 1.2470059880239521,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.2283,
+      "step": 833
+    },
+    {
+      "epoch": 1.248502994011976,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.232,
+      "step": 834
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.2278,
+      "step": 835
+    },
+    {
+      "epoch": 1.251497005988024,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.2336,
+      "step": 836
+    },
+    {
+      "epoch": 1.2529940119760479,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.218,
+      "step": 837
+    },
+    {
+      "epoch": 1.2544910179640718,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.2327,
+      "step": 838
+    },
+    {
+      "epoch": 1.2559880239520957,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.2223,
+      "step": 839
+    },
+    {
+      "epoch": 1.2574850299401197,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.2284,
+      "step": 840
+    },
+    {
+      "epoch": 1.2589820359281436,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.2221,
+      "step": 841
+    },
+    {
+      "epoch": 1.2604790419161676,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.2212,
+      "step": 842
+    },
+    {
+      "epoch": 1.2619760479041915,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.2208,
+      "step": 843
+    },
+    {
+      "epoch": 1.2634730538922156,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.2243,
+      "step": 844
+    },
+    {
+      "epoch": 1.2649700598802396,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.2203,
+      "step": 845
+    },
+    {
+      "epoch": 1.2664670658682635,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.2163,
+      "step": 846
+    },
+    {
+      "epoch": 1.2679640718562875,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.2256,
+      "step": 847
+    },
+    {
+      "epoch": 1.2694610778443114,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.2152,
+      "step": 848
+    },
+    {
+      "epoch": 1.2709580838323353,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.2169,
+      "step": 849
+    },
+    {
+      "epoch": 1.2724550898203593,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.2325,
+      "step": 850
+    },
+    {
+      "epoch": 1.2739520958083832,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.2178,
+      "step": 851
+    },
+    {
+      "epoch": 1.2754491017964071,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.2087,
+      "step": 852
+    },
+    {
+      "epoch": 1.276946107784431,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.2163,
+      "step": 853
+    },
+    {
+      "epoch": 1.278443113772455,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.2198,
+      "step": 854
+    },
+    {
+      "epoch": 1.2799401197604792,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.2201,
+      "step": 855
+    },
+    {
+      "epoch": 1.281437125748503,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.2171,
+      "step": 856
+    },
+    {
+      "epoch": 1.282934131736527,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0008,
+      "loss": 1.213,
+      "step": 857
+    },
+    {
+      "epoch": 1.284431137724551,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.2188,
+      "step": 858
+    },
+    {
+      "epoch": 1.285928143712575,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.2155,
+      "step": 859
+    },
+    {
+      "epoch": 1.2874251497005988,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.2163,
+      "step": 860
+    },
+    {
+      "epoch": 1.2889221556886228,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.2242,
+      "step": 861
+    },
+    {
+      "epoch": 1.2904191616766467,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.2176,
+      "step": 862
+    },
+    {
+      "epoch": 1.2919161676646707,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.2127,
+      "step": 863
+    },
+    {
+      "epoch": 1.2934131736526946,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.2145,
+      "step": 864
+    },
+    {
+      "epoch": 1.2949101796407185,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.2258,
+      "step": 865
+    },
+    {
+      "epoch": 1.2964071856287425,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.219,
+      "step": 866
+    },
+    {
+      "epoch": 1.2979041916167664,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.2148,
+      "step": 867
+    },
+    {
+      "epoch": 1.2994011976047903,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.215,
+      "step": 868
+    },
+    {
+      "epoch": 1.3008982035928143,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.207,
+      "step": 869
+    },
+    {
+      "epoch": 1.3023952095808382,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.2146,
+      "step": 870
+    },
+    {
+      "epoch": 1.3038922155688621,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0008,
+      "loss": 1.2173,
+      "step": 871
+    },
+    {
+      "epoch": 1.3053892215568863,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.2191,
+      "step": 872
+    },
+    {
+      "epoch": 1.3068862275449102,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.2156,
+      "step": 873
+    },
+    {
+      "epoch": 1.3083832335329342,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.2124,
+      "step": 874
+    },
+    {
+      "epoch": 1.3098802395209581,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.2191,
+      "step": 875
+    },
+    {
+      "epoch": 1.311377245508982,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.2193,
+      "step": 876
+    },
+    {
+      "epoch": 1.312874251497006,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.2146,
+      "step": 877
+    },
+    {
+      "epoch": 1.31437125748503,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.2121,
+      "step": 878
+    },
+    {
+      "epoch": 1.3158682634730539,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.2162,
+      "step": 879
+    },
+    {
+      "epoch": 1.3173652694610778,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.2154,
+      "step": 880
+    },
+    {
+      "epoch": 1.3188622754491017,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.2199,
+      "step": 881
+    },
+    {
+      "epoch": 1.3203592814371259,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.2234,
+      "step": 882
+    },
+    {
+      "epoch": 1.3218562874251498,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.222,
+      "step": 883
+    },
+    {
+      "epoch": 1.3233532934131738,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.2121,
+      "step": 884
+    },
+    {
+      "epoch": 1.3248502994011977,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.2114,
+      "step": 885
+    },
+    {
+      "epoch": 1.3263473053892216,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.2124,
+      "step": 886
+    },
+    {
+      "epoch": 1.3278443113772456,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.2143,
+      "step": 887
+    },
+    {
+      "epoch": 1.3293413173652695,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.2189,
+      "step": 888
+    },
+    {
+      "epoch": 1.3308383233532934,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.2093,
+      "step": 889
+    },
+    {
+      "epoch": 1.3323353293413174,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.2117,
+      "step": 890
+    },
+    {
+      "epoch": 1.3338323353293413,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.2157,
+      "step": 891
+    },
+    {
+      "epoch": 1.3353293413173652,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.2195,
+      "step": 892
+    },
+    {
+      "epoch": 1.3368263473053892,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1979,
+      "step": 893
+    },
+    {
+      "epoch": 1.3383233532934131,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.215,
+      "step": 894
+    },
+    {
+      "epoch": 1.339820359281437,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.2084,
+      "step": 895
+    },
+    {
+      "epoch": 1.341317365269461,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.2044,
+      "step": 896
+    },
+    {
+      "epoch": 1.342814371257485,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.2095,
+      "step": 897
+    },
+    {
+      "epoch": 1.3443113772455089,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.2104,
+      "step": 898
+    },
+    {
+      "epoch": 1.3458083832335328,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.2132,
+      "step": 899
+    },
+    {
+      "epoch": 1.347305389221557,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.2128,
+      "step": 900
+    },
+    {
+      "epoch": 1.348802395209581,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.2178,
+      "step": 901
+    },
+    {
+      "epoch": 1.3502994011976048,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.2073,
+      "step": 902
+    },
+    {
+      "epoch": 1.3517964071856288,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.2028,
+      "step": 903
+    },
+    {
+      "epoch": 1.3532934131736527,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.208,
+      "step": 904
+    },
+    {
+      "epoch": 1.3547904191616766,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.2178,
+      "step": 905
+    },
+    {
+      "epoch": 1.3562874251497006,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.2105,
+      "step": 906
+    },
+    {
+      "epoch": 1.3577844311377245,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.213,
+      "step": 907
+    },
+    {
+      "epoch": 1.3592814371257484,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.2116,
+      "step": 908
+    },
+    {
+      "epoch": 1.3607784431137724,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.2187,
+      "step": 909
+    },
+    {
+      "epoch": 1.3622754491017965,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0008,
+      "loss": 1.219,
+      "step": 910
+    },
+    {
+      "epoch": 1.3637724550898205,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.0008,
+      "loss": 1.2323,
+      "step": 911
+    },
+    {
+      "epoch": 1.3652694610778444,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0008,
+      "loss": 1.2959,
+      "step": 912
+    },
+    {
+      "epoch": 1.3667664670658684,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0008,
+      "loss": 1.2367,
+      "step": 913
+    },
+    {
+      "epoch": 1.3682634730538923,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0008,
+      "loss": 1.2636,
+      "step": 914
+    },
+    {
+      "epoch": 1.3697604790419162,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0008,
+      "loss": 1.2372,
+      "step": 915
+    },
+    {
+      "epoch": 1.3712574850299402,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0008,
+      "loss": 1.2443,
+      "step": 916
+    },
+    {
+      "epoch": 1.372754491017964,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0008,
+      "loss": 1.2336,
+      "step": 917
+    },
+    {
+      "epoch": 1.374251497005988,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0008,
+      "loss": 1.2485,
+      "step": 918
+    },
+    {
+      "epoch": 1.375748502994012,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0008,
+      "loss": 1.2288,
+      "step": 919
+    },
+    {
+      "epoch": 1.377245508982036,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0008,
+      "loss": 1.2385,
+      "step": 920
+    },
+    {
+      "epoch": 1.3787425149700598,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0008,
+      "loss": 1.2306,
+      "step": 921
+    },
+    {
+      "epoch": 1.3802395209580838,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0008,
+      "loss": 1.2302,
+      "step": 922
+    },
+    {
+      "epoch": 1.3817365269461077,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0008,
+      "loss": 1.2309,
+      "step": 923
+    },
+    {
+      "epoch": 1.3832335329341316,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0008,
+      "loss": 1.2244,
+      "step": 924
+    },
+    {
+      "epoch": 1.3847305389221556,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.2262,
+      "step": 925
+    },
+    {
+      "epoch": 1.3862275449101795,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.208,
+      "step": 926
+    },
+    {
+      "epoch": 1.3877245508982035,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0008,
+      "loss": 1.2209,
+      "step": 927
+    },
+    {
+      "epoch": 1.3892215568862276,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.2219,
+      "step": 928
+    },
+    {
+      "epoch": 1.3907185628742516,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.216,
+      "step": 929
+    },
+    {
+      "epoch": 1.3922155688622755,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.2252,
+      "step": 930
+    },
+    {
+      "epoch": 1.3937125748502994,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.2221,
+      "step": 931
+    },
+    {
+      "epoch": 1.3952095808383234,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.2244,
+      "step": 932
+    },
+    {
+      "epoch": 1.3967065868263473,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.2251,
+      "step": 933
+    },
+    {
+      "epoch": 1.3982035928143712,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.217,
+      "step": 934
+    },
+    {
+      "epoch": 1.3997005988023952,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.2183,
+      "step": 935
+    },
+    {
+      "epoch": 1.401197604790419,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.213,
+      "step": 936
+    },
+    {
+      "epoch": 1.402694610778443,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.2141,
+      "step": 937
+    },
+    {
+      "epoch": 1.4041916167664672,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.2129,
+      "step": 938
+    },
+    {
+      "epoch": 1.4056886227544911,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.2105,
+      "step": 939
+    },
+    {
+      "epoch": 1.407185628742515,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.2148,
+      "step": 940
+    },
+    {
+      "epoch": 1.408682634730539,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.2099,
+      "step": 941
+    },
+    {
+      "epoch": 1.410179640718563,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.2095,
+      "step": 942
+    },
+    {
+      "epoch": 1.4116766467065869,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.2154,
+      "step": 943
+    },
+    {
+      "epoch": 1.4131736526946108,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0008,
+      "loss": 1.217,
+      "step": 944
+    },
+    {
+      "epoch": 1.4146706586826348,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.2108,
+      "step": 945
+    },
+    {
+      "epoch": 1.4161676646706587,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0008,
+      "loss": 1.2119,
+      "step": 946
+    },
+    {
+      "epoch": 1.4176646706586826,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.2122,
+      "step": 947
+    },
+    {
+      "epoch": 1.4191616766467066,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.2096,
+      "step": 948
+    },
+    {
+      "epoch": 1.4206586826347305,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.2052,
+      "step": 949
+    },
+    {
+      "epoch": 1.4221556886227544,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0008,
+      "loss": 1.2113,
+      "step": 950
+    },
+    {
+      "epoch": 1.4236526946107784,
+      "grad_norm": 0.052734375,
+      "learning_rate": 0.0008,
+      "loss": 1.2125,
+      "step": 951
+    },
+    {
+      "epoch": 1.4251497005988023,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.2084,
+      "step": 952
+    },
+    {
+      "epoch": 1.4266467065868262,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.2112,
+      "step": 953
+    },
+    {
+      "epoch": 1.4281437125748502,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.2088,
+      "step": 954
+    },
+    {
+      "epoch": 1.4296407185628741,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.2017,
+      "step": 955
+    },
+    {
+      "epoch": 1.4311377245508983,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.2095,
+      "step": 956
+    },
+    {
+      "epoch": 1.4326347305389222,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0008,
+      "loss": 1.2201,
+      "step": 957
+    },
+    {
+      "epoch": 1.4341317365269461,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.2119,
+      "step": 958
+    },
+    {
+      "epoch": 1.43562874251497,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.2162,
+      "step": 959
+    },
+    {
+      "epoch": 1.437125748502994,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.2035,
+      "step": 960
+    },
+    {
+      "epoch": 1.438622754491018,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.2084,
+      "step": 961
+    },
+    {
+      "epoch": 1.4401197604790419,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.2118,
+      "step": 962
+    },
+    {
+      "epoch": 1.4416167664670658,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0008,
+      "loss": 1.2041,
+      "step": 963
+    },
+    {
+      "epoch": 1.4431137724550898,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.21,
+      "step": 964
+    },
+    {
+      "epoch": 1.4446107784431137,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0008,
+      "loss": 1.2105,
+      "step": 965
+    },
+    {
+      "epoch": 1.4461077844311379,
+      "grad_norm": 0.046875,
+      "learning_rate": 0.0008,
+      "loss": 1.2103,
+      "step": 966
+    },
+    {
+      "epoch": 1.4476047904191618,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0008,
+      "loss": 1.2033,
+      "step": 967
+    },
+    {
+      "epoch": 1.4491017964071857,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0008,
+      "loss": 1.2092,
+      "step": 968
+    },
+    {
+      "epoch": 1.4505988023952097,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0008,
+      "loss": 1.2062,
+      "step": 969
+    },
+    {
+      "epoch": 1.4520958083832336,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.2027,
+      "step": 970
+    },
+    {
+      "epoch": 1.4535928143712575,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.2035,
+      "step": 971
+    },
+    {
+      "epoch": 1.4550898203592815,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.2088,
+      "step": 972
+    },
+    {
+      "epoch": 1.4565868263473054,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.2028,
+      "step": 973
+    },
+    {
+      "epoch": 1.4580838323353293,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.2031,
+      "step": 974
+    },
+    {
+      "epoch": 1.4595808383233533,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.2023,
+      "step": 975
+    },
+    {
+      "epoch": 1.4610778443113772,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.205,
+      "step": 976
+    },
+    {
+      "epoch": 1.4625748502994012,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.2101,
+      "step": 977
+    },
+    {
+      "epoch": 1.464071856287425,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.2114,
+      "step": 978
+    },
+    {
+      "epoch": 1.465568862275449,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1998,
+      "step": 979
+    },
+    {
+      "epoch": 1.467065868263473,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.2041,
+      "step": 980
+    },
+    {
+      "epoch": 1.468562874251497,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0008,
+      "loss": 1.2183,
+      "step": 981
+    },
+    {
+      "epoch": 1.4700598802395208,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.2202,
+      "step": 982
+    },
+    {
+      "epoch": 1.471556886227545,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.2125,
+      "step": 983
+    },
+    {
+      "epoch": 1.473053892215569,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.2027,
+      "step": 984
+    },
+    {
+      "epoch": 1.4745508982035929,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.2034,
+      "step": 985
+    },
+    {
+      "epoch": 1.4760479041916168,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.2009,
+      "step": 986
+    },
+    {
+      "epoch": 1.4775449101796407,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.2107,
+      "step": 987
+    },
+    {
+      "epoch": 1.4790419161676647,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.2079,
+      "step": 988
+    },
+    {
+      "epoch": 1.4805389221556886,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.2113,
+      "step": 989
+    },
+    {
+      "epoch": 1.4820359281437125,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.2029,
+      "step": 990
+    },
+    {
+      "epoch": 1.4835329341317365,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.2102,
+      "step": 991
+    },
+    {
+      "epoch": 1.4850299401197604,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.2034,
+      "step": 992
+    },
+    {
+      "epoch": 1.4865269461077844,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.2087,
+      "step": 993
+    },
+    {
+      "epoch": 1.4880239520958085,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.2089,
+      "step": 994
+    },
+    {
+      "epoch": 1.4895209580838324,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.2062,
+      "step": 995
+    },
+    {
+      "epoch": 1.4910179640718564,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.2015,
+      "step": 996
+    },
+    {
+      "epoch": 1.4925149700598803,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.2039,
+      "step": 997
+    },
+    {
+      "epoch": 1.4940119760479043,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.2031,
+      "step": 998
+    },
+    {
+      "epoch": 1.4955089820359282,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.2136,
+      "step": 999
+    },
+    {
+      "epoch": 1.4970059880239521,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.2024,
+      "step": 1000
+    },
+    {
+      "epoch": 1.498502994011976,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0008,
+      "loss": 1.2109,
+      "step": 1001
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.2178,
+      "step": 1002
+    },
+    {
+      "epoch": 1.501497005988024,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.2105,
+      "step": 1003
+    },
+    {
+      "epoch": 1.5029940119760479,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.2054,
+      "step": 1004
+    },
+    {
+      "epoch": 1.5044910179640718,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.2028,
+      "step": 1005
+    },
+    {
+      "epoch": 1.5059880239520957,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.2002,
+      "step": 1006
+    },
+    {
+      "epoch": 1.5074850299401197,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.2092,
+      "step": 1007
+    },
+    {
+      "epoch": 1.5089820359281436,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.2041,
+      "step": 1008
+    },
+    {
+      "epoch": 1.5104790419161676,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.2079,
+      "step": 1009
+    },
+    {
+      "epoch": 1.5119760479041915,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1992,
+      "step": 1010
+    },
+    {
+      "epoch": 1.5134730538922154,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.2069,
+      "step": 1011
+    },
+    {
+      "epoch": 1.5149700598802394,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0008,
+      "loss": 1.2013,
+      "step": 1012
+    },
+    {
+      "epoch": 1.5164670658682635,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.2034,
+      "step": 1013
+    },
+    {
+      "epoch": 1.5179640718562875,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.2027,
+      "step": 1014
+    },
+    {
+      "epoch": 1.5194610778443114,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.2092,
+      "step": 1015
+    },
+    {
+      "epoch": 1.5209580838323353,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.2086,
+      "step": 1016
+    },
+    {
+      "epoch": 1.5224550898203593,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.2071,
+      "step": 1017
+    },
+    {
+      "epoch": 1.5239520958083832,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.2035,
+      "step": 1018
+    },
+    {
+      "epoch": 1.5254491017964071,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.203,
+      "step": 1019
+    },
+    {
+      "epoch": 1.5269461077844313,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.2158,
+      "step": 1020
+    },
+    {
+      "epoch": 1.5284431137724552,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.2044,
+      "step": 1021
+    },
+    {
+      "epoch": 1.5299401197604792,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.2039,
+      "step": 1022
+    },
+    {
+      "epoch": 1.531437125748503,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1994,
+      "step": 1023
+    },
+    {
+      "epoch": 1.532934131736527,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.2034,
+      "step": 1024
+    },
+    {
+      "epoch": 1.534431137724551,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0008,
+      "loss": 1.2066,
+      "step": 1025
+    },
+    {
+      "epoch": 1.535928143712575,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.2026,
+      "step": 1026
+    },
+    {
+      "epoch": 1.5374251497005988,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.2077,
+      "step": 1027
+    },
+    {
+      "epoch": 1.5389221556886228,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.2121,
+      "step": 1028
+    },
+    {
+      "epoch": 1.5404191616766467,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.2023,
+      "step": 1029
+    },
+    {
+      "epoch": 1.5419161676646707,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.2043,
+      "step": 1030
+    },
+    {
+      "epoch": 1.5434131736526946,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.2075,
+      "step": 1031
+    },
+    {
+      "epoch": 1.5449101796407185,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.2064,
+      "step": 1032
+    },
+    {
+      "epoch": 1.5464071856287425,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.2031,
+      "step": 1033
+    },
+    {
+      "epoch": 1.5479041916167664,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.2031,
+      "step": 1034
+    },
+    {
+      "epoch": 1.5494011976047903,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.2081,
+      "step": 1035
+    },
+    {
+      "epoch": 1.5508982035928143,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.2075,
+      "step": 1036
+    },
+    {
+      "epoch": 1.5523952095808382,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.2064,
+      "step": 1037
+    },
+    {
+      "epoch": 1.5538922155688621,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.2106,
+      "step": 1038
+    },
+    {
+      "epoch": 1.555389221556886,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.2055,
+      "step": 1039
+    },
+    {
+      "epoch": 1.55688622754491,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.2047,
+      "step": 1040
+    },
+    {
+      "epoch": 1.5583832335329342,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.2051,
+      "step": 1041
+    },
+    {
+      "epoch": 1.5598802395209581,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.2072,
+      "step": 1042
+    },
+    {
+      "epoch": 1.561377245508982,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.2055,
+      "step": 1043
+    },
+    {
+      "epoch": 1.562874251497006,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.2028,
+      "step": 1044
+    },
+    {
+      "epoch": 1.56437125748503,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.204,
+      "step": 1045
+    },
+    {
+      "epoch": 1.5658682634730539,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.2117,
+      "step": 1046
+    },
+    {
+      "epoch": 1.5673652694610778,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.2005,
+      "step": 1047
+    },
+    {
+      "epoch": 1.568862275449102,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.2032,
+      "step": 1048
+    },
+    {
+      "epoch": 1.5703592814371259,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.2109,
+      "step": 1049
+    },
+    {
+      "epoch": 1.5718562874251498,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.2036,
+      "step": 1050
+    },
+    {
+      "epoch": 1.5733532934131738,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1996,
+      "step": 1051
+    },
+    {
+      "epoch": 1.5748502994011977,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.2076,
+      "step": 1052
+    },
+    {
+      "epoch": 1.5763473053892216,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.2064,
+      "step": 1053
+    },
+    {
+      "epoch": 1.5778443113772456,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.2028,
+      "step": 1054
+    },
+    {
+      "epoch": 1.5793413173652695,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.2081,
+      "step": 1055
+    },
+    {
+      "epoch": 1.5808383233532934,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1996,
+      "step": 1056
+    },
+    {
+      "epoch": 1.5823353293413174,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.2042,
+      "step": 1057
+    },
+    {
+      "epoch": 1.5838323353293413,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1969,
+      "step": 1058
+    },
+    {
+      "epoch": 1.5853293413173652,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.2051,
+      "step": 1059
+    },
+    {
+      "epoch": 1.5868263473053892,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.2038,
+      "step": 1060
+    },
+    {
+      "epoch": 1.5883233532934131,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.2081,
+      "step": 1061
+    },
+    {
+      "epoch": 1.589820359281437,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.2055,
+      "step": 1062
+    },
+    {
+      "epoch": 1.591317365269461,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.202,
+      "step": 1063
+    },
+    {
+      "epoch": 1.592814371257485,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.2008,
+      "step": 1064
+    },
+    {
+      "epoch": 1.5943113772455089,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.2024,
+      "step": 1065
+    },
+    {
+      "epoch": 1.5958083832335328,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.2078,
+      "step": 1066
+    },
+    {
+      "epoch": 1.5973053892215567,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.2053,
+      "step": 1067
+    },
+    {
+      "epoch": 1.5988023952095807,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.2041,
+      "step": 1068
+    },
+    {
+      "epoch": 1.6002994011976048,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.2056,
+      "step": 1069
+    },
+    {
+      "epoch": 1.6017964071856288,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.2009,
+      "step": 1070
+    },
+    {
+      "epoch": 1.6032934131736527,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.211,
+      "step": 1071
+    },
+    {
+      "epoch": 1.6047904191616766,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.2129,
+      "step": 1072
+    },
+    {
+      "epoch": 1.6062874251497006,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.2025,
+      "step": 1073
+    },
+    {
+      "epoch": 1.6077844311377245,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.2019,
+      "step": 1074
+    },
+    {
+      "epoch": 1.6092814371257484,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.2059,
+      "step": 1075
+    },
+    {
+      "epoch": 1.6107784431137726,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.2036,
+      "step": 1076
+    },
+    {
+      "epoch": 1.6122754491017965,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.2065,
+      "step": 1077
+    },
+    {
+      "epoch": 1.6137724550898205,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.2114,
+      "step": 1078
+    },
+    {
+      "epoch": 1.6152694610778444,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.2037,
+      "step": 1079
+    },
+    {
+      "epoch": 1.6167664670658684,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.2015,
+      "step": 1080
+    },
+    {
+      "epoch": 1.6182634730538923,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.199,
+      "step": 1081
+    },
+    {
+      "epoch": 1.6197604790419162,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.2123,
+      "step": 1082
+    },
+    {
+      "epoch": 1.6212574850299402,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.2005,
+      "step": 1083
+    },
+    {
+      "epoch": 1.622754491017964,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1993,
+      "step": 1084
+    },
+    {
+      "epoch": 1.624251497005988,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.204,
+      "step": 1085
+    },
+    {
+      "epoch": 1.625748502994012,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.2042,
+      "step": 1086
+    },
+    {
+      "epoch": 1.627245508982036,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1993,
+      "step": 1087
+    },
+    {
+      "epoch": 1.6287425149700598,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.2052,
+      "step": 1088
+    },
+    {
+      "epoch": 1.6302395209580838,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.2036,
+      "step": 1089
+    },
+    {
+      "epoch": 1.6317365269461077,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1921,
+      "step": 1090
+    },
+    {
+      "epoch": 1.6332335329341316,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.2045,
+      "step": 1091
+    },
+    {
+      "epoch": 1.6347305389221556,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0008,
+      "loss": 1.2008,
+      "step": 1092
+    },
+    {
+      "epoch": 1.6362275449101795,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.204,
+      "step": 1093
+    },
+    {
+      "epoch": 1.6377245508982035,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1952,
+      "step": 1094
+    },
+    {
+      "epoch": 1.6392215568862274,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.201,
+      "step": 1095
+    },
+    {
+      "epoch": 1.6407185628742516,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1998,
+      "step": 1096
+    },
+    {
+      "epoch": 1.6422155688622755,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.2058,
+      "step": 1097
+    },
+    {
+      "epoch": 1.6437125748502994,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.207,
+      "step": 1098
+    },
+    {
+      "epoch": 1.6452095808383234,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.2026,
+      "step": 1099
+    },
+    {
+      "epoch": 1.6467065868263473,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1981,
+      "step": 1100
+    },
+    {
+      "epoch": 1.6482035928143712,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.2022,
+      "step": 1101
+    },
+    {
+      "epoch": 1.6497005988023952,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.2003,
+      "step": 1102
+    },
+    {
+      "epoch": 1.6511976047904193,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1976,
+      "step": 1103
+    },
+    {
+      "epoch": 1.6526946107784433,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.2066,
+      "step": 1104
+    },
+    {
+      "epoch": 1.6541916167664672,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.2024,
+      "step": 1105
+    },
+    {
+      "epoch": 1.6556886227544911,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.208,
+      "step": 1106
+    },
+    {
+      "epoch": 1.657185628742515,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1998,
+      "step": 1107
+    },
+    {
+      "epoch": 1.658682634730539,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.2023,
+      "step": 1108
+    },
+    {
+      "epoch": 1.660179640718563,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.2033,
+      "step": 1109
+    },
+    {
+      "epoch": 1.6616766467065869,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1999,
+      "step": 1110
+    },
+    {
+      "epoch": 1.6631736526946108,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0008,
+      "loss": 1.2061,
+      "step": 1111
+    },
+    {
+      "epoch": 1.6646706586826348,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.2035,
+      "step": 1112
+    },
+    {
+      "epoch": 1.6661676646706587,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1901,
+      "step": 1113
+    },
+    {
+      "epoch": 1.6676646706586826,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1957,
+      "step": 1114
+    },
+    {
+      "epoch": 1.6691616766467066,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1969,
+      "step": 1115
+    },
+    {
+      "epoch": 1.6706586826347305,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.2045,
+      "step": 1116
+    },
+    {
+      "epoch": 1.6721556886227544,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.2036,
+      "step": 1117
+    },
+    {
+      "epoch": 1.6736526946107784,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0008,
+      "loss": 1.2022,
+      "step": 1118
+    },
+    {
+      "epoch": 1.6751497005988023,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.2028,
+      "step": 1119
+    },
+    {
+      "epoch": 1.6766467065868262,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.2063,
+      "step": 1120
+    },
+    {
+      "epoch": 1.6781437125748502,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.203,
+      "step": 1121
+    },
+    {
+      "epoch": 1.6796407185628741,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.2046,
+      "step": 1122
+    },
+    {
+      "epoch": 1.681137724550898,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.2002,
+      "step": 1123
+    },
+    {
+      "epoch": 1.6826347305389222,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.2078,
+      "step": 1124
+    },
+    {
+      "epoch": 1.6841317365269461,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.2104,
+      "step": 1125
+    },
+    {
+      "epoch": 1.68562874251497,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.2006,
+      "step": 1126
+    },
+    {
+      "epoch": 1.687125748502994,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.207,
+      "step": 1127
+    },
+    {
+      "epoch": 1.688622754491018,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.2039,
+      "step": 1128
+    },
+    {
+      "epoch": 1.6901197604790419,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.2034,
+      "step": 1129
+    },
+    {
+      "epoch": 1.6916167664670658,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.2091,
+      "step": 1130
+    },
+    {
+      "epoch": 1.69311377245509,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1965,
+      "step": 1131
+    },
+    {
+      "epoch": 1.694610778443114,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.2039,
+      "step": 1132
+    },
+    {
+      "epoch": 1.6961077844311379,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1979,
+      "step": 1133
+    },
+    {
+      "epoch": 1.6976047904191618,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0008,
+      "loss": 1.2006,
+      "step": 1134
+    },
+    {
+      "epoch": 1.6991017964071857,
+      "grad_norm": 0.236328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1998,
+      "step": 1135
+    },
+    {
+      "epoch": 1.7005988023952097,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0008,
+      "loss": 1.2086,
+      "step": 1136
+    },
+    {
+      "epoch": 1.7020958083832336,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0008,
+      "loss": 1.2406,
+      "step": 1137
+    },
+    {
+      "epoch": 1.7035928143712575,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0008,
+      "loss": 1.2445,
+      "step": 1138
+    },
+    {
+      "epoch": 1.7050898203592815,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0008,
+      "loss": 1.2307,
+      "step": 1139
+    },
+    {
+      "epoch": 1.7065868263473054,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0008,
+      "loss": 1.2324,
+      "step": 1140
+    },
+    {
+      "epoch": 1.7080838323353293,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0008,
+      "loss": 1.2266,
+      "step": 1141
+    },
+    {
+      "epoch": 1.7095808383233533,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0008,
+      "loss": 1.2242,
+      "step": 1142
+    },
+    {
+      "epoch": 1.7110778443113772,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0008,
+      "loss": 1.2131,
+      "step": 1143
+    },
+    {
+      "epoch": 1.7125748502994012,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0008,
+      "loss": 1.2149,
+      "step": 1144
+    },
+    {
+      "epoch": 1.714071856287425,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0008,
+      "loss": 1.2205,
+      "step": 1145
+    },
+    {
+      "epoch": 1.715568862275449,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0008,
+      "loss": 1.2154,
+      "step": 1146
+    },
+    {
+      "epoch": 1.717065868263473,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0008,
+      "loss": 1.2186,
+      "step": 1147
+    },
+    {
+      "epoch": 1.718562874251497,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.2107,
+      "step": 1148
+    },
+    {
+      "epoch": 1.7200598802395208,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0008,
+      "loss": 1.2122,
+      "step": 1149
+    },
+    {
+      "epoch": 1.7215568862275448,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0008,
+      "loss": 1.2155,
+      "step": 1150
+    },
+    {
+      "epoch": 1.7230538922155687,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0008,
+      "loss": 1.2003,
+      "step": 1151
+    },
+    {
+      "epoch": 1.7245508982035929,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0008,
+      "loss": 1.2131,
+      "step": 1152
+    },
+    {
+      "epoch": 1.7260479041916168,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.2108,
+      "step": 1153
+    },
+    {
+      "epoch": 1.7275449101796407,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.2048,
+      "step": 1154
+    },
+    {
+      "epoch": 1.7290419161676647,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0008,
+      "loss": 1.2034,
+      "step": 1155
+    },
+    {
+      "epoch": 1.7305389221556886,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.2099,
+      "step": 1156
+    },
+    {
+      "epoch": 1.7320359281437125,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.2067,
+      "step": 1157
+    },
+    {
+      "epoch": 1.7335329341317365,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1989,
+      "step": 1158
+    },
+    {
+      "epoch": 1.7350299401197606,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.2037,
+      "step": 1159
+    },
+    {
+      "epoch": 1.7365269461077846,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1999,
+      "step": 1160
+    },
+    {
+      "epoch": 1.7380239520958085,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.2002,
+      "step": 1161
+    },
+    {
+      "epoch": 1.7395209580838324,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.2082,
+      "step": 1162
+    },
+    {
+      "epoch": 1.7410179640718564,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.2056,
+      "step": 1163
+    },
+    {
+      "epoch": 1.7425149700598803,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.204,
+      "step": 1164
+    },
+    {
+      "epoch": 1.7440119760479043,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.2007,
+      "step": 1165
+    },
+    {
+      "epoch": 1.7455089820359282,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.2006,
+      "step": 1166
+    },
+    {
+      "epoch": 1.7470059880239521,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.2037,
+      "step": 1167
+    },
+    {
+      "epoch": 1.748502994011976,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1902,
+      "step": 1168
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.2009,
+      "step": 1169
+    },
+    {
+      "epoch": 1.751497005988024,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1987,
+      "step": 1170
+    },
+    {
+      "epoch": 1.7529940119760479,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.2076,
+      "step": 1171
+    },
+    {
+      "epoch": 1.7544910179640718,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.2014,
+      "step": 1172
+    },
+    {
+      "epoch": 1.7559880239520957,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1927,
+      "step": 1173
+    },
+    {
+      "epoch": 1.7574850299401197,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.2002,
+      "step": 1174
+    },
+    {
+      "epoch": 1.7589820359281436,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.2018,
+      "step": 1175
+    },
+    {
+      "epoch": 1.7604790419161676,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1954,
+      "step": 1176
+    },
+    {
+      "epoch": 1.7619760479041915,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1957,
+      "step": 1177
+    },
+    {
+      "epoch": 1.7634730538922154,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.201,
+      "step": 1178
+    },
+    {
+      "epoch": 1.7649700598802394,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1982,
+      "step": 1179
+    },
+    {
+      "epoch": 1.7664670658682635,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1993,
+      "step": 1180
+    },
+    {
+      "epoch": 1.7679640718562875,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1968,
+      "step": 1181
+    },
+    {
+      "epoch": 1.7694610778443114,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.198,
+      "step": 1182
+    },
+    {
+      "epoch": 1.7709580838323353,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1968,
+      "step": 1183
+    },
+    {
+      "epoch": 1.7724550898203593,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.2056,
+      "step": 1184
+    },
+    {
+      "epoch": 1.7739520958083832,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1906,
+      "step": 1185
+    },
+    {
+      "epoch": 1.7754491017964071,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.206,
+      "step": 1186
+    },
+    {
+      "epoch": 1.7769461077844313,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1957,
+      "step": 1187
+    },
+    {
+      "epoch": 1.7784431137724552,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1938,
+      "step": 1188
+    },
+    {
+      "epoch": 1.7799401197604792,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1938,
+      "step": 1189
+    },
+    {
+      "epoch": 1.781437125748503,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.2022,
+      "step": 1190
+    },
+    {
+      "epoch": 1.782934131736527,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.2001,
+      "step": 1191
+    },
+    {
+      "epoch": 1.784431137724551,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.2002,
+      "step": 1192
+    },
+    {
+      "epoch": 1.785928143712575,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.2004,
+      "step": 1193
+    },
+    {
+      "epoch": 1.7874251497005988,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1993,
+      "step": 1194
+    },
+    {
+      "epoch": 1.7889221556886228,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.2006,
+      "step": 1195
+    },
+    {
+      "epoch": 1.7904191616766467,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.2053,
+      "step": 1196
+    },
+    {
+      "epoch": 1.7919161676646707,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.2009,
+      "step": 1197
+    },
+    {
+      "epoch": 1.7934131736526946,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1989,
+      "step": 1198
+    },
+    {
+      "epoch": 1.7949101796407185,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.199,
+      "step": 1199
+    },
+    {
+      "epoch": 1.7964071856287425,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1943,
+      "step": 1200
+    },
+    {
+      "epoch": 1.7979041916167664,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1942,
+      "step": 1201
+    },
+    {
+      "epoch": 1.7994011976047903,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1922,
+      "step": 1202
+    },
+    {
+      "epoch": 1.8008982035928143,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1993,
+      "step": 1203
+    },
+    {
+      "epoch": 1.8023952095808382,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0008,
+      "loss": 1.197,
+      "step": 1204
+    },
+    {
+      "epoch": 1.8038922155688621,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.2087,
+      "step": 1205
+    },
+    {
+      "epoch": 1.805389221556886,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1982,
+      "step": 1206
+    },
+    {
+      "epoch": 1.80688622754491,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1941,
+      "step": 1207
+    },
+    {
+      "epoch": 1.8083832335329342,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1965,
+      "step": 1208
+    },
+    {
+      "epoch": 1.8098802395209581,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.2009,
+      "step": 1209
+    },
+    {
+      "epoch": 1.811377245508982,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1945,
+      "step": 1210
+    },
+    {
+      "epoch": 1.812874251497006,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1946,
+      "step": 1211
+    },
+    {
+      "epoch": 1.81437125748503,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1955,
+      "step": 1212
+    },
+    {
+      "epoch": 1.8158682634730539,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.2027,
+      "step": 1213
+    },
+    {
+      "epoch": 1.8173652694610778,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.2017,
+      "step": 1214
+    },
+    {
+      "epoch": 1.818862275449102,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.2033,
+      "step": 1215
+    },
+    {
+      "epoch": 1.8203592814371259,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.2018,
+      "step": 1216
+    },
+    {
+      "epoch": 1.8218562874251498,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.2031,
+      "step": 1217
+    },
+    {
+      "epoch": 1.8233532934131738,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1977,
+      "step": 1218
+    },
+    {
+      "epoch": 1.8248502994011977,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1978,
+      "step": 1219
+    },
+    {
+      "epoch": 1.8263473053892216,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.2027,
+      "step": 1220
+    },
+    {
+      "epoch": 1.8278443113772456,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1979,
+      "step": 1221
+    },
+    {
+      "epoch": 1.8293413173652695,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.2007,
+      "step": 1222
+    },
+    {
+      "epoch": 1.8308383233532934,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.2038,
+      "step": 1223
+    },
+    {
+      "epoch": 1.8323353293413174,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.197,
+      "step": 1224
+    },
+    {
+      "epoch": 1.8338323353293413,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1969,
+      "step": 1225
+    },
+    {
+      "epoch": 1.8353293413173652,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.2045,
+      "step": 1226
+    },
+    {
+      "epoch": 1.8368263473053892,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1982,
+      "step": 1227
+    },
+    {
+      "epoch": 1.8383233532934131,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1956,
+      "step": 1228
+    },
+    {
+      "epoch": 1.839820359281437,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1997,
+      "step": 1229
+    },
+    {
+      "epoch": 1.841317365269461,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1935,
+      "step": 1230
+    },
+    {
+      "epoch": 1.842814371257485,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1979,
+      "step": 1231
+    },
+    {
+      "epoch": 1.8443113772455089,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1963,
+      "step": 1232
+    },
+    {
+      "epoch": 1.8458083832335328,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.2048,
+      "step": 1233
+    },
+    {
+      "epoch": 1.8473053892215567,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1986,
+      "step": 1234
+    },
+    {
+      "epoch": 1.8488023952095807,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1992,
+      "step": 1235
+    },
+    {
+      "epoch": 1.8502994011976048,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1962,
+      "step": 1236
+    },
+    {
+      "epoch": 1.8517964071856288,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1967,
+      "step": 1237
+    },
+    {
+      "epoch": 1.8532934131736527,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1933,
+      "step": 1238
+    },
+    {
+      "epoch": 1.8547904191616766,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1951,
+      "step": 1239
+    },
+    {
+      "epoch": 1.8562874251497006,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1929,
+      "step": 1240
+    },
+    {
+      "epoch": 1.8577844311377245,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1963,
+      "step": 1241
+    },
+    {
+      "epoch": 1.8592814371257484,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1939,
+      "step": 1242
+    },
+    {
+      "epoch": 1.8607784431137726,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1955,
+      "step": 1243
+    },
+    {
+      "epoch": 1.8622754491017965,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1972,
+      "step": 1244
+    },
+    {
+      "epoch": 1.8637724550898205,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.2039,
+      "step": 1245
+    },
+    {
+      "epoch": 1.8652694610778444,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1959,
+      "step": 1246
+    },
+    {
+      "epoch": 1.8667664670658684,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1993,
+      "step": 1247
+    },
+    {
+      "epoch": 1.8682634730538923,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1956,
+      "step": 1248
+    },
+    {
+      "epoch": 1.8697604790419162,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1985,
+      "step": 1249
+    },
+    {
+      "epoch": 1.8712574850299402,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1916,
+      "step": 1250
+    },
+    {
+      "epoch": 1.872754491017964,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1957,
+      "step": 1251
+    },
+    {
+      "epoch": 1.874251497005988,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1984,
+      "step": 1252
+    },
+    {
+      "epoch": 1.875748502994012,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1953,
+      "step": 1253
+    },
+    {
+      "epoch": 1.877245508982036,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.19,
+      "step": 1254
+    },
+    {
+      "epoch": 1.8787425149700598,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1997,
+      "step": 1255
+    },
+    {
+      "epoch": 1.8802395209580838,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1897,
+      "step": 1256
+    },
+    {
+      "epoch": 1.8817365269461077,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1943,
+      "step": 1257
+    },
+    {
+      "epoch": 1.8832335329341316,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1961,
+      "step": 1258
+    },
+    {
+      "epoch": 1.8847305389221556,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1967,
+      "step": 1259
+    },
+    {
+      "epoch": 1.8862275449101795,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1975,
+      "step": 1260
+    },
+    {
+      "epoch": 1.8877245508982035,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1958,
+      "step": 1261
+    },
+    {
+      "epoch": 1.8892215568862274,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1961,
+      "step": 1262
+    },
+    {
+      "epoch": 1.8907185628742516,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1865,
+      "step": 1263
+    },
+    {
+      "epoch": 1.8922155688622755,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1995,
+      "step": 1264
+    },
+    {
+      "epoch": 1.8937125748502994,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.195,
+      "step": 1265
+    },
+    {
+      "epoch": 1.8952095808383234,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1946,
+      "step": 1266
+    },
+    {
+      "epoch": 1.8967065868263473,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.2047,
+      "step": 1267
+    },
+    {
+      "epoch": 1.8982035928143712,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1946,
+      "step": 1268
+    },
+    {
+      "epoch": 1.8997005988023952,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1983,
+      "step": 1269
+    },
+    {
+      "epoch": 1.9011976047904193,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.197,
+      "step": 1270
+    },
+    {
+      "epoch": 1.9026946107784433,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.193,
+      "step": 1271
+    },
+    {
+      "epoch": 1.9041916167664672,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1948,
+      "step": 1272
+    },
+    {
+      "epoch": 1.9056886227544911,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1926,
+      "step": 1273
+    },
+    {
+      "epoch": 1.907185628742515,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1966,
+      "step": 1274
+    },
+    {
+      "epoch": 1.908682634730539,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.2019,
+      "step": 1275
+    },
+    {
+      "epoch": 1.910179640718563,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1955,
+      "step": 1276
+    },
+    {
+      "epoch": 1.9116766467065869,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1993,
+      "step": 1277
+    },
+    {
+      "epoch": 1.9131736526946108,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1926,
+      "step": 1278
+    },
+    {
+      "epoch": 1.9146706586826348,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1943,
+      "step": 1279
+    },
+    {
+      "epoch": 1.9161676646706587,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.2037,
+      "step": 1280
+    },
+    {
+      "epoch": 1.9176646706586826,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.194,
+      "step": 1281
+    },
+    {
+      "epoch": 1.9191616766467066,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1899,
+      "step": 1282
+    },
+    {
+      "epoch": 1.9206586826347305,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.2013,
+      "step": 1283
+    },
+    {
+      "epoch": 1.9221556886227544,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1933,
+      "step": 1284
+    },
+    {
+      "epoch": 1.9236526946107784,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1984,
+      "step": 1285
+    },
+    {
+      "epoch": 1.9251497005988023,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1964,
+      "step": 1286
+    },
+    {
+      "epoch": 1.9266467065868262,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.2036,
+      "step": 1287
+    },
+    {
+      "epoch": 1.9281437125748502,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1956,
+      "step": 1288
+    },
+    {
+      "epoch": 1.9296407185628741,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1956,
+      "step": 1289
+    },
+    {
+      "epoch": 1.931137724550898,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1928,
+      "step": 1290
+    },
+    {
+      "epoch": 1.9326347305389222,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1901,
+      "step": 1291
+    },
+    {
+      "epoch": 1.9341317365269461,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.2008,
+      "step": 1292
+    },
+    {
+      "epoch": 1.93562874251497,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1991,
+      "step": 1293
+    },
+    {
+      "epoch": 1.937125748502994,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1963,
+      "step": 1294
+    },
+    {
+      "epoch": 1.938622754491018,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1948,
+      "step": 1295
+    },
+    {
+      "epoch": 1.9401197604790419,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1986,
+      "step": 1296
+    },
+    {
+      "epoch": 1.9416167664670658,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1952,
+      "step": 1297
+    },
+    {
+      "epoch": 1.94311377245509,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.195,
+      "step": 1298
+    },
+    {
+      "epoch": 1.944610778443114,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.2005,
+      "step": 1299
+    },
+    {
+      "epoch": 1.9461077844311379,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1994,
+      "step": 1300
+    },
+    {
+      "epoch": 1.9476047904191618,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1986,
+      "step": 1301
+    },
+    {
+      "epoch": 1.9491017964071857,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1969,
+      "step": 1302
+    },
+    {
+      "epoch": 1.9505988023952097,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1933,
+      "step": 1303
+    },
+    {
+      "epoch": 1.9520958083832336,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1958,
+      "step": 1304
+    },
+    {
+      "epoch": 1.9535928143712575,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1988,
+      "step": 1305
+    },
+    {
+      "epoch": 1.9550898203592815,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1983,
+      "step": 1306
+    },
+    {
+      "epoch": 1.9565868263473054,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.2006,
+      "step": 1307
+    },
+    {
+      "epoch": 1.9580838323353293,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1964,
+      "step": 1308
+    },
+    {
+      "epoch": 1.9595808383233533,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1933,
+      "step": 1309
+    },
+    {
+      "epoch": 1.9610778443113772,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.2014,
+      "step": 1310
+    },
+    {
+      "epoch": 1.9625748502994012,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.195,
+      "step": 1311
+    },
+    {
+      "epoch": 1.964071856287425,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1968,
+      "step": 1312
+    },
+    {
+      "epoch": 1.965568862275449,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1899,
+      "step": 1313
+    },
+    {
+      "epoch": 1.967065868263473,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1916,
+      "step": 1314
+    },
+    {
+      "epoch": 1.968562874251497,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1935,
+      "step": 1315
+    },
+    {
+      "epoch": 1.9700598802395208,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1894,
+      "step": 1316
+    },
+    {
+      "epoch": 1.9715568862275448,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.194,
+      "step": 1317
+    },
+    {
+      "epoch": 1.9730538922155687,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1992,
+      "step": 1318
+    },
+    {
+      "epoch": 1.9745508982035929,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1932,
+      "step": 1319
+    },
+    {
+      "epoch": 1.9760479041916168,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1972,
+      "step": 1320
+    },
+    {
+      "epoch": 1.9775449101796407,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1878,
+      "step": 1321
+    },
+    {
+      "epoch": 1.9790419161676647,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1934,
+      "step": 1322
+    },
+    {
+      "epoch": 1.9805389221556886,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1848,
+      "step": 1323
+    },
+    {
+      "epoch": 1.9820359281437125,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1922,
+      "step": 1324
+    },
+    {
+      "epoch": 1.9835329341317365,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.2001,
+      "step": 1325
+    },
+    {
+      "epoch": 1.9850299401197606,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1906,
+      "step": 1326
+    },
+    {
+      "epoch": 1.9865269461077846,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1969,
+      "step": 1327
+    },
+    {
+      "epoch": 1.9880239520958085,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1859,
+      "step": 1328
+    },
+    {
+      "epoch": 1.9895209580838324,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1982,
+      "step": 1329
+    },
+    {
+      "epoch": 1.9910179640718564,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1932,
+      "step": 1330
+    },
+    {
+      "epoch": 1.9925149700598803,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1922,
+      "step": 1331
+    },
+    {
+      "epoch": 1.9940119760479043,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1961,
+      "step": 1332
+    },
+    {
+      "epoch": 1.9955089820359282,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.195,
+      "step": 1333
+    },
+    {
+      "epoch": 1.9970059880239521,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1998,
+      "step": 1334
+    },
+    {
+      "epoch": 1.998502994011976,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.2005,
+      "step": 1335
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.2005,
+      "step": 1336
+    },
+    {
+      "epoch": 2.001497005988024,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1958,
+      "step": 1337
+    },
+    {
+      "epoch": 2.002994011976048,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1954,
+      "step": 1338
+    },
+    {
+      "epoch": 2.004491017964072,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1946,
+      "step": 1339
+    },
+    {
+      "epoch": 2.0059880239520957,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.199,
+      "step": 1340
+    },
+    {
+      "epoch": 2.0074850299401197,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1871,
+      "step": 1341
+    },
+    {
+      "epoch": 2.0089820359281436,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1966,
+      "step": 1342
+    },
+    {
+      "epoch": 2.0104790419161676,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1905,
+      "step": 1343
+    },
+    {
+      "epoch": 2.0119760479041915,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.199,
+      "step": 1344
+    },
+    {
+      "epoch": 2.0134730538922154,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1907,
+      "step": 1345
+    },
+    {
+      "epoch": 2.0149700598802394,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.2011,
+      "step": 1346
+    },
+    {
+      "epoch": 2.0164670658682633,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1897,
+      "step": 1347
+    },
+    {
+      "epoch": 2.0179640718562872,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1983,
+      "step": 1348
+    },
+    {
+      "epoch": 2.019461077844311,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1901,
+      "step": 1349
+    },
+    {
+      "epoch": 2.020958083832335,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1962,
+      "step": 1350
+    },
+    {
+      "epoch": 2.0224550898203595,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.194,
+      "step": 1351
+    },
+    {
+      "epoch": 2.0239520958083834,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1909,
+      "step": 1352
+    },
+    {
+      "epoch": 2.0254491017964074,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.2032,
+      "step": 1353
+    },
+    {
+      "epoch": 2.0269461077844313,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1925,
+      "step": 1354
+    },
+    {
+      "epoch": 2.0284431137724552,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1884,
+      "step": 1355
+    },
+    {
+      "epoch": 2.029940119760479,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1977,
+      "step": 1356
+    },
+    {
+      "epoch": 2.031437125748503,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1943,
+      "step": 1357
+    },
+    {
+      "epoch": 2.032934131736527,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1926,
+      "step": 1358
+    },
+    {
+      "epoch": 2.034431137724551,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1863,
+      "step": 1359
+    },
+    {
+      "epoch": 2.035928143712575,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1936,
+      "step": 1360
+    },
+    {
+      "epoch": 2.037425149700599,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1934,
+      "step": 1361
+    },
+    {
+      "epoch": 2.038922155688623,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1982,
+      "step": 1362
+    },
+    {
+      "epoch": 2.0404191616766467,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1959,
+      "step": 1363
+    },
+    {
+      "epoch": 2.0419161676646707,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1966,
+      "step": 1364
+    },
+    {
+      "epoch": 2.0434131736526946,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1923,
+      "step": 1365
+    },
+    {
+      "epoch": 2.0449101796407185,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.194,
+      "step": 1366
+    },
+    {
+      "epoch": 2.0464071856287425,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1953,
+      "step": 1367
+    },
+    {
+      "epoch": 2.0479041916167664,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.193,
+      "step": 1368
+    },
+    {
+      "epoch": 2.0494011976047903,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.194,
+      "step": 1369
+    },
+    {
+      "epoch": 2.0508982035928143,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1945,
+      "step": 1370
+    },
+    {
+      "epoch": 2.052395209580838,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1969,
+      "step": 1371
+    },
+    {
+      "epoch": 2.053892215568862,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1943,
+      "step": 1372
+    },
+    {
+      "epoch": 2.055389221556886,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1939,
+      "step": 1373
+    },
+    {
+      "epoch": 2.05688622754491,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1909,
+      "step": 1374
+    },
+    {
+      "epoch": 2.058383233532934,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1855,
+      "step": 1375
+    },
+    {
+      "epoch": 2.059880239520958,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1896,
+      "step": 1376
+    },
+    {
+      "epoch": 2.061377245508982,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.188,
+      "step": 1377
+    },
+    {
+      "epoch": 2.062874251497006,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1959,
+      "step": 1378
+    },
+    {
+      "epoch": 2.06437125748503,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1956,
+      "step": 1379
+    },
+    {
+      "epoch": 2.065868263473054,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1839,
+      "step": 1380
+    },
+    {
+      "epoch": 2.067365269461078,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1939,
+      "step": 1381
+    },
+    {
+      "epoch": 2.068862275449102,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1906,
+      "step": 1382
+    },
+    {
+      "epoch": 2.070359281437126,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1945,
+      "step": 1383
+    },
+    {
+      "epoch": 2.07185628742515,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1945,
+      "step": 1384
+    },
+    {
+      "epoch": 2.0733532934131738,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1956,
+      "step": 1385
+    },
+    {
+      "epoch": 2.0748502994011977,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1963,
+      "step": 1386
+    },
+    {
+      "epoch": 2.0763473053892216,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1867,
+      "step": 1387
+    },
+    {
+      "epoch": 2.0778443113772456,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1957,
+      "step": 1388
+    },
+    {
+      "epoch": 2.0793413173652695,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1934,
+      "step": 1389
+    },
+    {
+      "epoch": 2.0808383233532934,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.193,
+      "step": 1390
+    },
+    {
+      "epoch": 2.0823353293413174,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1868,
+      "step": 1391
+    },
+    {
+      "epoch": 2.0838323353293413,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1901,
+      "step": 1392
+    },
+    {
+      "epoch": 2.0853293413173652,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1895,
+      "step": 1393
+    },
+    {
+      "epoch": 2.086826347305389,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1976,
+      "step": 1394
+    },
+    {
+      "epoch": 2.088323353293413,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1947,
+      "step": 1395
+    },
+    {
+      "epoch": 2.089820359281437,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1936,
+      "step": 1396
+    },
+    {
+      "epoch": 2.091317365269461,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1937,
+      "step": 1397
+    },
+    {
+      "epoch": 2.092814371257485,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.186,
+      "step": 1398
+    },
+    {
+      "epoch": 2.094311377245509,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1921,
+      "step": 1399
+    },
+    {
+      "epoch": 2.095808383233533,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1882,
+      "step": 1400
+    },
+    {
+      "epoch": 2.0973053892215567,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1893,
+      "step": 1401
+    },
+    {
+      "epoch": 2.0988023952095807,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1941,
+      "step": 1402
+    },
+    {
+      "epoch": 2.1002994011976046,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1909,
+      "step": 1403
+    },
+    {
+      "epoch": 2.1017964071856285,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1901,
+      "step": 1404
+    },
+    {
+      "epoch": 2.1032934131736525,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1978,
+      "step": 1405
+    },
+    {
+      "epoch": 2.1047904191616764,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1931,
+      "step": 1406
+    },
+    {
+      "epoch": 2.106287425149701,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1882,
+      "step": 1407
+    },
+    {
+      "epoch": 2.1077844311377247,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.186,
+      "step": 1408
+    },
+    {
+      "epoch": 2.1092814371257487,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1917,
+      "step": 1409
+    },
+    {
+      "epoch": 2.1107784431137726,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1915,
+      "step": 1410
+    },
+    {
+      "epoch": 2.1122754491017965,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1943,
+      "step": 1411
+    },
+    {
+      "epoch": 2.1137724550898205,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1916,
+      "step": 1412
+    },
+    {
+      "epoch": 2.1152694610778444,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1839,
+      "step": 1413
+    },
+    {
+      "epoch": 2.1167664670658684,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1864,
+      "step": 1414
+    },
+    {
+      "epoch": 2.1182634730538923,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1986,
+      "step": 1415
+    },
+    {
+      "epoch": 2.1197604790419162,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1941,
+      "step": 1416
+    },
+    {
+      "epoch": 2.12125748502994,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1936,
+      "step": 1417
+    },
+    {
+      "epoch": 2.122754491017964,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1944,
+      "step": 1418
+    },
+    {
+      "epoch": 2.124251497005988,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1981,
+      "step": 1419
+    },
+    {
+      "epoch": 2.125748502994012,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1893,
+      "step": 1420
+    },
+    {
+      "epoch": 2.127245508982036,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1952,
+      "step": 1421
+    },
+    {
+      "epoch": 2.12874251497006,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1851,
+      "step": 1422
+    },
+    {
+      "epoch": 2.1302395209580838,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1933,
+      "step": 1423
+    },
+    {
+      "epoch": 2.1317365269461077,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0008,
+      "loss": 1.1958,
+      "step": 1424
+    },
+    {
+      "epoch": 2.1332335329341316,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1903,
+      "step": 1425
+    },
+    {
+      "epoch": 2.1347305389221556,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0008,
+      "loss": 1.191,
+      "step": 1426
+    },
+    {
+      "epoch": 2.1362275449101795,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1894,
+      "step": 1427
+    },
+    {
+      "epoch": 2.1377245508982035,
+      "grad_norm": 0.228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1844,
+      "step": 1428
+    },
+    {
+      "epoch": 2.1392215568862274,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1981,
+      "step": 1429
+    },
+    {
+      "epoch": 2.1407185628742513,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.0008,
+      "loss": 1.205,
+      "step": 1430
+    },
+    {
+      "epoch": 2.1422155688622753,
+      "grad_norm": 1.875,
+      "learning_rate": 0.0008,
+      "loss": 1.2759,
+      "step": 1431
+    },
+    {
+      "epoch": 2.143712574850299,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0008,
+      "loss": 1.2266,
+      "step": 1432
+    },
+    {
+      "epoch": 2.1452095808383236,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.0008,
+      "loss": 1.2652,
+      "step": 1433
+    },
+    {
+      "epoch": 2.1467065868263475,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.2499,
+      "step": 1434
+    },
+    {
+      "epoch": 2.1482035928143715,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0008,
+      "loss": 1.2517,
+      "step": 1435
+    },
+    {
+      "epoch": 2.1497005988023954,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0008,
+      "loss": 1.2465,
+      "step": 1436
+    },
+    {
+      "epoch": 2.1511976047904193,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0008,
+      "loss": 1.2361,
+      "step": 1437
+    },
+    {
+      "epoch": 2.1526946107784433,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0008,
+      "loss": 1.2366,
+      "step": 1438
+    },
+    {
+      "epoch": 2.154191616766467,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0008,
+      "loss": 1.2336,
+      "step": 1439
+    },
+    {
+      "epoch": 2.155688622754491,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0008,
+      "loss": 1.23,
+      "step": 1440
+    },
+    {
+      "epoch": 2.157185628742515,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0008,
+      "loss": 1.2179,
+      "step": 1441
+    },
+    {
+      "epoch": 2.158682634730539,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0008,
+      "loss": 1.2181,
+      "step": 1442
+    },
+    {
+      "epoch": 2.160179640718563,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0008,
+      "loss": 1.2198,
+      "step": 1443
+    },
+    {
+      "epoch": 2.161676646706587,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0008,
+      "loss": 1.2163,
+      "step": 1444
+    },
+    {
+      "epoch": 2.163173652694611,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0008,
+      "loss": 1.212,
+      "step": 1445
+    },
+    {
+      "epoch": 2.1646706586826348,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0008,
+      "loss": 1.2124,
+      "step": 1446
+    },
+    {
+      "epoch": 2.1661676646706587,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0008,
+      "loss": 1.2083,
+      "step": 1447
+    },
+    {
+      "epoch": 2.1676646706586826,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0008,
+      "loss": 1.2074,
+      "step": 1448
+    },
+    {
+      "epoch": 2.1691616766467066,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0008,
+      "loss": 1.2028,
+      "step": 1449
+    },
+    {
+      "epoch": 2.1706586826347305,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0008,
+      "loss": 1.2135,
+      "step": 1450
+    },
+    {
+      "epoch": 2.1721556886227544,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.2023,
+      "step": 1451
+    },
+    {
+      "epoch": 2.1736526946107784,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.2095,
+      "step": 1452
+    },
+    {
+      "epoch": 2.1751497005988023,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.209,
+      "step": 1453
+    },
+    {
+      "epoch": 2.1766467065868262,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.2018,
+      "step": 1454
+    },
+    {
+      "epoch": 2.17814371257485,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1958,
+      "step": 1455
+    },
+    {
+      "epoch": 2.179640718562874,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1998,
+      "step": 1456
+    },
+    {
+      "epoch": 2.181137724550898,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.2,
+      "step": 1457
+    },
+    {
+      "epoch": 2.182634730538922,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.2027,
+      "step": 1458
+    },
+    {
+      "epoch": 2.184131736526946,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1967,
+      "step": 1459
+    },
+    {
+      "epoch": 2.18562874251497,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.2036,
+      "step": 1460
+    },
+    {
+      "epoch": 2.187125748502994,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1984,
+      "step": 1461
+    },
+    {
+      "epoch": 2.1886227544910177,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.2036,
+      "step": 1462
+    },
+    {
+      "epoch": 2.190119760479042,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.205,
+      "step": 1463
+    },
+    {
+      "epoch": 2.191616766467066,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1921,
+      "step": 1464
+    },
+    {
+      "epoch": 2.19311377245509,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.199,
+      "step": 1465
+    },
+    {
+      "epoch": 2.194610778443114,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1914,
+      "step": 1466
+    },
+    {
+      "epoch": 2.196107784431138,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1985,
+      "step": 1467
+    },
+    {
+      "epoch": 2.197604790419162,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.2006,
+      "step": 1468
+    },
+    {
+      "epoch": 2.1991017964071857,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1987,
+      "step": 1469
+    },
+    {
+      "epoch": 2.2005988023952097,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1985,
+      "step": 1470
+    },
+    {
+      "epoch": 2.2020958083832336,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1966,
+      "step": 1471
+    },
+    {
+      "epoch": 2.2035928143712575,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1943,
+      "step": 1472
+    },
+    {
+      "epoch": 2.2050898203592815,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.2011,
+      "step": 1473
+    },
+    {
+      "epoch": 2.2065868263473054,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1975,
+      "step": 1474
+    },
+    {
+      "epoch": 2.2080838323353293,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.2034,
+      "step": 1475
+    },
+    {
+      "epoch": 2.2095808383233533,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1873,
+      "step": 1476
+    },
+    {
+      "epoch": 2.211077844311377,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.2029,
+      "step": 1477
+    },
+    {
+      "epoch": 2.212574850299401,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1927,
+      "step": 1478
+    },
+    {
+      "epoch": 2.214071856287425,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.188,
+      "step": 1479
+    },
+    {
+      "epoch": 2.215568862275449,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1931,
+      "step": 1480
+    },
+    {
+      "epoch": 2.217065868263473,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1884,
+      "step": 1481
+    },
+    {
+      "epoch": 2.218562874251497,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1996,
+      "step": 1482
+    },
+    {
+      "epoch": 2.220059880239521,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1932,
+      "step": 1483
+    },
+    {
+      "epoch": 2.2215568862275448,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.197,
+      "step": 1484
+    },
+    {
+      "epoch": 2.2230538922155687,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1909,
+      "step": 1485
+    },
+    {
+      "epoch": 2.2245508982035926,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1947,
+      "step": 1486
+    },
+    {
+      "epoch": 2.2260479041916166,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1912,
+      "step": 1487
+    },
+    {
+      "epoch": 2.2275449101796405,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1913,
+      "step": 1488
+    },
+    {
+      "epoch": 2.229041916167665,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1919,
+      "step": 1489
+    },
+    {
+      "epoch": 2.230538922155689,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.187,
+      "step": 1490
+    },
+    {
+      "epoch": 2.2320359281437128,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1928,
+      "step": 1491
+    },
+    {
+      "epoch": 2.2335329341317367,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1785,
+      "step": 1492
+    },
+    {
+      "epoch": 2.2350299401197606,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1893,
+      "step": 1493
+    },
+    {
+      "epoch": 2.2365269461077846,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1801,
+      "step": 1494
+    },
+    {
+      "epoch": 2.2380239520958085,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1915,
+      "step": 1495
+    },
+    {
+      "epoch": 2.2395209580838324,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1852,
+      "step": 1496
+    },
+    {
+      "epoch": 2.2410179640718564,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1882,
+      "step": 1497
+    },
+    {
+      "epoch": 2.2425149700598803,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1948,
+      "step": 1498
+    },
+    {
+      "epoch": 2.2440119760479043,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1855,
+      "step": 1499
+    },
+    {
+      "epoch": 2.245508982035928,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1905,
+      "step": 1500
+    },
+    {
+      "epoch": 2.247005988023952,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1941,
+      "step": 1501
+    },
+    {
+      "epoch": 2.248502994011976,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1944,
+      "step": 1502
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1907,
+      "step": 1503
+    },
+    {
+      "epoch": 2.251497005988024,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1924,
+      "step": 1504
+    },
+    {
+      "epoch": 2.252994011976048,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.189,
+      "step": 1505
+    },
+    {
+      "epoch": 2.254491017964072,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1919,
+      "step": 1506
+    },
+    {
+      "epoch": 2.2559880239520957,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1871,
+      "step": 1507
+    },
+    {
+      "epoch": 2.2574850299401197,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.196,
+      "step": 1508
+    },
+    {
+      "epoch": 2.2589820359281436,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1979,
+      "step": 1509
+    },
+    {
+      "epoch": 2.2604790419161676,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1822,
+      "step": 1510
+    },
+    {
+      "epoch": 2.2619760479041915,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1863,
+      "step": 1511
+    },
+    {
+      "epoch": 2.2634730538922154,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1856,
+      "step": 1512
+    },
+    {
+      "epoch": 2.2649700598802394,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1891,
+      "step": 1513
+    },
+    {
+      "epoch": 2.2664670658682633,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1848,
+      "step": 1514
+    },
+    {
+      "epoch": 2.2679640718562872,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1944,
+      "step": 1515
+    },
+    {
+      "epoch": 2.269461077844311,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1957,
+      "step": 1516
+    },
+    {
+      "epoch": 2.270958083832335,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1876,
+      "step": 1517
+    },
+    {
+      "epoch": 2.272455089820359,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.19,
+      "step": 1518
+    },
+    {
+      "epoch": 2.2739520958083834,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1802,
+      "step": 1519
+    },
+    {
+      "epoch": 2.2754491017964074,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1838,
+      "step": 1520
+    },
+    {
+      "epoch": 2.2769461077844313,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1959,
+      "step": 1521
+    },
+    {
+      "epoch": 2.2784431137724552,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1993,
+      "step": 1522
+    },
+    {
+      "epoch": 2.279940119760479,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1901,
+      "step": 1523
+    },
+    {
+      "epoch": 2.281437125748503,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1815,
+      "step": 1524
+    },
+    {
+      "epoch": 2.282934131736527,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1943,
+      "step": 1525
+    },
+    {
+      "epoch": 2.284431137724551,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0008,
+      "loss": 1.2008,
+      "step": 1526
+    },
+    {
+      "epoch": 2.285928143712575,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1897,
+      "step": 1527
+    },
+    {
+      "epoch": 2.287425149700599,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.192,
+      "step": 1528
+    },
+    {
+      "epoch": 2.288922155688623,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.192,
+      "step": 1529
+    },
+    {
+      "epoch": 2.2904191616766467,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1904,
+      "step": 1530
+    },
+    {
+      "epoch": 2.2919161676646707,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1796,
+      "step": 1531
+    },
+    {
+      "epoch": 2.2934131736526946,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1839,
+      "step": 1532
+    },
+    {
+      "epoch": 2.2949101796407185,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1858,
+      "step": 1533
+    },
+    {
+      "epoch": 2.2964071856287425,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1899,
+      "step": 1534
+    },
+    {
+      "epoch": 2.2979041916167664,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.2006,
+      "step": 1535
+    },
+    {
+      "epoch": 2.2994011976047903,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1873,
+      "step": 1536
+    },
+    {
+      "epoch": 2.3008982035928143,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.187,
+      "step": 1537
+    },
+    {
+      "epoch": 2.302395209580838,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1934,
+      "step": 1538
+    },
+    {
+      "epoch": 2.303892215568862,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1892,
+      "step": 1539
+    },
+    {
+      "epoch": 2.305389221556886,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1885,
+      "step": 1540
+    },
+    {
+      "epoch": 2.30688622754491,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.197,
+      "step": 1541
+    },
+    {
+      "epoch": 2.308383233532934,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1912,
+      "step": 1542
+    },
+    {
+      "epoch": 2.309880239520958,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1864,
+      "step": 1543
+    },
+    {
+      "epoch": 2.3113772455089823,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1952,
+      "step": 1544
+    },
+    {
+      "epoch": 2.312874251497006,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1903,
+      "step": 1545
+    },
+    {
+      "epoch": 2.31437125748503,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1918,
+      "step": 1546
+    },
+    {
+      "epoch": 2.315868263473054,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1836,
+      "step": 1547
+    },
+    {
+      "epoch": 2.317365269461078,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.187,
+      "step": 1548
+    },
+    {
+      "epoch": 2.318862275449102,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1908,
+      "step": 1549
+    },
+    {
+      "epoch": 2.320359281437126,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1937,
+      "step": 1550
+    },
+    {
+      "epoch": 2.32185628742515,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1874,
+      "step": 1551
+    },
+    {
+      "epoch": 2.3233532934131738,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1874,
+      "step": 1552
+    },
+    {
+      "epoch": 2.3248502994011977,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1883,
+      "step": 1553
+    },
+    {
+      "epoch": 2.3263473053892216,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1901,
+      "step": 1554
+    },
+    {
+      "epoch": 2.3278443113772456,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1942,
+      "step": 1555
+    },
+    {
+      "epoch": 2.3293413173652695,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1838,
+      "step": 1556
+    },
+    {
+      "epoch": 2.3308383233532934,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.187,
+      "step": 1557
+    },
+    {
+      "epoch": 2.3323353293413174,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1858,
+      "step": 1558
+    },
+    {
+      "epoch": 2.3338323353293413,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1824,
+      "step": 1559
+    },
+    {
+      "epoch": 2.3353293413173652,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1842,
+      "step": 1560
+    },
+    {
+      "epoch": 2.336826347305389,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1931,
+      "step": 1561
+    },
+    {
+      "epoch": 2.338323353293413,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1898,
+      "step": 1562
+    },
+    {
+      "epoch": 2.339820359281437,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.185,
+      "step": 1563
+    },
+    {
+      "epoch": 2.341317365269461,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1816,
+      "step": 1564
+    },
+    {
+      "epoch": 2.342814371257485,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1845,
+      "step": 1565
+    },
+    {
+      "epoch": 2.344311377245509,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1906,
+      "step": 1566
+    },
+    {
+      "epoch": 2.345808383233533,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1877,
+      "step": 1567
+    },
+    {
+      "epoch": 2.3473053892215567,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1931,
+      "step": 1568
+    },
+    {
+      "epoch": 2.3488023952095807,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1833,
+      "step": 1569
+    },
+    {
+      "epoch": 2.3502994011976046,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1893,
+      "step": 1570
+    },
+    {
+      "epoch": 2.3517964071856285,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1978,
+      "step": 1571
+    },
+    {
+      "epoch": 2.3532934131736525,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1813,
+      "step": 1572
+    },
+    {
+      "epoch": 2.3547904191616764,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1875,
+      "step": 1573
+    },
+    {
+      "epoch": 2.3562874251497004,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1831,
+      "step": 1574
+    },
+    {
+      "epoch": 2.3577844311377247,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1852,
+      "step": 1575
+    },
+    {
+      "epoch": 2.3592814371257487,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1816,
+      "step": 1576
+    },
+    {
+      "epoch": 2.3607784431137726,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1909,
+      "step": 1577
+    },
+    {
+      "epoch": 2.3622754491017965,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1883,
+      "step": 1578
+    },
+    {
+      "epoch": 2.3637724550898205,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1891,
+      "step": 1579
+    },
+    {
+      "epoch": 2.3652694610778444,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.189,
+      "step": 1580
+    },
+    {
+      "epoch": 2.3667664670658684,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1867,
+      "step": 1581
+    },
+    {
+      "epoch": 2.3682634730538923,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1893,
+      "step": 1582
+    },
+    {
+      "epoch": 2.3697604790419162,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1884,
+      "step": 1583
+    },
+    {
+      "epoch": 2.37125748502994,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.182,
+      "step": 1584
+    },
+    {
+      "epoch": 2.372754491017964,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.187,
+      "step": 1585
+    },
+    {
+      "epoch": 2.374251497005988,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1887,
+      "step": 1586
+    },
+    {
+      "epoch": 2.375748502994012,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1878,
+      "step": 1587
+    },
+    {
+      "epoch": 2.377245508982036,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1938,
+      "step": 1588
+    },
+    {
+      "epoch": 2.37874251497006,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1822,
+      "step": 1589
+    },
+    {
+      "epoch": 2.3802395209580838,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1834,
+      "step": 1590
+    },
+    {
+      "epoch": 2.3817365269461077,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1851,
+      "step": 1591
+    },
+    {
+      "epoch": 2.3832335329341316,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1846,
+      "step": 1592
+    },
+    {
+      "epoch": 2.3847305389221556,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1839,
+      "step": 1593
+    },
+    {
+      "epoch": 2.3862275449101795,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0008,
+      "loss": 1.182,
+      "step": 1594
+    },
+    {
+      "epoch": 2.3877245508982035,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1845,
+      "step": 1595
+    },
+    {
+      "epoch": 2.3892215568862274,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1937,
+      "step": 1596
+    },
+    {
+      "epoch": 2.3907185628742513,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1867,
+      "step": 1597
+    },
+    {
+      "epoch": 2.3922155688622753,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1821,
+      "step": 1598
+    },
+    {
+      "epoch": 2.3937125748502996,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1908,
+      "step": 1599
+    },
+    {
+      "epoch": 2.3952095808383236,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1937,
+      "step": 1600
+    },
+    {
+      "epoch": 2.3967065868263475,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1865,
+      "step": 1601
+    },
+    {
+      "epoch": 2.3982035928143715,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1889,
+      "step": 1602
+    },
+    {
+      "epoch": 2.3997005988023954,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1921,
+      "step": 1603
+    },
+    {
+      "epoch": 2.4011976047904193,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1867,
+      "step": 1604
+    },
+    {
+      "epoch": 2.4026946107784433,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1819,
+      "step": 1605
+    },
+    {
+      "epoch": 2.404191616766467,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1875,
+      "step": 1606
+    },
+    {
+      "epoch": 2.405688622754491,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1847,
+      "step": 1607
+    },
+    {
+      "epoch": 2.407185628742515,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1846,
+      "step": 1608
+    },
+    {
+      "epoch": 2.408682634730539,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1948,
+      "step": 1609
+    },
+    {
+      "epoch": 2.410179640718563,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.186,
+      "step": 1610
+    },
+    {
+      "epoch": 2.411676646706587,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1886,
+      "step": 1611
+    },
+    {
+      "epoch": 2.413173652694611,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1887,
+      "step": 1612
+    },
+    {
+      "epoch": 2.4146706586826348,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1831,
+      "step": 1613
+    },
+    {
+      "epoch": 2.4161676646706587,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.183,
+      "step": 1614
+    },
+    {
+      "epoch": 2.4176646706586826,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1863,
+      "step": 1615
+    },
+    {
+      "epoch": 2.4191616766467066,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.187,
+      "step": 1616
+    },
+    {
+      "epoch": 2.4206586826347305,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1896,
+      "step": 1617
+    },
+    {
+      "epoch": 2.4221556886227544,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1889,
+      "step": 1618
+    },
+    {
+      "epoch": 2.4236526946107784,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1882,
+      "step": 1619
+    },
+    {
+      "epoch": 2.4251497005988023,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1835,
+      "step": 1620
+    },
+    {
+      "epoch": 2.4266467065868262,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1882,
+      "step": 1621
+    },
+    {
+      "epoch": 2.42814371257485,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1873,
+      "step": 1622
+    },
+    {
+      "epoch": 2.429640718562874,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1948,
+      "step": 1623
+    },
+    {
+      "epoch": 2.431137724550898,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1879,
+      "step": 1624
+    },
+    {
+      "epoch": 2.432634730538922,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1909,
+      "step": 1625
+    },
+    {
+      "epoch": 2.434131736526946,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.178,
+      "step": 1626
+    },
+    {
+      "epoch": 2.43562874251497,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1953,
+      "step": 1627
+    },
+    {
+      "epoch": 2.437125748502994,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1878,
+      "step": 1628
+    },
+    {
+      "epoch": 2.4386227544910177,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1837,
+      "step": 1629
+    },
+    {
+      "epoch": 2.440119760479042,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1779,
+      "step": 1630
+    },
+    {
+      "epoch": 2.441616766467066,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1887,
+      "step": 1631
+    },
+    {
+      "epoch": 2.44311377245509,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1849,
+      "step": 1632
+    },
+    {
+      "epoch": 2.444610778443114,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1887,
+      "step": 1633
+    },
+    {
+      "epoch": 2.446107784431138,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1871,
+      "step": 1634
+    },
+    {
+      "epoch": 2.447604790419162,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1879,
+      "step": 1635
+    },
+    {
+      "epoch": 2.4491017964071857,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1874,
+      "step": 1636
+    },
+    {
+      "epoch": 2.4505988023952097,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1824,
+      "step": 1637
+    },
+    {
+      "epoch": 2.4520958083832336,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1778,
+      "step": 1638
+    },
+    {
+      "epoch": 2.4535928143712575,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1872,
+      "step": 1639
+    },
+    {
+      "epoch": 2.4550898203592815,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1913,
+      "step": 1640
+    },
+    {
+      "epoch": 2.4565868263473054,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1838,
+      "step": 1641
+    },
+    {
+      "epoch": 2.4580838323353293,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1818,
+      "step": 1642
+    },
+    {
+      "epoch": 2.4595808383233533,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1771,
+      "step": 1643
+    },
+    {
+      "epoch": 2.461077844311377,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1915,
+      "step": 1644
+    },
+    {
+      "epoch": 2.462574850299401,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1867,
+      "step": 1645
+    },
+    {
+      "epoch": 2.464071856287425,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1905,
+      "step": 1646
+    },
+    {
+      "epoch": 2.465568862275449,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1843,
+      "step": 1647
+    },
+    {
+      "epoch": 2.467065868263473,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1892,
+      "step": 1648
+    },
+    {
+      "epoch": 2.468562874251497,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.185,
+      "step": 1649
+    },
+    {
+      "epoch": 2.470059880239521,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1863,
+      "step": 1650
+    },
+    {
+      "epoch": 2.4715568862275448,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1839,
+      "step": 1651
+    },
+    {
+      "epoch": 2.4730538922155687,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1917,
+      "step": 1652
+    },
+    {
+      "epoch": 2.4745508982035926,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1821,
+      "step": 1653
+    },
+    {
+      "epoch": 2.4760479041916166,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1858,
+      "step": 1654
+    },
+    {
+      "epoch": 2.477544910179641,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1864,
+      "step": 1655
+    },
+    {
+      "epoch": 2.479041916167665,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1803,
+      "step": 1656
+    },
+    {
+      "epoch": 2.480538922155689,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1925,
+      "step": 1657
+    },
+    {
+      "epoch": 2.4820359281437128,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1813,
+      "step": 1658
+    },
+    {
+      "epoch": 2.4835329341317367,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1904,
+      "step": 1659
+    },
+    {
+      "epoch": 2.4850299401197606,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1896,
+      "step": 1660
+    },
+    {
+      "epoch": 2.4865269461077846,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.188,
+      "step": 1661
+    },
+    {
+      "epoch": 2.4880239520958085,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1813,
+      "step": 1662
+    },
+    {
+      "epoch": 2.4895209580838324,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1888,
+      "step": 1663
+    },
+    {
+      "epoch": 2.4910179640718564,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1751,
+      "step": 1664
+    },
+    {
+      "epoch": 2.4925149700598803,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1868,
+      "step": 1665
+    },
+    {
+      "epoch": 2.4940119760479043,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1874,
+      "step": 1666
+    },
+    {
+      "epoch": 2.495508982035928,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1818,
+      "step": 1667
+    },
+    {
+      "epoch": 2.497005988023952,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1792,
+      "step": 1668
+    },
+    {
+      "epoch": 2.498502994011976,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1833,
+      "step": 1669
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1899,
+      "step": 1670
+    },
+    {
+      "epoch": 2.501497005988024,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1855,
+      "step": 1671
+    },
+    {
+      "epoch": 2.502994011976048,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1812,
+      "step": 1672
+    },
+    {
+      "epoch": 2.504491017964072,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1922,
+      "step": 1673
+    },
+    {
+      "epoch": 2.5059880239520957,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.187,
+      "step": 1674
+    },
+    {
+      "epoch": 2.5074850299401197,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1827,
+      "step": 1675
+    },
+    {
+      "epoch": 2.5089820359281436,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1886,
+      "step": 1676
+    },
+    {
+      "epoch": 2.5104790419161676,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1913,
+      "step": 1677
+    },
+    {
+      "epoch": 2.5119760479041915,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1848,
+      "step": 1678
+    },
+    {
+      "epoch": 2.5134730538922154,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1834,
+      "step": 1679
+    },
+    {
+      "epoch": 2.5149700598802394,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1853,
+      "step": 1680
+    },
+    {
+      "epoch": 2.5164670658682633,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.187,
+      "step": 1681
+    },
+    {
+      "epoch": 2.5179640718562872,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.177,
+      "step": 1682
+    },
+    {
+      "epoch": 2.519461077844311,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1822,
+      "step": 1683
+    },
+    {
+      "epoch": 2.520958083832335,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1879,
+      "step": 1684
+    },
+    {
+      "epoch": 2.522455089820359,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1877,
+      "step": 1685
+    },
+    {
+      "epoch": 2.523952095808383,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1873,
+      "step": 1686
+    },
+    {
+      "epoch": 2.525449101796407,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1849,
+      "step": 1687
+    },
+    {
+      "epoch": 2.5269461077844313,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1831,
+      "step": 1688
+    },
+    {
+      "epoch": 2.5284431137724552,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1909,
+      "step": 1689
+    },
+    {
+      "epoch": 2.529940119760479,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.18,
+      "step": 1690
+    },
+    {
+      "epoch": 2.531437125748503,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1869,
+      "step": 1691
+    },
+    {
+      "epoch": 2.532934131736527,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.19,
+      "step": 1692
+    },
+    {
+      "epoch": 2.534431137724551,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1832,
+      "step": 1693
+    },
+    {
+      "epoch": 2.535928143712575,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1862,
+      "step": 1694
+    },
+    {
+      "epoch": 2.537425149700599,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1863,
+      "step": 1695
+    },
+    {
+      "epoch": 2.538922155688623,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.192,
+      "step": 1696
+    },
+    {
+      "epoch": 2.5404191616766467,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1827,
+      "step": 1697
+    },
+    {
+      "epoch": 2.5419161676646707,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1887,
+      "step": 1698
+    },
+    {
+      "epoch": 2.5434131736526946,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1739,
+      "step": 1699
+    },
+    {
+      "epoch": 2.5449101796407185,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1846,
+      "step": 1700
+    },
+    {
+      "epoch": 2.5464071856287425,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1784,
+      "step": 1701
+    },
+    {
+      "epoch": 2.5479041916167664,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.187,
+      "step": 1702
+    },
+    {
+      "epoch": 2.5494011976047903,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1772,
+      "step": 1703
+    },
+    {
+      "epoch": 2.5508982035928143,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1907,
+      "step": 1704
+    },
+    {
+      "epoch": 2.552395209580838,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1876,
+      "step": 1705
+    },
+    {
+      "epoch": 2.553892215568862,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1839,
+      "step": 1706
+    },
+    {
+      "epoch": 2.555389221556886,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1796,
+      "step": 1707
+    },
+    {
+      "epoch": 2.55688622754491,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1888,
+      "step": 1708
+    },
+    {
+      "epoch": 2.5583832335329344,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1837,
+      "step": 1709
+    },
+    {
+      "epoch": 2.5598802395209583,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1887,
+      "step": 1710
+    },
+    {
+      "epoch": 2.5613772455089823,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1904,
+      "step": 1711
+    },
+    {
+      "epoch": 2.562874251497006,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1858,
+      "step": 1712
+    },
+    {
+      "epoch": 2.56437125748503,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1877,
+      "step": 1713
+    },
+    {
+      "epoch": 2.565868263473054,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1824,
+      "step": 1714
+    },
+    {
+      "epoch": 2.567365269461078,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1892,
+      "step": 1715
+    },
+    {
+      "epoch": 2.568862275449102,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1901,
+      "step": 1716
+    },
+    {
+      "epoch": 2.570359281437126,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1904,
+      "step": 1717
+    },
+    {
+      "epoch": 2.57185628742515,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1868,
+      "step": 1718
+    },
+    {
+      "epoch": 2.5733532934131738,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1862,
+      "step": 1719
+    },
+    {
+      "epoch": 2.5748502994011977,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1782,
+      "step": 1720
+    },
+    {
+      "epoch": 2.5763473053892216,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1874,
+      "step": 1721
+    },
+    {
+      "epoch": 2.5778443113772456,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.186,
+      "step": 1722
+    },
+    {
+      "epoch": 2.5793413173652695,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1846,
+      "step": 1723
+    },
+    {
+      "epoch": 2.5808383233532934,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1858,
+      "step": 1724
+    },
+    {
+      "epoch": 2.5823353293413174,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1868,
+      "step": 1725
+    },
+    {
+      "epoch": 2.5838323353293413,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1804,
+      "step": 1726
+    },
+    {
+      "epoch": 2.5853293413173652,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.182,
+      "step": 1727
+    },
+    {
+      "epoch": 2.586826347305389,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1889,
+      "step": 1728
+    },
+    {
+      "epoch": 2.588323353293413,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1837,
+      "step": 1729
+    },
+    {
+      "epoch": 2.589820359281437,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1867,
+      "step": 1730
+    },
+    {
+      "epoch": 2.591317365269461,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1822,
+      "step": 1731
+    },
+    {
+      "epoch": 2.592814371257485,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.181,
+      "step": 1732
+    },
+    {
+      "epoch": 2.594311377245509,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1816,
+      "step": 1733
+    },
+    {
+      "epoch": 2.595808383233533,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.191,
+      "step": 1734
+    },
+    {
+      "epoch": 2.5973053892215567,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1893,
+      "step": 1735
+    },
+    {
+      "epoch": 2.5988023952095807,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1843,
+      "step": 1736
+    },
+    {
+      "epoch": 2.6002994011976046,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1866,
+      "step": 1737
+    },
+    {
+      "epoch": 2.6017964071856285,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1834,
+      "step": 1738
+    },
+    {
+      "epoch": 2.6032934131736525,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1893,
+      "step": 1739
+    },
+    {
+      "epoch": 2.6047904191616764,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1811,
+      "step": 1740
+    },
+    {
+      "epoch": 2.6062874251497004,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.189,
+      "step": 1741
+    },
+    {
+      "epoch": 2.6077844311377243,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1893,
+      "step": 1742
+    },
+    {
+      "epoch": 2.6092814371257482,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1838,
+      "step": 1743
+    },
+    {
+      "epoch": 2.6107784431137726,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.181,
+      "step": 1744
+    },
+    {
+      "epoch": 2.6122754491017965,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1785,
+      "step": 1745
+    },
+    {
+      "epoch": 2.6137724550898205,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1845,
+      "step": 1746
+    },
+    {
+      "epoch": 2.6152694610778444,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1825,
+      "step": 1747
+    },
+    {
+      "epoch": 2.6167664670658684,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1812,
+      "step": 1748
+    },
+    {
+      "epoch": 2.6182634730538923,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1811,
+      "step": 1749
+    },
+    {
+      "epoch": 2.6197604790419162,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1755,
+      "step": 1750
+    },
+    {
+      "epoch": 2.62125748502994,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1808,
+      "step": 1751
+    },
+    {
+      "epoch": 2.622754491017964,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1755,
+      "step": 1752
+    },
+    {
+      "epoch": 2.624251497005988,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1854,
+      "step": 1753
+    },
+    {
+      "epoch": 2.625748502994012,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1842,
+      "step": 1754
+    },
+    {
+      "epoch": 2.627245508982036,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.187,
+      "step": 1755
+    },
+    {
+      "epoch": 2.62874251497006,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1825,
+      "step": 1756
+    },
+    {
+      "epoch": 2.6302395209580838,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1924,
+      "step": 1757
+    },
+    {
+      "epoch": 2.6317365269461077,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1828,
+      "step": 1758
+    },
+    {
+      "epoch": 2.6332335329341316,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1905,
+      "step": 1759
+    },
+    {
+      "epoch": 2.6347305389221556,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1835,
+      "step": 1760
+    },
+    {
+      "epoch": 2.6362275449101795,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1801,
+      "step": 1761
+    },
+    {
+      "epoch": 2.6377245508982035,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1871,
+      "step": 1762
+    },
+    {
+      "epoch": 2.6392215568862274,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1849,
+      "step": 1763
+    },
+    {
+      "epoch": 2.6407185628742518,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1822,
+      "step": 1764
+    },
+    {
+      "epoch": 2.6422155688622757,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1874,
+      "step": 1765
+    },
+    {
+      "epoch": 2.6437125748502996,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1794,
+      "step": 1766
+    },
+    {
+      "epoch": 2.6452095808383236,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1843,
+      "step": 1767
+    },
+    {
+      "epoch": 2.6467065868263475,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1862,
+      "step": 1768
+    },
+    {
+      "epoch": 2.6482035928143715,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1827,
+      "step": 1769
+    },
+    {
+      "epoch": 2.6497005988023954,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.184,
+      "step": 1770
+    },
+    {
+      "epoch": 2.6511976047904193,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1893,
+      "step": 1771
+    },
+    {
+      "epoch": 2.6526946107784433,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1823,
+      "step": 1772
+    },
+    {
+      "epoch": 2.654191616766467,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1814,
+      "step": 1773
+    },
+    {
+      "epoch": 2.655688622754491,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.185,
+      "step": 1774
+    },
+    {
+      "epoch": 2.657185628742515,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1882,
+      "step": 1775
+    },
+    {
+      "epoch": 2.658682634730539,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1881,
+      "step": 1776
+    },
+    {
+      "epoch": 2.660179640718563,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1912,
+      "step": 1777
+    },
+    {
+      "epoch": 2.661676646706587,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1795,
+      "step": 1778
+    },
+    {
+      "epoch": 2.663173652694611,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1891,
+      "step": 1779
+    },
+    {
+      "epoch": 2.6646706586826348,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1874,
+      "step": 1780
+    },
+    {
+      "epoch": 2.6661676646706587,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1862,
+      "step": 1781
+    },
+    {
+      "epoch": 2.6676646706586826,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1875,
+      "step": 1782
+    },
+    {
+      "epoch": 2.6691616766467066,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1809,
+      "step": 1783
+    },
+    {
+      "epoch": 2.6706586826347305,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1803,
+      "step": 1784
+    },
+    {
+      "epoch": 2.6721556886227544,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1839,
+      "step": 1785
+    },
+    {
+      "epoch": 2.6736526946107784,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1915,
+      "step": 1786
+    },
+    {
+      "epoch": 2.6751497005988023,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1808,
+      "step": 1787
+    },
+    {
+      "epoch": 2.6766467065868262,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1849,
+      "step": 1788
+    },
+    {
+      "epoch": 2.67814371257485,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1855,
+      "step": 1789
+    },
+    {
+      "epoch": 2.679640718562874,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1898,
+      "step": 1790
+    },
+    {
+      "epoch": 2.681137724550898,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1809,
+      "step": 1791
+    },
+    {
+      "epoch": 2.682634730538922,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1851,
+      "step": 1792
+    },
+    {
+      "epoch": 2.684131736526946,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1938,
+      "step": 1793
+    },
+    {
+      "epoch": 2.68562874251497,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1843,
+      "step": 1794
+    },
+    {
+      "epoch": 2.687125748502994,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1805,
+      "step": 1795
+    },
+    {
+      "epoch": 2.6886227544910177,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1851,
+      "step": 1796
+    },
+    {
+      "epoch": 2.6901197604790417,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1797,
+      "step": 1797
+    },
+    {
+      "epoch": 2.6916167664670656,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1814,
+      "step": 1798
+    },
+    {
+      "epoch": 2.69311377245509,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1902,
+      "step": 1799
+    },
+    {
+      "epoch": 2.694610778443114,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.179,
+      "step": 1800
+    },
+    {
+      "epoch": 2.696107784431138,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1783,
+      "step": 1801
+    },
+    {
+      "epoch": 2.697604790419162,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1787,
+      "step": 1802
+    },
+    {
+      "epoch": 2.6991017964071857,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1894,
+      "step": 1803
+    },
+    {
+      "epoch": 2.7005988023952097,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1855,
+      "step": 1804
+    },
+    {
+      "epoch": 2.7020958083832336,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1845,
+      "step": 1805
+    },
+    {
+      "epoch": 2.7035928143712575,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1827,
+      "step": 1806
+    },
+    {
+      "epoch": 2.7050898203592815,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1903,
+      "step": 1807
+    },
+    {
+      "epoch": 2.7065868263473054,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1753,
+      "step": 1808
+    },
+    {
+      "epoch": 2.7080838323353293,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.187,
+      "step": 1809
+    },
+    {
+      "epoch": 2.7095808383233533,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1876,
+      "step": 1810
+    },
+    {
+      "epoch": 2.711077844311377,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1766,
+      "step": 1811
+    },
+    {
+      "epoch": 2.712574850299401,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1816,
+      "step": 1812
+    },
+    {
+      "epoch": 2.714071856287425,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1825,
+      "step": 1813
+    },
+    {
+      "epoch": 2.715568862275449,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1904,
+      "step": 1814
+    },
+    {
+      "epoch": 2.717065868263473,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1805,
+      "step": 1815
+    },
+    {
+      "epoch": 2.718562874251497,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1896,
+      "step": 1816
+    },
+    {
+      "epoch": 2.720059880239521,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1796,
+      "step": 1817
+    },
+    {
+      "epoch": 2.7215568862275448,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1883,
+      "step": 1818
+    },
+    {
+      "epoch": 2.7230538922155687,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1784,
+      "step": 1819
+    },
+    {
+      "epoch": 2.724550898203593,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1865,
+      "step": 1820
+    },
+    {
+      "epoch": 2.726047904191617,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1855,
+      "step": 1821
+    },
+    {
+      "epoch": 2.727544910179641,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0008,
+      "loss": 1.185,
+      "step": 1822
+    },
+    {
+      "epoch": 2.729041916167665,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.185,
+      "step": 1823
+    },
+    {
+      "epoch": 2.730538922155689,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1836,
+      "step": 1824
+    },
+    {
+      "epoch": 2.7320359281437128,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1859,
+      "step": 1825
+    },
+    {
+      "epoch": 2.7335329341317367,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1846,
+      "step": 1826
+    },
+    {
+      "epoch": 2.7350299401197606,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1845,
+      "step": 1827
+    },
+    {
+      "epoch": 2.7365269461077846,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1817,
+      "step": 1828
+    },
+    {
+      "epoch": 2.7380239520958085,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1821,
+      "step": 1829
+    },
+    {
+      "epoch": 2.7395209580838324,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0008,
+      "loss": 1.183,
+      "step": 1830
+    },
+    {
+      "epoch": 2.7410179640718564,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1794,
+      "step": 1831
+    },
+    {
+      "epoch": 2.7425149700598803,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1798,
+      "step": 1832
+    },
+    {
+      "epoch": 2.7440119760479043,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1797,
+      "step": 1833
+    },
+    {
+      "epoch": 2.745508982035928,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1787,
+      "step": 1834
+    },
+    {
+      "epoch": 2.747005988023952,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1872,
+      "step": 1835
+    },
+    {
+      "epoch": 2.748502994011976,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1893,
+      "step": 1836
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1863,
+      "step": 1837
+    },
+    {
+      "epoch": 2.751497005988024,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1857,
+      "step": 1838
+    },
+    {
+      "epoch": 2.752994011976048,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1812,
+      "step": 1839
+    },
+    {
+      "epoch": 2.754491017964072,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1798,
+      "step": 1840
+    },
+    {
+      "epoch": 2.7559880239520957,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1804,
+      "step": 1841
+    },
+    {
+      "epoch": 2.7574850299401197,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1831,
+      "step": 1842
+    },
+    {
+      "epoch": 2.7589820359281436,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1801,
+      "step": 1843
+    },
+    {
+      "epoch": 2.7604790419161676,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1858,
+      "step": 1844
+    },
+    {
+      "epoch": 2.7619760479041915,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1807,
+      "step": 1845
+    },
+    {
+      "epoch": 2.7634730538922154,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1874,
+      "step": 1846
+    },
+    {
+      "epoch": 2.7649700598802394,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1818,
+      "step": 1847
+    },
+    {
+      "epoch": 2.7664670658682633,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1781,
+      "step": 1848
+    },
+    {
+      "epoch": 2.7679640718562872,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1867,
+      "step": 1849
+    },
+    {
+      "epoch": 2.769461077844311,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1813,
+      "step": 1850
+    },
+    {
+      "epoch": 2.770958083832335,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1781,
+      "step": 1851
+    },
+    {
+      "epoch": 2.772455089820359,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1795,
+      "step": 1852
+    },
+    {
+      "epoch": 2.773952095808383,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1887,
+      "step": 1853
+    },
+    {
+      "epoch": 2.775449101796407,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.18,
+      "step": 1854
+    },
+    {
+      "epoch": 2.7769461077844313,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1788,
+      "step": 1855
+    },
+    {
+      "epoch": 2.7784431137724552,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1769,
+      "step": 1856
+    },
+    {
+      "epoch": 2.779940119760479,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1805,
+      "step": 1857
+    },
+    {
+      "epoch": 2.781437125748503,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1799,
+      "step": 1858
+    },
+    {
+      "epoch": 2.782934131736527,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1787,
+      "step": 1859
+    },
+    {
+      "epoch": 2.784431137724551,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1841,
+      "step": 1860
+    },
+    {
+      "epoch": 2.785928143712575,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1733,
+      "step": 1861
+    },
+    {
+      "epoch": 2.787425149700599,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1865,
+      "step": 1862
+    },
+    {
+      "epoch": 2.788922155688623,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1776,
+      "step": 1863
+    },
+    {
+      "epoch": 2.7904191616766467,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1818,
+      "step": 1864
+    },
+    {
+      "epoch": 2.7919161676646707,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1833,
+      "step": 1865
+    },
+    {
+      "epoch": 2.7934131736526946,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1837,
+      "step": 1866
+    },
+    {
+      "epoch": 2.7949101796407185,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1839,
+      "step": 1867
+    },
+    {
+      "epoch": 2.7964071856287425,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1821,
+      "step": 1868
+    },
+    {
+      "epoch": 2.7979041916167664,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1764,
+      "step": 1869
+    },
+    {
+      "epoch": 2.7994011976047903,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1805,
+      "step": 1870
+    },
+    {
+      "epoch": 2.8008982035928143,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1782,
+      "step": 1871
+    },
+    {
+      "epoch": 2.802395209580838,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1757,
+      "step": 1872
+    },
+    {
+      "epoch": 2.803892215568862,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.19,
+      "step": 1873
+    },
+    {
+      "epoch": 2.805389221556886,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1822,
+      "step": 1874
+    },
+    {
+      "epoch": 2.80688622754491,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1874,
+      "step": 1875
+    },
+    {
+      "epoch": 2.8083832335329344,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1769,
+      "step": 1876
+    },
+    {
+      "epoch": 2.8098802395209583,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1805,
+      "step": 1877
+    },
+    {
+      "epoch": 2.8113772455089823,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1895,
+      "step": 1878
+    },
+    {
+      "epoch": 2.812874251497006,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1847,
+      "step": 1879
+    },
+    {
+      "epoch": 2.81437125748503,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1874,
+      "step": 1880
+    },
+    {
+      "epoch": 2.815868263473054,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.182,
+      "step": 1881
+    },
+    {
+      "epoch": 2.817365269461078,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1909,
+      "step": 1882
+    },
+    {
+      "epoch": 2.818862275449102,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1779,
+      "step": 1883
+    },
+    {
+      "epoch": 2.820359281437126,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1803,
+      "step": 1884
+    },
+    {
+      "epoch": 2.82185628742515,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1854,
+      "step": 1885
+    },
+    {
+      "epoch": 2.8233532934131738,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1795,
+      "step": 1886
+    },
+    {
+      "epoch": 2.8248502994011977,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1766,
+      "step": 1887
+    },
+    {
+      "epoch": 2.8263473053892216,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1869,
+      "step": 1888
+    },
+    {
+      "epoch": 2.8278443113772456,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0008,
+      "loss": 1.1896,
+      "step": 1889
+    },
+    {
+      "epoch": 2.8293413173652695,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.0008,
+      "loss": 1.2451,
+      "step": 1890
+    },
+    {
+      "epoch": 2.8308383233532934,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1999,
+      "step": 1891
+    },
+    {
+      "epoch": 2.8323353293413174,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0008,
+      "loss": 1.2046,
+      "step": 1892
+    },
+    {
+      "epoch": 2.8338323353293413,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0008,
+      "loss": 1.2139,
+      "step": 1893
+    },
+    {
+      "epoch": 2.8353293413173652,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0008,
+      "loss": 1.2034,
+      "step": 1894
+    },
+    {
+      "epoch": 2.836826347305389,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0008,
+      "loss": 1.2113,
+      "step": 1895
+    },
+    {
+      "epoch": 2.838323353293413,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0008,
+      "loss": 1.2191,
+      "step": 1896
+    },
+    {
+      "epoch": 2.839820359281437,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.2157,
+      "step": 1897
+    },
+    {
+      "epoch": 2.841317365269461,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0008,
+      "loss": 1.2156,
+      "step": 1898
+    },
+    {
+      "epoch": 2.842814371257485,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0008,
+      "loss": 1.2043,
+      "step": 1899
+    },
+    {
+      "epoch": 2.844311377245509,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0008,
+      "loss": 1.2027,
+      "step": 1900
+    },
+    {
+      "epoch": 2.845808383233533,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.2037,
+      "step": 1901
+    },
+    {
+      "epoch": 2.8473053892215567,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1945,
+      "step": 1902
+    },
+    {
+      "epoch": 2.8488023952095807,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1947,
+      "step": 1903
+    },
+    {
+      "epoch": 2.8502994011976046,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1927,
+      "step": 1904
+    },
+    {
+      "epoch": 2.8517964071856285,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1911,
+      "step": 1905
+    },
+    {
+      "epoch": 2.8532934131736525,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1888,
+      "step": 1906
+    },
+    {
+      "epoch": 2.8547904191616764,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1883,
+      "step": 1907
+    },
+    {
+      "epoch": 2.8562874251497004,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1916,
+      "step": 1908
+    },
+    {
+      "epoch": 2.8577844311377243,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1928,
+      "step": 1909
+    },
+    {
+      "epoch": 2.8592814371257482,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1925,
+      "step": 1910
+    },
+    {
+      "epoch": 2.8607784431137726,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1893,
+      "step": 1911
+    },
+    {
+      "epoch": 2.8622754491017965,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1854,
+      "step": 1912
+    },
+    {
+      "epoch": 2.8637724550898205,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1877,
+      "step": 1913
+    },
+    {
+      "epoch": 2.8652694610778444,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1842,
+      "step": 1914
+    },
+    {
+      "epoch": 2.8667664670658684,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1872,
+      "step": 1915
+    },
+    {
+      "epoch": 2.8682634730538923,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1879,
+      "step": 1916
+    },
+    {
+      "epoch": 2.8697604790419162,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1901,
+      "step": 1917
+    },
+    {
+      "epoch": 2.87125748502994,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1848,
+      "step": 1918
+    },
+    {
+      "epoch": 2.872754491017964,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1843,
+      "step": 1919
+    },
+    {
+      "epoch": 2.874251497005988,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1872,
+      "step": 1920
+    },
+    {
+      "epoch": 2.875748502994012,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1794,
+      "step": 1921
+    },
+    {
+      "epoch": 2.877245508982036,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1844,
+      "step": 1922
+    },
+    {
+      "epoch": 2.87874251497006,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1907,
+      "step": 1923
+    },
+    {
+      "epoch": 2.8802395209580838,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1878,
+      "step": 1924
+    },
+    {
+      "epoch": 2.8817365269461077,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1843,
+      "step": 1925
+    },
+    {
+      "epoch": 2.8832335329341316,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.178,
+      "step": 1926
+    },
+    {
+      "epoch": 2.8847305389221556,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1822,
+      "step": 1927
+    },
+    {
+      "epoch": 2.8862275449101795,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1834,
+      "step": 1928
+    },
+    {
+      "epoch": 2.8877245508982035,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.186,
+      "step": 1929
+    },
+    {
+      "epoch": 2.8892215568862274,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1909,
+      "step": 1930
+    },
+    {
+      "epoch": 2.8907185628742518,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1883,
+      "step": 1931
+    },
+    {
+      "epoch": 2.8922155688622757,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1848,
+      "step": 1932
+    },
+    {
+      "epoch": 2.8937125748502996,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.184,
+      "step": 1933
+    },
+    {
+      "epoch": 2.8952095808383236,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1867,
+      "step": 1934
+    },
+    {
+      "epoch": 2.8967065868263475,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1754,
+      "step": 1935
+    },
+    {
+      "epoch": 2.8982035928143715,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1803,
+      "step": 1936
+    },
+    {
+      "epoch": 2.8997005988023954,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1808,
+      "step": 1937
+    },
+    {
+      "epoch": 2.9011976047904193,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1799,
+      "step": 1938
+    },
+    {
+      "epoch": 2.9026946107784433,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.181,
+      "step": 1939
+    },
+    {
+      "epoch": 2.904191616766467,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1814,
+      "step": 1940
+    },
+    {
+      "epoch": 2.905688622754491,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1832,
+      "step": 1941
+    },
+    {
+      "epoch": 2.907185628742515,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1739,
+      "step": 1942
+    },
+    {
+      "epoch": 2.908682634730539,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1809,
+      "step": 1943
+    },
+    {
+      "epoch": 2.910179640718563,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1867,
+      "step": 1944
+    },
+    {
+      "epoch": 2.911676646706587,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1907,
+      "step": 1945
+    },
+    {
+      "epoch": 2.913173652694611,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1828,
+      "step": 1946
+    },
+    {
+      "epoch": 2.9146706586826348,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1797,
+      "step": 1947
+    },
+    {
+      "epoch": 2.9161676646706587,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1848,
+      "step": 1948
+    },
+    {
+      "epoch": 2.9176646706586826,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1761,
+      "step": 1949
+    },
+    {
+      "epoch": 2.9191616766467066,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1902,
+      "step": 1950
+    },
+    {
+      "epoch": 2.9206586826347305,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1825,
+      "step": 1951
+    },
+    {
+      "epoch": 2.9221556886227544,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1784,
+      "step": 1952
+    },
+    {
+      "epoch": 2.9236526946107784,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1792,
+      "step": 1953
+    },
+    {
+      "epoch": 2.9251497005988023,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.175,
+      "step": 1954
+    },
+    {
+      "epoch": 2.9266467065868262,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1842,
+      "step": 1955
+    },
+    {
+      "epoch": 2.92814371257485,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1811,
+      "step": 1956
+    },
+    {
+      "epoch": 2.929640718562874,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1818,
+      "step": 1957
+    },
+    {
+      "epoch": 2.931137724550898,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1782,
+      "step": 1958
+    },
+    {
+      "epoch": 2.932634730538922,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1768,
+      "step": 1959
+    },
+    {
+      "epoch": 2.934131736526946,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1846,
+      "step": 1960
+    },
+    {
+      "epoch": 2.93562874251497,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1858,
+      "step": 1961
+    },
+    {
+      "epoch": 2.937125748502994,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1895,
+      "step": 1962
+    },
+    {
+      "epoch": 2.9386227544910177,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1838,
+      "step": 1963
+    },
+    {
+      "epoch": 2.9401197604790417,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1748,
+      "step": 1964
+    },
+    {
+      "epoch": 2.9416167664670656,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1831,
+      "step": 1965
+    },
+    {
+      "epoch": 2.94311377245509,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1788,
+      "step": 1966
+    },
+    {
+      "epoch": 2.944610778443114,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1804,
+      "step": 1967
+    },
+    {
+      "epoch": 2.946107784431138,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1756,
+      "step": 1968
+    },
+    {
+      "epoch": 2.947604790419162,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1852,
+      "step": 1969
+    },
+    {
+      "epoch": 2.9491017964071857,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1794,
+      "step": 1970
+    },
+    {
+      "epoch": 2.9505988023952097,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1787,
+      "step": 1971
+    },
+    {
+      "epoch": 2.9520958083832336,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1844,
+      "step": 1972
+    },
+    {
+      "epoch": 2.9535928143712575,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.177,
+      "step": 1973
+    },
+    {
+      "epoch": 2.9550898203592815,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1811,
+      "step": 1974
+    },
+    {
+      "epoch": 2.9565868263473054,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1809,
+      "step": 1975
+    },
+    {
+      "epoch": 2.9580838323353293,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.183,
+      "step": 1976
+    },
+    {
+      "epoch": 2.9595808383233533,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1787,
+      "step": 1977
+    },
+    {
+      "epoch": 2.961077844311377,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1835,
+      "step": 1978
+    },
+    {
+      "epoch": 2.962574850299401,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1846,
+      "step": 1979
+    },
+    {
+      "epoch": 2.964071856287425,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1834,
+      "step": 1980
+    },
+    {
+      "epoch": 2.965568862275449,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1786,
+      "step": 1981
+    },
+    {
+      "epoch": 2.967065868263473,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1827,
+      "step": 1982
+    },
+    {
+      "epoch": 2.968562874251497,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1846,
+      "step": 1983
+    },
+    {
+      "epoch": 2.970059880239521,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1791,
+      "step": 1984
+    },
+    {
+      "epoch": 2.9715568862275448,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.179,
+      "step": 1985
+    },
+    {
+      "epoch": 2.9730538922155687,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1833,
+      "step": 1986
+    },
+    {
+      "epoch": 2.974550898203593,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1856,
+      "step": 1987
+    },
+    {
+      "epoch": 2.976047904191617,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1822,
+      "step": 1988
+    },
+    {
+      "epoch": 2.977544910179641,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1882,
+      "step": 1989
+    },
+    {
+      "epoch": 2.979041916167665,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.184,
+      "step": 1990
+    },
+    {
+      "epoch": 2.980538922155689,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1791,
+      "step": 1991
+    },
+    {
+      "epoch": 2.9820359281437128,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1781,
+      "step": 1992
+    },
+    {
+      "epoch": 2.9835329341317367,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1777,
+      "step": 1993
+    },
+    {
+      "epoch": 2.9850299401197606,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1788,
+      "step": 1994
+    },
+    {
+      "epoch": 2.9865269461077846,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.179,
+      "step": 1995
+    },
+    {
+      "epoch": 2.9880239520958085,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1746,
+      "step": 1996
+    },
+    {
+      "epoch": 2.9895209580838324,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1774,
+      "step": 1997
+    },
+    {
+      "epoch": 2.9910179640718564,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1793,
+      "step": 1998
+    },
+    {
+      "epoch": 2.9925149700598803,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1913,
+      "step": 1999
+    },
+    {
+      "epoch": 2.9940119760479043,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1788,
+      "step": 2000
+    },
+    {
+      "epoch": 2.995508982035928,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1812,
+      "step": 2001
+    },
+    {
+      "epoch": 2.997005988023952,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1774,
+      "step": 2002
+    },
+    {
+      "epoch": 2.998502994011976,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1888,
+      "step": 2003
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1848,
+      "step": 2004
+    },
+    {
+      "epoch": 3.001497005988024,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1775,
+      "step": 2005
+    },
+    {
+      "epoch": 3.002994011976048,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1796,
+      "step": 2006
+    },
+    {
+      "epoch": 3.004491017964072,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1843,
+      "step": 2007
+    },
+    {
+      "epoch": 3.0059880239520957,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1879,
+      "step": 2008
+    },
+    {
+      "epoch": 3.0074850299401197,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1827,
+      "step": 2009
+    },
+    {
+      "epoch": 3.0089820359281436,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1779,
+      "step": 2010
+    },
+    {
+      "epoch": 3.0104790419161676,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1726,
+      "step": 2011
+    },
+    {
+      "epoch": 3.0119760479041915,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1801,
+      "step": 2012
+    },
+    {
+      "epoch": 3.0134730538922154,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1811,
+      "step": 2013
+    },
+    {
+      "epoch": 3.0149700598802394,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1767,
+      "step": 2014
+    },
+    {
+      "epoch": 3.0164670658682633,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1843,
+      "step": 2015
+    },
+    {
+      "epoch": 3.0179640718562872,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1747,
+      "step": 2016
+    },
+    {
+      "epoch": 3.019461077844311,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1788,
+      "step": 2017
+    },
+    {
+      "epoch": 3.020958083832335,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1785,
+      "step": 2018
+    },
+    {
+      "epoch": 3.0224550898203595,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.171,
+      "step": 2019
+    },
+    {
+      "epoch": 3.0239520958083834,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1845,
+      "step": 2020
+    },
+    {
+      "epoch": 3.0254491017964074,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1837,
+      "step": 2021
+    },
+    {
+      "epoch": 3.0269461077844313,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1806,
+      "step": 2022
+    },
+    {
+      "epoch": 3.0284431137724552,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1809,
+      "step": 2023
+    },
+    {
+      "epoch": 3.029940119760479,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1848,
+      "step": 2024
+    },
+    {
+      "epoch": 3.031437125748503,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1782,
+      "step": 2025
+    },
+    {
+      "epoch": 3.032934131736527,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1786,
+      "step": 2026
+    },
+    {
+      "epoch": 3.034431137724551,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1764,
+      "step": 2027
+    },
+    {
+      "epoch": 3.035928143712575,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1857,
+      "step": 2028
+    },
+    {
+      "epoch": 3.037425149700599,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1756,
+      "step": 2029
+    },
+    {
+      "epoch": 3.038922155688623,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1794,
+      "step": 2030
+    },
+    {
+      "epoch": 3.0404191616766467,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.177,
+      "step": 2031
+    },
+    {
+      "epoch": 3.0419161676646707,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1807,
+      "step": 2032
+    },
+    {
+      "epoch": 3.0434131736526946,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1807,
+      "step": 2033
+    },
+    {
+      "epoch": 3.0449101796407185,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1773,
+      "step": 2034
+    },
+    {
+      "epoch": 3.0464071856287425,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1882,
+      "step": 2035
+    },
+    {
+      "epoch": 3.0479041916167664,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1763,
+      "step": 2036
+    },
+    {
+      "epoch": 3.0494011976047903,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1784,
+      "step": 2037
+    },
+    {
+      "epoch": 3.0508982035928143,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1741,
+      "step": 2038
+    },
+    {
+      "epoch": 3.052395209580838,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1798,
+      "step": 2039
+    },
+    {
+      "epoch": 3.053892215568862,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1907,
+      "step": 2040
+    },
+    {
+      "epoch": 3.055389221556886,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1814,
+      "step": 2041
+    },
+    {
+      "epoch": 3.05688622754491,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1848,
+      "step": 2042
+    },
+    {
+      "epoch": 3.058383233532934,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1797,
+      "step": 2043
+    },
+    {
+      "epoch": 3.059880239520958,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1734,
+      "step": 2044
+    },
+    {
+      "epoch": 3.061377245508982,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1807,
+      "step": 2045
+    },
+    {
+      "epoch": 3.062874251497006,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1824,
+      "step": 2046
+    },
+    {
+      "epoch": 3.06437125748503,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1758,
+      "step": 2047
+    },
+    {
+      "epoch": 3.065868263473054,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1851,
+      "step": 2048
+    },
+    {
+      "epoch": 3.067365269461078,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1798,
+      "step": 2049
+    },
+    {
+      "epoch": 3.068862275449102,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1795,
+      "step": 2050
+    },
+    {
+      "epoch": 3.070359281437126,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1779,
+      "step": 2051
+    },
+    {
+      "epoch": 3.07185628742515,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1715,
+      "step": 2052
+    },
+    {
+      "epoch": 3.0733532934131738,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1796,
+      "step": 2053
+    },
+    {
+      "epoch": 3.0748502994011977,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1826,
+      "step": 2054
+    },
+    {
+      "epoch": 3.0763473053892216,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1817,
+      "step": 2055
+    },
+    {
+      "epoch": 3.0778443113772456,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1742,
+      "step": 2056
+    },
+    {
+      "epoch": 3.0793413173652695,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1801,
+      "step": 2057
+    },
+    {
+      "epoch": 3.0808383233532934,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1813,
+      "step": 2058
+    },
+    {
+      "epoch": 3.0823353293413174,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1811,
+      "step": 2059
+    },
+    {
+      "epoch": 3.0838323353293413,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1775,
+      "step": 2060
+    },
+    {
+      "epoch": 3.0853293413173652,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1868,
+      "step": 2061
+    },
+    {
+      "epoch": 3.086826347305389,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1824,
+      "step": 2062
+    },
+    {
+      "epoch": 3.088323353293413,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1796,
+      "step": 2063
+    },
+    {
+      "epoch": 3.089820359281437,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1829,
+      "step": 2064
+    },
+    {
+      "epoch": 3.091317365269461,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1758,
+      "step": 2065
+    },
+    {
+      "epoch": 3.092814371257485,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1802,
+      "step": 2066
+    },
+    {
+      "epoch": 3.094311377245509,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1823,
+      "step": 2067
+    },
+    {
+      "epoch": 3.095808383233533,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1808,
+      "step": 2068
+    },
+    {
+      "epoch": 3.0973053892215567,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1794,
+      "step": 2069
+    },
+    {
+      "epoch": 3.0988023952095807,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1798,
+      "step": 2070
+    },
+    {
+      "epoch": 3.1002994011976046,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1805,
+      "step": 2071
+    },
+    {
+      "epoch": 3.1017964071856285,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1762,
+      "step": 2072
+    },
+    {
+      "epoch": 3.1032934131736525,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1714,
+      "step": 2073
+    },
+    {
+      "epoch": 3.1047904191616764,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1789,
+      "step": 2074
+    },
+    {
+      "epoch": 3.106287425149701,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1812,
+      "step": 2075
+    },
+    {
+      "epoch": 3.1077844311377247,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1792,
+      "step": 2076
+    },
+    {
+      "epoch": 3.1092814371257487,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.182,
+      "step": 2077
+    },
+    {
+      "epoch": 3.1107784431137726,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1872,
+      "step": 2078
+    },
+    {
+      "epoch": 3.1122754491017965,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1803,
+      "step": 2079
+    },
+    {
+      "epoch": 3.1137724550898205,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1795,
+      "step": 2080
+    },
+    {
+      "epoch": 3.1152694610778444,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1806,
+      "step": 2081
+    },
+    {
+      "epoch": 3.1167664670658684,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1792,
+      "step": 2082
+    },
+    {
+      "epoch": 3.1182634730538923,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1825,
+      "step": 2083
+    },
+    {
+      "epoch": 3.1197604790419162,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1732,
+      "step": 2084
+    },
+    {
+      "epoch": 3.12125748502994,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.182,
+      "step": 2085
+    },
+    {
+      "epoch": 3.122754491017964,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1848,
+      "step": 2086
+    },
+    {
+      "epoch": 3.124251497005988,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1771,
+      "step": 2087
+    },
+    {
+      "epoch": 3.125748502994012,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1805,
+      "step": 2088
+    },
+    {
+      "epoch": 3.127245508982036,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1764,
+      "step": 2089
+    },
+    {
+      "epoch": 3.12874251497006,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1797,
+      "step": 2090
+    },
+    {
+      "epoch": 3.1302395209580838,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1759,
+      "step": 2091
+    },
+    {
+      "epoch": 3.1317365269461077,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1763,
+      "step": 2092
+    },
+    {
+      "epoch": 3.1332335329341316,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1781,
+      "step": 2093
+    },
+    {
+      "epoch": 3.1347305389221556,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1765,
+      "step": 2094
+    },
+    {
+      "epoch": 3.1362275449101795,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1793,
+      "step": 2095
+    },
+    {
+      "epoch": 3.1377245508982035,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1753,
+      "step": 2096
+    },
+    {
+      "epoch": 3.1392215568862274,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1779,
+      "step": 2097
+    },
+    {
+      "epoch": 3.1407185628742513,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1788,
+      "step": 2098
+    },
+    {
+      "epoch": 3.1422155688622753,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1756,
+      "step": 2099
+    },
+    {
+      "epoch": 3.143712574850299,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1834,
+      "step": 2100
+    },
+    {
+      "epoch": 3.1452095808383236,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1811,
+      "step": 2101
+    },
+    {
+      "epoch": 3.1467065868263475,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1784,
+      "step": 2102
+    },
+    {
+      "epoch": 3.1482035928143715,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1851,
+      "step": 2103
+    },
+    {
+      "epoch": 3.1497005988023954,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1792,
+      "step": 2104
+    },
+    {
+      "epoch": 3.1511976047904193,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1804,
+      "step": 2105
+    },
+    {
+      "epoch": 3.1526946107784433,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1754,
+      "step": 2106
+    },
+    {
+      "epoch": 3.154191616766467,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1767,
+      "step": 2107
+    },
+    {
+      "epoch": 3.155688622754491,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.176,
+      "step": 2108
+    },
+    {
+      "epoch": 3.157185628742515,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1814,
+      "step": 2109
+    },
+    {
+      "epoch": 3.158682634730539,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1724,
+      "step": 2110
+    },
+    {
+      "epoch": 3.160179640718563,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1801,
+      "step": 2111
+    },
+    {
+      "epoch": 3.161676646706587,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1784,
+      "step": 2112
+    },
+    {
+      "epoch": 3.163173652694611,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1851,
+      "step": 2113
+    },
+    {
+      "epoch": 3.1646706586826348,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1815,
+      "step": 2114
+    },
+    {
+      "epoch": 3.1661676646706587,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1734,
+      "step": 2115
+    },
+    {
+      "epoch": 3.1676646706586826,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1785,
+      "step": 2116
+    },
+    {
+      "epoch": 3.1691616766467066,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1772,
+      "step": 2117
+    },
+    {
+      "epoch": 3.1706586826347305,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1814,
+      "step": 2118
+    },
+    {
+      "epoch": 3.1721556886227544,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1816,
+      "step": 2119
+    },
+    {
+      "epoch": 3.1736526946107784,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1771,
+      "step": 2120
+    },
+    {
+      "epoch": 3.1751497005988023,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1777,
+      "step": 2121
+    },
+    {
+      "epoch": 3.1766467065868262,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1839,
+      "step": 2122
+    },
+    {
+      "epoch": 3.17814371257485,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1765,
+      "step": 2123
+    },
+    {
+      "epoch": 3.179640718562874,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1825,
+      "step": 2124
+    },
+    {
+      "epoch": 3.181137724550898,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1825,
+      "step": 2125
+    },
+    {
+      "epoch": 3.182634730538922,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1772,
+      "step": 2126
+    },
+    {
+      "epoch": 3.184131736526946,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1787,
+      "step": 2127
+    },
+    {
+      "epoch": 3.18562874251497,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1821,
+      "step": 2128
+    },
+    {
+      "epoch": 3.187125748502994,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1792,
+      "step": 2129
+    },
+    {
+      "epoch": 3.1886227544910177,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1773,
+      "step": 2130
+    },
+    {
+      "epoch": 3.190119760479042,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.178,
+      "step": 2131
+    },
+    {
+      "epoch": 3.191616766467066,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1744,
+      "step": 2132
+    },
+    {
+      "epoch": 3.19311377245509,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1752,
+      "step": 2133
+    },
+    {
+      "epoch": 3.194610778443114,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1762,
+      "step": 2134
+    },
+    {
+      "epoch": 3.196107784431138,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1749,
+      "step": 2135
+    },
+    {
+      "epoch": 3.197604790419162,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1742,
+      "step": 2136
+    },
+    {
+      "epoch": 3.1991017964071857,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1725,
+      "step": 2137
+    },
+    {
+      "epoch": 3.2005988023952097,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1854,
+      "step": 2138
+    },
+    {
+      "epoch": 3.2020958083832336,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1817,
+      "step": 2139
+    },
+    {
+      "epoch": 3.2035928143712575,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1702,
+      "step": 2140
+    },
+    {
+      "epoch": 3.2050898203592815,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1785,
+      "step": 2141
+    },
+    {
+      "epoch": 3.2065868263473054,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1799,
+      "step": 2142
+    },
+    {
+      "epoch": 3.2080838323353293,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1773,
+      "step": 2143
+    },
+    {
+      "epoch": 3.2095808383233533,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1805,
+      "step": 2144
+    },
+    {
+      "epoch": 3.211077844311377,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1805,
+      "step": 2145
+    },
+    {
+      "epoch": 3.212574850299401,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.183,
+      "step": 2146
+    },
+    {
+      "epoch": 3.214071856287425,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.181,
+      "step": 2147
+    },
+    {
+      "epoch": 3.215568862275449,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1748,
+      "step": 2148
+    },
+    {
+      "epoch": 3.217065868263473,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1759,
+      "step": 2149
+    },
+    {
+      "epoch": 3.218562874251497,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1768,
+      "step": 2150
+    },
+    {
+      "epoch": 3.220059880239521,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1803,
+      "step": 2151
+    },
+    {
+      "epoch": 3.2215568862275448,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1827,
+      "step": 2152
+    },
+    {
+      "epoch": 3.2230538922155687,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1733,
+      "step": 2153
+    },
+    {
+      "epoch": 3.2245508982035926,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1779,
+      "step": 2154
+    },
+    {
+      "epoch": 3.2260479041916166,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1827,
+      "step": 2155
+    },
+    {
+      "epoch": 3.2275449101796405,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1818,
+      "step": 2156
+    },
+    {
+      "epoch": 3.229041916167665,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.178,
+      "step": 2157
+    },
+    {
+      "epoch": 3.230538922155689,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1727,
+      "step": 2158
+    },
+    {
+      "epoch": 3.2320359281437128,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1732,
+      "step": 2159
+    },
+    {
+      "epoch": 3.2335329341317367,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1828,
+      "step": 2160
+    },
+    {
+      "epoch": 3.2350299401197606,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1781,
+      "step": 2161
+    },
+    {
+      "epoch": 3.2365269461077846,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1762,
+      "step": 2162
+    },
+    {
+      "epoch": 3.2380239520958085,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1736,
+      "step": 2163
+    },
+    {
+      "epoch": 3.2395209580838324,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1843,
+      "step": 2164
+    },
+    {
+      "epoch": 3.2410179640718564,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1785,
+      "step": 2165
+    },
+    {
+      "epoch": 3.2425149700598803,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1746,
+      "step": 2166
+    },
+    {
+      "epoch": 3.2440119760479043,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1779,
+      "step": 2167
+    },
+    {
+      "epoch": 3.245508982035928,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1755,
+      "step": 2168
+    },
+    {
+      "epoch": 3.247005988023952,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1766,
+      "step": 2169
+    },
+    {
+      "epoch": 3.248502994011976,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1796,
+      "step": 2170
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1746,
+      "step": 2171
+    },
+    {
+      "epoch": 3.251497005988024,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1758,
+      "step": 2172
+    },
+    {
+      "epoch": 3.252994011976048,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1778,
+      "step": 2173
+    },
+    {
+      "epoch": 3.254491017964072,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1766,
+      "step": 2174
+    },
+    {
+      "epoch": 3.2559880239520957,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1801,
+      "step": 2175
+    },
+    {
+      "epoch": 3.2574850299401197,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1742,
+      "step": 2176
+    },
+    {
+      "epoch": 3.2589820359281436,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1758,
+      "step": 2177
+    },
+    {
+      "epoch": 3.2604790419161676,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1743,
+      "step": 2178
+    },
+    {
+      "epoch": 3.2619760479041915,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1859,
+      "step": 2179
+    },
+    {
+      "epoch": 3.2634730538922154,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1767,
+      "step": 2180
+    },
+    {
+      "epoch": 3.2649700598802394,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1796,
+      "step": 2181
+    },
+    {
+      "epoch": 3.2664670658682633,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1756,
+      "step": 2182
+    },
+    {
+      "epoch": 3.2679640718562872,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1789,
+      "step": 2183
+    },
+    {
+      "epoch": 3.269461077844311,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1796,
+      "step": 2184
+    },
+    {
+      "epoch": 3.270958083832335,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1811,
+      "step": 2185
+    },
+    {
+      "epoch": 3.272455089820359,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1732,
+      "step": 2186
+    },
+    {
+      "epoch": 3.2739520958083834,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1761,
+      "step": 2187
+    },
+    {
+      "epoch": 3.2754491017964074,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1772,
+      "step": 2188
+    },
+    {
+      "epoch": 3.2769461077844313,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1784,
+      "step": 2189
+    },
+    {
+      "epoch": 3.2784431137724552,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1743,
+      "step": 2190
+    },
+    {
+      "epoch": 3.279940119760479,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1662,
+      "step": 2191
+    },
+    {
+      "epoch": 3.281437125748503,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1671,
+      "step": 2192
+    },
+    {
+      "epoch": 3.282934131736527,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1824,
+      "step": 2193
+    },
+    {
+      "epoch": 3.284431137724551,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1787,
+      "step": 2194
+    },
+    {
+      "epoch": 3.285928143712575,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1704,
+      "step": 2195
+    },
+    {
+      "epoch": 3.287425149700599,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1758,
+      "step": 2196
+    },
+    {
+      "epoch": 3.288922155688623,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.173,
+      "step": 2197
+    },
+    {
+      "epoch": 3.2904191616766467,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1809,
+      "step": 2198
+    },
+    {
+      "epoch": 3.2919161676646707,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1722,
+      "step": 2199
+    },
+    {
+      "epoch": 3.2934131736526946,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1825,
+      "step": 2200
+    },
+    {
+      "epoch": 3.2949101796407185,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1796,
+      "step": 2201
+    },
+    {
+      "epoch": 3.2964071856287425,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1778,
+      "step": 2202
+    },
+    {
+      "epoch": 3.2979041916167664,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1735,
+      "step": 2203
+    },
+    {
+      "epoch": 3.2994011976047903,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.179,
+      "step": 2204
+    },
+    {
+      "epoch": 3.3008982035928143,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1829,
+      "step": 2205
+    },
+    {
+      "epoch": 3.302395209580838,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1733,
+      "step": 2206
+    },
+    {
+      "epoch": 3.303892215568862,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.177,
+      "step": 2207
+    },
+    {
+      "epoch": 3.305389221556886,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1741,
+      "step": 2208
+    },
+    {
+      "epoch": 3.30688622754491,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1789,
+      "step": 2209
+    },
+    {
+      "epoch": 3.308383233532934,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1835,
+      "step": 2210
+    },
+    {
+      "epoch": 3.309880239520958,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1797,
+      "step": 2211
+    },
+    {
+      "epoch": 3.3113772455089823,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1758,
+      "step": 2212
+    },
+    {
+      "epoch": 3.312874251497006,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1719,
+      "step": 2213
+    },
+    {
+      "epoch": 3.31437125748503,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1754,
+      "step": 2214
+    },
+    {
+      "epoch": 3.315868263473054,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1757,
+      "step": 2215
+    },
+    {
+      "epoch": 3.317365269461078,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1843,
+      "step": 2216
+    },
+    {
+      "epoch": 3.318862275449102,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1759,
+      "step": 2217
+    },
+    {
+      "epoch": 3.320359281437126,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1732,
+      "step": 2218
+    },
+    {
+      "epoch": 3.32185628742515,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1723,
+      "step": 2219
+    },
+    {
+      "epoch": 3.3233532934131738,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1794,
+      "step": 2220
+    },
+    {
+      "epoch": 3.3248502994011977,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1812,
+      "step": 2221
+    },
+    {
+      "epoch": 3.3263473053892216,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.18,
+      "step": 2222
+    },
+    {
+      "epoch": 3.3278443113772456,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1728,
+      "step": 2223
+    },
+    {
+      "epoch": 3.3293413173652695,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1739,
+      "step": 2224
+    },
+    {
+      "epoch": 3.3308383233532934,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1791,
+      "step": 2225
+    },
+    {
+      "epoch": 3.3323353293413174,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1732,
+      "step": 2226
+    },
+    {
+      "epoch": 3.3338323353293413,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1792,
+      "step": 2227
+    },
+    {
+      "epoch": 3.3353293413173652,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.175,
+      "step": 2228
+    },
+    {
+      "epoch": 3.336826347305389,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1789,
+      "step": 2229
+    },
+    {
+      "epoch": 3.338323353293413,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1707,
+      "step": 2230
+    },
+    {
+      "epoch": 3.339820359281437,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1766,
+      "step": 2231
+    },
+    {
+      "epoch": 3.341317365269461,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1738,
+      "step": 2232
+    },
+    {
+      "epoch": 3.342814371257485,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1719,
+      "step": 2233
+    },
+    {
+      "epoch": 3.344311377245509,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1797,
+      "step": 2234
+    },
+    {
+      "epoch": 3.345808383233533,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.178,
+      "step": 2235
+    },
+    {
+      "epoch": 3.3473053892215567,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1732,
+      "step": 2236
+    },
+    {
+      "epoch": 3.3488023952095807,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1717,
+      "step": 2237
+    },
+    {
+      "epoch": 3.3502994011976046,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1818,
+      "step": 2238
+    },
+    {
+      "epoch": 3.3517964071856285,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1789,
+      "step": 2239
+    },
+    {
+      "epoch": 3.3532934131736525,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1745,
+      "step": 2240
+    },
+    {
+      "epoch": 3.3547904191616764,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1743,
+      "step": 2241
+    },
+    {
+      "epoch": 3.3562874251497004,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1751,
+      "step": 2242
+    },
+    {
+      "epoch": 3.3577844311377247,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1774,
+      "step": 2243
+    },
+    {
+      "epoch": 3.3592814371257487,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1812,
+      "step": 2244
+    },
+    {
+      "epoch": 3.3607784431137726,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.172,
+      "step": 2245
+    },
+    {
+      "epoch": 3.3622754491017965,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1708,
+      "step": 2246
+    },
+    {
+      "epoch": 3.3637724550898205,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1784,
+      "step": 2247
+    },
+    {
+      "epoch": 3.3652694610778444,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1759,
+      "step": 2248
+    },
+    {
+      "epoch": 3.3667664670658684,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1759,
+      "step": 2249
+    },
+    {
+      "epoch": 3.3682634730538923,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.183,
+      "step": 2250
+    },
+    {
+      "epoch": 3.3697604790419162,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1802,
+      "step": 2251
+    },
+    {
+      "epoch": 3.37125748502994,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1726,
+      "step": 2252
+    },
+    {
+      "epoch": 3.372754491017964,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1756,
+      "step": 2253
+    },
+    {
+      "epoch": 3.374251497005988,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1774,
+      "step": 2254
+    },
+    {
+      "epoch": 3.375748502994012,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1728,
+      "step": 2255
+    },
+    {
+      "epoch": 3.377245508982036,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1756,
+      "step": 2256
+    },
+    {
+      "epoch": 3.37874251497006,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1718,
+      "step": 2257
+    },
+    {
+      "epoch": 3.3802395209580838,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1773,
+      "step": 2258
+    },
+    {
+      "epoch": 3.3817365269461077,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1834,
+      "step": 2259
+    },
+    {
+      "epoch": 3.3832335329341316,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1754,
+      "step": 2260
+    },
+    {
+      "epoch": 3.3847305389221556,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1675,
+      "step": 2261
+    },
+    {
+      "epoch": 3.3862275449101795,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0008,
+      "loss": 1.166,
+      "step": 2262
+    },
+    {
+      "epoch": 3.3877245508982035,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1785,
+      "step": 2263
+    },
+    {
+      "epoch": 3.3892215568862274,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1747,
+      "step": 2264
+    },
+    {
+      "epoch": 3.3907185628742513,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1714,
+      "step": 2265
+    },
+    {
+      "epoch": 3.3922155688622753,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0008,
+      "loss": 1.181,
+      "step": 2266
+    },
+    {
+      "epoch": 3.3937125748502996,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1781,
+      "step": 2267
+    },
+    {
+      "epoch": 3.3952095808383236,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1779,
+      "step": 2268
+    },
+    {
+      "epoch": 3.3967065868263475,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1732,
+      "step": 2269
+    },
+    {
+      "epoch": 3.3982035928143715,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1754,
+      "step": 2270
+    },
+    {
+      "epoch": 3.3997005988023954,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1756,
+      "step": 2271
+    },
+    {
+      "epoch": 3.4011976047904193,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1793,
+      "step": 2272
+    },
+    {
+      "epoch": 3.4026946107784433,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1767,
+      "step": 2273
+    },
+    {
+      "epoch": 3.404191616766467,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1695,
+      "step": 2274
+    },
+    {
+      "epoch": 3.405688622754491,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1709,
+      "step": 2275
+    },
+    {
+      "epoch": 3.407185628742515,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1738,
+      "step": 2276
+    },
+    {
+      "epoch": 3.408682634730539,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1665,
+      "step": 2277
+    },
+    {
+      "epoch": 3.410179640718563,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1753,
+      "step": 2278
+    },
+    {
+      "epoch": 3.411676646706587,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1796,
+      "step": 2279
+    },
+    {
+      "epoch": 3.413173652694611,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1796,
+      "step": 2280
+    },
+    {
+      "epoch": 3.4146706586826348,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1808,
+      "step": 2281
+    },
+    {
+      "epoch": 3.4161676646706587,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1722,
+      "step": 2282
+    },
+    {
+      "epoch": 3.4176646706586826,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1758,
+      "step": 2283
+    },
+    {
+      "epoch": 3.4191616766467066,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.173,
+      "step": 2284
+    },
+    {
+      "epoch": 3.4206586826347305,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1709,
+      "step": 2285
+    },
+    {
+      "epoch": 3.4221556886227544,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.178,
+      "step": 2286
+    },
+    {
+      "epoch": 3.4236526946107784,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1745,
+      "step": 2287
+    },
+    {
+      "epoch": 3.4251497005988023,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1794,
+      "step": 2288
+    },
+    {
+      "epoch": 3.4266467065868262,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1739,
+      "step": 2289
+    },
+    {
+      "epoch": 3.42814371257485,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1737,
+      "step": 2290
+    },
+    {
+      "epoch": 3.429640718562874,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1778,
+      "step": 2291
+    },
+    {
+      "epoch": 3.431137724550898,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1774,
+      "step": 2292
+    },
+    {
+      "epoch": 3.432634730538922,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1737,
+      "step": 2293
+    },
+    {
+      "epoch": 3.434131736526946,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1736,
+      "step": 2294
+    },
+    {
+      "epoch": 3.43562874251497,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1722,
+      "step": 2295
+    },
+    {
+      "epoch": 3.437125748502994,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1791,
+      "step": 2296
+    },
+    {
+      "epoch": 3.4386227544910177,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1789,
+      "step": 2297
+    },
+    {
+      "epoch": 3.440119760479042,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1798,
+      "step": 2298
+    },
+    {
+      "epoch": 3.441616766467066,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.176,
+      "step": 2299
+    },
+    {
+      "epoch": 3.44311377245509,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1806,
+      "step": 2300
+    },
+    {
+      "epoch": 3.444610778443114,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1718,
+      "step": 2301
+    },
+    {
+      "epoch": 3.446107784431138,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1687,
+      "step": 2302
+    },
+    {
+      "epoch": 3.447604790419162,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1742,
+      "step": 2303
+    },
+    {
+      "epoch": 3.4491017964071857,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1784,
+      "step": 2304
+    },
+    {
+      "epoch": 3.4505988023952097,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1756,
+      "step": 2305
+    },
+    {
+      "epoch": 3.4520958083832336,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1715,
+      "step": 2306
+    },
+    {
+      "epoch": 3.4535928143712575,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1709,
+      "step": 2307
+    },
+    {
+      "epoch": 3.4550898203592815,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1721,
+      "step": 2308
+    },
+    {
+      "epoch": 3.4565868263473054,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1842,
+      "step": 2309
+    },
+    {
+      "epoch": 3.4580838323353293,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1835,
+      "step": 2310
+    },
+    {
+      "epoch": 3.4595808383233533,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1761,
+      "step": 2311
+    },
+    {
+      "epoch": 3.461077844311377,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1756,
+      "step": 2312
+    },
+    {
+      "epoch": 3.462574850299401,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1753,
+      "step": 2313
+    },
+    {
+      "epoch": 3.464071856287425,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1773,
+      "step": 2314
+    },
+    {
+      "epoch": 3.465568862275449,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1696,
+      "step": 2315
+    },
+    {
+      "epoch": 3.467065868263473,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1746,
+      "step": 2316
+    },
+    {
+      "epoch": 3.468562874251497,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1728,
+      "step": 2317
+    },
+    {
+      "epoch": 3.470059880239521,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1763,
+      "step": 2318
+    },
+    {
+      "epoch": 3.4715568862275448,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1686,
+      "step": 2319
+    },
+    {
+      "epoch": 3.4730538922155687,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1728,
+      "step": 2320
+    },
+    {
+      "epoch": 3.4745508982035926,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1736,
+      "step": 2321
+    },
+    {
+      "epoch": 3.4760479041916166,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1753,
+      "step": 2322
+    },
+    {
+      "epoch": 3.477544910179641,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1808,
+      "step": 2323
+    },
+    {
+      "epoch": 3.479041916167665,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1715,
+      "step": 2324
+    },
+    {
+      "epoch": 3.480538922155689,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1839,
+      "step": 2325
+    },
+    {
+      "epoch": 3.4820359281437128,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1747,
+      "step": 2326
+    },
+    {
+      "epoch": 3.4835329341317367,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1778,
+      "step": 2327
+    },
+    {
+      "epoch": 3.4850299401197606,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1673,
+      "step": 2328
+    },
+    {
+      "epoch": 3.4865269461077846,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1832,
+      "step": 2329
+    },
+    {
+      "epoch": 3.4880239520958085,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.17,
+      "step": 2330
+    },
+    {
+      "epoch": 3.4895209580838324,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1751,
+      "step": 2331
+    },
+    {
+      "epoch": 3.4910179640718564,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.172,
+      "step": 2332
+    },
+    {
+      "epoch": 3.4925149700598803,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.172,
+      "step": 2333
+    },
+    {
+      "epoch": 3.4940119760479043,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1814,
+      "step": 2334
+    },
+    {
+      "epoch": 3.495508982035928,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1769,
+      "step": 2335
+    },
+    {
+      "epoch": 3.497005988023952,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1684,
+      "step": 2336
+    },
+    {
+      "epoch": 3.498502994011976,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1831,
+      "step": 2337
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1738,
+      "step": 2338
+    },
+    {
+      "epoch": 3.501497005988024,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1823,
+      "step": 2339
+    },
+    {
+      "epoch": 3.502994011976048,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1778,
+      "step": 2340
+    },
+    {
+      "epoch": 3.504491017964072,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1802,
+      "step": 2341
+    },
+    {
+      "epoch": 3.5059880239520957,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1745,
+      "step": 2342
+    },
+    {
+      "epoch": 3.5074850299401197,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1705,
+      "step": 2343
+    },
+    {
+      "epoch": 3.5089820359281436,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1705,
+      "step": 2344
+    },
+    {
+      "epoch": 3.5104790419161676,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1762,
+      "step": 2345
+    },
+    {
+      "epoch": 3.5119760479041915,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1699,
+      "step": 2346
+    },
+    {
+      "epoch": 3.5134730538922154,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1743,
+      "step": 2347
+    },
+    {
+      "epoch": 3.5149700598802394,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1783,
+      "step": 2348
+    },
+    {
+      "epoch": 3.5164670658682633,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1728,
+      "step": 2349
+    },
+    {
+      "epoch": 3.5179640718562872,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1739,
+      "step": 2350
+    },
+    {
+      "epoch": 3.519461077844311,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.178,
+      "step": 2351
+    },
+    {
+      "epoch": 3.520958083832335,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.184,
+      "step": 2352
+    },
+    {
+      "epoch": 3.522455089820359,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1738,
+      "step": 2353
+    },
+    {
+      "epoch": 3.523952095808383,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1737,
+      "step": 2354
+    },
+    {
+      "epoch": 3.525449101796407,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1819,
+      "step": 2355
+    },
+    {
+      "epoch": 3.5269461077844313,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1711,
+      "step": 2356
+    },
+    {
+      "epoch": 3.5284431137724552,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1685,
+      "step": 2357
+    },
+    {
+      "epoch": 3.529940119760479,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1759,
+      "step": 2358
+    },
+    {
+      "epoch": 3.531437125748503,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1716,
+      "step": 2359
+    },
+    {
+      "epoch": 3.532934131736527,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1773,
+      "step": 2360
+    },
+    {
+      "epoch": 3.534431137724551,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1782,
+      "step": 2361
+    },
+    {
+      "epoch": 3.535928143712575,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1748,
+      "step": 2362
+    },
+    {
+      "epoch": 3.537425149700599,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1752,
+      "step": 2363
+    },
+    {
+      "epoch": 3.538922155688623,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1691,
+      "step": 2364
+    },
+    {
+      "epoch": 3.5404191616766467,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1711,
+      "step": 2365
+    },
+    {
+      "epoch": 3.5419161676646707,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1825,
+      "step": 2366
+    },
+    {
+      "epoch": 3.5434131736526946,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1831,
+      "step": 2367
+    },
+    {
+      "epoch": 3.5449101796407185,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1745,
+      "step": 2368
+    },
+    {
+      "epoch": 3.5464071856287425,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1772,
+      "step": 2369
+    },
+    {
+      "epoch": 3.5479041916167664,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1674,
+      "step": 2370
+    },
+    {
+      "epoch": 3.5494011976047903,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1781,
+      "step": 2371
+    },
+    {
+      "epoch": 3.5508982035928143,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1764,
+      "step": 2372
+    },
+    {
+      "epoch": 3.552395209580838,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1799,
+      "step": 2373
+    },
+    {
+      "epoch": 3.553892215568862,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1786,
+      "step": 2374
+    },
+    {
+      "epoch": 3.555389221556886,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.178,
+      "step": 2375
+    },
+    {
+      "epoch": 3.55688622754491,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1824,
+      "step": 2376
+    },
+    {
+      "epoch": 3.5583832335329344,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1739,
+      "step": 2377
+    },
+    {
+      "epoch": 3.5598802395209583,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1771,
+      "step": 2378
+    },
+    {
+      "epoch": 3.5613772455089823,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1802,
+      "step": 2379
+    },
+    {
+      "epoch": 3.562874251497006,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1759,
+      "step": 2380
+    },
+    {
+      "epoch": 3.56437125748503,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1756,
+      "step": 2381
+    },
+    {
+      "epoch": 3.565868263473054,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.178,
+      "step": 2382
+    },
+    {
+      "epoch": 3.567365269461078,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1741,
+      "step": 2383
+    },
+    {
+      "epoch": 3.568862275449102,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1812,
+      "step": 2384
+    },
+    {
+      "epoch": 3.570359281437126,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.178,
+      "step": 2385
+    },
+    {
+      "epoch": 3.57185628742515,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1743,
+      "step": 2386
+    },
+    {
+      "epoch": 3.5733532934131738,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1732,
+      "step": 2387
+    },
+    {
+      "epoch": 3.5748502994011977,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1734,
+      "step": 2388
+    },
+    {
+      "epoch": 3.5763473053892216,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1696,
+      "step": 2389
+    },
+    {
+      "epoch": 3.5778443113772456,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1779,
+      "step": 2390
+    },
+    {
+      "epoch": 3.5793413173652695,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1778,
+      "step": 2391
+    },
+    {
+      "epoch": 3.5808383233532934,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1757,
+      "step": 2392
+    },
+    {
+      "epoch": 3.5823353293413174,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1754,
+      "step": 2393
+    },
+    {
+      "epoch": 3.5838323353293413,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1786,
+      "step": 2394
+    },
+    {
+      "epoch": 3.5853293413173652,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1803,
+      "step": 2395
+    },
+    {
+      "epoch": 3.586826347305389,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1738,
+      "step": 2396
+    },
+    {
+      "epoch": 3.588323353293413,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.165,
+      "step": 2397
+    },
+    {
+      "epoch": 3.589820359281437,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1724,
+      "step": 2398
+    },
+    {
+      "epoch": 3.591317365269461,
+      "grad_norm": 0.224609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1817,
+      "step": 2399
+    },
+    {
+      "epoch": 3.592814371257485,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1725,
+      "step": 2400
+    },
+    {
+      "epoch": 3.594311377245509,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1851,
+      "step": 2401
+    },
+    {
+      "epoch": 3.595808383233533,
+      "grad_norm": 1.0,
+      "learning_rate": 0.0008,
+      "loss": 1.1945,
+      "step": 2402
+    },
+    {
+      "epoch": 3.5973053892215567,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.0008,
+      "loss": 1.24,
+      "step": 2403
+    },
+    {
+      "epoch": 3.5988023952095807,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1938,
+      "step": 2404
+    },
+    {
+      "epoch": 3.6002994011976046,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0008,
+      "loss": 1.2344,
+      "step": 2405
+    },
+    {
+      "epoch": 3.6017964071856285,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0008,
+      "loss": 1.2008,
+      "step": 2406
+    },
+    {
+      "epoch": 3.6032934131736525,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0008,
+      "loss": 1.2244,
+      "step": 2407
+    },
+    {
+      "epoch": 3.6047904191616764,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0008,
+      "loss": 1.2083,
+      "step": 2408
+    },
+    {
+      "epoch": 3.6062874251497004,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0008,
+      "loss": 1.198,
+      "step": 2409
+    },
+    {
+      "epoch": 3.6077844311377243,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 0.0008,
+      "loss": 1.2014,
+      "step": 2410
+    },
+    {
+      "epoch": 3.6092814371257482,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1989,
+      "step": 2411
+    },
+    {
+      "epoch": 3.6107784431137726,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1992,
+      "step": 2412
+    },
+    {
+      "epoch": 3.6122754491017965,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1956,
+      "step": 2413
+    },
+    {
+      "epoch": 3.6137724550898205,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1881,
+      "step": 2414
+    },
+    {
+      "epoch": 3.6152694610778444,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0008,
+      "loss": 1.195,
+      "step": 2415
+    },
+    {
+      "epoch": 3.6167664670658684,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1901,
+      "step": 2416
+    },
+    {
+      "epoch": 3.6182634730538923,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1778,
+      "step": 2417
+    },
+    {
+      "epoch": 3.6197604790419162,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1927,
+      "step": 2418
+    },
+    {
+      "epoch": 3.62125748502994,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0008,
+      "loss": 1.184,
+      "step": 2419
+    },
+    {
+      "epoch": 3.622754491017964,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1833,
+      "step": 2420
+    },
+    {
+      "epoch": 3.624251497005988,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1809,
+      "step": 2421
+    },
+    {
+      "epoch": 3.625748502994012,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1795,
+      "step": 2422
+    },
+    {
+      "epoch": 3.627245508982036,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1864,
+      "step": 2423
+    },
+    {
+      "epoch": 3.62874251497006,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1775,
+      "step": 2424
+    },
+    {
+      "epoch": 3.6302395209580838,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.185,
+      "step": 2425
+    },
+    {
+      "epoch": 3.6317365269461077,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.185,
+      "step": 2426
+    },
+    {
+      "epoch": 3.6332335329341316,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1895,
+      "step": 2427
+    },
+    {
+      "epoch": 3.6347305389221556,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1788,
+      "step": 2428
+    },
+    {
+      "epoch": 3.6362275449101795,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1795,
+      "step": 2429
+    },
+    {
+      "epoch": 3.6377245508982035,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1752,
+      "step": 2430
+    },
+    {
+      "epoch": 3.6392215568862274,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1735,
+      "step": 2431
+    },
+    {
+      "epoch": 3.6407185628742518,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1855,
+      "step": 2432
+    },
+    {
+      "epoch": 3.6422155688622757,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1738,
+      "step": 2433
+    },
+    {
+      "epoch": 3.6437125748502996,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.175,
+      "step": 2434
+    },
+    {
+      "epoch": 3.6452095808383236,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1782,
+      "step": 2435
+    },
+    {
+      "epoch": 3.6467065868263475,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1755,
+      "step": 2436
+    },
+    {
+      "epoch": 3.6482035928143715,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1661,
+      "step": 2437
+    },
+    {
+      "epoch": 3.6497005988023954,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1763,
+      "step": 2438
+    },
+    {
+      "epoch": 3.6511976047904193,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.171,
+      "step": 2439
+    },
+    {
+      "epoch": 3.6526946107784433,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1772,
+      "step": 2440
+    },
+    {
+      "epoch": 3.654191616766467,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1738,
+      "step": 2441
+    },
+    {
+      "epoch": 3.655688622754491,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1767,
+      "step": 2442
+    },
+    {
+      "epoch": 3.657185628742515,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1754,
+      "step": 2443
+    },
+    {
+      "epoch": 3.658682634730539,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1735,
+      "step": 2444
+    },
+    {
+      "epoch": 3.660179640718563,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1733,
+      "step": 2445
+    },
+    {
+      "epoch": 3.661676646706587,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1792,
+      "step": 2446
+    },
+    {
+      "epoch": 3.663173652694611,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1738,
+      "step": 2447
+    },
+    {
+      "epoch": 3.6646706586826348,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1787,
+      "step": 2448
+    },
+    {
+      "epoch": 3.6661676646706587,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1723,
+      "step": 2449
+    },
+    {
+      "epoch": 3.6676646706586826,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1706,
+      "step": 2450
+    },
+    {
+      "epoch": 3.6691616766467066,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1735,
+      "step": 2451
+    },
+    {
+      "epoch": 3.6706586826347305,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1801,
+      "step": 2452
+    },
+    {
+      "epoch": 3.6721556886227544,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1686,
+      "step": 2453
+    },
+    {
+      "epoch": 3.6736526946107784,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1707,
+      "step": 2454
+    },
+    {
+      "epoch": 3.6751497005988023,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1685,
+      "step": 2455
+    },
+    {
+      "epoch": 3.6766467065868262,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1617,
+      "step": 2456
+    },
+    {
+      "epoch": 3.67814371257485,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1713,
+      "step": 2457
+    },
+    {
+      "epoch": 3.679640718562874,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.181,
+      "step": 2458
+    },
+    {
+      "epoch": 3.681137724550898,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1706,
+      "step": 2459
+    },
+    {
+      "epoch": 3.682634730538922,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1726,
+      "step": 2460
+    },
+    {
+      "epoch": 3.684131736526946,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1736,
+      "step": 2461
+    },
+    {
+      "epoch": 3.68562874251497,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1752,
+      "step": 2462
+    },
+    {
+      "epoch": 3.687125748502994,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1707,
+      "step": 2463
+    },
+    {
+      "epoch": 3.6886227544910177,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1779,
+      "step": 2464
+    },
+    {
+      "epoch": 3.6901197604790417,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1761,
+      "step": 2465
+    },
+    {
+      "epoch": 3.6916167664670656,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1791,
+      "step": 2466
+    },
+    {
+      "epoch": 3.69311377245509,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1753,
+      "step": 2467
+    },
+    {
+      "epoch": 3.694610778443114,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1748,
+      "step": 2468
+    },
+    {
+      "epoch": 3.696107784431138,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1829,
+      "step": 2469
+    },
+    {
+      "epoch": 3.697604790419162,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1736,
+      "step": 2470
+    },
+    {
+      "epoch": 3.6991017964071857,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1752,
+      "step": 2471
+    },
+    {
+      "epoch": 3.7005988023952097,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1702,
+      "step": 2472
+    },
+    {
+      "epoch": 3.7020958083832336,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.172,
+      "step": 2473
+    },
+    {
+      "epoch": 3.7035928143712575,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1714,
+      "step": 2474
+    },
+    {
+      "epoch": 3.7050898203592815,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1745,
+      "step": 2475
+    },
+    {
+      "epoch": 3.7065868263473054,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1675,
+      "step": 2476
+    },
+    {
+      "epoch": 3.7080838323353293,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1662,
+      "step": 2477
+    },
+    {
+      "epoch": 3.7095808383233533,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1772,
+      "step": 2478
+    },
+    {
+      "epoch": 3.711077844311377,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1769,
+      "step": 2479
+    },
+    {
+      "epoch": 3.712574850299401,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1733,
+      "step": 2480
+    },
+    {
+      "epoch": 3.714071856287425,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.176,
+      "step": 2481
+    },
+    {
+      "epoch": 3.715568862275449,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1835,
+      "step": 2482
+    },
+    {
+      "epoch": 3.717065868263473,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.182,
+      "step": 2483
+    },
+    {
+      "epoch": 3.718562874251497,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.173,
+      "step": 2484
+    },
+    {
+      "epoch": 3.720059880239521,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1762,
+      "step": 2485
+    },
+    {
+      "epoch": 3.7215568862275448,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1737,
+      "step": 2486
+    },
+    {
+      "epoch": 3.7230538922155687,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1685,
+      "step": 2487
+    },
+    {
+      "epoch": 3.724550898203593,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1731,
+      "step": 2488
+    },
+    {
+      "epoch": 3.726047904191617,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1713,
+      "step": 2489
+    },
+    {
+      "epoch": 3.727544910179641,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.168,
+      "step": 2490
+    },
+    {
+      "epoch": 3.729041916167665,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1633,
+      "step": 2491
+    },
+    {
+      "epoch": 3.730538922155689,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1759,
+      "step": 2492
+    },
+    {
+      "epoch": 3.7320359281437128,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.169,
+      "step": 2493
+    },
+    {
+      "epoch": 3.7335329341317367,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1729,
+      "step": 2494
+    },
+    {
+      "epoch": 3.7350299401197606,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1702,
+      "step": 2495
+    },
+    {
+      "epoch": 3.7365269461077846,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1731,
+      "step": 2496
+    },
+    {
+      "epoch": 3.7380239520958085,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1754,
+      "step": 2497
+    },
+    {
+      "epoch": 3.7395209580838324,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1674,
+      "step": 2498
+    },
+    {
+      "epoch": 3.7410179640718564,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1773,
+      "step": 2499
+    },
+    {
+      "epoch": 3.7425149700598803,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1686,
+      "step": 2500
+    },
+    {
+      "epoch": 3.7440119760479043,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1743,
+      "step": 2501
+    },
+    {
+      "epoch": 3.745508982035928,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1712,
+      "step": 2502
+    },
+    {
+      "epoch": 3.747005988023952,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1732,
+      "step": 2503
+    },
+    {
+      "epoch": 3.748502994011976,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1741,
+      "step": 2504
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1766,
+      "step": 2505
+    },
+    {
+      "epoch": 3.751497005988024,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1733,
+      "step": 2506
+    },
+    {
+      "epoch": 3.752994011976048,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1707,
+      "step": 2507
+    },
+    {
+      "epoch": 3.754491017964072,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1724,
+      "step": 2508
+    },
+    {
+      "epoch": 3.7559880239520957,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1689,
+      "step": 2509
+    },
+    {
+      "epoch": 3.7574850299401197,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1753,
+      "step": 2510
+    },
+    {
+      "epoch": 3.7589820359281436,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1746,
+      "step": 2511
+    },
+    {
+      "epoch": 3.7604790419161676,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1692,
+      "step": 2512
+    },
+    {
+      "epoch": 3.7619760479041915,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1814,
+      "step": 2513
+    },
+    {
+      "epoch": 3.7634730538922154,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1742,
+      "step": 2514
+    },
+    {
+      "epoch": 3.7649700598802394,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1716,
+      "step": 2515
+    },
+    {
+      "epoch": 3.7664670658682633,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1728,
+      "step": 2516
+    },
+    {
+      "epoch": 3.7679640718562872,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1756,
+      "step": 2517
+    },
+    {
+      "epoch": 3.769461077844311,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1734,
+      "step": 2518
+    },
+    {
+      "epoch": 3.770958083832335,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1751,
+      "step": 2519
+    },
+    {
+      "epoch": 3.772455089820359,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1694,
+      "step": 2520
+    },
+    {
+      "epoch": 3.773952095808383,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1784,
+      "step": 2521
+    },
+    {
+      "epoch": 3.775449101796407,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1704,
+      "step": 2522
+    },
+    {
+      "epoch": 3.7769461077844313,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1696,
+      "step": 2523
+    },
+    {
+      "epoch": 3.7784431137724552,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1694,
+      "step": 2524
+    },
+    {
+      "epoch": 3.779940119760479,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1702,
+      "step": 2525
+    },
+    {
+      "epoch": 3.781437125748503,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1701,
+      "step": 2526
+    },
+    {
+      "epoch": 3.782934131736527,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1714,
+      "step": 2527
+    },
+    {
+      "epoch": 3.784431137724551,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1746,
+      "step": 2528
+    },
+    {
+      "epoch": 3.785928143712575,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1776,
+      "step": 2529
+    },
+    {
+      "epoch": 3.787425149700599,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1674,
+      "step": 2530
+    },
+    {
+      "epoch": 3.788922155688623,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1682,
+      "step": 2531
+    },
+    {
+      "epoch": 3.7904191616766467,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1761,
+      "step": 2532
+    },
+    {
+      "epoch": 3.7919161676646707,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1727,
+      "step": 2533
+    },
+    {
+      "epoch": 3.7934131736526946,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1713,
+      "step": 2534
+    },
+    {
+      "epoch": 3.7949101796407185,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1706,
+      "step": 2535
+    },
+    {
+      "epoch": 3.7964071856287425,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1706,
+      "step": 2536
+    },
+    {
+      "epoch": 3.7979041916167664,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1829,
+      "step": 2537
+    },
+    {
+      "epoch": 3.7994011976047903,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.169,
+      "step": 2538
+    },
+    {
+      "epoch": 3.8008982035928143,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1814,
+      "step": 2539
+    },
+    {
+      "epoch": 3.802395209580838,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1663,
+      "step": 2540
+    },
+    {
+      "epoch": 3.803892215568862,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1709,
+      "step": 2541
+    },
+    {
+      "epoch": 3.805389221556886,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.167,
+      "step": 2542
+    },
+    {
+      "epoch": 3.80688622754491,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1723,
+      "step": 2543
+    },
+    {
+      "epoch": 3.8083832335329344,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1704,
+      "step": 2544
+    },
+    {
+      "epoch": 3.8098802395209583,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1757,
+      "step": 2545
+    },
+    {
+      "epoch": 3.8113772455089823,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1716,
+      "step": 2546
+    },
+    {
+      "epoch": 3.812874251497006,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.179,
+      "step": 2547
+    },
+    {
+      "epoch": 3.81437125748503,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1723,
+      "step": 2548
+    },
+    {
+      "epoch": 3.815868263473054,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1703,
+      "step": 2549
+    },
+    {
+      "epoch": 3.817365269461078,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1714,
+      "step": 2550
+    },
+    {
+      "epoch": 3.818862275449102,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.176,
+      "step": 2551
+    },
+    {
+      "epoch": 3.820359281437126,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1764,
+      "step": 2552
+    },
+    {
+      "epoch": 3.82185628742515,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.169,
+      "step": 2553
+    },
+    {
+      "epoch": 3.8233532934131738,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1763,
+      "step": 2554
+    },
+    {
+      "epoch": 3.8248502994011977,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1697,
+      "step": 2555
+    },
+    {
+      "epoch": 3.8263473053892216,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1669,
+      "step": 2556
+    },
+    {
+      "epoch": 3.8278443113772456,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1722,
+      "step": 2557
+    },
+    {
+      "epoch": 3.8293413173652695,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.165,
+      "step": 2558
+    },
+    {
+      "epoch": 3.8308383233532934,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1711,
+      "step": 2559
+    },
+    {
+      "epoch": 3.8323353293413174,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1667,
+      "step": 2560
+    },
+    {
+      "epoch": 3.8338323353293413,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1645,
+      "step": 2561
+    },
+    {
+      "epoch": 3.8353293413173652,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1736,
+      "step": 2562
+    },
+    {
+      "epoch": 3.836826347305389,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1743,
+      "step": 2563
+    },
+    {
+      "epoch": 3.838323353293413,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1659,
+      "step": 2564
+    },
+    {
+      "epoch": 3.839820359281437,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1734,
+      "step": 2565
+    },
+    {
+      "epoch": 3.841317365269461,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.176,
+      "step": 2566
+    },
+    {
+      "epoch": 3.842814371257485,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1731,
+      "step": 2567
+    },
+    {
+      "epoch": 3.844311377245509,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1751,
+      "step": 2568
+    },
+    {
+      "epoch": 3.845808383233533,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1704,
+      "step": 2569
+    },
+    {
+      "epoch": 3.8473053892215567,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1695,
+      "step": 2570
+    },
+    {
+      "epoch": 3.8488023952095807,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1724,
+      "step": 2571
+    },
+    {
+      "epoch": 3.8502994011976046,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1688,
+      "step": 2572
+    },
+    {
+      "epoch": 3.8517964071856285,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.171,
+      "step": 2573
+    },
+    {
+      "epoch": 3.8532934131736525,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1785,
+      "step": 2574
+    },
+    {
+      "epoch": 3.8547904191616764,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1703,
+      "step": 2575
+    },
+    {
+      "epoch": 3.8562874251497004,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.171,
+      "step": 2576
+    },
+    {
+      "epoch": 3.8577844311377243,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.172,
+      "step": 2577
+    },
+    {
+      "epoch": 3.8592814371257482,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1724,
+      "step": 2578
+    },
+    {
+      "epoch": 3.8607784431137726,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1699,
+      "step": 2579
+    },
+    {
+      "epoch": 3.8622754491017965,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1742,
+      "step": 2580
+    },
+    {
+      "epoch": 3.8637724550898205,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1762,
+      "step": 2581
+    },
+    {
+      "epoch": 3.8652694610778444,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1658,
+      "step": 2582
+    },
+    {
+      "epoch": 3.8667664670658684,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1766,
+      "step": 2583
+    },
+    {
+      "epoch": 3.8682634730538923,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1762,
+      "step": 2584
+    },
+    {
+      "epoch": 3.8697604790419162,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1842,
+      "step": 2585
+    },
+    {
+      "epoch": 3.87125748502994,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0008,
+      "loss": 1.2016,
+      "step": 2586
+    },
+    {
+      "epoch": 3.872754491017964,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.0008,
+      "loss": 1.195,
+      "step": 2587
+    },
+    {
+      "epoch": 3.874251497005988,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1825,
+      "step": 2588
+    },
+    {
+      "epoch": 3.875748502994012,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0008,
+      "loss": 1.195,
+      "step": 2589
+    },
+    {
+      "epoch": 3.877245508982036,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1849,
+      "step": 2590
+    },
+    {
+      "epoch": 3.87874251497006,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1929,
+      "step": 2591
+    },
+    {
+      "epoch": 3.8802395209580838,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1809,
+      "step": 2592
+    },
+    {
+      "epoch": 3.8817365269461077,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1813,
+      "step": 2593
+    },
+    {
+      "epoch": 3.8832335329341316,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1835,
+      "step": 2594
+    },
+    {
+      "epoch": 3.8847305389221556,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1786,
+      "step": 2595
+    },
+    {
+      "epoch": 3.8862275449101795,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1772,
+      "step": 2596
+    },
+    {
+      "epoch": 3.8877245508982035,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1751,
+      "step": 2597
+    },
+    {
+      "epoch": 3.8892215568862274,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1808,
+      "step": 2598
+    },
+    {
+      "epoch": 3.8907185628742518,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1784,
+      "step": 2599
+    },
+    {
+      "epoch": 3.8922155688622757,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1724,
+      "step": 2600
+    },
+    {
+      "epoch": 3.8937125748502996,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1768,
+      "step": 2601
+    },
+    {
+      "epoch": 3.8952095808383236,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1761,
+      "step": 2602
+    },
+    {
+      "epoch": 3.8967065868263475,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1711,
+      "step": 2603
+    },
+    {
+      "epoch": 3.8982035928143715,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1731,
+      "step": 2604
+    },
+    {
+      "epoch": 3.8997005988023954,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1782,
+      "step": 2605
+    },
+    {
+      "epoch": 3.9011976047904193,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1786,
+      "step": 2606
+    },
+    {
+      "epoch": 3.9026946107784433,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1726,
+      "step": 2607
+    },
+    {
+      "epoch": 3.904191616766467,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1703,
+      "step": 2608
+    },
+    {
+      "epoch": 3.905688622754491,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1773,
+      "step": 2609
+    },
+    {
+      "epoch": 3.907185628742515,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1703,
+      "step": 2610
+    },
+    {
+      "epoch": 3.908682634730539,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1656,
+      "step": 2611
+    },
+    {
+      "epoch": 3.910179640718563,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1732,
+      "step": 2612
+    },
+    {
+      "epoch": 3.911676646706587,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1735,
+      "step": 2613
+    },
+    {
+      "epoch": 3.913173652694611,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1686,
+      "step": 2614
+    },
+    {
+      "epoch": 3.9146706586826348,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1787,
+      "step": 2615
+    },
+    {
+      "epoch": 3.9161676646706587,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1739,
+      "step": 2616
+    },
+    {
+      "epoch": 3.9176646706586826,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1768,
+      "step": 2617
+    },
+    {
+      "epoch": 3.9191616766467066,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1775,
+      "step": 2618
+    },
+    {
+      "epoch": 3.9206586826347305,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1748,
+      "step": 2619
+    },
+    {
+      "epoch": 3.9221556886227544,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1675,
+      "step": 2620
+    },
+    {
+      "epoch": 3.9236526946107784,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1699,
+      "step": 2621
+    },
+    {
+      "epoch": 3.9251497005988023,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1734,
+      "step": 2622
+    },
+    {
+      "epoch": 3.9266467065868262,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1738,
+      "step": 2623
+    },
+    {
+      "epoch": 3.92814371257485,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1694,
+      "step": 2624
+    },
+    {
+      "epoch": 3.929640718562874,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1726,
+      "step": 2625
+    },
+    {
+      "epoch": 3.931137724550898,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1719,
+      "step": 2626
+    },
+    {
+      "epoch": 3.932634730538922,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1727,
+      "step": 2627
+    },
+    {
+      "epoch": 3.934131736526946,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.17,
+      "step": 2628
+    },
+    {
+      "epoch": 3.93562874251497,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1735,
+      "step": 2629
+    },
+    {
+      "epoch": 3.937125748502994,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1718,
+      "step": 2630
+    },
+    {
+      "epoch": 3.9386227544910177,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1733,
+      "step": 2631
+    },
+    {
+      "epoch": 3.9401197604790417,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1728,
+      "step": 2632
+    },
+    {
+      "epoch": 3.9416167664670656,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1714,
+      "step": 2633
+    },
+    {
+      "epoch": 3.94311377245509,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1749,
+      "step": 2634
+    },
+    {
+      "epoch": 3.944610778443114,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1653,
+      "step": 2635
+    },
+    {
+      "epoch": 3.946107784431138,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1714,
+      "step": 2636
+    },
+    {
+      "epoch": 3.947604790419162,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1722,
+      "step": 2637
+    },
+    {
+      "epoch": 3.9491017964071857,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1649,
+      "step": 2638
+    },
+    {
+      "epoch": 3.9505988023952097,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1731,
+      "step": 2639
+    },
+    {
+      "epoch": 3.9520958083832336,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1622,
+      "step": 2640
+    },
+    {
+      "epoch": 3.9535928143712575,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1654,
+      "step": 2641
+    },
+    {
+      "epoch": 3.9550898203592815,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1689,
+      "step": 2642
+    },
+    {
+      "epoch": 3.9565868263473054,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.169,
+      "step": 2643
+    },
+    {
+      "epoch": 3.9580838323353293,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1737,
+      "step": 2644
+    },
+    {
+      "epoch": 3.9595808383233533,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.175,
+      "step": 2645
+    },
+    {
+      "epoch": 3.961077844311377,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1711,
+      "step": 2646
+    },
+    {
+      "epoch": 3.962574850299401,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1717,
+      "step": 2647
+    },
+    {
+      "epoch": 3.964071856287425,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1686,
+      "step": 2648
+    },
+    {
+      "epoch": 3.965568862275449,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1654,
+      "step": 2649
+    },
+    {
+      "epoch": 3.967065868263473,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.178,
+      "step": 2650
+    },
+    {
+      "epoch": 3.968562874251497,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1751,
+      "step": 2651
+    },
+    {
+      "epoch": 3.970059880239521,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1692,
+      "step": 2652
+    },
+    {
+      "epoch": 3.9715568862275448,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1743,
+      "step": 2653
+    },
+    {
+      "epoch": 3.9730538922155687,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1795,
+      "step": 2654
+    },
+    {
+      "epoch": 3.974550898203593,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1774,
+      "step": 2655
+    },
+    {
+      "epoch": 3.976047904191617,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1705,
+      "step": 2656
+    },
+    {
+      "epoch": 3.977544910179641,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1787,
+      "step": 2657
+    },
+    {
+      "epoch": 3.979041916167665,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1726,
+      "step": 2658
+    },
+    {
+      "epoch": 3.980538922155689,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.165,
+      "step": 2659
+    },
+    {
+      "epoch": 3.9820359281437128,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1711,
+      "step": 2660
+    },
+    {
+      "epoch": 3.9835329341317367,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1704,
+      "step": 2661
+    },
+    {
+      "epoch": 3.9850299401197606,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.17,
+      "step": 2662
+    },
+    {
+      "epoch": 3.9865269461077846,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1733,
+      "step": 2663
+    },
+    {
+      "epoch": 3.9880239520958085,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.169,
+      "step": 2664
+    },
+    {
+      "epoch": 3.9895209580838324,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1696,
+      "step": 2665
+    },
+    {
+      "epoch": 3.9910179640718564,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.18,
+      "step": 2666
+    },
+    {
+      "epoch": 3.9925149700598803,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1663,
+      "step": 2667
+    },
+    {
+      "epoch": 3.9940119760479043,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1693,
+      "step": 2668
+    },
+    {
+      "epoch": 3.995508982035928,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1683,
+      "step": 2669
+    },
+    {
+      "epoch": 3.997005988023952,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1745,
+      "step": 2670
+    },
+    {
+      "epoch": 3.998502994011976,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1649,
+      "step": 2671
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1702,
+      "step": 2672
+    },
+    {
+      "epoch": 4.001497005988024,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1743,
+      "step": 2673
+    },
+    {
+      "epoch": 4.002994011976048,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1709,
+      "step": 2674
+    },
+    {
+      "epoch": 4.004491017964072,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1735,
+      "step": 2675
+    },
+    {
+      "epoch": 4.005988023952096,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1728,
+      "step": 2676
+    },
+    {
+      "epoch": 4.00748502994012,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1692,
+      "step": 2677
+    },
+    {
+      "epoch": 4.008982035928144,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.171,
+      "step": 2678
+    },
+    {
+      "epoch": 4.0104790419161676,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1659,
+      "step": 2679
+    },
+    {
+      "epoch": 4.0119760479041915,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1722,
+      "step": 2680
+    },
+    {
+      "epoch": 4.013473053892215,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1771,
+      "step": 2681
+    },
+    {
+      "epoch": 4.014970059880239,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1743,
+      "step": 2682
+    },
+    {
+      "epoch": 4.016467065868263,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.176,
+      "step": 2683
+    },
+    {
+      "epoch": 4.017964071856287,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1657,
+      "step": 2684
+    },
+    {
+      "epoch": 4.019461077844311,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1695,
+      "step": 2685
+    },
+    {
+      "epoch": 4.020958083832335,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1688,
+      "step": 2686
+    },
+    {
+      "epoch": 4.022455089820359,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1737,
+      "step": 2687
+    },
+    {
+      "epoch": 4.023952095808383,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1599,
+      "step": 2688
+    },
+    {
+      "epoch": 4.025449101796407,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.17,
+      "step": 2689
+    },
+    {
+      "epoch": 4.026946107784431,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1746,
+      "step": 2690
+    },
+    {
+      "epoch": 4.028443113772455,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1708,
+      "step": 2691
+    },
+    {
+      "epoch": 4.029940119760479,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1705,
+      "step": 2692
+    },
+    {
+      "epoch": 4.031437125748503,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1654,
+      "step": 2693
+    },
+    {
+      "epoch": 4.032934131736527,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1732,
+      "step": 2694
+    },
+    {
+      "epoch": 4.0344311377245505,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1728,
+      "step": 2695
+    },
+    {
+      "epoch": 4.0359281437125745,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1683,
+      "step": 2696
+    },
+    {
+      "epoch": 4.037425149700598,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1724,
+      "step": 2697
+    },
+    {
+      "epoch": 4.038922155688622,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1634,
+      "step": 2698
+    },
+    {
+      "epoch": 4.040419161676646,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1735,
+      "step": 2699
+    },
+    {
+      "epoch": 4.04191616766467,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1717,
+      "step": 2700
+    },
+    {
+      "epoch": 4.043413173652695,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1666,
+      "step": 2701
+    },
+    {
+      "epoch": 4.044910179640719,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1659,
+      "step": 2702
+    },
+    {
+      "epoch": 4.046407185628743,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1679,
+      "step": 2703
+    },
+    {
+      "epoch": 4.047904191616767,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1803,
+      "step": 2704
+    },
+    {
+      "epoch": 4.049401197604791,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1744,
+      "step": 2705
+    },
+    {
+      "epoch": 4.050898203592815,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.172,
+      "step": 2706
+    },
+    {
+      "epoch": 4.052395209580839,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1713,
+      "step": 2707
+    },
+    {
+      "epoch": 4.053892215568863,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1728,
+      "step": 2708
+    },
+    {
+      "epoch": 4.0553892215568865,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1743,
+      "step": 2709
+    },
+    {
+      "epoch": 4.0568862275449105,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1618,
+      "step": 2710
+    },
+    {
+      "epoch": 4.058383233532934,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1694,
+      "step": 2711
+    },
+    {
+      "epoch": 4.059880239520958,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1634,
+      "step": 2712
+    },
+    {
+      "epoch": 4.061377245508982,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1679,
+      "step": 2713
+    },
+    {
+      "epoch": 4.062874251497006,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1737,
+      "step": 2714
+    },
+    {
+      "epoch": 4.06437125748503,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1684,
+      "step": 2715
+    },
+    {
+      "epoch": 4.065868263473054,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1723,
+      "step": 2716
+    },
+    {
+      "epoch": 4.067365269461078,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.173,
+      "step": 2717
+    },
+    {
+      "epoch": 4.068862275449102,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1679,
+      "step": 2718
+    },
+    {
+      "epoch": 4.070359281437126,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1663,
+      "step": 2719
+    },
+    {
+      "epoch": 4.07185628742515,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1788,
+      "step": 2720
+    },
+    {
+      "epoch": 4.073353293413174,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1678,
+      "step": 2721
+    },
+    {
+      "epoch": 4.074850299401198,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1627,
+      "step": 2722
+    },
+    {
+      "epoch": 4.076347305389222,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1725,
+      "step": 2723
+    },
+    {
+      "epoch": 4.077844311377246,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1735,
+      "step": 2724
+    },
+    {
+      "epoch": 4.0793413173652695,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1627,
+      "step": 2725
+    },
+    {
+      "epoch": 4.080838323353293,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1658,
+      "step": 2726
+    },
+    {
+      "epoch": 4.082335329341317,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1705,
+      "step": 2727
+    },
+    {
+      "epoch": 4.083832335329341,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1732,
+      "step": 2728
+    },
+    {
+      "epoch": 4.085329341317365,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1729,
+      "step": 2729
+    },
+    {
+      "epoch": 4.086826347305389,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1642,
+      "step": 2730
+    },
+    {
+      "epoch": 4.088323353293413,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1716,
+      "step": 2731
+    },
+    {
+      "epoch": 4.089820359281437,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1649,
+      "step": 2732
+    },
+    {
+      "epoch": 4.091317365269461,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1698,
+      "step": 2733
+    },
+    {
+      "epoch": 4.092814371257485,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.17,
+      "step": 2734
+    },
+    {
+      "epoch": 4.094311377245509,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1727,
+      "step": 2735
+    },
+    {
+      "epoch": 4.095808383233533,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0008,
+      "loss": 1.17,
+      "step": 2736
+    },
+    {
+      "epoch": 4.097305389221557,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.164,
+      "step": 2737
+    },
+    {
+      "epoch": 4.098802395209581,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1721,
+      "step": 2738
+    },
+    {
+      "epoch": 4.100299401197605,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1679,
+      "step": 2739
+    },
+    {
+      "epoch": 4.1017964071856285,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1601,
+      "step": 2740
+    },
+    {
+      "epoch": 4.1032934131736525,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1654,
+      "step": 2741
+    },
+    {
+      "epoch": 4.104790419161676,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1728,
+      "step": 2742
+    },
+    {
+      "epoch": 4.1062874251497,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1683,
+      "step": 2743
+    },
+    {
+      "epoch": 4.107784431137724,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.169,
+      "step": 2744
+    },
+    {
+      "epoch": 4.109281437125748,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1686,
+      "step": 2745
+    },
+    {
+      "epoch": 4.110778443113772,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1742,
+      "step": 2746
+    },
+    {
+      "epoch": 4.112275449101796,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1683,
+      "step": 2747
+    },
+    {
+      "epoch": 4.11377245508982,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1677,
+      "step": 2748
+    },
+    {
+      "epoch": 4.115269461077844,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1623,
+      "step": 2749
+    },
+    {
+      "epoch": 4.116766467065868,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1598,
+      "step": 2750
+    },
+    {
+      "epoch": 4.118263473053892,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1741,
+      "step": 2751
+    },
+    {
+      "epoch": 4.119760479041916,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1774,
+      "step": 2752
+    },
+    {
+      "epoch": 4.12125748502994,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1661,
+      "step": 2753
+    },
+    {
+      "epoch": 4.122754491017964,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1627,
+      "step": 2754
+    },
+    {
+      "epoch": 4.124251497005988,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1728,
+      "step": 2755
+    },
+    {
+      "epoch": 4.125748502994012,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1683,
+      "step": 2756
+    },
+    {
+      "epoch": 4.127245508982036,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.168,
+      "step": 2757
+    },
+    {
+      "epoch": 4.12874251497006,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1746,
+      "step": 2758
+    },
+    {
+      "epoch": 4.130239520958084,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1751,
+      "step": 2759
+    },
+    {
+      "epoch": 4.131736526946108,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1747,
+      "step": 2760
+    },
+    {
+      "epoch": 4.133233532934132,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.162,
+      "step": 2761
+    },
+    {
+      "epoch": 4.134730538922156,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1757,
+      "step": 2762
+    },
+    {
+      "epoch": 4.13622754491018,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1781,
+      "step": 2763
+    },
+    {
+      "epoch": 4.137724550898204,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1706,
+      "step": 2764
+    },
+    {
+      "epoch": 4.139221556886228,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1634,
+      "step": 2765
+    },
+    {
+      "epoch": 4.140718562874252,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1711,
+      "step": 2766
+    },
+    {
+      "epoch": 4.142215568862276,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1717,
+      "step": 2767
+    },
+    {
+      "epoch": 4.1437125748503,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0008,
+      "loss": 1.175,
+      "step": 2768
+    },
+    {
+      "epoch": 4.145209580838324,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0008,
+      "loss": 1.172,
+      "step": 2769
+    },
+    {
+      "epoch": 4.1467065868263475,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.169,
+      "step": 2770
+    },
+    {
+      "epoch": 4.1482035928143715,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1679,
+      "step": 2771
+    },
+    {
+      "epoch": 4.149700598802395,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1721,
+      "step": 2772
+    },
+    {
+      "epoch": 4.151197604790419,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1694,
+      "step": 2773
+    },
+    {
+      "epoch": 4.152694610778443,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1631,
+      "step": 2774
+    },
+    {
+      "epoch": 4.154191616766467,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1742,
+      "step": 2775
+    },
+    {
+      "epoch": 4.155688622754491,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1761,
+      "step": 2776
+    },
+    {
+      "epoch": 4.157185628742515,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1733,
+      "step": 2777
+    },
+    {
+      "epoch": 4.158682634730539,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1678,
+      "step": 2778
+    },
+    {
+      "epoch": 4.160179640718563,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1673,
+      "step": 2779
+    },
+    {
+      "epoch": 4.161676646706587,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1646,
+      "step": 2780
+    },
+    {
+      "epoch": 4.163173652694611,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1773,
+      "step": 2781
+    },
+    {
+      "epoch": 4.164670658682635,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1657,
+      "step": 2782
+    },
+    {
+      "epoch": 4.166167664670659,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.166,
+      "step": 2783
+    },
+    {
+      "epoch": 4.167664670658683,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1711,
+      "step": 2784
+    },
+    {
+      "epoch": 4.169161676646707,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.17,
+      "step": 2785
+    },
+    {
+      "epoch": 4.1706586826347305,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.167,
+      "step": 2786
+    },
+    {
+      "epoch": 4.172155688622754,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1712,
+      "step": 2787
+    },
+    {
+      "epoch": 4.173652694610778,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1751,
+      "step": 2788
+    },
+    {
+      "epoch": 4.175149700598802,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1672,
+      "step": 2789
+    },
+    {
+      "epoch": 4.176646706586826,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1659,
+      "step": 2790
+    },
+    {
+      "epoch": 4.17814371257485,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1686,
+      "step": 2791
+    },
+    {
+      "epoch": 4.179640718562874,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1619,
+      "step": 2792
+    },
+    {
+      "epoch": 4.181137724550898,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0008,
+      "loss": 1.1705,
+      "step": 2793
+    },
+    {
+      "epoch": 4.182634730538922,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1692,
+      "step": 2794
+    },
+    {
+      "epoch": 4.184131736526946,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1704,
+      "step": 2795
+    },
+    {
+      "epoch": 4.18562874251497,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1636,
+      "step": 2796
+    },
+    {
+      "epoch": 4.187125748502994,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.167,
+      "step": 2797
+    },
+    {
+      "epoch": 4.188622754491018,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1683,
+      "step": 2798
+    },
+    {
+      "epoch": 4.190119760479042,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1659,
+      "step": 2799
+    },
+    {
+      "epoch": 4.191616766467066,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1709,
+      "step": 2800
+    },
+    {
+      "epoch": 4.1931137724550895,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1673,
+      "step": 2801
+    },
+    {
+      "epoch": 4.1946107784431135,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1694,
+      "step": 2802
+    },
+    {
+      "epoch": 4.196107784431137,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1635,
+      "step": 2803
+    },
+    {
+      "epoch": 4.197604790419161,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1705,
+      "step": 2804
+    },
+    {
+      "epoch": 4.199101796407185,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1666,
+      "step": 2805
+    },
+    {
+      "epoch": 4.200598802395209,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1708,
+      "step": 2806
+    },
+    {
+      "epoch": 4.202095808383233,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1731,
+      "step": 2807
+    },
+    {
+      "epoch": 4.203592814371257,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1685,
+      "step": 2808
+    },
+    {
+      "epoch": 4.205089820359281,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1714,
+      "step": 2809
+    },
+    {
+      "epoch": 4.206586826347305,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1667,
+      "step": 2810
+    },
+    {
+      "epoch": 4.20808383233533,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1683,
+      "step": 2811
+    },
+    {
+      "epoch": 4.209580838323353,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1689,
+      "step": 2812
+    },
+    {
+      "epoch": 4.211077844311378,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1737,
+      "step": 2813
+    },
+    {
+      "epoch": 4.212574850299402,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1765,
+      "step": 2814
+    },
+    {
+      "epoch": 4.2140718562874255,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.163,
+      "step": 2815
+    },
+    {
+      "epoch": 4.2155688622754495,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1668,
+      "step": 2816
+    },
+    {
+      "epoch": 4.217065868263473,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1711,
+      "step": 2817
+    },
+    {
+      "epoch": 4.218562874251497,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1665,
+      "step": 2818
+    },
+    {
+      "epoch": 4.220059880239521,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1682,
+      "step": 2819
+    },
+    {
+      "epoch": 4.221556886227545,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1746,
+      "step": 2820
+    },
+    {
+      "epoch": 4.223053892215569,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1689,
+      "step": 2821
+    },
+    {
+      "epoch": 4.224550898203593,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1718,
+      "step": 2822
+    },
+    {
+      "epoch": 4.226047904191617,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1764,
+      "step": 2823
+    },
+    {
+      "epoch": 4.227544910179641,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1692,
+      "step": 2824
+    },
+    {
+      "epoch": 4.229041916167665,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1677,
+      "step": 2825
+    },
+    {
+      "epoch": 4.230538922155689,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1702,
+      "step": 2826
+    },
+    {
+      "epoch": 4.232035928143713,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1735,
+      "step": 2827
+    },
+    {
+      "epoch": 4.233532934131737,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1707,
+      "step": 2828
+    },
+    {
+      "epoch": 4.235029940119761,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1731,
+      "step": 2829
+    },
+    {
+      "epoch": 4.236526946107785,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1787,
+      "step": 2830
+    },
+    {
+      "epoch": 4.2380239520958085,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1683,
+      "step": 2831
+    },
+    {
+      "epoch": 4.2395209580838324,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1771,
+      "step": 2832
+    },
+    {
+      "epoch": 4.241017964071856,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1733,
+      "step": 2833
+    },
+    {
+      "epoch": 4.24251497005988,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1766,
+      "step": 2834
+    },
+    {
+      "epoch": 4.244011976047904,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1673,
+      "step": 2835
+    },
+    {
+      "epoch": 4.245508982035928,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1706,
+      "step": 2836
+    },
+    {
+      "epoch": 4.247005988023952,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1692,
+      "step": 2837
+    },
+    {
+      "epoch": 4.248502994011976,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1663,
+      "step": 2838
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1647,
+      "step": 2839
+    },
+    {
+      "epoch": 4.251497005988024,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1691,
+      "step": 2840
+    },
+    {
+      "epoch": 4.252994011976048,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1724,
+      "step": 2841
+    },
+    {
+      "epoch": 4.254491017964072,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1672,
+      "step": 2842
+    },
+    {
+      "epoch": 4.255988023952096,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1668,
+      "step": 2843
+    },
+    {
+      "epoch": 4.25748502994012,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1674,
+      "step": 2844
+    },
+    {
+      "epoch": 4.258982035928144,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1707,
+      "step": 2845
+    },
+    {
+      "epoch": 4.2604790419161676,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1698,
+      "step": 2846
+    },
+    {
+      "epoch": 4.2619760479041915,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1739,
+      "step": 2847
+    },
+    {
+      "epoch": 4.263473053892215,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1635,
+      "step": 2848
+    },
+    {
+      "epoch": 4.264970059880239,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1747,
+      "step": 2849
+    },
+    {
+      "epoch": 4.266467065868263,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1627,
+      "step": 2850
+    },
+    {
+      "epoch": 4.267964071856287,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.171,
+      "step": 2851
+    },
+    {
+      "epoch": 4.269461077844311,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1747,
+      "step": 2852
+    },
+    {
+      "epoch": 4.270958083832335,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1705,
+      "step": 2853
+    },
+    {
+      "epoch": 4.272455089820359,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1663,
+      "step": 2854
+    },
+    {
+      "epoch": 4.273952095808383,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1695,
+      "step": 2855
+    },
+    {
+      "epoch": 4.275449101796407,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1771,
+      "step": 2856
+    },
+    {
+      "epoch": 4.276946107784431,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1617,
+      "step": 2857
+    },
+    {
+      "epoch": 4.278443113772455,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1652,
+      "step": 2858
+    },
+    {
+      "epoch": 4.279940119760479,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1666,
+      "step": 2859
+    },
+    {
+      "epoch": 4.281437125748503,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1716,
+      "step": 2860
+    },
+    {
+      "epoch": 4.282934131736527,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1644,
+      "step": 2861
+    },
+    {
+      "epoch": 4.2844311377245505,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1652,
+      "step": 2862
+    },
+    {
+      "epoch": 4.2859281437125745,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1697,
+      "step": 2863
+    },
+    {
+      "epoch": 4.287425149700598,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1657,
+      "step": 2864
+    },
+    {
+      "epoch": 4.288922155688622,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1701,
+      "step": 2865
+    },
+    {
+      "epoch": 4.290419161676647,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1689,
+      "step": 2866
+    },
+    {
+      "epoch": 4.29191616766467,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1747,
+      "step": 2867
+    },
+    {
+      "epoch": 4.293413173652695,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1674,
+      "step": 2868
+    },
+    {
+      "epoch": 4.294910179640719,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1676,
+      "step": 2869
+    },
+    {
+      "epoch": 4.296407185628743,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1756,
+      "step": 2870
+    },
+    {
+      "epoch": 4.297904191616767,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1726,
+      "step": 2871
+    },
+    {
+      "epoch": 4.299401197604791,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.174,
+      "step": 2872
+    },
+    {
+      "epoch": 4.300898203592815,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1635,
+      "step": 2873
+    },
+    {
+      "epoch": 4.302395209580839,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1755,
+      "step": 2874
+    },
+    {
+      "epoch": 4.303892215568863,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.166,
+      "step": 2875
+    },
+    {
+      "epoch": 4.3053892215568865,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1719,
+      "step": 2876
+    },
+    {
+      "epoch": 4.3068862275449105,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1654,
+      "step": 2877
+    },
+    {
+      "epoch": 4.308383233532934,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.162,
+      "step": 2878
+    },
+    {
+      "epoch": 4.309880239520958,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1643,
+      "step": 2879
+    },
+    {
+      "epoch": 4.311377245508982,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1743,
+      "step": 2880
+    },
+    {
+      "epoch": 4.312874251497006,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1688,
+      "step": 2881
+    },
+    {
+      "epoch": 4.31437125748503,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1735,
+      "step": 2882
+    },
+    {
+      "epoch": 4.315868263473054,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1629,
+      "step": 2883
+    },
+    {
+      "epoch": 4.317365269461078,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1697,
+      "step": 2884
+    },
+    {
+      "epoch": 4.318862275449102,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1617,
+      "step": 2885
+    },
+    {
+      "epoch": 4.320359281437126,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1727,
+      "step": 2886
+    },
+    {
+      "epoch": 4.32185628742515,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1721,
+      "step": 2887
+    },
+    {
+      "epoch": 4.323353293413174,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1629,
+      "step": 2888
+    },
+    {
+      "epoch": 4.324850299401198,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1695,
+      "step": 2889
+    },
+    {
+      "epoch": 4.326347305389222,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.162,
+      "step": 2890
+    },
+    {
+      "epoch": 4.327844311377246,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1658,
+      "step": 2891
+    },
+    {
+      "epoch": 4.3293413173652695,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1608,
+      "step": 2892
+    },
+    {
+      "epoch": 4.330838323353293,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1706,
+      "step": 2893
+    },
+    {
+      "epoch": 4.332335329341317,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1666,
+      "step": 2894
+    },
+    {
+      "epoch": 4.333832335329341,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1694,
+      "step": 2895
+    },
+    {
+      "epoch": 4.335329341317365,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1771,
+      "step": 2896
+    },
+    {
+      "epoch": 4.336826347305389,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1733,
+      "step": 2897
+    },
+    {
+      "epoch": 4.338323353293413,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1666,
+      "step": 2898
+    },
+    {
+      "epoch": 4.339820359281437,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1686,
+      "step": 2899
+    },
+    {
+      "epoch": 4.341317365269461,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1629,
+      "step": 2900
+    },
+    {
+      "epoch": 4.342814371257485,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1683,
+      "step": 2901
+    },
+    {
+      "epoch": 4.344311377245509,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1612,
+      "step": 2902
+    },
+    {
+      "epoch": 4.345808383233533,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1692,
+      "step": 2903
+    },
+    {
+      "epoch": 4.347305389221557,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1706,
+      "step": 2904
+    },
+    {
+      "epoch": 4.348802395209581,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1669,
+      "step": 2905
+    },
+    {
+      "epoch": 4.350299401197605,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1667,
+      "step": 2906
+    },
+    {
+      "epoch": 4.3517964071856285,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1667,
+      "step": 2907
+    },
+    {
+      "epoch": 4.3532934131736525,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1596,
+      "step": 2908
+    },
+    {
+      "epoch": 4.354790419161676,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1713,
+      "step": 2909
+    },
+    {
+      "epoch": 4.3562874251497,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1635,
+      "step": 2910
+    },
+    {
+      "epoch": 4.357784431137724,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1675,
+      "step": 2911
+    },
+    {
+      "epoch": 4.359281437125748,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.17,
+      "step": 2912
+    },
+    {
+      "epoch": 4.360778443113772,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1723,
+      "step": 2913
+    },
+    {
+      "epoch": 4.362275449101796,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1695,
+      "step": 2914
+    },
+    {
+      "epoch": 4.36377245508982,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.169,
+      "step": 2915
+    },
+    {
+      "epoch": 4.365269461077844,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1658,
+      "step": 2916
+    },
+    {
+      "epoch": 4.366766467065868,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1719,
+      "step": 2917
+    },
+    {
+      "epoch": 4.368263473053892,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0008,
+      "loss": 1.176,
+      "step": 2918
+    },
+    {
+      "epoch": 4.369760479041916,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.168,
+      "step": 2919
+    },
+    {
+      "epoch": 4.37125748502994,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1657,
+      "step": 2920
+    },
+    {
+      "epoch": 4.3727544910179645,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1696,
+      "step": 2921
+    },
+    {
+      "epoch": 4.374251497005988,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1687,
+      "step": 2922
+    },
+    {
+      "epoch": 4.375748502994012,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1703,
+      "step": 2923
+    },
+    {
+      "epoch": 4.3772455089820355,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1687,
+      "step": 2924
+    },
+    {
+      "epoch": 4.37874251497006,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1685,
+      "step": 2925
+    },
+    {
+      "epoch": 4.380239520958084,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1703,
+      "step": 2926
+    },
+    {
+      "epoch": 4.381736526946108,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1663,
+      "step": 2927
+    },
+    {
+      "epoch": 4.383233532934132,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1633,
+      "step": 2928
+    },
+    {
+      "epoch": 4.384730538922156,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1702,
+      "step": 2929
+    },
+    {
+      "epoch": 4.38622754491018,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1697,
+      "step": 2930
+    },
+    {
+      "epoch": 4.387724550898204,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1746,
+      "step": 2931
+    },
+    {
+      "epoch": 4.389221556886228,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1702,
+      "step": 2932
+    },
+    {
+      "epoch": 4.390718562874252,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1657,
+      "step": 2933
+    },
+    {
+      "epoch": 4.392215568862276,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1638,
+      "step": 2934
+    },
+    {
+      "epoch": 4.3937125748503,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1696,
+      "step": 2935
+    },
+    {
+      "epoch": 4.395209580838324,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1714,
+      "step": 2936
+    },
+    {
+      "epoch": 4.3967065868263475,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1649,
+      "step": 2937
+    },
+    {
+      "epoch": 4.3982035928143715,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.169,
+      "step": 2938
+    },
+    {
+      "epoch": 4.399700598802395,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1694,
+      "step": 2939
+    },
+    {
+      "epoch": 4.401197604790419,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1719,
+      "step": 2940
+    },
+    {
+      "epoch": 4.402694610778443,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1716,
+      "step": 2941
+    },
+    {
+      "epoch": 4.404191616766467,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1696,
+      "step": 2942
+    },
+    {
+      "epoch": 4.405688622754491,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1636,
+      "step": 2943
+    },
+    {
+      "epoch": 4.407185628742515,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1643,
+      "step": 2944
+    },
+    {
+      "epoch": 4.408682634730539,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1651,
+      "step": 2945
+    },
+    {
+      "epoch": 4.410179640718563,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1728,
+      "step": 2946
+    },
+    {
+      "epoch": 4.411676646706587,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1691,
+      "step": 2947
+    },
+    {
+      "epoch": 4.413173652694611,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1686,
+      "step": 2948
+    },
+    {
+      "epoch": 4.414670658682635,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1662,
+      "step": 2949
+    },
+    {
+      "epoch": 4.416167664670659,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1645,
+      "step": 2950
+    },
+    {
+      "epoch": 4.417664670658683,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1752,
+      "step": 2951
+    },
+    {
+      "epoch": 4.419161676646707,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1684,
+      "step": 2952
+    },
+    {
+      "epoch": 4.4206586826347305,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1658,
+      "step": 2953
+    },
+    {
+      "epoch": 4.422155688622754,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1729,
+      "step": 2954
+    },
+    {
+      "epoch": 4.423652694610778,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1759,
+      "step": 2955
+    },
+    {
+      "epoch": 4.425149700598802,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1629,
+      "step": 2956
+    },
+    {
+      "epoch": 4.426646706586826,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1652,
+      "step": 2957
+    },
+    {
+      "epoch": 4.42814371257485,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1695,
+      "step": 2958
+    },
+    {
+      "epoch": 4.429640718562874,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1694,
+      "step": 2959
+    },
+    {
+      "epoch": 4.431137724550898,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1651,
+      "step": 2960
+    },
+    {
+      "epoch": 4.432634730538922,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1707,
+      "step": 2961
+    },
+    {
+      "epoch": 4.434131736526946,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1767,
+      "step": 2962
+    },
+    {
+      "epoch": 4.43562874251497,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1626,
+      "step": 2963
+    },
+    {
+      "epoch": 4.437125748502994,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1639,
+      "step": 2964
+    },
+    {
+      "epoch": 4.438622754491018,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1654,
+      "step": 2965
+    },
+    {
+      "epoch": 4.440119760479042,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1645,
+      "step": 2966
+    },
+    {
+      "epoch": 4.441616766467066,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1643,
+      "step": 2967
+    },
+    {
+      "epoch": 4.4431137724550895,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1705,
+      "step": 2968
+    },
+    {
+      "epoch": 4.4446107784431135,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1651,
+      "step": 2969
+    },
+    {
+      "epoch": 4.446107784431137,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1666,
+      "step": 2970
+    },
+    {
+      "epoch": 4.447604790419161,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1667,
+      "step": 2971
+    },
+    {
+      "epoch": 4.449101796407185,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1727,
+      "step": 2972
+    },
+    {
+      "epoch": 4.450598802395209,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.173,
+      "step": 2973
+    },
+    {
+      "epoch": 4.452095808383233,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.172,
+      "step": 2974
+    },
+    {
+      "epoch": 4.453592814371257,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1671,
+      "step": 2975
+    },
+    {
+      "epoch": 4.455089820359281,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1698,
+      "step": 2976
+    },
+    {
+      "epoch": 4.456586826347305,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1648,
+      "step": 2977
+    },
+    {
+      "epoch": 4.45808383233533,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.168,
+      "step": 2978
+    },
+    {
+      "epoch": 4.459580838323353,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1632,
+      "step": 2979
+    },
+    {
+      "epoch": 4.461077844311378,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1646,
+      "step": 2980
+    },
+    {
+      "epoch": 4.462574850299402,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1554,
+      "step": 2981
+    },
+    {
+      "epoch": 4.4640718562874255,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.164,
+      "step": 2982
+    },
+    {
+      "epoch": 4.4655688622754495,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1702,
+      "step": 2983
+    },
+    {
+      "epoch": 4.467065868263473,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1629,
+      "step": 2984
+    },
+    {
+      "epoch": 4.468562874251497,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1664,
+      "step": 2985
+    },
+    {
+      "epoch": 4.470059880239521,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.167,
+      "step": 2986
+    },
+    {
+      "epoch": 4.471556886227545,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1611,
+      "step": 2987
+    },
+    {
+      "epoch": 4.473053892215569,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1733,
+      "step": 2988
+    },
+    {
+      "epoch": 4.474550898203593,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1699,
+      "step": 2989
+    },
+    {
+      "epoch": 4.476047904191617,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1655,
+      "step": 2990
+    },
+    {
+      "epoch": 4.477544910179641,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1698,
+      "step": 2991
+    },
+    {
+      "epoch": 4.479041916167665,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1669,
+      "step": 2992
+    },
+    {
+      "epoch": 4.480538922155689,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1706,
+      "step": 2993
+    },
+    {
+      "epoch": 4.482035928143713,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1751,
+      "step": 2994
+    },
+    {
+      "epoch": 4.483532934131737,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1688,
+      "step": 2995
+    },
+    {
+      "epoch": 4.485029940119761,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1696,
+      "step": 2996
+    },
+    {
+      "epoch": 4.486526946107785,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1613,
+      "step": 2997
+    },
+    {
+      "epoch": 4.4880239520958085,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1698,
+      "step": 2998
+    },
+    {
+      "epoch": 4.4895209580838324,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1695,
+      "step": 2999
+    },
+    {
+      "epoch": 4.491017964071856,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1731,
+      "step": 3000
+    },
+    {
+      "epoch": 4.49251497005988,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1676,
+      "step": 3001
+    },
+    {
+      "epoch": 4.494011976047904,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1759,
+      "step": 3002
+    },
+    {
+      "epoch": 4.495508982035928,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1696,
+      "step": 3003
+    },
+    {
+      "epoch": 4.497005988023952,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1733,
+      "step": 3004
+    },
+    {
+      "epoch": 4.498502994011976,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1715,
+      "step": 3005
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1671,
+      "step": 3006
+    },
+    {
+      "epoch": 4.501497005988024,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1656,
+      "step": 3007
+    },
+    {
+      "epoch": 4.502994011976048,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1629,
+      "step": 3008
+    },
+    {
+      "epoch": 4.504491017964072,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1709,
+      "step": 3009
+    },
+    {
+      "epoch": 4.505988023952096,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1643,
+      "step": 3010
+    },
+    {
+      "epoch": 4.50748502994012,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1655,
+      "step": 3011
+    },
+    {
+      "epoch": 4.508982035928144,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1735,
+      "step": 3012
+    },
+    {
+      "epoch": 4.5104790419161676,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1668,
+      "step": 3013
+    },
+    {
+      "epoch": 4.5119760479041915,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1678,
+      "step": 3014
+    },
+    {
+      "epoch": 4.513473053892215,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1619,
+      "step": 3015
+    },
+    {
+      "epoch": 4.514970059880239,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.163,
+      "step": 3016
+    },
+    {
+      "epoch": 4.516467065868263,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1639,
+      "step": 3017
+    },
+    {
+      "epoch": 4.517964071856287,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1697,
+      "step": 3018
+    },
+    {
+      "epoch": 4.519461077844311,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1655,
+      "step": 3019
+    },
+    {
+      "epoch": 4.520958083832335,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1649,
+      "step": 3020
+    },
+    {
+      "epoch": 4.522455089820359,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.175,
+      "step": 3021
+    },
+    {
+      "epoch": 4.523952095808383,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1683,
+      "step": 3022
+    },
+    {
+      "epoch": 4.525449101796407,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1725,
+      "step": 3023
+    },
+    {
+      "epoch": 4.526946107784431,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1617,
+      "step": 3024
+    },
+    {
+      "epoch": 4.528443113772455,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1688,
+      "step": 3025
+    },
+    {
+      "epoch": 4.529940119760479,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.166,
+      "step": 3026
+    },
+    {
+      "epoch": 4.531437125748503,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1679,
+      "step": 3027
+    },
+    {
+      "epoch": 4.532934131736527,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.173,
+      "step": 3028
+    },
+    {
+      "epoch": 4.5344311377245505,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1705,
+      "step": 3029
+    },
+    {
+      "epoch": 4.5359281437125745,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.167,
+      "step": 3030
+    },
+    {
+      "epoch": 4.537425149700599,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1682,
+      "step": 3031
+    },
+    {
+      "epoch": 4.538922155688622,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1636,
+      "step": 3032
+    },
+    {
+      "epoch": 4.540419161676647,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1608,
+      "step": 3033
+    },
+    {
+      "epoch": 4.54191616766467,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1652,
+      "step": 3034
+    },
+    {
+      "epoch": 4.543413173652695,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1697,
+      "step": 3035
+    },
+    {
+      "epoch": 4.544910179640718,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1692,
+      "step": 3036
+    },
+    {
+      "epoch": 4.546407185628743,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.169,
+      "step": 3037
+    },
+    {
+      "epoch": 4.547904191616767,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1739,
+      "step": 3038
+    },
+    {
+      "epoch": 4.549401197604791,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1746,
+      "step": 3039
+    },
+    {
+      "epoch": 4.550898203592815,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1607,
+      "step": 3040
+    },
+    {
+      "epoch": 4.552395209580839,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1612,
+      "step": 3041
+    },
+    {
+      "epoch": 4.553892215568863,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1604,
+      "step": 3042
+    },
+    {
+      "epoch": 4.5553892215568865,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1662,
+      "step": 3043
+    },
+    {
+      "epoch": 4.5568862275449105,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1652,
+      "step": 3044
+    },
+    {
+      "epoch": 4.558383233532934,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1689,
+      "step": 3045
+    },
+    {
+      "epoch": 4.559880239520958,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.174,
+      "step": 3046
+    },
+    {
+      "epoch": 4.561377245508982,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1716,
+      "step": 3047
+    },
+    {
+      "epoch": 4.562874251497006,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1642,
+      "step": 3048
+    },
+    {
+      "epoch": 4.56437125748503,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1681,
+      "step": 3049
+    },
+    {
+      "epoch": 4.565868263473054,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1729,
+      "step": 3050
+    },
+    {
+      "epoch": 4.567365269461078,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1577,
+      "step": 3051
+    },
+    {
+      "epoch": 4.568862275449102,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1674,
+      "step": 3052
+    },
+    {
+      "epoch": 4.570359281437126,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1718,
+      "step": 3053
+    },
+    {
+      "epoch": 4.57185628742515,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1728,
+      "step": 3054
+    },
+    {
+      "epoch": 4.573353293413174,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1677,
+      "step": 3055
+    },
+    {
+      "epoch": 4.574850299401198,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1677,
+      "step": 3056
+    },
+    {
+      "epoch": 4.576347305389222,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1714,
+      "step": 3057
+    },
+    {
+      "epoch": 4.577844311377246,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.173,
+      "step": 3058
+    },
+    {
+      "epoch": 4.5793413173652695,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1668,
+      "step": 3059
+    },
+    {
+      "epoch": 4.580838323353293,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1657,
+      "step": 3060
+    },
+    {
+      "epoch": 4.582335329341317,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1677,
+      "step": 3061
+    },
+    {
+      "epoch": 4.583832335329341,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1621,
+      "step": 3062
+    },
+    {
+      "epoch": 4.585329341317365,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1739,
+      "step": 3063
+    },
+    {
+      "epoch": 4.586826347305389,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1659,
+      "step": 3064
+    },
+    {
+      "epoch": 4.588323353293413,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1748,
+      "step": 3065
+    },
+    {
+      "epoch": 4.589820359281437,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1651,
+      "step": 3066
+    },
+    {
+      "epoch": 4.591317365269461,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.167,
+      "step": 3067
+    },
+    {
+      "epoch": 4.592814371257485,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1605,
+      "step": 3068
+    },
+    {
+      "epoch": 4.594311377245509,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1631,
+      "step": 3069
+    },
+    {
+      "epoch": 4.595808383233533,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1609,
+      "step": 3070
+    },
+    {
+      "epoch": 4.597305389221557,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1661,
+      "step": 3071
+    },
+    {
+      "epoch": 4.598802395209581,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1749,
+      "step": 3072
+    },
+    {
+      "epoch": 4.600299401197605,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1701,
+      "step": 3073
+    },
+    {
+      "epoch": 4.6017964071856285,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1682,
+      "step": 3074
+    },
+    {
+      "epoch": 4.6032934131736525,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1558,
+      "step": 3075
+    },
+    {
+      "epoch": 4.604790419161676,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1672,
+      "step": 3076
+    },
+    {
+      "epoch": 4.6062874251497,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1664,
+      "step": 3077
+    },
+    {
+      "epoch": 4.607784431137724,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1641,
+      "step": 3078
+    },
+    {
+      "epoch": 4.609281437125748,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1697,
+      "step": 3079
+    },
+    {
+      "epoch": 4.610778443113772,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1636,
+      "step": 3080
+    },
+    {
+      "epoch": 4.612275449101796,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1635,
+      "step": 3081
+    },
+    {
+      "epoch": 4.61377245508982,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1706,
+      "step": 3082
+    },
+    {
+      "epoch": 4.615269461077844,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1737,
+      "step": 3083
+    },
+    {
+      "epoch": 4.616766467065868,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1653,
+      "step": 3084
+    },
+    {
+      "epoch": 4.618263473053892,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1644,
+      "step": 3085
+    },
+    {
+      "epoch": 4.619760479041916,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1644,
+      "step": 3086
+    },
+    {
+      "epoch": 4.62125748502994,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1737,
+      "step": 3087
+    },
+    {
+      "epoch": 4.6227544910179645,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1691,
+      "step": 3088
+    },
+    {
+      "epoch": 4.624251497005988,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1714,
+      "step": 3089
+    },
+    {
+      "epoch": 4.625748502994012,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1664,
+      "step": 3090
+    },
+    {
+      "epoch": 4.6272455089820355,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1725,
+      "step": 3091
+    },
+    {
+      "epoch": 4.62874251497006,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1666,
+      "step": 3092
+    },
+    {
+      "epoch": 4.630239520958084,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1665,
+      "step": 3093
+    },
+    {
+      "epoch": 4.631736526946108,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.171,
+      "step": 3094
+    },
+    {
+      "epoch": 4.633233532934132,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1646,
+      "step": 3095
+    },
+    {
+      "epoch": 4.634730538922156,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1675,
+      "step": 3096
+    },
+    {
+      "epoch": 4.63622754491018,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1668,
+      "step": 3097
+    },
+    {
+      "epoch": 4.637724550898204,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1624,
+      "step": 3098
+    },
+    {
+      "epoch": 4.639221556886228,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1661,
+      "step": 3099
+    },
+    {
+      "epoch": 4.640718562874252,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1693,
+      "step": 3100
+    },
+    {
+      "epoch": 4.642215568862276,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1591,
+      "step": 3101
+    },
+    {
+      "epoch": 4.6437125748503,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1649,
+      "step": 3102
+    },
+    {
+      "epoch": 4.645209580838324,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1666,
+      "step": 3103
+    },
+    {
+      "epoch": 4.6467065868263475,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1674,
+      "step": 3104
+    },
+    {
+      "epoch": 4.6482035928143715,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1692,
+      "step": 3105
+    },
+    {
+      "epoch": 4.649700598802395,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1646,
+      "step": 3106
+    },
+    {
+      "epoch": 4.651197604790419,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.162,
+      "step": 3107
+    },
+    {
+      "epoch": 4.652694610778443,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1639,
+      "step": 3108
+    },
+    {
+      "epoch": 4.654191616766467,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1649,
+      "step": 3109
+    },
+    {
+      "epoch": 4.655688622754491,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1653,
+      "step": 3110
+    },
+    {
+      "epoch": 4.657185628742515,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1706,
+      "step": 3111
+    },
+    {
+      "epoch": 4.658682634730539,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1681,
+      "step": 3112
+    },
+    {
+      "epoch": 4.660179640718563,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1627,
+      "step": 3113
+    },
+    {
+      "epoch": 4.661676646706587,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1665,
+      "step": 3114
+    },
+    {
+      "epoch": 4.663173652694611,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1681,
+      "step": 3115
+    },
+    {
+      "epoch": 4.664670658682635,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.162,
+      "step": 3116
+    },
+    {
+      "epoch": 4.666167664670659,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1684,
+      "step": 3117
+    },
+    {
+      "epoch": 4.667664670658683,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1635,
+      "step": 3118
+    },
+    {
+      "epoch": 4.669161676646707,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1626,
+      "step": 3119
+    },
+    {
+      "epoch": 4.6706586826347305,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1644,
+      "step": 3120
+    },
+    {
+      "epoch": 4.672155688622754,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1724,
+      "step": 3121
+    },
+    {
+      "epoch": 4.673652694610778,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1717,
+      "step": 3122
+    },
+    {
+      "epoch": 4.675149700598802,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1676,
+      "step": 3123
+    },
+    {
+      "epoch": 4.676646706586826,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1738,
+      "step": 3124
+    },
+    {
+      "epoch": 4.67814371257485,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1696,
+      "step": 3125
+    },
+    {
+      "epoch": 4.679640718562874,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.169,
+      "step": 3126
+    },
+    {
+      "epoch": 4.681137724550898,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1619,
+      "step": 3127
+    },
+    {
+      "epoch": 4.682634730538922,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1707,
+      "step": 3128
+    },
+    {
+      "epoch": 4.684131736526946,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1588,
+      "step": 3129
+    },
+    {
+      "epoch": 4.68562874251497,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.168,
+      "step": 3130
+    },
+    {
+      "epoch": 4.687125748502994,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1681,
+      "step": 3131
+    },
+    {
+      "epoch": 4.688622754491018,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1706,
+      "step": 3132
+    },
+    {
+      "epoch": 4.690119760479042,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1674,
+      "step": 3133
+    },
+    {
+      "epoch": 4.691616766467066,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1641,
+      "step": 3134
+    },
+    {
+      "epoch": 4.6931137724550895,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1655,
+      "step": 3135
+    },
+    {
+      "epoch": 4.6946107784431135,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1663,
+      "step": 3136
+    },
+    {
+      "epoch": 4.696107784431137,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1751,
+      "step": 3137
+    },
+    {
+      "epoch": 4.697604790419161,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1675,
+      "step": 3138
+    },
+    {
+      "epoch": 4.699101796407185,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1548,
+      "step": 3139
+    },
+    {
+      "epoch": 4.700598802395209,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1629,
+      "step": 3140
+    },
+    {
+      "epoch": 4.702095808383233,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1635,
+      "step": 3141
+    },
+    {
+      "epoch": 4.703592814371257,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1649,
+      "step": 3142
+    },
+    {
+      "epoch": 4.705089820359282,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1619,
+      "step": 3143
+    },
+    {
+      "epoch": 4.706586826347305,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1636,
+      "step": 3144
+    },
+    {
+      "epoch": 4.70808383233533,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1755,
+      "step": 3145
+    },
+    {
+      "epoch": 4.709580838323353,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1597,
+      "step": 3146
+    },
+    {
+      "epoch": 4.711077844311378,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1711,
+      "step": 3147
+    },
+    {
+      "epoch": 4.712574850299401,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1664,
+      "step": 3148
+    },
+    {
+      "epoch": 4.7140718562874255,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1692,
+      "step": 3149
+    },
+    {
+      "epoch": 4.7155688622754495,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.161,
+      "step": 3150
+    },
+    {
+      "epoch": 4.717065868263473,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1669,
+      "step": 3151
+    },
+    {
+      "epoch": 4.718562874251497,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1654,
+      "step": 3152
+    },
+    {
+      "epoch": 4.720059880239521,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1579,
+      "step": 3153
+    },
+    {
+      "epoch": 4.721556886227545,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1623,
+      "step": 3154
+    },
+    {
+      "epoch": 4.723053892215569,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1585,
+      "step": 3155
+    },
+    {
+      "epoch": 4.724550898203593,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.165,
+      "step": 3156
+    },
+    {
+      "epoch": 4.726047904191617,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1652,
+      "step": 3157
+    },
+    {
+      "epoch": 4.727544910179641,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1679,
+      "step": 3158
+    },
+    {
+      "epoch": 4.729041916167665,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1629,
+      "step": 3159
+    },
+    {
+      "epoch": 4.730538922155689,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1618,
+      "step": 3160
+    },
+    {
+      "epoch": 4.732035928143713,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1634,
+      "step": 3161
+    },
+    {
+      "epoch": 4.733532934131737,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1709,
+      "step": 3162
+    },
+    {
+      "epoch": 4.735029940119761,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1709,
+      "step": 3163
+    },
+    {
+      "epoch": 4.736526946107785,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1647,
+      "step": 3164
+    },
+    {
+      "epoch": 4.7380239520958085,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1702,
+      "step": 3165
+    },
+    {
+      "epoch": 4.7395209580838324,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.168,
+      "step": 3166
+    },
+    {
+      "epoch": 4.741017964071856,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1602,
+      "step": 3167
+    },
+    {
+      "epoch": 4.74251497005988,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1724,
+      "step": 3168
+    },
+    {
+      "epoch": 4.744011976047904,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1743,
+      "step": 3169
+    },
+    {
+      "epoch": 4.745508982035928,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1657,
+      "step": 3170
+    },
+    {
+      "epoch": 4.747005988023952,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1656,
+      "step": 3171
+    },
+    {
+      "epoch": 4.748502994011976,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1647,
+      "step": 3172
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1656,
+      "step": 3173
+    },
+    {
+      "epoch": 4.751497005988024,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1684,
+      "step": 3174
+    },
+    {
+      "epoch": 4.752994011976048,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1633,
+      "step": 3175
+    },
+    {
+      "epoch": 4.754491017964072,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1675,
+      "step": 3176
+    },
+    {
+      "epoch": 4.755988023952096,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.168,
+      "step": 3177
+    },
+    {
+      "epoch": 4.75748502994012,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1658,
+      "step": 3178
+    },
+    {
+      "epoch": 4.758982035928144,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1692,
+      "step": 3179
+    },
+    {
+      "epoch": 4.7604790419161676,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1705,
+      "step": 3180
+    },
+    {
+      "epoch": 4.7619760479041915,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1725,
+      "step": 3181
+    },
+    {
+      "epoch": 4.763473053892215,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1713,
+      "step": 3182
+    },
+    {
+      "epoch": 4.764970059880239,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1662,
+      "step": 3183
+    },
+    {
+      "epoch": 4.766467065868263,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1702,
+      "step": 3184
+    },
+    {
+      "epoch": 4.767964071856287,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1666,
+      "step": 3185
+    },
+    {
+      "epoch": 4.769461077844311,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1624,
+      "step": 3186
+    },
+    {
+      "epoch": 4.770958083832335,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1699,
+      "step": 3187
+    },
+    {
+      "epoch": 4.772455089820359,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1688,
+      "step": 3188
+    },
+    {
+      "epoch": 4.773952095808383,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.164,
+      "step": 3189
+    },
+    {
+      "epoch": 4.775449101796407,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1676,
+      "step": 3190
+    },
+    {
+      "epoch": 4.776946107784431,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1743,
+      "step": 3191
+    },
+    {
+      "epoch": 4.778443113772455,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1705,
+      "step": 3192
+    },
+    {
+      "epoch": 4.779940119760479,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1645,
+      "step": 3193
+    },
+    {
+      "epoch": 4.781437125748503,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1635,
+      "step": 3194
+    },
+    {
+      "epoch": 4.782934131736527,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1635,
+      "step": 3195
+    },
+    {
+      "epoch": 4.7844311377245505,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1697,
+      "step": 3196
+    },
+    {
+      "epoch": 4.7859281437125745,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1662,
+      "step": 3197
+    },
+    {
+      "epoch": 4.787425149700599,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1712,
+      "step": 3198
+    },
+    {
+      "epoch": 4.788922155688622,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1702,
+      "step": 3199
+    },
+    {
+      "epoch": 4.790419161676647,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1625,
+      "step": 3200
+    },
+    {
+      "epoch": 4.79191616766467,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1623,
+      "step": 3201
+    },
+    {
+      "epoch": 4.793413173652695,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1633,
+      "step": 3202
+    },
+    {
+      "epoch": 4.794910179640718,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.159,
+      "step": 3203
+    },
+    {
+      "epoch": 4.796407185628743,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1644,
+      "step": 3204
+    },
+    {
+      "epoch": 4.797904191616767,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1658,
+      "step": 3205
+    },
+    {
+      "epoch": 4.799401197604791,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1697,
+      "step": 3206
+    },
+    {
+      "epoch": 4.800898203592815,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.165,
+      "step": 3207
+    },
+    {
+      "epoch": 4.802395209580839,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1577,
+      "step": 3208
+    },
+    {
+      "epoch": 4.803892215568863,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.167,
+      "step": 3209
+    },
+    {
+      "epoch": 4.8053892215568865,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1696,
+      "step": 3210
+    },
+    {
+      "epoch": 4.8068862275449105,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1668,
+      "step": 3211
+    },
+    {
+      "epoch": 4.808383233532934,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1681,
+      "step": 3212
+    },
+    {
+      "epoch": 4.809880239520958,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1683,
+      "step": 3213
+    },
+    {
+      "epoch": 4.811377245508982,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1688,
+      "step": 3214
+    },
+    {
+      "epoch": 4.812874251497006,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1655,
+      "step": 3215
+    },
+    {
+      "epoch": 4.81437125748503,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1653,
+      "step": 3216
+    },
+    {
+      "epoch": 4.815868263473054,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1656,
+      "step": 3217
+    },
+    {
+      "epoch": 4.817365269461078,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1673,
+      "step": 3218
+    },
+    {
+      "epoch": 4.818862275449102,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1707,
+      "step": 3219
+    },
+    {
+      "epoch": 4.820359281437126,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1685,
+      "step": 3220
+    },
+    {
+      "epoch": 4.82185628742515,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1656,
+      "step": 3221
+    },
+    {
+      "epoch": 4.823353293413174,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1634,
+      "step": 3222
+    },
+    {
+      "epoch": 4.824850299401198,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1673,
+      "step": 3223
+    },
+    {
+      "epoch": 4.826347305389222,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.161,
+      "step": 3224
+    },
+    {
+      "epoch": 4.827844311377246,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.162,
+      "step": 3225
+    },
+    {
+      "epoch": 4.8293413173652695,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1656,
+      "step": 3226
+    },
+    {
+      "epoch": 4.830838323353293,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1593,
+      "step": 3227
+    },
+    {
+      "epoch": 4.832335329341317,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1696,
+      "step": 3228
+    },
+    {
+      "epoch": 4.833832335329341,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1686,
+      "step": 3229
+    },
+    {
+      "epoch": 4.835329341317365,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1619,
+      "step": 3230
+    },
+    {
+      "epoch": 4.836826347305389,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.162,
+      "step": 3231
+    },
+    {
+      "epoch": 4.838323353293413,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1608,
+      "step": 3232
+    },
+    {
+      "epoch": 4.839820359281437,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1682,
+      "step": 3233
+    },
+    {
+      "epoch": 4.841317365269461,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1605,
+      "step": 3234
+    },
+    {
+      "epoch": 4.842814371257485,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1587,
+      "step": 3235
+    },
+    {
+      "epoch": 4.844311377245509,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1689,
+      "step": 3236
+    },
+    {
+      "epoch": 4.845808383233533,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1647,
+      "step": 3237
+    },
+    {
+      "epoch": 4.847305389221557,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1666,
+      "step": 3238
+    },
+    {
+      "epoch": 4.848802395209581,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1671,
+      "step": 3239
+    },
+    {
+      "epoch": 4.850299401197605,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.165,
+      "step": 3240
+    },
+    {
+      "epoch": 4.8517964071856285,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1635,
+      "step": 3241
+    },
+    {
+      "epoch": 4.8532934131736525,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1659,
+      "step": 3242
+    },
+    {
+      "epoch": 4.854790419161676,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1631,
+      "step": 3243
+    },
+    {
+      "epoch": 4.8562874251497,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1643,
+      "step": 3244
+    },
+    {
+      "epoch": 4.857784431137724,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1673,
+      "step": 3245
+    },
+    {
+      "epoch": 4.859281437125748,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.156,
+      "step": 3246
+    },
+    {
+      "epoch": 4.860778443113772,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.161,
+      "step": 3247
+    },
+    {
+      "epoch": 4.862275449101796,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1672,
+      "step": 3248
+    },
+    {
+      "epoch": 4.86377245508982,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1671,
+      "step": 3249
+    },
+    {
+      "epoch": 4.865269461077844,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1615,
+      "step": 3250
+    },
+    {
+      "epoch": 4.866766467065868,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1596,
+      "step": 3251
+    },
+    {
+      "epoch": 4.868263473053892,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1671,
+      "step": 3252
+    },
+    {
+      "epoch": 4.869760479041916,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1658,
+      "step": 3253
+    },
+    {
+      "epoch": 4.87125748502994,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1654,
+      "step": 3254
+    },
+    {
+      "epoch": 4.8727544910179645,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1701,
+      "step": 3255
+    },
+    {
+      "epoch": 4.874251497005988,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.173,
+      "step": 3256
+    },
+    {
+      "epoch": 4.875748502994012,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1671,
+      "step": 3257
+    },
+    {
+      "epoch": 4.8772455089820355,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1647,
+      "step": 3258
+    },
+    {
+      "epoch": 4.87874251497006,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.166,
+      "step": 3259
+    },
+    {
+      "epoch": 4.880239520958084,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1652,
+      "step": 3260
+    },
+    {
+      "epoch": 4.881736526946108,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.168,
+      "step": 3261
+    },
+    {
+      "epoch": 4.883233532934132,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1606,
+      "step": 3262
+    },
+    {
+      "epoch": 4.884730538922156,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1633,
+      "step": 3263
+    },
+    {
+      "epoch": 4.88622754491018,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1619,
+      "step": 3264
+    },
+    {
+      "epoch": 4.887724550898204,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1683,
+      "step": 3265
+    },
+    {
+      "epoch": 4.889221556886228,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1594,
+      "step": 3266
+    },
+    {
+      "epoch": 4.890718562874252,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1672,
+      "step": 3267
+    },
+    {
+      "epoch": 4.892215568862276,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1696,
+      "step": 3268
+    },
+    {
+      "epoch": 4.8937125748503,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1656,
+      "step": 3269
+    },
+    {
+      "epoch": 4.895209580838324,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1642,
+      "step": 3270
+    },
+    {
+      "epoch": 4.8967065868263475,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1682,
+      "step": 3271
+    },
+    {
+      "epoch": 4.8982035928143715,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1652,
+      "step": 3272
+    },
+    {
+      "epoch": 4.899700598802395,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.159,
+      "step": 3273
+    },
+    {
+      "epoch": 4.901197604790419,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1628,
+      "step": 3274
+    },
+    {
+      "epoch": 4.902694610778443,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1656,
+      "step": 3275
+    },
+    {
+      "epoch": 4.904191616766467,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.162,
+      "step": 3276
+    },
+    {
+      "epoch": 4.905688622754491,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1629,
+      "step": 3277
+    },
+    {
+      "epoch": 4.907185628742515,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1627,
+      "step": 3278
+    },
+    {
+      "epoch": 4.908682634730539,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1672,
+      "step": 3279
+    },
+    {
+      "epoch": 4.910179640718563,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1626,
+      "step": 3280
+    },
+    {
+      "epoch": 4.911676646706587,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1627,
+      "step": 3281
+    },
+    {
+      "epoch": 4.913173652694611,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1646,
+      "step": 3282
+    },
+    {
+      "epoch": 4.914670658682635,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1653,
+      "step": 3283
+    },
+    {
+      "epoch": 4.916167664670659,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1558,
+      "step": 3284
+    },
+    {
+      "epoch": 4.917664670658683,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1644,
+      "step": 3285
+    },
+    {
+      "epoch": 4.919161676646707,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.174,
+      "step": 3286
+    },
+    {
+      "epoch": 4.9206586826347305,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1694,
+      "step": 3287
+    },
+    {
+      "epoch": 4.922155688622754,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1636,
+      "step": 3288
+    },
+    {
+      "epoch": 4.923652694610778,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1747,
+      "step": 3289
+    },
+    {
+      "epoch": 4.925149700598802,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1673,
+      "step": 3290
+    },
+    {
+      "epoch": 4.926646706586826,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1656,
+      "step": 3291
+    },
+    {
+      "epoch": 4.92814371257485,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.16,
+      "step": 3292
+    },
+    {
+      "epoch": 4.929640718562874,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1713,
+      "step": 3293
+    },
+    {
+      "epoch": 4.931137724550898,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1664,
+      "step": 3294
+    },
+    {
+      "epoch": 4.932634730538922,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1691,
+      "step": 3295
+    },
+    {
+      "epoch": 4.934131736526946,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1655,
+      "step": 3296
+    },
+    {
+      "epoch": 4.93562874251497,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1623,
+      "step": 3297
+    },
+    {
+      "epoch": 4.937125748502994,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1657,
+      "step": 3298
+    },
+    {
+      "epoch": 4.938622754491018,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1668,
+      "step": 3299
+    },
+    {
+      "epoch": 4.940119760479042,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1704,
+      "step": 3300
+    },
+    {
+      "epoch": 4.941616766467066,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0008,
+      "loss": 1.166,
+      "step": 3301
+    },
+    {
+      "epoch": 4.9431137724550895,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1714,
+      "step": 3302
+    },
+    {
+      "epoch": 4.9446107784431135,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1577,
+      "step": 3303
+    },
+    {
+      "epoch": 4.946107784431137,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1643,
+      "step": 3304
+    },
+    {
+      "epoch": 4.947604790419161,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1711,
+      "step": 3305
+    },
+    {
+      "epoch": 4.949101796407185,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.158,
+      "step": 3306
+    },
+    {
+      "epoch": 4.950598802395209,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1663,
+      "step": 3307
+    },
+    {
+      "epoch": 4.952095808383233,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1611,
+      "step": 3308
+    },
+    {
+      "epoch": 4.953592814371257,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.173,
+      "step": 3309
+    },
+    {
+      "epoch": 4.955089820359282,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1589,
+      "step": 3310
+    },
+    {
+      "epoch": 4.956586826347305,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.164,
+      "step": 3311
+    },
+    {
+      "epoch": 4.95808383233533,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1672,
+      "step": 3312
+    },
+    {
+      "epoch": 4.959580838323353,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1626,
+      "step": 3313
+    },
+    {
+      "epoch": 4.961077844311378,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1711,
+      "step": 3314
+    },
+    {
+      "epoch": 4.962574850299401,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1571,
+      "step": 3315
+    },
+    {
+      "epoch": 4.9640718562874255,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.169,
+      "step": 3316
+    },
+    {
+      "epoch": 4.9655688622754495,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.17,
+      "step": 3317
+    },
+    {
+      "epoch": 4.967065868263473,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1691,
+      "step": 3318
+    },
+    {
+      "epoch": 4.968562874251497,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1671,
+      "step": 3319
+    },
+    {
+      "epoch": 4.970059880239521,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1712,
+      "step": 3320
+    },
+    {
+      "epoch": 4.971556886227545,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1657,
+      "step": 3321
+    },
+    {
+      "epoch": 4.973053892215569,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1502,
+      "step": 3322
+    },
+    {
+      "epoch": 4.974550898203593,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1623,
+      "step": 3323
+    },
+    {
+      "epoch": 4.976047904191617,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1674,
+      "step": 3324
+    },
+    {
+      "epoch": 4.977544910179641,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1745,
+      "step": 3325
+    },
+    {
+      "epoch": 4.979041916167665,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1629,
+      "step": 3326
+    },
+    {
+      "epoch": 4.980538922155689,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1611,
+      "step": 3327
+    },
+    {
+      "epoch": 4.982035928143713,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.165,
+      "step": 3328
+    },
+    {
+      "epoch": 4.983532934131737,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1593,
+      "step": 3329
+    },
+    {
+      "epoch": 4.985029940119761,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.166,
+      "step": 3330
+    },
+    {
+      "epoch": 4.986526946107785,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1601,
+      "step": 3331
+    },
+    {
+      "epoch": 4.9880239520958085,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1646,
+      "step": 3332
+    },
+    {
+      "epoch": 4.9895209580838324,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1647,
+      "step": 3333
+    },
+    {
+      "epoch": 4.991017964071856,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.166,
+      "step": 3334
+    },
+    {
+      "epoch": 4.99251497005988,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1666,
+      "step": 3335
+    },
+    {
+      "epoch": 4.994011976047904,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.164,
+      "step": 3336
+    },
+    {
+      "epoch": 4.995508982035928,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1696,
+      "step": 3337
+    },
+    {
+      "epoch": 4.997005988023952,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1686,
+      "step": 3338
+    },
+    {
+      "epoch": 4.998502994011976,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1641,
+      "step": 3339
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1668,
+      "step": 3340
+    },
+    {
+      "epoch": 5.001497005988024,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1582,
+      "step": 3341
+    },
+    {
+      "epoch": 5.002994011976048,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1601,
+      "step": 3342
+    },
+    {
+      "epoch": 5.004491017964072,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1642,
+      "step": 3343
+    },
+    {
+      "epoch": 5.005988023952096,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1676,
+      "step": 3344
+    },
+    {
+      "epoch": 5.00748502994012,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1717,
+      "step": 3345
+    },
+    {
+      "epoch": 5.008982035928144,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1622,
+      "step": 3346
+    },
+    {
+      "epoch": 5.0104790419161676,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1643,
+      "step": 3347
+    },
+    {
+      "epoch": 5.0119760479041915,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1692,
+      "step": 3348
+    },
+    {
+      "epoch": 5.013473053892215,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1638,
+      "step": 3349
+    },
+    {
+      "epoch": 5.014970059880239,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1645,
+      "step": 3350
+    },
+    {
+      "epoch": 5.016467065868263,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1701,
+      "step": 3351
+    },
+    {
+      "epoch": 5.017964071856287,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.157,
+      "step": 3352
+    },
+    {
+      "epoch": 5.019461077844311,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1619,
+      "step": 3353
+    },
+    {
+      "epoch": 5.020958083832335,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1655,
+      "step": 3354
+    },
+    {
+      "epoch": 5.022455089820359,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1573,
+      "step": 3355
+    },
+    {
+      "epoch": 5.023952095808383,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1762,
+      "step": 3356
+    },
+    {
+      "epoch": 5.025449101796407,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1698,
+      "step": 3357
+    },
+    {
+      "epoch": 5.026946107784431,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.161,
+      "step": 3358
+    },
+    {
+      "epoch": 5.028443113772455,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1639,
+      "step": 3359
+    },
+    {
+      "epoch": 5.029940119760479,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.168,
+      "step": 3360
+    },
+    {
+      "epoch": 5.031437125748503,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1661,
+      "step": 3361
+    },
+    {
+      "epoch": 5.032934131736527,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1606,
+      "step": 3362
+    },
+    {
+      "epoch": 5.0344311377245505,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1598,
+      "step": 3363
+    },
+    {
+      "epoch": 5.0359281437125745,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1623,
+      "step": 3364
+    },
+    {
+      "epoch": 5.037425149700598,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1632,
+      "step": 3365
+    },
+    {
+      "epoch": 5.038922155688622,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1572,
+      "step": 3366
+    },
+    {
+      "epoch": 5.040419161676646,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1713,
+      "step": 3367
+    },
+    {
+      "epoch": 5.04191616766467,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.166,
+      "step": 3368
+    },
+    {
+      "epoch": 5.043413173652695,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1691,
+      "step": 3369
+    },
+    {
+      "epoch": 5.044910179640719,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1717,
+      "step": 3370
+    },
+    {
+      "epoch": 5.046407185628743,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1659,
+      "step": 3371
+    },
+    {
+      "epoch": 5.047904191616767,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1641,
+      "step": 3372
+    },
+    {
+      "epoch": 5.049401197604791,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1662,
+      "step": 3373
+    },
+    {
+      "epoch": 5.050898203592815,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1693,
+      "step": 3374
+    },
+    {
+      "epoch": 5.052395209580839,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1679,
+      "step": 3375
+    },
+    {
+      "epoch": 5.053892215568863,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1643,
+      "step": 3376
+    },
+    {
+      "epoch": 5.0553892215568865,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1663,
+      "step": 3377
+    },
+    {
+      "epoch": 5.0568862275449105,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1619,
+      "step": 3378
+    },
+    {
+      "epoch": 5.058383233532934,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1623,
+      "step": 3379
+    },
+    {
+      "epoch": 5.059880239520958,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1594,
+      "step": 3380
+    },
+    {
+      "epoch": 5.061377245508982,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1635,
+      "step": 3381
+    },
+    {
+      "epoch": 5.062874251497006,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1653,
+      "step": 3382
+    },
+    {
+      "epoch": 5.06437125748503,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1686,
+      "step": 3383
+    },
+    {
+      "epoch": 5.065868263473054,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1659,
+      "step": 3384
+    },
+    {
+      "epoch": 5.067365269461078,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1645,
+      "step": 3385
+    },
+    {
+      "epoch": 5.068862275449102,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1551,
+      "step": 3386
+    },
+    {
+      "epoch": 5.070359281437126,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1613,
+      "step": 3387
+    },
+    {
+      "epoch": 5.07185628742515,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1648,
+      "step": 3388
+    },
+    {
+      "epoch": 5.073353293413174,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1601,
+      "step": 3389
+    },
+    {
+      "epoch": 5.074850299401198,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1708,
+      "step": 3390
+    },
+    {
+      "epoch": 5.076347305389222,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1598,
+      "step": 3391
+    },
+    {
+      "epoch": 5.077844311377246,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1688,
+      "step": 3392
+    },
+    {
+      "epoch": 5.0793413173652695,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1632,
+      "step": 3393
+    },
+    {
+      "epoch": 5.080838323353293,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1649,
+      "step": 3394
+    },
+    {
+      "epoch": 5.082335329341317,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.159,
+      "step": 3395
+    },
+    {
+      "epoch": 5.083832335329341,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.164,
+      "step": 3396
+    },
+    {
+      "epoch": 5.085329341317365,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1611,
+      "step": 3397
+    },
+    {
+      "epoch": 5.086826347305389,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1651,
+      "step": 3398
+    },
+    {
+      "epoch": 5.088323353293413,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1558,
+      "step": 3399
+    },
+    {
+      "epoch": 5.089820359281437,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1615,
+      "step": 3400
+    },
+    {
+      "epoch": 5.091317365269461,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1693,
+      "step": 3401
+    },
+    {
+      "epoch": 5.092814371257485,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.168,
+      "step": 3402
+    },
+    {
+      "epoch": 5.094311377245509,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1744,
+      "step": 3403
+    },
+    {
+      "epoch": 5.095808383233533,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1562,
+      "step": 3404
+    },
+    {
+      "epoch": 5.097305389221557,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1656,
+      "step": 3405
+    },
+    {
+      "epoch": 5.098802395209581,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1655,
+      "step": 3406
+    },
+    {
+      "epoch": 5.100299401197605,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1549,
+      "step": 3407
+    },
+    {
+      "epoch": 5.1017964071856285,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.168,
+      "step": 3408
+    },
+    {
+      "epoch": 5.1032934131736525,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1604,
+      "step": 3409
+    },
+    {
+      "epoch": 5.104790419161676,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1647,
+      "step": 3410
+    },
+    {
+      "epoch": 5.1062874251497,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1709,
+      "step": 3411
+    },
+    {
+      "epoch": 5.107784431137724,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1618,
+      "step": 3412
+    },
+    {
+      "epoch": 5.109281437125748,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1612,
+      "step": 3413
+    },
+    {
+      "epoch": 5.110778443113772,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1657,
+      "step": 3414
+    },
+    {
+      "epoch": 5.112275449101796,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1679,
+      "step": 3415
+    },
+    {
+      "epoch": 5.11377245508982,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.165,
+      "step": 3416
+    },
+    {
+      "epoch": 5.115269461077844,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1633,
+      "step": 3417
+    },
+    {
+      "epoch": 5.116766467065868,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1663,
+      "step": 3418
+    },
+    {
+      "epoch": 5.118263473053892,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1645,
+      "step": 3419
+    },
+    {
+      "epoch": 5.119760479041916,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1636,
+      "step": 3420
+    },
+    {
+      "epoch": 5.12125748502994,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1633,
+      "step": 3421
+    },
+    {
+      "epoch": 5.122754491017964,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1624,
+      "step": 3422
+    },
+    {
+      "epoch": 5.124251497005988,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1629,
+      "step": 3423
+    },
+    {
+      "epoch": 5.125748502994012,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1685,
+      "step": 3424
+    },
+    {
+      "epoch": 5.127245508982036,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1576,
+      "step": 3425
+    },
+    {
+      "epoch": 5.12874251497006,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1643,
+      "step": 3426
+    },
+    {
+      "epoch": 5.130239520958084,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1566,
+      "step": 3427
+    },
+    {
+      "epoch": 5.131736526946108,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1652,
+      "step": 3428
+    },
+    {
+      "epoch": 5.133233532934132,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1653,
+      "step": 3429
+    },
+    {
+      "epoch": 5.134730538922156,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1596,
+      "step": 3430
+    },
+    {
+      "epoch": 5.13622754491018,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1574,
+      "step": 3431
+    },
+    {
+      "epoch": 5.137724550898204,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1614,
+      "step": 3432
+    },
+    {
+      "epoch": 5.139221556886228,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1617,
+      "step": 3433
+    },
+    {
+      "epoch": 5.140718562874252,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1699,
+      "step": 3434
+    },
+    {
+      "epoch": 5.142215568862276,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1679,
+      "step": 3435
+    },
+    {
+      "epoch": 5.1437125748503,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1617,
+      "step": 3436
+    },
+    {
+      "epoch": 5.145209580838324,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1602,
+      "step": 3437
+    },
+    {
+      "epoch": 5.1467065868263475,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1669,
+      "step": 3438
+    },
+    {
+      "epoch": 5.1482035928143715,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1605,
+      "step": 3439
+    },
+    {
+      "epoch": 5.149700598802395,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1646,
+      "step": 3440
+    },
+    {
+      "epoch": 5.151197604790419,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1663,
+      "step": 3441
+    },
+    {
+      "epoch": 5.152694610778443,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1648,
+      "step": 3442
+    },
+    {
+      "epoch": 5.154191616766467,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1627,
+      "step": 3443
+    },
+    {
+      "epoch": 5.155688622754491,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1548,
+      "step": 3444
+    },
+    {
+      "epoch": 5.157185628742515,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1664,
+      "step": 3445
+    },
+    {
+      "epoch": 5.158682634730539,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1631,
+      "step": 3446
+    },
+    {
+      "epoch": 5.160179640718563,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1651,
+      "step": 3447
+    },
+    {
+      "epoch": 5.161676646706587,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1661,
+      "step": 3448
+    },
+    {
+      "epoch": 5.163173652694611,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1638,
+      "step": 3449
+    },
+    {
+      "epoch": 5.164670658682635,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1748,
+      "step": 3450
+    },
+    {
+      "epoch": 5.166167664670659,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1658,
+      "step": 3451
+    },
+    {
+      "epoch": 5.167664670658683,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1654,
+      "step": 3452
+    },
+    {
+      "epoch": 5.169161676646707,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1674,
+      "step": 3453
+    },
+    {
+      "epoch": 5.1706586826347305,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1601,
+      "step": 3454
+    },
+    {
+      "epoch": 5.172155688622754,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1613,
+      "step": 3455
+    },
+    {
+      "epoch": 5.173652694610778,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.162,
+      "step": 3456
+    },
+    {
+      "epoch": 5.175149700598802,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.169,
+      "step": 3457
+    },
+    {
+      "epoch": 5.176646706586826,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1645,
+      "step": 3458
+    },
+    {
+      "epoch": 5.17814371257485,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.17,
+      "step": 3459
+    },
+    {
+      "epoch": 5.179640718562874,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1683,
+      "step": 3460
+    },
+    {
+      "epoch": 5.181137724550898,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1598,
+      "step": 3461
+    },
+    {
+      "epoch": 5.182634730538922,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1662,
+      "step": 3462
+    },
+    {
+      "epoch": 5.184131736526946,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1665,
+      "step": 3463
+    },
+    {
+      "epoch": 5.18562874251497,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1683,
+      "step": 3464
+    },
+    {
+      "epoch": 5.187125748502994,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1614,
+      "step": 3465
+    },
+    {
+      "epoch": 5.188622754491018,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1681,
+      "step": 3466
+    },
+    {
+      "epoch": 5.190119760479042,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1698,
+      "step": 3467
+    },
+    {
+      "epoch": 5.191616766467066,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1688,
+      "step": 3468
+    },
+    {
+      "epoch": 5.1931137724550895,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1646,
+      "step": 3469
+    },
+    {
+      "epoch": 5.1946107784431135,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1642,
+      "step": 3470
+    },
+    {
+      "epoch": 5.196107784431137,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1675,
+      "step": 3471
+    },
+    {
+      "epoch": 5.197604790419161,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1649,
+      "step": 3472
+    },
+    {
+      "epoch": 5.199101796407185,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1636,
+      "step": 3473
+    },
+    {
+      "epoch": 5.200598802395209,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.161,
+      "step": 3474
+    },
+    {
+      "epoch": 5.202095808383233,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1595,
+      "step": 3475
+    },
+    {
+      "epoch": 5.203592814371257,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1689,
+      "step": 3476
+    },
+    {
+      "epoch": 5.205089820359281,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1621,
+      "step": 3477
+    },
+    {
+      "epoch": 5.206586826347305,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1589,
+      "step": 3478
+    },
+    {
+      "epoch": 5.20808383233533,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1658,
+      "step": 3479
+    },
+    {
+      "epoch": 5.209580838323353,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1533,
+      "step": 3480
+    },
+    {
+      "epoch": 5.211077844311378,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1597,
+      "step": 3481
+    },
+    {
+      "epoch": 5.212574850299402,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1648,
+      "step": 3482
+    },
+    {
+      "epoch": 5.2140718562874255,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1689,
+      "step": 3483
+    },
+    {
+      "epoch": 5.2155688622754495,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1699,
+      "step": 3484
+    },
+    {
+      "epoch": 5.217065868263473,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1702,
+      "step": 3485
+    },
+    {
+      "epoch": 5.218562874251497,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.161,
+      "step": 3486
+    },
+    {
+      "epoch": 5.220059880239521,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.164,
+      "step": 3487
+    },
+    {
+      "epoch": 5.221556886227545,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1633,
+      "step": 3488
+    },
+    {
+      "epoch": 5.223053892215569,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1556,
+      "step": 3489
+    },
+    {
+      "epoch": 5.224550898203593,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1647,
+      "step": 3490
+    },
+    {
+      "epoch": 5.226047904191617,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1664,
+      "step": 3491
+    },
+    {
+      "epoch": 5.227544910179641,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1613,
+      "step": 3492
+    },
+    {
+      "epoch": 5.229041916167665,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1663,
+      "step": 3493
+    },
+    {
+      "epoch": 5.230538922155689,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1717,
+      "step": 3494
+    },
+    {
+      "epoch": 5.232035928143713,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1624,
+      "step": 3495
+    },
+    {
+      "epoch": 5.233532934131737,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1604,
+      "step": 3496
+    },
+    {
+      "epoch": 5.235029940119761,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1654,
+      "step": 3497
+    },
+    {
+      "epoch": 5.236526946107785,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1693,
+      "step": 3498
+    },
+    {
+      "epoch": 5.2380239520958085,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1629,
+      "step": 3499
+    },
+    {
+      "epoch": 5.2395209580838324,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1571,
+      "step": 3500
+    },
+    {
+      "epoch": 5.241017964071856,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1609,
+      "step": 3501
+    },
+    {
+      "epoch": 5.24251497005988,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1646,
+      "step": 3502
+    },
+    {
+      "epoch": 5.244011976047904,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.168,
+      "step": 3503
+    },
+    {
+      "epoch": 5.245508982035928,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1593,
+      "step": 3504
+    },
+    {
+      "epoch": 5.247005988023952,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1621,
+      "step": 3505
+    },
+    {
+      "epoch": 5.248502994011976,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1685,
+      "step": 3506
+    },
+    {
+      "epoch": 5.25,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1639,
+      "step": 3507
+    },
+    {
+      "epoch": 5.251497005988024,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0008,
+      "loss": 1.162,
+      "step": 3508
+    },
+    {
+      "epoch": 5.252994011976048,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1606,
+      "step": 3509
+    },
+    {
+      "epoch": 5.254491017964072,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1686,
+      "step": 3510
+    },
+    {
+      "epoch": 5.255988023952096,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1604,
+      "step": 3511
+    },
+    {
+      "epoch": 5.25748502994012,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1664,
+      "step": 3512
+    },
+    {
+      "epoch": 5.258982035928144,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1625,
+      "step": 3513
+    },
+    {
+      "epoch": 5.2604790419161676,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1613,
+      "step": 3514
+    },
+    {
+      "epoch": 5.2619760479041915,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1609,
+      "step": 3515
+    },
+    {
+      "epoch": 5.263473053892215,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.162,
+      "step": 3516
+    },
+    {
+      "epoch": 5.264970059880239,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1599,
+      "step": 3517
+    },
+    {
+      "epoch": 5.266467065868263,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1589,
+      "step": 3518
+    },
+    {
+      "epoch": 5.267964071856287,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1591,
+      "step": 3519
+    },
+    {
+      "epoch": 5.269461077844311,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.16,
+      "step": 3520
+    },
+    {
+      "epoch": 5.270958083832335,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1616,
+      "step": 3521
+    },
+    {
+      "epoch": 5.272455089820359,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1593,
+      "step": 3522
+    },
+    {
+      "epoch": 5.273952095808383,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1685,
+      "step": 3523
+    },
+    {
+      "epoch": 5.275449101796407,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1647,
+      "step": 3524
+    },
+    {
+      "epoch": 5.276946107784431,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.167,
+      "step": 3525
+    },
+    {
+      "epoch": 5.278443113772455,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1673,
+      "step": 3526
+    },
+    {
+      "epoch": 5.279940119760479,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.167,
+      "step": 3527
+    },
+    {
+      "epoch": 5.281437125748503,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1655,
+      "step": 3528
+    },
+    {
+      "epoch": 5.282934131736527,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1609,
+      "step": 3529
+    },
+    {
+      "epoch": 5.2844311377245505,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1625,
+      "step": 3530
+    },
+    {
+      "epoch": 5.2859281437125745,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1706,
+      "step": 3531
+    },
+    {
+      "epoch": 5.287425149700598,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1591,
+      "step": 3532
+    },
+    {
+      "epoch": 5.288922155688622,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1649,
+      "step": 3533
+    },
+    {
+      "epoch": 5.290419161676647,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1608,
+      "step": 3534
+    },
+    {
+      "epoch": 5.29191616766467,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1651,
+      "step": 3535
+    },
+    {
+      "epoch": 5.293413173652695,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1627,
+      "step": 3536
+    },
+    {
+      "epoch": 5.294910179640719,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1631,
+      "step": 3537
+    },
+    {
+      "epoch": 5.296407185628743,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1578,
+      "step": 3538
+    },
+    {
+      "epoch": 5.297904191616767,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1618,
+      "step": 3539
+    },
+    {
+      "epoch": 5.299401197604791,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1603,
+      "step": 3540
+    },
+    {
+      "epoch": 5.300898203592815,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1582,
+      "step": 3541
+    },
+    {
+      "epoch": 5.302395209580839,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1542,
+      "step": 3542
+    },
+    {
+      "epoch": 5.303892215568863,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1653,
+      "step": 3543
+    },
+    {
+      "epoch": 5.3053892215568865,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1629,
+      "step": 3544
+    },
+    {
+      "epoch": 5.3068862275449105,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1613,
+      "step": 3545
+    },
+    {
+      "epoch": 5.308383233532934,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1638,
+      "step": 3546
+    },
+    {
+      "epoch": 5.309880239520958,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1701,
+      "step": 3547
+    },
+    {
+      "epoch": 5.311377245508982,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1565,
+      "step": 3548
+    },
+    {
+      "epoch": 5.312874251497006,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1663,
+      "step": 3549
+    },
+    {
+      "epoch": 5.31437125748503,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1564,
+      "step": 3550
+    },
+    {
+      "epoch": 5.315868263473054,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1608,
+      "step": 3551
+    },
+    {
+      "epoch": 5.317365269461078,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1611,
+      "step": 3552
+    },
+    {
+      "epoch": 5.318862275449102,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.164,
+      "step": 3553
+    },
+    {
+      "epoch": 5.320359281437126,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.151,
+      "step": 3554
+    },
+    {
+      "epoch": 5.32185628742515,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.161,
+      "step": 3555
+    },
+    {
+      "epoch": 5.323353293413174,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1612,
+      "step": 3556
+    },
+    {
+      "epoch": 5.324850299401198,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1654,
+      "step": 3557
+    },
+    {
+      "epoch": 5.326347305389222,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1617,
+      "step": 3558
+    },
+    {
+      "epoch": 5.327844311377246,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1703,
+      "step": 3559
+    },
+    {
+      "epoch": 5.3293413173652695,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.165,
+      "step": 3560
+    },
+    {
+      "epoch": 5.330838323353293,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1632,
+      "step": 3561
+    },
+    {
+      "epoch": 5.332335329341317,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1688,
+      "step": 3562
+    },
+    {
+      "epoch": 5.333832335329341,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1617,
+      "step": 3563
+    },
+    {
+      "epoch": 5.335329341317365,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1612,
+      "step": 3564
+    },
+    {
+      "epoch": 5.336826347305389,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1638,
+      "step": 3565
+    },
+    {
+      "epoch": 5.338323353293413,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1617,
+      "step": 3566
+    },
+    {
+      "epoch": 5.339820359281437,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1609,
+      "step": 3567
+    },
+    {
+      "epoch": 5.341317365269461,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1656,
+      "step": 3568
+    },
+    {
+      "epoch": 5.342814371257485,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1664,
+      "step": 3569
+    },
+    {
+      "epoch": 5.344311377245509,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1604,
+      "step": 3570
+    },
+    {
+      "epoch": 5.345808383233533,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1661,
+      "step": 3571
+    },
+    {
+      "epoch": 5.347305389221557,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1591,
+      "step": 3572
+    },
+    {
+      "epoch": 5.348802395209581,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1593,
+      "step": 3573
+    },
+    {
+      "epoch": 5.350299401197605,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1634,
+      "step": 3574
+    },
+    {
+      "epoch": 5.3517964071856285,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1578,
+      "step": 3575
+    },
+    {
+      "epoch": 5.3532934131736525,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1638,
+      "step": 3576
+    },
+    {
+      "epoch": 5.354790419161676,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1669,
+      "step": 3577
+    },
+    {
+      "epoch": 5.3562874251497,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1682,
+      "step": 3578
+    },
+    {
+      "epoch": 5.357784431137724,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1599,
+      "step": 3579
+    },
+    {
+      "epoch": 5.359281437125748,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1619,
+      "step": 3580
+    },
+    {
+      "epoch": 5.360778443113772,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1601,
+      "step": 3581
+    },
+    {
+      "epoch": 5.362275449101796,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1634,
+      "step": 3582
+    },
+    {
+      "epoch": 5.36377245508982,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1597,
+      "step": 3583
+    },
+    {
+      "epoch": 5.365269461077844,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.164,
+      "step": 3584
+    },
+    {
+      "epoch": 5.366766467065868,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1637,
+      "step": 3585
+    },
+    {
+      "epoch": 5.368263473053892,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1577,
+      "step": 3586
+    },
+    {
+      "epoch": 5.369760479041916,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1649,
+      "step": 3587
+    },
+    {
+      "epoch": 5.37125748502994,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.16,
+      "step": 3588
+    },
+    {
+      "epoch": 5.3727544910179645,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1668,
+      "step": 3589
+    },
+    {
+      "epoch": 5.374251497005988,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1652,
+      "step": 3590
+    },
+    {
+      "epoch": 5.375748502994012,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1658,
+      "step": 3591
+    },
+    {
+      "epoch": 5.3772455089820355,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1648,
+      "step": 3592
+    },
+    {
+      "epoch": 5.37874251497006,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1591,
+      "step": 3593
+    },
+    {
+      "epoch": 5.380239520958084,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1663,
+      "step": 3594
+    },
+    {
+      "epoch": 5.381736526946108,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1676,
+      "step": 3595
+    },
+    {
+      "epoch": 5.383233532934132,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.166,
+      "step": 3596
+    },
+    {
+      "epoch": 5.384730538922156,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1707,
+      "step": 3597
+    },
+    {
+      "epoch": 5.38622754491018,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1614,
+      "step": 3598
+    },
+    {
+      "epoch": 5.387724550898204,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1671,
+      "step": 3599
+    },
+    {
+      "epoch": 5.389221556886228,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1648,
+      "step": 3600
+    },
+    {
+      "epoch": 5.390718562874252,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1592,
+      "step": 3601
+    },
+    {
+      "epoch": 5.392215568862276,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1697,
+      "step": 3602
+    },
+    {
+      "epoch": 5.3937125748503,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1602,
+      "step": 3603
+    },
+    {
+      "epoch": 5.395209580838324,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1581,
+      "step": 3604
+    },
+    {
+      "epoch": 5.3967065868263475,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1601,
+      "step": 3605
+    },
+    {
+      "epoch": 5.3982035928143715,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1603,
+      "step": 3606
+    },
+    {
+      "epoch": 5.399700598802395,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1604,
+      "step": 3607
+    },
+    {
+      "epoch": 5.401197604790419,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1654,
+      "step": 3608
+    },
+    {
+      "epoch": 5.402694610778443,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1605,
+      "step": 3609
+    },
+    {
+      "epoch": 5.404191616766467,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1627,
+      "step": 3610
+    },
+    {
+      "epoch": 5.405688622754491,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.161,
+      "step": 3611
+    },
+    {
+      "epoch": 5.407185628742515,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1615,
+      "step": 3612
+    },
+    {
+      "epoch": 5.408682634730539,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1618,
+      "step": 3613
+    },
+    {
+      "epoch": 5.410179640718563,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1614,
+      "step": 3614
+    },
+    {
+      "epoch": 5.411676646706587,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1614,
+      "step": 3615
+    },
+    {
+      "epoch": 5.413173652694611,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1651,
+      "step": 3616
+    },
+    {
+      "epoch": 5.414670658682635,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1678,
+      "step": 3617
+    },
+    {
+      "epoch": 5.416167664670659,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1681,
+      "step": 3618
+    },
+    {
+      "epoch": 5.417664670658683,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1568,
+      "step": 3619
+    },
+    {
+      "epoch": 5.419161676646707,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1649,
+      "step": 3620
+    },
+    {
+      "epoch": 5.4206586826347305,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1609,
+      "step": 3621
+    },
+    {
+      "epoch": 5.422155688622754,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1628,
+      "step": 3622
+    },
+    {
+      "epoch": 5.423652694610778,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1583,
+      "step": 3623
+    },
+    {
+      "epoch": 5.425149700598802,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1661,
+      "step": 3624
+    },
+    {
+      "epoch": 5.426646706586826,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1665,
+      "step": 3625
+    },
+    {
+      "epoch": 5.42814371257485,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1563,
+      "step": 3626
+    },
+    {
+      "epoch": 5.429640718562874,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1615,
+      "step": 3627
+    },
+    {
+      "epoch": 5.431137724550898,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1607,
+      "step": 3628
+    },
+    {
+      "epoch": 5.432634730538922,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1594,
+      "step": 3629
+    },
+    {
+      "epoch": 5.434131736526946,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1639,
+      "step": 3630
+    },
+    {
+      "epoch": 5.43562874251497,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0008,
+      "loss": 1.164,
+      "step": 3631
+    },
+    {
+      "epoch": 5.437125748502994,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1699,
+      "step": 3632
+    },
+    {
+      "epoch": 5.438622754491018,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1591,
+      "step": 3633
+    },
+    {
+      "epoch": 5.440119760479042,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1636,
+      "step": 3634
+    },
+    {
+      "epoch": 5.441616766467066,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1656,
+      "step": 3635
+    },
+    {
+      "epoch": 5.4431137724550895,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1653,
+      "step": 3636
+    },
+    {
+      "epoch": 5.4446107784431135,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1597,
+      "step": 3637
+    },
+    {
+      "epoch": 5.446107784431137,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1642,
+      "step": 3638
+    },
+    {
+      "epoch": 5.447604790419161,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1644,
+      "step": 3639
+    },
+    {
+      "epoch": 5.449101796407185,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1583,
+      "step": 3640
+    },
+    {
+      "epoch": 5.450598802395209,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1602,
+      "step": 3641
+    },
+    {
+      "epoch": 5.452095808383233,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1621,
+      "step": 3642
+    },
+    {
+      "epoch": 5.453592814371257,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1619,
+      "step": 3643
+    },
+    {
+      "epoch": 5.455089820359281,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1694,
+      "step": 3644
+    },
+    {
+      "epoch": 5.456586826347305,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.155,
+      "step": 3645
+    },
+    {
+      "epoch": 5.45808383233533,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1615,
+      "step": 3646
+    },
+    {
+      "epoch": 5.459580838323353,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1636,
+      "step": 3647
+    },
+    {
+      "epoch": 5.461077844311378,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1646,
+      "step": 3648
+    },
+    {
+      "epoch": 5.462574850299402,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1626,
+      "step": 3649
+    },
+    {
+      "epoch": 5.4640718562874255,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1647,
+      "step": 3650
+    },
+    {
+      "epoch": 5.4655688622754495,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1595,
+      "step": 3651
+    },
+    {
+      "epoch": 5.467065868263473,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1606,
+      "step": 3652
+    },
+    {
+      "epoch": 5.468562874251497,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1686,
+      "step": 3653
+    },
+    {
+      "epoch": 5.470059880239521,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1621,
+      "step": 3654
+    },
+    {
+      "epoch": 5.471556886227545,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1684,
+      "step": 3655
+    },
+    {
+      "epoch": 5.473053892215569,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1673,
+      "step": 3656
+    },
+    {
+      "epoch": 5.474550898203593,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1642,
+      "step": 3657
+    },
+    {
+      "epoch": 5.476047904191617,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1574,
+      "step": 3658
+    },
+    {
+      "epoch": 5.477544910179641,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.168,
+      "step": 3659
+    },
+    {
+      "epoch": 5.479041916167665,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1696,
+      "step": 3660
+    },
+    {
+      "epoch": 5.480538922155689,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1573,
+      "step": 3661
+    },
+    {
+      "epoch": 5.482035928143713,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1641,
+      "step": 3662
+    },
+    {
+      "epoch": 5.483532934131737,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1705,
+      "step": 3663
+    },
+    {
+      "epoch": 5.485029940119761,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1591,
+      "step": 3664
+    },
+    {
+      "epoch": 5.486526946107785,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1655,
+      "step": 3665
+    },
+    {
+      "epoch": 5.4880239520958085,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1506,
+      "step": 3666
+    },
+    {
+      "epoch": 5.4895209580838324,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1628,
+      "step": 3667
+    },
+    {
+      "epoch": 5.491017964071856,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1562,
+      "step": 3668
+    },
+    {
+      "epoch": 5.49251497005988,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1635,
+      "step": 3669
+    },
+    {
+      "epoch": 5.494011976047904,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1597,
+      "step": 3670
+    },
+    {
+      "epoch": 5.495508982035928,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1623,
+      "step": 3671
+    },
+    {
+      "epoch": 5.497005988023952,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1664,
+      "step": 3672
+    },
+    {
+      "epoch": 5.498502994011976,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.168,
+      "step": 3673
+    },
+    {
+      "epoch": 5.5,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1535,
+      "step": 3674
+    },
+    {
+      "epoch": 5.501497005988024,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1642,
+      "step": 3675
+    },
+    {
+      "epoch": 5.502994011976048,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1645,
+      "step": 3676
+    },
+    {
+      "epoch": 5.504491017964072,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1649,
+      "step": 3677
+    },
+    {
+      "epoch": 5.505988023952096,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1598,
+      "step": 3678
+    },
+    {
+      "epoch": 5.50748502994012,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1593,
+      "step": 3679
+    },
+    {
+      "epoch": 5.508982035928144,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1621,
+      "step": 3680
+    },
+    {
+      "epoch": 5.5104790419161676,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1632,
+      "step": 3681
+    },
+    {
+      "epoch": 5.5119760479041915,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.163,
+      "step": 3682
+    },
+    {
+      "epoch": 5.513473053892215,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1634,
+      "step": 3683
+    },
+    {
+      "epoch": 5.514970059880239,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1671,
+      "step": 3684
+    },
+    {
+      "epoch": 5.516467065868263,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1605,
+      "step": 3685
+    },
+    {
+      "epoch": 5.517964071856287,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1554,
+      "step": 3686
+    },
+    {
+      "epoch": 5.519461077844311,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1644,
+      "step": 3687
+    },
+    {
+      "epoch": 5.520958083832335,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1652,
+      "step": 3688
+    },
+    {
+      "epoch": 5.522455089820359,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1689,
+      "step": 3689
+    },
+    {
+      "epoch": 5.523952095808383,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1651,
+      "step": 3690
+    },
+    {
+      "epoch": 5.525449101796407,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1607,
+      "step": 3691
+    },
+    {
+      "epoch": 5.526946107784431,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1604,
+      "step": 3692
+    },
+    {
+      "epoch": 5.528443113772455,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1669,
+      "step": 3693
+    },
+    {
+      "epoch": 5.529940119760479,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1605,
+      "step": 3694
+    },
+    {
+      "epoch": 5.531437125748503,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1616,
+      "step": 3695
+    },
+    {
+      "epoch": 5.532934131736527,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1616,
+      "step": 3696
+    },
+    {
+      "epoch": 5.5344311377245505,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1579,
+      "step": 3697
+    },
+    {
+      "epoch": 5.5359281437125745,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1638,
+      "step": 3698
+    },
+    {
+      "epoch": 5.537425149700599,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1664,
+      "step": 3699
+    },
+    {
+      "epoch": 5.538922155688622,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1693,
+      "step": 3700
+    },
+    {
+      "epoch": 5.540419161676647,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1677,
+      "step": 3701
+    },
+    {
+      "epoch": 5.54191616766467,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1621,
+      "step": 3702
+    },
+    {
+      "epoch": 5.543413173652695,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1655,
+      "step": 3703
+    },
+    {
+      "epoch": 5.544910179640718,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1678,
+      "step": 3704
+    },
+    {
+      "epoch": 5.546407185628743,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1631,
+      "step": 3705
+    },
+    {
+      "epoch": 5.547904191616767,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1542,
+      "step": 3706
+    },
+    {
+      "epoch": 5.549401197604791,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0008,
+      "loss": 1.1527,
+      "step": 3707
+    },
+    {
+      "epoch": 5.550898203592815,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1576,
+      "step": 3708
+    },
+    {
+      "epoch": 5.552395209580839,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1701,
+      "step": 3709
+    },
+    {
+      "epoch": 5.553892215568863,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.164,
+      "step": 3710
+    },
+    {
+      "epoch": 5.5553892215568865,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1592,
+      "step": 3711
+    },
+    {
+      "epoch": 5.5568862275449105,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1642,
+      "step": 3712
+    },
+    {
+      "epoch": 5.558383233532934,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.158,
+      "step": 3713
+    },
+    {
+      "epoch": 5.559880239520958,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1669,
+      "step": 3714
+    },
+    {
+      "epoch": 5.561377245508982,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1589,
+      "step": 3715
+    },
+    {
+      "epoch": 5.562874251497006,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1643,
+      "step": 3716
+    },
+    {
+      "epoch": 5.56437125748503,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1649,
+      "step": 3717
+    },
+    {
+      "epoch": 5.565868263473054,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1606,
+      "step": 3718
+    },
+    {
+      "epoch": 5.567365269461078,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1619,
+      "step": 3719
+    },
+    {
+      "epoch": 5.568862275449102,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1637,
+      "step": 3720
+    },
+    {
+      "epoch": 5.570359281437126,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1626,
+      "step": 3721
+    },
+    {
+      "epoch": 5.57185628742515,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1546,
+      "step": 3722
+    },
+    {
+      "epoch": 5.573353293413174,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1584,
+      "step": 3723
+    },
+    {
+      "epoch": 5.574850299401198,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1656,
+      "step": 3724
+    },
+    {
+      "epoch": 5.576347305389222,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1576,
+      "step": 3725
+    },
+    {
+      "epoch": 5.577844311377246,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1649,
+      "step": 3726
+    },
+    {
+      "epoch": 5.5793413173652695,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1622,
+      "step": 3727
+    },
+    {
+      "epoch": 5.580838323353293,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1682,
+      "step": 3728
+    },
+    {
+      "epoch": 5.582335329341317,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1683,
+      "step": 3729
+    },
+    {
+      "epoch": 5.583832335329341,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1581,
+      "step": 3730
+    },
+    {
+      "epoch": 5.585329341317365,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1588,
+      "step": 3731
+    },
+    {
+      "epoch": 5.586826347305389,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1562,
+      "step": 3732
+    },
+    {
+      "epoch": 5.588323353293413,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1583,
+      "step": 3733
+    },
+    {
+      "epoch": 5.589820359281437,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1668,
+      "step": 3734
+    },
+    {
+      "epoch": 5.591317365269461,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.156,
+      "step": 3735
+    },
+    {
+      "epoch": 5.592814371257485,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1615,
+      "step": 3736
+    },
+    {
+      "epoch": 5.594311377245509,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1632,
+      "step": 3737
+    },
+    {
+      "epoch": 5.595808383233533,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1571,
+      "step": 3738
+    },
+    {
+      "epoch": 5.597305389221557,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.156,
+      "step": 3739
+    },
+    {
+      "epoch": 5.598802395209581,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1601,
+      "step": 3740
+    },
+    {
+      "epoch": 5.600299401197605,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1663,
+      "step": 3741
+    },
+    {
+      "epoch": 5.6017964071856285,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.164,
+      "step": 3742
+    },
+    {
+      "epoch": 5.6032934131736525,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1562,
+      "step": 3743
+    },
+    {
+      "epoch": 5.604790419161676,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1632,
+      "step": 3744
+    },
+    {
+      "epoch": 5.6062874251497,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1627,
+      "step": 3745
+    },
+    {
+      "epoch": 5.607784431137724,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1606,
+      "step": 3746
+    },
+    {
+      "epoch": 5.609281437125748,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1651,
+      "step": 3747
+    },
+    {
+      "epoch": 5.610778443113772,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.165,
+      "step": 3748
+    },
+    {
+      "epoch": 5.612275449101796,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1647,
+      "step": 3749
+    },
+    {
+      "epoch": 5.61377245508982,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1639,
+      "step": 3750
+    },
+    {
+      "epoch": 5.615269461077844,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1606,
+      "step": 3751
+    },
+    {
+      "epoch": 5.616766467065868,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1589,
+      "step": 3752
+    },
+    {
+      "epoch": 5.618263473053892,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1644,
+      "step": 3753
+    },
+    {
+      "epoch": 5.619760479041916,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1664,
+      "step": 3754
+    },
+    {
+      "epoch": 5.62125748502994,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1665,
+      "step": 3755
+    },
+    {
+      "epoch": 5.6227544910179645,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1632,
+      "step": 3756
+    },
+    {
+      "epoch": 5.624251497005988,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1623,
+      "step": 3757
+    },
+    {
+      "epoch": 5.625748502994012,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1548,
+      "step": 3758
+    },
+    {
+      "epoch": 5.6272455089820355,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1658,
+      "step": 3759
+    },
+    {
+      "epoch": 5.62874251497006,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1544,
+      "step": 3760
+    },
+    {
+      "epoch": 5.630239520958084,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1625,
+      "step": 3761
+    },
+    {
+      "epoch": 5.631736526946108,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1593,
+      "step": 3762
+    },
+    {
+      "epoch": 5.633233532934132,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.159,
+      "step": 3763
+    },
+    {
+      "epoch": 5.634730538922156,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1634,
+      "step": 3764
+    },
+    {
+      "epoch": 5.63622754491018,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1559,
+      "step": 3765
+    },
+    {
+      "epoch": 5.637724550898204,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1667,
+      "step": 3766
+    },
+    {
+      "epoch": 5.639221556886228,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1633,
+      "step": 3767
+    },
+    {
+      "epoch": 5.640718562874252,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1593,
+      "step": 3768
+    },
+    {
+      "epoch": 5.642215568862276,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1652,
+      "step": 3769
+    },
+    {
+      "epoch": 5.6437125748503,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1628,
+      "step": 3770
+    },
+    {
+      "epoch": 5.645209580838324,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1647,
+      "step": 3771
+    },
+    {
+      "epoch": 5.6467065868263475,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1598,
+      "step": 3772
+    },
+    {
+      "epoch": 5.6482035928143715,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1643,
+      "step": 3773
+    },
+    {
+      "epoch": 5.649700598802395,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1568,
+      "step": 3774
+    },
+    {
+      "epoch": 5.651197604790419,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1623,
+      "step": 3775
+    },
+    {
+      "epoch": 5.652694610778443,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1624,
+      "step": 3776
+    },
+    {
+      "epoch": 5.654191616766467,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1646,
+      "step": 3777
+    },
+    {
+      "epoch": 5.655688622754491,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1565,
+      "step": 3778
+    },
+    {
+      "epoch": 5.657185628742515,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1603,
+      "step": 3779
+    },
+    {
+      "epoch": 5.658682634730539,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1607,
+      "step": 3780
+    },
+    {
+      "epoch": 5.660179640718563,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.159,
+      "step": 3781
+    },
+    {
+      "epoch": 5.661676646706587,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1696,
+      "step": 3782
+    },
+    {
+      "epoch": 5.663173652694611,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1634,
+      "step": 3783
+    },
+    {
+      "epoch": 5.664670658682635,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1617,
+      "step": 3784
+    },
+    {
+      "epoch": 5.666167664670659,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1559,
+      "step": 3785
+    },
+    {
+      "epoch": 5.667664670658683,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1617,
+      "step": 3786
+    },
+    {
+      "epoch": 5.669161676646707,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.156,
+      "step": 3787
+    },
+    {
+      "epoch": 5.6706586826347305,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1612,
+      "step": 3788
+    },
+    {
+      "epoch": 5.672155688622754,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1607,
+      "step": 3789
+    },
+    {
+      "epoch": 5.673652694610778,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1665,
+      "step": 3790
+    },
+    {
+      "epoch": 5.675149700598802,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1597,
+      "step": 3791
+    },
+    {
+      "epoch": 5.676646706586826,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1583,
+      "step": 3792
+    },
+    {
+      "epoch": 5.67814371257485,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1632,
+      "step": 3793
+    },
+    {
+      "epoch": 5.679640718562874,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1554,
+      "step": 3794
+    },
+    {
+      "epoch": 5.681137724550898,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1606,
+      "step": 3795
+    },
+    {
+      "epoch": 5.682634730538922,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1648,
+      "step": 3796
+    },
+    {
+      "epoch": 5.684131736526946,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1644,
+      "step": 3797
+    },
+    {
+      "epoch": 5.68562874251497,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1587,
+      "step": 3798
+    },
+    {
+      "epoch": 5.687125748502994,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1617,
+      "step": 3799
+    },
+    {
+      "epoch": 5.688622754491018,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.159,
+      "step": 3800
+    },
+    {
+      "epoch": 5.690119760479042,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1505,
+      "step": 3801
+    },
+    {
+      "epoch": 5.691616766467066,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1673,
+      "step": 3802
+    },
+    {
+      "epoch": 5.6931137724550895,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1621,
+      "step": 3803
+    },
+    {
+      "epoch": 5.6946107784431135,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.162,
+      "step": 3804
+    },
+    {
+      "epoch": 5.696107784431137,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1663,
+      "step": 3805
+    },
+    {
+      "epoch": 5.697604790419161,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1538,
+      "step": 3806
+    },
+    {
+      "epoch": 5.699101796407185,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1562,
+      "step": 3807
+    },
+    {
+      "epoch": 5.700598802395209,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1507,
+      "step": 3808
+    },
+    {
+      "epoch": 5.702095808383233,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1628,
+      "step": 3809
+    },
+    {
+      "epoch": 5.703592814371257,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1615,
+      "step": 3810
+    },
+    {
+      "epoch": 5.705089820359282,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1652,
+      "step": 3811
+    },
+    {
+      "epoch": 5.706586826347305,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1568,
+      "step": 3812
+    },
+    {
+      "epoch": 5.70808383233533,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1552,
+      "step": 3813
+    },
+    {
+      "epoch": 5.709580838323353,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1607,
+      "step": 3814
+    },
+    {
+      "epoch": 5.711077844311378,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.161,
+      "step": 3815
+    },
+    {
+      "epoch": 5.712574850299401,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.159,
+      "step": 3816
+    },
+    {
+      "epoch": 5.7140718562874255,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1604,
+      "step": 3817
+    },
+    {
+      "epoch": 5.7155688622754495,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1636,
+      "step": 3818
+    },
+    {
+      "epoch": 5.717065868263473,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1586,
+      "step": 3819
+    },
+    {
+      "epoch": 5.718562874251497,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1592,
+      "step": 3820
+    },
+    {
+      "epoch": 5.720059880239521,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1522,
+      "step": 3821
+    },
+    {
+      "epoch": 5.721556886227545,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1616,
+      "step": 3822
+    },
+    {
+      "epoch": 5.723053892215569,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1573,
+      "step": 3823
+    },
+    {
+      "epoch": 5.724550898203593,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1559,
+      "step": 3824
+    },
+    {
+      "epoch": 5.726047904191617,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1603,
+      "step": 3825
+    },
+    {
+      "epoch": 5.727544910179641,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1643,
+      "step": 3826
+    },
+    {
+      "epoch": 5.729041916167665,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1673,
+      "step": 3827
+    },
+    {
+      "epoch": 5.730538922155689,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1622,
+      "step": 3828
+    },
+    {
+      "epoch": 5.732035928143713,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1656,
+      "step": 3829
+    },
+    {
+      "epoch": 5.733532934131737,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1578,
+      "step": 3830
+    },
+    {
+      "epoch": 5.735029940119761,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1616,
+      "step": 3831
+    },
+    {
+      "epoch": 5.736526946107785,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1616,
+      "step": 3832
+    },
+    {
+      "epoch": 5.7380239520958085,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1657,
+      "step": 3833
+    },
+    {
+      "epoch": 5.7395209580838324,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1581,
+      "step": 3834
+    },
+    {
+      "epoch": 5.741017964071856,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1697,
+      "step": 3835
+    },
+    {
+      "epoch": 5.74251497005988,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1644,
+      "step": 3836
+    },
+    {
+      "epoch": 5.744011976047904,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.156,
+      "step": 3837
+    },
+    {
+      "epoch": 5.745508982035928,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1641,
+      "step": 3838
+    },
+    {
+      "epoch": 5.747005988023952,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1617,
+      "step": 3839
+    },
+    {
+      "epoch": 5.748502994011976,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.158,
+      "step": 3840
+    },
+    {
+      "epoch": 5.75,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1574,
+      "step": 3841
+    },
+    {
+      "epoch": 5.751497005988024,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1579,
+      "step": 3842
+    },
+    {
+      "epoch": 5.752994011976048,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1619,
+      "step": 3843
+    },
+    {
+      "epoch": 5.754491017964072,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1687,
+      "step": 3844
+    },
+    {
+      "epoch": 5.755988023952096,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1596,
+      "step": 3845
+    },
+    {
+      "epoch": 5.75748502994012,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1679,
+      "step": 3846
+    },
+    {
+      "epoch": 5.758982035928144,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1653,
+      "step": 3847
+    },
+    {
+      "epoch": 5.7604790419161676,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1633,
+      "step": 3848
+    },
+    {
+      "epoch": 5.7619760479041915,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1629,
+      "step": 3849
+    },
+    {
+      "epoch": 5.763473053892215,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1679,
+      "step": 3850
+    },
+    {
+      "epoch": 5.764970059880239,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1606,
+      "step": 3851
+    },
+    {
+      "epoch": 5.766467065868263,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.164,
+      "step": 3852
+    },
+    {
+      "epoch": 5.767964071856287,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1652,
+      "step": 3853
+    },
+    {
+      "epoch": 5.769461077844311,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1572,
+      "step": 3854
+    },
+    {
+      "epoch": 5.770958083832335,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1592,
+      "step": 3855
+    },
+    {
+      "epoch": 5.772455089820359,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1532,
+      "step": 3856
+    },
+    {
+      "epoch": 5.773952095808383,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1672,
+      "step": 3857
+    },
+    {
+      "epoch": 5.775449101796407,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1631,
+      "step": 3858
+    },
+    {
+      "epoch": 5.776946107784431,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.158,
+      "step": 3859
+    },
+    {
+      "epoch": 5.778443113772455,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1605,
+      "step": 3860
+    },
+    {
+      "epoch": 5.779940119760479,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1592,
+      "step": 3861
+    },
+    {
+      "epoch": 5.781437125748503,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1546,
+      "step": 3862
+    },
+    {
+      "epoch": 5.782934131736527,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1582,
+      "step": 3863
+    },
+    {
+      "epoch": 5.7844311377245505,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1626,
+      "step": 3864
+    },
+    {
+      "epoch": 5.7859281437125745,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1553,
+      "step": 3865
+    },
+    {
+      "epoch": 5.787425149700599,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1667,
+      "step": 3866
+    },
+    {
+      "epoch": 5.788922155688622,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1556,
+      "step": 3867
+    },
+    {
+      "epoch": 5.790419161676647,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1645,
+      "step": 3868
+    },
+    {
+      "epoch": 5.79191616766467,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1613,
+      "step": 3869
+    },
+    {
+      "epoch": 5.793413173652695,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1553,
+      "step": 3870
+    },
+    {
+      "epoch": 5.794910179640718,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1563,
+      "step": 3871
+    },
+    {
+      "epoch": 5.796407185628743,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1603,
+      "step": 3872
+    },
+    {
+      "epoch": 5.797904191616767,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1715,
+      "step": 3873
+    },
+    {
+      "epoch": 5.799401197604791,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1677,
+      "step": 3874
+    },
+    {
+      "epoch": 5.800898203592815,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1706,
+      "step": 3875
+    },
+    {
+      "epoch": 5.802395209580839,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1504,
+      "step": 3876
+    },
+    {
+      "epoch": 5.803892215568863,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1596,
+      "step": 3877
+    },
+    {
+      "epoch": 5.8053892215568865,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1551,
+      "step": 3878
+    },
+    {
+      "epoch": 5.8068862275449105,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1641,
+      "step": 3879
+    },
+    {
+      "epoch": 5.808383233532934,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1584,
+      "step": 3880
+    },
+    {
+      "epoch": 5.809880239520958,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1567,
+      "step": 3881
+    },
+    {
+      "epoch": 5.811377245508982,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1616,
+      "step": 3882
+    },
+    {
+      "epoch": 5.812874251497006,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1648,
+      "step": 3883
+    },
+    {
+      "epoch": 5.81437125748503,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1638,
+      "step": 3884
+    },
+    {
+      "epoch": 5.815868263473054,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1607,
+      "step": 3885
+    },
+    {
+      "epoch": 5.817365269461078,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1576,
+      "step": 3886
+    },
+    {
+      "epoch": 5.818862275449102,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1538,
+      "step": 3887
+    },
+    {
+      "epoch": 5.820359281437126,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1573,
+      "step": 3888
+    },
+    {
+      "epoch": 5.82185628742515,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1583,
+      "step": 3889
+    },
+    {
+      "epoch": 5.823353293413174,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1534,
+      "step": 3890
+    },
+    {
+      "epoch": 5.824850299401198,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1625,
+      "step": 3891
+    },
+    {
+      "epoch": 5.826347305389222,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1598,
+      "step": 3892
+    },
+    {
+      "epoch": 5.827844311377246,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1574,
+      "step": 3893
+    },
+    {
+      "epoch": 5.8293413173652695,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1588,
+      "step": 3894
+    },
+    {
+      "epoch": 5.830838323353293,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1598,
+      "step": 3895
+    },
+    {
+      "epoch": 5.832335329341317,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1593,
+      "step": 3896
+    },
+    {
+      "epoch": 5.833832335329341,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1567,
+      "step": 3897
+    },
+    {
+      "epoch": 5.835329341317365,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1662,
+      "step": 3898
+    },
+    {
+      "epoch": 5.836826347305389,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1574,
+      "step": 3899
+    },
+    {
+      "epoch": 5.838323353293413,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1649,
+      "step": 3900
+    },
+    {
+      "epoch": 5.839820359281437,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1617,
+      "step": 3901
+    },
+    {
+      "epoch": 5.841317365269461,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1585,
+      "step": 3902
+    },
+    {
+      "epoch": 5.842814371257485,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1551,
+      "step": 3903
+    },
+    {
+      "epoch": 5.844311377245509,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1671,
+      "step": 3904
+    },
+    {
+      "epoch": 5.845808383233533,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1593,
+      "step": 3905
+    },
+    {
+      "epoch": 5.847305389221557,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.155,
+      "step": 3906
+    },
+    {
+      "epoch": 5.848802395209581,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1632,
+      "step": 3907
+    },
+    {
+      "epoch": 5.850299401197605,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1621,
+      "step": 3908
+    },
+    {
+      "epoch": 5.8517964071856285,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1519,
+      "step": 3909
+    },
+    {
+      "epoch": 5.8532934131736525,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1551,
+      "step": 3910
+    },
+    {
+      "epoch": 5.854790419161676,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.162,
+      "step": 3911
+    },
+    {
+      "epoch": 5.8562874251497,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1597,
+      "step": 3912
+    },
+    {
+      "epoch": 5.857784431137724,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1573,
+      "step": 3913
+    },
+    {
+      "epoch": 5.859281437125748,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1592,
+      "step": 3914
+    },
+    {
+      "epoch": 5.860778443113772,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1652,
+      "step": 3915
+    },
+    {
+      "epoch": 5.862275449101796,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.152,
+      "step": 3916
+    },
+    {
+      "epoch": 5.86377245508982,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1704,
+      "step": 3917
+    },
+    {
+      "epoch": 5.865269461077844,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1555,
+      "step": 3918
+    },
+    {
+      "epoch": 5.866766467065868,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.16,
+      "step": 3919
+    },
+    {
+      "epoch": 5.868263473053892,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1634,
+      "step": 3920
+    },
+    {
+      "epoch": 5.869760479041916,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1541,
+      "step": 3921
+    },
+    {
+      "epoch": 5.87125748502994,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1633,
+      "step": 3922
+    },
+    {
+      "epoch": 5.8727544910179645,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1583,
+      "step": 3923
+    },
+    {
+      "epoch": 5.874251497005988,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1674,
+      "step": 3924
+    },
+    {
+      "epoch": 5.875748502994012,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1681,
+      "step": 3925
+    },
+    {
+      "epoch": 5.8772455089820355,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1585,
+      "step": 3926
+    },
+    {
+      "epoch": 5.87874251497006,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1612,
+      "step": 3927
+    },
+    {
+      "epoch": 5.880239520958084,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1623,
+      "step": 3928
+    },
+    {
+      "epoch": 5.881736526946108,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.158,
+      "step": 3929
+    },
+    {
+      "epoch": 5.883233532934132,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1597,
+      "step": 3930
+    },
+    {
+      "epoch": 5.884730538922156,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1675,
+      "step": 3931
+    },
+    {
+      "epoch": 5.88622754491018,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1615,
+      "step": 3932
+    },
+    {
+      "epoch": 5.887724550898204,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1586,
+      "step": 3933
+    },
+    {
+      "epoch": 5.889221556886228,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1611,
+      "step": 3934
+    },
+    {
+      "epoch": 5.890718562874252,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1604,
+      "step": 3935
+    },
+    {
+      "epoch": 5.892215568862276,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1597,
+      "step": 3936
+    },
+    {
+      "epoch": 5.8937125748503,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1578,
+      "step": 3937
+    },
+    {
+      "epoch": 5.895209580838324,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1637,
+      "step": 3938
+    },
+    {
+      "epoch": 5.8967065868263475,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1591,
+      "step": 3939
+    },
+    {
+      "epoch": 5.8982035928143715,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1578,
+      "step": 3940
+    },
+    {
+      "epoch": 5.899700598802395,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1556,
+      "step": 3941
+    },
+    {
+      "epoch": 5.901197604790419,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1552,
+      "step": 3942
+    },
+    {
+      "epoch": 5.902694610778443,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1572,
+      "step": 3943
+    },
+    {
+      "epoch": 5.904191616766467,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1694,
+      "step": 3944
+    },
+    {
+      "epoch": 5.905688622754491,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1636,
+      "step": 3945
+    },
+    {
+      "epoch": 5.907185628742515,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1643,
+      "step": 3946
+    },
+    {
+      "epoch": 5.908682634730539,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1578,
+      "step": 3947
+    },
+    {
+      "epoch": 5.910179640718563,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.168,
+      "step": 3948
+    },
+    {
+      "epoch": 5.911676646706587,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1656,
+      "step": 3949
+    },
+    {
+      "epoch": 5.913173652694611,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1555,
+      "step": 3950
+    },
+    {
+      "epoch": 5.914670658682635,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0008,
+      "loss": 1.1697,
+      "step": 3951
+    },
+    {
+      "epoch": 5.916167664670659,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1532,
+      "step": 3952
+    },
+    {
+      "epoch": 5.917664670658683,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1595,
+      "step": 3953
+    },
+    {
+      "epoch": 5.919161676646707,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1596,
+      "step": 3954
+    },
+    {
+      "epoch": 5.9206586826347305,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1592,
+      "step": 3955
+    },
+    {
+      "epoch": 5.922155688622754,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1684,
+      "step": 3956
+    },
+    {
+      "epoch": 5.923652694610778,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0008,
+      "loss": 1.163,
+      "step": 3957
+    },
+    {
+      "epoch": 5.925149700598802,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1562,
+      "step": 3958
+    },
+    {
+      "epoch": 5.926646706586826,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1629,
+      "step": 3959
+    },
+    {
+      "epoch": 5.92814371257485,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1519,
+      "step": 3960
+    },
+    {
+      "epoch": 5.929640718562874,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1534,
+      "step": 3961
+    },
+    {
+      "epoch": 5.931137724550898,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1634,
+      "step": 3962
+    },
+    {
+      "epoch": 5.932634730538922,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1582,
+      "step": 3963
+    },
+    {
+      "epoch": 5.934131736526946,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1585,
+      "step": 3964
+    },
+    {
+      "epoch": 5.93562874251497,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1548,
+      "step": 3965
+    },
+    {
+      "epoch": 5.937125748502994,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1657,
+      "step": 3966
+    },
+    {
+      "epoch": 5.938622754491018,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1551,
+      "step": 3967
+    },
+    {
+      "epoch": 5.940119760479042,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1623,
+      "step": 3968
+    },
+    {
+      "epoch": 5.941616766467066,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1614,
+      "step": 3969
+    },
+    {
+      "epoch": 5.9431137724550895,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.156,
+      "step": 3970
+    },
+    {
+      "epoch": 5.9446107784431135,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1635,
+      "step": 3971
+    },
+    {
+      "epoch": 5.946107784431137,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1616,
+      "step": 3972
+    },
+    {
+      "epoch": 5.947604790419161,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1511,
+      "step": 3973
+    },
+    {
+      "epoch": 5.949101796407185,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1573,
+      "step": 3974
+    },
+    {
+      "epoch": 5.950598802395209,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1654,
+      "step": 3975
+    },
+    {
+      "epoch": 5.952095808383233,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1559,
+      "step": 3976
+    },
+    {
+      "epoch": 5.953592814371257,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1572,
+      "step": 3977
+    },
+    {
+      "epoch": 5.955089820359282,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.169,
+      "step": 3978
+    },
+    {
+      "epoch": 5.956586826347305,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1567,
+      "step": 3979
+    },
+    {
+      "epoch": 5.95808383233533,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1571,
+      "step": 3980
+    },
+    {
+      "epoch": 5.959580838323353,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1643,
+      "step": 3981
+    },
+    {
+      "epoch": 5.961077844311378,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1559,
+      "step": 3982
+    },
+    {
+      "epoch": 5.962574850299401,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1634,
+      "step": 3983
+    },
+    {
+      "epoch": 5.9640718562874255,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1617,
+      "step": 3984
+    },
+    {
+      "epoch": 5.9655688622754495,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.152,
+      "step": 3985
+    },
+    {
+      "epoch": 5.967065868263473,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1487,
+      "step": 3986
+    },
+    {
+      "epoch": 5.968562874251497,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1566,
+      "step": 3987
+    },
+    {
+      "epoch": 5.970059880239521,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1679,
+      "step": 3988
+    },
+    {
+      "epoch": 5.971556886227545,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1588,
+      "step": 3989
+    },
+    {
+      "epoch": 5.973053892215569,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1609,
+      "step": 3990
+    },
+    {
+      "epoch": 5.974550898203593,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1604,
+      "step": 3991
+    },
+    {
+      "epoch": 5.976047904191617,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1558,
+      "step": 3992
+    },
+    {
+      "epoch": 5.977544910179641,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1585,
+      "step": 3993
+    },
+    {
+      "epoch": 5.979041916167665,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1647,
+      "step": 3994
+    },
+    {
+      "epoch": 5.980538922155689,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1654,
+      "step": 3995
+    },
+    {
+      "epoch": 5.982035928143713,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1579,
+      "step": 3996
+    },
+    {
+      "epoch": 5.983532934131737,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1494,
+      "step": 3997
+    },
+    {
+      "epoch": 5.985029940119761,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1647,
+      "step": 3998
+    },
+    {
+      "epoch": 5.986526946107785,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1605,
+      "step": 3999
+    },
+    {
+      "epoch": 5.9880239520958085,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1595,
+      "step": 4000
+    },
+    {
+      "epoch": 5.9895209580838324,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1553,
+      "step": 4001
+    },
+    {
+      "epoch": 5.991017964071856,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1653,
+      "step": 4002
+    },
+    {
+      "epoch": 5.99251497005988,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1578,
+      "step": 4003
+    },
+    {
+      "epoch": 5.994011976047904,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1634,
+      "step": 4004
+    },
+    {
+      "epoch": 5.995508982035928,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1597,
+      "step": 4005
+    },
+    {
+      "epoch": 5.997005988023952,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1563,
+      "step": 4006
+    },
+    {
+      "epoch": 5.998502994011976,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1578,
+      "step": 4007
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1599,
+      "step": 4008
+    },
+    {
+      "epoch": 6.001497005988024,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1583,
+      "step": 4009
+    },
+    {
+      "epoch": 6.002994011976048,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1622,
+      "step": 4010
+    },
+    {
+      "epoch": 6.004491017964072,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1628,
+      "step": 4011
+    },
+    {
+      "epoch": 6.005988023952096,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1581,
+      "step": 4012
+    },
+    {
+      "epoch": 6.00748502994012,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1613,
+      "step": 4013
+    },
+    {
+      "epoch": 6.008982035928144,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.157,
+      "step": 4014
+    },
+    {
+      "epoch": 6.0104790419161676,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1613,
+      "step": 4015
+    },
+    {
+      "epoch": 6.0119760479041915,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1556,
+      "step": 4016
+    },
+    {
+      "epoch": 6.013473053892215,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1612,
+      "step": 4017
+    },
+    {
+      "epoch": 6.014970059880239,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1658,
+      "step": 4018
+    },
+    {
+      "epoch": 6.016467065868263,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1598,
+      "step": 4019
+    },
+    {
+      "epoch": 6.017964071856287,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1642,
+      "step": 4020
+    },
+    {
+      "epoch": 6.019461077844311,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1587,
+      "step": 4021
+    },
+    {
+      "epoch": 6.020958083832335,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1517,
+      "step": 4022
+    },
+    {
+      "epoch": 6.022455089820359,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1482,
+      "step": 4023
+    },
+    {
+      "epoch": 6.023952095808383,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1591,
+      "step": 4024
+    },
+    {
+      "epoch": 6.025449101796407,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1593,
+      "step": 4025
+    },
+    {
+      "epoch": 6.026946107784431,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1562,
+      "step": 4026
+    },
+    {
+      "epoch": 6.028443113772455,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1614,
+      "step": 4027
+    },
+    {
+      "epoch": 6.029940119760479,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1642,
+      "step": 4028
+    },
+    {
+      "epoch": 6.031437125748503,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1615,
+      "step": 4029
+    },
+    {
+      "epoch": 6.032934131736527,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1633,
+      "step": 4030
+    },
+    {
+      "epoch": 6.0344311377245505,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1578,
+      "step": 4031
+    },
+    {
+      "epoch": 6.0359281437125745,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.163,
+      "step": 4032
+    },
+    {
+      "epoch": 6.037425149700598,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1579,
+      "step": 4033
+    },
+    {
+      "epoch": 6.038922155688622,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1572,
+      "step": 4034
+    },
+    {
+      "epoch": 6.040419161676646,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1578,
+      "step": 4035
+    },
+    {
+      "epoch": 6.04191616766467,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1609,
+      "step": 4036
+    },
+    {
+      "epoch": 6.043413173652695,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1549,
+      "step": 4037
+    },
+    {
+      "epoch": 6.044910179640719,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1605,
+      "step": 4038
+    },
+    {
+      "epoch": 6.046407185628743,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1623,
+      "step": 4039
+    },
+    {
+      "epoch": 6.047904191616767,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1644,
+      "step": 4040
+    },
+    {
+      "epoch": 6.049401197604791,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1594,
+      "step": 4041
+    },
+    {
+      "epoch": 6.050898203592815,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1622,
+      "step": 4042
+    },
+    {
+      "epoch": 6.052395209580839,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1634,
+      "step": 4043
+    },
+    {
+      "epoch": 6.053892215568863,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1616,
+      "step": 4044
+    },
+    {
+      "epoch": 6.0553892215568865,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1536,
+      "step": 4045
+    },
+    {
+      "epoch": 6.0568862275449105,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1619,
+      "step": 4046
+    },
+    {
+      "epoch": 6.058383233532934,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1552,
+      "step": 4047
+    },
+    {
+      "epoch": 6.059880239520958,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1615,
+      "step": 4048
+    },
+    {
+      "epoch": 6.061377245508982,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.156,
+      "step": 4049
+    },
+    {
+      "epoch": 6.062874251497006,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1609,
+      "step": 4050
+    },
+    {
+      "epoch": 6.06437125748503,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1583,
+      "step": 4051
+    },
+    {
+      "epoch": 6.065868263473054,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1619,
+      "step": 4052
+    },
+    {
+      "epoch": 6.067365269461078,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1634,
+      "step": 4053
+    },
+    {
+      "epoch": 6.068862275449102,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.155,
+      "step": 4054
+    },
+    {
+      "epoch": 6.070359281437126,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1533,
+      "step": 4055
+    },
+    {
+      "epoch": 6.07185628742515,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1556,
+      "step": 4056
+    },
+    {
+      "epoch": 6.073353293413174,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1571,
+      "step": 4057
+    },
+    {
+      "epoch": 6.074850299401198,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1549,
+      "step": 4058
+    },
+    {
+      "epoch": 6.076347305389222,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1566,
+      "step": 4059
+    },
+    {
+      "epoch": 6.077844311377246,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.153,
+      "step": 4060
+    },
+    {
+      "epoch": 6.0793413173652695,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1599,
+      "step": 4061
+    },
+    {
+      "epoch": 6.080838323353293,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1551,
+      "step": 4062
+    },
+    {
+      "epoch": 6.082335329341317,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1638,
+      "step": 4063
+    },
+    {
+      "epoch": 6.083832335329341,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1595,
+      "step": 4064
+    },
+    {
+      "epoch": 6.085329341317365,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1586,
+      "step": 4065
+    },
+    {
+      "epoch": 6.086826347305389,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1531,
+      "step": 4066
+    },
+    {
+      "epoch": 6.088323353293413,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1613,
+      "step": 4067
+    },
+    {
+      "epoch": 6.089820359281437,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1633,
+      "step": 4068
+    },
+    {
+      "epoch": 6.091317365269461,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1643,
+      "step": 4069
+    },
+    {
+      "epoch": 6.092814371257485,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1601,
+      "step": 4070
+    },
+    {
+      "epoch": 6.094311377245509,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1603,
+      "step": 4071
+    },
+    {
+      "epoch": 6.095808383233533,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1591,
+      "step": 4072
+    },
+    {
+      "epoch": 6.097305389221557,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1577,
+      "step": 4073
+    },
+    {
+      "epoch": 6.098802395209581,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1616,
+      "step": 4074
+    },
+    {
+      "epoch": 6.100299401197605,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1556,
+      "step": 4075
+    },
+    {
+      "epoch": 6.1017964071856285,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1556,
+      "step": 4076
+    },
+    {
+      "epoch": 6.1032934131736525,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1613,
+      "step": 4077
+    },
+    {
+      "epoch": 6.104790419161676,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.162,
+      "step": 4078
+    },
+    {
+      "epoch": 6.1062874251497,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1545,
+      "step": 4079
+    },
+    {
+      "epoch": 6.107784431137724,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1586,
+      "step": 4080
+    },
+    {
+      "epoch": 6.109281437125748,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1573,
+      "step": 4081
+    },
+    {
+      "epoch": 6.110778443113772,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1625,
+      "step": 4082
+    },
+    {
+      "epoch": 6.112275449101796,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.164,
+      "step": 4083
+    },
+    {
+      "epoch": 6.11377245508982,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1633,
+      "step": 4084
+    },
+    {
+      "epoch": 6.115269461077844,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.158,
+      "step": 4085
+    },
+    {
+      "epoch": 6.116766467065868,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1571,
+      "step": 4086
+    },
+    {
+      "epoch": 6.118263473053892,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1652,
+      "step": 4087
+    },
+    {
+      "epoch": 6.119760479041916,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1579,
+      "step": 4088
+    },
+    {
+      "epoch": 6.12125748502994,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1617,
+      "step": 4089
+    },
+    {
+      "epoch": 6.122754491017964,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1536,
+      "step": 4090
+    },
+    {
+      "epoch": 6.124251497005988,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1629,
+      "step": 4091
+    },
+    {
+      "epoch": 6.125748502994012,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1603,
+      "step": 4092
+    },
+    {
+      "epoch": 6.127245508982036,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.157,
+      "step": 4093
+    },
+    {
+      "epoch": 6.12874251497006,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1591,
+      "step": 4094
+    },
+    {
+      "epoch": 6.130239520958084,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1612,
+      "step": 4095
+    },
+    {
+      "epoch": 6.131736526946108,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1609,
+      "step": 4096
+    },
+    {
+      "epoch": 6.133233532934132,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1586,
+      "step": 4097
+    },
+    {
+      "epoch": 6.134730538922156,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1631,
+      "step": 4098
+    },
+    {
+      "epoch": 6.13622754491018,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1514,
+      "step": 4099
+    },
+    {
+      "epoch": 6.137724550898204,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1637,
+      "step": 4100
+    },
+    {
+      "epoch": 6.139221556886228,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1605,
+      "step": 4101
+    },
+    {
+      "epoch": 6.140718562874252,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1572,
+      "step": 4102
+    },
+    {
+      "epoch": 6.142215568862276,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.156,
+      "step": 4103
+    },
+    {
+      "epoch": 6.1437125748503,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1548,
+      "step": 4104
+    },
+    {
+      "epoch": 6.145209580838324,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1609,
+      "step": 4105
+    },
+    {
+      "epoch": 6.1467065868263475,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1599,
+      "step": 4106
+    },
+    {
+      "epoch": 6.1482035928143715,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1641,
+      "step": 4107
+    },
+    {
+      "epoch": 6.149700598802395,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1629,
+      "step": 4108
+    },
+    {
+      "epoch": 6.151197604790419,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1637,
+      "step": 4109
+    },
+    {
+      "epoch": 6.152694610778443,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1605,
+      "step": 4110
+    },
+    {
+      "epoch": 6.154191616766467,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1499,
+      "step": 4111
+    },
+    {
+      "epoch": 6.155688622754491,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1546,
+      "step": 4112
+    },
+    {
+      "epoch": 6.157185628742515,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1532,
+      "step": 4113
+    },
+    {
+      "epoch": 6.158682634730539,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.159,
+      "step": 4114
+    },
+    {
+      "epoch": 6.160179640718563,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1627,
+      "step": 4115
+    },
+    {
+      "epoch": 6.161676646706587,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1577,
+      "step": 4116
+    },
+    {
+      "epoch": 6.163173652694611,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1562,
+      "step": 4117
+    },
+    {
+      "epoch": 6.164670658682635,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1543,
+      "step": 4118
+    },
+    {
+      "epoch": 6.166167664670659,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1603,
+      "step": 4119
+    },
+    {
+      "epoch": 6.167664670658683,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1603,
+      "step": 4120
+    },
+    {
+      "epoch": 6.169161676646707,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1574,
+      "step": 4121
+    },
+    {
+      "epoch": 6.1706586826347305,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1577,
+      "step": 4122
+    },
+    {
+      "epoch": 6.172155688622754,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1594,
+      "step": 4123
+    },
+    {
+      "epoch": 6.173652694610778,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1538,
+      "step": 4124
+    },
+    {
+      "epoch": 6.175149700598802,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1553,
+      "step": 4125
+    },
+    {
+      "epoch": 6.176646706586826,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1514,
+      "step": 4126
+    },
+    {
+      "epoch": 6.17814371257485,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1688,
+      "step": 4127
+    },
+    {
+      "epoch": 6.179640718562874,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1612,
+      "step": 4128
+    },
+    {
+      "epoch": 6.181137724550898,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1569,
+      "step": 4129
+    },
+    {
+      "epoch": 6.182634730538922,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1568,
+      "step": 4130
+    },
+    {
+      "epoch": 6.184131736526946,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1585,
+      "step": 4131
+    },
+    {
+      "epoch": 6.18562874251497,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1519,
+      "step": 4132
+    },
+    {
+      "epoch": 6.187125748502994,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1607,
+      "step": 4133
+    },
+    {
+      "epoch": 6.188622754491018,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.163,
+      "step": 4134
+    },
+    {
+      "epoch": 6.190119760479042,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1564,
+      "step": 4135
+    },
+    {
+      "epoch": 6.191616766467066,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1604,
+      "step": 4136
+    },
+    {
+      "epoch": 6.1931137724550895,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1622,
+      "step": 4137
+    },
+    {
+      "epoch": 6.1946107784431135,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1555,
+      "step": 4138
+    },
+    {
+      "epoch": 6.196107784431137,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.16,
+      "step": 4139
+    },
+    {
+      "epoch": 6.197604790419161,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1532,
+      "step": 4140
+    },
+    {
+      "epoch": 6.199101796407185,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1596,
+      "step": 4141
+    },
+    {
+      "epoch": 6.200598802395209,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1615,
+      "step": 4142
+    },
+    {
+      "epoch": 6.202095808383233,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1552,
+      "step": 4143
+    },
+    {
+      "epoch": 6.203592814371257,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.158,
+      "step": 4144
+    },
+    {
+      "epoch": 6.205089820359281,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1635,
+      "step": 4145
+    },
+    {
+      "epoch": 6.206586826347305,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1614,
+      "step": 4146
+    },
+    {
+      "epoch": 6.20808383233533,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1643,
+      "step": 4147
+    },
+    {
+      "epoch": 6.209580838323353,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.154,
+      "step": 4148
+    },
+    {
+      "epoch": 6.211077844311378,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1586,
+      "step": 4149
+    },
+    {
+      "epoch": 6.212574850299402,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1617,
+      "step": 4150
+    },
+    {
+      "epoch": 6.2140718562874255,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.168,
+      "step": 4151
+    },
+    {
+      "epoch": 6.2155688622754495,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1502,
+      "step": 4152
+    },
+    {
+      "epoch": 6.217065868263473,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1571,
+      "step": 4153
+    },
+    {
+      "epoch": 6.218562874251497,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1568,
+      "step": 4154
+    },
+    {
+      "epoch": 6.220059880239521,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1536,
+      "step": 4155
+    },
+    {
+      "epoch": 6.221556886227545,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1604,
+      "step": 4156
+    },
+    {
+      "epoch": 6.223053892215569,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1538,
+      "step": 4157
+    },
+    {
+      "epoch": 6.224550898203593,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1544,
+      "step": 4158
+    },
+    {
+      "epoch": 6.226047904191617,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.16,
+      "step": 4159
+    },
+    {
+      "epoch": 6.227544910179641,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1569,
+      "step": 4160
+    },
+    {
+      "epoch": 6.229041916167665,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1669,
+      "step": 4161
+    },
+    {
+      "epoch": 6.230538922155689,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1599,
+      "step": 4162
+    },
+    {
+      "epoch": 6.232035928143713,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1568,
+      "step": 4163
+    },
+    {
+      "epoch": 6.233532934131737,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1545,
+      "step": 4164
+    },
+    {
+      "epoch": 6.235029940119761,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1507,
+      "step": 4165
+    },
+    {
+      "epoch": 6.236526946107785,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1606,
+      "step": 4166
+    },
+    {
+      "epoch": 6.2380239520958085,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.152,
+      "step": 4167
+    },
+    {
+      "epoch": 6.2395209580838324,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1574,
+      "step": 4168
+    },
+    {
+      "epoch": 6.241017964071856,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1648,
+      "step": 4169
+    },
+    {
+      "epoch": 6.24251497005988,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1556,
+      "step": 4170
+    },
+    {
+      "epoch": 6.244011976047904,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.157,
+      "step": 4171
+    },
+    {
+      "epoch": 6.245508982035928,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1563,
+      "step": 4172
+    },
+    {
+      "epoch": 6.247005988023952,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1592,
+      "step": 4173
+    },
+    {
+      "epoch": 6.248502994011976,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1605,
+      "step": 4174
+    },
+    {
+      "epoch": 6.25,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1616,
+      "step": 4175
+    },
+    {
+      "epoch": 6.251497005988024,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1542,
+      "step": 4176
+    },
+    {
+      "epoch": 6.252994011976048,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.163,
+      "step": 4177
+    },
+    {
+      "epoch": 6.254491017964072,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1551,
+      "step": 4178
+    },
+    {
+      "epoch": 6.255988023952096,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1624,
+      "step": 4179
+    },
+    {
+      "epoch": 6.25748502994012,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1611,
+      "step": 4180
+    },
+    {
+      "epoch": 6.258982035928144,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1646,
+      "step": 4181
+    },
+    {
+      "epoch": 6.2604790419161676,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1584,
+      "step": 4182
+    },
+    {
+      "epoch": 6.2619760479041915,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1599,
+      "step": 4183
+    },
+    {
+      "epoch": 6.263473053892215,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1591,
+      "step": 4184
+    },
+    {
+      "epoch": 6.264970059880239,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1612,
+      "step": 4185
+    },
+    {
+      "epoch": 6.266467065868263,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1562,
+      "step": 4186
+    },
+    {
+      "epoch": 6.267964071856287,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1572,
+      "step": 4187
+    },
+    {
+      "epoch": 6.269461077844311,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1636,
+      "step": 4188
+    },
+    {
+      "epoch": 6.270958083832335,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1574,
+      "step": 4189
+    },
+    {
+      "epoch": 6.272455089820359,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1566,
+      "step": 4190
+    },
+    {
+      "epoch": 6.273952095808383,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1634,
+      "step": 4191
+    },
+    {
+      "epoch": 6.275449101796407,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1664,
+      "step": 4192
+    },
+    {
+      "epoch": 6.276946107784431,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1589,
+      "step": 4193
+    },
+    {
+      "epoch": 6.278443113772455,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1503,
+      "step": 4194
+    },
+    {
+      "epoch": 6.279940119760479,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1613,
+      "step": 4195
+    },
+    {
+      "epoch": 6.281437125748503,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1525,
+      "step": 4196
+    },
+    {
+      "epoch": 6.282934131736527,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1668,
+      "step": 4197
+    },
+    {
+      "epoch": 6.2844311377245505,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.166,
+      "step": 4198
+    },
+    {
+      "epoch": 6.2859281437125745,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1543,
+      "step": 4199
+    },
+    {
+      "epoch": 6.287425149700598,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1628,
+      "step": 4200
+    },
+    {
+      "epoch": 6.288922155688622,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1582,
+      "step": 4201
+    },
+    {
+      "epoch": 6.290419161676647,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.163,
+      "step": 4202
+    },
+    {
+      "epoch": 6.29191616766467,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1582,
+      "step": 4203
+    },
+    {
+      "epoch": 6.293413173652695,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1568,
+      "step": 4204
+    },
+    {
+      "epoch": 6.294910179640719,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1603,
+      "step": 4205
+    },
+    {
+      "epoch": 6.296407185628743,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1553,
+      "step": 4206
+    },
+    {
+      "epoch": 6.297904191616767,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1602,
+      "step": 4207
+    },
+    {
+      "epoch": 6.299401197604791,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1601,
+      "step": 4208
+    },
+    {
+      "epoch": 6.300898203592815,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1569,
+      "step": 4209
+    },
+    {
+      "epoch": 6.302395209580839,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1573,
+      "step": 4210
+    },
+    {
+      "epoch": 6.303892215568863,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1609,
+      "step": 4211
+    },
+    {
+      "epoch": 6.3053892215568865,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1631,
+      "step": 4212
+    },
+    {
+      "epoch": 6.3068862275449105,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1578,
+      "step": 4213
+    },
+    {
+      "epoch": 6.308383233532934,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1635,
+      "step": 4214
+    },
+    {
+      "epoch": 6.309880239520958,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1605,
+      "step": 4215
+    },
+    {
+      "epoch": 6.311377245508982,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1609,
+      "step": 4216
+    },
+    {
+      "epoch": 6.312874251497006,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1568,
+      "step": 4217
+    },
+    {
+      "epoch": 6.31437125748503,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1547,
+      "step": 4218
+    },
+    {
+      "epoch": 6.315868263473054,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1561,
+      "step": 4219
+    },
+    {
+      "epoch": 6.317365269461078,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1547,
+      "step": 4220
+    },
+    {
+      "epoch": 6.318862275449102,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1526,
+      "step": 4221
+    },
+    {
+      "epoch": 6.320359281437126,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1536,
+      "step": 4222
+    },
+    {
+      "epoch": 6.32185628742515,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1523,
+      "step": 4223
+    },
+    {
+      "epoch": 6.323353293413174,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1604,
+      "step": 4224
+    },
+    {
+      "epoch": 6.324850299401198,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1599,
+      "step": 4225
+    },
+    {
+      "epoch": 6.326347305389222,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1569,
+      "step": 4226
+    },
+    {
+      "epoch": 6.327844311377246,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1596,
+      "step": 4227
+    },
+    {
+      "epoch": 6.3293413173652695,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.155,
+      "step": 4228
+    },
+    {
+      "epoch": 6.330838323353293,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1531,
+      "step": 4229
+    },
+    {
+      "epoch": 6.332335329341317,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1573,
+      "step": 4230
+    },
+    {
+      "epoch": 6.333832335329341,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1622,
+      "step": 4231
+    },
+    {
+      "epoch": 6.335329341317365,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1609,
+      "step": 4232
+    },
+    {
+      "epoch": 6.336826347305389,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1575,
+      "step": 4233
+    },
+    {
+      "epoch": 6.338323353293413,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1594,
+      "step": 4234
+    },
+    {
+      "epoch": 6.339820359281437,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1535,
+      "step": 4235
+    },
+    {
+      "epoch": 6.341317365269461,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1585,
+      "step": 4236
+    },
+    {
+      "epoch": 6.342814371257485,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1631,
+      "step": 4237
+    },
+    {
+      "epoch": 6.344311377245509,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1523,
+      "step": 4238
+    },
+    {
+      "epoch": 6.345808383233533,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1514,
+      "step": 4239
+    },
+    {
+      "epoch": 6.347305389221557,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1608,
+      "step": 4240
+    },
+    {
+      "epoch": 6.348802395209581,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1496,
+      "step": 4241
+    },
+    {
+      "epoch": 6.350299401197605,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1597,
+      "step": 4242
+    },
+    {
+      "epoch": 6.3517964071856285,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1648,
+      "step": 4243
+    },
+    {
+      "epoch": 6.3532934131736525,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1621,
+      "step": 4244
+    },
+    {
+      "epoch": 6.354790419161676,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.16,
+      "step": 4245
+    },
+    {
+      "epoch": 6.3562874251497,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1579,
+      "step": 4246
+    },
+    {
+      "epoch": 6.357784431137724,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1555,
+      "step": 4247
+    },
+    {
+      "epoch": 6.359281437125748,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1642,
+      "step": 4248
+    },
+    {
+      "epoch": 6.360778443113772,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1527,
+      "step": 4249
+    },
+    {
+      "epoch": 6.362275449101796,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1577,
+      "step": 4250
+    },
+    {
+      "epoch": 6.36377245508982,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.164,
+      "step": 4251
+    },
+    {
+      "epoch": 6.365269461077844,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1638,
+      "step": 4252
+    },
+    {
+      "epoch": 6.366766467065868,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1522,
+      "step": 4253
+    },
+    {
+      "epoch": 6.368263473053892,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1581,
+      "step": 4254
+    },
+    {
+      "epoch": 6.369760479041916,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1551,
+      "step": 4255
+    },
+    {
+      "epoch": 6.37125748502994,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1614,
+      "step": 4256
+    },
+    {
+      "epoch": 6.3727544910179645,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1607,
+      "step": 4257
+    },
+    {
+      "epoch": 6.374251497005988,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1656,
+      "step": 4258
+    },
+    {
+      "epoch": 6.375748502994012,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1595,
+      "step": 4259
+    },
+    {
+      "epoch": 6.3772455089820355,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1634,
+      "step": 4260
+    },
+    {
+      "epoch": 6.37874251497006,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1592,
+      "step": 4261
+    },
+    {
+      "epoch": 6.380239520958084,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1552,
+      "step": 4262
+    },
+    {
+      "epoch": 6.381736526946108,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1594,
+      "step": 4263
+    },
+    {
+      "epoch": 6.383233532934132,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1537,
+      "step": 4264
+    },
+    {
+      "epoch": 6.384730538922156,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1587,
+      "step": 4265
+    },
+    {
+      "epoch": 6.38622754491018,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1563,
+      "step": 4266
+    },
+    {
+      "epoch": 6.387724550898204,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1584,
+      "step": 4267
+    },
+    {
+      "epoch": 6.389221556886228,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1535,
+      "step": 4268
+    },
+    {
+      "epoch": 6.390718562874252,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1542,
+      "step": 4269
+    },
+    {
+      "epoch": 6.392215568862276,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1551,
+      "step": 4270
+    },
+    {
+      "epoch": 6.3937125748503,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1549,
+      "step": 4271
+    },
+    {
+      "epoch": 6.395209580838324,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1618,
+      "step": 4272
+    },
+    {
+      "epoch": 6.3967065868263475,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1636,
+      "step": 4273
+    },
+    {
+      "epoch": 6.3982035928143715,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1565,
+      "step": 4274
+    },
+    {
+      "epoch": 6.399700598802395,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1576,
+      "step": 4275
+    },
+    {
+      "epoch": 6.401197604790419,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1573,
+      "step": 4276
+    },
+    {
+      "epoch": 6.402694610778443,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1584,
+      "step": 4277
+    },
+    {
+      "epoch": 6.404191616766467,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1567,
+      "step": 4278
+    },
+    {
+      "epoch": 6.405688622754491,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1526,
+      "step": 4279
+    },
+    {
+      "epoch": 6.407185628742515,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1588,
+      "step": 4280
+    },
+    {
+      "epoch": 6.408682634730539,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1481,
+      "step": 4281
+    },
+    {
+      "epoch": 6.410179640718563,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.158,
+      "step": 4282
+    },
+    {
+      "epoch": 6.411676646706587,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1634,
+      "step": 4283
+    },
+    {
+      "epoch": 6.413173652694611,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1627,
+      "step": 4284
+    },
+    {
+      "epoch": 6.414670658682635,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1573,
+      "step": 4285
+    },
+    {
+      "epoch": 6.416167664670659,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1544,
+      "step": 4286
+    },
+    {
+      "epoch": 6.417664670658683,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1604,
+      "step": 4287
+    },
+    {
+      "epoch": 6.419161676646707,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1597,
+      "step": 4288
+    },
+    {
+      "epoch": 6.4206586826347305,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1538,
+      "step": 4289
+    },
+    {
+      "epoch": 6.422155688622754,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1552,
+      "step": 4290
+    },
+    {
+      "epoch": 6.423652694610778,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1526,
+      "step": 4291
+    },
+    {
+      "epoch": 6.425149700598802,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1494,
+      "step": 4292
+    },
+    {
+      "epoch": 6.426646706586826,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1622,
+      "step": 4293
+    },
+    {
+      "epoch": 6.42814371257485,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1595,
+      "step": 4294
+    },
+    {
+      "epoch": 6.429640718562874,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.16,
+      "step": 4295
+    },
+    {
+      "epoch": 6.431137724550898,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1626,
+      "step": 4296
+    },
+    {
+      "epoch": 6.432634730538922,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1581,
+      "step": 4297
+    },
+    {
+      "epoch": 6.434131736526946,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1586,
+      "step": 4298
+    },
+    {
+      "epoch": 6.43562874251497,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1548,
+      "step": 4299
+    },
+    {
+      "epoch": 6.437125748502994,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1582,
+      "step": 4300
+    },
+    {
+      "epoch": 6.438622754491018,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1614,
+      "step": 4301
+    },
+    {
+      "epoch": 6.440119760479042,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1546,
+      "step": 4302
+    },
+    {
+      "epoch": 6.441616766467066,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1541,
+      "step": 4303
+    },
+    {
+      "epoch": 6.4431137724550895,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1555,
+      "step": 4304
+    },
+    {
+      "epoch": 6.4446107784431135,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1659,
+      "step": 4305
+    },
+    {
+      "epoch": 6.446107784431137,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1582,
+      "step": 4306
+    },
+    {
+      "epoch": 6.447604790419161,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1545,
+      "step": 4307
+    },
+    {
+      "epoch": 6.449101796407185,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1563,
+      "step": 4308
+    },
+    {
+      "epoch": 6.450598802395209,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.154,
+      "step": 4309
+    },
+    {
+      "epoch": 6.452095808383233,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1584,
+      "step": 4310
+    },
+    {
+      "epoch": 6.453592814371257,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1546,
+      "step": 4311
+    },
+    {
+      "epoch": 6.455089820359281,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1559,
+      "step": 4312
+    },
+    {
+      "epoch": 6.456586826347305,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1598,
+      "step": 4313
+    },
+    {
+      "epoch": 6.45808383233533,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.155,
+      "step": 4314
+    },
+    {
+      "epoch": 6.459580838323353,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1643,
+      "step": 4315
+    },
+    {
+      "epoch": 6.461077844311378,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1656,
+      "step": 4316
+    },
+    {
+      "epoch": 6.462574850299402,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.154,
+      "step": 4317
+    },
+    {
+      "epoch": 6.4640718562874255,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1538,
+      "step": 4318
+    },
+    {
+      "epoch": 6.4655688622754495,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1549,
+      "step": 4319
+    },
+    {
+      "epoch": 6.467065868263473,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1515,
+      "step": 4320
+    },
+    {
+      "epoch": 6.468562874251497,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1612,
+      "step": 4321
+    },
+    {
+      "epoch": 6.470059880239521,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1575,
+      "step": 4322
+    },
+    {
+      "epoch": 6.471556886227545,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1539,
+      "step": 4323
+    },
+    {
+      "epoch": 6.473053892215569,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1593,
+      "step": 4324
+    },
+    {
+      "epoch": 6.474550898203593,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1573,
+      "step": 4325
+    },
+    {
+      "epoch": 6.476047904191617,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1544,
+      "step": 4326
+    },
+    {
+      "epoch": 6.477544910179641,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1603,
+      "step": 4327
+    },
+    {
+      "epoch": 6.479041916167665,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.152,
+      "step": 4328
+    },
+    {
+      "epoch": 6.480538922155689,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1593,
+      "step": 4329
+    },
+    {
+      "epoch": 6.482035928143713,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1549,
+      "step": 4330
+    },
+    {
+      "epoch": 6.483532934131737,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1569,
+      "step": 4331
+    },
+    {
+      "epoch": 6.485029940119761,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1608,
+      "step": 4332
+    },
+    {
+      "epoch": 6.486526946107785,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1566,
+      "step": 4333
+    },
+    {
+      "epoch": 6.4880239520958085,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1527,
+      "step": 4334
+    },
+    {
+      "epoch": 6.4895209580838324,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.155,
+      "step": 4335
+    },
+    {
+      "epoch": 6.491017964071856,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1571,
+      "step": 4336
+    },
+    {
+      "epoch": 6.49251497005988,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.164,
+      "step": 4337
+    },
+    {
+      "epoch": 6.494011976047904,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1586,
+      "step": 4338
+    },
+    {
+      "epoch": 6.495508982035928,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1599,
+      "step": 4339
+    },
+    {
+      "epoch": 6.497005988023952,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1585,
+      "step": 4340
+    },
+    {
+      "epoch": 6.498502994011976,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1538,
+      "step": 4341
+    },
+    {
+      "epoch": 6.5,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1678,
+      "step": 4342
+    },
+    {
+      "epoch": 6.501497005988024,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1581,
+      "step": 4343
+    },
+    {
+      "epoch": 6.502994011976048,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1525,
+      "step": 4344
+    },
+    {
+      "epoch": 6.504491017964072,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1574,
+      "step": 4345
+    },
+    {
+      "epoch": 6.505988023952096,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1606,
+      "step": 4346
+    },
+    {
+      "epoch": 6.50748502994012,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1603,
+      "step": 4347
+    },
+    {
+      "epoch": 6.508982035928144,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1531,
+      "step": 4348
+    },
+    {
+      "epoch": 6.5104790419161676,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1566,
+      "step": 4349
+    },
+    {
+      "epoch": 6.5119760479041915,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.167,
+      "step": 4350
+    },
+    {
+      "epoch": 6.513473053892215,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1611,
+      "step": 4351
+    },
+    {
+      "epoch": 6.514970059880239,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1631,
+      "step": 4352
+    },
+    {
+      "epoch": 6.516467065868263,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1601,
+      "step": 4353
+    },
+    {
+      "epoch": 6.517964071856287,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1552,
+      "step": 4354
+    },
+    {
+      "epoch": 6.519461077844311,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1639,
+      "step": 4355
+    },
+    {
+      "epoch": 6.520958083832335,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1594,
+      "step": 4356
+    },
+    {
+      "epoch": 6.522455089820359,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1558,
+      "step": 4357
+    },
+    {
+      "epoch": 6.523952095808383,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1518,
+      "step": 4358
+    },
+    {
+      "epoch": 6.525449101796407,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.161,
+      "step": 4359
+    },
+    {
+      "epoch": 6.526946107784431,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1512,
+      "step": 4360
+    },
+    {
+      "epoch": 6.528443113772455,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1539,
+      "step": 4361
+    },
+    {
+      "epoch": 6.529940119760479,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1594,
+      "step": 4362
+    },
+    {
+      "epoch": 6.531437125748503,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1534,
+      "step": 4363
+    },
+    {
+      "epoch": 6.532934131736527,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1594,
+      "step": 4364
+    },
+    {
+      "epoch": 6.5344311377245505,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1609,
+      "step": 4365
+    },
+    {
+      "epoch": 6.5359281437125745,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1527,
+      "step": 4366
+    },
+    {
+      "epoch": 6.537425149700599,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1621,
+      "step": 4367
+    },
+    {
+      "epoch": 6.538922155688622,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1558,
+      "step": 4368
+    },
+    {
+      "epoch": 6.540419161676647,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1523,
+      "step": 4369
+    },
+    {
+      "epoch": 6.54191616766467,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0008,
+      "loss": 1.16,
+      "step": 4370
+    },
+    {
+      "epoch": 6.543413173652695,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1581,
+      "step": 4371
+    },
+    {
+      "epoch": 6.544910179640718,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1609,
+      "step": 4372
+    },
+    {
+      "epoch": 6.546407185628743,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1549,
+      "step": 4373
+    },
+    {
+      "epoch": 6.547904191616767,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1553,
+      "step": 4374
+    },
+    {
+      "epoch": 6.549401197604791,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1587,
+      "step": 4375
+    },
+    {
+      "epoch": 6.550898203592815,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1594,
+      "step": 4376
+    },
+    {
+      "epoch": 6.552395209580839,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1549,
+      "step": 4377
+    },
+    {
+      "epoch": 6.553892215568863,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0008,
+      "loss": 1.1624,
+      "step": 4378
+    },
+    {
+      "epoch": 6.5553892215568865,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1482,
+      "step": 4379
+    },
+    {
+      "epoch": 6.5568862275449105,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1549,
+      "step": 4380
+    },
+    {
+      "epoch": 6.558383233532934,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1581,
+      "step": 4381
+    },
+    {
+      "epoch": 6.559880239520958,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1574,
+      "step": 4382
+    },
+    {
+      "epoch": 6.561377245508982,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1543,
+      "step": 4383
+    },
+    {
+      "epoch": 6.562874251497006,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1551,
+      "step": 4384
+    },
+    {
+      "epoch": 6.56437125748503,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1602,
+      "step": 4385
+    },
+    {
+      "epoch": 6.565868263473054,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1543,
+      "step": 4386
+    },
+    {
+      "epoch": 6.567365269461078,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1471,
+      "step": 4387
+    },
+    {
+      "epoch": 6.568862275449102,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1591,
+      "step": 4388
+    },
+    {
+      "epoch": 6.570359281437126,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1577,
+      "step": 4389
+    },
+    {
+      "epoch": 6.57185628742515,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1595,
+      "step": 4390
+    },
+    {
+      "epoch": 6.573353293413174,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1568,
+      "step": 4391
+    },
+    {
+      "epoch": 6.574850299401198,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1532,
+      "step": 4392
+    },
+    {
+      "epoch": 6.576347305389222,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1553,
+      "step": 4393
+    },
+    {
+      "epoch": 6.577844311377246,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1607,
+      "step": 4394
+    },
+    {
+      "epoch": 6.5793413173652695,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1612,
+      "step": 4395
+    },
+    {
+      "epoch": 6.580838323353293,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1565,
+      "step": 4396
+    },
+    {
+      "epoch": 6.582335329341317,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1573,
+      "step": 4397
+    },
+    {
+      "epoch": 6.583832335329341,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1559,
+      "step": 4398
+    },
+    {
+      "epoch": 6.585329341317365,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1592,
+      "step": 4399
+    },
+    {
+      "epoch": 6.586826347305389,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1546,
+      "step": 4400
+    },
+    {
+      "epoch": 6.588323353293413,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1595,
+      "step": 4401
+    },
+    {
+      "epoch": 6.589820359281437,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1601,
+      "step": 4402
+    },
+    {
+      "epoch": 6.591317365269461,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1617,
+      "step": 4403
+    },
+    {
+      "epoch": 6.592814371257485,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1613,
+      "step": 4404
+    },
+    {
+      "epoch": 6.594311377245509,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1602,
+      "step": 4405
+    },
+    {
+      "epoch": 6.595808383233533,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1538,
+      "step": 4406
+    },
+    {
+      "epoch": 6.597305389221557,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1544,
+      "step": 4407
+    },
+    {
+      "epoch": 6.598802395209581,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.168,
+      "step": 4408
+    },
+    {
+      "epoch": 6.600299401197605,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1551,
+      "step": 4409
+    },
+    {
+      "epoch": 6.6017964071856285,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1592,
+      "step": 4410
+    },
+    {
+      "epoch": 6.6032934131736525,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1539,
+      "step": 4411
+    },
+    {
+      "epoch": 6.604790419161676,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1595,
+      "step": 4412
+    },
+    {
+      "epoch": 6.6062874251497,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1621,
+      "step": 4413
+    },
+    {
+      "epoch": 6.607784431137724,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1509,
+      "step": 4414
+    },
+    {
+      "epoch": 6.609281437125748,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1551,
+      "step": 4415
+    },
+    {
+      "epoch": 6.610778443113772,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1618,
+      "step": 4416
+    },
+    {
+      "epoch": 6.612275449101796,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1591,
+      "step": 4417
+    },
+    {
+      "epoch": 6.61377245508982,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1645,
+      "step": 4418
+    },
+    {
+      "epoch": 6.615269461077844,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1576,
+      "step": 4419
+    },
+    {
+      "epoch": 6.616766467065868,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1515,
+      "step": 4420
+    },
+    {
+      "epoch": 6.618263473053892,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1526,
+      "step": 4421
+    },
+    {
+      "epoch": 6.619760479041916,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1571,
+      "step": 4422
+    },
+    {
+      "epoch": 6.62125748502994,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1571,
+      "step": 4423
+    },
+    {
+      "epoch": 6.6227544910179645,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1573,
+      "step": 4424
+    },
+    {
+      "epoch": 6.624251497005988,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.159,
+      "step": 4425
+    },
+    {
+      "epoch": 6.625748502994012,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1497,
+      "step": 4426
+    },
+    {
+      "epoch": 6.6272455089820355,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1553,
+      "step": 4427
+    },
+    {
+      "epoch": 6.62874251497006,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1602,
+      "step": 4428
+    },
+    {
+      "epoch": 6.630239520958084,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1577,
+      "step": 4429
+    },
+    {
+      "epoch": 6.631736526946108,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1624,
+      "step": 4430
+    },
+    {
+      "epoch": 6.633233532934132,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1586,
+      "step": 4431
+    },
+    {
+      "epoch": 6.634730538922156,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1607,
+      "step": 4432
+    },
+    {
+      "epoch": 6.63622754491018,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1593,
+      "step": 4433
+    },
+    {
+      "epoch": 6.637724550898204,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.159,
+      "step": 4434
+    },
+    {
+      "epoch": 6.639221556886228,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.161,
+      "step": 4435
+    },
+    {
+      "epoch": 6.640718562874252,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1548,
+      "step": 4436
+    },
+    {
+      "epoch": 6.642215568862276,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.16,
+      "step": 4437
+    },
+    {
+      "epoch": 6.6437125748503,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1565,
+      "step": 4438
+    },
+    {
+      "epoch": 6.645209580838324,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1585,
+      "step": 4439
+    },
+    {
+      "epoch": 6.6467065868263475,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1621,
+      "step": 4440
+    },
+    {
+      "epoch": 6.6482035928143715,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1557,
+      "step": 4441
+    },
+    {
+      "epoch": 6.649700598802395,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1554,
+      "step": 4442
+    },
+    {
+      "epoch": 6.651197604790419,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1594,
+      "step": 4443
+    },
+    {
+      "epoch": 6.652694610778443,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1563,
+      "step": 4444
+    },
+    {
+      "epoch": 6.654191616766467,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1597,
+      "step": 4445
+    },
+    {
+      "epoch": 6.655688622754491,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1533,
+      "step": 4446
+    },
+    {
+      "epoch": 6.657185628742515,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1591,
+      "step": 4447
+    },
+    {
+      "epoch": 6.658682634730539,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1534,
+      "step": 4448
+    },
+    {
+      "epoch": 6.660179640718563,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1552,
+      "step": 4449
+    },
+    {
+      "epoch": 6.661676646706587,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1556,
+      "step": 4450
+    },
+    {
+      "epoch": 6.663173652694611,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.156,
+      "step": 4451
+    },
+    {
+      "epoch": 6.664670658682635,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1491,
+      "step": 4452
+    },
+    {
+      "epoch": 6.666167664670659,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1492,
+      "step": 4453
+    },
+    {
+      "epoch": 6.667664670658683,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1548,
+      "step": 4454
+    },
+    {
+      "epoch": 6.669161676646707,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1529,
+      "step": 4455
+    },
+    {
+      "epoch": 6.6706586826347305,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1581,
+      "step": 4456
+    },
+    {
+      "epoch": 6.672155688622754,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1643,
+      "step": 4457
+    },
+    {
+      "epoch": 6.673652694610778,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1506,
+      "step": 4458
+    },
+    {
+      "epoch": 6.675149700598802,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1503,
+      "step": 4459
+    },
+    {
+      "epoch": 6.676646706586826,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1637,
+      "step": 4460
+    },
+    {
+      "epoch": 6.67814371257485,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1612,
+      "step": 4461
+    },
+    {
+      "epoch": 6.679640718562874,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1584,
+      "step": 4462
+    },
+    {
+      "epoch": 6.681137724550898,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.154,
+      "step": 4463
+    },
+    {
+      "epoch": 6.682634730538922,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1519,
+      "step": 4464
+    },
+    {
+      "epoch": 6.684131736526946,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0008,
+      "loss": 1.161,
+      "step": 4465
+    },
+    {
+      "epoch": 6.68562874251497,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1594,
+      "step": 4466
+    },
+    {
+      "epoch": 6.687125748502994,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1658,
+      "step": 4467
+    },
+    {
+      "epoch": 6.688622754491018,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1567,
+      "step": 4468
+    },
+    {
+      "epoch": 6.690119760479042,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1535,
+      "step": 4469
+    },
+    {
+      "epoch": 6.691616766467066,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1498,
+      "step": 4470
+    },
+    {
+      "epoch": 6.6931137724550895,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.153,
+      "step": 4471
+    },
+    {
+      "epoch": 6.6946107784431135,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.159,
+      "step": 4472
+    },
+    {
+      "epoch": 6.696107784431137,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1575,
+      "step": 4473
+    },
+    {
+      "epoch": 6.697604790419161,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1593,
+      "step": 4474
+    },
+    {
+      "epoch": 6.699101796407185,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1571,
+      "step": 4475
+    },
+    {
+      "epoch": 6.700598802395209,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1571,
+      "step": 4476
+    },
+    {
+      "epoch": 6.702095808383233,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1659,
+      "step": 4477
+    },
+    {
+      "epoch": 6.703592814371257,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.155,
+      "step": 4478
+    },
+    {
+      "epoch": 6.705089820359282,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1515,
+      "step": 4479
+    },
+    {
+      "epoch": 6.706586826347305,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1579,
+      "step": 4480
+    },
+    {
+      "epoch": 6.70808383233533,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1581,
+      "step": 4481
+    },
+    {
+      "epoch": 6.709580838323353,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1598,
+      "step": 4482
+    },
+    {
+      "epoch": 6.711077844311378,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1664,
+      "step": 4483
+    },
+    {
+      "epoch": 6.712574850299401,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1508,
+      "step": 4484
+    },
+    {
+      "epoch": 6.7140718562874255,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1535,
+      "step": 4485
+    },
+    {
+      "epoch": 6.7155688622754495,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1566,
+      "step": 4486
+    },
+    {
+      "epoch": 6.717065868263473,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1643,
+      "step": 4487
+    },
+    {
+      "epoch": 6.718562874251497,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1602,
+      "step": 4488
+    },
+    {
+      "epoch": 6.720059880239521,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1543,
+      "step": 4489
+    },
+    {
+      "epoch": 6.721556886227545,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1533,
+      "step": 4490
+    },
+    {
+      "epoch": 6.723053892215569,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1545,
+      "step": 4491
+    },
+    {
+      "epoch": 6.724550898203593,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1579,
+      "step": 4492
+    },
+    {
+      "epoch": 6.726047904191617,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1598,
+      "step": 4493
+    },
+    {
+      "epoch": 6.727544910179641,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1546,
+      "step": 4494
+    },
+    {
+      "epoch": 6.729041916167665,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 4495
+    },
+    {
+      "epoch": 6.730538922155689,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1524,
+      "step": 4496
+    },
+    {
+      "epoch": 6.732035928143713,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1575,
+      "step": 4497
+    },
+    {
+      "epoch": 6.733532934131737,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1545,
+      "step": 4498
+    },
+    {
+      "epoch": 6.735029940119761,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1511,
+      "step": 4499
+    },
+    {
+      "epoch": 6.736526946107785,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1617,
+      "step": 4500
+    },
+    {
+      "epoch": 6.7380239520958085,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0008,
+      "loss": 1.16,
+      "step": 4501
+    },
+    {
+      "epoch": 6.7395209580838324,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1516,
+      "step": 4502
+    },
+    {
+      "epoch": 6.741017964071856,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1585,
+      "step": 4503
+    },
+    {
+      "epoch": 6.74251497005988,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1543,
+      "step": 4504
+    },
+    {
+      "epoch": 6.744011976047904,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.162,
+      "step": 4505
+    },
+    {
+      "epoch": 6.745508982035928,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1656,
+      "step": 4506
+    },
+    {
+      "epoch": 6.747005988023952,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1579,
+      "step": 4507
+    },
+    {
+      "epoch": 6.748502994011976,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1574,
+      "step": 4508
+    },
+    {
+      "epoch": 6.75,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1523,
+      "step": 4509
+    },
+    {
+      "epoch": 6.751497005988024,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 4510
+    },
+    {
+      "epoch": 6.752994011976048,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1555,
+      "step": 4511
+    },
+    {
+      "epoch": 6.754491017964072,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.158,
+      "step": 4512
+    },
+    {
+      "epoch": 6.755988023952096,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1506,
+      "step": 4513
+    },
+    {
+      "epoch": 6.75748502994012,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1531,
+      "step": 4514
+    },
+    {
+      "epoch": 6.758982035928144,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1534,
+      "step": 4515
+    },
+    {
+      "epoch": 6.7604790419161676,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1555,
+      "step": 4516
+    },
+    {
+      "epoch": 6.7619760479041915,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1594,
+      "step": 4517
+    },
+    {
+      "epoch": 6.763473053892215,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1586,
+      "step": 4518
+    },
+    {
+      "epoch": 6.764970059880239,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1582,
+      "step": 4519
+    },
+    {
+      "epoch": 6.766467065868263,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.151,
+      "step": 4520
+    },
+    {
+      "epoch": 6.767964071856287,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1528,
+      "step": 4521
+    },
+    {
+      "epoch": 6.769461077844311,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1633,
+      "step": 4522
+    },
+    {
+      "epoch": 6.770958083832335,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1611,
+      "step": 4523
+    },
+    {
+      "epoch": 6.772455089820359,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1554,
+      "step": 4524
+    },
+    {
+      "epoch": 6.773952095808383,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1598,
+      "step": 4525
+    },
+    {
+      "epoch": 6.775449101796407,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1608,
+      "step": 4526
+    },
+    {
+      "epoch": 6.776946107784431,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.161,
+      "step": 4527
+    },
+    {
+      "epoch": 6.778443113772455,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1546,
+      "step": 4528
+    },
+    {
+      "epoch": 6.779940119760479,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1599,
+      "step": 4529
+    },
+    {
+      "epoch": 6.781437125748503,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1596,
+      "step": 4530
+    },
+    {
+      "epoch": 6.782934131736527,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1571,
+      "step": 4531
+    },
+    {
+      "epoch": 6.7844311377245505,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1494,
+      "step": 4532
+    },
+    {
+      "epoch": 6.7859281437125745,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1545,
+      "step": 4533
+    },
+    {
+      "epoch": 6.787425149700599,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1542,
+      "step": 4534
+    },
+    {
+      "epoch": 6.788922155688622,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1573,
+      "step": 4535
+    },
+    {
+      "epoch": 6.790419161676647,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1541,
+      "step": 4536
+    },
+    {
+      "epoch": 6.79191616766467,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1551,
+      "step": 4537
+    },
+    {
+      "epoch": 6.793413173652695,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1603,
+      "step": 4538
+    },
+    {
+      "epoch": 6.794910179640718,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1561,
+      "step": 4539
+    },
+    {
+      "epoch": 6.796407185628743,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.154,
+      "step": 4540
+    },
+    {
+      "epoch": 6.797904191616767,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1503,
+      "step": 4541
+    },
+    {
+      "epoch": 6.799401197604791,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1623,
+      "step": 4542
+    },
+    {
+      "epoch": 6.800898203592815,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1581,
+      "step": 4543
+    },
+    {
+      "epoch": 6.802395209580839,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1547,
+      "step": 4544
+    },
+    {
+      "epoch": 6.803892215568863,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1584,
+      "step": 4545
+    },
+    {
+      "epoch": 6.8053892215568865,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1583,
+      "step": 4546
+    },
+    {
+      "epoch": 6.8068862275449105,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1564,
+      "step": 4547
+    },
+    {
+      "epoch": 6.808383233532934,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1581,
+      "step": 4548
+    },
+    {
+      "epoch": 6.809880239520958,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1533,
+      "step": 4549
+    },
+    {
+      "epoch": 6.811377245508982,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1602,
+      "step": 4550
+    },
+    {
+      "epoch": 6.812874251497006,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1553,
+      "step": 4551
+    },
+    {
+      "epoch": 6.81437125748503,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1543,
+      "step": 4552
+    },
+    {
+      "epoch": 6.815868263473054,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1487,
+      "step": 4553
+    },
+    {
+      "epoch": 6.817365269461078,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1563,
+      "step": 4554
+    },
+    {
+      "epoch": 6.818862275449102,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1582,
+      "step": 4555
+    },
+    {
+      "epoch": 6.820359281437126,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1554,
+      "step": 4556
+    },
+    {
+      "epoch": 6.82185628742515,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1592,
+      "step": 4557
+    },
+    {
+      "epoch": 6.823353293413174,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1589,
+      "step": 4558
+    },
+    {
+      "epoch": 6.824850299401198,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1565,
+      "step": 4559
+    },
+    {
+      "epoch": 6.826347305389222,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1527,
+      "step": 4560
+    },
+    {
+      "epoch": 6.827844311377246,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1562,
+      "step": 4561
+    },
+    {
+      "epoch": 6.8293413173652695,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.158,
+      "step": 4562
+    },
+    {
+      "epoch": 6.830838323353293,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1606,
+      "step": 4563
+    },
+    {
+      "epoch": 6.832335329341317,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1528,
+      "step": 4564
+    },
+    {
+      "epoch": 6.833832335329341,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.156,
+      "step": 4565
+    },
+    {
+      "epoch": 6.835329341317365,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1537,
+      "step": 4566
+    },
+    {
+      "epoch": 6.836826347305389,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1559,
+      "step": 4567
+    },
+    {
+      "epoch": 6.838323353293413,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1572,
+      "step": 4568
+    },
+    {
+      "epoch": 6.839820359281437,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1549,
+      "step": 4569
+    },
+    {
+      "epoch": 6.841317365269461,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1569,
+      "step": 4570
+    },
+    {
+      "epoch": 6.842814371257485,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1552,
+      "step": 4571
+    },
+    {
+      "epoch": 6.844311377245509,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1471,
+      "step": 4572
+    },
+    {
+      "epoch": 6.845808383233533,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1608,
+      "step": 4573
+    },
+    {
+      "epoch": 6.847305389221557,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1536,
+      "step": 4574
+    },
+    {
+      "epoch": 6.848802395209581,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1531,
+      "step": 4575
+    },
+    {
+      "epoch": 6.850299401197605,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1484,
+      "step": 4576
+    },
+    {
+      "epoch": 6.8517964071856285,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1551,
+      "step": 4577
+    },
+    {
+      "epoch": 6.8532934131736525,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.153,
+      "step": 4578
+    },
+    {
+      "epoch": 6.854790419161676,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.156,
+      "step": 4579
+    },
+    {
+      "epoch": 6.8562874251497,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1617,
+      "step": 4580
+    },
+    {
+      "epoch": 6.857784431137724,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1581,
+      "step": 4581
+    },
+    {
+      "epoch": 6.859281437125748,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1596,
+      "step": 4582
+    },
+    {
+      "epoch": 6.860778443113772,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.16,
+      "step": 4583
+    },
+    {
+      "epoch": 6.862275449101796,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1532,
+      "step": 4584
+    },
+    {
+      "epoch": 6.86377245508982,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1509,
+      "step": 4585
+    },
+    {
+      "epoch": 6.865269461077844,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1598,
+      "step": 4586
+    },
+    {
+      "epoch": 6.866766467065868,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1492,
+      "step": 4587
+    },
+    {
+      "epoch": 6.868263473053892,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1619,
+      "step": 4588
+    },
+    {
+      "epoch": 6.869760479041916,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1594,
+      "step": 4589
+    },
+    {
+      "epoch": 6.87125748502994,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0008,
+      "loss": 1.1551,
+      "step": 4590
+    },
+    {
+      "epoch": 6.8727544910179645,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.161,
+      "step": 4591
+    },
+    {
+      "epoch": 6.874251497005988,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.153,
+      "step": 4592
+    },
+    {
+      "epoch": 6.875748502994012,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1583,
+      "step": 4593
+    },
+    {
+      "epoch": 6.8772455089820355,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1583,
+      "step": 4594
+    },
+    {
+      "epoch": 6.87874251497006,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1581,
+      "step": 4595
+    },
+    {
+      "epoch": 6.880239520958084,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1581,
+      "step": 4596
+    },
+    {
+      "epoch": 6.881736526946108,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1548,
+      "step": 4597
+    },
+    {
+      "epoch": 6.883233532934132,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1519,
+      "step": 4598
+    },
+    {
+      "epoch": 6.884730538922156,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1505,
+      "step": 4599
+    },
+    {
+      "epoch": 6.88622754491018,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1521,
+      "step": 4600
+    },
+    {
+      "epoch": 6.887724550898204,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1516,
+      "step": 4601
+    },
+    {
+      "epoch": 6.889221556886228,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1588,
+      "step": 4602
+    },
+    {
+      "epoch": 6.890718562874252,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1535,
+      "step": 4603
+    },
+    {
+      "epoch": 6.892215568862276,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1571,
+      "step": 4604
+    },
+    {
+      "epoch": 6.8937125748503,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1528,
+      "step": 4605
+    },
+    {
+      "epoch": 6.895209580838324,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1529,
+      "step": 4606
+    },
+    {
+      "epoch": 6.8967065868263475,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1538,
+      "step": 4607
+    },
+    {
+      "epoch": 6.8982035928143715,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1546,
+      "step": 4608
+    },
+    {
+      "epoch": 6.899700598802395,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1516,
+      "step": 4609
+    },
+    {
+      "epoch": 6.901197604790419,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.15,
+      "step": 4610
+    },
+    {
+      "epoch": 6.902694610778443,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1537,
+      "step": 4611
+    },
+    {
+      "epoch": 6.904191616766467,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1603,
+      "step": 4612
+    },
+    {
+      "epoch": 6.905688622754491,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0008,
+      "loss": 1.158,
+      "step": 4613
+    },
+    {
+      "epoch": 6.907185628742515,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1535,
+      "step": 4614
+    },
+    {
+      "epoch": 6.908682634730539,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1538,
+      "step": 4615
+    },
+    {
+      "epoch": 6.910179640718563,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1578,
+      "step": 4616
+    },
+    {
+      "epoch": 6.911676646706587,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1522,
+      "step": 4617
+    },
+    {
+      "epoch": 6.913173652694611,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1618,
+      "step": 4618
+    },
+    {
+      "epoch": 6.914670658682635,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1523,
+      "step": 4619
+    },
+    {
+      "epoch": 6.916167664670659,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.157,
+      "step": 4620
+    },
+    {
+      "epoch": 6.917664670658683,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1597,
+      "step": 4621
+    },
+    {
+      "epoch": 6.919161676646707,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1594,
+      "step": 4622
+    },
+    {
+      "epoch": 6.9206586826347305,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.161,
+      "step": 4623
+    },
+    {
+      "epoch": 6.922155688622754,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1541,
+      "step": 4624
+    },
+    {
+      "epoch": 6.923652694610778,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1573,
+      "step": 4625
+    },
+    {
+      "epoch": 6.925149700598802,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1644,
+      "step": 4626
+    },
+    {
+      "epoch": 6.926646706586826,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1543,
+      "step": 4627
+    },
+    {
+      "epoch": 6.92814371257485,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1557,
+      "step": 4628
+    },
+    {
+      "epoch": 6.929640718562874,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1597,
+      "step": 4629
+    },
+    {
+      "epoch": 6.931137724550898,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.151,
+      "step": 4630
+    },
+    {
+      "epoch": 6.932634730538922,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1566,
+      "step": 4631
+    },
+    {
+      "epoch": 6.934131736526946,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1605,
+      "step": 4632
+    },
+    {
+      "epoch": 6.93562874251497,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.159,
+      "step": 4633
+    },
+    {
+      "epoch": 6.937125748502994,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.153,
+      "step": 4634
+    },
+    {
+      "epoch": 6.938622754491018,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1532,
+      "step": 4635
+    },
+    {
+      "epoch": 6.940119760479042,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1555,
+      "step": 4636
+    },
+    {
+      "epoch": 6.941616766467066,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1646,
+      "step": 4637
+    },
+    {
+      "epoch": 6.9431137724550895,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1598,
+      "step": 4638
+    },
+    {
+      "epoch": 6.9446107784431135,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1531,
+      "step": 4639
+    },
+    {
+      "epoch": 6.946107784431137,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1601,
+      "step": 4640
+    },
+    {
+      "epoch": 6.947604790419161,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1565,
+      "step": 4641
+    },
+    {
+      "epoch": 6.949101796407185,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.153,
+      "step": 4642
+    },
+    {
+      "epoch": 6.950598802395209,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1601,
+      "step": 4643
+    },
+    {
+      "epoch": 6.952095808383233,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1555,
+      "step": 4644
+    },
+    {
+      "epoch": 6.953592814371257,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.16,
+      "step": 4645
+    },
+    {
+      "epoch": 6.955089820359282,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1599,
+      "step": 4646
+    },
+    {
+      "epoch": 6.956586826347305,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1533,
+      "step": 4647
+    },
+    {
+      "epoch": 6.95808383233533,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1597,
+      "step": 4648
+    },
+    {
+      "epoch": 6.959580838323353,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1539,
+      "step": 4649
+    },
+    {
+      "epoch": 6.961077844311378,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1583,
+      "step": 4650
+    },
+    {
+      "epoch": 6.962574850299401,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1547,
+      "step": 4651
+    },
+    {
+      "epoch": 6.9640718562874255,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.16,
+      "step": 4652
+    },
+    {
+      "epoch": 6.9655688622754495,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1578,
+      "step": 4653
+    },
+    {
+      "epoch": 6.967065868263473,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1522,
+      "step": 4654
+    },
+    {
+      "epoch": 6.968562874251497,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1494,
+      "step": 4655
+    },
+    {
+      "epoch": 6.970059880239521,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1634,
+      "step": 4656
+    },
+    {
+      "epoch": 6.971556886227545,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1506,
+      "step": 4657
+    },
+    {
+      "epoch": 6.973053892215569,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1566,
+      "step": 4658
+    },
+    {
+      "epoch": 6.974550898203593,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1505,
+      "step": 4659
+    },
+    {
+      "epoch": 6.976047904191617,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1625,
+      "step": 4660
+    },
+    {
+      "epoch": 6.977544910179641,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1603,
+      "step": 4661
+    },
+    {
+      "epoch": 6.979041916167665,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.157,
+      "step": 4662
+    },
+    {
+      "epoch": 6.980538922155689,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1574,
+      "step": 4663
+    },
+    {
+      "epoch": 6.982035928143713,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1509,
+      "step": 4664
+    },
+    {
+      "epoch": 6.983532934131737,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1619,
+      "step": 4665
+    },
+    {
+      "epoch": 6.985029940119761,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1596,
+      "step": 4666
+    },
+    {
+      "epoch": 6.986526946107785,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1537,
+      "step": 4667
+    },
+    {
+      "epoch": 6.9880239520958085,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.157,
+      "step": 4668
+    },
+    {
+      "epoch": 6.9895209580838324,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1564,
+      "step": 4669
+    },
+    {
+      "epoch": 6.991017964071856,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1537,
+      "step": 4670
+    },
+    {
+      "epoch": 6.99251497005988,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1597,
+      "step": 4671
+    },
+    {
+      "epoch": 6.994011976047904,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1584,
+      "step": 4672
+    },
+    {
+      "epoch": 6.995508982035928,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1563,
+      "step": 4673
+    },
+    {
+      "epoch": 6.997005988023952,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1561,
+      "step": 4674
+    },
+    {
+      "epoch": 6.998502994011976,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1582,
+      "step": 4675
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1549,
+      "step": 4676
+    },
+    {
+      "epoch": 7.001497005988024,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1562,
+      "step": 4677
+    },
+    {
+      "epoch": 7.002994011976048,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1631,
+      "step": 4678
+    },
+    {
+      "epoch": 7.004491017964072,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1555,
+      "step": 4679
+    },
+    {
+      "epoch": 7.005988023952096,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1591,
+      "step": 4680
+    },
+    {
+      "epoch": 7.00748502994012,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1508,
+      "step": 4681
+    },
+    {
+      "epoch": 7.008982035928144,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1619,
+      "step": 4682
+    },
+    {
+      "epoch": 7.0104790419161676,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1511,
+      "step": 4683
+    },
+    {
+      "epoch": 7.0119760479041915,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1594,
+      "step": 4684
+    },
+    {
+      "epoch": 7.013473053892215,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1527,
+      "step": 4685
+    },
+    {
+      "epoch": 7.014970059880239,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.148,
+      "step": 4686
+    },
+    {
+      "epoch": 7.016467065868263,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1663,
+      "step": 4687
+    },
+    {
+      "epoch": 7.017964071856287,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1596,
+      "step": 4688
+    },
+    {
+      "epoch": 7.019461077844311,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1545,
+      "step": 4689
+    },
+    {
+      "epoch": 7.020958083832335,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1554,
+      "step": 4690
+    },
+    {
+      "epoch": 7.022455089820359,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1544,
+      "step": 4691
+    },
+    {
+      "epoch": 7.023952095808383,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1513,
+      "step": 4692
+    },
+    {
+      "epoch": 7.025449101796407,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1531,
+      "step": 4693
+    },
+    {
+      "epoch": 7.026946107784431,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0008,
+      "loss": 1.155,
+      "step": 4694
+    },
+    {
+      "epoch": 7.028443113772455,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1591,
+      "step": 4695
+    },
+    {
+      "epoch": 7.029940119760479,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1539,
+      "step": 4696
+    },
+    {
+      "epoch": 7.031437125748503,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1614,
+      "step": 4697
+    },
+    {
+      "epoch": 7.032934131736527,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1565,
+      "step": 4698
+    },
+    {
+      "epoch": 7.0344311377245505,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1586,
+      "step": 4699
+    },
+    {
+      "epoch": 7.0359281437125745,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1545,
+      "step": 4700
+    },
+    {
+      "epoch": 7.037425149700598,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1591,
+      "step": 4701
+    },
+    {
+      "epoch": 7.038922155688622,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1565,
+      "step": 4702
+    },
+    {
+      "epoch": 7.040419161676646,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 4703
+    },
+    {
+      "epoch": 7.04191616766467,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1548,
+      "step": 4704
+    },
+    {
+      "epoch": 7.043413173652695,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1565,
+      "step": 4705
+    },
+    {
+      "epoch": 7.044910179640719,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1562,
+      "step": 4706
+    },
+    {
+      "epoch": 7.046407185628743,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 4707
+    },
+    {
+      "epoch": 7.047904191616767,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1601,
+      "step": 4708
+    },
+    {
+      "epoch": 7.049401197604791,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1557,
+      "step": 4709
+    },
+    {
+      "epoch": 7.050898203592815,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1563,
+      "step": 4710
+    },
+    {
+      "epoch": 7.052395209580839,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1543,
+      "step": 4711
+    },
+    {
+      "epoch": 7.053892215568863,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 4712
+    },
+    {
+      "epoch": 7.0553892215568865,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1536,
+      "step": 4713
+    },
+    {
+      "epoch": 7.0568862275449105,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1559,
+      "step": 4714
+    },
+    {
+      "epoch": 7.058383233532934,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1525,
+      "step": 4715
+    },
+    {
+      "epoch": 7.059880239520958,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1574,
+      "step": 4716
+    },
+    {
+      "epoch": 7.061377245508982,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1513,
+      "step": 4717
+    },
+    {
+      "epoch": 7.062874251497006,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1463,
+      "step": 4718
+    },
+    {
+      "epoch": 7.06437125748503,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1553,
+      "step": 4719
+    },
+    {
+      "epoch": 7.065868263473054,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1536,
+      "step": 4720
+    },
+    {
+      "epoch": 7.067365269461078,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1573,
+      "step": 4721
+    },
+    {
+      "epoch": 7.068862275449102,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1471,
+      "step": 4722
+    },
+    {
+      "epoch": 7.070359281437126,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1593,
+      "step": 4723
+    },
+    {
+      "epoch": 7.07185628742515,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1575,
+      "step": 4724
+    },
+    {
+      "epoch": 7.073353293413174,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1511,
+      "step": 4725
+    },
+    {
+      "epoch": 7.074850299401198,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1533,
+      "step": 4726
+    },
+    {
+      "epoch": 7.076347305389222,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1526,
+      "step": 4727
+    },
+    {
+      "epoch": 7.077844311377246,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1583,
+      "step": 4728
+    },
+    {
+      "epoch": 7.0793413173652695,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1575,
+      "step": 4729
+    },
+    {
+      "epoch": 7.080838323353293,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1613,
+      "step": 4730
+    },
+    {
+      "epoch": 7.082335329341317,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1535,
+      "step": 4731
+    },
+    {
+      "epoch": 7.083832335329341,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1531,
+      "step": 4732
+    },
+    {
+      "epoch": 7.085329341317365,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1531,
+      "step": 4733
+    },
+    {
+      "epoch": 7.086826347305389,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1531,
+      "step": 4734
+    },
+    {
+      "epoch": 7.088323353293413,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1573,
+      "step": 4735
+    },
+    {
+      "epoch": 7.089820359281437,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1488,
+      "step": 4736
+    },
+    {
+      "epoch": 7.091317365269461,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1579,
+      "step": 4737
+    },
+    {
+      "epoch": 7.092814371257485,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 4738
+    },
+    {
+      "epoch": 7.094311377245509,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1503,
+      "step": 4739
+    },
+    {
+      "epoch": 7.095808383233533,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1583,
+      "step": 4740
+    },
+    {
+      "epoch": 7.097305389221557,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.155,
+      "step": 4741
+    },
+    {
+      "epoch": 7.098802395209581,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.161,
+      "step": 4742
+    },
+    {
+      "epoch": 7.100299401197605,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1586,
+      "step": 4743
+    },
+    {
+      "epoch": 7.1017964071856285,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1537,
+      "step": 4744
+    },
+    {
+      "epoch": 7.1032934131736525,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1561,
+      "step": 4745
+    },
+    {
+      "epoch": 7.104790419161676,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1508,
+      "step": 4746
+    },
+    {
+      "epoch": 7.1062874251497,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1526,
+      "step": 4747
+    },
+    {
+      "epoch": 7.107784431137724,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1542,
+      "step": 4748
+    },
+    {
+      "epoch": 7.109281437125748,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1536,
+      "step": 4749
+    },
+    {
+      "epoch": 7.110778443113772,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1584,
+      "step": 4750
+    },
+    {
+      "epoch": 7.112275449101796,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1592,
+      "step": 4751
+    },
+    {
+      "epoch": 7.11377245508982,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.154,
+      "step": 4752
+    },
+    {
+      "epoch": 7.115269461077844,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1497,
+      "step": 4753
+    },
+    {
+      "epoch": 7.116766467065868,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1592,
+      "step": 4754
+    },
+    {
+      "epoch": 7.118263473053892,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1583,
+      "step": 4755
+    },
+    {
+      "epoch": 7.119760479041916,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1606,
+      "step": 4756
+    },
+    {
+      "epoch": 7.12125748502994,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1548,
+      "step": 4757
+    },
+    {
+      "epoch": 7.122754491017964,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0008,
+      "loss": 1.162,
+      "step": 4758
+    },
+    {
+      "epoch": 7.124251497005988,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1553,
+      "step": 4759
+    },
+    {
+      "epoch": 7.125748502994012,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1515,
+      "step": 4760
+    },
+    {
+      "epoch": 7.127245508982036,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1557,
+      "step": 4761
+    },
+    {
+      "epoch": 7.12874251497006,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1598,
+      "step": 4762
+    },
+    {
+      "epoch": 7.130239520958084,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1617,
+      "step": 4763
+    },
+    {
+      "epoch": 7.131736526946108,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1573,
+      "step": 4764
+    },
+    {
+      "epoch": 7.133233532934132,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1538,
+      "step": 4765
+    },
+    {
+      "epoch": 7.134730538922156,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1546,
+      "step": 4766
+    },
+    {
+      "epoch": 7.13622754491018,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1484,
+      "step": 4767
+    },
+    {
+      "epoch": 7.137724550898204,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1601,
+      "step": 4768
+    },
+    {
+      "epoch": 7.139221556886228,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1497,
+      "step": 4769
+    },
+    {
+      "epoch": 7.140718562874252,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1504,
+      "step": 4770
+    },
+    {
+      "epoch": 7.142215568862276,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1535,
+      "step": 4771
+    },
+    {
+      "epoch": 7.1437125748503,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1549,
+      "step": 4772
+    },
+    {
+      "epoch": 7.145209580838324,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1534,
+      "step": 4773
+    },
+    {
+      "epoch": 7.1467065868263475,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1618,
+      "step": 4774
+    },
+    {
+      "epoch": 7.1482035928143715,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1539,
+      "step": 4775
+    },
+    {
+      "epoch": 7.149700598802395,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1516,
+      "step": 4776
+    },
+    {
+      "epoch": 7.151197604790419,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1566,
+      "step": 4777
+    },
+    {
+      "epoch": 7.152694610778443,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1668,
+      "step": 4778
+    },
+    {
+      "epoch": 7.154191616766467,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1537,
+      "step": 4779
+    },
+    {
+      "epoch": 7.155688622754491,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.157,
+      "step": 4780
+    },
+    {
+      "epoch": 7.157185628742515,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1592,
+      "step": 4781
+    },
+    {
+      "epoch": 7.158682634730539,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1589,
+      "step": 4782
+    },
+    {
+      "epoch": 7.160179640718563,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1576,
+      "step": 4783
+    },
+    {
+      "epoch": 7.161676646706587,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1563,
+      "step": 4784
+    },
+    {
+      "epoch": 7.163173652694611,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1587,
+      "step": 4785
+    },
+    {
+      "epoch": 7.164670658682635,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1552,
+      "step": 4786
+    },
+    {
+      "epoch": 7.166167664670659,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1552,
+      "step": 4787
+    },
+    {
+      "epoch": 7.167664670658683,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1511,
+      "step": 4788
+    },
+    {
+      "epoch": 7.169161676646707,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1534,
+      "step": 4789
+    },
+    {
+      "epoch": 7.1706586826347305,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1609,
+      "step": 4790
+    },
+    {
+      "epoch": 7.172155688622754,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1517,
+      "step": 4791
+    },
+    {
+      "epoch": 7.173652694610778,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1526,
+      "step": 4792
+    },
+    {
+      "epoch": 7.175149700598802,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1555,
+      "step": 4793
+    },
+    {
+      "epoch": 7.176646706586826,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1546,
+      "step": 4794
+    },
+    {
+      "epoch": 7.17814371257485,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1558,
+      "step": 4795
+    },
+    {
+      "epoch": 7.179640718562874,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1475,
+      "step": 4796
+    },
+    {
+      "epoch": 7.181137724550898,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1588,
+      "step": 4797
+    },
+    {
+      "epoch": 7.182634730538922,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1516,
+      "step": 4798
+    },
+    {
+      "epoch": 7.184131736526946,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1535,
+      "step": 4799
+    },
+    {
+      "epoch": 7.18562874251497,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1521,
+      "step": 4800
+    },
+    {
+      "epoch": 7.187125748502994,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1533,
+      "step": 4801
+    },
+    {
+      "epoch": 7.188622754491018,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.159,
+      "step": 4802
+    },
+    {
+      "epoch": 7.190119760479042,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1566,
+      "step": 4803
+    },
+    {
+      "epoch": 7.191616766467066,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1496,
+      "step": 4804
+    },
+    {
+      "epoch": 7.1931137724550895,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1581,
+      "step": 4805
+    },
+    {
+      "epoch": 7.1946107784431135,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1511,
+      "step": 4806
+    },
+    {
+      "epoch": 7.196107784431137,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1503,
+      "step": 4807
+    },
+    {
+      "epoch": 7.197604790419161,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1568,
+      "step": 4808
+    },
+    {
+      "epoch": 7.199101796407185,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1549,
+      "step": 4809
+    },
+    {
+      "epoch": 7.200598802395209,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1572,
+      "step": 4810
+    },
+    {
+      "epoch": 7.202095808383233,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1534,
+      "step": 4811
+    },
+    {
+      "epoch": 7.203592814371257,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1575,
+      "step": 4812
+    },
+    {
+      "epoch": 7.205089820359281,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1495,
+      "step": 4813
+    },
+    {
+      "epoch": 7.206586826347305,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1621,
+      "step": 4814
+    },
+    {
+      "epoch": 7.20808383233533,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1536,
+      "step": 4815
+    },
+    {
+      "epoch": 7.209580838323353,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.157,
+      "step": 4816
+    },
+    {
+      "epoch": 7.211077844311378,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1567,
+      "step": 4817
+    },
+    {
+      "epoch": 7.212574850299402,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1612,
+      "step": 4818
+    },
+    {
+      "epoch": 7.2140718562874255,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1609,
+      "step": 4819
+    },
+    {
+      "epoch": 7.2155688622754495,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1537,
+      "step": 4820
+    },
+    {
+      "epoch": 7.217065868263473,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1504,
+      "step": 4821
+    },
+    {
+      "epoch": 7.218562874251497,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1577,
+      "step": 4822
+    },
+    {
+      "epoch": 7.220059880239521,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1554,
+      "step": 4823
+    },
+    {
+      "epoch": 7.221556886227545,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1477,
+      "step": 4824
+    },
+    {
+      "epoch": 7.223053892215569,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1529,
+      "step": 4825
+    },
+    {
+      "epoch": 7.224550898203593,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1577,
+      "step": 4826
+    },
+    {
+      "epoch": 7.226047904191617,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1526,
+      "step": 4827
+    },
+    {
+      "epoch": 7.227544910179641,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1559,
+      "step": 4828
+    },
+    {
+      "epoch": 7.229041916167665,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1482,
+      "step": 4829
+    },
+    {
+      "epoch": 7.230538922155689,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 4830
+    },
+    {
+      "epoch": 7.232035928143713,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1556,
+      "step": 4831
+    },
+    {
+      "epoch": 7.233532934131737,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1531,
+      "step": 4832
+    },
+    {
+      "epoch": 7.235029940119761,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1585,
+      "step": 4833
+    },
+    {
+      "epoch": 7.236526946107785,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1571,
+      "step": 4834
+    },
+    {
+      "epoch": 7.2380239520958085,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1588,
+      "step": 4835
+    },
+    {
+      "epoch": 7.2395209580838324,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1494,
+      "step": 4836
+    },
+    {
+      "epoch": 7.241017964071856,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1565,
+      "step": 4837
+    },
+    {
+      "epoch": 7.24251497005988,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1589,
+      "step": 4838
+    },
+    {
+      "epoch": 7.244011976047904,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1554,
+      "step": 4839
+    },
+    {
+      "epoch": 7.245508982035928,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1621,
+      "step": 4840
+    },
+    {
+      "epoch": 7.247005988023952,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1549,
+      "step": 4841
+    },
+    {
+      "epoch": 7.248502994011976,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1494,
+      "step": 4842
+    },
+    {
+      "epoch": 7.25,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.158,
+      "step": 4843
+    },
+    {
+      "epoch": 7.251497005988024,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1578,
+      "step": 4844
+    },
+    {
+      "epoch": 7.252994011976048,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1535,
+      "step": 4845
+    },
+    {
+      "epoch": 7.254491017964072,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1522,
+      "step": 4846
+    },
+    {
+      "epoch": 7.255988023952096,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1529,
+      "step": 4847
+    },
+    {
+      "epoch": 7.25748502994012,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1543,
+      "step": 4848
+    },
+    {
+      "epoch": 7.258982035928144,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1499,
+      "step": 4849
+    },
+    {
+      "epoch": 7.2604790419161676,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1507,
+      "step": 4850
+    },
+    {
+      "epoch": 7.2619760479041915,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1522,
+      "step": 4851
+    },
+    {
+      "epoch": 7.263473053892215,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1579,
+      "step": 4852
+    },
+    {
+      "epoch": 7.264970059880239,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1562,
+      "step": 4853
+    },
+    {
+      "epoch": 7.266467065868263,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1587,
+      "step": 4854
+    },
+    {
+      "epoch": 7.267964071856287,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1567,
+      "step": 4855
+    },
+    {
+      "epoch": 7.269461077844311,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1528,
+      "step": 4856
+    },
+    {
+      "epoch": 7.270958083832335,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 4857
+    },
+    {
+      "epoch": 7.272455089820359,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1577,
+      "step": 4858
+    },
+    {
+      "epoch": 7.273952095808383,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1507,
+      "step": 4859
+    },
+    {
+      "epoch": 7.275449101796407,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1529,
+      "step": 4860
+    },
+    {
+      "epoch": 7.276946107784431,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1519,
+      "step": 4861
+    },
+    {
+      "epoch": 7.278443113772455,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1512,
+      "step": 4862
+    },
+    {
+      "epoch": 7.279940119760479,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1605,
+      "step": 4863
+    },
+    {
+      "epoch": 7.281437125748503,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1574,
+      "step": 4864
+    },
+    {
+      "epoch": 7.282934131736527,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1557,
+      "step": 4865
+    },
+    {
+      "epoch": 7.2844311377245505,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1523,
+      "step": 4866
+    },
+    {
+      "epoch": 7.2859281437125745,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1522,
+      "step": 4867
+    },
+    {
+      "epoch": 7.287425149700598,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1508,
+      "step": 4868
+    },
+    {
+      "epoch": 7.288922155688622,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1553,
+      "step": 4869
+    },
+    {
+      "epoch": 7.290419161676647,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1511,
+      "step": 4870
+    },
+    {
+      "epoch": 7.29191616766467,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1581,
+      "step": 4871
+    },
+    {
+      "epoch": 7.293413173652695,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1597,
+      "step": 4872
+    },
+    {
+      "epoch": 7.294910179640719,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1465,
+      "step": 4873
+    },
+    {
+      "epoch": 7.296407185628743,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1543,
+      "step": 4874
+    },
+    {
+      "epoch": 7.297904191616767,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1465,
+      "step": 4875
+    },
+    {
+      "epoch": 7.299401197604791,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.152,
+      "step": 4876
+    },
+    {
+      "epoch": 7.300898203592815,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1587,
+      "step": 4877
+    },
+    {
+      "epoch": 7.302395209580839,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1497,
+      "step": 4878
+    },
+    {
+      "epoch": 7.303892215568863,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1579,
+      "step": 4879
+    },
+    {
+      "epoch": 7.3053892215568865,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1602,
+      "step": 4880
+    },
+    {
+      "epoch": 7.3068862275449105,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1535,
+      "step": 4881
+    },
+    {
+      "epoch": 7.308383233532934,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1496,
+      "step": 4882
+    },
+    {
+      "epoch": 7.309880239520958,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.155,
+      "step": 4883
+    },
+    {
+      "epoch": 7.311377245508982,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1537,
+      "step": 4884
+    },
+    {
+      "epoch": 7.312874251497006,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1519,
+      "step": 4885
+    },
+    {
+      "epoch": 7.31437125748503,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1578,
+      "step": 4886
+    },
+    {
+      "epoch": 7.315868263473054,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1556,
+      "step": 4887
+    },
+    {
+      "epoch": 7.317365269461078,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1555,
+      "step": 4888
+    },
+    {
+      "epoch": 7.318862275449102,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1554,
+      "step": 4889
+    },
+    {
+      "epoch": 7.320359281437126,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1546,
+      "step": 4890
+    },
+    {
+      "epoch": 7.32185628742515,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1576,
+      "step": 4891
+    },
+    {
+      "epoch": 7.323353293413174,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1531,
+      "step": 4892
+    },
+    {
+      "epoch": 7.324850299401198,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1521,
+      "step": 4893
+    },
+    {
+      "epoch": 7.326347305389222,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1536,
+      "step": 4894
+    },
+    {
+      "epoch": 7.327844311377246,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.15,
+      "step": 4895
+    },
+    {
+      "epoch": 7.3293413173652695,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1538,
+      "step": 4896
+    },
+    {
+      "epoch": 7.330838323353293,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1549,
+      "step": 4897
+    },
+    {
+      "epoch": 7.332335329341317,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.159,
+      "step": 4898
+    },
+    {
+      "epoch": 7.333832335329341,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1529,
+      "step": 4899
+    },
+    {
+      "epoch": 7.335329341317365,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1522,
+      "step": 4900
+    },
+    {
+      "epoch": 7.336826347305389,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1528,
+      "step": 4901
+    },
+    {
+      "epoch": 7.338323353293413,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.158,
+      "step": 4902
+    },
+    {
+      "epoch": 7.339820359281437,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1536,
+      "step": 4903
+    },
+    {
+      "epoch": 7.341317365269461,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0008,
+      "loss": 1.1512,
+      "step": 4904
+    },
+    {
+      "epoch": 7.342814371257485,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1505,
+      "step": 4905
+    },
+    {
+      "epoch": 7.344311377245509,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1591,
+      "step": 4906
+    },
+    {
+      "epoch": 7.345808383233533,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1577,
+      "step": 4907
+    },
+    {
+      "epoch": 7.347305389221557,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1586,
+      "step": 4908
+    },
+    {
+      "epoch": 7.348802395209581,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1479,
+      "step": 4909
+    },
+    {
+      "epoch": 7.350299401197605,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1548,
+      "step": 4910
+    },
+    {
+      "epoch": 7.3517964071856285,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1536,
+      "step": 4911
+    },
+    {
+      "epoch": 7.3532934131736525,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1571,
+      "step": 4912
+    },
+    {
+      "epoch": 7.354790419161676,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1525,
+      "step": 4913
+    },
+    {
+      "epoch": 7.3562874251497,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1569,
+      "step": 4914
+    },
+    {
+      "epoch": 7.357784431137724,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1532,
+      "step": 4915
+    },
+    {
+      "epoch": 7.359281437125748,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1605,
+      "step": 4916
+    },
+    {
+      "epoch": 7.360778443113772,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.155,
+      "step": 4917
+    },
+    {
+      "epoch": 7.362275449101796,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1516,
+      "step": 4918
+    },
+    {
+      "epoch": 7.36377245508982,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1624,
+      "step": 4919
+    },
+    {
+      "epoch": 7.365269461077844,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1643,
+      "step": 4920
+    },
+    {
+      "epoch": 7.366766467065868,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.154,
+      "step": 4921
+    },
+    {
+      "epoch": 7.368263473053892,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1529,
+      "step": 4922
+    },
+    {
+      "epoch": 7.369760479041916,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1547,
+      "step": 4923
+    },
+    {
+      "epoch": 7.37125748502994,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1528,
+      "step": 4924
+    },
+    {
+      "epoch": 7.3727544910179645,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1562,
+      "step": 4925
+    },
+    {
+      "epoch": 7.374251497005988,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1527,
+      "step": 4926
+    },
+    {
+      "epoch": 7.375748502994012,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1542,
+      "step": 4927
+    },
+    {
+      "epoch": 7.3772455089820355,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1605,
+      "step": 4928
+    },
+    {
+      "epoch": 7.37874251497006,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.157,
+      "step": 4929
+    },
+    {
+      "epoch": 7.380239520958084,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1589,
+      "step": 4930
+    },
+    {
+      "epoch": 7.381736526946108,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1571,
+      "step": 4931
+    },
+    {
+      "epoch": 7.383233532934132,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1519,
+      "step": 4932
+    },
+    {
+      "epoch": 7.384730538922156,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1595,
+      "step": 4933
+    },
+    {
+      "epoch": 7.38622754491018,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1609,
+      "step": 4934
+    },
+    {
+      "epoch": 7.387724550898204,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1568,
+      "step": 4935
+    },
+    {
+      "epoch": 7.389221556886228,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1547,
+      "step": 4936
+    },
+    {
+      "epoch": 7.390718562874252,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1574,
+      "step": 4937
+    },
+    {
+      "epoch": 7.392215568862276,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1567,
+      "step": 4938
+    },
+    {
+      "epoch": 7.3937125748503,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1606,
+      "step": 4939
+    },
+    {
+      "epoch": 7.395209580838324,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1541,
+      "step": 4940
+    },
+    {
+      "epoch": 7.3967065868263475,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1536,
+      "step": 4941
+    },
+    {
+      "epoch": 7.3982035928143715,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.156,
+      "step": 4942
+    },
+    {
+      "epoch": 7.399700598802395,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1561,
+      "step": 4943
+    },
+    {
+      "epoch": 7.401197604790419,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1527,
+      "step": 4944
+    },
+    {
+      "epoch": 7.402694610778443,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1549,
+      "step": 4945
+    },
+    {
+      "epoch": 7.404191616766467,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1525,
+      "step": 4946
+    },
+    {
+      "epoch": 7.405688622754491,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1526,
+      "step": 4947
+    },
+    {
+      "epoch": 7.407185628742515,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1601,
+      "step": 4948
+    },
+    {
+      "epoch": 7.408682634730539,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1565,
+      "step": 4949
+    },
+    {
+      "epoch": 7.410179640718563,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1547,
+      "step": 4950
+    },
+    {
+      "epoch": 7.411676646706587,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1625,
+      "step": 4951
+    },
+    {
+      "epoch": 7.413173652694611,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1565,
+      "step": 4952
+    },
+    {
+      "epoch": 7.414670658682635,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1525,
+      "step": 4953
+    },
+    {
+      "epoch": 7.416167664670659,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1553,
+      "step": 4954
+    },
+    {
+      "epoch": 7.417664670658683,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1555,
+      "step": 4955
+    },
+    {
+      "epoch": 7.419161676646707,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1535,
+      "step": 4956
+    },
+    {
+      "epoch": 7.4206586826347305,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1544,
+      "step": 4957
+    },
+    {
+      "epoch": 7.422155688622754,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1524,
+      "step": 4958
+    },
+    {
+      "epoch": 7.423652694610778,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1517,
+      "step": 4959
+    },
+    {
+      "epoch": 7.425149700598802,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1557,
+      "step": 4960
+    },
+    {
+      "epoch": 7.426646706586826,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1564,
+      "step": 4961
+    },
+    {
+      "epoch": 7.42814371257485,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1551,
+      "step": 4962
+    },
+    {
+      "epoch": 7.429640718562874,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1564,
+      "step": 4963
+    },
+    {
+      "epoch": 7.431137724550898,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1555,
+      "step": 4964
+    },
+    {
+      "epoch": 7.432634730538922,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1546,
+      "step": 4965
+    },
+    {
+      "epoch": 7.434131736526946,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1525,
+      "step": 4966
+    },
+    {
+      "epoch": 7.43562874251497,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1527,
+      "step": 4967
+    },
+    {
+      "epoch": 7.437125748502994,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1528,
+      "step": 4968
+    },
+    {
+      "epoch": 7.438622754491018,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1581,
+      "step": 4969
+    },
+    {
+      "epoch": 7.440119760479042,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1575,
+      "step": 4970
+    },
+    {
+      "epoch": 7.441616766467066,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1509,
+      "step": 4971
+    },
+    {
+      "epoch": 7.4431137724550895,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1593,
+      "step": 4972
+    },
+    {
+      "epoch": 7.4446107784431135,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1572,
+      "step": 4973
+    },
+    {
+      "epoch": 7.446107784431137,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1537,
+      "step": 4974
+    },
+    {
+      "epoch": 7.447604790419161,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1556,
+      "step": 4975
+    },
+    {
+      "epoch": 7.449101796407185,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1523,
+      "step": 4976
+    },
+    {
+      "epoch": 7.450598802395209,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1577,
+      "step": 4977
+    },
+    {
+      "epoch": 7.452095808383233,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1539,
+      "step": 4978
+    },
+    {
+      "epoch": 7.453592814371257,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1581,
+      "step": 4979
+    },
+    {
+      "epoch": 7.455089820359281,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1529,
+      "step": 4980
+    },
+    {
+      "epoch": 7.456586826347305,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.154,
+      "step": 4981
+    },
+    {
+      "epoch": 7.45808383233533,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1578,
+      "step": 4982
+    },
+    {
+      "epoch": 7.459580838323353,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1541,
+      "step": 4983
+    },
+    {
+      "epoch": 7.461077844311378,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1547,
+      "step": 4984
+    },
+    {
+      "epoch": 7.462574850299402,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1493,
+      "step": 4985
+    },
+    {
+      "epoch": 7.4640718562874255,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1526,
+      "step": 4986
+    },
+    {
+      "epoch": 7.4655688622754495,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1585,
+      "step": 4987
+    },
+    {
+      "epoch": 7.467065868263473,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.151,
+      "step": 4988
+    },
+    {
+      "epoch": 7.468562874251497,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.148,
+      "step": 4989
+    },
+    {
+      "epoch": 7.470059880239521,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1526,
+      "step": 4990
+    },
+    {
+      "epoch": 7.471556886227545,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1556,
+      "step": 4991
+    },
+    {
+      "epoch": 7.473053892215569,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1493,
+      "step": 4992
+    },
+    {
+      "epoch": 7.474550898203593,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1503,
+      "step": 4993
+    },
+    {
+      "epoch": 7.476047904191617,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1577,
+      "step": 4994
+    },
+    {
+      "epoch": 7.477544910179641,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1559,
+      "step": 4995
+    },
+    {
+      "epoch": 7.479041916167665,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.161,
+      "step": 4996
+    },
+    {
+      "epoch": 7.480538922155689,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1527,
+      "step": 4997
+    },
+    {
+      "epoch": 7.482035928143713,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1498,
+      "step": 4998
+    },
+    {
+      "epoch": 7.483532934131737,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1608,
+      "step": 4999
+    },
+    {
+      "epoch": 7.485029940119761,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1498,
+      "step": 5000
+    },
+    {
+      "epoch": 7.486526946107785,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1603,
+      "step": 5001
+    },
+    {
+      "epoch": 7.4880239520958085,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1504,
+      "step": 5002
+    },
+    {
+      "epoch": 7.4895209580838324,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1586,
+      "step": 5003
+    },
+    {
+      "epoch": 7.491017964071856,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1525,
+      "step": 5004
+    },
+    {
+      "epoch": 7.49251497005988,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1533,
+      "step": 5005
+    },
+    {
+      "epoch": 7.494011976047904,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1571,
+      "step": 5006
+    },
+    {
+      "epoch": 7.495508982035928,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1496,
+      "step": 5007
+    },
+    {
+      "epoch": 7.497005988023952,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1509,
+      "step": 5008
+    },
+    {
+      "epoch": 7.498502994011976,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1565,
+      "step": 5009
+    },
+    {
+      "epoch": 7.5,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1549,
+      "step": 5010
+    },
+    {
+      "epoch": 7.501497005988024,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1522,
+      "step": 5011
+    },
+    {
+      "epoch": 7.502994011976048,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1506,
+      "step": 5012
+    },
+    {
+      "epoch": 7.504491017964072,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1601,
+      "step": 5013
+    },
+    {
+      "epoch": 7.505988023952096,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1518,
+      "step": 5014
+    },
+    {
+      "epoch": 7.50748502994012,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1482,
+      "step": 5015
+    },
+    {
+      "epoch": 7.508982035928144,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1543,
+      "step": 5016
+    },
+    {
+      "epoch": 7.5104790419161676,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1515,
+      "step": 5017
+    },
+    {
+      "epoch": 7.5119760479041915,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1535,
+      "step": 5018
+    },
+    {
+      "epoch": 7.513473053892215,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1588,
+      "step": 5019
+    },
+    {
+      "epoch": 7.514970059880239,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1551,
+      "step": 5020
+    },
+    {
+      "epoch": 7.516467065868263,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1561,
+      "step": 5021
+    },
+    {
+      "epoch": 7.517964071856287,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1597,
+      "step": 5022
+    },
+    {
+      "epoch": 7.519461077844311,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1508,
+      "step": 5023
+    },
+    {
+      "epoch": 7.520958083832335,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1607,
+      "step": 5024
+    },
+    {
+      "epoch": 7.522455089820359,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1558,
+      "step": 5025
+    },
+    {
+      "epoch": 7.523952095808383,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1571,
+      "step": 5026
+    },
+    {
+      "epoch": 7.525449101796407,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 5027
+    },
+    {
+      "epoch": 7.526946107784431,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1558,
+      "step": 5028
+    },
+    {
+      "epoch": 7.528443113772455,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 5029
+    },
+    {
+      "epoch": 7.529940119760479,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1594,
+      "step": 5030
+    },
+    {
+      "epoch": 7.531437125748503,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1532,
+      "step": 5031
+    },
+    {
+      "epoch": 7.532934131736527,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1543,
+      "step": 5032
+    },
+    {
+      "epoch": 7.5344311377245505,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1494,
+      "step": 5033
+    },
+    {
+      "epoch": 7.5359281437125745,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.151,
+      "step": 5034
+    },
+    {
+      "epoch": 7.537425149700599,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1571,
+      "step": 5035
+    },
+    {
+      "epoch": 7.538922155688622,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1534,
+      "step": 5036
+    },
+    {
+      "epoch": 7.540419161676647,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.154,
+      "step": 5037
+    },
+    {
+      "epoch": 7.54191616766467,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.153,
+      "step": 5038
+    },
+    {
+      "epoch": 7.543413173652695,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1552,
+      "step": 5039
+    },
+    {
+      "epoch": 7.544910179640718,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1517,
+      "step": 5040
+    },
+    {
+      "epoch": 7.546407185628743,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1564,
+      "step": 5041
+    },
+    {
+      "epoch": 7.547904191616767,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1511,
+      "step": 5042
+    },
+    {
+      "epoch": 7.549401197604791,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1621,
+      "step": 5043
+    },
+    {
+      "epoch": 7.550898203592815,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1553,
+      "step": 5044
+    },
+    {
+      "epoch": 7.552395209580839,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1554,
+      "step": 5045
+    },
+    {
+      "epoch": 7.553892215568863,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1514,
+      "step": 5046
+    },
+    {
+      "epoch": 7.5553892215568865,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1526,
+      "step": 5047
+    },
+    {
+      "epoch": 7.5568862275449105,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1536,
+      "step": 5048
+    },
+    {
+      "epoch": 7.558383233532934,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1518,
+      "step": 5049
+    },
+    {
+      "epoch": 7.559880239520958,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1516,
+      "step": 5050
+    },
+    {
+      "epoch": 7.561377245508982,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1502,
+      "step": 5051
+    },
+    {
+      "epoch": 7.562874251497006,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1632,
+      "step": 5052
+    },
+    {
+      "epoch": 7.56437125748503,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1573,
+      "step": 5053
+    },
+    {
+      "epoch": 7.565868263473054,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1562,
+      "step": 5054
+    },
+    {
+      "epoch": 7.567365269461078,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1554,
+      "step": 5055
+    },
+    {
+      "epoch": 7.568862275449102,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1486,
+      "step": 5056
+    },
+    {
+      "epoch": 7.570359281437126,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1515,
+      "step": 5057
+    },
+    {
+      "epoch": 7.57185628742515,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1573,
+      "step": 5058
+    },
+    {
+      "epoch": 7.573353293413174,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1492,
+      "step": 5059
+    },
+    {
+      "epoch": 7.574850299401198,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1541,
+      "step": 5060
+    },
+    {
+      "epoch": 7.576347305389222,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1565,
+      "step": 5061
+    },
+    {
+      "epoch": 7.577844311377246,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1528,
+      "step": 5062
+    },
+    {
+      "epoch": 7.5793413173652695,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1573,
+      "step": 5063
+    },
+    {
+      "epoch": 7.580838323353293,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1485,
+      "step": 5064
+    },
+    {
+      "epoch": 7.582335329341317,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.158,
+      "step": 5065
+    },
+    {
+      "epoch": 7.583832335329341,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1546,
+      "step": 5066
+    },
+    {
+      "epoch": 7.585329341317365,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1594,
+      "step": 5067
+    },
+    {
+      "epoch": 7.586826347305389,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1464,
+      "step": 5068
+    },
+    {
+      "epoch": 7.588323353293413,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1609,
+      "step": 5069
+    },
+    {
+      "epoch": 7.589820359281437,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1507,
+      "step": 5070
+    },
+    {
+      "epoch": 7.591317365269461,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1529,
+      "step": 5071
+    },
+    {
+      "epoch": 7.592814371257485,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1569,
+      "step": 5072
+    },
+    {
+      "epoch": 7.594311377245509,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1539,
+      "step": 5073
+    },
+    {
+      "epoch": 7.595808383233533,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1567,
+      "step": 5074
+    },
+    {
+      "epoch": 7.597305389221557,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1541,
+      "step": 5075
+    },
+    {
+      "epoch": 7.598802395209581,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1552,
+      "step": 5076
+    },
+    {
+      "epoch": 7.600299401197605,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1561,
+      "step": 5077
+    },
+    {
+      "epoch": 7.6017964071856285,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1497,
+      "step": 5078
+    },
+    {
+      "epoch": 7.6032934131736525,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1526,
+      "step": 5079
+    },
+    {
+      "epoch": 7.604790419161676,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1509,
+      "step": 5080
+    },
+    {
+      "epoch": 7.6062874251497,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1587,
+      "step": 5081
+    },
+    {
+      "epoch": 7.607784431137724,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1575,
+      "step": 5082
+    },
+    {
+      "epoch": 7.609281437125748,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1558,
+      "step": 5083
+    },
+    {
+      "epoch": 7.610778443113772,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1531,
+      "step": 5084
+    },
+    {
+      "epoch": 7.612275449101796,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1543,
+      "step": 5085
+    },
+    {
+      "epoch": 7.61377245508982,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1513,
+      "step": 5086
+    },
+    {
+      "epoch": 7.615269461077844,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1526,
+      "step": 5087
+    },
+    {
+      "epoch": 7.616766467065868,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1542,
+      "step": 5088
+    },
+    {
+      "epoch": 7.618263473053892,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1566,
+      "step": 5089
+    },
+    {
+      "epoch": 7.619760479041916,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1576,
+      "step": 5090
+    },
+    {
+      "epoch": 7.62125748502994,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1569,
+      "step": 5091
+    },
+    {
+      "epoch": 7.6227544910179645,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1499,
+      "step": 5092
+    },
+    {
+      "epoch": 7.624251497005988,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1519,
+      "step": 5093
+    },
+    {
+      "epoch": 7.625748502994012,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1571,
+      "step": 5094
+    },
+    {
+      "epoch": 7.6272455089820355,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1593,
+      "step": 5095
+    },
+    {
+      "epoch": 7.62874251497006,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1534,
+      "step": 5096
+    },
+    {
+      "epoch": 7.630239520958084,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1558,
+      "step": 5097
+    },
+    {
+      "epoch": 7.631736526946108,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1581,
+      "step": 5098
+    },
+    {
+      "epoch": 7.633233532934132,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1498,
+      "step": 5099
+    },
+    {
+      "epoch": 7.634730538922156,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1544,
+      "step": 5100
+    },
+    {
+      "epoch": 7.63622754491018,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1543,
+      "step": 5101
+    },
+    {
+      "epoch": 7.637724550898204,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1543,
+      "step": 5102
+    },
+    {
+      "epoch": 7.639221556886228,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1584,
+      "step": 5103
+    },
+    {
+      "epoch": 7.640718562874252,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1593,
+      "step": 5104
+    },
+    {
+      "epoch": 7.642215568862276,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1502,
+      "step": 5105
+    },
+    {
+      "epoch": 7.6437125748503,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1489,
+      "step": 5106
+    },
+    {
+      "epoch": 7.645209580838324,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1513,
+      "step": 5107
+    },
+    {
+      "epoch": 7.6467065868263475,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1491,
+      "step": 5108
+    },
+    {
+      "epoch": 7.6482035928143715,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1459,
+      "step": 5109
+    },
+    {
+      "epoch": 7.649700598802395,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.153,
+      "step": 5110
+    },
+    {
+      "epoch": 7.651197604790419,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1512,
+      "step": 5111
+    },
+    {
+      "epoch": 7.652694610778443,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1509,
+      "step": 5112
+    },
+    {
+      "epoch": 7.654191616766467,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1611,
+      "step": 5113
+    },
+    {
+      "epoch": 7.655688622754491,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1487,
+      "step": 5114
+    },
+    {
+      "epoch": 7.657185628742515,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1462,
+      "step": 5115
+    },
+    {
+      "epoch": 7.658682634730539,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.149,
+      "step": 5116
+    },
+    {
+      "epoch": 7.660179640718563,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1611,
+      "step": 5117
+    },
+    {
+      "epoch": 7.661676646706587,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1514,
+      "step": 5118
+    },
+    {
+      "epoch": 7.663173652694611,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1589,
+      "step": 5119
+    },
+    {
+      "epoch": 7.664670658682635,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1586,
+      "step": 5120
+    },
+    {
+      "epoch": 7.666167664670659,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1599,
+      "step": 5121
+    },
+    {
+      "epoch": 7.667664670658683,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1518,
+      "step": 5122
+    },
+    {
+      "epoch": 7.669161676646707,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 5123
+    },
+    {
+      "epoch": 7.6706586826347305,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1525,
+      "step": 5124
+    },
+    {
+      "epoch": 7.672155688622754,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 5125
+    },
+    {
+      "epoch": 7.673652694610778,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1471,
+      "step": 5126
+    },
+    {
+      "epoch": 7.675149700598802,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1518,
+      "step": 5127
+    },
+    {
+      "epoch": 7.676646706586826,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 5128
+    },
+    {
+      "epoch": 7.67814371257485,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1522,
+      "step": 5129
+    },
+    {
+      "epoch": 7.679640718562874,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1531,
+      "step": 5130
+    },
+    {
+      "epoch": 7.681137724550898,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1545,
+      "step": 5131
+    },
+    {
+      "epoch": 7.682634730538922,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1594,
+      "step": 5132
+    },
+    {
+      "epoch": 7.684131736526946,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1525,
+      "step": 5133
+    },
+    {
+      "epoch": 7.68562874251497,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1554,
+      "step": 5134
+    },
+    {
+      "epoch": 7.687125748502994,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 5135
+    },
+    {
+      "epoch": 7.688622754491018,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1553,
+      "step": 5136
+    },
+    {
+      "epoch": 7.690119760479042,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1538,
+      "step": 5137
+    },
+    {
+      "epoch": 7.691616766467066,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1503,
+      "step": 5138
+    },
+    {
+      "epoch": 7.6931137724550895,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1522,
+      "step": 5139
+    },
+    {
+      "epoch": 7.6946107784431135,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1517,
+      "step": 5140
+    },
+    {
+      "epoch": 7.696107784431137,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1503,
+      "step": 5141
+    },
+    {
+      "epoch": 7.697604790419161,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1572,
+      "step": 5142
+    },
+    {
+      "epoch": 7.699101796407185,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1553,
+      "step": 5143
+    },
+    {
+      "epoch": 7.700598802395209,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1573,
+      "step": 5144
+    },
+    {
+      "epoch": 7.702095808383233,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1563,
+      "step": 5145
+    },
+    {
+      "epoch": 7.703592814371257,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1511,
+      "step": 5146
+    },
+    {
+      "epoch": 7.705089820359282,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1617,
+      "step": 5147
+    },
+    {
+      "epoch": 7.706586826347305,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1532,
+      "step": 5148
+    },
+    {
+      "epoch": 7.70808383233533,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1493,
+      "step": 5149
+    },
+    {
+      "epoch": 7.709580838323353,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1498,
+      "step": 5150
+    },
+    {
+      "epoch": 7.711077844311378,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1504,
+      "step": 5151
+    },
+    {
+      "epoch": 7.712574850299401,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1604,
+      "step": 5152
+    },
+    {
+      "epoch": 7.7140718562874255,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1522,
+      "step": 5153
+    },
+    {
+      "epoch": 7.7155688622754495,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1548,
+      "step": 5154
+    },
+    {
+      "epoch": 7.717065868263473,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1512,
+      "step": 5155
+    },
+    {
+      "epoch": 7.718562874251497,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1542,
+      "step": 5156
+    },
+    {
+      "epoch": 7.720059880239521,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1612,
+      "step": 5157
+    },
+    {
+      "epoch": 7.721556886227545,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1566,
+      "step": 5158
+    },
+    {
+      "epoch": 7.723053892215569,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1565,
+      "step": 5159
+    },
+    {
+      "epoch": 7.724550898203593,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1512,
+      "step": 5160
+    },
+    {
+      "epoch": 7.726047904191617,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1535,
+      "step": 5161
+    },
+    {
+      "epoch": 7.727544910179641,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1553,
+      "step": 5162
+    },
+    {
+      "epoch": 7.729041916167665,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1561,
+      "step": 5163
+    },
+    {
+      "epoch": 7.730538922155689,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1491,
+      "step": 5164
+    },
+    {
+      "epoch": 7.732035928143713,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1479,
+      "step": 5165
+    },
+    {
+      "epoch": 7.733532934131737,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1564,
+      "step": 5166
+    },
+    {
+      "epoch": 7.735029940119761,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1539,
+      "step": 5167
+    },
+    {
+      "epoch": 7.736526946107785,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1571,
+      "step": 5168
+    },
+    {
+      "epoch": 7.7380239520958085,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1563,
+      "step": 5169
+    },
+    {
+      "epoch": 7.7395209580838324,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1483,
+      "step": 5170
+    },
+    {
+      "epoch": 7.741017964071856,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1503,
+      "step": 5171
+    },
+    {
+      "epoch": 7.74251497005988,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1554,
+      "step": 5172
+    },
+    {
+      "epoch": 7.744011976047904,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1518,
+      "step": 5173
+    },
+    {
+      "epoch": 7.745508982035928,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1496,
+      "step": 5174
+    },
+    {
+      "epoch": 7.747005988023952,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1557,
+      "step": 5175
+    },
+    {
+      "epoch": 7.748502994011976,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1551,
+      "step": 5176
+    },
+    {
+      "epoch": 7.75,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1593,
+      "step": 5177
+    },
+    {
+      "epoch": 7.751497005988024,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 5178
+    },
+    {
+      "epoch": 7.752994011976048,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1509,
+      "step": 5179
+    },
+    {
+      "epoch": 7.754491017964072,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1499,
+      "step": 5180
+    },
+    {
+      "epoch": 7.755988023952096,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.146,
+      "step": 5181
+    },
+    {
+      "epoch": 7.75748502994012,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1512,
+      "step": 5182
+    },
+    {
+      "epoch": 7.758982035928144,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1612,
+      "step": 5183
+    },
+    {
+      "epoch": 7.7604790419161676,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1576,
+      "step": 5184
+    },
+    {
+      "epoch": 7.7619760479041915,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1499,
+      "step": 5185
+    },
+    {
+      "epoch": 7.763473053892215,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1542,
+      "step": 5186
+    },
+    {
+      "epoch": 7.764970059880239,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1563,
+      "step": 5187
+    },
+    {
+      "epoch": 7.766467065868263,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1486,
+      "step": 5188
+    },
+    {
+      "epoch": 7.767964071856287,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1566,
+      "step": 5189
+    },
+    {
+      "epoch": 7.769461077844311,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1623,
+      "step": 5190
+    },
+    {
+      "epoch": 7.770958083832335,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1573,
+      "step": 5191
+    },
+    {
+      "epoch": 7.772455089820359,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1501,
+      "step": 5192
+    },
+    {
+      "epoch": 7.773952095808383,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1585,
+      "step": 5193
+    },
+    {
+      "epoch": 7.775449101796407,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1511,
+      "step": 5194
+    },
+    {
+      "epoch": 7.776946107784431,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1552,
+      "step": 5195
+    },
+    {
+      "epoch": 7.778443113772455,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1528,
+      "step": 5196
+    },
+    {
+      "epoch": 7.779940119760479,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1542,
+      "step": 5197
+    },
+    {
+      "epoch": 7.781437125748503,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1513,
+      "step": 5198
+    },
+    {
+      "epoch": 7.782934131736527,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1529,
+      "step": 5199
+    },
+    {
+      "epoch": 7.7844311377245505,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1542,
+      "step": 5200
+    },
+    {
+      "epoch": 7.7859281437125745,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.155,
+      "step": 5201
+    },
+    {
+      "epoch": 7.787425149700599,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1542,
+      "step": 5202
+    },
+    {
+      "epoch": 7.788922155688622,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1554,
+      "step": 5203
+    },
+    {
+      "epoch": 7.790419161676647,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 5204
+    },
+    {
+      "epoch": 7.79191616766467,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1557,
+      "step": 5205
+    },
+    {
+      "epoch": 7.793413173652695,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1521,
+      "step": 5206
+    },
+    {
+      "epoch": 7.794910179640718,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1564,
+      "step": 5207
+    },
+    {
+      "epoch": 7.796407185628743,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.146,
+      "step": 5208
+    },
+    {
+      "epoch": 7.797904191616767,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1574,
+      "step": 5209
+    },
+    {
+      "epoch": 7.799401197604791,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1561,
+      "step": 5210
+    },
+    {
+      "epoch": 7.800898203592815,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1508,
+      "step": 5211
+    },
+    {
+      "epoch": 7.802395209580839,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1517,
+      "step": 5212
+    },
+    {
+      "epoch": 7.803892215568863,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1525,
+      "step": 5213
+    },
+    {
+      "epoch": 7.8053892215568865,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1514,
+      "step": 5214
+    },
+    {
+      "epoch": 7.8068862275449105,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1517,
+      "step": 5215
+    },
+    {
+      "epoch": 7.808383233532934,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 5216
+    },
+    {
+      "epoch": 7.809880239520958,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1537,
+      "step": 5217
+    },
+    {
+      "epoch": 7.811377245508982,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1505,
+      "step": 5218
+    },
+    {
+      "epoch": 7.812874251497006,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1565,
+      "step": 5219
+    },
+    {
+      "epoch": 7.81437125748503,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1511,
+      "step": 5220
+    },
+    {
+      "epoch": 7.815868263473054,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1471,
+      "step": 5221
+    },
+    {
+      "epoch": 7.817365269461078,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1552,
+      "step": 5222
+    },
+    {
+      "epoch": 7.818862275449102,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 5223
+    },
+    {
+      "epoch": 7.820359281437126,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1499,
+      "step": 5224
+    },
+    {
+      "epoch": 7.82185628742515,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1558,
+      "step": 5225
+    },
+    {
+      "epoch": 7.823353293413174,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1515,
+      "step": 5226
+    },
+    {
+      "epoch": 7.824850299401198,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1523,
+      "step": 5227
+    },
+    {
+      "epoch": 7.826347305389222,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1542,
+      "step": 5228
+    },
+    {
+      "epoch": 7.827844311377246,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1532,
+      "step": 5229
+    },
+    {
+      "epoch": 7.8293413173652695,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.152,
+      "step": 5230
+    },
+    {
+      "epoch": 7.830838323353293,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1518,
+      "step": 5231
+    },
+    {
+      "epoch": 7.832335329341317,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.151,
+      "step": 5232
+    },
+    {
+      "epoch": 7.833832335329341,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1586,
+      "step": 5233
+    },
+    {
+      "epoch": 7.835329341317365,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1484,
+      "step": 5234
+    },
+    {
+      "epoch": 7.836826347305389,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1487,
+      "step": 5235
+    },
+    {
+      "epoch": 7.838323353293413,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1498,
+      "step": 5236
+    },
+    {
+      "epoch": 7.839820359281437,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1504,
+      "step": 5237
+    },
+    {
+      "epoch": 7.841317365269461,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1531,
+      "step": 5238
+    },
+    {
+      "epoch": 7.842814371257485,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1551,
+      "step": 5239
+    },
+    {
+      "epoch": 7.844311377245509,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 5240
+    },
+    {
+      "epoch": 7.845808383233533,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1506,
+      "step": 5241
+    },
+    {
+      "epoch": 7.847305389221557,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1521,
+      "step": 5242
+    },
+    {
+      "epoch": 7.848802395209581,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1473,
+      "step": 5243
+    },
+    {
+      "epoch": 7.850299401197605,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1546,
+      "step": 5244
+    },
+    {
+      "epoch": 7.8517964071856285,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1568,
+      "step": 5245
+    },
+    {
+      "epoch": 7.8532934131736525,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1507,
+      "step": 5246
+    },
+    {
+      "epoch": 7.854790419161676,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 5247
+    },
+    {
+      "epoch": 7.8562874251497,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1583,
+      "step": 5248
+    },
+    {
+      "epoch": 7.857784431137724,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1497,
+      "step": 5249
+    },
+    {
+      "epoch": 7.859281437125748,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 5250
+    },
+    {
+      "epoch": 7.860778443113772,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1493,
+      "step": 5251
+    },
+    {
+      "epoch": 7.862275449101796,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1561,
+      "step": 5252
+    },
+    {
+      "epoch": 7.86377245508982,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1557,
+      "step": 5253
+    },
+    {
+      "epoch": 7.865269461077844,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1601,
+      "step": 5254
+    },
+    {
+      "epoch": 7.866766467065868,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1528,
+      "step": 5255
+    },
+    {
+      "epoch": 7.868263473053892,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1501,
+      "step": 5256
+    },
+    {
+      "epoch": 7.869760479041916,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1586,
+      "step": 5257
+    },
+    {
+      "epoch": 7.87125748502994,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1545,
+      "step": 5258
+    },
+    {
+      "epoch": 7.8727544910179645,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1576,
+      "step": 5259
+    },
+    {
+      "epoch": 7.874251497005988,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1532,
+      "step": 5260
+    },
+    {
+      "epoch": 7.875748502994012,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1481,
+      "step": 5261
+    },
+    {
+      "epoch": 7.8772455089820355,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1514,
+      "step": 5262
+    },
+    {
+      "epoch": 7.87874251497006,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1514,
+      "step": 5263
+    },
+    {
+      "epoch": 7.880239520958084,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1575,
+      "step": 5264
+    },
+    {
+      "epoch": 7.881736526946108,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1581,
+      "step": 5265
+    },
+    {
+      "epoch": 7.883233532934132,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1552,
+      "step": 5266
+    },
+    {
+      "epoch": 7.884730538922156,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1505,
+      "step": 5267
+    },
+    {
+      "epoch": 7.88622754491018,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1509,
+      "step": 5268
+    },
+    {
+      "epoch": 7.887724550898204,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.152,
+      "step": 5269
+    },
+    {
+      "epoch": 7.889221556886228,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1553,
+      "step": 5270
+    },
+    {
+      "epoch": 7.890718562874252,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1582,
+      "step": 5271
+    },
+    {
+      "epoch": 7.892215568862276,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.156,
+      "step": 5272
+    },
+    {
+      "epoch": 7.8937125748503,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1514,
+      "step": 5273
+    },
+    {
+      "epoch": 7.895209580838324,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1585,
+      "step": 5274
+    },
+    {
+      "epoch": 7.8967065868263475,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1558,
+      "step": 5275
+    },
+    {
+      "epoch": 7.8982035928143715,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1551,
+      "step": 5276
+    },
+    {
+      "epoch": 7.899700598802395,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 5277
+    },
+    {
+      "epoch": 7.901197604790419,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1516,
+      "step": 5278
+    },
+    {
+      "epoch": 7.902694610778443,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1582,
+      "step": 5279
+    },
+    {
+      "epoch": 7.904191616766467,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.153,
+      "step": 5280
+    },
+    {
+      "epoch": 7.905688622754491,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1525,
+      "step": 5281
+    },
+    {
+      "epoch": 7.907185628742515,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1518,
+      "step": 5282
+    },
+    {
+      "epoch": 7.908682634730539,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1496,
+      "step": 5283
+    },
+    {
+      "epoch": 7.910179640718563,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1518,
+      "step": 5284
+    },
+    {
+      "epoch": 7.911676646706587,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1541,
+      "step": 5285
+    },
+    {
+      "epoch": 7.913173652694611,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1513,
+      "step": 5286
+    },
+    {
+      "epoch": 7.914670658682635,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1522,
+      "step": 5287
+    },
+    {
+      "epoch": 7.916167664670659,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1534,
+      "step": 5288
+    },
+    {
+      "epoch": 7.917664670658683,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1559,
+      "step": 5289
+    },
+    {
+      "epoch": 7.919161676646707,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1511,
+      "step": 5290
+    },
+    {
+      "epoch": 7.9206586826347305,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1527,
+      "step": 5291
+    },
+    {
+      "epoch": 7.922155688622754,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1553,
+      "step": 5292
+    },
+    {
+      "epoch": 7.923652694610778,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1525,
+      "step": 5293
+    },
+    {
+      "epoch": 7.925149700598802,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1488,
+      "step": 5294
+    },
+    {
+      "epoch": 7.926646706586826,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1505,
+      "step": 5295
+    },
+    {
+      "epoch": 7.92814371257485,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.152,
+      "step": 5296
+    },
+    {
+      "epoch": 7.929640718562874,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1554,
+      "step": 5297
+    },
+    {
+      "epoch": 7.931137724550898,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1462,
+      "step": 5298
+    },
+    {
+      "epoch": 7.932634730538922,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 5299
+    },
+    {
+      "epoch": 7.934131736526946,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1528,
+      "step": 5300
+    },
+    {
+      "epoch": 7.93562874251497,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1516,
+      "step": 5301
+    },
+    {
+      "epoch": 7.937125748502994,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 5302
+    },
+    {
+      "epoch": 7.938622754491018,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 5303
+    },
+    {
+      "epoch": 7.940119760479042,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1446,
+      "step": 5304
+    },
+    {
+      "epoch": 7.941616766467066,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1502,
+      "step": 5305
+    },
+    {
+      "epoch": 7.9431137724550895,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1485,
+      "step": 5306
+    },
+    {
+      "epoch": 7.9446107784431135,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1468,
+      "step": 5307
+    },
+    {
+      "epoch": 7.946107784431137,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1496,
+      "step": 5308
+    },
+    {
+      "epoch": 7.947604790419161,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1501,
+      "step": 5309
+    },
+    {
+      "epoch": 7.949101796407185,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1516,
+      "step": 5310
+    },
+    {
+      "epoch": 7.950598802395209,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.152,
+      "step": 5311
+    },
+    {
+      "epoch": 7.952095808383233,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1606,
+      "step": 5312
+    },
+    {
+      "epoch": 7.953592814371257,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1541,
+      "step": 5313
+    },
+    {
+      "epoch": 7.955089820359282,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1521,
+      "step": 5314
+    },
+    {
+      "epoch": 7.956586826347305,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 5315
+    },
+    {
+      "epoch": 7.95808383233533,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1562,
+      "step": 5316
+    },
+    {
+      "epoch": 7.959580838323353,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.154,
+      "step": 5317
+    },
+    {
+      "epoch": 7.961077844311378,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1518,
+      "step": 5318
+    },
+    {
+      "epoch": 7.962574850299401,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1548,
+      "step": 5319
+    },
+    {
+      "epoch": 7.9640718562874255,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1507,
+      "step": 5320
+    },
+    {
+      "epoch": 7.9655688622754495,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1539,
+      "step": 5321
+    },
+    {
+      "epoch": 7.967065868263473,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1498,
+      "step": 5322
+    },
+    {
+      "epoch": 7.968562874251497,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1556,
+      "step": 5323
+    },
+    {
+      "epoch": 7.970059880239521,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1482,
+      "step": 5324
+    },
+    {
+      "epoch": 7.971556886227545,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1545,
+      "step": 5325
+    },
+    {
+      "epoch": 7.973053892215569,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1581,
+      "step": 5326
+    },
+    {
+      "epoch": 7.974550898203593,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1533,
+      "step": 5327
+    },
+    {
+      "epoch": 7.976047904191617,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1531,
+      "step": 5328
+    },
+    {
+      "epoch": 7.977544910179641,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1522,
+      "step": 5329
+    },
+    {
+      "epoch": 7.979041916167665,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1603,
+      "step": 5330
+    },
+    {
+      "epoch": 7.980538922155689,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1488,
+      "step": 5331
+    },
+    {
+      "epoch": 7.982035928143713,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.153,
+      "step": 5332
+    },
+    {
+      "epoch": 7.983532934131737,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1584,
+      "step": 5333
+    },
+    {
+      "epoch": 7.985029940119761,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 5334
+    },
+    {
+      "epoch": 7.986526946107785,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1529,
+      "step": 5335
+    },
+    {
+      "epoch": 7.9880239520958085,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1563,
+      "step": 5336
+    },
+    {
+      "epoch": 7.9895209580838324,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1513,
+      "step": 5337
+    },
+    {
+      "epoch": 7.991017964071856,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1549,
+      "step": 5338
+    },
+    {
+      "epoch": 7.99251497005988,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1519,
+      "step": 5339
+    },
+    {
+      "epoch": 7.994011976047904,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1513,
+      "step": 5340
+    },
+    {
+      "epoch": 7.995508982035928,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1549,
+      "step": 5341
+    },
+    {
+      "epoch": 7.997005988023952,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 5342
+    },
+    {
+      "epoch": 7.998502994011976,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 5343
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1479,
+      "step": 5344
+    },
+    {
+      "epoch": 8.001497005988025,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 5345
+    },
+    {
+      "epoch": 8.002994011976048,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1516,
+      "step": 5346
+    },
+    {
+      "epoch": 8.004491017964073,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1493,
+      "step": 5347
+    },
+    {
+      "epoch": 8.005988023952096,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 5348
+    },
+    {
+      "epoch": 8.00748502994012,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1553,
+      "step": 5349
+    },
+    {
+      "epoch": 8.008982035928144,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1521,
+      "step": 5350
+    },
+    {
+      "epoch": 8.010479041916168,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1549,
+      "step": 5351
+    },
+    {
+      "epoch": 8.011976047904191,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1572,
+      "step": 5352
+    },
+    {
+      "epoch": 8.013473053892216,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1505,
+      "step": 5353
+    },
+    {
+      "epoch": 8.01497005988024,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1518,
+      "step": 5354
+    },
+    {
+      "epoch": 8.016467065868264,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1525,
+      "step": 5355
+    },
+    {
+      "epoch": 8.017964071856287,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1541,
+      "step": 5356
+    },
+    {
+      "epoch": 8.019461077844312,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1514,
+      "step": 5357
+    },
+    {
+      "epoch": 8.020958083832335,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1521,
+      "step": 5358
+    },
+    {
+      "epoch": 8.02245508982036,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1489,
+      "step": 5359
+    },
+    {
+      "epoch": 8.023952095808383,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1494,
+      "step": 5360
+    },
+    {
+      "epoch": 8.025449101796408,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1511,
+      "step": 5361
+    },
+    {
+      "epoch": 8.02694610778443,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1567,
+      "step": 5362
+    },
+    {
+      "epoch": 8.028443113772456,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1485,
+      "step": 5363
+    },
+    {
+      "epoch": 8.029940119760479,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1529,
+      "step": 5364
+    },
+    {
+      "epoch": 8.031437125748504,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1472,
+      "step": 5365
+    },
+    {
+      "epoch": 8.032934131736527,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.156,
+      "step": 5366
+    },
+    {
+      "epoch": 8.034431137724551,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1605,
+      "step": 5367
+    },
+    {
+      "epoch": 8.035928143712574,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1501,
+      "step": 5368
+    },
+    {
+      "epoch": 8.0374251497006,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.15,
+      "step": 5369
+    },
+    {
+      "epoch": 8.038922155688622,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1559,
+      "step": 5370
+    },
+    {
+      "epoch": 8.040419161676647,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1546,
+      "step": 5371
+    },
+    {
+      "epoch": 8.04191616766467,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 5372
+    },
+    {
+      "epoch": 8.043413173652695,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1459,
+      "step": 5373
+    },
+    {
+      "epoch": 8.044910179640718,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1536,
+      "step": 5374
+    },
+    {
+      "epoch": 8.046407185628743,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 5375
+    },
+    {
+      "epoch": 8.047904191616766,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1511,
+      "step": 5376
+    },
+    {
+      "epoch": 8.04940119760479,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1578,
+      "step": 5377
+    },
+    {
+      "epoch": 8.050898203592814,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1604,
+      "step": 5378
+    },
+    {
+      "epoch": 8.052395209580839,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1497,
+      "step": 5379
+    },
+    {
+      "epoch": 8.053892215568862,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.144,
+      "step": 5380
+    },
+    {
+      "epoch": 8.055389221556887,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1554,
+      "step": 5381
+    },
+    {
+      "epoch": 8.05688622754491,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1545,
+      "step": 5382
+    },
+    {
+      "epoch": 8.058383233532934,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.158,
+      "step": 5383
+    },
+    {
+      "epoch": 8.059880239520957,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1507,
+      "step": 5384
+    },
+    {
+      "epoch": 8.061377245508982,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1529,
+      "step": 5385
+    },
+    {
+      "epoch": 8.062874251497005,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1544,
+      "step": 5386
+    },
+    {
+      "epoch": 8.06437125748503,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 5387
+    },
+    {
+      "epoch": 8.065868263473053,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1517,
+      "step": 5388
+    },
+    {
+      "epoch": 8.067365269461078,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 5389
+    },
+    {
+      "epoch": 8.068862275449101,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 5390
+    },
+    {
+      "epoch": 8.070359281437126,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 5391
+    },
+    {
+      "epoch": 8.071856287425149,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1503,
+      "step": 5392
+    },
+    {
+      "epoch": 8.073353293413174,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1553,
+      "step": 5393
+    },
+    {
+      "epoch": 8.074850299401197,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1554,
+      "step": 5394
+    },
+    {
+      "epoch": 8.076347305389222,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1499,
+      "step": 5395
+    },
+    {
+      "epoch": 8.077844311377245,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1489,
+      "step": 5396
+    },
+    {
+      "epoch": 8.07934131736527,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.149,
+      "step": 5397
+    },
+    {
+      "epoch": 8.080838323353293,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1492,
+      "step": 5398
+    },
+    {
+      "epoch": 8.082335329341317,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1553,
+      "step": 5399
+    },
+    {
+      "epoch": 8.08383233532934,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1495,
+      "step": 5400
+    },
+    {
+      "epoch": 8.085329341317365,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1517,
+      "step": 5401
+    },
+    {
+      "epoch": 8.08682634730539,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1506,
+      "step": 5402
+    },
+    {
+      "epoch": 8.088323353293413,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 5403
+    },
+    {
+      "epoch": 8.089820359281438,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1527,
+      "step": 5404
+    },
+    {
+      "epoch": 8.091317365269461,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 5405
+    },
+    {
+      "epoch": 8.092814371257486,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1481,
+      "step": 5406
+    },
+    {
+      "epoch": 8.094311377245509,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1546,
+      "step": 5407
+    },
+    {
+      "epoch": 8.095808383233534,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1541,
+      "step": 5408
+    },
+    {
+      "epoch": 8.097305389221557,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 5409
+    },
+    {
+      "epoch": 8.098802395209582,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1569,
+      "step": 5410
+    },
+    {
+      "epoch": 8.100299401197605,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1499,
+      "step": 5411
+    },
+    {
+      "epoch": 8.10179640718563,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1511,
+      "step": 5412
+    },
+    {
+      "epoch": 8.103293413173652,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1558,
+      "step": 5413
+    },
+    {
+      "epoch": 8.104790419161677,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1492,
+      "step": 5414
+    },
+    {
+      "epoch": 8.1062874251497,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1491,
+      "step": 5415
+    },
+    {
+      "epoch": 8.107784431137725,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1527,
+      "step": 5416
+    },
+    {
+      "epoch": 8.109281437125748,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1494,
+      "step": 5417
+    },
+    {
+      "epoch": 8.110778443113773,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1574,
+      "step": 5418
+    },
+    {
+      "epoch": 8.112275449101796,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1582,
+      "step": 5419
+    },
+    {
+      "epoch": 8.113772455089821,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1508,
+      "step": 5420
+    },
+    {
+      "epoch": 8.115269461077844,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1577,
+      "step": 5421
+    },
+    {
+      "epoch": 8.116766467065869,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 5422
+    },
+    {
+      "epoch": 8.118263473053892,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1602,
+      "step": 5423
+    },
+    {
+      "epoch": 8.119760479041917,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1535,
+      "step": 5424
+    },
+    {
+      "epoch": 8.12125748502994,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 5425
+    },
+    {
+      "epoch": 8.122754491017965,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1475,
+      "step": 5426
+    },
+    {
+      "epoch": 8.124251497005988,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.152,
+      "step": 5427
+    },
+    {
+      "epoch": 8.125748502994012,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.152,
+      "step": 5428
+    },
+    {
+      "epoch": 8.127245508982035,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1501,
+      "step": 5429
+    },
+    {
+      "epoch": 8.12874251497006,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1518,
+      "step": 5430
+    },
+    {
+      "epoch": 8.130239520958083,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1504,
+      "step": 5431
+    },
+    {
+      "epoch": 8.131736526946108,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1496,
+      "step": 5432
+    },
+    {
+      "epoch": 8.133233532934131,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1576,
+      "step": 5433
+    },
+    {
+      "epoch": 8.134730538922156,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1551,
+      "step": 5434
+    },
+    {
+      "epoch": 8.136227544910179,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1482,
+      "step": 5435
+    },
+    {
+      "epoch": 8.137724550898204,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.155,
+      "step": 5436
+    },
+    {
+      "epoch": 8.139221556886227,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1517,
+      "step": 5437
+    },
+    {
+      "epoch": 8.140718562874252,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1565,
+      "step": 5438
+    },
+    {
+      "epoch": 8.142215568862275,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1551,
+      "step": 5439
+    },
+    {
+      "epoch": 8.1437125748503,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1553,
+      "step": 5440
+    },
+    {
+      "epoch": 8.145209580838323,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1542,
+      "step": 5441
+    },
+    {
+      "epoch": 8.146706586826348,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0008,
+      "loss": 1.1539,
+      "step": 5442
+    },
+    {
+      "epoch": 8.14820359281437,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0008,
+      "loss": 1.157,
+      "step": 5443
+    },
+    {
+      "epoch": 8.149700598802395,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.152,
+      "step": 5444
+    },
+    {
+      "epoch": 8.151197604790418,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1522,
+      "step": 5445
+    },
+    {
+      "epoch": 8.152694610778443,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1497,
+      "step": 5446
+    },
+    {
+      "epoch": 8.154191616766466,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0008,
+      "loss": 1.158,
+      "step": 5447
+    },
+    {
+      "epoch": 8.155688622754491,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1523,
+      "step": 5448
+    },
+    {
+      "epoch": 8.157185628742514,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1533,
+      "step": 5449
+    },
+    {
+      "epoch": 8.158682634730539,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0008,
+      "loss": 1.1855,
+      "step": 5450
+    },
+    {
+      "epoch": 8.160179640718562,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1727,
+      "step": 5451
+    },
+    {
+      "epoch": 8.161676646706587,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0008,
+      "loss": 1.161,
+      "step": 5452
+    },
+    {
+      "epoch": 8.16317365269461,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0008,
+      "loss": 1.1676,
+      "step": 5453
+    },
+    {
+      "epoch": 8.164670658682635,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1607,
+      "step": 5454
+    },
+    {
+      "epoch": 8.16616766467066,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1718,
+      "step": 5455
+    },
+    {
+      "epoch": 8.167664670658683,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1675,
+      "step": 5456
+    },
+    {
+      "epoch": 8.169161676646707,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1579,
+      "step": 5457
+    },
+    {
+      "epoch": 8.17065868263473,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0008,
+      "loss": 1.162,
+      "step": 5458
+    },
+    {
+      "epoch": 8.172155688622755,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1661,
+      "step": 5459
+    },
+    {
+      "epoch": 8.173652694610778,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1658,
+      "step": 5460
+    },
+    {
+      "epoch": 8.175149700598803,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0008,
+      "loss": 1.1597,
+      "step": 5461
+    },
+    {
+      "epoch": 8.176646706586826,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1678,
+      "step": 5462
+    },
+    {
+      "epoch": 8.178143712574851,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1571,
+      "step": 5463
+    },
+    {
+      "epoch": 8.179640718562874,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1588,
+      "step": 5464
+    },
+    {
+      "epoch": 8.181137724550899,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1642,
+      "step": 5465
+    },
+    {
+      "epoch": 8.182634730538922,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1643,
+      "step": 5466
+    },
+    {
+      "epoch": 8.184131736526947,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1508,
+      "step": 5467
+    },
+    {
+      "epoch": 8.18562874251497,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1541,
+      "step": 5468
+    },
+    {
+      "epoch": 8.187125748502995,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1534,
+      "step": 5469
+    },
+    {
+      "epoch": 8.188622754491018,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1591,
+      "step": 5470
+    },
+    {
+      "epoch": 8.190119760479043,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1576,
+      "step": 5471
+    },
+    {
+      "epoch": 8.191616766467066,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0008,
+      "loss": 1.15,
+      "step": 5472
+    },
+    {
+      "epoch": 8.19311377245509,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1573,
+      "step": 5473
+    },
+    {
+      "epoch": 8.194610778443113,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1593,
+      "step": 5474
+    },
+    {
+      "epoch": 8.196107784431138,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1545,
+      "step": 5475
+    },
+    {
+      "epoch": 8.197604790419161,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1544,
+      "step": 5476
+    },
+    {
+      "epoch": 8.199101796407186,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1593,
+      "step": 5477
+    },
+    {
+      "epoch": 8.20059880239521,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1527,
+      "step": 5478
+    },
+    {
+      "epoch": 8.202095808383234,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1515,
+      "step": 5479
+    },
+    {
+      "epoch": 8.203592814371257,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1555,
+      "step": 5480
+    },
+    {
+      "epoch": 8.205089820359282,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1517,
+      "step": 5481
+    },
+    {
+      "epoch": 8.206586826347305,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1513,
+      "step": 5482
+    },
+    {
+      "epoch": 8.20808383233533,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1593,
+      "step": 5483
+    },
+    {
+      "epoch": 8.209580838323353,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1579,
+      "step": 5484
+    },
+    {
+      "epoch": 8.211077844311378,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1579,
+      "step": 5485
+    },
+    {
+      "epoch": 8.2125748502994,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1486,
+      "step": 5486
+    },
+    {
+      "epoch": 8.214071856287426,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1512,
+      "step": 5487
+    },
+    {
+      "epoch": 8.215568862275449,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1549,
+      "step": 5488
+    },
+    {
+      "epoch": 8.217065868263473,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1556,
+      "step": 5489
+    },
+    {
+      "epoch": 8.218562874251496,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1563,
+      "step": 5490
+    },
+    {
+      "epoch": 8.220059880239521,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1567,
+      "step": 5491
+    },
+    {
+      "epoch": 8.221556886227544,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1577,
+      "step": 5492
+    },
+    {
+      "epoch": 8.22305389221557,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1519,
+      "step": 5493
+    },
+    {
+      "epoch": 8.224550898203592,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1469,
+      "step": 5494
+    },
+    {
+      "epoch": 8.226047904191617,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1571,
+      "step": 5495
+    },
+    {
+      "epoch": 8.22754491017964,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1567,
+      "step": 5496
+    },
+    {
+      "epoch": 8.229041916167665,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1536,
+      "step": 5497
+    },
+    {
+      "epoch": 8.230538922155688,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1527,
+      "step": 5498
+    },
+    {
+      "epoch": 8.232035928143713,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1541,
+      "step": 5499
+    },
+    {
+      "epoch": 8.233532934131736,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1611,
+      "step": 5500
+    },
+    {
+      "epoch": 8.23502994011976,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1486,
+      "step": 5501
+    },
+    {
+      "epoch": 8.236526946107784,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1554,
+      "step": 5502
+    },
+    {
+      "epoch": 8.238023952095809,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1534,
+      "step": 5503
+    },
+    {
+      "epoch": 8.239520958083832,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1541,
+      "step": 5504
+    },
+    {
+      "epoch": 8.241017964071856,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1515,
+      "step": 5505
+    },
+    {
+      "epoch": 8.24251497005988,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1506,
+      "step": 5506
+    },
+    {
+      "epoch": 8.244011976047904,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1526,
+      "step": 5507
+    },
+    {
+      "epoch": 8.245508982035927,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1509,
+      "step": 5508
+    },
+    {
+      "epoch": 8.247005988023952,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1493,
+      "step": 5509
+    },
+    {
+      "epoch": 8.248502994011975,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1502,
+      "step": 5510
+    },
+    {
+      "epoch": 8.25,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1541,
+      "step": 5511
+    },
+    {
+      "epoch": 8.251497005988025,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.156,
+      "step": 5512
+    },
+    {
+      "epoch": 8.252994011976048,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1565,
+      "step": 5513
+    },
+    {
+      "epoch": 8.254491017964073,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1494,
+      "step": 5514
+    },
+    {
+      "epoch": 8.255988023952096,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1593,
+      "step": 5515
+    },
+    {
+      "epoch": 8.25748502994012,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 5516
+    },
+    {
+      "epoch": 8.258982035928144,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1524,
+      "step": 5517
+    },
+    {
+      "epoch": 8.260479041916168,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1567,
+      "step": 5518
+    },
+    {
+      "epoch": 8.261976047904191,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1553,
+      "step": 5519
+    },
+    {
+      "epoch": 8.263473053892216,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.148,
+      "step": 5520
+    },
+    {
+      "epoch": 8.26497005988024,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 5521
+    },
+    {
+      "epoch": 8.266467065868264,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1544,
+      "step": 5522
+    },
+    {
+      "epoch": 8.267964071856287,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1565,
+      "step": 5523
+    },
+    {
+      "epoch": 8.269461077844312,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1589,
+      "step": 5524
+    },
+    {
+      "epoch": 8.270958083832335,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1472,
+      "step": 5525
+    },
+    {
+      "epoch": 8.27245508982036,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1517,
+      "step": 5526
+    },
+    {
+      "epoch": 8.273952095808383,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1571,
+      "step": 5527
+    },
+    {
+      "epoch": 8.275449101796408,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1532,
+      "step": 5528
+    },
+    {
+      "epoch": 8.27694610778443,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1547,
+      "step": 5529
+    },
+    {
+      "epoch": 8.278443113772456,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1555,
+      "step": 5530
+    },
+    {
+      "epoch": 8.279940119760479,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1502,
+      "step": 5531
+    },
+    {
+      "epoch": 8.281437125748504,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 5532
+    },
+    {
+      "epoch": 8.282934131736527,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1547,
+      "step": 5533
+    },
+    {
+      "epoch": 8.284431137724551,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1564,
+      "step": 5534
+    },
+    {
+      "epoch": 8.285928143712574,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1552,
+      "step": 5535
+    },
+    {
+      "epoch": 8.2874251497006,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1491,
+      "step": 5536
+    },
+    {
+      "epoch": 8.288922155688622,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1544,
+      "step": 5537
+    },
+    {
+      "epoch": 8.290419161676647,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1574,
+      "step": 5538
+    },
+    {
+      "epoch": 8.29191616766467,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1564,
+      "step": 5539
+    },
+    {
+      "epoch": 8.293413173652695,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1497,
+      "step": 5540
+    },
+    {
+      "epoch": 8.294910179640718,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1504,
+      "step": 5541
+    },
+    {
+      "epoch": 8.296407185628743,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1528,
+      "step": 5542
+    },
+    {
+      "epoch": 8.297904191616766,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1551,
+      "step": 5543
+    },
+    {
+      "epoch": 8.29940119760479,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 5544
+    },
+    {
+      "epoch": 8.300898203592814,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1516,
+      "step": 5545
+    },
+    {
+      "epoch": 8.302395209580839,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1491,
+      "step": 5546
+    },
+    {
+      "epoch": 8.303892215568862,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1527,
+      "step": 5547
+    },
+    {
+      "epoch": 8.305389221556887,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 5548
+    },
+    {
+      "epoch": 8.30688622754491,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1493,
+      "step": 5549
+    },
+    {
+      "epoch": 8.308383233532934,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1559,
+      "step": 5550
+    },
+    {
+      "epoch": 8.309880239520957,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1513,
+      "step": 5551
+    },
+    {
+      "epoch": 8.311377245508982,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1523,
+      "step": 5552
+    },
+    {
+      "epoch": 8.312874251497005,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1485,
+      "step": 5553
+    },
+    {
+      "epoch": 8.31437125748503,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1471,
+      "step": 5554
+    },
+    {
+      "epoch": 8.315868263473053,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1499,
+      "step": 5555
+    },
+    {
+      "epoch": 8.317365269461078,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.153,
+      "step": 5556
+    },
+    {
+      "epoch": 8.318862275449101,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1568,
+      "step": 5557
+    },
+    {
+      "epoch": 8.320359281437126,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.152,
+      "step": 5558
+    },
+    {
+      "epoch": 8.321856287425149,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1586,
+      "step": 5559
+    },
+    {
+      "epoch": 8.323353293413174,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1481,
+      "step": 5560
+    },
+    {
+      "epoch": 8.324850299401197,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1534,
+      "step": 5561
+    },
+    {
+      "epoch": 8.326347305389222,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1518,
+      "step": 5562
+    },
+    {
+      "epoch": 8.327844311377245,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1518,
+      "step": 5563
+    },
+    {
+      "epoch": 8.32934131736527,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 5564
+    },
+    {
+      "epoch": 8.330838323353294,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1486,
+      "step": 5565
+    },
+    {
+      "epoch": 8.332335329341317,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1489,
+      "step": 5566
+    },
+    {
+      "epoch": 8.33383233532934,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1555,
+      "step": 5567
+    },
+    {
+      "epoch": 8.335329341317365,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1509,
+      "step": 5568
+    },
+    {
+      "epoch": 8.33682634730539,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1534,
+      "step": 5569
+    },
+    {
+      "epoch": 8.338323353293413,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1462,
+      "step": 5570
+    },
+    {
+      "epoch": 8.339820359281438,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1486,
+      "step": 5571
+    },
+    {
+      "epoch": 8.341317365269461,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1485,
+      "step": 5572
+    },
+    {
+      "epoch": 8.342814371257486,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1505,
+      "step": 5573
+    },
+    {
+      "epoch": 8.344311377245509,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1522,
+      "step": 5574
+    },
+    {
+      "epoch": 8.345808383233534,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1529,
+      "step": 5575
+    },
+    {
+      "epoch": 8.347305389221557,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.15,
+      "step": 5576
+    },
+    {
+      "epoch": 8.348802395209582,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 5577
+    },
+    {
+      "epoch": 8.350299401197605,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1472,
+      "step": 5578
+    },
+    {
+      "epoch": 8.35179640718563,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1547,
+      "step": 5579
+    },
+    {
+      "epoch": 8.353293413173652,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1489,
+      "step": 5580
+    },
+    {
+      "epoch": 8.354790419161677,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.153,
+      "step": 5581
+    },
+    {
+      "epoch": 8.3562874251497,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1581,
+      "step": 5582
+    },
+    {
+      "epoch": 8.357784431137725,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.157,
+      "step": 5583
+    },
+    {
+      "epoch": 8.359281437125748,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 5584
+    },
+    {
+      "epoch": 8.360778443113773,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.157,
+      "step": 5585
+    },
+    {
+      "epoch": 8.362275449101796,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1498,
+      "step": 5586
+    },
+    {
+      "epoch": 8.363772455089821,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1492,
+      "step": 5587
+    },
+    {
+      "epoch": 8.365269461077844,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 5588
+    },
+    {
+      "epoch": 8.366766467065869,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1513,
+      "step": 5589
+    },
+    {
+      "epoch": 8.368263473053892,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1554,
+      "step": 5590
+    },
+    {
+      "epoch": 8.369760479041917,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.154,
+      "step": 5591
+    },
+    {
+      "epoch": 8.37125748502994,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1544,
+      "step": 5592
+    },
+    {
+      "epoch": 8.372754491017965,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 5593
+    },
+    {
+      "epoch": 8.374251497005988,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 5594
+    },
+    {
+      "epoch": 8.375748502994012,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1476,
+      "step": 5595
+    },
+    {
+      "epoch": 8.377245508982035,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1573,
+      "step": 5596
+    },
+    {
+      "epoch": 8.37874251497006,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1473,
+      "step": 5597
+    },
+    {
+      "epoch": 8.380239520958083,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1523,
+      "step": 5598
+    },
+    {
+      "epoch": 8.381736526946108,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1473,
+      "step": 5599
+    },
+    {
+      "epoch": 8.383233532934131,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1486,
+      "step": 5600
+    },
+    {
+      "epoch": 8.384730538922156,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1549,
+      "step": 5601
+    },
+    {
+      "epoch": 8.386227544910179,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1504,
+      "step": 5602
+    },
+    {
+      "epoch": 8.387724550898204,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1488,
+      "step": 5603
+    },
+    {
+      "epoch": 8.389221556886227,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1531,
+      "step": 5604
+    },
+    {
+      "epoch": 8.390718562874252,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 5605
+    },
+    {
+      "epoch": 8.392215568862275,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1507,
+      "step": 5606
+    },
+    {
+      "epoch": 8.3937125748503,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1525,
+      "step": 5607
+    },
+    {
+      "epoch": 8.395209580838323,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1492,
+      "step": 5608
+    },
+    {
+      "epoch": 8.396706586826348,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1487,
+      "step": 5609
+    },
+    {
+      "epoch": 8.39820359281437,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 5610
+    },
+    {
+      "epoch": 8.399700598802395,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1528,
+      "step": 5611
+    },
+    {
+      "epoch": 8.401197604790418,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1514,
+      "step": 5612
+    },
+    {
+      "epoch": 8.402694610778443,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1494,
+      "step": 5613
+    },
+    {
+      "epoch": 8.404191616766466,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1539,
+      "step": 5614
+    },
+    {
+      "epoch": 8.405688622754491,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1501,
+      "step": 5615
+    },
+    {
+      "epoch": 8.407185628742514,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1563,
+      "step": 5616
+    },
+    {
+      "epoch": 8.408682634730539,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1519,
+      "step": 5617
+    },
+    {
+      "epoch": 8.410179640718562,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1488,
+      "step": 5618
+    },
+    {
+      "epoch": 8.411676646706587,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1523,
+      "step": 5619
+    },
+    {
+      "epoch": 8.41317365269461,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1416,
+      "step": 5620
+    },
+    {
+      "epoch": 8.414670658682635,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1507,
+      "step": 5621
+    },
+    {
+      "epoch": 8.41616766467066,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1473,
+      "step": 5622
+    },
+    {
+      "epoch": 8.417664670658683,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1524,
+      "step": 5623
+    },
+    {
+      "epoch": 8.419161676646706,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1498,
+      "step": 5624
+    },
+    {
+      "epoch": 8.42065868263473,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1524,
+      "step": 5625
+    },
+    {
+      "epoch": 8.422155688622755,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1534,
+      "step": 5626
+    },
+    {
+      "epoch": 8.423652694610778,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 5627
+    },
+    {
+      "epoch": 8.425149700598803,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1517,
+      "step": 5628
+    },
+    {
+      "epoch": 8.426646706586826,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1534,
+      "step": 5629
+    },
+    {
+      "epoch": 8.428143712574851,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1504,
+      "step": 5630
+    },
+    {
+      "epoch": 8.429640718562874,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1539,
+      "step": 5631
+    },
+    {
+      "epoch": 8.431137724550899,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1491,
+      "step": 5632
+    },
+    {
+      "epoch": 8.432634730538922,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1558,
+      "step": 5633
+    },
+    {
+      "epoch": 8.434131736526947,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1562,
+      "step": 5634
+    },
+    {
+      "epoch": 8.43562874251497,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1487,
+      "step": 5635
+    },
+    {
+      "epoch": 8.437125748502995,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1565,
+      "step": 5636
+    },
+    {
+      "epoch": 8.438622754491018,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1532,
+      "step": 5637
+    },
+    {
+      "epoch": 8.440119760479043,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1482,
+      "step": 5638
+    },
+    {
+      "epoch": 8.441616766467066,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.149,
+      "step": 5639
+    },
+    {
+      "epoch": 8.44311377245509,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1571,
+      "step": 5640
+    },
+    {
+      "epoch": 8.444610778443113,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1571,
+      "step": 5641
+    },
+    {
+      "epoch": 8.446107784431138,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1557,
+      "step": 5642
+    },
+    {
+      "epoch": 8.447604790419161,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1536,
+      "step": 5643
+    },
+    {
+      "epoch": 8.449101796407186,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1581,
+      "step": 5644
+    },
+    {
+      "epoch": 8.45059880239521,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1584,
+      "step": 5645
+    },
+    {
+      "epoch": 8.452095808383234,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1471,
+      "step": 5646
+    },
+    {
+      "epoch": 8.453592814371257,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1532,
+      "step": 5647
+    },
+    {
+      "epoch": 8.455089820359282,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1581,
+      "step": 5648
+    },
+    {
+      "epoch": 8.456586826347305,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1542,
+      "step": 5649
+    },
+    {
+      "epoch": 8.45808383233533,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 5650
+    },
+    {
+      "epoch": 8.459580838323353,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1511,
+      "step": 5651
+    },
+    {
+      "epoch": 8.461077844311378,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1556,
+      "step": 5652
+    },
+    {
+      "epoch": 8.4625748502994,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.153,
+      "step": 5653
+    },
+    {
+      "epoch": 8.464071856287426,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1537,
+      "step": 5654
+    },
+    {
+      "epoch": 8.465568862275449,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1505,
+      "step": 5655
+    },
+    {
+      "epoch": 8.467065868263473,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1428,
+      "step": 5656
+    },
+    {
+      "epoch": 8.468562874251496,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 5657
+    },
+    {
+      "epoch": 8.470059880239521,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1479,
+      "step": 5658
+    },
+    {
+      "epoch": 8.471556886227544,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1569,
+      "step": 5659
+    },
+    {
+      "epoch": 8.47305389221557,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1533,
+      "step": 5660
+    },
+    {
+      "epoch": 8.474550898203592,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1509,
+      "step": 5661
+    },
+    {
+      "epoch": 8.476047904191617,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1473,
+      "step": 5662
+    },
+    {
+      "epoch": 8.47754491017964,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1553,
+      "step": 5663
+    },
+    {
+      "epoch": 8.479041916167665,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1464,
+      "step": 5664
+    },
+    {
+      "epoch": 8.480538922155688,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1563,
+      "step": 5665
+    },
+    {
+      "epoch": 8.482035928143713,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1521,
+      "step": 5666
+    },
+    {
+      "epoch": 8.483532934131736,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1492,
+      "step": 5667
+    },
+    {
+      "epoch": 8.48502994011976,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.151,
+      "step": 5668
+    },
+    {
+      "epoch": 8.486526946107784,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1552,
+      "step": 5669
+    },
+    {
+      "epoch": 8.488023952095809,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 5670
+    },
+    {
+      "epoch": 8.489520958083832,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1578,
+      "step": 5671
+    },
+    {
+      "epoch": 8.491017964071856,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1505,
+      "step": 5672
+    },
+    {
+      "epoch": 8.49251497005988,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1484,
+      "step": 5673
+    },
+    {
+      "epoch": 8.494011976047904,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1488,
+      "step": 5674
+    },
+    {
+      "epoch": 8.495508982035927,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.151,
+      "step": 5675
+    },
+    {
+      "epoch": 8.497005988023952,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1497,
+      "step": 5676
+    },
+    {
+      "epoch": 8.498502994011975,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1541,
+      "step": 5677
+    },
+    {
+      "epoch": 8.5,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1582,
+      "step": 5678
+    },
+    {
+      "epoch": 8.501497005988025,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1533,
+      "step": 5679
+    },
+    {
+      "epoch": 8.502994011976048,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1493,
+      "step": 5680
+    },
+    {
+      "epoch": 8.504491017964071,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1544,
+      "step": 5681
+    },
+    {
+      "epoch": 8.505988023952096,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1498,
+      "step": 5682
+    },
+    {
+      "epoch": 8.50748502994012,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1507,
+      "step": 5683
+    },
+    {
+      "epoch": 8.508982035928144,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 5684
+    },
+    {
+      "epoch": 8.510479041916168,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 5685
+    },
+    {
+      "epoch": 8.511976047904191,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1501,
+      "step": 5686
+    },
+    {
+      "epoch": 8.513473053892216,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1503,
+      "step": 5687
+    },
+    {
+      "epoch": 8.51497005988024,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1521,
+      "step": 5688
+    },
+    {
+      "epoch": 8.516467065868264,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1551,
+      "step": 5689
+    },
+    {
+      "epoch": 8.517964071856287,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1501,
+      "step": 5690
+    },
+    {
+      "epoch": 8.519461077844312,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1469,
+      "step": 5691
+    },
+    {
+      "epoch": 8.520958083832335,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1495,
+      "step": 5692
+    },
+    {
+      "epoch": 8.52245508982036,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1549,
+      "step": 5693
+    },
+    {
+      "epoch": 8.523952095808383,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.154,
+      "step": 5694
+    },
+    {
+      "epoch": 8.525449101796408,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1568,
+      "step": 5695
+    },
+    {
+      "epoch": 8.52694610778443,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1582,
+      "step": 5696
+    },
+    {
+      "epoch": 8.528443113772456,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1446,
+      "step": 5697
+    },
+    {
+      "epoch": 8.529940119760479,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1585,
+      "step": 5698
+    },
+    {
+      "epoch": 8.531437125748504,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1491,
+      "step": 5699
+    },
+    {
+      "epoch": 8.532934131736527,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1443,
+      "step": 5700
+    },
+    {
+      "epoch": 8.534431137724551,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1557,
+      "step": 5701
+    },
+    {
+      "epoch": 8.535928143712574,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1503,
+      "step": 5702
+    },
+    {
+      "epoch": 8.5374251497006,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1495,
+      "step": 5703
+    },
+    {
+      "epoch": 8.538922155688622,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1521,
+      "step": 5704
+    },
+    {
+      "epoch": 8.540419161676647,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1544,
+      "step": 5705
+    },
+    {
+      "epoch": 8.54191616766467,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1512,
+      "step": 5706
+    },
+    {
+      "epoch": 8.543413173652695,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1512,
+      "step": 5707
+    },
+    {
+      "epoch": 8.544910179640718,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1498,
+      "step": 5708
+    },
+    {
+      "epoch": 8.546407185628743,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.147,
+      "step": 5709
+    },
+    {
+      "epoch": 8.547904191616766,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 5710
+    },
+    {
+      "epoch": 8.54940119760479,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1486,
+      "step": 5711
+    },
+    {
+      "epoch": 8.550898203592814,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1469,
+      "step": 5712
+    },
+    {
+      "epoch": 8.552395209580839,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.155,
+      "step": 5713
+    },
+    {
+      "epoch": 8.553892215568862,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1618,
+      "step": 5714
+    },
+    {
+      "epoch": 8.555389221556887,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1589,
+      "step": 5715
+    },
+    {
+      "epoch": 8.55688622754491,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1527,
+      "step": 5716
+    },
+    {
+      "epoch": 8.558383233532934,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.148,
+      "step": 5717
+    },
+    {
+      "epoch": 8.559880239520957,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1568,
+      "step": 5718
+    },
+    {
+      "epoch": 8.561377245508982,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1543,
+      "step": 5719
+    },
+    {
+      "epoch": 8.562874251497005,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1536,
+      "step": 5720
+    },
+    {
+      "epoch": 8.56437125748503,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1504,
+      "step": 5721
+    },
+    {
+      "epoch": 8.565868263473053,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.15,
+      "step": 5722
+    },
+    {
+      "epoch": 8.567365269461078,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.154,
+      "step": 5723
+    },
+    {
+      "epoch": 8.568862275449101,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1512,
+      "step": 5724
+    },
+    {
+      "epoch": 8.570359281437126,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1513,
+      "step": 5725
+    },
+    {
+      "epoch": 8.571856287425149,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1517,
+      "step": 5726
+    },
+    {
+      "epoch": 8.573353293413174,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1509,
+      "step": 5727
+    },
+    {
+      "epoch": 8.574850299401197,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.153,
+      "step": 5728
+    },
+    {
+      "epoch": 8.576347305389222,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1501,
+      "step": 5729
+    },
+    {
+      "epoch": 8.577844311377245,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1547,
+      "step": 5730
+    },
+    {
+      "epoch": 8.57934131736527,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1552,
+      "step": 5731
+    },
+    {
+      "epoch": 8.580838323353294,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1497,
+      "step": 5732
+    },
+    {
+      "epoch": 8.582335329341317,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1513,
+      "step": 5733
+    },
+    {
+      "epoch": 8.58383233532934,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0008,
+      "loss": 1.151,
+      "step": 5734
+    },
+    {
+      "epoch": 8.585329341317365,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 5735
+    },
+    {
+      "epoch": 8.58682634730539,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 5736
+    },
+    {
+      "epoch": 8.588323353293413,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1446,
+      "step": 5737
+    },
+    {
+      "epoch": 8.589820359281438,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.149,
+      "step": 5738
+    },
+    {
+      "epoch": 8.591317365269461,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1565,
+      "step": 5739
+    },
+    {
+      "epoch": 8.592814371257486,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 5740
+    },
+    {
+      "epoch": 8.594311377245509,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 5741
+    },
+    {
+      "epoch": 8.595808383233534,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1559,
+      "step": 5742
+    },
+    {
+      "epoch": 8.597305389221557,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1506,
+      "step": 5743
+    },
+    {
+      "epoch": 8.598802395209582,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1542,
+      "step": 5744
+    },
+    {
+      "epoch": 8.600299401197605,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1483,
+      "step": 5745
+    },
+    {
+      "epoch": 8.60179640718563,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 5746
+    },
+    {
+      "epoch": 8.603293413173652,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1457,
+      "step": 5747
+    },
+    {
+      "epoch": 8.604790419161677,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1513,
+      "step": 5748
+    },
+    {
+      "epoch": 8.6062874251497,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1472,
+      "step": 5749
+    },
+    {
+      "epoch": 8.607784431137725,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1496,
+      "step": 5750
+    },
+    {
+      "epoch": 8.609281437125748,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1503,
+      "step": 5751
+    },
+    {
+      "epoch": 8.610778443113773,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 5752
+    },
+    {
+      "epoch": 8.612275449101796,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1484,
+      "step": 5753
+    },
+    {
+      "epoch": 8.613772455089821,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.156,
+      "step": 5754
+    },
+    {
+      "epoch": 8.615269461077844,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1499,
+      "step": 5755
+    },
+    {
+      "epoch": 8.616766467065869,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1475,
+      "step": 5756
+    },
+    {
+      "epoch": 8.618263473053892,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1487,
+      "step": 5757
+    },
+    {
+      "epoch": 8.619760479041917,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1507,
+      "step": 5758
+    },
+    {
+      "epoch": 8.62125748502994,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1488,
+      "step": 5759
+    },
+    {
+      "epoch": 8.622754491017965,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1506,
+      "step": 5760
+    },
+    {
+      "epoch": 8.624251497005988,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1508,
+      "step": 5761
+    },
+    {
+      "epoch": 8.625748502994012,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1503,
+      "step": 5762
+    },
+    {
+      "epoch": 8.627245508982035,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 5763
+    },
+    {
+      "epoch": 8.62874251497006,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1567,
+      "step": 5764
+    },
+    {
+      "epoch": 8.630239520958083,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1519,
+      "step": 5765
+    },
+    {
+      "epoch": 8.631736526946108,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1489,
+      "step": 5766
+    },
+    {
+      "epoch": 8.633233532934131,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1604,
+      "step": 5767
+    },
+    {
+      "epoch": 8.634730538922156,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1516,
+      "step": 5768
+    },
+    {
+      "epoch": 8.636227544910179,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1492,
+      "step": 5769
+    },
+    {
+      "epoch": 8.637724550898204,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.148,
+      "step": 5770
+    },
+    {
+      "epoch": 8.639221556886227,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1454,
+      "step": 5771
+    },
+    {
+      "epoch": 8.640718562874252,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 5772
+    },
+    {
+      "epoch": 8.642215568862275,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1507,
+      "step": 5773
+    },
+    {
+      "epoch": 8.6437125748503,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1552,
+      "step": 5774
+    },
+    {
+      "epoch": 8.645209580838323,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 5775
+    },
+    {
+      "epoch": 8.646706586826348,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 5776
+    },
+    {
+      "epoch": 8.64820359281437,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1539,
+      "step": 5777
+    },
+    {
+      "epoch": 8.649700598802395,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1553,
+      "step": 5778
+    },
+    {
+      "epoch": 8.651197604790418,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 5779
+    },
+    {
+      "epoch": 8.652694610778443,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1518,
+      "step": 5780
+    },
+    {
+      "epoch": 8.654191616766466,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1473,
+      "step": 5781
+    },
+    {
+      "epoch": 8.655688622754491,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1477,
+      "step": 5782
+    },
+    {
+      "epoch": 8.657185628742514,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1498,
+      "step": 5783
+    },
+    {
+      "epoch": 8.658682634730539,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1515,
+      "step": 5784
+    },
+    {
+      "epoch": 8.660179640718562,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 5785
+    },
+    {
+      "epoch": 8.661676646706587,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.15,
+      "step": 5786
+    },
+    {
+      "epoch": 8.66317365269461,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.158,
+      "step": 5787
+    },
+    {
+      "epoch": 8.664670658682635,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.15,
+      "step": 5788
+    },
+    {
+      "epoch": 8.66616766467066,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.15,
+      "step": 5789
+    },
+    {
+      "epoch": 8.667664670658683,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1457,
+      "step": 5790
+    },
+    {
+      "epoch": 8.669161676646706,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1472,
+      "step": 5791
+    },
+    {
+      "epoch": 8.67065868263473,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1513,
+      "step": 5792
+    },
+    {
+      "epoch": 8.672155688622755,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1501,
+      "step": 5793
+    },
+    {
+      "epoch": 8.673652694610778,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1484,
+      "step": 5794
+    },
+    {
+      "epoch": 8.675149700598803,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1516,
+      "step": 5795
+    },
+    {
+      "epoch": 8.676646706586826,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1515,
+      "step": 5796
+    },
+    {
+      "epoch": 8.678143712574851,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.155,
+      "step": 5797
+    },
+    {
+      "epoch": 8.679640718562874,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1472,
+      "step": 5798
+    },
+    {
+      "epoch": 8.681137724550899,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.155,
+      "step": 5799
+    },
+    {
+      "epoch": 8.682634730538922,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 5800
+    },
+    {
+      "epoch": 8.684131736526947,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1572,
+      "step": 5801
+    },
+    {
+      "epoch": 8.68562874251497,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1469,
+      "step": 5802
+    },
+    {
+      "epoch": 8.687125748502995,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 5803
+    },
+    {
+      "epoch": 8.688622754491018,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 5804
+    },
+    {
+      "epoch": 8.690119760479043,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 5805
+    },
+    {
+      "epoch": 8.691616766467066,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.153,
+      "step": 5806
+    },
+    {
+      "epoch": 8.69311377245509,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1537,
+      "step": 5807
+    },
+    {
+      "epoch": 8.694610778443113,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 5808
+    },
+    {
+      "epoch": 8.696107784431138,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1486,
+      "step": 5809
+    },
+    {
+      "epoch": 8.697604790419161,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 5810
+    },
+    {
+      "epoch": 8.699101796407186,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1515,
+      "step": 5811
+    },
+    {
+      "epoch": 8.70059880239521,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1511,
+      "step": 5812
+    },
+    {
+      "epoch": 8.702095808383234,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1521,
+      "step": 5813
+    },
+    {
+      "epoch": 8.703592814371257,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1502,
+      "step": 5814
+    },
+    {
+      "epoch": 8.705089820359282,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1508,
+      "step": 5815
+    },
+    {
+      "epoch": 8.706586826347305,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.146,
+      "step": 5816
+    },
+    {
+      "epoch": 8.70808383233533,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 5817
+    },
+    {
+      "epoch": 8.709580838323353,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1532,
+      "step": 5818
+    },
+    {
+      "epoch": 8.711077844311378,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1482,
+      "step": 5819
+    },
+    {
+      "epoch": 8.7125748502994,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1485,
+      "step": 5820
+    },
+    {
+      "epoch": 8.714071856287426,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.157,
+      "step": 5821
+    },
+    {
+      "epoch": 8.715568862275449,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.155,
+      "step": 5822
+    },
+    {
+      "epoch": 8.717065868263473,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 5823
+    },
+    {
+      "epoch": 8.718562874251496,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.15,
+      "step": 5824
+    },
+    {
+      "epoch": 8.720059880239521,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1495,
+      "step": 5825
+    },
+    {
+      "epoch": 8.721556886227544,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.15,
+      "step": 5826
+    },
+    {
+      "epoch": 8.72305389221557,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1487,
+      "step": 5827
+    },
+    {
+      "epoch": 8.724550898203592,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1496,
+      "step": 5828
+    },
+    {
+      "epoch": 8.726047904191617,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 5829
+    },
+    {
+      "epoch": 8.72754491017964,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1494,
+      "step": 5830
+    },
+    {
+      "epoch": 8.729041916167665,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1492,
+      "step": 5831
+    },
+    {
+      "epoch": 8.730538922155688,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1538,
+      "step": 5832
+    },
+    {
+      "epoch": 8.732035928143713,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1534,
+      "step": 5833
+    },
+    {
+      "epoch": 8.733532934131736,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1464,
+      "step": 5834
+    },
+    {
+      "epoch": 8.73502994011976,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 5835
+    },
+    {
+      "epoch": 8.736526946107784,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1531,
+      "step": 5836
+    },
+    {
+      "epoch": 8.738023952095809,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1541,
+      "step": 5837
+    },
+    {
+      "epoch": 8.739520958083832,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1505,
+      "step": 5838
+    },
+    {
+      "epoch": 8.741017964071856,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 5839
+    },
+    {
+      "epoch": 8.74251497005988,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1575,
+      "step": 5840
+    },
+    {
+      "epoch": 8.744011976047904,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1534,
+      "step": 5841
+    },
+    {
+      "epoch": 8.745508982035929,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1514,
+      "step": 5842
+    },
+    {
+      "epoch": 8.747005988023952,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 5843
+    },
+    {
+      "epoch": 8.748502994011975,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1469,
+      "step": 5844
+    },
+    {
+      "epoch": 8.75,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1486,
+      "step": 5845
+    },
+    {
+      "epoch": 8.751497005988025,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1485,
+      "step": 5846
+    },
+    {
+      "epoch": 8.752994011976048,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1507,
+      "step": 5847
+    },
+    {
+      "epoch": 8.754491017964071,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1491,
+      "step": 5848
+    },
+    {
+      "epoch": 8.755988023952096,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1504,
+      "step": 5849
+    },
+    {
+      "epoch": 8.75748502994012,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 5850
+    },
+    {
+      "epoch": 8.758982035928144,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1499,
+      "step": 5851
+    },
+    {
+      "epoch": 8.760479041916168,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1556,
+      "step": 5852
+    },
+    {
+      "epoch": 8.761976047904191,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1484,
+      "step": 5853
+    },
+    {
+      "epoch": 8.763473053892216,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1536,
+      "step": 5854
+    },
+    {
+      "epoch": 8.76497005988024,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1485,
+      "step": 5855
+    },
+    {
+      "epoch": 8.766467065868264,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1493,
+      "step": 5856
+    },
+    {
+      "epoch": 8.767964071856287,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1511,
+      "step": 5857
+    },
+    {
+      "epoch": 8.769461077844312,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 5858
+    },
+    {
+      "epoch": 8.770958083832335,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1536,
+      "step": 5859
+    },
+    {
+      "epoch": 8.77245508982036,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1544,
+      "step": 5860
+    },
+    {
+      "epoch": 8.773952095808383,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1555,
+      "step": 5861
+    },
+    {
+      "epoch": 8.775449101796408,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1465,
+      "step": 5862
+    },
+    {
+      "epoch": 8.77694610778443,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.158,
+      "step": 5863
+    },
+    {
+      "epoch": 8.778443113772456,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1471,
+      "step": 5864
+    },
+    {
+      "epoch": 8.779940119760479,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1502,
+      "step": 5865
+    },
+    {
+      "epoch": 8.781437125748504,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0008,
+      "loss": 1.1498,
+      "step": 5866
+    },
+    {
+      "epoch": 8.782934131736527,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1505,
+      "step": 5867
+    },
+    {
+      "epoch": 8.784431137724551,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.153,
+      "step": 5868
+    },
+    {
+      "epoch": 8.785928143712574,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 5869
+    },
+    {
+      "epoch": 8.7874251497006,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 5870
+    },
+    {
+      "epoch": 8.788922155688622,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1562,
+      "step": 5871
+    },
+    {
+      "epoch": 8.790419161676647,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1555,
+      "step": 5872
+    },
+    {
+      "epoch": 8.79191616766467,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 5873
+    },
+    {
+      "epoch": 8.793413173652695,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1514,
+      "step": 5874
+    },
+    {
+      "epoch": 8.794910179640718,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1552,
+      "step": 5875
+    },
+    {
+      "epoch": 8.796407185628743,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1485,
+      "step": 5876
+    },
+    {
+      "epoch": 8.797904191616766,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1542,
+      "step": 5877
+    },
+    {
+      "epoch": 8.79940119760479,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1525,
+      "step": 5878
+    },
+    {
+      "epoch": 8.800898203592814,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1472,
+      "step": 5879
+    },
+    {
+      "epoch": 8.802395209580839,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1497,
+      "step": 5880
+    },
+    {
+      "epoch": 8.803892215568862,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1531,
+      "step": 5881
+    },
+    {
+      "epoch": 8.805389221556887,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1555,
+      "step": 5882
+    },
+    {
+      "epoch": 8.80688622754491,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1483,
+      "step": 5883
+    },
+    {
+      "epoch": 8.808383233532934,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1504,
+      "step": 5884
+    },
+    {
+      "epoch": 8.809880239520957,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1491,
+      "step": 5885
+    },
+    {
+      "epoch": 8.811377245508982,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1511,
+      "step": 5886
+    },
+    {
+      "epoch": 8.812874251497005,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1457,
+      "step": 5887
+    },
+    {
+      "epoch": 8.81437125748503,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 5888
+    },
+    {
+      "epoch": 8.815868263473053,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1566,
+      "step": 5889
+    },
+    {
+      "epoch": 8.817365269461078,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1491,
+      "step": 5890
+    },
+    {
+      "epoch": 8.818862275449101,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1493,
+      "step": 5891
+    },
+    {
+      "epoch": 8.820359281437126,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1512,
+      "step": 5892
+    },
+    {
+      "epoch": 8.821856287425149,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1588,
+      "step": 5893
+    },
+    {
+      "epoch": 8.823353293413174,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 5894
+    },
+    {
+      "epoch": 8.824850299401197,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1486,
+      "step": 5895
+    },
+    {
+      "epoch": 8.826347305389222,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1494,
+      "step": 5896
+    },
+    {
+      "epoch": 8.827844311377245,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1469,
+      "step": 5897
+    },
+    {
+      "epoch": 8.82934131736527,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1518,
+      "step": 5898
+    },
+    {
+      "epoch": 8.830838323353294,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.15,
+      "step": 5899
+    },
+    {
+      "epoch": 8.832335329341317,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1508,
+      "step": 5900
+    },
+    {
+      "epoch": 8.83383233532934,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 5901
+    },
+    {
+      "epoch": 8.835329341317365,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1518,
+      "step": 5902
+    },
+    {
+      "epoch": 8.83682634730539,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1496,
+      "step": 5903
+    },
+    {
+      "epoch": 8.838323353293413,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1489,
+      "step": 5904
+    },
+    {
+      "epoch": 8.839820359281438,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0008,
+      "loss": 1.153,
+      "step": 5905
+    },
+    {
+      "epoch": 8.841317365269461,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1475,
+      "step": 5906
+    },
+    {
+      "epoch": 8.842814371257486,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1523,
+      "step": 5907
+    },
+    {
+      "epoch": 8.844311377245509,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1556,
+      "step": 5908
+    },
+    {
+      "epoch": 8.845808383233534,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1521,
+      "step": 5909
+    },
+    {
+      "epoch": 8.847305389221557,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.153,
+      "step": 5910
+    },
+    {
+      "epoch": 8.848802395209582,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1468,
+      "step": 5911
+    },
+    {
+      "epoch": 8.850299401197605,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1502,
+      "step": 5912
+    },
+    {
+      "epoch": 8.85179640718563,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1572,
+      "step": 5913
+    },
+    {
+      "epoch": 8.853293413173652,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1499,
+      "step": 5914
+    },
+    {
+      "epoch": 8.854790419161677,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1468,
+      "step": 5915
+    },
+    {
+      "epoch": 8.8562874251497,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1477,
+      "step": 5916
+    },
+    {
+      "epoch": 8.857784431137725,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1491,
+      "step": 5917
+    },
+    {
+      "epoch": 8.859281437125748,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1539,
+      "step": 5918
+    },
+    {
+      "epoch": 8.860778443113773,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0008,
+      "loss": 1.157,
+      "step": 5919
+    },
+    {
+      "epoch": 8.862275449101796,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1505,
+      "step": 5920
+    },
+    {
+      "epoch": 8.863772455089821,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1522,
+      "step": 5921
+    },
+    {
+      "epoch": 8.865269461077844,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1479,
+      "step": 5922
+    },
+    {
+      "epoch": 8.866766467065869,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1513,
+      "step": 5923
+    },
+    {
+      "epoch": 8.868263473053892,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 5924
+    },
+    {
+      "epoch": 8.869760479041917,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1486,
+      "step": 5925
+    },
+    {
+      "epoch": 8.87125748502994,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1487,
+      "step": 5926
+    },
+    {
+      "epoch": 8.872754491017965,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 5927
+    },
+    {
+      "epoch": 8.874251497005988,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1503,
+      "step": 5928
+    },
+    {
+      "epoch": 8.875748502994012,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1482,
+      "step": 5929
+    },
+    {
+      "epoch": 8.877245508982035,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1515,
+      "step": 5930
+    },
+    {
+      "epoch": 8.87874251497006,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1562,
+      "step": 5931
+    },
+    {
+      "epoch": 8.880239520958083,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1515,
+      "step": 5932
+    },
+    {
+      "epoch": 8.881736526946108,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.147,
+      "step": 5933
+    },
+    {
+      "epoch": 8.883233532934131,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1477,
+      "step": 5934
+    },
+    {
+      "epoch": 8.884730538922156,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.15,
+      "step": 5935
+    },
+    {
+      "epoch": 8.886227544910179,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0008,
+      "loss": 1.154,
+      "step": 5936
+    },
+    {
+      "epoch": 8.887724550898204,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1493,
+      "step": 5937
+    },
+    {
+      "epoch": 8.889221556886227,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 5938
+    },
+    {
+      "epoch": 8.890718562874252,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1489,
+      "step": 5939
+    },
+    {
+      "epoch": 8.892215568862275,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1535,
+      "step": 5940
+    },
+    {
+      "epoch": 8.8937125748503,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1527,
+      "step": 5941
+    },
+    {
+      "epoch": 8.895209580838323,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1529,
+      "step": 5942
+    },
+    {
+      "epoch": 8.896706586826348,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1513,
+      "step": 5943
+    },
+    {
+      "epoch": 8.89820359281437,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1531,
+      "step": 5944
+    },
+    {
+      "epoch": 8.899700598802395,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1522,
+      "step": 5945
+    },
+    {
+      "epoch": 8.901197604790418,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 5946
+    },
+    {
+      "epoch": 8.902694610778443,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1465,
+      "step": 5947
+    },
+    {
+      "epoch": 8.904191616766466,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1519,
+      "step": 5948
+    },
+    {
+      "epoch": 8.905688622754491,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 5949
+    },
+    {
+      "epoch": 8.907185628742514,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 5950
+    },
+    {
+      "epoch": 8.908682634730539,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1563,
+      "step": 5951
+    },
+    {
+      "epoch": 8.910179640718562,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1566,
+      "step": 5952
+    },
+    {
+      "epoch": 8.911676646706587,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1518,
+      "step": 5953
+    },
+    {
+      "epoch": 8.91317365269461,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1534,
+      "step": 5954
+    },
+    {
+      "epoch": 8.914670658682635,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1558,
+      "step": 5955
+    },
+    {
+      "epoch": 8.91616766467066,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1513,
+      "step": 5956
+    },
+    {
+      "epoch": 8.917664670658683,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1492,
+      "step": 5957
+    },
+    {
+      "epoch": 8.919161676646706,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1505,
+      "step": 5958
+    },
+    {
+      "epoch": 8.92065868263473,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1516,
+      "step": 5959
+    },
+    {
+      "epoch": 8.922155688622755,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1505,
+      "step": 5960
+    },
+    {
+      "epoch": 8.923652694610778,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1479,
+      "step": 5961
+    },
+    {
+      "epoch": 8.925149700598803,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1429,
+      "step": 5962
+    },
+    {
+      "epoch": 8.926646706586826,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1523,
+      "step": 5963
+    },
+    {
+      "epoch": 8.928143712574851,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1463,
+      "step": 5964
+    },
+    {
+      "epoch": 8.929640718562874,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 5965
+    },
+    {
+      "epoch": 8.931137724550899,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1476,
+      "step": 5966
+    },
+    {
+      "epoch": 8.932634730538922,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 5967
+    },
+    {
+      "epoch": 8.934131736526947,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1482,
+      "step": 5968
+    },
+    {
+      "epoch": 8.93562874251497,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1501,
+      "step": 5969
+    },
+    {
+      "epoch": 8.937125748502995,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1475,
+      "step": 5970
+    },
+    {
+      "epoch": 8.938622754491018,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1513,
+      "step": 5971
+    },
+    {
+      "epoch": 8.940119760479043,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1521,
+      "step": 5972
+    },
+    {
+      "epoch": 8.941616766467066,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1555,
+      "step": 5973
+    },
+    {
+      "epoch": 8.94311377245509,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1535,
+      "step": 5974
+    },
+    {
+      "epoch": 8.944610778443113,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 5975
+    },
+    {
+      "epoch": 8.946107784431138,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1508,
+      "step": 5976
+    },
+    {
+      "epoch": 8.947604790419161,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1509,
+      "step": 5977
+    },
+    {
+      "epoch": 8.949101796407186,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1481,
+      "step": 5978
+    },
+    {
+      "epoch": 8.95059880239521,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 5979
+    },
+    {
+      "epoch": 8.952095808383234,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.144,
+      "step": 5980
+    },
+    {
+      "epoch": 8.953592814371257,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 5981
+    },
+    {
+      "epoch": 8.955089820359282,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1504,
+      "step": 5982
+    },
+    {
+      "epoch": 8.956586826347305,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.16,
+      "step": 5983
+    },
+    {
+      "epoch": 8.95808383233533,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1553,
+      "step": 5984
+    },
+    {
+      "epoch": 8.959580838323353,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1506,
+      "step": 5985
+    },
+    {
+      "epoch": 8.961077844311378,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1524,
+      "step": 5986
+    },
+    {
+      "epoch": 8.9625748502994,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1522,
+      "step": 5987
+    },
+    {
+      "epoch": 8.964071856287426,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 5988
+    },
+    {
+      "epoch": 8.965568862275449,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1485,
+      "step": 5989
+    },
+    {
+      "epoch": 8.967065868263473,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.155,
+      "step": 5990
+    },
+    {
+      "epoch": 8.968562874251496,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 5991
+    },
+    {
+      "epoch": 8.970059880239521,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1563,
+      "step": 5992
+    },
+    {
+      "epoch": 8.971556886227544,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 5993
+    },
+    {
+      "epoch": 8.97305389221557,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1537,
+      "step": 5994
+    },
+    {
+      "epoch": 8.974550898203592,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 5995
+    },
+    {
+      "epoch": 8.976047904191617,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1496,
+      "step": 5996
+    },
+    {
+      "epoch": 8.97754491017964,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1554,
+      "step": 5997
+    },
+    {
+      "epoch": 8.979041916167665,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1557,
+      "step": 5998
+    },
+    {
+      "epoch": 8.980538922155688,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1521,
+      "step": 5999
+    },
+    {
+      "epoch": 8.982035928143713,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 6000
+    },
+    {
+      "epoch": 8.983532934131736,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1517,
+      "step": 6001
+    },
+    {
+      "epoch": 8.98502994011976,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.153,
+      "step": 6002
+    },
+    {
+      "epoch": 8.986526946107784,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1506,
+      "step": 6003
+    },
+    {
+      "epoch": 8.988023952095809,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1492,
+      "step": 6004
+    },
+    {
+      "epoch": 8.989520958083832,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 6005
+    },
+    {
+      "epoch": 8.991017964071856,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 6006
+    },
+    {
+      "epoch": 8.99251497005988,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.149,
+      "step": 6007
+    },
+    {
+      "epoch": 8.994011976047904,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 6008
+    },
+    {
+      "epoch": 8.995508982035929,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 6009
+    },
+    {
+      "epoch": 8.997005988023952,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1459,
+      "step": 6010
+    },
+    {
+      "epoch": 8.998502994011975,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1472,
+      "step": 6011
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 6012
+    },
+    {
+      "epoch": 9.001497005988025,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1485,
+      "step": 6013
+    },
+    {
+      "epoch": 9.002994011976048,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1527,
+      "step": 6014
+    },
+    {
+      "epoch": 9.004491017964073,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1485,
+      "step": 6015
+    },
+    {
+      "epoch": 9.005988023952096,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 6016
+    },
+    {
+      "epoch": 9.00748502994012,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1446,
+      "step": 6017
+    },
+    {
+      "epoch": 9.008982035928144,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1446,
+      "step": 6018
+    },
+    {
+      "epoch": 9.010479041916168,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1507,
+      "step": 6019
+    },
+    {
+      "epoch": 9.011976047904191,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1513,
+      "step": 6020
+    },
+    {
+      "epoch": 9.013473053892216,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1475,
+      "step": 6021
+    },
+    {
+      "epoch": 9.01497005988024,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 6022
+    },
+    {
+      "epoch": 9.016467065868264,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1495,
+      "step": 6023
+    },
+    {
+      "epoch": 9.017964071856287,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1563,
+      "step": 6024
+    },
+    {
+      "epoch": 9.019461077844312,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1521,
+      "step": 6025
+    },
+    {
+      "epoch": 9.020958083832335,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1486,
+      "step": 6026
+    },
+    {
+      "epoch": 9.02245508982036,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1519,
+      "step": 6027
+    },
+    {
+      "epoch": 9.023952095808383,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1488,
+      "step": 6028
+    },
+    {
+      "epoch": 9.025449101796408,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1462,
+      "step": 6029
+    },
+    {
+      "epoch": 9.02694610778443,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 6030
+    },
+    {
+      "epoch": 9.028443113772456,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1512,
+      "step": 6031
+    },
+    {
+      "epoch": 9.029940119760479,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1476,
+      "step": 6032
+    },
+    {
+      "epoch": 9.031437125748504,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1503,
+      "step": 6033
+    },
+    {
+      "epoch": 9.032934131736527,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1481,
+      "step": 6034
+    },
+    {
+      "epoch": 9.034431137724551,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1484,
+      "step": 6035
+    },
+    {
+      "epoch": 9.035928143712574,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1539,
+      "step": 6036
+    },
+    {
+      "epoch": 9.0374251497006,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 6037
+    },
+    {
+      "epoch": 9.038922155688622,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1482,
+      "step": 6038
+    },
+    {
+      "epoch": 9.040419161676647,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1558,
+      "step": 6039
+    },
+    {
+      "epoch": 9.04191616766467,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 6040
+    },
+    {
+      "epoch": 9.043413173652695,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1496,
+      "step": 6041
+    },
+    {
+      "epoch": 9.044910179640718,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.146,
+      "step": 6042
+    },
+    {
+      "epoch": 9.046407185628743,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1496,
+      "step": 6043
+    },
+    {
+      "epoch": 9.047904191616766,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1468,
+      "step": 6044
+    },
+    {
+      "epoch": 9.04940119760479,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1504,
+      "step": 6045
+    },
+    {
+      "epoch": 9.050898203592814,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1533,
+      "step": 6046
+    },
+    {
+      "epoch": 9.052395209580839,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1485,
+      "step": 6047
+    },
+    {
+      "epoch": 9.053892215568862,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1496,
+      "step": 6048
+    },
+    {
+      "epoch": 9.055389221556887,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1497,
+      "step": 6049
+    },
+    {
+      "epoch": 9.05688622754491,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1509,
+      "step": 6050
+    },
+    {
+      "epoch": 9.058383233532934,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 6051
+    },
+    {
+      "epoch": 9.059880239520957,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1359,
+      "step": 6052
+    },
+    {
+      "epoch": 9.061377245508982,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1521,
+      "step": 6053
+    },
+    {
+      "epoch": 9.062874251497005,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1528,
+      "step": 6054
+    },
+    {
+      "epoch": 9.06437125748503,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 6055
+    },
+    {
+      "epoch": 9.065868263473053,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1457,
+      "step": 6056
+    },
+    {
+      "epoch": 9.067365269461078,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1464,
+      "step": 6057
+    },
+    {
+      "epoch": 9.068862275449101,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 6058
+    },
+    {
+      "epoch": 9.070359281437126,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1485,
+      "step": 6059
+    },
+    {
+      "epoch": 9.071856287425149,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1493,
+      "step": 6060
+    },
+    {
+      "epoch": 9.073353293413174,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 6061
+    },
+    {
+      "epoch": 9.074850299401197,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1497,
+      "step": 6062
+    },
+    {
+      "epoch": 9.076347305389222,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1583,
+      "step": 6063
+    },
+    {
+      "epoch": 9.077844311377245,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 6064
+    },
+    {
+      "epoch": 9.07934131736527,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 6065
+    },
+    {
+      "epoch": 9.080838323353293,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 6066
+    },
+    {
+      "epoch": 9.082335329341317,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1545,
+      "step": 6067
+    },
+    {
+      "epoch": 9.08383233532934,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1534,
+      "step": 6068
+    },
+    {
+      "epoch": 9.085329341317365,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1601,
+      "step": 6069
+    },
+    {
+      "epoch": 9.08682634730539,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1495,
+      "step": 6070
+    },
+    {
+      "epoch": 9.088323353293413,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1472,
+      "step": 6071
+    },
+    {
+      "epoch": 9.089820359281438,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1492,
+      "step": 6072
+    },
+    {
+      "epoch": 9.091317365269461,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1494,
+      "step": 6073
+    },
+    {
+      "epoch": 9.092814371257486,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1547,
+      "step": 6074
+    },
+    {
+      "epoch": 9.094311377245509,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1479,
+      "step": 6075
+    },
+    {
+      "epoch": 9.095808383233534,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1526,
+      "step": 6076
+    },
+    {
+      "epoch": 9.097305389221557,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1548,
+      "step": 6077
+    },
+    {
+      "epoch": 9.098802395209582,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1477,
+      "step": 6078
+    },
+    {
+      "epoch": 9.100299401197605,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 6079
+    },
+    {
+      "epoch": 9.10179640718563,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1546,
+      "step": 6080
+    },
+    {
+      "epoch": 9.103293413173652,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 6081
+    },
+    {
+      "epoch": 9.104790419161677,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1502,
+      "step": 6082
+    },
+    {
+      "epoch": 9.1062874251497,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 6083
+    },
+    {
+      "epoch": 9.107784431137725,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1468,
+      "step": 6084
+    },
+    {
+      "epoch": 9.109281437125748,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1506,
+      "step": 6085
+    },
+    {
+      "epoch": 9.110778443113773,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1477,
+      "step": 6086
+    },
+    {
+      "epoch": 9.112275449101796,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 6087
+    },
+    {
+      "epoch": 9.113772455089821,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 6088
+    },
+    {
+      "epoch": 9.115269461077844,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1593,
+      "step": 6089
+    },
+    {
+      "epoch": 9.116766467065869,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1517,
+      "step": 6090
+    },
+    {
+      "epoch": 9.118263473053892,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 6091
+    },
+    {
+      "epoch": 9.119760479041917,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1557,
+      "step": 6092
+    },
+    {
+      "epoch": 9.12125748502994,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1469,
+      "step": 6093
+    },
+    {
+      "epoch": 9.122754491017965,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1515,
+      "step": 6094
+    },
+    {
+      "epoch": 9.124251497005988,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1503,
+      "step": 6095
+    },
+    {
+      "epoch": 9.125748502994012,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1503,
+      "step": 6096
+    },
+    {
+      "epoch": 9.127245508982035,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 6097
+    },
+    {
+      "epoch": 9.12874251497006,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1541,
+      "step": 6098
+    },
+    {
+      "epoch": 9.130239520958083,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1507,
+      "step": 6099
+    },
+    {
+      "epoch": 9.131736526946108,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 6100
+    },
+    {
+      "epoch": 9.133233532934131,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0008,
+      "loss": 1.153,
+      "step": 6101
+    },
+    {
+      "epoch": 9.134730538922156,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 6102
+    },
+    {
+      "epoch": 9.136227544910179,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1488,
+      "step": 6103
+    },
+    {
+      "epoch": 9.137724550898204,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1498,
+      "step": 6104
+    },
+    {
+      "epoch": 9.139221556886227,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1462,
+      "step": 6105
+    },
+    {
+      "epoch": 9.140718562874252,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1578,
+      "step": 6106
+    },
+    {
+      "epoch": 9.142215568862275,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1524,
+      "step": 6107
+    },
+    {
+      "epoch": 9.1437125748503,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1556,
+      "step": 6108
+    },
+    {
+      "epoch": 9.145209580838323,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1516,
+      "step": 6109
+    },
+    {
+      "epoch": 9.146706586826348,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1469,
+      "step": 6110
+    },
+    {
+      "epoch": 9.14820359281437,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1515,
+      "step": 6111
+    },
+    {
+      "epoch": 9.149700598802395,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1535,
+      "step": 6112
+    },
+    {
+      "epoch": 9.151197604790418,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1524,
+      "step": 6113
+    },
+    {
+      "epoch": 9.152694610778443,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 6114
+    },
+    {
+      "epoch": 9.154191616766466,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1481,
+      "step": 6115
+    },
+    {
+      "epoch": 9.155688622754491,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1488,
+      "step": 6116
+    },
+    {
+      "epoch": 9.157185628742514,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1464,
+      "step": 6117
+    },
+    {
+      "epoch": 9.158682634730539,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1508,
+      "step": 6118
+    },
+    {
+      "epoch": 9.160179640718562,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1531,
+      "step": 6119
+    },
+    {
+      "epoch": 9.161676646706587,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1477,
+      "step": 6120
+    },
+    {
+      "epoch": 9.16317365269461,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 6121
+    },
+    {
+      "epoch": 9.164670658682635,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1506,
+      "step": 6122
+    },
+    {
+      "epoch": 9.16616766467066,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1487,
+      "step": 6123
+    },
+    {
+      "epoch": 9.167664670658683,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1502,
+      "step": 6124
+    },
+    {
+      "epoch": 9.169161676646707,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 6125
+    },
+    {
+      "epoch": 9.17065868263473,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1558,
+      "step": 6126
+    },
+    {
+      "epoch": 9.172155688622755,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1531,
+      "step": 6127
+    },
+    {
+      "epoch": 9.173652694610778,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1492,
+      "step": 6128
+    },
+    {
+      "epoch": 9.175149700598803,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1495,
+      "step": 6129
+    },
+    {
+      "epoch": 9.176646706586826,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1531,
+      "step": 6130
+    },
+    {
+      "epoch": 9.178143712574851,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1484,
+      "step": 6131
+    },
+    {
+      "epoch": 9.179640718562874,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 6132
+    },
+    {
+      "epoch": 9.181137724550899,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1551,
+      "step": 6133
+    },
+    {
+      "epoch": 9.182634730538922,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.151,
+      "step": 6134
+    },
+    {
+      "epoch": 9.184131736526947,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1463,
+      "step": 6135
+    },
+    {
+      "epoch": 9.18562874251497,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1511,
+      "step": 6136
+    },
+    {
+      "epoch": 9.187125748502995,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1496,
+      "step": 6137
+    },
+    {
+      "epoch": 9.188622754491018,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1531,
+      "step": 6138
+    },
+    {
+      "epoch": 9.190119760479043,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1485,
+      "step": 6139
+    },
+    {
+      "epoch": 9.191616766467066,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 6140
+    },
+    {
+      "epoch": 9.19311377245509,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1535,
+      "step": 6141
+    },
+    {
+      "epoch": 9.194610778443113,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1547,
+      "step": 6142
+    },
+    {
+      "epoch": 9.196107784431138,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1555,
+      "step": 6143
+    },
+    {
+      "epoch": 9.197604790419161,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1548,
+      "step": 6144
+    },
+    {
+      "epoch": 9.199101796407186,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 6145
+    },
+    {
+      "epoch": 9.20059880239521,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 6146
+    },
+    {
+      "epoch": 9.202095808383234,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.151,
+      "step": 6147
+    },
+    {
+      "epoch": 9.203592814371257,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 6148
+    },
+    {
+      "epoch": 9.205089820359282,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1504,
+      "step": 6149
+    },
+    {
+      "epoch": 9.206586826347305,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1462,
+      "step": 6150
+    },
+    {
+      "epoch": 9.20808383233533,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.146,
+      "step": 6151
+    },
+    {
+      "epoch": 9.209580838323353,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1485,
+      "step": 6152
+    },
+    {
+      "epoch": 9.211077844311378,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 6153
+    },
+    {
+      "epoch": 9.2125748502994,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1551,
+      "step": 6154
+    },
+    {
+      "epoch": 9.214071856287426,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 6155
+    },
+    {
+      "epoch": 9.215568862275449,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1522,
+      "step": 6156
+    },
+    {
+      "epoch": 9.217065868263473,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1485,
+      "step": 6157
+    },
+    {
+      "epoch": 9.218562874251496,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1522,
+      "step": 6158
+    },
+    {
+      "epoch": 9.220059880239521,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1518,
+      "step": 6159
+    },
+    {
+      "epoch": 9.221556886227544,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1508,
+      "step": 6160
+    },
+    {
+      "epoch": 9.22305389221557,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1537,
+      "step": 6161
+    },
+    {
+      "epoch": 9.224550898203592,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 6162
+    },
+    {
+      "epoch": 9.226047904191617,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 6163
+    },
+    {
+      "epoch": 9.22754491017964,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1472,
+      "step": 6164
+    },
+    {
+      "epoch": 9.229041916167665,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1537,
+      "step": 6165
+    },
+    {
+      "epoch": 9.230538922155688,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1488,
+      "step": 6166
+    },
+    {
+      "epoch": 9.232035928143713,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1476,
+      "step": 6167
+    },
+    {
+      "epoch": 9.233532934131736,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1476,
+      "step": 6168
+    },
+    {
+      "epoch": 9.23502994011976,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1464,
+      "step": 6169
+    },
+    {
+      "epoch": 9.236526946107784,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1496,
+      "step": 6170
+    },
+    {
+      "epoch": 9.238023952095809,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 6171
+    },
+    {
+      "epoch": 9.239520958083832,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1537,
+      "step": 6172
+    },
+    {
+      "epoch": 9.241017964071856,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1481,
+      "step": 6173
+    },
+    {
+      "epoch": 9.24251497005988,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 6174
+    },
+    {
+      "epoch": 9.244011976047904,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1534,
+      "step": 6175
+    },
+    {
+      "epoch": 9.245508982035927,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1485,
+      "step": 6176
+    },
+    {
+      "epoch": 9.247005988023952,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1522,
+      "step": 6177
+    },
+    {
+      "epoch": 9.248502994011975,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 6178
+    },
+    {
+      "epoch": 9.25,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1462,
+      "step": 6179
+    },
+    {
+      "epoch": 9.251497005988025,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1536,
+      "step": 6180
+    },
+    {
+      "epoch": 9.252994011976048,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1485,
+      "step": 6181
+    },
+    {
+      "epoch": 9.254491017964073,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1553,
+      "step": 6182
+    },
+    {
+      "epoch": 9.255988023952096,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1465,
+      "step": 6183
+    },
+    {
+      "epoch": 9.25748502994012,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 6184
+    },
+    {
+      "epoch": 9.258982035928144,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1493,
+      "step": 6185
+    },
+    {
+      "epoch": 9.260479041916168,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 6186
+    },
+    {
+      "epoch": 9.261976047904191,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 6187
+    },
+    {
+      "epoch": 9.263473053892216,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1511,
+      "step": 6188
+    },
+    {
+      "epoch": 9.26497005988024,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1544,
+      "step": 6189
+    },
+    {
+      "epoch": 9.266467065868264,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1511,
+      "step": 6190
+    },
+    {
+      "epoch": 9.267964071856287,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 6191
+    },
+    {
+      "epoch": 9.269461077844312,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1484,
+      "step": 6192
+    },
+    {
+      "epoch": 9.270958083832335,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1492,
+      "step": 6193
+    },
+    {
+      "epoch": 9.27245508982036,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1502,
+      "step": 6194
+    },
+    {
+      "epoch": 9.273952095808383,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1513,
+      "step": 6195
+    },
+    {
+      "epoch": 9.275449101796408,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1488,
+      "step": 6196
+    },
+    {
+      "epoch": 9.27694610778443,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 6197
+    },
+    {
+      "epoch": 9.278443113772456,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1511,
+      "step": 6198
+    },
+    {
+      "epoch": 9.279940119760479,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1477,
+      "step": 6199
+    },
+    {
+      "epoch": 9.281437125748504,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1549,
+      "step": 6200
+    },
+    {
+      "epoch": 9.282934131736527,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1515,
+      "step": 6201
+    },
+    {
+      "epoch": 9.284431137724551,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1481,
+      "step": 6202
+    },
+    {
+      "epoch": 9.285928143712574,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1454,
+      "step": 6203
+    },
+    {
+      "epoch": 9.2874251497006,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 6204
+    },
+    {
+      "epoch": 9.288922155688622,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1602,
+      "step": 6205
+    },
+    {
+      "epoch": 9.290419161676647,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.149,
+      "step": 6206
+    },
+    {
+      "epoch": 9.29191616766467,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 6207
+    },
+    {
+      "epoch": 9.293413173652695,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1485,
+      "step": 6208
+    },
+    {
+      "epoch": 9.294910179640718,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1493,
+      "step": 6209
+    },
+    {
+      "epoch": 9.296407185628743,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1463,
+      "step": 6210
+    },
+    {
+      "epoch": 9.297904191616766,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1479,
+      "step": 6211
+    },
+    {
+      "epoch": 9.29940119760479,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1488,
+      "step": 6212
+    },
+    {
+      "epoch": 9.300898203592814,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1476,
+      "step": 6213
+    },
+    {
+      "epoch": 9.302395209580839,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1465,
+      "step": 6214
+    },
+    {
+      "epoch": 9.303892215568862,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1468,
+      "step": 6215
+    },
+    {
+      "epoch": 9.305389221556887,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1491,
+      "step": 6216
+    },
+    {
+      "epoch": 9.30688622754491,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1428,
+      "step": 6217
+    },
+    {
+      "epoch": 9.308383233532934,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1479,
+      "step": 6218
+    },
+    {
+      "epoch": 9.309880239520957,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 6219
+    },
+    {
+      "epoch": 9.311377245508982,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1536,
+      "step": 6220
+    },
+    {
+      "epoch": 9.312874251497005,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 6221
+    },
+    {
+      "epoch": 9.31437125748503,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 6222
+    },
+    {
+      "epoch": 9.315868263473053,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1457,
+      "step": 6223
+    },
+    {
+      "epoch": 9.317365269461078,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1522,
+      "step": 6224
+    },
+    {
+      "epoch": 9.318862275449101,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1539,
+      "step": 6225
+    },
+    {
+      "epoch": 9.320359281437126,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1517,
+      "step": 6226
+    },
+    {
+      "epoch": 9.321856287425149,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1487,
+      "step": 6227
+    },
+    {
+      "epoch": 9.323353293413174,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1491,
+      "step": 6228
+    },
+    {
+      "epoch": 9.324850299401197,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 6229
+    },
+    {
+      "epoch": 9.326347305389222,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1505,
+      "step": 6230
+    },
+    {
+      "epoch": 9.327844311377245,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1514,
+      "step": 6231
+    },
+    {
+      "epoch": 9.32934131736527,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 6232
+    },
+    {
+      "epoch": 9.330838323353294,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1507,
+      "step": 6233
+    },
+    {
+      "epoch": 9.332335329341317,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1529,
+      "step": 6234
+    },
+    {
+      "epoch": 9.33383233532934,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 6235
+    },
+    {
+      "epoch": 9.335329341317365,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1535,
+      "step": 6236
+    },
+    {
+      "epoch": 9.33682634730539,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1495,
+      "step": 6237
+    },
+    {
+      "epoch": 9.338323353293413,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 6238
+    },
+    {
+      "epoch": 9.339820359281438,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1497,
+      "step": 6239
+    },
+    {
+      "epoch": 9.341317365269461,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1484,
+      "step": 6240
+    },
+    {
+      "epoch": 9.342814371257486,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.147,
+      "step": 6241
+    },
+    {
+      "epoch": 9.344311377245509,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 6242
+    },
+    {
+      "epoch": 9.345808383233534,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1477,
+      "step": 6243
+    },
+    {
+      "epoch": 9.347305389221557,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1499,
+      "step": 6244
+    },
+    {
+      "epoch": 9.348802395209582,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 6245
+    },
+    {
+      "epoch": 9.350299401197605,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1492,
+      "step": 6246
+    },
+    {
+      "epoch": 9.35179640718563,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 6247
+    },
+    {
+      "epoch": 9.353293413173652,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 6248
+    },
+    {
+      "epoch": 9.354790419161677,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1483,
+      "step": 6249
+    },
+    {
+      "epoch": 9.3562874251497,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1483,
+      "step": 6250
+    },
+    {
+      "epoch": 9.357784431137725,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 6251
+    },
+    {
+      "epoch": 9.359281437125748,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1523,
+      "step": 6252
+    },
+    {
+      "epoch": 9.360778443113773,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1521,
+      "step": 6253
+    },
+    {
+      "epoch": 9.362275449101796,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1538,
+      "step": 6254
+    },
+    {
+      "epoch": 9.363772455089821,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1507,
+      "step": 6255
+    },
+    {
+      "epoch": 9.365269461077844,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1498,
+      "step": 6256
+    },
+    {
+      "epoch": 9.366766467065869,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.149,
+      "step": 6257
+    },
+    {
+      "epoch": 9.368263473053892,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.148,
+      "step": 6258
+    },
+    {
+      "epoch": 9.369760479041917,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1528,
+      "step": 6259
+    },
+    {
+      "epoch": 9.37125748502994,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 6260
+    },
+    {
+      "epoch": 9.372754491017965,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1476,
+      "step": 6261
+    },
+    {
+      "epoch": 9.374251497005988,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 6262
+    },
+    {
+      "epoch": 9.375748502994012,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1547,
+      "step": 6263
+    },
+    {
+      "epoch": 9.377245508982035,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 6264
+    },
+    {
+      "epoch": 9.37874251497006,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1509,
+      "step": 6265
+    },
+    {
+      "epoch": 9.380239520958083,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 6266
+    },
+    {
+      "epoch": 9.381736526946108,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1507,
+      "step": 6267
+    },
+    {
+      "epoch": 9.383233532934131,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.152,
+      "step": 6268
+    },
+    {
+      "epoch": 9.384730538922156,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1499,
+      "step": 6269
+    },
+    {
+      "epoch": 9.386227544910179,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1463,
+      "step": 6270
+    },
+    {
+      "epoch": 9.387724550898204,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1497,
+      "step": 6271
+    },
+    {
+      "epoch": 9.389221556886227,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1492,
+      "step": 6272
+    },
+    {
+      "epoch": 9.390718562874252,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1525,
+      "step": 6273
+    },
+    {
+      "epoch": 9.392215568862275,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.149,
+      "step": 6274
+    },
+    {
+      "epoch": 9.3937125748503,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1516,
+      "step": 6275
+    },
+    {
+      "epoch": 9.395209580838323,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1434,
+      "step": 6276
+    },
+    {
+      "epoch": 9.396706586826348,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1528,
+      "step": 6277
+    },
+    {
+      "epoch": 9.39820359281437,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.153,
+      "step": 6278
+    },
+    {
+      "epoch": 9.399700598802395,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.144,
+      "step": 6279
+    },
+    {
+      "epoch": 9.401197604790418,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1465,
+      "step": 6280
+    },
+    {
+      "epoch": 9.402694610778443,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1483,
+      "step": 6281
+    },
+    {
+      "epoch": 9.404191616766466,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 6282
+    },
+    {
+      "epoch": 9.405688622754491,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 6283
+    },
+    {
+      "epoch": 9.407185628742514,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 6284
+    },
+    {
+      "epoch": 9.408682634730539,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1473,
+      "step": 6285
+    },
+    {
+      "epoch": 9.410179640718562,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 6286
+    },
+    {
+      "epoch": 9.411676646706587,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1469,
+      "step": 6287
+    },
+    {
+      "epoch": 9.41317365269461,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1533,
+      "step": 6288
+    },
+    {
+      "epoch": 9.414670658682635,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 6289
+    },
+    {
+      "epoch": 9.41616766467066,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 6290
+    },
+    {
+      "epoch": 9.417664670658683,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1464,
+      "step": 6291
+    },
+    {
+      "epoch": 9.419161676646706,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1503,
+      "step": 6292
+    },
+    {
+      "epoch": 9.42065868263473,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1536,
+      "step": 6293
+    },
+    {
+      "epoch": 9.422155688622755,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1446,
+      "step": 6294
+    },
+    {
+      "epoch": 9.423652694610778,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1566,
+      "step": 6295
+    },
+    {
+      "epoch": 9.425149700598803,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1492,
+      "step": 6296
+    },
+    {
+      "epoch": 9.426646706586826,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1499,
+      "step": 6297
+    },
+    {
+      "epoch": 9.428143712574851,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1528,
+      "step": 6298
+    },
+    {
+      "epoch": 9.429640718562874,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1472,
+      "step": 6299
+    },
+    {
+      "epoch": 9.431137724550899,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1465,
+      "step": 6300
+    },
+    {
+      "epoch": 9.432634730538922,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 6301
+    },
+    {
+      "epoch": 9.434131736526947,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 6302
+    },
+    {
+      "epoch": 9.43562874251497,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1521,
+      "step": 6303
+    },
+    {
+      "epoch": 9.437125748502995,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1491,
+      "step": 6304
+    },
+    {
+      "epoch": 9.438622754491018,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1506,
+      "step": 6305
+    },
+    {
+      "epoch": 9.440119760479043,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 6306
+    },
+    {
+      "epoch": 9.441616766467066,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1541,
+      "step": 6307
+    },
+    {
+      "epoch": 9.44311377245509,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 6308
+    },
+    {
+      "epoch": 9.444610778443113,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.147,
+      "step": 6309
+    },
+    {
+      "epoch": 9.446107784431138,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.15,
+      "step": 6310
+    },
+    {
+      "epoch": 9.447604790419161,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1509,
+      "step": 6311
+    },
+    {
+      "epoch": 9.449101796407186,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1563,
+      "step": 6312
+    },
+    {
+      "epoch": 9.45059880239521,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1548,
+      "step": 6313
+    },
+    {
+      "epoch": 9.452095808383234,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1504,
+      "step": 6314
+    },
+    {
+      "epoch": 9.453592814371257,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1546,
+      "step": 6315
+    },
+    {
+      "epoch": 9.455089820359282,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1477,
+      "step": 6316
+    },
+    {
+      "epoch": 9.456586826347305,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1509,
+      "step": 6317
+    },
+    {
+      "epoch": 9.45808383233533,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1488,
+      "step": 6318
+    },
+    {
+      "epoch": 9.459580838323353,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1504,
+      "step": 6319
+    },
+    {
+      "epoch": 9.461077844311378,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.15,
+      "step": 6320
+    },
+    {
+      "epoch": 9.4625748502994,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 6321
+    },
+    {
+      "epoch": 9.464071856287426,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1508,
+      "step": 6322
+    },
+    {
+      "epoch": 9.465568862275449,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1513,
+      "step": 6323
+    },
+    {
+      "epoch": 9.467065868263473,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1476,
+      "step": 6324
+    },
+    {
+      "epoch": 9.468562874251496,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1465,
+      "step": 6325
+    },
+    {
+      "epoch": 9.470059880239521,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1523,
+      "step": 6326
+    },
+    {
+      "epoch": 9.471556886227544,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1506,
+      "step": 6327
+    },
+    {
+      "epoch": 9.47305389221557,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1462,
+      "step": 6328
+    },
+    {
+      "epoch": 9.474550898203592,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1485,
+      "step": 6329
+    },
+    {
+      "epoch": 9.476047904191617,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1462,
+      "step": 6330
+    },
+    {
+      "epoch": 9.47754491017964,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1484,
+      "step": 6331
+    },
+    {
+      "epoch": 9.479041916167665,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 6332
+    },
+    {
+      "epoch": 9.480538922155688,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 6333
+    },
+    {
+      "epoch": 9.482035928143713,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1495,
+      "step": 6334
+    },
+    {
+      "epoch": 9.483532934131736,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1561,
+      "step": 6335
+    },
+    {
+      "epoch": 9.48502994011976,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1462,
+      "step": 6336
+    },
+    {
+      "epoch": 9.486526946107784,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1494,
+      "step": 6337
+    },
+    {
+      "epoch": 9.488023952095809,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 6338
+    },
+    {
+      "epoch": 9.489520958083832,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 6339
+    },
+    {
+      "epoch": 9.491017964071856,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1476,
+      "step": 6340
+    },
+    {
+      "epoch": 9.49251497005988,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 6341
+    },
+    {
+      "epoch": 9.494011976047904,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 6342
+    },
+    {
+      "epoch": 9.495508982035927,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1516,
+      "step": 6343
+    },
+    {
+      "epoch": 9.497005988023952,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 6344
+    },
+    {
+      "epoch": 9.498502994011975,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1554,
+      "step": 6345
+    },
+    {
+      "epoch": 9.5,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1529,
+      "step": 6346
+    },
+    {
+      "epoch": 9.501497005988025,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 6347
+    },
+    {
+      "epoch": 9.502994011976048,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1505,
+      "step": 6348
+    },
+    {
+      "epoch": 9.504491017964071,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1457,
+      "step": 6349
+    },
+    {
+      "epoch": 9.505988023952096,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1464,
+      "step": 6350
+    },
+    {
+      "epoch": 9.50748502994012,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 6351
+    },
+    {
+      "epoch": 9.508982035928144,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1536,
+      "step": 6352
+    },
+    {
+      "epoch": 9.510479041916168,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.153,
+      "step": 6353
+    },
+    {
+      "epoch": 9.511976047904191,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1536,
+      "step": 6354
+    },
+    {
+      "epoch": 9.513473053892216,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 6355
+    },
+    {
+      "epoch": 9.51497005988024,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 6356
+    },
+    {
+      "epoch": 9.516467065868264,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1502,
+      "step": 6357
+    },
+    {
+      "epoch": 9.517964071856287,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.152,
+      "step": 6358
+    },
+    {
+      "epoch": 9.519461077844312,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 6359
+    },
+    {
+      "epoch": 9.520958083832335,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1471,
+      "step": 6360
+    },
+    {
+      "epoch": 9.52245508982036,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1476,
+      "step": 6361
+    },
+    {
+      "epoch": 9.523952095808383,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1496,
+      "step": 6362
+    },
+    {
+      "epoch": 9.525449101796408,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 6363
+    },
+    {
+      "epoch": 9.52694610778443,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 6364
+    },
+    {
+      "epoch": 9.528443113772456,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1491,
+      "step": 6365
+    },
+    {
+      "epoch": 9.529940119760479,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1514,
+      "step": 6366
+    },
+    {
+      "epoch": 9.531437125748504,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1476,
+      "step": 6367
+    },
+    {
+      "epoch": 9.532934131736527,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1595,
+      "step": 6368
+    },
+    {
+      "epoch": 9.534431137724551,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 6369
+    },
+    {
+      "epoch": 9.535928143712574,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1529,
+      "step": 6370
+    },
+    {
+      "epoch": 9.5374251497006,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1457,
+      "step": 6371
+    },
+    {
+      "epoch": 9.538922155688622,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1502,
+      "step": 6372
+    },
+    {
+      "epoch": 9.540419161676647,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 6373
+    },
+    {
+      "epoch": 9.54191616766467,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1481,
+      "step": 6374
+    },
+    {
+      "epoch": 9.543413173652695,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1487,
+      "step": 6375
+    },
+    {
+      "epoch": 9.544910179640718,
+      "grad_norm": 0.19140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1488,
+      "step": 6376
+    },
+    {
+      "epoch": 9.546407185628743,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1521,
+      "step": 6377
+    },
+    {
+      "epoch": 9.547904191616766,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 6378
+    },
+    {
+      "epoch": 9.54940119760479,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1528,
+      "step": 6379
+    },
+    {
+      "epoch": 9.550898203592814,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1486,
+      "step": 6380
+    },
+    {
+      "epoch": 9.552395209580839,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1534,
+      "step": 6381
+    },
+    {
+      "epoch": 9.553892215568862,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1472,
+      "step": 6382
+    },
+    {
+      "epoch": 9.555389221556887,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1469,
+      "step": 6383
+    },
+    {
+      "epoch": 9.55688622754491,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 6384
+    },
+    {
+      "epoch": 9.558383233532934,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1543,
+      "step": 6385
+    },
+    {
+      "epoch": 9.559880239520957,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1479,
+      "step": 6386
+    },
+    {
+      "epoch": 9.561377245508982,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1446,
+      "step": 6387
+    },
+    {
+      "epoch": 9.562874251497005,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.151,
+      "step": 6388
+    },
+    {
+      "epoch": 9.56437125748503,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.149,
+      "step": 6389
+    },
+    {
+      "epoch": 9.565868263473053,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1507,
+      "step": 6390
+    },
+    {
+      "epoch": 9.567365269461078,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 6391
+    },
+    {
+      "epoch": 9.568862275449101,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 6392
+    },
+    {
+      "epoch": 9.570359281437126,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1511,
+      "step": 6393
+    },
+    {
+      "epoch": 9.571856287425149,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1516,
+      "step": 6394
+    },
+    {
+      "epoch": 9.573353293413174,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1533,
+      "step": 6395
+    },
+    {
+      "epoch": 9.574850299401197,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1472,
+      "step": 6396
+    },
+    {
+      "epoch": 9.576347305389222,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1505,
+      "step": 6397
+    },
+    {
+      "epoch": 9.577844311377245,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1492,
+      "step": 6398
+    },
+    {
+      "epoch": 9.57934131736527,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.144,
+      "step": 6399
+    },
+    {
+      "epoch": 9.580838323353294,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1504,
+      "step": 6400
+    },
+    {
+      "epoch": 9.582335329341317,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1561,
+      "step": 6401
+    },
+    {
+      "epoch": 9.58383233532934,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 6402
+    },
+    {
+      "epoch": 9.585329341317365,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.149,
+      "step": 6403
+    },
+    {
+      "epoch": 9.58682634730539,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1508,
+      "step": 6404
+    },
+    {
+      "epoch": 9.588323353293413,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1493,
+      "step": 6405
+    },
+    {
+      "epoch": 9.589820359281438,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1529,
+      "step": 6406
+    },
+    {
+      "epoch": 9.591317365269461,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1547,
+      "step": 6407
+    },
+    {
+      "epoch": 9.592814371257486,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1502,
+      "step": 6408
+    },
+    {
+      "epoch": 9.594311377245509,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 6409
+    },
+    {
+      "epoch": 9.595808383233534,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1503,
+      "step": 6410
+    },
+    {
+      "epoch": 9.597305389221557,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 6411
+    },
+    {
+      "epoch": 9.598802395209582,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1479,
+      "step": 6412
+    },
+    {
+      "epoch": 9.600299401197605,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1509,
+      "step": 6413
+    },
+    {
+      "epoch": 9.60179640718563,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1454,
+      "step": 6414
+    },
+    {
+      "epoch": 9.603293413173652,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1494,
+      "step": 6415
+    },
+    {
+      "epoch": 9.604790419161677,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.151,
+      "step": 6416
+    },
+    {
+      "epoch": 9.6062874251497,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1486,
+      "step": 6417
+    },
+    {
+      "epoch": 9.607784431137725,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1536,
+      "step": 6418
+    },
+    {
+      "epoch": 9.609281437125748,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1529,
+      "step": 6419
+    },
+    {
+      "epoch": 9.610778443113773,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1573,
+      "step": 6420
+    },
+    {
+      "epoch": 9.612275449101796,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1429,
+      "step": 6421
+    },
+    {
+      "epoch": 9.613772455089821,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1494,
+      "step": 6422
+    },
+    {
+      "epoch": 9.615269461077844,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.148,
+      "step": 6423
+    },
+    {
+      "epoch": 9.616766467065869,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.147,
+      "step": 6424
+    },
+    {
+      "epoch": 9.618263473053892,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 6425
+    },
+    {
+      "epoch": 9.619760479041917,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1495,
+      "step": 6426
+    },
+    {
+      "epoch": 9.62125748502994,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1527,
+      "step": 6427
+    },
+    {
+      "epoch": 9.622754491017965,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.15,
+      "step": 6428
+    },
+    {
+      "epoch": 9.624251497005988,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1491,
+      "step": 6429
+    },
+    {
+      "epoch": 9.625748502994012,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 6430
+    },
+    {
+      "epoch": 9.627245508982035,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 6431
+    },
+    {
+      "epoch": 9.62874251497006,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 6432
+    },
+    {
+      "epoch": 9.630239520958083,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1528,
+      "step": 6433
+    },
+    {
+      "epoch": 9.631736526946108,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.15,
+      "step": 6434
+    },
+    {
+      "epoch": 9.633233532934131,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1508,
+      "step": 6435
+    },
+    {
+      "epoch": 9.634730538922156,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1469,
+      "step": 6436
+    },
+    {
+      "epoch": 9.636227544910179,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 6437
+    },
+    {
+      "epoch": 9.637724550898204,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1496,
+      "step": 6438
+    },
+    {
+      "epoch": 9.639221556886227,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1514,
+      "step": 6439
+    },
+    {
+      "epoch": 9.640718562874252,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 6440
+    },
+    {
+      "epoch": 9.642215568862275,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1506,
+      "step": 6441
+    },
+    {
+      "epoch": 9.6437125748503,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1533,
+      "step": 6442
+    },
+    {
+      "epoch": 9.645209580838323,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 6443
+    },
+    {
+      "epoch": 9.646706586826348,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1524,
+      "step": 6444
+    },
+    {
+      "epoch": 9.64820359281437,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1497,
+      "step": 6445
+    },
+    {
+      "epoch": 9.649700598802395,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1524,
+      "step": 6446
+    },
+    {
+      "epoch": 9.651197604790418,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1475,
+      "step": 6447
+    },
+    {
+      "epoch": 9.652694610778443,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1482,
+      "step": 6448
+    },
+    {
+      "epoch": 9.654191616766466,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1499,
+      "step": 6449
+    },
+    {
+      "epoch": 9.655688622754491,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1554,
+      "step": 6450
+    },
+    {
+      "epoch": 9.657185628742514,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1495,
+      "step": 6451
+    },
+    {
+      "epoch": 9.658682634730539,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1481,
+      "step": 6452
+    },
+    {
+      "epoch": 9.660179640718562,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 6453
+    },
+    {
+      "epoch": 9.661676646706587,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 6454
+    },
+    {
+      "epoch": 9.66317365269461,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1506,
+      "step": 6455
+    },
+    {
+      "epoch": 9.664670658682635,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1495,
+      "step": 6456
+    },
+    {
+      "epoch": 9.66616766467066,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1534,
+      "step": 6457
+    },
+    {
+      "epoch": 9.667664670658683,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 6458
+    },
+    {
+      "epoch": 9.669161676646706,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.147,
+      "step": 6459
+    },
+    {
+      "epoch": 9.67065868263473,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1502,
+      "step": 6460
+    },
+    {
+      "epoch": 9.672155688622755,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1504,
+      "step": 6461
+    },
+    {
+      "epoch": 9.673652694610778,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 6462
+    },
+    {
+      "epoch": 9.675149700598803,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0008,
+      "loss": 1.148,
+      "step": 6463
+    },
+    {
+      "epoch": 9.676646706586826,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.154,
+      "step": 6464
+    },
+    {
+      "epoch": 9.678143712574851,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.15,
+      "step": 6465
+    },
+    {
+      "epoch": 9.679640718562874,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 6466
+    },
+    {
+      "epoch": 9.681137724550899,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 6467
+    },
+    {
+      "epoch": 9.682634730538922,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1516,
+      "step": 6468
+    },
+    {
+      "epoch": 9.684131736526947,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1468,
+      "step": 6469
+    },
+    {
+      "epoch": 9.68562874251497,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1531,
+      "step": 6470
+    },
+    {
+      "epoch": 9.687125748502995,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1536,
+      "step": 6471
+    },
+    {
+      "epoch": 9.688622754491018,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 6472
+    },
+    {
+      "epoch": 9.690119760479043,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1506,
+      "step": 6473
+    },
+    {
+      "epoch": 9.691616766467066,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 6474
+    },
+    {
+      "epoch": 9.69311377245509,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1529,
+      "step": 6475
+    },
+    {
+      "epoch": 9.694610778443113,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 6476
+    },
+    {
+      "epoch": 9.696107784431138,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1497,
+      "step": 6477
+    },
+    {
+      "epoch": 9.697604790419161,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1486,
+      "step": 6478
+    },
+    {
+      "epoch": 9.699101796407186,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 6479
+    },
+    {
+      "epoch": 9.70059880239521,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 6480
+    },
+    {
+      "epoch": 9.702095808383234,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1513,
+      "step": 6481
+    },
+    {
+      "epoch": 9.703592814371257,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1539,
+      "step": 6482
+    },
+    {
+      "epoch": 9.705089820359282,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 6483
+    },
+    {
+      "epoch": 9.706586826347305,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 6484
+    },
+    {
+      "epoch": 9.70808383233533,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 6485
+    },
+    {
+      "epoch": 9.709580838323353,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1497,
+      "step": 6486
+    },
+    {
+      "epoch": 9.711077844311378,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 6487
+    },
+    {
+      "epoch": 9.7125748502994,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.152,
+      "step": 6488
+    },
+    {
+      "epoch": 9.714071856287426,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1462,
+      "step": 6489
+    },
+    {
+      "epoch": 9.715568862275449,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1524,
+      "step": 6490
+    },
+    {
+      "epoch": 9.717065868263473,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1506,
+      "step": 6491
+    },
+    {
+      "epoch": 9.718562874251496,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1446,
+      "step": 6492
+    },
+    {
+      "epoch": 9.720059880239521,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.152,
+      "step": 6493
+    },
+    {
+      "epoch": 9.721556886227544,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1498,
+      "step": 6494
+    },
+    {
+      "epoch": 9.72305389221557,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1547,
+      "step": 6495
+    },
+    {
+      "epoch": 9.724550898203592,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1498,
+      "step": 6496
+    },
+    {
+      "epoch": 9.726047904191617,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.15,
+      "step": 6497
+    },
+    {
+      "epoch": 9.72754491017964,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1546,
+      "step": 6498
+    },
+    {
+      "epoch": 9.729041916167665,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1518,
+      "step": 6499
+    },
+    {
+      "epoch": 9.730538922155688,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1491,
+      "step": 6500
+    },
+    {
+      "epoch": 9.732035928143713,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.152,
+      "step": 6501
+    },
+    {
+      "epoch": 9.733532934131736,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 6502
+    },
+    {
+      "epoch": 9.73502994011976,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1529,
+      "step": 6503
+    },
+    {
+      "epoch": 9.736526946107784,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1463,
+      "step": 6504
+    },
+    {
+      "epoch": 9.738023952095809,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1503,
+      "step": 6505
+    },
+    {
+      "epoch": 9.739520958083832,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1482,
+      "step": 6506
+    },
+    {
+      "epoch": 9.741017964071856,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1531,
+      "step": 6507
+    },
+    {
+      "epoch": 9.74251497005988,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1524,
+      "step": 6508
+    },
+    {
+      "epoch": 9.744011976047904,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1553,
+      "step": 6509
+    },
+    {
+      "epoch": 9.745508982035929,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 6510
+    },
+    {
+      "epoch": 9.747005988023952,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 6511
+    },
+    {
+      "epoch": 9.748502994011975,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1495,
+      "step": 6512
+    },
+    {
+      "epoch": 9.75,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1472,
+      "step": 6513
+    },
+    {
+      "epoch": 9.751497005988025,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1488,
+      "step": 6514
+    },
+    {
+      "epoch": 9.752994011976048,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 6515
+    },
+    {
+      "epoch": 9.754491017964071,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 6516
+    },
+    {
+      "epoch": 9.755988023952096,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1501,
+      "step": 6517
+    },
+    {
+      "epoch": 9.75748502994012,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1465,
+      "step": 6518
+    },
+    {
+      "epoch": 9.758982035928144,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 6519
+    },
+    {
+      "epoch": 9.760479041916168,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 6520
+    },
+    {
+      "epoch": 9.761976047904191,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1484,
+      "step": 6521
+    },
+    {
+      "epoch": 9.763473053892216,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.146,
+      "step": 6522
+    },
+    {
+      "epoch": 9.76497005988024,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 6523
+    },
+    {
+      "epoch": 9.766467065868264,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1504,
+      "step": 6524
+    },
+    {
+      "epoch": 9.767964071856287,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1496,
+      "step": 6525
+    },
+    {
+      "epoch": 9.769461077844312,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 6526
+    },
+    {
+      "epoch": 9.770958083832335,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 6527
+    },
+    {
+      "epoch": 9.77245508982036,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 6528
+    },
+    {
+      "epoch": 9.773952095808383,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1392,
+      "step": 6529
+    },
+    {
+      "epoch": 9.775449101796408,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 6530
+    },
+    {
+      "epoch": 9.77694610778443,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 6531
+    },
+    {
+      "epoch": 9.778443113772456,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 6532
+    },
+    {
+      "epoch": 9.779940119760479,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 6533
+    },
+    {
+      "epoch": 9.781437125748504,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1434,
+      "step": 6534
+    },
+    {
+      "epoch": 9.782934131736527,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1525,
+      "step": 6535
+    },
+    {
+      "epoch": 9.784431137724551,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1519,
+      "step": 6536
+    },
+    {
+      "epoch": 9.785928143712574,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1483,
+      "step": 6537
+    },
+    {
+      "epoch": 9.7874251497006,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1547,
+      "step": 6538
+    },
+    {
+      "epoch": 9.788922155688622,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1502,
+      "step": 6539
+    },
+    {
+      "epoch": 9.790419161676647,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1443,
+      "step": 6540
+    },
+    {
+      "epoch": 9.79191616766467,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1515,
+      "step": 6541
+    },
+    {
+      "epoch": 9.793413173652695,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1434,
+      "step": 6542
+    },
+    {
+      "epoch": 9.794910179640718,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1519,
+      "step": 6543
+    },
+    {
+      "epoch": 9.796407185628743,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1464,
+      "step": 6544
+    },
+    {
+      "epoch": 9.797904191616766,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1543,
+      "step": 6545
+    },
+    {
+      "epoch": 9.79940119760479,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1457,
+      "step": 6546
+    },
+    {
+      "epoch": 9.800898203592814,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1472,
+      "step": 6547
+    },
+    {
+      "epoch": 9.802395209580839,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 6548
+    },
+    {
+      "epoch": 9.803892215568862,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1492,
+      "step": 6549
+    },
+    {
+      "epoch": 9.805389221556887,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1488,
+      "step": 6550
+    },
+    {
+      "epoch": 9.80688622754491,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 6551
+    },
+    {
+      "epoch": 9.808383233532934,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1475,
+      "step": 6552
+    },
+    {
+      "epoch": 9.809880239520957,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 6553
+    },
+    {
+      "epoch": 9.811377245508982,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1492,
+      "step": 6554
+    },
+    {
+      "epoch": 9.812874251497005,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 6555
+    },
+    {
+      "epoch": 9.81437125748503,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1392,
+      "step": 6556
+    },
+    {
+      "epoch": 9.815868263473053,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1544,
+      "step": 6557
+    },
+    {
+      "epoch": 9.817365269461078,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 6558
+    },
+    {
+      "epoch": 9.818862275449101,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 6559
+    },
+    {
+      "epoch": 9.820359281437126,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1509,
+      "step": 6560
+    },
+    {
+      "epoch": 9.821856287425149,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1569,
+      "step": 6561
+    },
+    {
+      "epoch": 9.823353293413174,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1481,
+      "step": 6562
+    },
+    {
+      "epoch": 9.824850299401197,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.149,
+      "step": 6563
+    },
+    {
+      "epoch": 9.826347305389222,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1517,
+      "step": 6564
+    },
+    {
+      "epoch": 9.827844311377245,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 6565
+    },
+    {
+      "epoch": 9.82934131736527,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1462,
+      "step": 6566
+    },
+    {
+      "epoch": 9.830838323353294,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1443,
+      "step": 6567
+    },
+    {
+      "epoch": 9.832335329341317,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1482,
+      "step": 6568
+    },
+    {
+      "epoch": 9.83383233532934,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1503,
+      "step": 6569
+    },
+    {
+      "epoch": 9.835329341317365,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1428,
+      "step": 6570
+    },
+    {
+      "epoch": 9.83682634730539,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1446,
+      "step": 6571
+    },
+    {
+      "epoch": 9.838323353293413,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1527,
+      "step": 6572
+    },
+    {
+      "epoch": 9.839820359281438,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1519,
+      "step": 6573
+    },
+    {
+      "epoch": 9.841317365269461,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 6574
+    },
+    {
+      "epoch": 9.842814371257486,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.146,
+      "step": 6575
+    },
+    {
+      "epoch": 9.844311377245509,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 6576
+    },
+    {
+      "epoch": 9.845808383233534,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1503,
+      "step": 6577
+    },
+    {
+      "epoch": 9.847305389221557,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 6578
+    },
+    {
+      "epoch": 9.848802395209582,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.156,
+      "step": 6579
+    },
+    {
+      "epoch": 9.850299401197605,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1491,
+      "step": 6580
+    },
+    {
+      "epoch": 9.85179640718563,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.153,
+      "step": 6581
+    },
+    {
+      "epoch": 9.853293413173652,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1416,
+      "step": 6582
+    },
+    {
+      "epoch": 9.854790419161677,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.147,
+      "step": 6583
+    },
+    {
+      "epoch": 9.8562874251497,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 6584
+    },
+    {
+      "epoch": 9.857784431137725,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1558,
+      "step": 6585
+    },
+    {
+      "epoch": 9.859281437125748,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1487,
+      "step": 6586
+    },
+    {
+      "epoch": 9.860778443113773,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 6587
+    },
+    {
+      "epoch": 9.862275449101796,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1469,
+      "step": 6588
+    },
+    {
+      "epoch": 9.863772455089821,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1525,
+      "step": 6589
+    },
+    {
+      "epoch": 9.865269461077844,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1485,
+      "step": 6590
+    },
+    {
+      "epoch": 9.866766467065869,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 6591
+    },
+    {
+      "epoch": 9.868263473053892,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.146,
+      "step": 6592
+    },
+    {
+      "epoch": 9.869760479041917,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1475,
+      "step": 6593
+    },
+    {
+      "epoch": 9.87125748502994,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 6594
+    },
+    {
+      "epoch": 9.872754491017965,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1513,
+      "step": 6595
+    },
+    {
+      "epoch": 9.874251497005988,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1521,
+      "step": 6596
+    },
+    {
+      "epoch": 9.875748502994012,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1469,
+      "step": 6597
+    },
+    {
+      "epoch": 9.877245508982035,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1529,
+      "step": 6598
+    },
+    {
+      "epoch": 9.87874251497006,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1485,
+      "step": 6599
+    },
+    {
+      "epoch": 9.880239520958083,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1482,
+      "step": 6600
+    },
+    {
+      "epoch": 9.881736526946108,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 6601
+    },
+    {
+      "epoch": 9.883233532934131,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1468,
+      "step": 6602
+    },
+    {
+      "epoch": 9.884730538922156,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1513,
+      "step": 6603
+    },
+    {
+      "epoch": 9.886227544910179,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1471,
+      "step": 6604
+    },
+    {
+      "epoch": 9.887724550898204,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 6605
+    },
+    {
+      "epoch": 9.889221556886227,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1416,
+      "step": 6606
+    },
+    {
+      "epoch": 9.890718562874252,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1488,
+      "step": 6607
+    },
+    {
+      "epoch": 9.892215568862275,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1528,
+      "step": 6608
+    },
+    {
+      "epoch": 9.8937125748503,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 6609
+    },
+    {
+      "epoch": 9.895209580838323,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 6610
+    },
+    {
+      "epoch": 9.896706586826348,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 6611
+    },
+    {
+      "epoch": 9.89820359281437,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 6612
+    },
+    {
+      "epoch": 9.899700598802395,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 6613
+    },
+    {
+      "epoch": 9.901197604790418,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 6614
+    },
+    {
+      "epoch": 9.902694610778443,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1496,
+      "step": 6615
+    },
+    {
+      "epoch": 9.904191616766466,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1472,
+      "step": 6616
+    },
+    {
+      "epoch": 9.905688622754491,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 6617
+    },
+    {
+      "epoch": 9.907185628742514,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 6618
+    },
+    {
+      "epoch": 9.908682634730539,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 6619
+    },
+    {
+      "epoch": 9.910179640718562,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1501,
+      "step": 6620
+    },
+    {
+      "epoch": 9.911676646706587,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.149,
+      "step": 6621
+    },
+    {
+      "epoch": 9.91317365269461,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 6622
+    },
+    {
+      "epoch": 9.914670658682635,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 6623
+    },
+    {
+      "epoch": 9.91616766467066,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 6624
+    },
+    {
+      "epoch": 9.917664670658683,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1383,
+      "step": 6625
+    },
+    {
+      "epoch": 9.919161676646706,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1507,
+      "step": 6626
+    },
+    {
+      "epoch": 9.92065868263473,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.149,
+      "step": 6627
+    },
+    {
+      "epoch": 9.922155688622755,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 6628
+    },
+    {
+      "epoch": 9.923652694610778,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1502,
+      "step": 6629
+    },
+    {
+      "epoch": 9.925149700598803,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1522,
+      "step": 6630
+    },
+    {
+      "epoch": 9.926646706586826,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1494,
+      "step": 6631
+    },
+    {
+      "epoch": 9.928143712574851,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1499,
+      "step": 6632
+    },
+    {
+      "epoch": 9.929640718562874,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1543,
+      "step": 6633
+    },
+    {
+      "epoch": 9.931137724550899,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1469,
+      "step": 6634
+    },
+    {
+      "epoch": 9.932634730538922,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1521,
+      "step": 6635
+    },
+    {
+      "epoch": 9.934131736526947,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1494,
+      "step": 6636
+    },
+    {
+      "epoch": 9.93562874251497,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1475,
+      "step": 6637
+    },
+    {
+      "epoch": 9.937125748502995,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 6638
+    },
+    {
+      "epoch": 9.938622754491018,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 6639
+    },
+    {
+      "epoch": 9.940119760479043,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1503,
+      "step": 6640
+    },
+    {
+      "epoch": 9.941616766467066,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1446,
+      "step": 6641
+    },
+    {
+      "epoch": 9.94311377245509,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1512,
+      "step": 6642
+    },
+    {
+      "epoch": 9.944610778443113,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1475,
+      "step": 6643
+    },
+    {
+      "epoch": 9.946107784431138,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 6644
+    },
+    {
+      "epoch": 9.947604790419161,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.149,
+      "step": 6645
+    },
+    {
+      "epoch": 9.949101796407186,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1491,
+      "step": 6646
+    },
+    {
+      "epoch": 9.95059880239521,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1484,
+      "step": 6647
+    },
+    {
+      "epoch": 9.952095808383234,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1515,
+      "step": 6648
+    },
+    {
+      "epoch": 9.953592814371257,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 6649
+    },
+    {
+      "epoch": 9.955089820359282,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1392,
+      "step": 6650
+    },
+    {
+      "epoch": 9.956586826347305,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1523,
+      "step": 6651
+    },
+    {
+      "epoch": 9.95808383233533,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1506,
+      "step": 6652
+    },
+    {
+      "epoch": 9.959580838323353,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 6653
+    },
+    {
+      "epoch": 9.961077844311378,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 6654
+    },
+    {
+      "epoch": 9.9625748502994,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1505,
+      "step": 6655
+    },
+    {
+      "epoch": 9.964071856287426,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 6656
+    },
+    {
+      "epoch": 9.965568862275449,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1472,
+      "step": 6657
+    },
+    {
+      "epoch": 9.967065868263473,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1443,
+      "step": 6658
+    },
+    {
+      "epoch": 9.968562874251496,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1525,
+      "step": 6659
+    },
+    {
+      "epoch": 9.970059880239521,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1514,
+      "step": 6660
+    },
+    {
+      "epoch": 9.971556886227544,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 6661
+    },
+    {
+      "epoch": 9.97305389221557,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1457,
+      "step": 6662
+    },
+    {
+      "epoch": 9.974550898203592,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 6663
+    },
+    {
+      "epoch": 9.976047904191617,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1464,
+      "step": 6664
+    },
+    {
+      "epoch": 9.97754491017964,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1487,
+      "step": 6665
+    },
+    {
+      "epoch": 9.979041916167665,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1377,
+      "step": 6666
+    },
+    {
+      "epoch": 9.980538922155688,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1529,
+      "step": 6667
+    },
+    {
+      "epoch": 9.982035928143713,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1482,
+      "step": 6668
+    },
+    {
+      "epoch": 9.983532934131736,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 6669
+    },
+    {
+      "epoch": 9.98502994011976,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 6670
+    },
+    {
+      "epoch": 9.986526946107784,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 6671
+    },
+    {
+      "epoch": 9.988023952095809,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1491,
+      "step": 6672
+    },
+    {
+      "epoch": 9.989520958083832,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1507,
+      "step": 6673
+    },
+    {
+      "epoch": 9.991017964071856,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1489,
+      "step": 6674
+    },
+    {
+      "epoch": 9.99251497005988,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1475,
+      "step": 6675
+    },
+    {
+      "epoch": 9.994011976047904,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1514,
+      "step": 6676
+    },
+    {
+      "epoch": 9.995508982035929,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1473,
+      "step": 6677
+    },
+    {
+      "epoch": 9.997005988023952,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1484,
+      "step": 6678
+    },
+    {
+      "epoch": 9.998502994011975,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 6679
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1508,
+      "step": 6680
+    },
+    {
+      "epoch": 10.001497005988025,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1464,
+      "step": 6681
+    },
+    {
+      "epoch": 10.002994011976048,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 6682
+    },
+    {
+      "epoch": 10.004491017964073,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 6683
+    },
+    {
+      "epoch": 10.005988023952096,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 6684
+    },
+    {
+      "epoch": 10.00748502994012,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1464,
+      "step": 6685
+    },
+    {
+      "epoch": 10.008982035928144,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 6686
+    },
+    {
+      "epoch": 10.010479041916168,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1457,
+      "step": 6687
+    },
+    {
+      "epoch": 10.011976047904191,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1471,
+      "step": 6688
+    },
+    {
+      "epoch": 10.013473053892216,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.147,
+      "step": 6689
+    },
+    {
+      "epoch": 10.01497005988024,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 6690
+    },
+    {
+      "epoch": 10.016467065868264,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 6691
+    },
+    {
+      "epoch": 10.017964071856287,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1505,
+      "step": 6692
+    },
+    {
+      "epoch": 10.019461077844312,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1372,
+      "step": 6693
+    },
+    {
+      "epoch": 10.020958083832335,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1507,
+      "step": 6694
+    },
+    {
+      "epoch": 10.02245508982036,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1481,
+      "step": 6695
+    },
+    {
+      "epoch": 10.023952095808383,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 6696
+    },
+    {
+      "epoch": 10.025449101796408,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 6697
+    },
+    {
+      "epoch": 10.02694610778443,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1416,
+      "step": 6698
+    },
+    {
+      "epoch": 10.028443113772456,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 6699
+    },
+    {
+      "epoch": 10.029940119760479,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 6700
+    },
+    {
+      "epoch": 10.031437125748504,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1527,
+      "step": 6701
+    },
+    {
+      "epoch": 10.032934131736527,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 6702
+    },
+    {
+      "epoch": 10.034431137724551,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1462,
+      "step": 6703
+    },
+    {
+      "epoch": 10.035928143712574,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1518,
+      "step": 6704
+    },
+    {
+      "epoch": 10.0374251497006,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 6705
+    },
+    {
+      "epoch": 10.038922155688622,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 6706
+    },
+    {
+      "epoch": 10.040419161676647,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1462,
+      "step": 6707
+    },
+    {
+      "epoch": 10.04191616766467,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1503,
+      "step": 6708
+    },
+    {
+      "epoch": 10.043413173652695,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1481,
+      "step": 6709
+    },
+    {
+      "epoch": 10.044910179640718,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 6710
+    },
+    {
+      "epoch": 10.046407185628743,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 6711
+    },
+    {
+      "epoch": 10.047904191616766,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1575,
+      "step": 6712
+    },
+    {
+      "epoch": 10.04940119760479,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1482,
+      "step": 6713
+    },
+    {
+      "epoch": 10.050898203592814,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 6714
+    },
+    {
+      "epoch": 10.052395209580839,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1472,
+      "step": 6715
+    },
+    {
+      "epoch": 10.053892215568862,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1563,
+      "step": 6716
+    },
+    {
+      "epoch": 10.055389221556887,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1493,
+      "step": 6717
+    },
+    {
+      "epoch": 10.05688622754491,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 6718
+    },
+    {
+      "epoch": 10.058383233532934,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 6719
+    },
+    {
+      "epoch": 10.059880239520957,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 6720
+    },
+    {
+      "epoch": 10.061377245508982,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1479,
+      "step": 6721
+    },
+    {
+      "epoch": 10.062874251497005,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 6722
+    },
+    {
+      "epoch": 10.06437125748503,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 6723
+    },
+    {
+      "epoch": 10.065868263473053,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 6724
+    },
+    {
+      "epoch": 10.067365269461078,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 6725
+    },
+    {
+      "epoch": 10.068862275449101,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 6726
+    },
+    {
+      "epoch": 10.070359281437126,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 6727
+    },
+    {
+      "epoch": 10.071856287425149,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1511,
+      "step": 6728
+    },
+    {
+      "epoch": 10.073353293413174,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1511,
+      "step": 6729
+    },
+    {
+      "epoch": 10.074850299401197,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.149,
+      "step": 6730
+    },
+    {
+      "epoch": 10.076347305389222,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 6731
+    },
+    {
+      "epoch": 10.077844311377245,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1584,
+      "step": 6732
+    },
+    {
+      "epoch": 10.07934131736527,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1383,
+      "step": 6733
+    },
+    {
+      "epoch": 10.080838323353293,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1512,
+      "step": 6734
+    },
+    {
+      "epoch": 10.082335329341317,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.146,
+      "step": 6735
+    },
+    {
+      "epoch": 10.08383233532934,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1554,
+      "step": 6736
+    },
+    {
+      "epoch": 10.085329341317365,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1446,
+      "step": 6737
+    },
+    {
+      "epoch": 10.08682634730539,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1475,
+      "step": 6738
+    },
+    {
+      "epoch": 10.088323353293413,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1484,
+      "step": 6739
+    },
+    {
+      "epoch": 10.089820359281438,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1487,
+      "step": 6740
+    },
+    {
+      "epoch": 10.091317365269461,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1488,
+      "step": 6741
+    },
+    {
+      "epoch": 10.092814371257486,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1368,
+      "step": 6742
+    },
+    {
+      "epoch": 10.094311377245509,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 6743
+    },
+    {
+      "epoch": 10.095808383233534,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1476,
+      "step": 6744
+    },
+    {
+      "epoch": 10.097305389221557,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 6745
+    },
+    {
+      "epoch": 10.098802395209582,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1381,
+      "step": 6746
+    },
+    {
+      "epoch": 10.100299401197605,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 6747
+    },
+    {
+      "epoch": 10.10179640718563,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1513,
+      "step": 6748
+    },
+    {
+      "epoch": 10.103293413173652,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1393,
+      "step": 6749
+    },
+    {
+      "epoch": 10.104790419161677,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1485,
+      "step": 6750
+    },
+    {
+      "epoch": 10.1062874251497,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1517,
+      "step": 6751
+    },
+    {
+      "epoch": 10.107784431137725,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1514,
+      "step": 6752
+    },
+    {
+      "epoch": 10.109281437125748,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1469,
+      "step": 6753
+    },
+    {
+      "epoch": 10.110778443113773,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1511,
+      "step": 6754
+    },
+    {
+      "epoch": 10.112275449101796,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 6755
+    },
+    {
+      "epoch": 10.113772455089821,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 6756
+    },
+    {
+      "epoch": 10.115269461077844,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 6757
+    },
+    {
+      "epoch": 10.116766467065869,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 6758
+    },
+    {
+      "epoch": 10.118263473053892,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 6759
+    },
+    {
+      "epoch": 10.119760479041917,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 6760
+    },
+    {
+      "epoch": 10.12125748502994,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.149,
+      "step": 6761
+    },
+    {
+      "epoch": 10.122754491017965,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1496,
+      "step": 6762
+    },
+    {
+      "epoch": 10.124251497005988,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1459,
+      "step": 6763
+    },
+    {
+      "epoch": 10.125748502994012,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1473,
+      "step": 6764
+    },
+    {
+      "epoch": 10.127245508982035,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 6765
+    },
+    {
+      "epoch": 10.12874251497006,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 6766
+    },
+    {
+      "epoch": 10.130239520958083,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 6767
+    },
+    {
+      "epoch": 10.131736526946108,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1473,
+      "step": 6768
+    },
+    {
+      "epoch": 10.133233532934131,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 6769
+    },
+    {
+      "epoch": 10.134730538922156,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 6770
+    },
+    {
+      "epoch": 10.136227544910179,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1524,
+      "step": 6771
+    },
+    {
+      "epoch": 10.137724550898204,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 6772
+    },
+    {
+      "epoch": 10.139221556886227,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 6773
+    },
+    {
+      "epoch": 10.140718562874252,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1531,
+      "step": 6774
+    },
+    {
+      "epoch": 10.142215568862275,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1484,
+      "step": 6775
+    },
+    {
+      "epoch": 10.1437125748503,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1494,
+      "step": 6776
+    },
+    {
+      "epoch": 10.145209580838323,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 6777
+    },
+    {
+      "epoch": 10.146706586826348,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 6778
+    },
+    {
+      "epoch": 10.14820359281437,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 6779
+    },
+    {
+      "epoch": 10.149700598802395,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1524,
+      "step": 6780
+    },
+    {
+      "epoch": 10.151197604790418,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1487,
+      "step": 6781
+    },
+    {
+      "epoch": 10.152694610778443,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 6782
+    },
+    {
+      "epoch": 10.154191616766466,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.153,
+      "step": 6783
+    },
+    {
+      "epoch": 10.155688622754491,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1483,
+      "step": 6784
+    },
+    {
+      "epoch": 10.157185628742514,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 6785
+    },
+    {
+      "epoch": 10.158682634730539,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1497,
+      "step": 6786
+    },
+    {
+      "epoch": 10.160179640718562,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1495,
+      "step": 6787
+    },
+    {
+      "epoch": 10.161676646706587,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.147,
+      "step": 6788
+    },
+    {
+      "epoch": 10.16317365269461,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1488,
+      "step": 6789
+    },
+    {
+      "epoch": 10.164670658682635,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1476,
+      "step": 6790
+    },
+    {
+      "epoch": 10.16616766467066,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1499,
+      "step": 6791
+    },
+    {
+      "epoch": 10.167664670658683,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.149,
+      "step": 6792
+    },
+    {
+      "epoch": 10.169161676646707,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 6793
+    },
+    {
+      "epoch": 10.17065868263473,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 6794
+    },
+    {
+      "epoch": 10.172155688622755,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0008,
+      "loss": 1.146,
+      "step": 6795
+    },
+    {
+      "epoch": 10.173652694610778,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 6796
+    },
+    {
+      "epoch": 10.175149700598803,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 6797
+    },
+    {
+      "epoch": 10.176646706586826,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 6798
+    },
+    {
+      "epoch": 10.178143712574851,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1486,
+      "step": 6799
+    },
+    {
+      "epoch": 10.179640718562874,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 6800
+    },
+    {
+      "epoch": 10.181137724550899,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1472,
+      "step": 6801
+    },
+    {
+      "epoch": 10.182634730538922,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 6802
+    },
+    {
+      "epoch": 10.184131736526947,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.154,
+      "step": 6803
+    },
+    {
+      "epoch": 10.18562874251497,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 6804
+    },
+    {
+      "epoch": 10.187125748502995,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 6805
+    },
+    {
+      "epoch": 10.188622754491018,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 6806
+    },
+    {
+      "epoch": 10.190119760479043,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1416,
+      "step": 6807
+    },
+    {
+      "epoch": 10.191616766467066,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 6808
+    },
+    {
+      "epoch": 10.19311377245509,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1534,
+      "step": 6809
+    },
+    {
+      "epoch": 10.194610778443113,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1484,
+      "step": 6810
+    },
+    {
+      "epoch": 10.196107784431138,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1499,
+      "step": 6811
+    },
+    {
+      "epoch": 10.197604790419161,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1454,
+      "step": 6812
+    },
+    {
+      "epoch": 10.199101796407186,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1462,
+      "step": 6813
+    },
+    {
+      "epoch": 10.20059880239521,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1482,
+      "step": 6814
+    },
+    {
+      "epoch": 10.202095808383234,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 6815
+    },
+    {
+      "epoch": 10.203592814371257,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.151,
+      "step": 6816
+    },
+    {
+      "epoch": 10.205089820359282,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 6817
+    },
+    {
+      "epoch": 10.206586826347305,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1454,
+      "step": 6818
+    },
+    {
+      "epoch": 10.20808383233533,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.147,
+      "step": 6819
+    },
+    {
+      "epoch": 10.209580838323353,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1428,
+      "step": 6820
+    },
+    {
+      "epoch": 10.211077844311378,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 6821
+    },
+    {
+      "epoch": 10.2125748502994,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1473,
+      "step": 6822
+    },
+    {
+      "epoch": 10.214071856287426,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 6823
+    },
+    {
+      "epoch": 10.215568862275449,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 6824
+    },
+    {
+      "epoch": 10.217065868263473,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1459,
+      "step": 6825
+    },
+    {
+      "epoch": 10.218562874251496,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1499,
+      "step": 6826
+    },
+    {
+      "epoch": 10.220059880239521,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.151,
+      "step": 6827
+    },
+    {
+      "epoch": 10.221556886227544,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 6828
+    },
+    {
+      "epoch": 10.22305389221557,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1528,
+      "step": 6829
+    },
+    {
+      "epoch": 10.224550898203592,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1364,
+      "step": 6830
+    },
+    {
+      "epoch": 10.226047904191617,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1446,
+      "step": 6831
+    },
+    {
+      "epoch": 10.22754491017964,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 6832
+    },
+    {
+      "epoch": 10.229041916167665,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1454,
+      "step": 6833
+    },
+    {
+      "epoch": 10.230538922155688,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 6834
+    },
+    {
+      "epoch": 10.232035928143713,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1486,
+      "step": 6835
+    },
+    {
+      "epoch": 10.233532934131736,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 6836
+    },
+    {
+      "epoch": 10.23502994011976,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 6837
+    },
+    {
+      "epoch": 10.236526946107784,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 6838
+    },
+    {
+      "epoch": 10.238023952095809,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1494,
+      "step": 6839
+    },
+    {
+      "epoch": 10.239520958083832,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1535,
+      "step": 6840
+    },
+    {
+      "epoch": 10.241017964071856,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 6841
+    },
+    {
+      "epoch": 10.24251497005988,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 6842
+    },
+    {
+      "epoch": 10.244011976047904,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 6843
+    },
+    {
+      "epoch": 10.245508982035927,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1483,
+      "step": 6844
+    },
+    {
+      "epoch": 10.247005988023952,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 6845
+    },
+    {
+      "epoch": 10.248502994011975,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1482,
+      "step": 6846
+    },
+    {
+      "epoch": 10.25,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1472,
+      "step": 6847
+    },
+    {
+      "epoch": 10.251497005988025,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 6848
+    },
+    {
+      "epoch": 10.252994011976048,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.15,
+      "step": 6849
+    },
+    {
+      "epoch": 10.254491017964073,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 6850
+    },
+    {
+      "epoch": 10.255988023952096,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 6851
+    },
+    {
+      "epoch": 10.25748502994012,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.152,
+      "step": 6852
+    },
+    {
+      "epoch": 10.258982035928144,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1482,
+      "step": 6853
+    },
+    {
+      "epoch": 10.260479041916168,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 6854
+    },
+    {
+      "epoch": 10.261976047904191,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 6855
+    },
+    {
+      "epoch": 10.263473053892216,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 6856
+    },
+    {
+      "epoch": 10.26497005988024,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 6857
+    },
+    {
+      "epoch": 10.266467065868264,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1515,
+      "step": 6858
+    },
+    {
+      "epoch": 10.267964071856287,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 6859
+    },
+    {
+      "epoch": 10.269461077844312,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1546,
+      "step": 6860
+    },
+    {
+      "epoch": 10.270958083832335,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 6861
+    },
+    {
+      "epoch": 10.27245508982036,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 6862
+    },
+    {
+      "epoch": 10.273952095808383,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 6863
+    },
+    {
+      "epoch": 10.275449101796408,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1472,
+      "step": 6864
+    },
+    {
+      "epoch": 10.27694610778443,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1502,
+      "step": 6865
+    },
+    {
+      "epoch": 10.278443113772456,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.152,
+      "step": 6866
+    },
+    {
+      "epoch": 10.279940119760479,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 6867
+    },
+    {
+      "epoch": 10.281437125748504,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1454,
+      "step": 6868
+    },
+    {
+      "epoch": 10.282934131736527,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 6869
+    },
+    {
+      "epoch": 10.284431137724551,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1491,
+      "step": 6870
+    },
+    {
+      "epoch": 10.285928143712574,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1513,
+      "step": 6871
+    },
+    {
+      "epoch": 10.2874251497006,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 6872
+    },
+    {
+      "epoch": 10.288922155688622,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.14,
+      "step": 6873
+    },
+    {
+      "epoch": 10.290419161676647,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 6874
+    },
+    {
+      "epoch": 10.29191616766467,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1505,
+      "step": 6875
+    },
+    {
+      "epoch": 10.293413173652695,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1518,
+      "step": 6876
+    },
+    {
+      "epoch": 10.294910179640718,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1509,
+      "step": 6877
+    },
+    {
+      "epoch": 10.296407185628743,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 6878
+    },
+    {
+      "epoch": 10.297904191616766,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 6879
+    },
+    {
+      "epoch": 10.29940119760479,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1519,
+      "step": 6880
+    },
+    {
+      "epoch": 10.300898203592814,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1508,
+      "step": 6881
+    },
+    {
+      "epoch": 10.302395209580839,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1531,
+      "step": 6882
+    },
+    {
+      "epoch": 10.303892215568862,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1499,
+      "step": 6883
+    },
+    {
+      "epoch": 10.305389221556887,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 6884
+    },
+    {
+      "epoch": 10.30688622754491,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1386,
+      "step": 6885
+    },
+    {
+      "epoch": 10.308383233532934,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 6886
+    },
+    {
+      "epoch": 10.309880239520957,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1515,
+      "step": 6887
+    },
+    {
+      "epoch": 10.311377245508982,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 6888
+    },
+    {
+      "epoch": 10.312874251497005,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 6889
+    },
+    {
+      "epoch": 10.31437125748503,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 6890
+    },
+    {
+      "epoch": 10.315868263473053,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1459,
+      "step": 6891
+    },
+    {
+      "epoch": 10.317365269461078,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 6892
+    },
+    {
+      "epoch": 10.318862275449101,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1465,
+      "step": 6893
+    },
+    {
+      "epoch": 10.320359281437126,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1536,
+      "step": 6894
+    },
+    {
+      "epoch": 10.321856287425149,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1463,
+      "step": 6895
+    },
+    {
+      "epoch": 10.323353293413174,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 6896
+    },
+    {
+      "epoch": 10.324850299401197,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 6897
+    },
+    {
+      "epoch": 10.326347305389222,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 6898
+    },
+    {
+      "epoch": 10.327844311377245,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 6899
+    },
+    {
+      "epoch": 10.32934131736527,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1488,
+      "step": 6900
+    },
+    {
+      "epoch": 10.330838323353294,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1513,
+      "step": 6901
+    },
+    {
+      "epoch": 10.332335329341317,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 6902
+    },
+    {
+      "epoch": 10.33383233532934,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.152,
+      "step": 6903
+    },
+    {
+      "epoch": 10.335329341317365,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 6904
+    },
+    {
+      "epoch": 10.33682634730539,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0008,
+      "loss": 1.1491,
+      "step": 6905
+    },
+    {
+      "epoch": 10.338323353293413,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1463,
+      "step": 6906
+    },
+    {
+      "epoch": 10.339820359281438,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1457,
+      "step": 6907
+    },
+    {
+      "epoch": 10.341317365269461,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 6908
+    },
+    {
+      "epoch": 10.342814371257486,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1475,
+      "step": 6909
+    },
+    {
+      "epoch": 10.344311377245509,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1545,
+      "step": 6910
+    },
+    {
+      "epoch": 10.345808383233534,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 6911
+    },
+    {
+      "epoch": 10.347305389221557,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 6912
+    },
+    {
+      "epoch": 10.348802395209582,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1457,
+      "step": 6913
+    },
+    {
+      "epoch": 10.350299401197605,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 6914
+    },
+    {
+      "epoch": 10.35179640718563,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1493,
+      "step": 6915
+    },
+    {
+      "epoch": 10.353293413173652,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1494,
+      "step": 6916
+    },
+    {
+      "epoch": 10.354790419161677,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1454,
+      "step": 6917
+    },
+    {
+      "epoch": 10.3562874251497,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 6918
+    },
+    {
+      "epoch": 10.357784431137725,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1491,
+      "step": 6919
+    },
+    {
+      "epoch": 10.359281437125748,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1505,
+      "step": 6920
+    },
+    {
+      "epoch": 10.360778443113773,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1493,
+      "step": 6921
+    },
+    {
+      "epoch": 10.362275449101796,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 6922
+    },
+    {
+      "epoch": 10.363772455089821,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1476,
+      "step": 6923
+    },
+    {
+      "epoch": 10.365269461077844,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 6924
+    },
+    {
+      "epoch": 10.366766467065869,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 6925
+    },
+    {
+      "epoch": 10.368263473053892,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.14,
+      "step": 6926
+    },
+    {
+      "epoch": 10.369760479041917,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 6927
+    },
+    {
+      "epoch": 10.37125748502994,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 6928
+    },
+    {
+      "epoch": 10.372754491017965,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 6929
+    },
+    {
+      "epoch": 10.374251497005988,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 6930
+    },
+    {
+      "epoch": 10.375748502994012,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1356,
+      "step": 6931
+    },
+    {
+      "epoch": 10.377245508982035,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1518,
+      "step": 6932
+    },
+    {
+      "epoch": 10.37874251497006,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.151,
+      "step": 6933
+    },
+    {
+      "epoch": 10.380239520958083,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1443,
+      "step": 6934
+    },
+    {
+      "epoch": 10.381736526946108,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1515,
+      "step": 6935
+    },
+    {
+      "epoch": 10.383233532934131,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1513,
+      "step": 6936
+    },
+    {
+      "epoch": 10.384730538922156,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.147,
+      "step": 6937
+    },
+    {
+      "epoch": 10.386227544910179,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1477,
+      "step": 6938
+    },
+    {
+      "epoch": 10.387724550898204,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.151,
+      "step": 6939
+    },
+    {
+      "epoch": 10.389221556886227,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1454,
+      "step": 6940
+    },
+    {
+      "epoch": 10.390718562874252,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1413,
+      "step": 6941
+    },
+    {
+      "epoch": 10.392215568862275,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 6942
+    },
+    {
+      "epoch": 10.3937125748503,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 6943
+    },
+    {
+      "epoch": 10.395209580838323,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1551,
+      "step": 6944
+    },
+    {
+      "epoch": 10.396706586826348,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1546,
+      "step": 6945
+    },
+    {
+      "epoch": 10.39820359281437,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1487,
+      "step": 6946
+    },
+    {
+      "epoch": 10.399700598802395,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1475,
+      "step": 6947
+    },
+    {
+      "epoch": 10.401197604790418,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 6948
+    },
+    {
+      "epoch": 10.402694610778443,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1484,
+      "step": 6949
+    },
+    {
+      "epoch": 10.404191616766466,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1473,
+      "step": 6950
+    },
+    {
+      "epoch": 10.405688622754491,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1428,
+      "step": 6951
+    },
+    {
+      "epoch": 10.407185628742514,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1508,
+      "step": 6952
+    },
+    {
+      "epoch": 10.408682634730539,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1507,
+      "step": 6953
+    },
+    {
+      "epoch": 10.410179640718562,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.155,
+      "step": 6954
+    },
+    {
+      "epoch": 10.411676646706587,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 6955
+    },
+    {
+      "epoch": 10.41317365269461,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 6956
+    },
+    {
+      "epoch": 10.414670658682635,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 6957
+    },
+    {
+      "epoch": 10.41616766467066,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 6958
+    },
+    {
+      "epoch": 10.417664670658683,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 6959
+    },
+    {
+      "epoch": 10.419161676646706,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 6960
+    },
+    {
+      "epoch": 10.42065868263473,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.151,
+      "step": 6961
+    },
+    {
+      "epoch": 10.422155688622755,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 6962
+    },
+    {
+      "epoch": 10.423652694610778,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 6963
+    },
+    {
+      "epoch": 10.425149700598803,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1393,
+      "step": 6964
+    },
+    {
+      "epoch": 10.426646706586826,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 6965
+    },
+    {
+      "epoch": 10.428143712574851,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1515,
+      "step": 6966
+    },
+    {
+      "epoch": 10.429640718562874,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 6967
+    },
+    {
+      "epoch": 10.431137724550899,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1477,
+      "step": 6968
+    },
+    {
+      "epoch": 10.432634730538922,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 6969
+    },
+    {
+      "epoch": 10.434131736526947,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 6970
+    },
+    {
+      "epoch": 10.43562874251497,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1387,
+      "step": 6971
+    },
+    {
+      "epoch": 10.437125748502995,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 6972
+    },
+    {
+      "epoch": 10.438622754491018,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 6973
+    },
+    {
+      "epoch": 10.440119760479043,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 6974
+    },
+    {
+      "epoch": 10.441616766467066,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1526,
+      "step": 6975
+    },
+    {
+      "epoch": 10.44311377245509,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 6976
+    },
+    {
+      "epoch": 10.444610778443113,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 6977
+    },
+    {
+      "epoch": 10.446107784431138,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 6978
+    },
+    {
+      "epoch": 10.447604790419161,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 6979
+    },
+    {
+      "epoch": 10.449101796407186,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 6980
+    },
+    {
+      "epoch": 10.45059880239521,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 6981
+    },
+    {
+      "epoch": 10.452095808383234,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 6982
+    },
+    {
+      "epoch": 10.453592814371257,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1494,
+      "step": 6983
+    },
+    {
+      "epoch": 10.455089820359282,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1485,
+      "step": 6984
+    },
+    {
+      "epoch": 10.456586826347305,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 6985
+    },
+    {
+      "epoch": 10.45808383233533,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 6986
+    },
+    {
+      "epoch": 10.459580838323353,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 6987
+    },
+    {
+      "epoch": 10.461077844311378,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1428,
+      "step": 6988
+    },
+    {
+      "epoch": 10.4625748502994,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1473,
+      "step": 6989
+    },
+    {
+      "epoch": 10.464071856287426,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1501,
+      "step": 6990
+    },
+    {
+      "epoch": 10.465568862275449,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0008,
+      "loss": 1.1496,
+      "step": 6991
+    },
+    {
+      "epoch": 10.467065868263473,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1479,
+      "step": 6992
+    },
+    {
+      "epoch": 10.468562874251496,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 6993
+    },
+    {
+      "epoch": 10.470059880239521,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 6994
+    },
+    {
+      "epoch": 10.471556886227544,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1572,
+      "step": 6995
+    },
+    {
+      "epoch": 10.47305389221557,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 6996
+    },
+    {
+      "epoch": 10.474550898203592,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 6997
+    },
+    {
+      "epoch": 10.476047904191617,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 6998
+    },
+    {
+      "epoch": 10.47754491017964,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 6999
+    },
+    {
+      "epoch": 10.479041916167665,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1375,
+      "step": 7000
+    },
+    {
+      "epoch": 10.480538922155688,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1498,
+      "step": 7001
+    },
+    {
+      "epoch": 10.482035928143713,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.14,
+      "step": 7002
+    },
+    {
+      "epoch": 10.483532934131736,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 7003
+    },
+    {
+      "epoch": 10.48502994011976,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1471,
+      "step": 7004
+    },
+    {
+      "epoch": 10.486526946107784,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1468,
+      "step": 7005
+    },
+    {
+      "epoch": 10.488023952095809,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1512,
+      "step": 7006
+    },
+    {
+      "epoch": 10.489520958083832,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 7007
+    },
+    {
+      "epoch": 10.491017964071856,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1484,
+      "step": 7008
+    },
+    {
+      "epoch": 10.49251497005988,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.149,
+      "step": 7009
+    },
+    {
+      "epoch": 10.494011976047904,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.14,
+      "step": 7010
+    },
+    {
+      "epoch": 10.495508982035927,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 7011
+    },
+    {
+      "epoch": 10.497005988023952,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 7012
+    },
+    {
+      "epoch": 10.498502994011975,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 7013
+    },
+    {
+      "epoch": 10.5,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 7014
+    },
+    {
+      "epoch": 10.501497005988025,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 7015
+    },
+    {
+      "epoch": 10.502994011976048,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.148,
+      "step": 7016
+    },
+    {
+      "epoch": 10.504491017964071,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 7017
+    },
+    {
+      "epoch": 10.505988023952096,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 7018
+    },
+    {
+      "epoch": 10.50748502994012,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 7019
+    },
+    {
+      "epoch": 10.508982035928144,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 7020
+    },
+    {
+      "epoch": 10.510479041916168,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 7021
+    },
+    {
+      "epoch": 10.511976047904191,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1504,
+      "step": 7022
+    },
+    {
+      "epoch": 10.513473053892216,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 7023
+    },
+    {
+      "epoch": 10.51497005988024,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1506,
+      "step": 7024
+    },
+    {
+      "epoch": 10.516467065868264,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1518,
+      "step": 7025
+    },
+    {
+      "epoch": 10.517964071856287,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 7026
+    },
+    {
+      "epoch": 10.519461077844312,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1413,
+      "step": 7027
+    },
+    {
+      "epoch": 10.520958083832335,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1526,
+      "step": 7028
+    },
+    {
+      "epoch": 10.52245508982036,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 7029
+    },
+    {
+      "epoch": 10.523952095808383,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1495,
+      "step": 7030
+    },
+    {
+      "epoch": 10.525449101796408,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 7031
+    },
+    {
+      "epoch": 10.52694610778443,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1477,
+      "step": 7032
+    },
+    {
+      "epoch": 10.528443113772456,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1506,
+      "step": 7033
+    },
+    {
+      "epoch": 10.529940119760479,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1457,
+      "step": 7034
+    },
+    {
+      "epoch": 10.531437125748504,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0008,
+      "loss": 1.153,
+      "step": 7035
+    },
+    {
+      "epoch": 10.532934131736527,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 7036
+    },
+    {
+      "epoch": 10.534431137724551,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 7037
+    },
+    {
+      "epoch": 10.535928143712574,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1429,
+      "step": 7038
+    },
+    {
+      "epoch": 10.5374251497006,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 7039
+    },
+    {
+      "epoch": 10.538922155688622,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1499,
+      "step": 7040
+    },
+    {
+      "epoch": 10.540419161676647,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1463,
+      "step": 7041
+    },
+    {
+      "epoch": 10.54191616766467,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 7042
+    },
+    {
+      "epoch": 10.543413173652695,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1499,
+      "step": 7043
+    },
+    {
+      "epoch": 10.544910179640718,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 7044
+    },
+    {
+      "epoch": 10.546407185628743,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 7045
+    },
+    {
+      "epoch": 10.547904191616766,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 7046
+    },
+    {
+      "epoch": 10.54940119760479,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1376,
+      "step": 7047
+    },
+    {
+      "epoch": 10.550898203592814,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 7048
+    },
+    {
+      "epoch": 10.552395209580839,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1392,
+      "step": 7049
+    },
+    {
+      "epoch": 10.553892215568862,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 7050
+    },
+    {
+      "epoch": 10.555389221556887,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 7051
+    },
+    {
+      "epoch": 10.55688622754491,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 7052
+    },
+    {
+      "epoch": 10.558383233532934,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 7053
+    },
+    {
+      "epoch": 10.559880239520957,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 7054
+    },
+    {
+      "epoch": 10.561377245508982,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1498,
+      "step": 7055
+    },
+    {
+      "epoch": 10.562874251497005,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 7056
+    },
+    {
+      "epoch": 10.56437125748503,
+      "grad_norm": 0.240234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1476,
+      "step": 7057
+    },
+    {
+      "epoch": 10.565868263473053,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 7058
+    },
+    {
+      "epoch": 10.567365269461078,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1484,
+      "step": 7059
+    },
+    {
+      "epoch": 10.568862275449101,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0008,
+      "loss": 1.15,
+      "step": 7060
+    },
+    {
+      "epoch": 10.570359281437126,
+      "grad_norm": 0.22265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1556,
+      "step": 7061
+    },
+    {
+      "epoch": 10.571856287425149,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1548,
+      "step": 7062
+    },
+    {
+      "epoch": 10.573353293413174,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 7063
+    },
+    {
+      "epoch": 10.574850299401197,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0008,
+      "loss": 1.1503,
+      "step": 7064
+    },
+    {
+      "epoch": 10.576347305389222,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1491,
+      "step": 7065
+    },
+    {
+      "epoch": 10.577844311377245,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1516,
+      "step": 7066
+    },
+    {
+      "epoch": 10.57934131736527,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1539,
+      "step": 7067
+    },
+    {
+      "epoch": 10.580838323353294,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1531,
+      "step": 7068
+    },
+    {
+      "epoch": 10.582335329341317,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1469,
+      "step": 7069
+    },
+    {
+      "epoch": 10.58383233532934,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1496,
+      "step": 7070
+    },
+    {
+      "epoch": 10.585329341317365,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1434,
+      "step": 7071
+    },
+    {
+      "epoch": 10.58682634730539,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1503,
+      "step": 7072
+    },
+    {
+      "epoch": 10.588323353293413,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 7073
+    },
+    {
+      "epoch": 10.589820359281438,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 7074
+    },
+    {
+      "epoch": 10.591317365269461,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 7075
+    },
+    {
+      "epoch": 10.592814371257486,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.15,
+      "step": 7076
+    },
+    {
+      "epoch": 10.594311377245509,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1413,
+      "step": 7077
+    },
+    {
+      "epoch": 10.595808383233534,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.147,
+      "step": 7078
+    },
+    {
+      "epoch": 10.597305389221557,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1495,
+      "step": 7079
+    },
+    {
+      "epoch": 10.598802395209582,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1446,
+      "step": 7080
+    },
+    {
+      "epoch": 10.600299401197605,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1443,
+      "step": 7081
+    },
+    {
+      "epoch": 10.60179640718563,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.14,
+      "step": 7082
+    },
+    {
+      "epoch": 10.603293413173652,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 7083
+    },
+    {
+      "epoch": 10.604790419161677,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1508,
+      "step": 7084
+    },
+    {
+      "epoch": 10.6062874251497,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1384,
+      "step": 7085
+    },
+    {
+      "epoch": 10.607784431137725,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1486,
+      "step": 7086
+    },
+    {
+      "epoch": 10.609281437125748,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1471,
+      "step": 7087
+    },
+    {
+      "epoch": 10.610778443113773,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 7088
+    },
+    {
+      "epoch": 10.612275449101796,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 7089
+    },
+    {
+      "epoch": 10.613772455089821,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 7090
+    },
+    {
+      "epoch": 10.615269461077844,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1482,
+      "step": 7091
+    },
+    {
+      "epoch": 10.616766467065869,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1536,
+      "step": 7092
+    },
+    {
+      "epoch": 10.618263473053892,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 7093
+    },
+    {
+      "epoch": 10.619760479041917,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 7094
+    },
+    {
+      "epoch": 10.62125748502994,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 7095
+    },
+    {
+      "epoch": 10.622754491017965,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1537,
+      "step": 7096
+    },
+    {
+      "epoch": 10.624251497005988,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1526,
+      "step": 7097
+    },
+    {
+      "epoch": 10.625748502994012,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1506,
+      "step": 7098
+    },
+    {
+      "epoch": 10.627245508982035,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 7099
+    },
+    {
+      "epoch": 10.62874251497006,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1501,
+      "step": 7100
+    },
+    {
+      "epoch": 10.630239520958083,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 7101
+    },
+    {
+      "epoch": 10.631736526946108,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 7102
+    },
+    {
+      "epoch": 10.633233532934131,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 7103
+    },
+    {
+      "epoch": 10.634730538922156,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1468,
+      "step": 7104
+    },
+    {
+      "epoch": 10.636227544910179,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 7105
+    },
+    {
+      "epoch": 10.637724550898204,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 7106
+    },
+    {
+      "epoch": 10.639221556886227,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1471,
+      "step": 7107
+    },
+    {
+      "epoch": 10.640718562874252,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1463,
+      "step": 7108
+    },
+    {
+      "epoch": 10.642215568862275,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1492,
+      "step": 7109
+    },
+    {
+      "epoch": 10.6437125748503,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1489,
+      "step": 7110
+    },
+    {
+      "epoch": 10.645209580838323,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 7111
+    },
+    {
+      "epoch": 10.646706586826348,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 7112
+    },
+    {
+      "epoch": 10.64820359281437,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 7113
+    },
+    {
+      "epoch": 10.649700598802395,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 7114
+    },
+    {
+      "epoch": 10.651197604790418,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1484,
+      "step": 7115
+    },
+    {
+      "epoch": 10.652694610778443,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1462,
+      "step": 7116
+    },
+    {
+      "epoch": 10.654191616766466,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1533,
+      "step": 7117
+    },
+    {
+      "epoch": 10.655688622754491,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1477,
+      "step": 7118
+    },
+    {
+      "epoch": 10.657185628742514,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 7119
+    },
+    {
+      "epoch": 10.658682634730539,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1491,
+      "step": 7120
+    },
+    {
+      "epoch": 10.660179640718562,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 7121
+    },
+    {
+      "epoch": 10.661676646706587,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1535,
+      "step": 7122
+    },
+    {
+      "epoch": 10.66317365269461,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.148,
+      "step": 7123
+    },
+    {
+      "epoch": 10.664670658682635,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 7124
+    },
+    {
+      "epoch": 10.66616766467066,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1454,
+      "step": 7125
+    },
+    {
+      "epoch": 10.667664670658683,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 7126
+    },
+    {
+      "epoch": 10.669161676646706,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1493,
+      "step": 7127
+    },
+    {
+      "epoch": 10.67065868263473,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 7128
+    },
+    {
+      "epoch": 10.672155688622755,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1357,
+      "step": 7129
+    },
+    {
+      "epoch": 10.673652694610778,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 7130
+    },
+    {
+      "epoch": 10.675149700598803,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 7131
+    },
+    {
+      "epoch": 10.676646706586826,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1428,
+      "step": 7132
+    },
+    {
+      "epoch": 10.678143712574851,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 7133
+    },
+    {
+      "epoch": 10.679640718562874,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1471,
+      "step": 7134
+    },
+    {
+      "epoch": 10.681137724550899,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 7135
+    },
+    {
+      "epoch": 10.682634730538922,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1475,
+      "step": 7136
+    },
+    {
+      "epoch": 10.684131736526947,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1434,
+      "step": 7137
+    },
+    {
+      "epoch": 10.68562874251497,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 7138
+    },
+    {
+      "epoch": 10.687125748502995,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1503,
+      "step": 7139
+    },
+    {
+      "epoch": 10.688622754491018,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 7140
+    },
+    {
+      "epoch": 10.690119760479043,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 7141
+    },
+    {
+      "epoch": 10.691616766467066,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 7142
+    },
+    {
+      "epoch": 10.69311377245509,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.149,
+      "step": 7143
+    },
+    {
+      "epoch": 10.694610778443113,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 7144
+    },
+    {
+      "epoch": 10.696107784431138,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 7145
+    },
+    {
+      "epoch": 10.697604790419161,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 7146
+    },
+    {
+      "epoch": 10.699101796407186,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1387,
+      "step": 7147
+    },
+    {
+      "epoch": 10.70059880239521,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1379,
+      "step": 7148
+    },
+    {
+      "epoch": 10.702095808383234,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 7149
+    },
+    {
+      "epoch": 10.703592814371257,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 7150
+    },
+    {
+      "epoch": 10.705089820359282,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 7151
+    },
+    {
+      "epoch": 10.706586826347305,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 7152
+    },
+    {
+      "epoch": 10.70808383233533,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1463,
+      "step": 7153
+    },
+    {
+      "epoch": 10.709580838323353,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 7154
+    },
+    {
+      "epoch": 10.711077844311378,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 7155
+    },
+    {
+      "epoch": 10.7125748502994,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 7156
+    },
+    {
+      "epoch": 10.714071856287426,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1446,
+      "step": 7157
+    },
+    {
+      "epoch": 10.715568862275449,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 7158
+    },
+    {
+      "epoch": 10.717065868263473,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1498,
+      "step": 7159
+    },
+    {
+      "epoch": 10.718562874251496,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 7160
+    },
+    {
+      "epoch": 10.720059880239521,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 7161
+    },
+    {
+      "epoch": 10.721556886227544,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1472,
+      "step": 7162
+    },
+    {
+      "epoch": 10.72305389221557,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 7163
+    },
+    {
+      "epoch": 10.724550898203592,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1464,
+      "step": 7164
+    },
+    {
+      "epoch": 10.726047904191617,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 7165
+    },
+    {
+      "epoch": 10.72754491017964,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1476,
+      "step": 7166
+    },
+    {
+      "epoch": 10.729041916167665,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1434,
+      "step": 7167
+    },
+    {
+      "epoch": 10.730538922155688,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 7168
+    },
+    {
+      "epoch": 10.732035928143713,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1471,
+      "step": 7169
+    },
+    {
+      "epoch": 10.733532934131736,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 7170
+    },
+    {
+      "epoch": 10.73502994011976,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1498,
+      "step": 7171
+    },
+    {
+      "epoch": 10.736526946107784,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1464,
+      "step": 7172
+    },
+    {
+      "epoch": 10.738023952095809,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0008,
+      "loss": 1.147,
+      "step": 7173
+    },
+    {
+      "epoch": 10.739520958083832,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1513,
+      "step": 7174
+    },
+    {
+      "epoch": 10.741017964071856,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1428,
+      "step": 7175
+    },
+    {
+      "epoch": 10.74251497005988,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 7176
+    },
+    {
+      "epoch": 10.744011976047904,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 7177
+    },
+    {
+      "epoch": 10.745508982035929,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1501,
+      "step": 7178
+    },
+    {
+      "epoch": 10.747005988023952,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.146,
+      "step": 7179
+    },
+    {
+      "epoch": 10.748502994011975,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 7180
+    },
+    {
+      "epoch": 10.75,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 7181
+    },
+    {
+      "epoch": 10.751497005988025,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 7182
+    },
+    {
+      "epoch": 10.752994011976048,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 7183
+    },
+    {
+      "epoch": 10.754491017964071,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1468,
+      "step": 7184
+    },
+    {
+      "epoch": 10.755988023952096,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1471,
+      "step": 7185
+    },
+    {
+      "epoch": 10.75748502994012,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1472,
+      "step": 7186
+    },
+    {
+      "epoch": 10.758982035928144,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 7187
+    },
+    {
+      "epoch": 10.760479041916168,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 7188
+    },
+    {
+      "epoch": 10.761976047904191,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1504,
+      "step": 7189
+    },
+    {
+      "epoch": 10.763473053892216,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1413,
+      "step": 7190
+    },
+    {
+      "epoch": 10.76497005988024,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.146,
+      "step": 7191
+    },
+    {
+      "epoch": 10.766467065868264,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 7192
+    },
+    {
+      "epoch": 10.767964071856287,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.151,
+      "step": 7193
+    },
+    {
+      "epoch": 10.769461077844312,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 7194
+    },
+    {
+      "epoch": 10.770958083832335,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1495,
+      "step": 7195
+    },
+    {
+      "epoch": 10.77245508982036,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1507,
+      "step": 7196
+    },
+    {
+      "epoch": 10.773952095808383,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1462,
+      "step": 7197
+    },
+    {
+      "epoch": 10.775449101796408,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 7198
+    },
+    {
+      "epoch": 10.77694610778443,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1485,
+      "step": 7199
+    },
+    {
+      "epoch": 10.778443113772456,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 7200
+    },
+    {
+      "epoch": 10.779940119760479,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 7201
+    },
+    {
+      "epoch": 10.781437125748504,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 7202
+    },
+    {
+      "epoch": 10.782934131736527,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1457,
+      "step": 7203
+    },
+    {
+      "epoch": 10.784431137724551,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 7204
+    },
+    {
+      "epoch": 10.785928143712574,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 7205
+    },
+    {
+      "epoch": 10.7874251497006,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1473,
+      "step": 7206
+    },
+    {
+      "epoch": 10.788922155688622,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 7207
+    },
+    {
+      "epoch": 10.790419161676647,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1475,
+      "step": 7208
+    },
+    {
+      "epoch": 10.79191616766467,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1512,
+      "step": 7209
+    },
+    {
+      "epoch": 10.793413173652695,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 7210
+    },
+    {
+      "epoch": 10.794910179640718,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 7211
+    },
+    {
+      "epoch": 10.796407185628743,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1373,
+      "step": 7212
+    },
+    {
+      "epoch": 10.797904191616766,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1454,
+      "step": 7213
+    },
+    {
+      "epoch": 10.79940119760479,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 7214
+    },
+    {
+      "epoch": 10.800898203592814,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 7215
+    },
+    {
+      "epoch": 10.802395209580839,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1462,
+      "step": 7216
+    },
+    {
+      "epoch": 10.803892215568862,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1468,
+      "step": 7217
+    },
+    {
+      "epoch": 10.805389221556887,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.144,
+      "step": 7218
+    },
+    {
+      "epoch": 10.80688622754491,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1487,
+      "step": 7219
+    },
+    {
+      "epoch": 10.808383233532934,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1416,
+      "step": 7220
+    },
+    {
+      "epoch": 10.809880239520957,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1472,
+      "step": 7221
+    },
+    {
+      "epoch": 10.811377245508982,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1464,
+      "step": 7222
+    },
+    {
+      "epoch": 10.812874251497005,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 7223
+    },
+    {
+      "epoch": 10.81437125748503,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1536,
+      "step": 7224
+    },
+    {
+      "epoch": 10.815868263473053,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1377,
+      "step": 7225
+    },
+    {
+      "epoch": 10.817365269461078,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1475,
+      "step": 7226
+    },
+    {
+      "epoch": 10.818862275449101,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1486,
+      "step": 7227
+    },
+    {
+      "epoch": 10.820359281437126,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1464,
+      "step": 7228
+    },
+    {
+      "epoch": 10.821856287425149,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1416,
+      "step": 7229
+    },
+    {
+      "epoch": 10.823353293413174,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1468,
+      "step": 7230
+    },
+    {
+      "epoch": 10.824850299401197,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1483,
+      "step": 7231
+    },
+    {
+      "epoch": 10.826347305389222,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 7232
+    },
+    {
+      "epoch": 10.827844311377245,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1457,
+      "step": 7233
+    },
+    {
+      "epoch": 10.82934131736527,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 7234
+    },
+    {
+      "epoch": 10.830838323353294,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 7235
+    },
+    {
+      "epoch": 10.832335329341317,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1488,
+      "step": 7236
+    },
+    {
+      "epoch": 10.83383233532934,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1459,
+      "step": 7237
+    },
+    {
+      "epoch": 10.835329341317365,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 7238
+    },
+    {
+      "epoch": 10.83682634730539,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1471,
+      "step": 7239
+    },
+    {
+      "epoch": 10.838323353293413,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1518,
+      "step": 7240
+    },
+    {
+      "epoch": 10.839820359281438,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 7241
+    },
+    {
+      "epoch": 10.841317365269461,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.152,
+      "step": 7242
+    },
+    {
+      "epoch": 10.842814371257486,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 7243
+    },
+    {
+      "epoch": 10.844311377245509,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1462,
+      "step": 7244
+    },
+    {
+      "epoch": 10.845808383233534,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 7245
+    },
+    {
+      "epoch": 10.847305389221557,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 7246
+    },
+    {
+      "epoch": 10.848802395209582,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 7247
+    },
+    {
+      "epoch": 10.850299401197605,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1393,
+      "step": 7248
+    },
+    {
+      "epoch": 10.85179640718563,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 7249
+    },
+    {
+      "epoch": 10.853293413173652,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 7250
+    },
+    {
+      "epoch": 10.854790419161677,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 7251
+    },
+    {
+      "epoch": 10.8562874251497,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 7252
+    },
+    {
+      "epoch": 10.857784431137725,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 7253
+    },
+    {
+      "epoch": 10.859281437125748,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 7254
+    },
+    {
+      "epoch": 10.860778443113773,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 7255
+    },
+    {
+      "epoch": 10.862275449101796,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 7256
+    },
+    {
+      "epoch": 10.863772455089821,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1492,
+      "step": 7257
+    },
+    {
+      "epoch": 10.865269461077844,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 7258
+    },
+    {
+      "epoch": 10.866766467065869,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1336,
+      "step": 7259
+    },
+    {
+      "epoch": 10.868263473053892,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1498,
+      "step": 7260
+    },
+    {
+      "epoch": 10.869760479041917,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 7261
+    },
+    {
+      "epoch": 10.87125748502994,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1471,
+      "step": 7262
+    },
+    {
+      "epoch": 10.872754491017965,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1517,
+      "step": 7263
+    },
+    {
+      "epoch": 10.874251497005988,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 7264
+    },
+    {
+      "epoch": 10.875748502994012,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 7265
+    },
+    {
+      "epoch": 10.877245508982035,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1462,
+      "step": 7266
+    },
+    {
+      "epoch": 10.87874251497006,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 7267
+    },
+    {
+      "epoch": 10.880239520958083,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 7268
+    },
+    {
+      "epoch": 10.881736526946108,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 7269
+    },
+    {
+      "epoch": 10.883233532934131,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.148,
+      "step": 7270
+    },
+    {
+      "epoch": 10.884730538922156,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 7271
+    },
+    {
+      "epoch": 10.886227544910179,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 7272
+    },
+    {
+      "epoch": 10.887724550898204,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1484,
+      "step": 7273
+    },
+    {
+      "epoch": 10.889221556886227,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 7274
+    },
+    {
+      "epoch": 10.890718562874252,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1469,
+      "step": 7275
+    },
+    {
+      "epoch": 10.892215568862275,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 7276
+    },
+    {
+      "epoch": 10.8937125748503,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 7277
+    },
+    {
+      "epoch": 10.895209580838323,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 7278
+    },
+    {
+      "epoch": 10.896706586826348,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 7279
+    },
+    {
+      "epoch": 10.89820359281437,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1509,
+      "step": 7280
+    },
+    {
+      "epoch": 10.899700598802395,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 7281
+    },
+    {
+      "epoch": 10.901197604790418,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1495,
+      "step": 7282
+    },
+    {
+      "epoch": 10.902694610778443,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 7283
+    },
+    {
+      "epoch": 10.904191616766466,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 7284
+    },
+    {
+      "epoch": 10.905688622754491,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1469,
+      "step": 7285
+    },
+    {
+      "epoch": 10.907185628742514,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.144,
+      "step": 7286
+    },
+    {
+      "epoch": 10.908682634730539,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 7287
+    },
+    {
+      "epoch": 10.910179640718562,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 7288
+    },
+    {
+      "epoch": 10.911676646706587,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1513,
+      "step": 7289
+    },
+    {
+      "epoch": 10.91317365269461,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1395,
+      "step": 7290
+    },
+    {
+      "epoch": 10.914670658682635,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1465,
+      "step": 7291
+    },
+    {
+      "epoch": 10.91616766467066,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 7292
+    },
+    {
+      "epoch": 10.917664670658683,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 7293
+    },
+    {
+      "epoch": 10.919161676646706,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1459,
+      "step": 7294
+    },
+    {
+      "epoch": 10.92065868263473,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 7295
+    },
+    {
+      "epoch": 10.922155688622755,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 7296
+    },
+    {
+      "epoch": 10.923652694610778,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.15,
+      "step": 7297
+    },
+    {
+      "epoch": 10.925149700598803,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1369,
+      "step": 7298
+    },
+    {
+      "epoch": 10.926646706586826,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 7299
+    },
+    {
+      "epoch": 10.928143712574851,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1536,
+      "step": 7300
+    },
+    {
+      "epoch": 10.929640718562874,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1485,
+      "step": 7301
+    },
+    {
+      "epoch": 10.931137724550899,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 7302
+    },
+    {
+      "epoch": 10.932634730538922,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 7303
+    },
+    {
+      "epoch": 10.934131736526947,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 7304
+    },
+    {
+      "epoch": 10.93562874251497,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 7305
+    },
+    {
+      "epoch": 10.937125748502995,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1429,
+      "step": 7306
+    },
+    {
+      "epoch": 10.938622754491018,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1468,
+      "step": 7307
+    },
+    {
+      "epoch": 10.940119760479043,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1454,
+      "step": 7308
+    },
+    {
+      "epoch": 10.941616766467066,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 7309
+    },
+    {
+      "epoch": 10.94311377245509,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 7310
+    },
+    {
+      "epoch": 10.944610778443113,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 7311
+    },
+    {
+      "epoch": 10.946107784431138,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 7312
+    },
+    {
+      "epoch": 10.947604790419161,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 7313
+    },
+    {
+      "epoch": 10.949101796407186,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 7314
+    },
+    {
+      "epoch": 10.95059880239521,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1513,
+      "step": 7315
+    },
+    {
+      "epoch": 10.952095808383234,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 7316
+    },
+    {
+      "epoch": 10.953592814371257,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 7317
+    },
+    {
+      "epoch": 10.955089820359282,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 7318
+    },
+    {
+      "epoch": 10.956586826347305,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 7319
+    },
+    {
+      "epoch": 10.95808383233533,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1459,
+      "step": 7320
+    },
+    {
+      "epoch": 10.959580838323353,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 7321
+    },
+    {
+      "epoch": 10.961077844311378,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1476,
+      "step": 7322
+    },
+    {
+      "epoch": 10.9625748502994,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 7323
+    },
+    {
+      "epoch": 10.964071856287426,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1483,
+      "step": 7324
+    },
+    {
+      "epoch": 10.965568862275449,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 7325
+    },
+    {
+      "epoch": 10.967065868263473,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 7326
+    },
+    {
+      "epoch": 10.968562874251496,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 7327
+    },
+    {
+      "epoch": 10.970059880239521,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1376,
+      "step": 7328
+    },
+    {
+      "epoch": 10.971556886227544,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 7329
+    },
+    {
+      "epoch": 10.97305389221557,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 7330
+    },
+    {
+      "epoch": 10.974550898203592,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 7331
+    },
+    {
+      "epoch": 10.976047904191617,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 7332
+    },
+    {
+      "epoch": 10.97754491017964,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1469,
+      "step": 7333
+    },
+    {
+      "epoch": 10.979041916167665,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1514,
+      "step": 7334
+    },
+    {
+      "epoch": 10.980538922155688,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 7335
+    },
+    {
+      "epoch": 10.982035928143713,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1378,
+      "step": 7336
+    },
+    {
+      "epoch": 10.983532934131736,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 7337
+    },
+    {
+      "epoch": 10.98502994011976,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 7338
+    },
+    {
+      "epoch": 10.986526946107784,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1484,
+      "step": 7339
+    },
+    {
+      "epoch": 10.988023952095809,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1473,
+      "step": 7340
+    },
+    {
+      "epoch": 10.989520958083832,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1475,
+      "step": 7341
+    },
+    {
+      "epoch": 10.991017964071856,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 7342
+    },
+    {
+      "epoch": 10.99251497005988,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 7343
+    },
+    {
+      "epoch": 10.994011976047904,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 7344
+    },
+    {
+      "epoch": 10.995508982035929,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 7345
+    },
+    {
+      "epoch": 10.997005988023952,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 7346
+    },
+    {
+      "epoch": 10.998502994011975,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1384,
+      "step": 7347
+    },
+    {
+      "epoch": 11.0,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 7348
+    },
+    {
+      "epoch": 11.001497005988025,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 7349
+    },
+    {
+      "epoch": 11.002994011976048,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 7350
+    },
+    {
+      "epoch": 11.004491017964073,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 7351
+    },
+    {
+      "epoch": 11.005988023952096,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1493,
+      "step": 7352
+    },
+    {
+      "epoch": 11.00748502994012,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1477,
+      "step": 7353
+    },
+    {
+      "epoch": 11.008982035928144,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 7354
+    },
+    {
+      "epoch": 11.010479041916168,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1495,
+      "step": 7355
+    },
+    {
+      "epoch": 11.011976047904191,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 7356
+    },
+    {
+      "epoch": 11.013473053892216,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1416,
+      "step": 7357
+    },
+    {
+      "epoch": 11.01497005988024,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 7358
+    },
+    {
+      "epoch": 11.016467065868264,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 7359
+    },
+    {
+      "epoch": 11.017964071856287,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 7360
+    },
+    {
+      "epoch": 11.019461077844312,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 7361
+    },
+    {
+      "epoch": 11.020958083832335,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 7362
+    },
+    {
+      "epoch": 11.02245508982036,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1492,
+      "step": 7363
+    },
+    {
+      "epoch": 11.023952095808383,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 7364
+    },
+    {
+      "epoch": 11.025449101796408,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1469,
+      "step": 7365
+    },
+    {
+      "epoch": 11.02694610778443,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 7366
+    },
+    {
+      "epoch": 11.028443113772456,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 7367
+    },
+    {
+      "epoch": 11.029940119760479,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 7368
+    },
+    {
+      "epoch": 11.031437125748504,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1386,
+      "step": 7369
+    },
+    {
+      "epoch": 11.032934131736527,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1501,
+      "step": 7370
+    },
+    {
+      "epoch": 11.034431137724551,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 7371
+    },
+    {
+      "epoch": 11.035928143712574,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1369,
+      "step": 7372
+    },
+    {
+      "epoch": 11.0374251497006,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 7373
+    },
+    {
+      "epoch": 11.038922155688622,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 7374
+    },
+    {
+      "epoch": 11.040419161676647,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1365,
+      "step": 7375
+    },
+    {
+      "epoch": 11.04191616766467,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 7376
+    },
+    {
+      "epoch": 11.043413173652695,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1462,
+      "step": 7377
+    },
+    {
+      "epoch": 11.044910179640718,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 7378
+    },
+    {
+      "epoch": 11.046407185628743,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 7379
+    },
+    {
+      "epoch": 11.047904191616766,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.146,
+      "step": 7380
+    },
+    {
+      "epoch": 11.04940119760479,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1526,
+      "step": 7381
+    },
+    {
+      "epoch": 11.050898203592814,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1386,
+      "step": 7382
+    },
+    {
+      "epoch": 11.052395209580839,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1376,
+      "step": 7383
+    },
+    {
+      "epoch": 11.053892215568862,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1526,
+      "step": 7384
+    },
+    {
+      "epoch": 11.055389221556887,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 7385
+    },
+    {
+      "epoch": 11.05688622754491,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 0.0008,
+      "loss": 1.146,
+      "step": 7386
+    },
+    {
+      "epoch": 11.058383233532934,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 7387
+    },
+    {
+      "epoch": 11.059880239520957,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0008,
+      "loss": 1.136,
+      "step": 7388
+    },
+    {
+      "epoch": 11.061377245508982,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1465,
+      "step": 7389
+    },
+    {
+      "epoch": 11.062874251497005,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1463,
+      "step": 7390
+    },
+    {
+      "epoch": 11.06437125748503,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.151,
+      "step": 7391
+    },
+    {
+      "epoch": 11.065868263473053,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1463,
+      "step": 7392
+    },
+    {
+      "epoch": 11.067365269461078,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 7393
+    },
+    {
+      "epoch": 11.068862275449101,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.15,
+      "step": 7394
+    },
+    {
+      "epoch": 11.070359281437126,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 7395
+    },
+    {
+      "epoch": 11.071856287425149,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 7396
+    },
+    {
+      "epoch": 11.073353293413174,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 7397
+    },
+    {
+      "epoch": 11.074850299401197,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 7398
+    },
+    {
+      "epoch": 11.076347305389222,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 7399
+    },
+    {
+      "epoch": 11.077844311377245,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1428,
+      "step": 7400
+    },
+    {
+      "epoch": 11.07934131736527,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1494,
+      "step": 7401
+    },
+    {
+      "epoch": 11.080838323353293,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 7402
+    },
+    {
+      "epoch": 11.082335329341317,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1502,
+      "step": 7403
+    },
+    {
+      "epoch": 11.08383233532934,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1481,
+      "step": 7404
+    },
+    {
+      "epoch": 11.085329341317365,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 7405
+    },
+    {
+      "epoch": 11.08682634730539,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 7406
+    },
+    {
+      "epoch": 11.088323353293413,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.147,
+      "step": 7407
+    },
+    {
+      "epoch": 11.089820359281438,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1443,
+      "step": 7408
+    },
+    {
+      "epoch": 11.091317365269461,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1495,
+      "step": 7409
+    },
+    {
+      "epoch": 11.092814371257486,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1446,
+      "step": 7410
+    },
+    {
+      "epoch": 11.094311377245509,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1468,
+      "step": 7411
+    },
+    {
+      "epoch": 11.095808383233534,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 7412
+    },
+    {
+      "epoch": 11.097305389221557,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 7413
+    },
+    {
+      "epoch": 11.098802395209582,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1494,
+      "step": 7414
+    },
+    {
+      "epoch": 11.100299401197605,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 7415
+    },
+    {
+      "epoch": 11.10179640718563,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1383,
+      "step": 7416
+    },
+    {
+      "epoch": 11.103293413173652,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1392,
+      "step": 7417
+    },
+    {
+      "epoch": 11.104790419161677,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 7418
+    },
+    {
+      "epoch": 11.1062874251497,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1443,
+      "step": 7419
+    },
+    {
+      "epoch": 11.107784431137725,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 7420
+    },
+    {
+      "epoch": 11.109281437125748,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1365,
+      "step": 7421
+    },
+    {
+      "epoch": 11.110778443113773,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 7422
+    },
+    {
+      "epoch": 11.112275449101796,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 7423
+    },
+    {
+      "epoch": 11.113772455089821,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 7424
+    },
+    {
+      "epoch": 11.115269461077844,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 7425
+    },
+    {
+      "epoch": 11.116766467065869,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1517,
+      "step": 7426
+    },
+    {
+      "epoch": 11.118263473053892,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 7427
+    },
+    {
+      "epoch": 11.119760479041917,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1443,
+      "step": 7428
+    },
+    {
+      "epoch": 11.12125748502994,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 7429
+    },
+    {
+      "epoch": 11.122754491017965,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 7430
+    },
+    {
+      "epoch": 11.124251497005988,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 7431
+    },
+    {
+      "epoch": 11.125748502994012,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1475,
+      "step": 7432
+    },
+    {
+      "epoch": 11.127245508982035,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1479,
+      "step": 7433
+    },
+    {
+      "epoch": 11.12874251497006,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 7434
+    },
+    {
+      "epoch": 11.130239520958083,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1479,
+      "step": 7435
+    },
+    {
+      "epoch": 11.131736526946108,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 7436
+    },
+    {
+      "epoch": 11.133233532934131,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 7437
+    },
+    {
+      "epoch": 11.134730538922156,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1484,
+      "step": 7438
+    },
+    {
+      "epoch": 11.136227544910179,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 7439
+    },
+    {
+      "epoch": 11.137724550898204,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1472,
+      "step": 7440
+    },
+    {
+      "epoch": 11.139221556886227,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1443,
+      "step": 7441
+    },
+    {
+      "epoch": 11.140718562874252,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 7442
+    },
+    {
+      "epoch": 11.142215568862275,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 7443
+    },
+    {
+      "epoch": 11.1437125748503,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1487,
+      "step": 7444
+    },
+    {
+      "epoch": 11.145209580838323,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1515,
+      "step": 7445
+    },
+    {
+      "epoch": 11.146706586826348,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1387,
+      "step": 7446
+    },
+    {
+      "epoch": 11.14820359281437,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1477,
+      "step": 7447
+    },
+    {
+      "epoch": 11.149700598802395,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 7448
+    },
+    {
+      "epoch": 11.151197604790418,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1368,
+      "step": 7449
+    },
+    {
+      "epoch": 11.152694610778443,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1481,
+      "step": 7450
+    },
+    {
+      "epoch": 11.154191616766466,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1481,
+      "step": 7451
+    },
+    {
+      "epoch": 11.155688622754491,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 7452
+    },
+    {
+      "epoch": 11.157185628742514,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 7453
+    },
+    {
+      "epoch": 11.158682634730539,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 7454
+    },
+    {
+      "epoch": 11.160179640718562,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1533,
+      "step": 7455
+    },
+    {
+      "epoch": 11.161676646706587,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.147,
+      "step": 7456
+    },
+    {
+      "epoch": 11.16317365269461,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1459,
+      "step": 7457
+    },
+    {
+      "epoch": 11.164670658682635,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 7458
+    },
+    {
+      "epoch": 11.16616766467066,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1556,
+      "step": 7459
+    },
+    {
+      "epoch": 11.167664670658683,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 7460
+    },
+    {
+      "epoch": 11.169161676646707,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 7461
+    },
+    {
+      "epoch": 11.17065868263473,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 7462
+    },
+    {
+      "epoch": 11.172155688622755,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.144,
+      "step": 7463
+    },
+    {
+      "epoch": 11.173652694610778,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 7464
+    },
+    {
+      "epoch": 11.175149700598803,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1464,
+      "step": 7465
+    },
+    {
+      "epoch": 11.176646706586826,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 7466
+    },
+    {
+      "epoch": 11.178143712574851,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1493,
+      "step": 7467
+    },
+    {
+      "epoch": 11.179640718562874,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 7468
+    },
+    {
+      "epoch": 11.181137724550899,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 7469
+    },
+    {
+      "epoch": 11.182634730538922,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 7470
+    },
+    {
+      "epoch": 11.184131736526947,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1506,
+      "step": 7471
+    },
+    {
+      "epoch": 11.18562874251497,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1477,
+      "step": 7472
+    },
+    {
+      "epoch": 11.187125748502995,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 7473
+    },
+    {
+      "epoch": 11.188622754491018,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 7474
+    },
+    {
+      "epoch": 11.190119760479043,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 7475
+    },
+    {
+      "epoch": 11.191616766467066,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1515,
+      "step": 7476
+    },
+    {
+      "epoch": 11.19311377245509,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 7477
+    },
+    {
+      "epoch": 11.194610778443113,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1475,
+      "step": 7478
+    },
+    {
+      "epoch": 11.196107784431138,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 7479
+    },
+    {
+      "epoch": 11.197604790419161,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 7480
+    },
+    {
+      "epoch": 11.199101796407186,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1486,
+      "step": 7481
+    },
+    {
+      "epoch": 11.20059880239521,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 7482
+    },
+    {
+      "epoch": 11.202095808383234,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1428,
+      "step": 7483
+    },
+    {
+      "epoch": 11.203592814371257,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 7484
+    },
+    {
+      "epoch": 11.205089820359282,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 7485
+    },
+    {
+      "epoch": 11.206586826347305,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.15,
+      "step": 7486
+    },
+    {
+      "epoch": 11.20808383233533,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 7487
+    },
+    {
+      "epoch": 11.209580838323353,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 7488
+    },
+    {
+      "epoch": 11.211077844311378,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 7489
+    },
+    {
+      "epoch": 11.2125748502994,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1471,
+      "step": 7490
+    },
+    {
+      "epoch": 11.214071856287426,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1443,
+      "step": 7491
+    },
+    {
+      "epoch": 11.215568862275449,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 7492
+    },
+    {
+      "epoch": 11.217065868263473,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 7493
+    },
+    {
+      "epoch": 11.218562874251496,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1367,
+      "step": 7494
+    },
+    {
+      "epoch": 11.220059880239521,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1523,
+      "step": 7495
+    },
+    {
+      "epoch": 11.221556886227544,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 7496
+    },
+    {
+      "epoch": 11.22305389221557,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 7497
+    },
+    {
+      "epoch": 11.224550898203592,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1469,
+      "step": 7498
+    },
+    {
+      "epoch": 11.226047904191617,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1459,
+      "step": 7499
+    },
+    {
+      "epoch": 11.22754491017964,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.146,
+      "step": 7500
+    },
+    {
+      "epoch": 11.229041916167665,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 7501
+    },
+    {
+      "epoch": 11.230538922155688,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 7502
+    },
+    {
+      "epoch": 11.232035928143713,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 7503
+    },
+    {
+      "epoch": 11.233532934131736,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1366,
+      "step": 7504
+    },
+    {
+      "epoch": 11.23502994011976,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 7505
+    },
+    {
+      "epoch": 11.236526946107784,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 7506
+    },
+    {
+      "epoch": 11.238023952095809,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 7507
+    },
+    {
+      "epoch": 11.239520958083832,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1541,
+      "step": 7508
+    },
+    {
+      "epoch": 11.241017964071856,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1386,
+      "step": 7509
+    },
+    {
+      "epoch": 11.24251497005988,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 7510
+    },
+    {
+      "epoch": 11.244011976047904,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1494,
+      "step": 7511
+    },
+    {
+      "epoch": 11.245508982035927,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1387,
+      "step": 7512
+    },
+    {
+      "epoch": 11.247005988023952,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1489,
+      "step": 7513
+    },
+    {
+      "epoch": 11.248502994011975,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1505,
+      "step": 7514
+    },
+    {
+      "epoch": 11.25,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1377,
+      "step": 7515
+    },
+    {
+      "epoch": 11.251497005988025,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1468,
+      "step": 7516
+    },
+    {
+      "epoch": 11.252994011976048,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 7517
+    },
+    {
+      "epoch": 11.254491017964073,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1434,
+      "step": 7518
+    },
+    {
+      "epoch": 11.255988023952096,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1366,
+      "step": 7519
+    },
+    {
+      "epoch": 11.25748502994012,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1464,
+      "step": 7520
+    },
+    {
+      "epoch": 11.258982035928144,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1468,
+      "step": 7521
+    },
+    {
+      "epoch": 11.260479041916168,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 7522
+    },
+    {
+      "epoch": 11.261976047904191,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1494,
+      "step": 7523
+    },
+    {
+      "epoch": 11.263473053892216,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 7524
+    },
+    {
+      "epoch": 11.26497005988024,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 7525
+    },
+    {
+      "epoch": 11.266467065868264,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.148,
+      "step": 7526
+    },
+    {
+      "epoch": 11.267964071856287,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1334,
+      "step": 7527
+    },
+    {
+      "epoch": 11.269461077844312,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 7528
+    },
+    {
+      "epoch": 11.270958083832335,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1465,
+      "step": 7529
+    },
+    {
+      "epoch": 11.27245508982036,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 7530
+    },
+    {
+      "epoch": 11.273952095808383,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 7531
+    },
+    {
+      "epoch": 11.275449101796408,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 7532
+    },
+    {
+      "epoch": 11.27694610778443,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 7533
+    },
+    {
+      "epoch": 11.278443113772456,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 7534
+    },
+    {
+      "epoch": 11.279940119760479,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 7535
+    },
+    {
+      "epoch": 11.281437125748504,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1351,
+      "step": 7536
+    },
+    {
+      "epoch": 11.282934131736527,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1429,
+      "step": 7537
+    },
+    {
+      "epoch": 11.284431137724551,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.144,
+      "step": 7538
+    },
+    {
+      "epoch": 11.285928143712574,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1434,
+      "step": 7539
+    },
+    {
+      "epoch": 11.2874251497006,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 7540
+    },
+    {
+      "epoch": 11.288922155688622,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 7541
+    },
+    {
+      "epoch": 11.290419161676647,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 7542
+    },
+    {
+      "epoch": 11.29191616766467,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 7543
+    },
+    {
+      "epoch": 11.293413173652695,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1532,
+      "step": 7544
+    },
+    {
+      "epoch": 11.294910179640718,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1496,
+      "step": 7545
+    },
+    {
+      "epoch": 11.296407185628743,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 7546
+    },
+    {
+      "epoch": 11.297904191616766,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 7547
+    },
+    {
+      "epoch": 11.29940119760479,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 7548
+    },
+    {
+      "epoch": 11.300898203592814,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 7549
+    },
+    {
+      "epoch": 11.302395209580839,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 7550
+    },
+    {
+      "epoch": 11.303892215568862,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1506,
+      "step": 7551
+    },
+    {
+      "epoch": 11.305389221556887,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1368,
+      "step": 7552
+    },
+    {
+      "epoch": 11.30688622754491,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 7553
+    },
+    {
+      "epoch": 11.308383233532934,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1457,
+      "step": 7554
+    },
+    {
+      "epoch": 11.309880239520957,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 7555
+    },
+    {
+      "epoch": 11.311377245508982,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1348,
+      "step": 7556
+    },
+    {
+      "epoch": 11.312874251497005,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 7557
+    },
+    {
+      "epoch": 11.31437125748503,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 7558
+    },
+    {
+      "epoch": 11.315868263473053,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 7559
+    },
+    {
+      "epoch": 11.317365269461078,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 7560
+    },
+    {
+      "epoch": 11.318862275449101,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1361,
+      "step": 7561
+    },
+    {
+      "epoch": 11.320359281437126,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 7562
+    },
+    {
+      "epoch": 11.321856287425149,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1477,
+      "step": 7563
+    },
+    {
+      "epoch": 11.323353293413174,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 7564
+    },
+    {
+      "epoch": 11.324850299401197,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 7565
+    },
+    {
+      "epoch": 11.326347305389222,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 7566
+    },
+    {
+      "epoch": 11.327844311377245,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1387,
+      "step": 7567
+    },
+    {
+      "epoch": 11.32934131736527,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 7568
+    },
+    {
+      "epoch": 11.330838323353294,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 7569
+    },
+    {
+      "epoch": 11.332335329341317,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1446,
+      "step": 7570
+    },
+    {
+      "epoch": 11.33383233532934,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 7571
+    },
+    {
+      "epoch": 11.335329341317365,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 7572
+    },
+    {
+      "epoch": 11.33682634730539,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 7573
+    },
+    {
+      "epoch": 11.338323353293413,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1489,
+      "step": 7574
+    },
+    {
+      "epoch": 11.339820359281438,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1457,
+      "step": 7575
+    },
+    {
+      "epoch": 11.341317365269461,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 7576
+    },
+    {
+      "epoch": 11.342814371257486,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 7577
+    },
+    {
+      "epoch": 11.344311377245509,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 7578
+    },
+    {
+      "epoch": 11.345808383233534,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 7579
+    },
+    {
+      "epoch": 11.347305389221557,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 7580
+    },
+    {
+      "epoch": 11.348802395209582,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 7581
+    },
+    {
+      "epoch": 11.350299401197605,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 7582
+    },
+    {
+      "epoch": 11.35179640718563,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 7583
+    },
+    {
+      "epoch": 11.353293413173652,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1491,
+      "step": 7584
+    },
+    {
+      "epoch": 11.354790419161677,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1428,
+      "step": 7585
+    },
+    {
+      "epoch": 11.3562874251497,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1473,
+      "step": 7586
+    },
+    {
+      "epoch": 11.357784431137725,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 7587
+    },
+    {
+      "epoch": 11.359281437125748,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1416,
+      "step": 7588
+    },
+    {
+      "epoch": 11.360778443113773,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 7589
+    },
+    {
+      "epoch": 11.362275449101796,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 7590
+    },
+    {
+      "epoch": 11.363772455089821,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 7591
+    },
+    {
+      "epoch": 11.365269461077844,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 7592
+    },
+    {
+      "epoch": 11.366766467065869,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 7593
+    },
+    {
+      "epoch": 11.368263473053892,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 7594
+    },
+    {
+      "epoch": 11.369760479041917,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1459,
+      "step": 7595
+    },
+    {
+      "epoch": 11.37125748502994,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1362,
+      "step": 7596
+    },
+    {
+      "epoch": 11.372754491017965,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1393,
+      "step": 7597
+    },
+    {
+      "epoch": 11.374251497005988,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1496,
+      "step": 7598
+    },
+    {
+      "epoch": 11.375748502994012,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1496,
+      "step": 7599
+    },
+    {
+      "epoch": 11.377245508982035,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 7600
+    },
+    {
+      "epoch": 11.37874251497006,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 7601
+    },
+    {
+      "epoch": 11.380239520958083,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1459,
+      "step": 7602
+    },
+    {
+      "epoch": 11.381736526946108,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 7603
+    },
+    {
+      "epoch": 11.383233532934131,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1367,
+      "step": 7604
+    },
+    {
+      "epoch": 11.384730538922156,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 7605
+    },
+    {
+      "epoch": 11.386227544910179,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 7606
+    },
+    {
+      "epoch": 11.387724550898204,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 7607
+    },
+    {
+      "epoch": 11.389221556886227,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 7608
+    },
+    {
+      "epoch": 11.390718562874252,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 7609
+    },
+    {
+      "epoch": 11.392215568862275,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1502,
+      "step": 7610
+    },
+    {
+      "epoch": 11.3937125748503,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1469,
+      "step": 7611
+    },
+    {
+      "epoch": 11.395209580838323,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 7612
+    },
+    {
+      "epoch": 11.396706586826348,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1381,
+      "step": 7613
+    },
+    {
+      "epoch": 11.39820359281437,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 7614
+    },
+    {
+      "epoch": 11.399700598802395,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1484,
+      "step": 7615
+    },
+    {
+      "epoch": 11.401197604790418,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 7616
+    },
+    {
+      "epoch": 11.402694610778443,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 7617
+    },
+    {
+      "epoch": 11.404191616766466,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1464,
+      "step": 7618
+    },
+    {
+      "epoch": 11.405688622754491,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 7619
+    },
+    {
+      "epoch": 11.407185628742514,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 7620
+    },
+    {
+      "epoch": 11.408682634730539,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 7621
+    },
+    {
+      "epoch": 11.410179640718562,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 7622
+    },
+    {
+      "epoch": 11.411676646706587,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 7623
+    },
+    {
+      "epoch": 11.41317365269461,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 7624
+    },
+    {
+      "epoch": 11.414670658682635,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1386,
+      "step": 7625
+    },
+    {
+      "epoch": 11.41616766467066,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 7626
+    },
+    {
+      "epoch": 11.417664670658683,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 7627
+    },
+    {
+      "epoch": 11.419161676646706,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 7628
+    },
+    {
+      "epoch": 11.42065868263473,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1471,
+      "step": 7629
+    },
+    {
+      "epoch": 11.422155688622755,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1464,
+      "step": 7630
+    },
+    {
+      "epoch": 11.423652694610778,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1386,
+      "step": 7631
+    },
+    {
+      "epoch": 11.425149700598803,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1339,
+      "step": 7632
+    },
+    {
+      "epoch": 11.426646706586826,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1363,
+      "step": 7633
+    },
+    {
+      "epoch": 11.428143712574851,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 7634
+    },
+    {
+      "epoch": 11.429640718562874,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1475,
+      "step": 7635
+    },
+    {
+      "epoch": 11.431137724550899,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1475,
+      "step": 7636
+    },
+    {
+      "epoch": 11.432634730538922,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 7637
+    },
+    {
+      "epoch": 11.434131736526947,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 7638
+    },
+    {
+      "epoch": 11.43562874251497,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1459,
+      "step": 7639
+    },
+    {
+      "epoch": 11.437125748502995,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 7640
+    },
+    {
+      "epoch": 11.438622754491018,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1392,
+      "step": 7641
+    },
+    {
+      "epoch": 11.440119760479043,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 7642
+    },
+    {
+      "epoch": 11.441616766467066,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 7643
+    },
+    {
+      "epoch": 11.44311377245509,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1459,
+      "step": 7644
+    },
+    {
+      "epoch": 11.444610778443113,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 7645
+    },
+    {
+      "epoch": 11.446107784431138,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 7646
+    },
+    {
+      "epoch": 11.447604790419161,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 7647
+    },
+    {
+      "epoch": 11.449101796407186,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 7648
+    },
+    {
+      "epoch": 11.45059880239521,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 7649
+    },
+    {
+      "epoch": 11.452095808383234,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1416,
+      "step": 7650
+    },
+    {
+      "epoch": 11.453592814371257,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 7651
+    },
+    {
+      "epoch": 11.455089820359282,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1468,
+      "step": 7652
+    },
+    {
+      "epoch": 11.456586826347305,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1477,
+      "step": 7653
+    },
+    {
+      "epoch": 11.45808383233533,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1507,
+      "step": 7654
+    },
+    {
+      "epoch": 11.459580838323353,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 7655
+    },
+    {
+      "epoch": 11.461077844311378,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 7656
+    },
+    {
+      "epoch": 11.4625748502994,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 7657
+    },
+    {
+      "epoch": 11.464071856287426,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1457,
+      "step": 7658
+    },
+    {
+      "epoch": 11.465568862275449,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 7659
+    },
+    {
+      "epoch": 11.467065868263473,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1483,
+      "step": 7660
+    },
+    {
+      "epoch": 11.468562874251496,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 7661
+    },
+    {
+      "epoch": 11.470059880239521,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 7662
+    },
+    {
+      "epoch": 11.471556886227544,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 7663
+    },
+    {
+      "epoch": 11.47305389221557,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 7664
+    },
+    {
+      "epoch": 11.474550898203592,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 7665
+    },
+    {
+      "epoch": 11.476047904191617,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1379,
+      "step": 7666
+    },
+    {
+      "epoch": 11.47754491017964,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1413,
+      "step": 7667
+    },
+    {
+      "epoch": 11.479041916167665,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 7668
+    },
+    {
+      "epoch": 11.480538922155688,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 7669
+    },
+    {
+      "epoch": 11.482035928143713,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 7670
+    },
+    {
+      "epoch": 11.483532934131736,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 7671
+    },
+    {
+      "epoch": 11.48502994011976,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 7672
+    },
+    {
+      "epoch": 11.486526946107784,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 7673
+    },
+    {
+      "epoch": 11.488023952095809,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1393,
+      "step": 7674
+    },
+    {
+      "epoch": 11.489520958083832,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1462,
+      "step": 7675
+    },
+    {
+      "epoch": 11.491017964071856,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 7676
+    },
+    {
+      "epoch": 11.49251497005988,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 7677
+    },
+    {
+      "epoch": 11.494011976047904,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 7678
+    },
+    {
+      "epoch": 11.495508982035927,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1493,
+      "step": 7679
+    },
+    {
+      "epoch": 11.497005988023952,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 7680
+    },
+    {
+      "epoch": 11.498502994011975,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1491,
+      "step": 7681
+    },
+    {
+      "epoch": 11.5,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 7682
+    },
+    {
+      "epoch": 11.501497005988025,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 7683
+    },
+    {
+      "epoch": 11.502994011976048,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 7684
+    },
+    {
+      "epoch": 11.504491017964071,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 7685
+    },
+    {
+      "epoch": 11.505988023952096,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 7686
+    },
+    {
+      "epoch": 11.50748502994012,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 7687
+    },
+    {
+      "epoch": 11.508982035928144,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 7688
+    },
+    {
+      "epoch": 11.510479041916168,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1387,
+      "step": 7689
+    },
+    {
+      "epoch": 11.511976047904191,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1471,
+      "step": 7690
+    },
+    {
+      "epoch": 11.513473053892216,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.146,
+      "step": 7691
+    },
+    {
+      "epoch": 11.51497005988024,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1477,
+      "step": 7692
+    },
+    {
+      "epoch": 11.516467065868264,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 7693
+    },
+    {
+      "epoch": 11.517964071856287,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 7694
+    },
+    {
+      "epoch": 11.519461077844312,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1428,
+      "step": 7695
+    },
+    {
+      "epoch": 11.520958083832335,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 7696
+    },
+    {
+      "epoch": 11.52245508982036,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.135,
+      "step": 7697
+    },
+    {
+      "epoch": 11.523952095808383,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 7698
+    },
+    {
+      "epoch": 11.525449101796408,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1498,
+      "step": 7699
+    },
+    {
+      "epoch": 11.52694610778443,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1491,
+      "step": 7700
+    },
+    {
+      "epoch": 11.528443113772456,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1364,
+      "step": 7701
+    },
+    {
+      "epoch": 11.529940119760479,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1454,
+      "step": 7702
+    },
+    {
+      "epoch": 11.531437125748504,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 7703
+    },
+    {
+      "epoch": 11.532934131736527,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1485,
+      "step": 7704
+    },
+    {
+      "epoch": 11.534431137724551,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 7705
+    },
+    {
+      "epoch": 11.535928143712574,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1384,
+      "step": 7706
+    },
+    {
+      "epoch": 11.5374251497006,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 7707
+    },
+    {
+      "epoch": 11.538922155688622,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1481,
+      "step": 7708
+    },
+    {
+      "epoch": 11.540419161676647,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 7709
+    },
+    {
+      "epoch": 11.54191616766467,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 7710
+    },
+    {
+      "epoch": 11.543413173652695,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 7711
+    },
+    {
+      "epoch": 11.544910179640718,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 7712
+    },
+    {
+      "epoch": 11.546407185628743,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 7713
+    },
+    {
+      "epoch": 11.547904191616766,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 7714
+    },
+    {
+      "epoch": 11.54940119760479,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 7715
+    },
+    {
+      "epoch": 11.550898203592814,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 7716
+    },
+    {
+      "epoch": 11.552395209580839,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 7717
+    },
+    {
+      "epoch": 11.553892215568862,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 7718
+    },
+    {
+      "epoch": 11.555389221556887,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 7719
+    },
+    {
+      "epoch": 11.55688622754491,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 7720
+    },
+    {
+      "epoch": 11.558383233532934,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 7721
+    },
+    {
+      "epoch": 11.559880239520957,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 7722
+    },
+    {
+      "epoch": 11.561377245508982,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 7723
+    },
+    {
+      "epoch": 11.562874251497005,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.147,
+      "step": 7724
+    },
+    {
+      "epoch": 11.56437125748503,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 7725
+    },
+    {
+      "epoch": 11.565868263473053,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1519,
+      "step": 7726
+    },
+    {
+      "epoch": 11.567365269461078,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 7727
+    },
+    {
+      "epoch": 11.568862275449101,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 7728
+    },
+    {
+      "epoch": 11.570359281437126,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 7729
+    },
+    {
+      "epoch": 11.571856287425149,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 7730
+    },
+    {
+      "epoch": 11.573353293413174,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 7731
+    },
+    {
+      "epoch": 11.574850299401197,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 7732
+    },
+    {
+      "epoch": 11.576347305389222,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 7733
+    },
+    {
+      "epoch": 11.577844311377245,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1487,
+      "step": 7734
+    },
+    {
+      "epoch": 11.57934131736527,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 7735
+    },
+    {
+      "epoch": 11.580838323353294,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1363,
+      "step": 7736
+    },
+    {
+      "epoch": 11.582335329341317,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1384,
+      "step": 7737
+    },
+    {
+      "epoch": 11.58383233532934,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 7738
+    },
+    {
+      "epoch": 11.585329341317365,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1361,
+      "step": 7739
+    },
+    {
+      "epoch": 11.58682634730539,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1329,
+      "step": 7740
+    },
+    {
+      "epoch": 11.588323353293413,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1477,
+      "step": 7741
+    },
+    {
+      "epoch": 11.589820359281438,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1378,
+      "step": 7742
+    },
+    {
+      "epoch": 11.591317365269461,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 7743
+    },
+    {
+      "epoch": 11.592814371257486,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1477,
+      "step": 7744
+    },
+    {
+      "epoch": 11.594311377245509,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 7745
+    },
+    {
+      "epoch": 11.595808383233534,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 7746
+    },
+    {
+      "epoch": 11.597305389221557,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 7747
+    },
+    {
+      "epoch": 11.598802395209582,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 7748
+    },
+    {
+      "epoch": 11.600299401197605,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1379,
+      "step": 7749
+    },
+    {
+      "epoch": 11.60179640718563,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1464,
+      "step": 7750
+    },
+    {
+      "epoch": 11.603293413173652,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 7751
+    },
+    {
+      "epoch": 11.604790419161677,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 7752
+    },
+    {
+      "epoch": 11.6062874251497,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1471,
+      "step": 7753
+    },
+    {
+      "epoch": 11.607784431137725,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1368,
+      "step": 7754
+    },
+    {
+      "epoch": 11.609281437125748,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 7755
+    },
+    {
+      "epoch": 11.610778443113773,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 7756
+    },
+    {
+      "epoch": 11.612275449101796,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 7757
+    },
+    {
+      "epoch": 11.613772455089821,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 7758
+    },
+    {
+      "epoch": 11.615269461077844,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1395,
+      "step": 7759
+    },
+    {
+      "epoch": 11.616766467065869,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 7760
+    },
+    {
+      "epoch": 11.618263473053892,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 7761
+    },
+    {
+      "epoch": 11.619760479041917,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 7762
+    },
+    {
+      "epoch": 11.62125748502994,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1446,
+      "step": 7763
+    },
+    {
+      "epoch": 11.622754491017965,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.144,
+      "step": 7764
+    },
+    {
+      "epoch": 11.624251497005988,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 7765
+    },
+    {
+      "epoch": 11.625748502994012,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1524,
+      "step": 7766
+    },
+    {
+      "epoch": 11.627245508982035,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1465,
+      "step": 7767
+    },
+    {
+      "epoch": 11.62874251497006,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1477,
+      "step": 7768
+    },
+    {
+      "epoch": 11.630239520958083,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 7769
+    },
+    {
+      "epoch": 11.631736526946108,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1473,
+      "step": 7770
+    },
+    {
+      "epoch": 11.633233532934131,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1377,
+      "step": 7771
+    },
+    {
+      "epoch": 11.634730538922156,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 7772
+    },
+    {
+      "epoch": 11.636227544910179,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1454,
+      "step": 7773
+    },
+    {
+      "epoch": 11.637724550898204,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 7774
+    },
+    {
+      "epoch": 11.639221556886227,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1459,
+      "step": 7775
+    },
+    {
+      "epoch": 11.640718562874252,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 7776
+    },
+    {
+      "epoch": 11.642215568862275,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 7777
+    },
+    {
+      "epoch": 11.6437125748503,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 7778
+    },
+    {
+      "epoch": 11.645209580838323,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1361,
+      "step": 7779
+    },
+    {
+      "epoch": 11.646706586826348,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 7780
+    },
+    {
+      "epoch": 11.64820359281437,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 7781
+    },
+    {
+      "epoch": 11.649700598802395,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 7782
+    },
+    {
+      "epoch": 11.651197604790418,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1497,
+      "step": 7783
+    },
+    {
+      "epoch": 11.652694610778443,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1496,
+      "step": 7784
+    },
+    {
+      "epoch": 11.654191616766466,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1371,
+      "step": 7785
+    },
+    {
+      "epoch": 11.655688622754491,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1473,
+      "step": 7786
+    },
+    {
+      "epoch": 11.657185628742514,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 7787
+    },
+    {
+      "epoch": 11.658682634730539,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1491,
+      "step": 7788
+    },
+    {
+      "epoch": 11.660179640718562,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 7789
+    },
+    {
+      "epoch": 11.661676646706587,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1465,
+      "step": 7790
+    },
+    {
+      "epoch": 11.66317365269461,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 7791
+    },
+    {
+      "epoch": 11.664670658682635,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 7792
+    },
+    {
+      "epoch": 11.66616766467066,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 7793
+    },
+    {
+      "epoch": 11.667664670658683,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1454,
+      "step": 7794
+    },
+    {
+      "epoch": 11.669161676646706,
+      "grad_norm": 0.0546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 7795
+    },
+    {
+      "epoch": 11.67065868263473,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 7796
+    },
+    {
+      "epoch": 11.672155688622755,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 7797
+    },
+    {
+      "epoch": 11.673652694610778,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 7798
+    },
+    {
+      "epoch": 11.675149700598803,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1348,
+      "step": 7799
+    },
+    {
+      "epoch": 11.676646706586826,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1386,
+      "step": 7800
+    },
+    {
+      "epoch": 11.678143712574851,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1454,
+      "step": 7801
+    },
+    {
+      "epoch": 11.679640718562874,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 7802
+    },
+    {
+      "epoch": 11.681137724550899,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 7803
+    },
+    {
+      "epoch": 11.682634730538922,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1473,
+      "step": 7804
+    },
+    {
+      "epoch": 11.684131736526947,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 7805
+    },
+    {
+      "epoch": 11.68562874251497,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 7806
+    },
+    {
+      "epoch": 11.687125748502995,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1351,
+      "step": 7807
+    },
+    {
+      "epoch": 11.688622754491018,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 7808
+    },
+    {
+      "epoch": 11.690119760479043,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 7809
+    },
+    {
+      "epoch": 11.691616766467066,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1378,
+      "step": 7810
+    },
+    {
+      "epoch": 11.69311377245509,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 7811
+    },
+    {
+      "epoch": 11.694610778443113,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1495,
+      "step": 7812
+    },
+    {
+      "epoch": 11.696107784431138,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.147,
+      "step": 7813
+    },
+    {
+      "epoch": 11.697604790419161,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 7814
+    },
+    {
+      "epoch": 11.699101796407186,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 7815
+    },
+    {
+      "epoch": 11.70059880239521,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 7816
+    },
+    {
+      "epoch": 11.702095808383234,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1548,
+      "step": 7817
+    },
+    {
+      "epoch": 11.703592814371257,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1489,
+      "step": 7818
+    },
+    {
+      "epoch": 11.705089820359282,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 7819
+    },
+    {
+      "epoch": 11.706586826347305,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 7820
+    },
+    {
+      "epoch": 11.70808383233533,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 7821
+    },
+    {
+      "epoch": 11.709580838323353,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 7822
+    },
+    {
+      "epoch": 11.711077844311378,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 7823
+    },
+    {
+      "epoch": 11.7125748502994,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 7824
+    },
+    {
+      "epoch": 11.714071856287426,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 7825
+    },
+    {
+      "epoch": 11.715568862275449,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1473,
+      "step": 7826
+    },
+    {
+      "epoch": 11.717065868263473,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 7827
+    },
+    {
+      "epoch": 11.718562874251496,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1491,
+      "step": 7828
+    },
+    {
+      "epoch": 11.720059880239521,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1462,
+      "step": 7829
+    },
+    {
+      "epoch": 11.721556886227544,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 7830
+    },
+    {
+      "epoch": 11.72305389221557,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.148,
+      "step": 7831
+    },
+    {
+      "epoch": 11.724550898203592,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 7832
+    },
+    {
+      "epoch": 11.726047904191617,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 7833
+    },
+    {
+      "epoch": 11.72754491017964,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 7834
+    },
+    {
+      "epoch": 11.729041916167665,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1383,
+      "step": 7835
+    },
+    {
+      "epoch": 11.730538922155688,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 7836
+    },
+    {
+      "epoch": 11.732035928143713,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 7837
+    },
+    {
+      "epoch": 11.733532934131736,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 7838
+    },
+    {
+      "epoch": 11.73502994011976,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 7839
+    },
+    {
+      "epoch": 11.736526946107784,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 7840
+    },
+    {
+      "epoch": 11.738023952095809,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1359,
+      "step": 7841
+    },
+    {
+      "epoch": 11.739520958083832,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 7842
+    },
+    {
+      "epoch": 11.741017964071856,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1351,
+      "step": 7843
+    },
+    {
+      "epoch": 11.74251497005988,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 7844
+    },
+    {
+      "epoch": 11.744011976047904,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 7845
+    },
+    {
+      "epoch": 11.745508982035929,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 7846
+    },
+    {
+      "epoch": 11.747005988023952,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1416,
+      "step": 7847
+    },
+    {
+      "epoch": 11.748502994011975,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1454,
+      "step": 7848
+    },
+    {
+      "epoch": 11.75,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 7849
+    },
+    {
+      "epoch": 11.751497005988025,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 7850
+    },
+    {
+      "epoch": 11.752994011976048,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 7851
+    },
+    {
+      "epoch": 11.754491017964071,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1376,
+      "step": 7852
+    },
+    {
+      "epoch": 11.755988023952096,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 7853
+    },
+    {
+      "epoch": 11.75748502994012,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.14,
+      "step": 7854
+    },
+    {
+      "epoch": 11.758982035928144,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 7855
+    },
+    {
+      "epoch": 11.760479041916168,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 7856
+    },
+    {
+      "epoch": 11.761976047904191,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 7857
+    },
+    {
+      "epoch": 11.763473053892216,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1457,
+      "step": 7858
+    },
+    {
+      "epoch": 11.76497005988024,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 7859
+    },
+    {
+      "epoch": 11.766467065868264,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 7860
+    },
+    {
+      "epoch": 11.767964071856287,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1457,
+      "step": 7861
+    },
+    {
+      "epoch": 11.769461077844312,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1434,
+      "step": 7862
+    },
+    {
+      "epoch": 11.770958083832335,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 7863
+    },
+    {
+      "epoch": 11.77245508982036,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 7864
+    },
+    {
+      "epoch": 11.773952095808383,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1377,
+      "step": 7865
+    },
+    {
+      "epoch": 11.775449101796408,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1487,
+      "step": 7866
+    },
+    {
+      "epoch": 11.77694610778443,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1471,
+      "step": 7867
+    },
+    {
+      "epoch": 11.778443113772456,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 7868
+    },
+    {
+      "epoch": 11.779940119760479,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1483,
+      "step": 7869
+    },
+    {
+      "epoch": 11.781437125748504,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 7870
+    },
+    {
+      "epoch": 11.782934131736527,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1378,
+      "step": 7871
+    },
+    {
+      "epoch": 11.784431137724551,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1486,
+      "step": 7872
+    },
+    {
+      "epoch": 11.785928143712574,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1469,
+      "step": 7873
+    },
+    {
+      "epoch": 11.7874251497006,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 7874
+    },
+    {
+      "epoch": 11.788922155688622,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 7875
+    },
+    {
+      "epoch": 11.790419161676647,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1471,
+      "step": 7876
+    },
+    {
+      "epoch": 11.79191616766467,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 7877
+    },
+    {
+      "epoch": 11.793413173652695,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 7878
+    },
+    {
+      "epoch": 11.794910179640718,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 7879
+    },
+    {
+      "epoch": 11.796407185628743,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1454,
+      "step": 7880
+    },
+    {
+      "epoch": 11.797904191616766,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 7881
+    },
+    {
+      "epoch": 11.79940119760479,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 7882
+    },
+    {
+      "epoch": 11.800898203592814,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1465,
+      "step": 7883
+    },
+    {
+      "epoch": 11.802395209580839,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1518,
+      "step": 7884
+    },
+    {
+      "epoch": 11.803892215568862,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1376,
+      "step": 7885
+    },
+    {
+      "epoch": 11.805389221556887,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 7886
+    },
+    {
+      "epoch": 11.80688622754491,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1376,
+      "step": 7887
+    },
+    {
+      "epoch": 11.808383233532934,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1508,
+      "step": 7888
+    },
+    {
+      "epoch": 11.809880239520957,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 7889
+    },
+    {
+      "epoch": 11.811377245508982,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 7890
+    },
+    {
+      "epoch": 11.812874251497005,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 7891
+    },
+    {
+      "epoch": 11.81437125748503,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 7892
+    },
+    {
+      "epoch": 11.815868263473053,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.147,
+      "step": 7893
+    },
+    {
+      "epoch": 11.817365269461078,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1366,
+      "step": 7894
+    },
+    {
+      "epoch": 11.818862275449101,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 7895
+    },
+    {
+      "epoch": 11.820359281437126,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 7896
+    },
+    {
+      "epoch": 11.821856287425149,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 7897
+    },
+    {
+      "epoch": 11.823353293413174,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1462,
+      "step": 7898
+    },
+    {
+      "epoch": 11.824850299401197,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1465,
+      "step": 7899
+    },
+    {
+      "epoch": 11.826347305389222,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 7900
+    },
+    {
+      "epoch": 11.827844311377245,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1393,
+      "step": 7901
+    },
+    {
+      "epoch": 11.82934131736527,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1384,
+      "step": 7902
+    },
+    {
+      "epoch": 11.830838323353294,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1494,
+      "step": 7903
+    },
+    {
+      "epoch": 11.832335329341317,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 7904
+    },
+    {
+      "epoch": 11.83383233532934,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 7905
+    },
+    {
+      "epoch": 11.835329341317365,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 7906
+    },
+    {
+      "epoch": 11.83682634730539,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1434,
+      "step": 7907
+    },
+    {
+      "epoch": 11.838323353293413,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 7908
+    },
+    {
+      "epoch": 11.839820359281438,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 7909
+    },
+    {
+      "epoch": 11.841317365269461,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.137,
+      "step": 7910
+    },
+    {
+      "epoch": 11.842814371257486,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 7911
+    },
+    {
+      "epoch": 11.844311377245509,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1481,
+      "step": 7912
+    },
+    {
+      "epoch": 11.845808383233534,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.146,
+      "step": 7913
+    },
+    {
+      "epoch": 11.847305389221557,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 7914
+    },
+    {
+      "epoch": 11.848802395209582,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 7915
+    },
+    {
+      "epoch": 11.850299401197605,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 7916
+    },
+    {
+      "epoch": 11.85179640718563,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 7917
+    },
+    {
+      "epoch": 11.853293413173652,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 7918
+    },
+    {
+      "epoch": 11.854790419161677,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 7919
+    },
+    {
+      "epoch": 11.8562874251497,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 7920
+    },
+    {
+      "epoch": 11.857784431137725,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 7921
+    },
+    {
+      "epoch": 11.859281437125748,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 7922
+    },
+    {
+      "epoch": 11.860778443113773,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1459,
+      "step": 7923
+    },
+    {
+      "epoch": 11.862275449101796,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1446,
+      "step": 7924
+    },
+    {
+      "epoch": 11.863772455089821,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1503,
+      "step": 7925
+    },
+    {
+      "epoch": 11.865269461077844,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 7926
+    },
+    {
+      "epoch": 11.866766467065869,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1505,
+      "step": 7927
+    },
+    {
+      "epoch": 11.868263473053892,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 7928
+    },
+    {
+      "epoch": 11.869760479041917,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 7929
+    },
+    {
+      "epoch": 11.87125748502994,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1465,
+      "step": 7930
+    },
+    {
+      "epoch": 11.872754491017965,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1379,
+      "step": 7931
+    },
+    {
+      "epoch": 11.874251497005988,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 7932
+    },
+    {
+      "epoch": 11.875748502994012,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1434,
+      "step": 7933
+    },
+    {
+      "epoch": 11.877245508982035,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 7934
+    },
+    {
+      "epoch": 11.87874251497006,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 7935
+    },
+    {
+      "epoch": 11.880239520958083,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1434,
+      "step": 7936
+    },
+    {
+      "epoch": 11.881736526946108,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 7937
+    },
+    {
+      "epoch": 11.883233532934131,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 7938
+    },
+    {
+      "epoch": 11.884730538922156,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 7939
+    },
+    {
+      "epoch": 11.886227544910179,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1443,
+      "step": 7940
+    },
+    {
+      "epoch": 11.887724550898204,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 7941
+    },
+    {
+      "epoch": 11.889221556886227,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 7942
+    },
+    {
+      "epoch": 11.890718562874252,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1337,
+      "step": 7943
+    },
+    {
+      "epoch": 11.892215568862275,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 7944
+    },
+    {
+      "epoch": 11.8937125748503,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 7945
+    },
+    {
+      "epoch": 11.895209580838323,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1413,
+      "step": 7946
+    },
+    {
+      "epoch": 11.896706586826348,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1434,
+      "step": 7947
+    },
+    {
+      "epoch": 11.89820359281437,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1369,
+      "step": 7948
+    },
+    {
+      "epoch": 11.899700598802395,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.144,
+      "step": 7949
+    },
+    {
+      "epoch": 11.901197604790418,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 7950
+    },
+    {
+      "epoch": 11.902694610778443,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 7951
+    },
+    {
+      "epoch": 11.904191616766466,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 7952
+    },
+    {
+      "epoch": 11.905688622754491,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1384,
+      "step": 7953
+    },
+    {
+      "epoch": 11.907185628742514,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1434,
+      "step": 7954
+    },
+    {
+      "epoch": 11.908682634730539,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 7955
+    },
+    {
+      "epoch": 11.910179640718562,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 7956
+    },
+    {
+      "epoch": 11.911676646706587,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 7957
+    },
+    {
+      "epoch": 11.91317365269461,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1446,
+      "step": 7958
+    },
+    {
+      "epoch": 11.914670658682635,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 7959
+    },
+    {
+      "epoch": 11.91616766467066,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 7960
+    },
+    {
+      "epoch": 11.917664670658683,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 7961
+    },
+    {
+      "epoch": 11.919161676646706,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 7962
+    },
+    {
+      "epoch": 11.92065868263473,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 7963
+    },
+    {
+      "epoch": 11.922155688622755,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1364,
+      "step": 7964
+    },
+    {
+      "epoch": 11.923652694610778,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1348,
+      "step": 7965
+    },
+    {
+      "epoch": 11.925149700598803,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 7966
+    },
+    {
+      "epoch": 11.926646706586826,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1372,
+      "step": 7967
+    },
+    {
+      "epoch": 11.928143712574851,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 7968
+    },
+    {
+      "epoch": 11.929640718562874,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1386,
+      "step": 7969
+    },
+    {
+      "epoch": 11.931137724550899,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 7970
+    },
+    {
+      "epoch": 11.932634730538922,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 7971
+    },
+    {
+      "epoch": 11.934131736526947,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1487,
+      "step": 7972
+    },
+    {
+      "epoch": 11.93562874251497,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 7973
+    },
+    {
+      "epoch": 11.937125748502995,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1472,
+      "step": 7974
+    },
+    {
+      "epoch": 11.938622754491018,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 7975
+    },
+    {
+      "epoch": 11.940119760479043,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 7976
+    },
+    {
+      "epoch": 11.941616766467066,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 7977
+    },
+    {
+      "epoch": 11.94311377245509,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1387,
+      "step": 7978
+    },
+    {
+      "epoch": 11.944610778443113,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1465,
+      "step": 7979
+    },
+    {
+      "epoch": 11.946107784431138,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.144,
+      "step": 7980
+    },
+    {
+      "epoch": 11.947604790419161,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1489,
+      "step": 7981
+    },
+    {
+      "epoch": 11.949101796407186,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 7982
+    },
+    {
+      "epoch": 11.95059880239521,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 7983
+    },
+    {
+      "epoch": 11.952095808383234,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1383,
+      "step": 7984
+    },
+    {
+      "epoch": 11.953592814371257,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 7985
+    },
+    {
+      "epoch": 11.955089820359282,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 7986
+    },
+    {
+      "epoch": 11.956586826347305,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1384,
+      "step": 7987
+    },
+    {
+      "epoch": 11.95808383233533,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1446,
+      "step": 7988
+    },
+    {
+      "epoch": 11.959580838323353,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 7989
+    },
+    {
+      "epoch": 11.961077844311378,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1357,
+      "step": 7990
+    },
+    {
+      "epoch": 11.9625748502994,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 7991
+    },
+    {
+      "epoch": 11.964071856287426,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1429,
+      "step": 7992
+    },
+    {
+      "epoch": 11.965568862275449,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1496,
+      "step": 7993
+    },
+    {
+      "epoch": 11.967065868263473,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 7994
+    },
+    {
+      "epoch": 11.968562874251496,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 7995
+    },
+    {
+      "epoch": 11.970059880239521,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1428,
+      "step": 7996
+    },
+    {
+      "epoch": 11.971556886227544,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1493,
+      "step": 7997
+    },
+    {
+      "epoch": 11.97305389221557,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1395,
+      "step": 7998
+    },
+    {
+      "epoch": 11.974550898203592,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 7999
+    },
+    {
+      "epoch": 11.976047904191617,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 8000
+    },
+    {
+      "epoch": 11.97754491017964,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 8001
+    },
+    {
+      "epoch": 11.979041916167665,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 8002
+    },
+    {
+      "epoch": 11.980538922155688,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 8003
+    },
+    {
+      "epoch": 11.982035928143713,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1366,
+      "step": 8004
+    },
+    {
+      "epoch": 11.983532934131736,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 8005
+    },
+    {
+      "epoch": 11.98502994011976,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 8006
+    },
+    {
+      "epoch": 11.986526946107784,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 8007
+    },
+    {
+      "epoch": 11.988023952095809,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 8008
+    },
+    {
+      "epoch": 11.989520958083832,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 8009
+    },
+    {
+      "epoch": 11.991017964071856,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1429,
+      "step": 8010
+    },
+    {
+      "epoch": 11.99251497005988,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 8011
+    },
+    {
+      "epoch": 11.994011976047904,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1457,
+      "step": 8012
+    },
+    {
+      "epoch": 11.995508982035929,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 8013
+    },
+    {
+      "epoch": 11.997005988023952,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1387,
+      "step": 8014
+    },
+    {
+      "epoch": 11.998502994011975,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 8015
+    },
+    {
+      "epoch": 12.0,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 8016
+    },
+    {
+      "epoch": 12.001497005988025,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 8017
+    },
+    {
+      "epoch": 12.002994011976048,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 8018
+    },
+    {
+      "epoch": 12.004491017964073,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1443,
+      "step": 8019
+    },
+    {
+      "epoch": 12.005988023952096,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 8020
+    },
+    {
+      "epoch": 12.00748502994012,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.137,
+      "step": 8021
+    },
+    {
+      "epoch": 12.008982035928144,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 8022
+    },
+    {
+      "epoch": 12.010479041916168,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1493,
+      "step": 8023
+    },
+    {
+      "epoch": 12.011976047904191,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 8024
+    },
+    {
+      "epoch": 12.013473053892216,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1362,
+      "step": 8025
+    },
+    {
+      "epoch": 12.01497005988024,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 8026
+    },
+    {
+      "epoch": 12.016467065868264,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1384,
+      "step": 8027
+    },
+    {
+      "epoch": 12.017964071856287,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 8028
+    },
+    {
+      "epoch": 12.019461077844312,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 8029
+    },
+    {
+      "epoch": 12.020958083832335,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 8030
+    },
+    {
+      "epoch": 12.02245508982036,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 8031
+    },
+    {
+      "epoch": 12.023952095808383,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1446,
+      "step": 8032
+    },
+    {
+      "epoch": 12.025449101796408,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 8033
+    },
+    {
+      "epoch": 12.02694610778443,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 8034
+    },
+    {
+      "epoch": 12.028443113772456,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 8035
+    },
+    {
+      "epoch": 12.029940119760479,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 8036
+    },
+    {
+      "epoch": 12.031437125748504,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 8037
+    },
+    {
+      "epoch": 12.032934131736527,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 8038
+    },
+    {
+      "epoch": 12.034431137724551,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1352,
+      "step": 8039
+    },
+    {
+      "epoch": 12.035928143712574,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 8040
+    },
+    {
+      "epoch": 12.0374251497006,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 8041
+    },
+    {
+      "epoch": 12.038922155688622,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 8042
+    },
+    {
+      "epoch": 12.040419161676647,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 8043
+    },
+    {
+      "epoch": 12.04191616766467,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 8044
+    },
+    {
+      "epoch": 12.043413173652695,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 8045
+    },
+    {
+      "epoch": 12.044910179640718,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.136,
+      "step": 8046
+    },
+    {
+      "epoch": 12.046407185628743,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 8047
+    },
+    {
+      "epoch": 12.047904191616766,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 8048
+    },
+    {
+      "epoch": 12.04940119760479,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 8049
+    },
+    {
+      "epoch": 12.050898203592814,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 8050
+    },
+    {
+      "epoch": 12.052395209580839,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1498,
+      "step": 8051
+    },
+    {
+      "epoch": 12.053892215568862,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 8052
+    },
+    {
+      "epoch": 12.055389221556887,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1463,
+      "step": 8053
+    },
+    {
+      "epoch": 12.05688622754491,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 8054
+    },
+    {
+      "epoch": 12.058383233532934,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 8055
+    },
+    {
+      "epoch": 12.059880239520957,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 8056
+    },
+    {
+      "epoch": 12.061377245508982,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 8057
+    },
+    {
+      "epoch": 12.062874251497005,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 8058
+    },
+    {
+      "epoch": 12.06437125748503,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 8059
+    },
+    {
+      "epoch": 12.065868263473053,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 8060
+    },
+    {
+      "epoch": 12.067365269461078,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 8061
+    },
+    {
+      "epoch": 12.068862275449101,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1507,
+      "step": 8062
+    },
+    {
+      "epoch": 12.070359281437126,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 8063
+    },
+    {
+      "epoch": 12.071856287425149,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 8064
+    },
+    {
+      "epoch": 12.073353293413174,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 8065
+    },
+    {
+      "epoch": 12.074850299401197,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1347,
+      "step": 8066
+    },
+    {
+      "epoch": 12.076347305389222,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1351,
+      "step": 8067
+    },
+    {
+      "epoch": 12.077844311377245,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 8068
+    },
+    {
+      "epoch": 12.07934131736527,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1383,
+      "step": 8069
+    },
+    {
+      "epoch": 12.080838323353293,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1462,
+      "step": 8070
+    },
+    {
+      "epoch": 12.082335329341317,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1376,
+      "step": 8071
+    },
+    {
+      "epoch": 12.08383233532934,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 8072
+    },
+    {
+      "epoch": 12.085329341317365,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 8073
+    },
+    {
+      "epoch": 12.08682634730539,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 8074
+    },
+    {
+      "epoch": 12.088323353293413,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 8075
+    },
+    {
+      "epoch": 12.089820359281438,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1387,
+      "step": 8076
+    },
+    {
+      "epoch": 12.091317365269461,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1384,
+      "step": 8077
+    },
+    {
+      "epoch": 12.092814371257486,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 8078
+    },
+    {
+      "epoch": 12.094311377245509,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1387,
+      "step": 8079
+    },
+    {
+      "epoch": 12.095808383233534,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 8080
+    },
+    {
+      "epoch": 12.097305389221557,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1428,
+      "step": 8081
+    },
+    {
+      "epoch": 12.098802395209582,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 8082
+    },
+    {
+      "epoch": 12.100299401197605,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1368,
+      "step": 8083
+    },
+    {
+      "epoch": 12.10179640718563,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 8084
+    },
+    {
+      "epoch": 12.103293413173652,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 8085
+    },
+    {
+      "epoch": 12.104790419161677,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1483,
+      "step": 8086
+    },
+    {
+      "epoch": 12.1062874251497,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 8087
+    },
+    {
+      "epoch": 12.107784431137725,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 8088
+    },
+    {
+      "epoch": 12.109281437125748,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1429,
+      "step": 8089
+    },
+    {
+      "epoch": 12.110778443113773,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 8090
+    },
+    {
+      "epoch": 12.112275449101796,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 8091
+    },
+    {
+      "epoch": 12.113772455089821,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 8092
+    },
+    {
+      "epoch": 12.115269461077844,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1392,
+      "step": 8093
+    },
+    {
+      "epoch": 12.116766467065869,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 8094
+    },
+    {
+      "epoch": 12.118263473053892,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 8095
+    },
+    {
+      "epoch": 12.119760479041917,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 8096
+    },
+    {
+      "epoch": 12.12125748502994,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1395,
+      "step": 8097
+    },
+    {
+      "epoch": 12.122754491017965,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 8098
+    },
+    {
+      "epoch": 12.124251497005988,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 8099
+    },
+    {
+      "epoch": 12.125748502994012,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1386,
+      "step": 8100
+    },
+    {
+      "epoch": 12.127245508982035,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1443,
+      "step": 8101
+    },
+    {
+      "epoch": 12.12874251497006,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1497,
+      "step": 8102
+    },
+    {
+      "epoch": 12.130239520958083,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 8103
+    },
+    {
+      "epoch": 12.131736526946108,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1475,
+      "step": 8104
+    },
+    {
+      "epoch": 12.133233532934131,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1443,
+      "step": 8105
+    },
+    {
+      "epoch": 12.134730538922156,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1465,
+      "step": 8106
+    },
+    {
+      "epoch": 12.136227544910179,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.146,
+      "step": 8107
+    },
+    {
+      "epoch": 12.137724550898204,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1372,
+      "step": 8108
+    },
+    {
+      "epoch": 12.139221556886227,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1381,
+      "step": 8109
+    },
+    {
+      "epoch": 12.140718562874252,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 8110
+    },
+    {
+      "epoch": 12.142215568862275,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 8111
+    },
+    {
+      "epoch": 12.1437125748503,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 8112
+    },
+    {
+      "epoch": 12.145209580838323,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 8113
+    },
+    {
+      "epoch": 12.146706586826348,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 8114
+    },
+    {
+      "epoch": 12.14820359281437,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 8115
+    },
+    {
+      "epoch": 12.149700598802395,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1361,
+      "step": 8116
+    },
+    {
+      "epoch": 12.151197604790418,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 8117
+    },
+    {
+      "epoch": 12.152694610778443,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 8118
+    },
+    {
+      "epoch": 12.154191616766466,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.144,
+      "step": 8119
+    },
+    {
+      "epoch": 12.155688622754491,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1462,
+      "step": 8120
+    },
+    {
+      "epoch": 12.157185628742514,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 8121
+    },
+    {
+      "epoch": 12.158682634730539,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 8122
+    },
+    {
+      "epoch": 12.160179640718562,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1378,
+      "step": 8123
+    },
+    {
+      "epoch": 12.161676646706587,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 8124
+    },
+    {
+      "epoch": 12.16317365269461,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 8125
+    },
+    {
+      "epoch": 12.164670658682635,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 8126
+    },
+    {
+      "epoch": 12.16616766467066,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1395,
+      "step": 8127
+    },
+    {
+      "epoch": 12.167664670658683,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1486,
+      "step": 8128
+    },
+    {
+      "epoch": 12.169161676646707,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 8129
+    },
+    {
+      "epoch": 12.17065868263473,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 8130
+    },
+    {
+      "epoch": 12.172155688622755,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1384,
+      "step": 8131
+    },
+    {
+      "epoch": 12.173652694610778,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1481,
+      "step": 8132
+    },
+    {
+      "epoch": 12.175149700598803,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 8133
+    },
+    {
+      "epoch": 12.176646706586826,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 8134
+    },
+    {
+      "epoch": 12.178143712574851,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 8135
+    },
+    {
+      "epoch": 12.179640718562874,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 8136
+    },
+    {
+      "epoch": 12.181137724550899,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 8137
+    },
+    {
+      "epoch": 12.182634730538922,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 8138
+    },
+    {
+      "epoch": 12.184131736526947,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1371,
+      "step": 8139
+    },
+    {
+      "epoch": 12.18562874251497,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1429,
+      "step": 8140
+    },
+    {
+      "epoch": 12.187125748502995,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 8141
+    },
+    {
+      "epoch": 12.188622754491018,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0008,
+      "loss": 1.134,
+      "step": 8142
+    },
+    {
+      "epoch": 12.190119760479043,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1459,
+      "step": 8143
+    },
+    {
+      "epoch": 12.191616766467066,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 8144
+    },
+    {
+      "epoch": 12.19311377245509,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1468,
+      "step": 8145
+    },
+    {
+      "epoch": 12.194610778443113,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1395,
+      "step": 8146
+    },
+    {
+      "epoch": 12.196107784431138,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1384,
+      "step": 8147
+    },
+    {
+      "epoch": 12.197604790419161,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 8148
+    },
+    {
+      "epoch": 12.199101796407186,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1364,
+      "step": 8149
+    },
+    {
+      "epoch": 12.20059880239521,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 8150
+    },
+    {
+      "epoch": 12.202095808383234,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 8151
+    },
+    {
+      "epoch": 12.203592814371257,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 8152
+    },
+    {
+      "epoch": 12.205089820359282,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.137,
+      "step": 8153
+    },
+    {
+      "epoch": 12.206586826347305,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1477,
+      "step": 8154
+    },
+    {
+      "epoch": 12.20808383233533,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1371,
+      "step": 8155
+    },
+    {
+      "epoch": 12.209580838323353,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.144,
+      "step": 8156
+    },
+    {
+      "epoch": 12.211077844311378,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 8157
+    },
+    {
+      "epoch": 12.2125748502994,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 8158
+    },
+    {
+      "epoch": 12.214071856287426,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 8159
+    },
+    {
+      "epoch": 12.215568862275449,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 8160
+    },
+    {
+      "epoch": 12.217065868263473,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1416,
+      "step": 8161
+    },
+    {
+      "epoch": 12.218562874251496,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1323,
+      "step": 8162
+    },
+    {
+      "epoch": 12.220059880239521,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.136,
+      "step": 8163
+    },
+    {
+      "epoch": 12.221556886227544,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 8164
+    },
+    {
+      "epoch": 12.22305389221557,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 8165
+    },
+    {
+      "epoch": 12.224550898203592,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1379,
+      "step": 8166
+    },
+    {
+      "epoch": 12.226047904191617,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 8167
+    },
+    {
+      "epoch": 12.22754491017964,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1351,
+      "step": 8168
+    },
+    {
+      "epoch": 12.229041916167665,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.144,
+      "step": 8169
+    },
+    {
+      "epoch": 12.230538922155688,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 8170
+    },
+    {
+      "epoch": 12.232035928143713,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 8171
+    },
+    {
+      "epoch": 12.233532934131736,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 8172
+    },
+    {
+      "epoch": 12.23502994011976,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 8173
+    },
+    {
+      "epoch": 12.236526946107784,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1459,
+      "step": 8174
+    },
+    {
+      "epoch": 12.238023952095809,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 8175
+    },
+    {
+      "epoch": 12.239520958083832,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1383,
+      "step": 8176
+    },
+    {
+      "epoch": 12.241017964071856,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 8177
+    },
+    {
+      "epoch": 12.24251497005988,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 8178
+    },
+    {
+      "epoch": 12.244011976047904,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 8179
+    },
+    {
+      "epoch": 12.245508982035927,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.14,
+      "step": 8180
+    },
+    {
+      "epoch": 12.247005988023952,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 8181
+    },
+    {
+      "epoch": 12.248502994011975,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 8182
+    },
+    {
+      "epoch": 12.25,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 8183
+    },
+    {
+      "epoch": 12.251497005988025,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 8184
+    },
+    {
+      "epoch": 12.252994011976048,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.14,
+      "step": 8185
+    },
+    {
+      "epoch": 12.254491017964073,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 8186
+    },
+    {
+      "epoch": 12.255988023952096,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 8187
+    },
+    {
+      "epoch": 12.25748502994012,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1413,
+      "step": 8188
+    },
+    {
+      "epoch": 12.258982035928144,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1465,
+      "step": 8189
+    },
+    {
+      "epoch": 12.260479041916168,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1454,
+      "step": 8190
+    },
+    {
+      "epoch": 12.261976047904191,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 8191
+    },
+    {
+      "epoch": 12.263473053892216,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1339,
+      "step": 8192
+    },
+    {
+      "epoch": 12.26497005988024,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1465,
+      "step": 8193
+    },
+    {
+      "epoch": 12.266467065868264,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1371,
+      "step": 8194
+    },
+    {
+      "epoch": 12.267964071856287,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1386,
+      "step": 8195
+    },
+    {
+      "epoch": 12.269461077844312,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 8196
+    },
+    {
+      "epoch": 12.270958083832335,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0008,
+      "loss": 1.134,
+      "step": 8197
+    },
+    {
+      "epoch": 12.27245508982036,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.147,
+      "step": 8198
+    },
+    {
+      "epoch": 12.273952095808383,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 8199
+    },
+    {
+      "epoch": 12.275449101796408,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 8200
+    },
+    {
+      "epoch": 12.27694610778443,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 8201
+    },
+    {
+      "epoch": 12.278443113772456,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 8202
+    },
+    {
+      "epoch": 12.279940119760479,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1375,
+      "step": 8203
+    },
+    {
+      "epoch": 12.281437125748504,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1371,
+      "step": 8204
+    },
+    {
+      "epoch": 12.282934131736527,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 8205
+    },
+    {
+      "epoch": 12.284431137724551,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1361,
+      "step": 8206
+    },
+    {
+      "epoch": 12.285928143712574,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 8207
+    },
+    {
+      "epoch": 12.2874251497006,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1392,
+      "step": 8208
+    },
+    {
+      "epoch": 12.288922155688622,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 8209
+    },
+    {
+      "epoch": 12.290419161676647,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.14,
+      "step": 8210
+    },
+    {
+      "epoch": 12.29191616766467,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 8211
+    },
+    {
+      "epoch": 12.293413173652695,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 8212
+    },
+    {
+      "epoch": 12.294910179640718,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 8213
+    },
+    {
+      "epoch": 12.296407185628743,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 8214
+    },
+    {
+      "epoch": 12.297904191616766,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 8215
+    },
+    {
+      "epoch": 12.29940119760479,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1471,
+      "step": 8216
+    },
+    {
+      "epoch": 12.300898203592814,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 8217
+    },
+    {
+      "epoch": 12.302395209580839,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 8218
+    },
+    {
+      "epoch": 12.303892215568862,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1498,
+      "step": 8219
+    },
+    {
+      "epoch": 12.305389221556887,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1372,
+      "step": 8220
+    },
+    {
+      "epoch": 12.30688622754491,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1341,
+      "step": 8221
+    },
+    {
+      "epoch": 12.308383233532934,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1323,
+      "step": 8222
+    },
+    {
+      "epoch": 12.309880239520957,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1369,
+      "step": 8223
+    },
+    {
+      "epoch": 12.311377245508982,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 8224
+    },
+    {
+      "epoch": 12.312874251497005,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 8225
+    },
+    {
+      "epoch": 12.31437125748503,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 8226
+    },
+    {
+      "epoch": 12.315868263473053,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 8227
+    },
+    {
+      "epoch": 12.317365269461078,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 8228
+    },
+    {
+      "epoch": 12.318862275449101,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1381,
+      "step": 8229
+    },
+    {
+      "epoch": 12.320359281437126,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1364,
+      "step": 8230
+    },
+    {
+      "epoch": 12.321856287425149,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 8231
+    },
+    {
+      "epoch": 12.323353293413174,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1365,
+      "step": 8232
+    },
+    {
+      "epoch": 12.324850299401197,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 8233
+    },
+    {
+      "epoch": 12.326347305389222,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 8234
+    },
+    {
+      "epoch": 12.327844311377245,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1379,
+      "step": 8235
+    },
+    {
+      "epoch": 12.32934131736527,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1392,
+      "step": 8236
+    },
+    {
+      "epoch": 12.330838323353294,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1486,
+      "step": 8237
+    },
+    {
+      "epoch": 12.332335329341317,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 8238
+    },
+    {
+      "epoch": 12.33383233532934,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1375,
+      "step": 8239
+    },
+    {
+      "epoch": 12.335329341317365,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 8240
+    },
+    {
+      "epoch": 12.33682634730539,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 8241
+    },
+    {
+      "epoch": 12.338323353293413,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 8242
+    },
+    {
+      "epoch": 12.339820359281438,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 8243
+    },
+    {
+      "epoch": 12.341317365269461,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 8244
+    },
+    {
+      "epoch": 12.342814371257486,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 8245
+    },
+    {
+      "epoch": 12.344311377245509,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1383,
+      "step": 8246
+    },
+    {
+      "epoch": 12.345808383233534,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 8247
+    },
+    {
+      "epoch": 12.347305389221557,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1459,
+      "step": 8248
+    },
+    {
+      "epoch": 12.348802395209582,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 8249
+    },
+    {
+      "epoch": 12.350299401197605,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1379,
+      "step": 8250
+    },
+    {
+      "epoch": 12.35179640718563,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 8251
+    },
+    {
+      "epoch": 12.353293413173652,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1387,
+      "step": 8252
+    },
+    {
+      "epoch": 12.354790419161677,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 8253
+    },
+    {
+      "epoch": 12.3562874251497,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 8254
+    },
+    {
+      "epoch": 12.357784431137725,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 8255
+    },
+    {
+      "epoch": 12.359281437125748,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 8256
+    },
+    {
+      "epoch": 12.360778443113773,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 8257
+    },
+    {
+      "epoch": 12.362275449101796,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1496,
+      "step": 8258
+    },
+    {
+      "epoch": 12.363772455089821,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1383,
+      "step": 8259
+    },
+    {
+      "epoch": 12.365269461077844,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 8260
+    },
+    {
+      "epoch": 12.366766467065869,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 8261
+    },
+    {
+      "epoch": 12.368263473053892,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1375,
+      "step": 8262
+    },
+    {
+      "epoch": 12.369760479041917,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.136,
+      "step": 8263
+    },
+    {
+      "epoch": 12.37125748502994,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 8264
+    },
+    {
+      "epoch": 12.372754491017965,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1358,
+      "step": 8265
+    },
+    {
+      "epoch": 12.374251497005988,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1351,
+      "step": 8266
+    },
+    {
+      "epoch": 12.375748502994012,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 8267
+    },
+    {
+      "epoch": 12.377245508982035,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 8268
+    },
+    {
+      "epoch": 12.37874251497006,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 8269
+    },
+    {
+      "epoch": 12.380239520958083,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 8270
+    },
+    {
+      "epoch": 12.381736526946108,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 8271
+    },
+    {
+      "epoch": 12.383233532934131,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 8272
+    },
+    {
+      "epoch": 12.384730538922156,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1459,
+      "step": 8273
+    },
+    {
+      "epoch": 12.386227544910179,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 8274
+    },
+    {
+      "epoch": 12.387724550898204,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.144,
+      "step": 8275
+    },
+    {
+      "epoch": 12.389221556886227,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 8276
+    },
+    {
+      "epoch": 12.390718562874252,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 8277
+    },
+    {
+      "epoch": 12.392215568862275,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 8278
+    },
+    {
+      "epoch": 12.3937125748503,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1457,
+      "step": 8279
+    },
+    {
+      "epoch": 12.395209580838323,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 8280
+    },
+    {
+      "epoch": 12.396706586826348,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1491,
+      "step": 8281
+    },
+    {
+      "epoch": 12.39820359281437,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 8282
+    },
+    {
+      "epoch": 12.399700598802395,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1492,
+      "step": 8283
+    },
+    {
+      "epoch": 12.401197604790418,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 8284
+    },
+    {
+      "epoch": 12.402694610778443,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1387,
+      "step": 8285
+    },
+    {
+      "epoch": 12.404191616766466,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 8286
+    },
+    {
+      "epoch": 12.405688622754491,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1372,
+      "step": 8287
+    },
+    {
+      "epoch": 12.407185628742514,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 8288
+    },
+    {
+      "epoch": 12.408682634730539,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 8289
+    },
+    {
+      "epoch": 12.410179640718562,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 8290
+    },
+    {
+      "epoch": 12.411676646706587,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 8291
+    },
+    {
+      "epoch": 12.41317365269461,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1369,
+      "step": 8292
+    },
+    {
+      "epoch": 12.414670658682635,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1372,
+      "step": 8293
+    },
+    {
+      "epoch": 12.41616766467066,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 8294
+    },
+    {
+      "epoch": 12.417664670658683,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 8295
+    },
+    {
+      "epoch": 12.419161676646706,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1499,
+      "step": 8296
+    },
+    {
+      "epoch": 12.42065868263473,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 8297
+    },
+    {
+      "epoch": 12.422155688622755,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1434,
+      "step": 8298
+    },
+    {
+      "epoch": 12.423652694610778,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 8299
+    },
+    {
+      "epoch": 12.425149700598803,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 8300
+    },
+    {
+      "epoch": 12.426646706586826,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 8301
+    },
+    {
+      "epoch": 12.428143712574851,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 8302
+    },
+    {
+      "epoch": 12.429640718562874,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 8303
+    },
+    {
+      "epoch": 12.431137724550899,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 8304
+    },
+    {
+      "epoch": 12.432634730538922,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 8305
+    },
+    {
+      "epoch": 12.434131736526947,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 8306
+    },
+    {
+      "epoch": 12.43562874251497,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 8307
+    },
+    {
+      "epoch": 12.437125748502995,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 8308
+    },
+    {
+      "epoch": 12.438622754491018,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1351,
+      "step": 8309
+    },
+    {
+      "epoch": 12.440119760479043,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1507,
+      "step": 8310
+    },
+    {
+      "epoch": 12.441616766467066,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 8311
+    },
+    {
+      "epoch": 12.44311377245509,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 8312
+    },
+    {
+      "epoch": 12.444610778443113,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 8313
+    },
+    {
+      "epoch": 12.446107784431138,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 8314
+    },
+    {
+      "epoch": 12.447604790419161,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1368,
+      "step": 8315
+    },
+    {
+      "epoch": 12.449101796407186,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 8316
+    },
+    {
+      "epoch": 12.45059880239521,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 8317
+    },
+    {
+      "epoch": 12.452095808383234,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 8318
+    },
+    {
+      "epoch": 12.453592814371257,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1372,
+      "step": 8319
+    },
+    {
+      "epoch": 12.455089820359282,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 8320
+    },
+    {
+      "epoch": 12.456586826347305,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 8321
+    },
+    {
+      "epoch": 12.45808383233533,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1352,
+      "step": 8322
+    },
+    {
+      "epoch": 12.459580838323353,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1392,
+      "step": 8323
+    },
+    {
+      "epoch": 12.461077844311378,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 8324
+    },
+    {
+      "epoch": 12.4625748502994,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 8325
+    },
+    {
+      "epoch": 12.464071856287426,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 8326
+    },
+    {
+      "epoch": 12.465568862275449,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1371,
+      "step": 8327
+    },
+    {
+      "epoch": 12.467065868263473,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 8328
+    },
+    {
+      "epoch": 12.468562874251496,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 8329
+    },
+    {
+      "epoch": 12.470059880239521,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 8330
+    },
+    {
+      "epoch": 12.471556886227544,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 8331
+    },
+    {
+      "epoch": 12.47305389221557,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 8332
+    },
+    {
+      "epoch": 12.474550898203592,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 8333
+    },
+    {
+      "epoch": 12.476047904191617,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 8334
+    },
+    {
+      "epoch": 12.47754491017964,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1379,
+      "step": 8335
+    },
+    {
+      "epoch": 12.479041916167665,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 8336
+    },
+    {
+      "epoch": 12.480538922155688,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 8337
+    },
+    {
+      "epoch": 12.482035928143713,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 8338
+    },
+    {
+      "epoch": 12.483532934131736,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 8339
+    },
+    {
+      "epoch": 12.48502994011976,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 8340
+    },
+    {
+      "epoch": 12.486526946107784,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 8341
+    },
+    {
+      "epoch": 12.488023952095809,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1377,
+      "step": 8342
+    },
+    {
+      "epoch": 12.489520958083832,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1337,
+      "step": 8343
+    },
+    {
+      "epoch": 12.491017964071856,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1413,
+      "step": 8344
+    },
+    {
+      "epoch": 12.49251497005988,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1463,
+      "step": 8345
+    },
+    {
+      "epoch": 12.494011976047904,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1472,
+      "step": 8346
+    },
+    {
+      "epoch": 12.495508982035927,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 8347
+    },
+    {
+      "epoch": 12.497005988023952,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 8348
+    },
+    {
+      "epoch": 12.498502994011975,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1434,
+      "step": 8349
+    },
+    {
+      "epoch": 12.5,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 8350
+    },
+    {
+      "epoch": 12.501497005988025,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 8351
+    },
+    {
+      "epoch": 12.502994011976048,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1392,
+      "step": 8352
+    },
+    {
+      "epoch": 12.504491017964071,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 8353
+    },
+    {
+      "epoch": 12.505988023952096,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1475,
+      "step": 8354
+    },
+    {
+      "epoch": 12.50748502994012,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 8355
+    },
+    {
+      "epoch": 12.508982035928144,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1454,
+      "step": 8356
+    },
+    {
+      "epoch": 12.510479041916168,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 8357
+    },
+    {
+      "epoch": 12.511976047904191,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1352,
+      "step": 8358
+    },
+    {
+      "epoch": 12.513473053892216,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 8359
+    },
+    {
+      "epoch": 12.51497005988024,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 8360
+    },
+    {
+      "epoch": 12.516467065868264,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 8361
+    },
+    {
+      "epoch": 12.517964071856287,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 8362
+    },
+    {
+      "epoch": 12.519461077844312,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 8363
+    },
+    {
+      "epoch": 12.520958083832335,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 8364
+    },
+    {
+      "epoch": 12.52245508982036,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 8365
+    },
+    {
+      "epoch": 12.523952095808383,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1527,
+      "step": 8366
+    },
+    {
+      "epoch": 12.525449101796408,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 8367
+    },
+    {
+      "epoch": 12.52694610778443,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1341,
+      "step": 8368
+    },
+    {
+      "epoch": 12.528443113772456,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1481,
+      "step": 8369
+    },
+    {
+      "epoch": 12.529940119760479,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 8370
+    },
+    {
+      "epoch": 12.531437125748504,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 8371
+    },
+    {
+      "epoch": 12.532934131736527,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 8372
+    },
+    {
+      "epoch": 12.534431137724551,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 8373
+    },
+    {
+      "epoch": 12.535928143712574,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 8374
+    },
+    {
+      "epoch": 12.5374251497006,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 8375
+    },
+    {
+      "epoch": 12.538922155688622,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1395,
+      "step": 8376
+    },
+    {
+      "epoch": 12.540419161676647,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 8377
+    },
+    {
+      "epoch": 12.54191616766467,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 8378
+    },
+    {
+      "epoch": 12.543413173652695,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 8379
+    },
+    {
+      "epoch": 12.544910179640718,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.144,
+      "step": 8380
+    },
+    {
+      "epoch": 12.546407185628743,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 8381
+    },
+    {
+      "epoch": 12.547904191616766,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0008,
+      "loss": 1.14,
+      "step": 8382
+    },
+    {
+      "epoch": 12.54940119760479,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 8383
+    },
+    {
+      "epoch": 12.550898203592814,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1361,
+      "step": 8384
+    },
+    {
+      "epoch": 12.552395209580839,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.135,
+      "step": 8385
+    },
+    {
+      "epoch": 12.553892215568862,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 8386
+    },
+    {
+      "epoch": 12.555389221556887,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1476,
+      "step": 8387
+    },
+    {
+      "epoch": 12.55688622754491,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 8388
+    },
+    {
+      "epoch": 12.558383233532934,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1294,
+      "step": 8389
+    },
+    {
+      "epoch": 12.559880239520957,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 8390
+    },
+    {
+      "epoch": 12.561377245508982,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 8391
+    },
+    {
+      "epoch": 12.562874251497005,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 8392
+    },
+    {
+      "epoch": 12.56437125748503,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 8393
+    },
+    {
+      "epoch": 12.565868263473053,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1375,
+      "step": 8394
+    },
+    {
+      "epoch": 12.567365269461078,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 8395
+    },
+    {
+      "epoch": 12.568862275449101,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1468,
+      "step": 8396
+    },
+    {
+      "epoch": 12.570359281437126,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1454,
+      "step": 8397
+    },
+    {
+      "epoch": 12.571856287425149,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 8398
+    },
+    {
+      "epoch": 12.573353293413174,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 8399
+    },
+    {
+      "epoch": 12.574850299401197,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 8400
+    },
+    {
+      "epoch": 12.576347305389222,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 8401
+    },
+    {
+      "epoch": 12.577844311377245,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 8402
+    },
+    {
+      "epoch": 12.57934131736527,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0008,
+      "loss": 1.146,
+      "step": 8403
+    },
+    {
+      "epoch": 12.580838323353294,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1471,
+      "step": 8404
+    },
+    {
+      "epoch": 12.582335329341317,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1366,
+      "step": 8405
+    },
+    {
+      "epoch": 12.58383233532934,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 8406
+    },
+    {
+      "epoch": 12.585329341317365,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 8407
+    },
+    {
+      "epoch": 12.58682634730539,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1509,
+      "step": 8408
+    },
+    {
+      "epoch": 12.588323353293413,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 8409
+    },
+    {
+      "epoch": 12.589820359281438,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 8410
+    },
+    {
+      "epoch": 12.591317365269461,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1434,
+      "step": 8411
+    },
+    {
+      "epoch": 12.592814371257486,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 8412
+    },
+    {
+      "epoch": 12.594311377245509,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 8413
+    },
+    {
+      "epoch": 12.595808383233534,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 8414
+    },
+    {
+      "epoch": 12.597305389221557,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 8415
+    },
+    {
+      "epoch": 12.598802395209582,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 8416
+    },
+    {
+      "epoch": 12.600299401197605,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1428,
+      "step": 8417
+    },
+    {
+      "epoch": 12.60179640718563,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 8418
+    },
+    {
+      "epoch": 12.603293413173652,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 8419
+    },
+    {
+      "epoch": 12.604790419161677,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 8420
+    },
+    {
+      "epoch": 12.6062874251497,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 8421
+    },
+    {
+      "epoch": 12.607784431137725,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1454,
+      "step": 8422
+    },
+    {
+      "epoch": 12.609281437125748,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 8423
+    },
+    {
+      "epoch": 12.610778443113773,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1463,
+      "step": 8424
+    },
+    {
+      "epoch": 12.612275449101796,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1443,
+      "step": 8425
+    },
+    {
+      "epoch": 12.613772455089821,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 8426
+    },
+    {
+      "epoch": 12.615269461077844,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 8427
+    },
+    {
+      "epoch": 12.616766467065869,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 8428
+    },
+    {
+      "epoch": 12.618263473053892,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1413,
+      "step": 8429
+    },
+    {
+      "epoch": 12.619760479041917,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1446,
+      "step": 8430
+    },
+    {
+      "epoch": 12.62125748502994,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 8431
+    },
+    {
+      "epoch": 12.622754491017965,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.144,
+      "step": 8432
+    },
+    {
+      "epoch": 12.624251497005988,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1363,
+      "step": 8433
+    },
+    {
+      "epoch": 12.625748502994012,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1362,
+      "step": 8434
+    },
+    {
+      "epoch": 12.627245508982035,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 8435
+    },
+    {
+      "epoch": 12.62874251497006,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1376,
+      "step": 8436
+    },
+    {
+      "epoch": 12.630239520958083,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1344,
+      "step": 8437
+    },
+    {
+      "epoch": 12.631736526946108,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1365,
+      "step": 8438
+    },
+    {
+      "epoch": 12.633233532934131,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 8439
+    },
+    {
+      "epoch": 12.634730538922156,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1356,
+      "step": 8440
+    },
+    {
+      "epoch": 12.636227544910179,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 8441
+    },
+    {
+      "epoch": 12.637724550898204,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 8442
+    },
+    {
+      "epoch": 12.639221556886227,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1395,
+      "step": 8443
+    },
+    {
+      "epoch": 12.640718562874252,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.134,
+      "step": 8444
+    },
+    {
+      "epoch": 12.642215568862275,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 8445
+    },
+    {
+      "epoch": 12.6437125748503,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1367,
+      "step": 8446
+    },
+    {
+      "epoch": 12.645209580838323,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 8447
+    },
+    {
+      "epoch": 12.646706586826348,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 8448
+    },
+    {
+      "epoch": 12.64820359281437,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.134,
+      "step": 8449
+    },
+    {
+      "epoch": 12.649700598802395,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 8450
+    },
+    {
+      "epoch": 12.651197604790418,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 8451
+    },
+    {
+      "epoch": 12.652694610778443,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1363,
+      "step": 8452
+    },
+    {
+      "epoch": 12.654191616766466,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 8453
+    },
+    {
+      "epoch": 12.655688622754491,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 8454
+    },
+    {
+      "epoch": 12.657185628742514,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1386,
+      "step": 8455
+    },
+    {
+      "epoch": 12.658682634730539,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 8456
+    },
+    {
+      "epoch": 12.660179640718562,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 8457
+    },
+    {
+      "epoch": 12.661676646706587,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 8458
+    },
+    {
+      "epoch": 12.66317365269461,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 8459
+    },
+    {
+      "epoch": 12.664670658682635,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1446,
+      "step": 8460
+    },
+    {
+      "epoch": 12.66616766467066,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 8461
+    },
+    {
+      "epoch": 12.667664670658683,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 8462
+    },
+    {
+      "epoch": 12.669161676646706,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 8463
+    },
+    {
+      "epoch": 12.67065868263473,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1301,
+      "step": 8464
+    },
+    {
+      "epoch": 12.672155688622755,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 8465
+    },
+    {
+      "epoch": 12.673652694610778,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1428,
+      "step": 8466
+    },
+    {
+      "epoch": 12.675149700598803,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1351,
+      "step": 8467
+    },
+    {
+      "epoch": 12.676646706586826,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1416,
+      "step": 8468
+    },
+    {
+      "epoch": 12.678143712574851,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1381,
+      "step": 8469
+    },
+    {
+      "epoch": 12.679640718562874,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1475,
+      "step": 8470
+    },
+    {
+      "epoch": 12.681137724550899,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1373,
+      "step": 8471
+    },
+    {
+      "epoch": 12.682634730538922,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1395,
+      "step": 8472
+    },
+    {
+      "epoch": 12.684131736526947,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1393,
+      "step": 8473
+    },
+    {
+      "epoch": 12.68562874251497,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 8474
+    },
+    {
+      "epoch": 12.687125748502995,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 8475
+    },
+    {
+      "epoch": 12.688622754491018,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 8476
+    },
+    {
+      "epoch": 12.690119760479043,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 8477
+    },
+    {
+      "epoch": 12.691616766467066,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.14,
+      "step": 8478
+    },
+    {
+      "epoch": 12.69311377245509,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1327,
+      "step": 8479
+    },
+    {
+      "epoch": 12.694610778443113,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 8480
+    },
+    {
+      "epoch": 12.696107784431138,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 8481
+    },
+    {
+      "epoch": 12.697604790419161,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1366,
+      "step": 8482
+    },
+    {
+      "epoch": 12.699101796407186,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 8483
+    },
+    {
+      "epoch": 12.70059880239521,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 8484
+    },
+    {
+      "epoch": 12.702095808383234,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1476,
+      "step": 8485
+    },
+    {
+      "epoch": 12.703592814371257,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1465,
+      "step": 8486
+    },
+    {
+      "epoch": 12.705089820359282,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 8487
+    },
+    {
+      "epoch": 12.706586826347305,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 8488
+    },
+    {
+      "epoch": 12.70808383233533,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 8489
+    },
+    {
+      "epoch": 12.709580838323353,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 8490
+    },
+    {
+      "epoch": 12.711077844311378,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 8491
+    },
+    {
+      "epoch": 12.7125748502994,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 8492
+    },
+    {
+      "epoch": 12.714071856287426,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 8493
+    },
+    {
+      "epoch": 12.715568862275449,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1346,
+      "step": 8494
+    },
+    {
+      "epoch": 12.717065868263473,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1464,
+      "step": 8495
+    },
+    {
+      "epoch": 12.718562874251496,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1357,
+      "step": 8496
+    },
+    {
+      "epoch": 12.720059880239521,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1502,
+      "step": 8497
+    },
+    {
+      "epoch": 12.721556886227544,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 8498
+    },
+    {
+      "epoch": 12.72305389221557,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 8499
+    },
+    {
+      "epoch": 12.724550898203592,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 8500
+    },
+    {
+      "epoch": 12.726047904191617,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 8501
+    },
+    {
+      "epoch": 12.72754491017964,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 8502
+    },
+    {
+      "epoch": 12.729041916167665,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1386,
+      "step": 8503
+    },
+    {
+      "epoch": 12.730538922155688,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1308,
+      "step": 8504
+    },
+    {
+      "epoch": 12.732035928143713,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 8505
+    },
+    {
+      "epoch": 12.733532934131736,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 8506
+    },
+    {
+      "epoch": 12.73502994011976,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 8507
+    },
+    {
+      "epoch": 12.736526946107784,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1347,
+      "step": 8508
+    },
+    {
+      "epoch": 12.738023952095809,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 8509
+    },
+    {
+      "epoch": 12.739520958083832,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 8510
+    },
+    {
+      "epoch": 12.741017964071856,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 8511
+    },
+    {
+      "epoch": 12.74251497005988,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 8512
+    },
+    {
+      "epoch": 12.744011976047904,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 8513
+    },
+    {
+      "epoch": 12.745508982035929,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 8514
+    },
+    {
+      "epoch": 12.747005988023952,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 8515
+    },
+    {
+      "epoch": 12.748502994011975,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 8516
+    },
+    {
+      "epoch": 12.75,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 8517
+    },
+    {
+      "epoch": 12.751497005988025,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 8518
+    },
+    {
+      "epoch": 12.752994011976048,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1361,
+      "step": 8519
+    },
+    {
+      "epoch": 12.754491017964071,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 8520
+    },
+    {
+      "epoch": 12.755988023952096,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 8521
+    },
+    {
+      "epoch": 12.75748502994012,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 8522
+    },
+    {
+      "epoch": 12.758982035928144,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1383,
+      "step": 8523
+    },
+    {
+      "epoch": 12.760479041916168,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1355,
+      "step": 8524
+    },
+    {
+      "epoch": 12.761976047904191,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1316,
+      "step": 8525
+    },
+    {
+      "epoch": 12.763473053892216,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 8526
+    },
+    {
+      "epoch": 12.76497005988024,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1481,
+      "step": 8527
+    },
+    {
+      "epoch": 12.766467065868264,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 8528
+    },
+    {
+      "epoch": 12.767964071856287,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 8529
+    },
+    {
+      "epoch": 12.769461077844312,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 8530
+    },
+    {
+      "epoch": 12.770958083832335,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 8531
+    },
+    {
+      "epoch": 12.77245508982036,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 8532
+    },
+    {
+      "epoch": 12.773952095808383,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 8533
+    },
+    {
+      "epoch": 12.775449101796408,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 8534
+    },
+    {
+      "epoch": 12.77694610778443,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1463,
+      "step": 8535
+    },
+    {
+      "epoch": 12.778443113772456,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 8536
+    },
+    {
+      "epoch": 12.779940119760479,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1471,
+      "step": 8537
+    },
+    {
+      "epoch": 12.781437125748504,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1428,
+      "step": 8538
+    },
+    {
+      "epoch": 12.782934131736527,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1376,
+      "step": 8539
+    },
+    {
+      "epoch": 12.784431137724551,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 8540
+    },
+    {
+      "epoch": 12.785928143712574,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 8541
+    },
+    {
+      "epoch": 12.7874251497006,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1463,
+      "step": 8542
+    },
+    {
+      "epoch": 12.788922155688622,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 8543
+    },
+    {
+      "epoch": 12.790419161676647,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 8544
+    },
+    {
+      "epoch": 12.79191616766467,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 8545
+    },
+    {
+      "epoch": 12.793413173652695,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 8546
+    },
+    {
+      "epoch": 12.794910179640718,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 8547
+    },
+    {
+      "epoch": 12.796407185628743,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1413,
+      "step": 8548
+    },
+    {
+      "epoch": 12.797904191616766,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1483,
+      "step": 8549
+    },
+    {
+      "epoch": 12.79940119760479,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 8550
+    },
+    {
+      "epoch": 12.800898203592814,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 8551
+    },
+    {
+      "epoch": 12.802395209580839,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 8552
+    },
+    {
+      "epoch": 12.803892215568862,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 8553
+    },
+    {
+      "epoch": 12.805389221556887,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1365,
+      "step": 8554
+    },
+    {
+      "epoch": 12.80688622754491,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.137,
+      "step": 8555
+    },
+    {
+      "epoch": 12.808383233532934,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 8556
+    },
+    {
+      "epoch": 12.809880239520957,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.144,
+      "step": 8557
+    },
+    {
+      "epoch": 12.811377245508982,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 8558
+    },
+    {
+      "epoch": 12.812874251497005,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 8559
+    },
+    {
+      "epoch": 12.81437125748503,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1362,
+      "step": 8560
+    },
+    {
+      "epoch": 12.815868263473053,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 8561
+    },
+    {
+      "epoch": 12.817365269461078,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 8562
+    },
+    {
+      "epoch": 12.818862275449101,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 8563
+    },
+    {
+      "epoch": 12.820359281437126,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1336,
+      "step": 8564
+    },
+    {
+      "epoch": 12.821856287425149,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 8565
+    },
+    {
+      "epoch": 12.823353293413174,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 8566
+    },
+    {
+      "epoch": 12.824850299401197,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 8567
+    },
+    {
+      "epoch": 12.826347305389222,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 8568
+    },
+    {
+      "epoch": 12.827844311377245,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1476,
+      "step": 8569
+    },
+    {
+      "epoch": 12.82934131736527,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 8570
+    },
+    {
+      "epoch": 12.830838323353294,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1363,
+      "step": 8571
+    },
+    {
+      "epoch": 12.832335329341317,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 8572
+    },
+    {
+      "epoch": 12.83383233532934,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1297,
+      "step": 8573
+    },
+    {
+      "epoch": 12.835329341317365,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 8574
+    },
+    {
+      "epoch": 12.83682634730539,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 8575
+    },
+    {
+      "epoch": 12.838323353293413,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 8576
+    },
+    {
+      "epoch": 12.839820359281438,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 8577
+    },
+    {
+      "epoch": 12.841317365269461,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 8578
+    },
+    {
+      "epoch": 12.842814371257486,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 8579
+    },
+    {
+      "epoch": 12.844311377245509,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 8580
+    },
+    {
+      "epoch": 12.845808383233534,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1475,
+      "step": 8581
+    },
+    {
+      "epoch": 12.847305389221557,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 8582
+    },
+    {
+      "epoch": 12.848802395209582,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 8583
+    },
+    {
+      "epoch": 12.850299401197605,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 8584
+    },
+    {
+      "epoch": 12.85179640718563,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1384,
+      "step": 8585
+    },
+    {
+      "epoch": 12.853293413173652,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 8586
+    },
+    {
+      "epoch": 12.854790419161677,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 8587
+    },
+    {
+      "epoch": 12.8562874251497,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1486,
+      "step": 8588
+    },
+    {
+      "epoch": 12.857784431137725,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1376,
+      "step": 8589
+    },
+    {
+      "epoch": 12.859281437125748,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1443,
+      "step": 8590
+    },
+    {
+      "epoch": 12.860778443113773,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 8591
+    },
+    {
+      "epoch": 12.862275449101796,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 8592
+    },
+    {
+      "epoch": 12.863772455089821,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 8593
+    },
+    {
+      "epoch": 12.865269461077844,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1446,
+      "step": 8594
+    },
+    {
+      "epoch": 12.866766467065869,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1384,
+      "step": 8595
+    },
+    {
+      "epoch": 12.868263473053892,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1416,
+      "step": 8596
+    },
+    {
+      "epoch": 12.869760479041917,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.14,
+      "step": 8597
+    },
+    {
+      "epoch": 12.87125748502994,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1387,
+      "step": 8598
+    },
+    {
+      "epoch": 12.872754491017965,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 8599
+    },
+    {
+      "epoch": 12.874251497005988,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1384,
+      "step": 8600
+    },
+    {
+      "epoch": 12.875748502994012,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 8601
+    },
+    {
+      "epoch": 12.877245508982035,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 8602
+    },
+    {
+      "epoch": 12.87874251497006,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1381,
+      "step": 8603
+    },
+    {
+      "epoch": 12.880239520958083,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 8604
+    },
+    {
+      "epoch": 12.881736526946108,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 8605
+    },
+    {
+      "epoch": 12.883233532934131,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1428,
+      "step": 8606
+    },
+    {
+      "epoch": 12.884730538922156,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 8607
+    },
+    {
+      "epoch": 12.886227544910179,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 8608
+    },
+    {
+      "epoch": 12.887724550898204,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 8609
+    },
+    {
+      "epoch": 12.889221556886227,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1298,
+      "step": 8610
+    },
+    {
+      "epoch": 12.890718562874252,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 8611
+    },
+    {
+      "epoch": 12.892215568862275,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 8612
+    },
+    {
+      "epoch": 12.8937125748503,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1334,
+      "step": 8613
+    },
+    {
+      "epoch": 12.895209580838323,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 8614
+    },
+    {
+      "epoch": 12.896706586826348,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1378,
+      "step": 8615
+    },
+    {
+      "epoch": 12.89820359281437,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 8616
+    },
+    {
+      "epoch": 12.899700598802395,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1372,
+      "step": 8617
+    },
+    {
+      "epoch": 12.901197604790418,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 8618
+    },
+    {
+      "epoch": 12.902694610778443,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1336,
+      "step": 8619
+    },
+    {
+      "epoch": 12.904191616766466,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1366,
+      "step": 8620
+    },
+    {
+      "epoch": 12.905688622754491,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 8621
+    },
+    {
+      "epoch": 12.907185628742514,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 8622
+    },
+    {
+      "epoch": 12.908682634730539,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 8623
+    },
+    {
+      "epoch": 12.910179640718562,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 8624
+    },
+    {
+      "epoch": 12.911676646706587,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 8625
+    },
+    {
+      "epoch": 12.91317365269461,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 8626
+    },
+    {
+      "epoch": 12.914670658682635,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1302,
+      "step": 8627
+    },
+    {
+      "epoch": 12.91616766467066,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1358,
+      "step": 8628
+    },
+    {
+      "epoch": 12.917664670658683,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1443,
+      "step": 8629
+    },
+    {
+      "epoch": 12.919161676646706,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1429,
+      "step": 8630
+    },
+    {
+      "epoch": 12.92065868263473,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 8631
+    },
+    {
+      "epoch": 12.922155688622755,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1471,
+      "step": 8632
+    },
+    {
+      "epoch": 12.923652694610778,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1492,
+      "step": 8633
+    },
+    {
+      "epoch": 12.925149700598803,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 8634
+    },
+    {
+      "epoch": 12.926646706586826,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 8635
+    },
+    {
+      "epoch": 12.928143712574851,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1356,
+      "step": 8636
+    },
+    {
+      "epoch": 12.929640718562874,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 8637
+    },
+    {
+      "epoch": 12.931137724550899,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1372,
+      "step": 8638
+    },
+    {
+      "epoch": 12.932634730538922,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 8639
+    },
+    {
+      "epoch": 12.934131736526947,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1369,
+      "step": 8640
+    },
+    {
+      "epoch": 12.93562874251497,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 8641
+    },
+    {
+      "epoch": 12.937125748502995,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 8642
+    },
+    {
+      "epoch": 12.938622754491018,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1373,
+      "step": 8643
+    },
+    {
+      "epoch": 12.940119760479043,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1354,
+      "step": 8644
+    },
+    {
+      "epoch": 12.941616766467066,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1349,
+      "step": 8645
+    },
+    {
+      "epoch": 12.94311377245509,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 8646
+    },
+    {
+      "epoch": 12.944610778443113,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 8647
+    },
+    {
+      "epoch": 12.946107784431138,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 8648
+    },
+    {
+      "epoch": 12.947604790419161,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.134,
+      "step": 8649
+    },
+    {
+      "epoch": 12.949101796407186,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 8650
+    },
+    {
+      "epoch": 12.95059880239521,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1443,
+      "step": 8651
+    },
+    {
+      "epoch": 12.952095808383234,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1376,
+      "step": 8652
+    },
+    {
+      "epoch": 12.953592814371257,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1392,
+      "step": 8653
+    },
+    {
+      "epoch": 12.955089820359282,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 8654
+    },
+    {
+      "epoch": 12.956586826347305,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1428,
+      "step": 8655
+    },
+    {
+      "epoch": 12.95808383233533,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 8656
+    },
+    {
+      "epoch": 12.959580838323353,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 8657
+    },
+    {
+      "epoch": 12.961077844311378,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 8658
+    },
+    {
+      "epoch": 12.9625748502994,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.136,
+      "step": 8659
+    },
+    {
+      "epoch": 12.964071856287426,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1386,
+      "step": 8660
+    },
+    {
+      "epoch": 12.965568862275449,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 8661
+    },
+    {
+      "epoch": 12.967065868263473,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1443,
+      "step": 8662
+    },
+    {
+      "epoch": 12.968562874251496,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 8663
+    },
+    {
+      "epoch": 12.970059880239521,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1429,
+      "step": 8664
+    },
+    {
+      "epoch": 12.971556886227544,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 8665
+    },
+    {
+      "epoch": 12.97305389221557,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 8666
+    },
+    {
+      "epoch": 12.974550898203592,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 8667
+    },
+    {
+      "epoch": 12.976047904191617,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1372,
+      "step": 8668
+    },
+    {
+      "epoch": 12.97754491017964,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1351,
+      "step": 8669
+    },
+    {
+      "epoch": 12.979041916167665,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 8670
+    },
+    {
+      "epoch": 12.980538922155688,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 8671
+    },
+    {
+      "epoch": 12.982035928143713,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 8672
+    },
+    {
+      "epoch": 12.983532934131736,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 8673
+    },
+    {
+      "epoch": 12.98502994011976,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1443,
+      "step": 8674
+    },
+    {
+      "epoch": 12.986526946107784,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1416,
+      "step": 8675
+    },
+    {
+      "epoch": 12.988023952095809,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 8676
+    },
+    {
+      "epoch": 12.989520958083832,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1317,
+      "step": 8677
+    },
+    {
+      "epoch": 12.991017964071856,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 8678
+    },
+    {
+      "epoch": 12.99251497005988,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 8679
+    },
+    {
+      "epoch": 12.994011976047904,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.146,
+      "step": 8680
+    },
+    {
+      "epoch": 12.995508982035929,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1354,
+      "step": 8681
+    },
+    {
+      "epoch": 12.997005988023952,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1413,
+      "step": 8682
+    },
+    {
+      "epoch": 12.998502994011975,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 8683
+    },
+    {
+      "epoch": 13.0,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 8684
+    },
+    {
+      "epoch": 13.001497005988025,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 8685
+    },
+    {
+      "epoch": 13.002994011976048,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1372,
+      "step": 8686
+    },
+    {
+      "epoch": 13.004491017964073,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1372,
+      "step": 8687
+    },
+    {
+      "epoch": 13.005988023952096,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 8688
+    },
+    {
+      "epoch": 13.00748502994012,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1393,
+      "step": 8689
+    },
+    {
+      "epoch": 13.008982035928144,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1348,
+      "step": 8690
+    },
+    {
+      "epoch": 13.010479041916168,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 8691
+    },
+    {
+      "epoch": 13.011976047904191,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 8692
+    },
+    {
+      "epoch": 13.013473053892216,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 8693
+    },
+    {
+      "epoch": 13.01497005988024,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1395,
+      "step": 8694
+    },
+    {
+      "epoch": 13.016467065868264,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.144,
+      "step": 8695
+    },
+    {
+      "epoch": 13.017964071856287,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 8696
+    },
+    {
+      "epoch": 13.019461077844312,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.137,
+      "step": 8697
+    },
+    {
+      "epoch": 13.020958083832335,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1331,
+      "step": 8698
+    },
+    {
+      "epoch": 13.02245508982036,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1335,
+      "step": 8699
+    },
+    {
+      "epoch": 13.023952095808383,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 8700
+    },
+    {
+      "epoch": 13.025449101796408,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1335,
+      "step": 8701
+    },
+    {
+      "epoch": 13.02694610778443,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 8702
+    },
+    {
+      "epoch": 13.028443113772456,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 8703
+    },
+    {
+      "epoch": 13.029940119760479,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1361,
+      "step": 8704
+    },
+    {
+      "epoch": 13.031437125748504,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 8705
+    },
+    {
+      "epoch": 13.032934131736527,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1428,
+      "step": 8706
+    },
+    {
+      "epoch": 13.034431137724551,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 8707
+    },
+    {
+      "epoch": 13.035928143712574,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 8708
+    },
+    {
+      "epoch": 13.0374251497006,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 8709
+    },
+    {
+      "epoch": 13.038922155688622,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1344,
+      "step": 8710
+    },
+    {
+      "epoch": 13.040419161676647,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1349,
+      "step": 8711
+    },
+    {
+      "epoch": 13.04191616766467,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1392,
+      "step": 8712
+    },
+    {
+      "epoch": 13.043413173652695,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 8713
+    },
+    {
+      "epoch": 13.044910179640718,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 8714
+    },
+    {
+      "epoch": 13.046407185628743,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1358,
+      "step": 8715
+    },
+    {
+      "epoch": 13.047904191616766,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 8716
+    },
+    {
+      "epoch": 13.04940119760479,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 8717
+    },
+    {
+      "epoch": 13.050898203592814,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 8718
+    },
+    {
+      "epoch": 13.052395209580839,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 8719
+    },
+    {
+      "epoch": 13.053892215568862,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 8720
+    },
+    {
+      "epoch": 13.055389221556887,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 8721
+    },
+    {
+      "epoch": 13.05688622754491,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1354,
+      "step": 8722
+    },
+    {
+      "epoch": 13.058383233532934,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 8723
+    },
+    {
+      "epoch": 13.059880239520957,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.14,
+      "step": 8724
+    },
+    {
+      "epoch": 13.061377245508982,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 8725
+    },
+    {
+      "epoch": 13.062874251497005,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 8726
+    },
+    {
+      "epoch": 13.06437125748503,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1463,
+      "step": 8727
+    },
+    {
+      "epoch": 13.065868263473053,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1319,
+      "step": 8728
+    },
+    {
+      "epoch": 13.067365269461078,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 8729
+    },
+    {
+      "epoch": 13.068862275449101,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.133,
+      "step": 8730
+    },
+    {
+      "epoch": 13.070359281437126,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1369,
+      "step": 8731
+    },
+    {
+      "epoch": 13.071856287425149,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 8732
+    },
+    {
+      "epoch": 13.073353293413174,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 8733
+    },
+    {
+      "epoch": 13.074850299401197,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1469,
+      "step": 8734
+    },
+    {
+      "epoch": 13.076347305389222,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1368,
+      "step": 8735
+    },
+    {
+      "epoch": 13.077844311377245,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1459,
+      "step": 8736
+    },
+    {
+      "epoch": 13.07934131736527,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 8737
+    },
+    {
+      "epoch": 13.080838323353293,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 8738
+    },
+    {
+      "epoch": 13.082335329341317,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 8739
+    },
+    {
+      "epoch": 13.08383233532934,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1359,
+      "step": 8740
+    },
+    {
+      "epoch": 13.085329341317365,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 8741
+    },
+    {
+      "epoch": 13.08682634730539,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 8742
+    },
+    {
+      "epoch": 13.088323353293413,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 8743
+    },
+    {
+      "epoch": 13.089820359281438,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1471,
+      "step": 8744
+    },
+    {
+      "epoch": 13.091317365269461,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1358,
+      "step": 8745
+    },
+    {
+      "epoch": 13.092814371257486,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 8746
+    },
+    {
+      "epoch": 13.094311377245509,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1361,
+      "step": 8747
+    },
+    {
+      "epoch": 13.095808383233534,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1357,
+      "step": 8748
+    },
+    {
+      "epoch": 13.097305389221557,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1378,
+      "step": 8749
+    },
+    {
+      "epoch": 13.098802395209582,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 8750
+    },
+    {
+      "epoch": 13.100299401197605,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1357,
+      "step": 8751
+    },
+    {
+      "epoch": 13.10179640718563,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1464,
+      "step": 8752
+    },
+    {
+      "epoch": 13.103293413173652,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 8753
+    },
+    {
+      "epoch": 13.104790419161677,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1351,
+      "step": 8754
+    },
+    {
+      "epoch": 13.1062874251497,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1355,
+      "step": 8755
+    },
+    {
+      "epoch": 13.107784431137725,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1343,
+      "step": 8756
+    },
+    {
+      "epoch": 13.109281437125748,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1341,
+      "step": 8757
+    },
+    {
+      "epoch": 13.110778443113773,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 8758
+    },
+    {
+      "epoch": 13.112275449101796,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1476,
+      "step": 8759
+    },
+    {
+      "epoch": 13.113772455089821,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 8760
+    },
+    {
+      "epoch": 13.115269461077844,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 8761
+    },
+    {
+      "epoch": 13.116766467065869,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 8762
+    },
+    {
+      "epoch": 13.118263473053892,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1484,
+      "step": 8763
+    },
+    {
+      "epoch": 13.119760479041917,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 8764
+    },
+    {
+      "epoch": 13.12125748502994,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1443,
+      "step": 8765
+    },
+    {
+      "epoch": 13.122754491017965,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 8766
+    },
+    {
+      "epoch": 13.124251497005988,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 8767
+    },
+    {
+      "epoch": 13.125748502994012,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 8768
+    },
+    {
+      "epoch": 13.127245508982035,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1378,
+      "step": 8769
+    },
+    {
+      "epoch": 13.12874251497006,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1327,
+      "step": 8770
+    },
+    {
+      "epoch": 13.130239520958083,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 8771
+    },
+    {
+      "epoch": 13.131736526946108,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.135,
+      "step": 8772
+    },
+    {
+      "epoch": 13.133233532934131,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 8773
+    },
+    {
+      "epoch": 13.134730538922156,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1457,
+      "step": 8774
+    },
+    {
+      "epoch": 13.136227544910179,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 8775
+    },
+    {
+      "epoch": 13.137724550898204,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1364,
+      "step": 8776
+    },
+    {
+      "epoch": 13.139221556886227,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1342,
+      "step": 8777
+    },
+    {
+      "epoch": 13.140718562874252,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1362,
+      "step": 8778
+    },
+    {
+      "epoch": 13.142215568862275,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 8779
+    },
+    {
+      "epoch": 13.1437125748503,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1357,
+      "step": 8780
+    },
+    {
+      "epoch": 13.145209580838323,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1392,
+      "step": 8781
+    },
+    {
+      "epoch": 13.146706586826348,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1464,
+      "step": 8782
+    },
+    {
+      "epoch": 13.14820359281437,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1457,
+      "step": 8783
+    },
+    {
+      "epoch": 13.149700598802395,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.136,
+      "step": 8784
+    },
+    {
+      "epoch": 13.151197604790418,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1413,
+      "step": 8785
+    },
+    {
+      "epoch": 13.152694610778443,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 8786
+    },
+    {
+      "epoch": 13.154191616766466,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1366,
+      "step": 8787
+    },
+    {
+      "epoch": 13.155688622754491,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 8788
+    },
+    {
+      "epoch": 13.157185628742514,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 8789
+    },
+    {
+      "epoch": 13.158682634730539,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.144,
+      "step": 8790
+    },
+    {
+      "epoch": 13.160179640718562,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.135,
+      "step": 8791
+    },
+    {
+      "epoch": 13.161676646706587,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 8792
+    },
+    {
+      "epoch": 13.16317365269461,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1454,
+      "step": 8793
+    },
+    {
+      "epoch": 13.164670658682635,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1347,
+      "step": 8794
+    },
+    {
+      "epoch": 13.16616766467066,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 8795
+    },
+    {
+      "epoch": 13.167664670658683,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 8796
+    },
+    {
+      "epoch": 13.169161676646707,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 8797
+    },
+    {
+      "epoch": 13.17065868263473,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1344,
+      "step": 8798
+    },
+    {
+      "epoch": 13.172155688622755,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 8799
+    },
+    {
+      "epoch": 13.173652694610778,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 8800
+    },
+    {
+      "epoch": 13.175149700598803,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 8801
+    },
+    {
+      "epoch": 13.176646706586826,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1387,
+      "step": 8802
+    },
+    {
+      "epoch": 13.178143712574851,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 8803
+    },
+    {
+      "epoch": 13.179640718562874,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 8804
+    },
+    {
+      "epoch": 13.181137724550899,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1348,
+      "step": 8805
+    },
+    {
+      "epoch": 13.182634730538922,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1387,
+      "step": 8806
+    },
+    {
+      "epoch": 13.184131736526947,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 8807
+    },
+    {
+      "epoch": 13.18562874251497,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 8808
+    },
+    {
+      "epoch": 13.187125748502995,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1376,
+      "step": 8809
+    },
+    {
+      "epoch": 13.188622754491018,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 8810
+    },
+    {
+      "epoch": 13.190119760479043,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 8811
+    },
+    {
+      "epoch": 13.191616766467066,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1312,
+      "step": 8812
+    },
+    {
+      "epoch": 13.19311377245509,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1359,
+      "step": 8813
+    },
+    {
+      "epoch": 13.194610778443113,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1351,
+      "step": 8814
+    },
+    {
+      "epoch": 13.196107784431138,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1413,
+      "step": 8815
+    },
+    {
+      "epoch": 13.197604790419161,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 8816
+    },
+    {
+      "epoch": 13.199101796407186,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 8817
+    },
+    {
+      "epoch": 13.20059880239521,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 8818
+    },
+    {
+      "epoch": 13.202095808383234,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1378,
+      "step": 8819
+    },
+    {
+      "epoch": 13.203592814371257,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 8820
+    },
+    {
+      "epoch": 13.205089820359282,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.14,
+      "step": 8821
+    },
+    {
+      "epoch": 13.206586826347305,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1367,
+      "step": 8822
+    },
+    {
+      "epoch": 13.20808383233533,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1454,
+      "step": 8823
+    },
+    {
+      "epoch": 13.209580838323353,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1434,
+      "step": 8824
+    },
+    {
+      "epoch": 13.211077844311378,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1381,
+      "step": 8825
+    },
+    {
+      "epoch": 13.2125748502994,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 8826
+    },
+    {
+      "epoch": 13.214071856287426,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 8827
+    },
+    {
+      "epoch": 13.215568862275449,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1345,
+      "step": 8828
+    },
+    {
+      "epoch": 13.217065868263473,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1483,
+      "step": 8829
+    },
+    {
+      "epoch": 13.218562874251496,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 8830
+    },
+    {
+      "epoch": 13.220059880239521,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1336,
+      "step": 8831
+    },
+    {
+      "epoch": 13.221556886227544,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 8832
+    },
+    {
+      "epoch": 13.22305389221557,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 8833
+    },
+    {
+      "epoch": 13.224550898203592,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 8834
+    },
+    {
+      "epoch": 13.226047904191617,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1387,
+      "step": 8835
+    },
+    {
+      "epoch": 13.22754491017964,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 8836
+    },
+    {
+      "epoch": 13.229041916167665,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 8837
+    },
+    {
+      "epoch": 13.230538922155688,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.14,
+      "step": 8838
+    },
+    {
+      "epoch": 13.232035928143713,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 8839
+    },
+    {
+      "epoch": 13.233532934131736,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1383,
+      "step": 8840
+    },
+    {
+      "epoch": 13.23502994011976,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 8841
+    },
+    {
+      "epoch": 13.236526946107784,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 8842
+    },
+    {
+      "epoch": 13.238023952095809,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 8843
+    },
+    {
+      "epoch": 13.239520958083832,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1393,
+      "step": 8844
+    },
+    {
+      "epoch": 13.241017964071856,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1376,
+      "step": 8845
+    },
+    {
+      "epoch": 13.24251497005988,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1386,
+      "step": 8846
+    },
+    {
+      "epoch": 13.244011976047904,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 8847
+    },
+    {
+      "epoch": 13.245508982035927,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1481,
+      "step": 8848
+    },
+    {
+      "epoch": 13.247005988023952,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 8849
+    },
+    {
+      "epoch": 13.248502994011975,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 8850
+    },
+    {
+      "epoch": 13.25,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 8851
+    },
+    {
+      "epoch": 13.251497005988025,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 8852
+    },
+    {
+      "epoch": 13.252994011976048,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1323,
+      "step": 8853
+    },
+    {
+      "epoch": 13.254491017964073,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 8854
+    },
+    {
+      "epoch": 13.255988023952096,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1371,
+      "step": 8855
+    },
+    {
+      "epoch": 13.25748502994012,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 8856
+    },
+    {
+      "epoch": 13.258982035928144,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 8857
+    },
+    {
+      "epoch": 13.260479041916168,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 8858
+    },
+    {
+      "epoch": 13.261976047904191,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 8859
+    },
+    {
+      "epoch": 13.263473053892216,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 8860
+    },
+    {
+      "epoch": 13.26497005988024,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1354,
+      "step": 8861
+    },
+    {
+      "epoch": 13.266467065868264,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 8862
+    },
+    {
+      "epoch": 13.267964071856287,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1429,
+      "step": 8863
+    },
+    {
+      "epoch": 13.269461077844312,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.135,
+      "step": 8864
+    },
+    {
+      "epoch": 13.270958083832335,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 8865
+    },
+    {
+      "epoch": 13.27245508982036,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 8866
+    },
+    {
+      "epoch": 13.273952095808383,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 8867
+    },
+    {
+      "epoch": 13.275449101796408,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 8868
+    },
+    {
+      "epoch": 13.27694610778443,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.134,
+      "step": 8869
+    },
+    {
+      "epoch": 13.278443113772456,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 8870
+    },
+    {
+      "epoch": 13.279940119760479,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 8871
+    },
+    {
+      "epoch": 13.281437125748504,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1327,
+      "step": 8872
+    },
+    {
+      "epoch": 13.282934131736527,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.132,
+      "step": 8873
+    },
+    {
+      "epoch": 13.284431137724551,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1346,
+      "step": 8874
+    },
+    {
+      "epoch": 13.285928143712574,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.136,
+      "step": 8875
+    },
+    {
+      "epoch": 13.2874251497006,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1342,
+      "step": 8876
+    },
+    {
+      "epoch": 13.288922155688622,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.134,
+      "step": 8877
+    },
+    {
+      "epoch": 13.290419161676647,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 8878
+    },
+    {
+      "epoch": 13.29191616766467,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 8879
+    },
+    {
+      "epoch": 13.293413173652695,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 8880
+    },
+    {
+      "epoch": 13.294910179640718,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 8881
+    },
+    {
+      "epoch": 13.296407185628743,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 8882
+    },
+    {
+      "epoch": 13.297904191616766,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1372,
+      "step": 8883
+    },
+    {
+      "epoch": 13.29940119760479,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 8884
+    },
+    {
+      "epoch": 13.300898203592814,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1335,
+      "step": 8885
+    },
+    {
+      "epoch": 13.302395209580839,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 8886
+    },
+    {
+      "epoch": 13.303892215568862,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.146,
+      "step": 8887
+    },
+    {
+      "epoch": 13.305389221556887,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 8888
+    },
+    {
+      "epoch": 13.30688622754491,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1381,
+      "step": 8889
+    },
+    {
+      "epoch": 13.308383233532934,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 8890
+    },
+    {
+      "epoch": 13.309880239520957,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1487,
+      "step": 8891
+    },
+    {
+      "epoch": 13.311377245508982,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 8892
+    },
+    {
+      "epoch": 13.312874251497005,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 8893
+    },
+    {
+      "epoch": 13.31437125748503,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1319,
+      "step": 8894
+    },
+    {
+      "epoch": 13.315868263473053,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1294,
+      "step": 8895
+    },
+    {
+      "epoch": 13.317365269461078,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 8896
+    },
+    {
+      "epoch": 13.318862275449101,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1371,
+      "step": 8897
+    },
+    {
+      "epoch": 13.320359281437126,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1472,
+      "step": 8898
+    },
+    {
+      "epoch": 13.321856287425149,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 8899
+    },
+    {
+      "epoch": 13.323353293413174,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.135,
+      "step": 8900
+    },
+    {
+      "epoch": 13.324850299401197,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 8901
+    },
+    {
+      "epoch": 13.326347305389222,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 8902
+    },
+    {
+      "epoch": 13.327844311377245,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 8903
+    },
+    {
+      "epoch": 13.32934131736527,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1367,
+      "step": 8904
+    },
+    {
+      "epoch": 13.330838323353294,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 8905
+    },
+    {
+      "epoch": 13.332335329341317,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 8906
+    },
+    {
+      "epoch": 13.33383233532934,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1362,
+      "step": 8907
+    },
+    {
+      "epoch": 13.335329341317365,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1487,
+      "step": 8908
+    },
+    {
+      "epoch": 13.33682634730539,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 8909
+    },
+    {
+      "epoch": 13.338323353293413,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 8910
+    },
+    {
+      "epoch": 13.339820359281438,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 8911
+    },
+    {
+      "epoch": 13.341317365269461,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 8912
+    },
+    {
+      "epoch": 13.342814371257486,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1378,
+      "step": 8913
+    },
+    {
+      "epoch": 13.344311377245509,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 8914
+    },
+    {
+      "epoch": 13.345808383233534,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1362,
+      "step": 8915
+    },
+    {
+      "epoch": 13.347305389221557,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1395,
+      "step": 8916
+    },
+    {
+      "epoch": 13.348802395209582,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 8917
+    },
+    {
+      "epoch": 13.350299401197605,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 8918
+    },
+    {
+      "epoch": 13.35179640718563,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 8919
+    },
+    {
+      "epoch": 13.353293413173652,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 8920
+    },
+    {
+      "epoch": 13.354790419161677,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1462,
+      "step": 8921
+    },
+    {
+      "epoch": 13.3562874251497,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 8922
+    },
+    {
+      "epoch": 13.357784431137725,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1359,
+      "step": 8923
+    },
+    {
+      "epoch": 13.359281437125748,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1364,
+      "step": 8924
+    },
+    {
+      "epoch": 13.360778443113773,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1318,
+      "step": 8925
+    },
+    {
+      "epoch": 13.362275449101796,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1377,
+      "step": 8926
+    },
+    {
+      "epoch": 13.363772455089821,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1357,
+      "step": 8927
+    },
+    {
+      "epoch": 13.365269461077844,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 8928
+    },
+    {
+      "epoch": 13.366766467065869,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 8929
+    },
+    {
+      "epoch": 13.368263473053892,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1484,
+      "step": 8930
+    },
+    {
+      "epoch": 13.369760479041917,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 8931
+    },
+    {
+      "epoch": 13.37125748502994,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1393,
+      "step": 8932
+    },
+    {
+      "epoch": 13.372754491017965,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 8933
+    },
+    {
+      "epoch": 13.374251497005988,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1364,
+      "step": 8934
+    },
+    {
+      "epoch": 13.375748502994012,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 8935
+    },
+    {
+      "epoch": 13.377245508982035,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1353,
+      "step": 8936
+    },
+    {
+      "epoch": 13.37874251497006,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1434,
+      "step": 8937
+    },
+    {
+      "epoch": 13.380239520958083,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1325,
+      "step": 8938
+    },
+    {
+      "epoch": 13.381736526946108,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 8939
+    },
+    {
+      "epoch": 13.383233532934131,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 8940
+    },
+    {
+      "epoch": 13.384730538922156,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1383,
+      "step": 8941
+    },
+    {
+      "epoch": 13.386227544910179,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 8942
+    },
+    {
+      "epoch": 13.387724550898204,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1375,
+      "step": 8943
+    },
+    {
+      "epoch": 13.389221556886227,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1373,
+      "step": 8944
+    },
+    {
+      "epoch": 13.390718562874252,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 8945
+    },
+    {
+      "epoch": 13.392215568862275,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 8946
+    },
+    {
+      "epoch": 13.3937125748503,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1387,
+      "step": 8947
+    },
+    {
+      "epoch": 13.395209580838323,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1376,
+      "step": 8948
+    },
+    {
+      "epoch": 13.396706586826348,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1434,
+      "step": 8949
+    },
+    {
+      "epoch": 13.39820359281437,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 8950
+    },
+    {
+      "epoch": 13.399700598802395,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 8951
+    },
+    {
+      "epoch": 13.401197604790418,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 8952
+    },
+    {
+      "epoch": 13.402694610778443,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.147,
+      "step": 8953
+    },
+    {
+      "epoch": 13.404191616766466,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1434,
+      "step": 8954
+    },
+    {
+      "epoch": 13.405688622754491,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 8955
+    },
+    {
+      "epoch": 13.407185628742514,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 8956
+    },
+    {
+      "epoch": 13.408682634730539,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1311,
+      "step": 8957
+    },
+    {
+      "epoch": 13.410179640718562,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 8958
+    },
+    {
+      "epoch": 13.411676646706587,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1366,
+      "step": 8959
+    },
+    {
+      "epoch": 13.41317365269461,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 8960
+    },
+    {
+      "epoch": 13.414670658682635,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 8961
+    },
+    {
+      "epoch": 13.41616766467066,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 8962
+    },
+    {
+      "epoch": 13.417664670658683,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1366,
+      "step": 8963
+    },
+    {
+      "epoch": 13.419161676646706,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.131,
+      "step": 8964
+    },
+    {
+      "epoch": 13.42065868263473,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 8965
+    },
+    {
+      "epoch": 13.422155688622755,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1377,
+      "step": 8966
+    },
+    {
+      "epoch": 13.423652694610778,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 8967
+    },
+    {
+      "epoch": 13.425149700598803,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1352,
+      "step": 8968
+    },
+    {
+      "epoch": 13.426646706586826,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1355,
+      "step": 8969
+    },
+    {
+      "epoch": 13.428143712574851,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1357,
+      "step": 8970
+    },
+    {
+      "epoch": 13.429640718562874,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.135,
+      "step": 8971
+    },
+    {
+      "epoch": 13.431137724550899,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 8972
+    },
+    {
+      "epoch": 13.432634730538922,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 8973
+    },
+    {
+      "epoch": 13.434131736526947,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1372,
+      "step": 8974
+    },
+    {
+      "epoch": 13.43562874251497,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 8975
+    },
+    {
+      "epoch": 13.437125748502995,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1367,
+      "step": 8976
+    },
+    {
+      "epoch": 13.438622754491018,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1357,
+      "step": 8977
+    },
+    {
+      "epoch": 13.440119760479043,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 8978
+    },
+    {
+      "epoch": 13.441616766467066,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 8979
+    },
+    {
+      "epoch": 13.44311377245509,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 8980
+    },
+    {
+      "epoch": 13.444610778443113,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1377,
+      "step": 8981
+    },
+    {
+      "epoch": 13.446107784431138,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1372,
+      "step": 8982
+    },
+    {
+      "epoch": 13.447604790419161,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 8983
+    },
+    {
+      "epoch": 13.449101796407186,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1363,
+      "step": 8984
+    },
+    {
+      "epoch": 13.45059880239521,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1459,
+      "step": 8985
+    },
+    {
+      "epoch": 13.452095808383234,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 8986
+    },
+    {
+      "epoch": 13.453592814371257,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 8987
+    },
+    {
+      "epoch": 13.455089820359282,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 8988
+    },
+    {
+      "epoch": 13.456586826347305,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 8989
+    },
+    {
+      "epoch": 13.45808383233533,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.144,
+      "step": 8990
+    },
+    {
+      "epoch": 13.459580838323353,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1395,
+      "step": 8991
+    },
+    {
+      "epoch": 13.461077844311378,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 8992
+    },
+    {
+      "epoch": 13.4625748502994,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 8993
+    },
+    {
+      "epoch": 13.464071856287426,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 8994
+    },
+    {
+      "epoch": 13.465568862275449,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 8995
+    },
+    {
+      "epoch": 13.467065868263473,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1377,
+      "step": 8996
+    },
+    {
+      "epoch": 13.468562874251496,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 8997
+    },
+    {
+      "epoch": 13.470059880239521,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 8998
+    },
+    {
+      "epoch": 13.471556886227544,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1429,
+      "step": 8999
+    },
+    {
+      "epoch": 13.47305389221557,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1345,
+      "step": 9000
+    },
+    {
+      "epoch": 13.474550898203592,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 9001
+    },
+    {
+      "epoch": 13.476047904191617,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1375,
+      "step": 9002
+    },
+    {
+      "epoch": 13.47754491017964,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 9003
+    },
+    {
+      "epoch": 13.479041916167665,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 9004
+    },
+    {
+      "epoch": 13.480538922155688,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 9005
+    },
+    {
+      "epoch": 13.482035928143713,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 9006
+    },
+    {
+      "epoch": 13.483532934131736,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1338,
+      "step": 9007
+    },
+    {
+      "epoch": 13.48502994011976,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 9008
+    },
+    {
+      "epoch": 13.486526946107784,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1339,
+      "step": 9009
+    },
+    {
+      "epoch": 13.488023952095809,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 9010
+    },
+    {
+      "epoch": 13.489520958083832,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1413,
+      "step": 9011
+    },
+    {
+      "epoch": 13.491017964071856,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 9012
+    },
+    {
+      "epoch": 13.49251497005988,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 9013
+    },
+    {
+      "epoch": 13.494011976047904,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1367,
+      "step": 9014
+    },
+    {
+      "epoch": 13.495508982035927,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 9015
+    },
+    {
+      "epoch": 13.497005988023952,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1413,
+      "step": 9016
+    },
+    {
+      "epoch": 13.498502994011975,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1353,
+      "step": 9017
+    },
+    {
+      "epoch": 13.5,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 9018
+    },
+    {
+      "epoch": 13.501497005988025,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1346,
+      "step": 9019
+    },
+    {
+      "epoch": 13.502994011976048,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1366,
+      "step": 9020
+    },
+    {
+      "epoch": 13.504491017964071,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 9021
+    },
+    {
+      "epoch": 13.505988023952096,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1357,
+      "step": 9022
+    },
+    {
+      "epoch": 13.50748502994012,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1359,
+      "step": 9023
+    },
+    {
+      "epoch": 13.508982035928144,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1344,
+      "step": 9024
+    },
+    {
+      "epoch": 13.510479041916168,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 9025
+    },
+    {
+      "epoch": 13.511976047904191,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 9026
+    },
+    {
+      "epoch": 13.513473053892216,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 9027
+    },
+    {
+      "epoch": 13.51497005988024,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 9028
+    },
+    {
+      "epoch": 13.516467065868264,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1372,
+      "step": 9029
+    },
+    {
+      "epoch": 13.517964071856287,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1379,
+      "step": 9030
+    },
+    {
+      "epoch": 13.519461077844312,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 9031
+    },
+    {
+      "epoch": 13.520958083832335,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1395,
+      "step": 9032
+    },
+    {
+      "epoch": 13.52245508982036,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1416,
+      "step": 9033
+    },
+    {
+      "epoch": 13.523952095808383,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 9034
+    },
+    {
+      "epoch": 13.525449101796408,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 9035
+    },
+    {
+      "epoch": 13.52694610778443,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1373,
+      "step": 9036
+    },
+    {
+      "epoch": 13.528443113772456,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1333,
+      "step": 9037
+    },
+    {
+      "epoch": 13.529940119760479,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 9038
+    },
+    {
+      "epoch": 13.531437125748504,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 9039
+    },
+    {
+      "epoch": 13.532934131736527,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 9040
+    },
+    {
+      "epoch": 13.534431137724551,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1342,
+      "step": 9041
+    },
+    {
+      "epoch": 13.535928143712574,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 9042
+    },
+    {
+      "epoch": 13.5374251497006,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.147,
+      "step": 9043
+    },
+    {
+      "epoch": 13.538922155688622,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 9044
+    },
+    {
+      "epoch": 13.540419161676647,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 9045
+    },
+    {
+      "epoch": 13.54191616766467,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1365,
+      "step": 9046
+    },
+    {
+      "epoch": 13.543413173652695,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1342,
+      "step": 9047
+    },
+    {
+      "epoch": 13.544910179640718,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 9048
+    },
+    {
+      "epoch": 13.546407185628743,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1349,
+      "step": 9049
+    },
+    {
+      "epoch": 13.547904191616766,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1386,
+      "step": 9050
+    },
+    {
+      "epoch": 13.54940119760479,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1356,
+      "step": 9051
+    },
+    {
+      "epoch": 13.550898203592814,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 9052
+    },
+    {
+      "epoch": 13.552395209580839,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1293,
+      "step": 9053
+    },
+    {
+      "epoch": 13.553892215568862,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1358,
+      "step": 9054
+    },
+    {
+      "epoch": 13.555389221556887,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 9055
+    },
+    {
+      "epoch": 13.55688622754491,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 9056
+    },
+    {
+      "epoch": 13.558383233532934,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 9057
+    },
+    {
+      "epoch": 13.559880239520957,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1443,
+      "step": 9058
+    },
+    {
+      "epoch": 13.561377245508982,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 9059
+    },
+    {
+      "epoch": 13.562874251497005,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1369,
+      "step": 9060
+    },
+    {
+      "epoch": 13.56437125748503,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1366,
+      "step": 9061
+    },
+    {
+      "epoch": 13.565868263473053,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 9062
+    },
+    {
+      "epoch": 13.567365269461078,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1416,
+      "step": 9063
+    },
+    {
+      "epoch": 13.568862275449101,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 9064
+    },
+    {
+      "epoch": 13.570359281437126,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1344,
+      "step": 9065
+    },
+    {
+      "epoch": 13.571856287425149,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1365,
+      "step": 9066
+    },
+    {
+      "epoch": 13.573353293413174,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 9067
+    },
+    {
+      "epoch": 13.574850299401197,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1364,
+      "step": 9068
+    },
+    {
+      "epoch": 13.576347305389222,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1373,
+      "step": 9069
+    },
+    {
+      "epoch": 13.577844311377245,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1343,
+      "step": 9070
+    },
+    {
+      "epoch": 13.57934131736527,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 9071
+    },
+    {
+      "epoch": 13.580838323353294,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1323,
+      "step": 9072
+    },
+    {
+      "epoch": 13.582335329341317,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 9073
+    },
+    {
+      "epoch": 13.58383233532934,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1356,
+      "step": 9074
+    },
+    {
+      "epoch": 13.585329341317365,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1368,
+      "step": 9075
+    },
+    {
+      "epoch": 13.58682634730539,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 9076
+    },
+    {
+      "epoch": 13.588323353293413,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1335,
+      "step": 9077
+    },
+    {
+      "epoch": 13.589820359281438,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 9078
+    },
+    {
+      "epoch": 13.591317365269461,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 9079
+    },
+    {
+      "epoch": 13.592814371257486,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1428,
+      "step": 9080
+    },
+    {
+      "epoch": 13.594311377245509,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 9081
+    },
+    {
+      "epoch": 13.595808383233534,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 9082
+    },
+    {
+      "epoch": 13.597305389221557,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1375,
+      "step": 9083
+    },
+    {
+      "epoch": 13.598802395209582,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 9084
+    },
+    {
+      "epoch": 13.600299401197605,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 9085
+    },
+    {
+      "epoch": 13.60179640718563,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 9086
+    },
+    {
+      "epoch": 13.603293413173652,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.13,
+      "step": 9087
+    },
+    {
+      "epoch": 13.604790419161677,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 9088
+    },
+    {
+      "epoch": 13.6062874251497,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1377,
+      "step": 9089
+    },
+    {
+      "epoch": 13.607784431137725,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1353,
+      "step": 9090
+    },
+    {
+      "epoch": 13.609281437125748,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1343,
+      "step": 9091
+    },
+    {
+      "epoch": 13.610778443113773,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 9092
+    },
+    {
+      "epoch": 13.612275449101796,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 9093
+    },
+    {
+      "epoch": 13.613772455089821,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1392,
+      "step": 9094
+    },
+    {
+      "epoch": 13.615269461077844,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 9095
+    },
+    {
+      "epoch": 13.616766467065869,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1312,
+      "step": 9096
+    },
+    {
+      "epoch": 13.618263473053892,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1372,
+      "step": 9097
+    },
+    {
+      "epoch": 13.619760479041917,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 9098
+    },
+    {
+      "epoch": 13.62125748502994,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1381,
+      "step": 9099
+    },
+    {
+      "epoch": 13.622754491017965,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 9100
+    },
+    {
+      "epoch": 13.624251497005988,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 9101
+    },
+    {
+      "epoch": 13.625748502994012,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 9102
+    },
+    {
+      "epoch": 13.627245508982035,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1378,
+      "step": 9103
+    },
+    {
+      "epoch": 13.62874251497006,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1338,
+      "step": 9104
+    },
+    {
+      "epoch": 13.630239520958083,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1357,
+      "step": 9105
+    },
+    {
+      "epoch": 13.631736526946108,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 9106
+    },
+    {
+      "epoch": 13.633233532934131,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 9107
+    },
+    {
+      "epoch": 13.634730538922156,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 9108
+    },
+    {
+      "epoch": 13.636227544910179,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 9109
+    },
+    {
+      "epoch": 13.637724550898204,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1362,
+      "step": 9110
+    },
+    {
+      "epoch": 13.639221556886227,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1328,
+      "step": 9111
+    },
+    {
+      "epoch": 13.640718562874252,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1367,
+      "step": 9112
+    },
+    {
+      "epoch": 13.642215568862275,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.137,
+      "step": 9113
+    },
+    {
+      "epoch": 13.6437125748503,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1384,
+      "step": 9114
+    },
+    {
+      "epoch": 13.645209580838323,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 9115
+    },
+    {
+      "epoch": 13.646706586826348,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1428,
+      "step": 9116
+    },
+    {
+      "epoch": 13.64820359281437,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1361,
+      "step": 9117
+    },
+    {
+      "epoch": 13.649700598802395,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1413,
+      "step": 9118
+    },
+    {
+      "epoch": 13.651197604790418,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.137,
+      "step": 9119
+    },
+    {
+      "epoch": 13.652694610778443,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1358,
+      "step": 9120
+    },
+    {
+      "epoch": 13.654191616766466,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 9121
+    },
+    {
+      "epoch": 13.655688622754491,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1365,
+      "step": 9122
+    },
+    {
+      "epoch": 13.657185628742514,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1383,
+      "step": 9123
+    },
+    {
+      "epoch": 13.658682634730539,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 9124
+    },
+    {
+      "epoch": 13.660179640718562,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1354,
+      "step": 9125
+    },
+    {
+      "epoch": 13.661676646706587,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 9126
+    },
+    {
+      "epoch": 13.66317365269461,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 9127
+    },
+    {
+      "epoch": 13.664670658682635,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1355,
+      "step": 9128
+    },
+    {
+      "epoch": 13.66616766467066,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1381,
+      "step": 9129
+    },
+    {
+      "epoch": 13.667664670658683,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1491,
+      "step": 9130
+    },
+    {
+      "epoch": 13.669161676646706,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0008,
+      "loss": 1.152,
+      "step": 9131
+    },
+    {
+      "epoch": 13.67065868263473,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1754,
+      "step": 9132
+    },
+    {
+      "epoch": 13.672155688622755,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0008,
+      "loss": 1.155,
+      "step": 9133
+    },
+    {
+      "epoch": 13.673652694610778,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1507,
+      "step": 9134
+    },
+    {
+      "epoch": 13.675149700598803,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1575,
+      "step": 9135
+    },
+    {
+      "epoch": 13.676646706586826,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1395,
+      "step": 9136
+    },
+    {
+      "epoch": 13.678143712574851,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1566,
+      "step": 9137
+    },
+    {
+      "epoch": 13.679640718562874,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1481,
+      "step": 9138
+    },
+    {
+      "epoch": 13.681137724550899,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1555,
+      "step": 9139
+    },
+    {
+      "epoch": 13.682634730538922,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0008,
+      "loss": 1.155,
+      "step": 9140
+    },
+    {
+      "epoch": 13.684131736526947,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1502,
+      "step": 9141
+    },
+    {
+      "epoch": 13.68562874251497,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 9142
+    },
+    {
+      "epoch": 13.687125748502995,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 9143
+    },
+    {
+      "epoch": 13.688622754491018,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1551,
+      "step": 9144
+    },
+    {
+      "epoch": 13.690119760479043,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 9145
+    },
+    {
+      "epoch": 13.691616766467066,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1416,
+      "step": 9146
+    },
+    {
+      "epoch": 13.69311377245509,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 9147
+    },
+    {
+      "epoch": 13.694610778443113,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.148,
+      "step": 9148
+    },
+    {
+      "epoch": 13.696107784431138,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 9149
+    },
+    {
+      "epoch": 13.697604790419161,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1429,
+      "step": 9150
+    },
+    {
+      "epoch": 13.699101796407186,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1487,
+      "step": 9151
+    },
+    {
+      "epoch": 13.70059880239521,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1416,
+      "step": 9152
+    },
+    {
+      "epoch": 13.702095808383234,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1483,
+      "step": 9153
+    },
+    {
+      "epoch": 13.703592814371257,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1328,
+      "step": 9154
+    },
+    {
+      "epoch": 13.705089820359282,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1369,
+      "step": 9155
+    },
+    {
+      "epoch": 13.706586826347305,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 9156
+    },
+    {
+      "epoch": 13.70808383233533,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1471,
+      "step": 9157
+    },
+    {
+      "epoch": 13.709580838323353,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1443,
+      "step": 9158
+    },
+    {
+      "epoch": 13.711077844311378,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 9159
+    },
+    {
+      "epoch": 13.7125748502994,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 9160
+    },
+    {
+      "epoch": 13.714071856287426,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1352,
+      "step": 9161
+    },
+    {
+      "epoch": 13.715568862275449,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 9162
+    },
+    {
+      "epoch": 13.717065868263473,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 9163
+    },
+    {
+      "epoch": 13.718562874251496,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 9164
+    },
+    {
+      "epoch": 13.720059880239521,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1519,
+      "step": 9165
+    },
+    {
+      "epoch": 13.721556886227544,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 9166
+    },
+    {
+      "epoch": 13.72305389221557,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 9167
+    },
+    {
+      "epoch": 13.724550898203592,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.14,
+      "step": 9168
+    },
+    {
+      "epoch": 13.726047904191617,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1366,
+      "step": 9169
+    },
+    {
+      "epoch": 13.72754491017964,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 9170
+    },
+    {
+      "epoch": 13.729041916167665,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 9171
+    },
+    {
+      "epoch": 13.730538922155688,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 9172
+    },
+    {
+      "epoch": 13.732035928143713,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 9173
+    },
+    {
+      "epoch": 13.733532934131736,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 9174
+    },
+    {
+      "epoch": 13.73502994011976,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 9175
+    },
+    {
+      "epoch": 13.736526946107784,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 9176
+    },
+    {
+      "epoch": 13.738023952095809,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1393,
+      "step": 9177
+    },
+    {
+      "epoch": 13.739520958083832,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 9178
+    },
+    {
+      "epoch": 13.741017964071856,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1413,
+      "step": 9179
+    },
+    {
+      "epoch": 13.74251497005988,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1384,
+      "step": 9180
+    },
+    {
+      "epoch": 13.744011976047904,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 9181
+    },
+    {
+      "epoch": 13.745508982035929,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 9182
+    },
+    {
+      "epoch": 13.747005988023952,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1343,
+      "step": 9183
+    },
+    {
+      "epoch": 13.748502994011975,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 9184
+    },
+    {
+      "epoch": 13.75,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.134,
+      "step": 9185
+    },
+    {
+      "epoch": 13.751497005988025,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 9186
+    },
+    {
+      "epoch": 13.752994011976048,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 9187
+    },
+    {
+      "epoch": 13.754491017964071,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 9188
+    },
+    {
+      "epoch": 13.755988023952096,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 9189
+    },
+    {
+      "epoch": 13.75748502994012,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1342,
+      "step": 9190
+    },
+    {
+      "epoch": 13.758982035928144,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 9191
+    },
+    {
+      "epoch": 13.760479041916168,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 9192
+    },
+    {
+      "epoch": 13.761976047904191,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1321,
+      "step": 9193
+    },
+    {
+      "epoch": 13.763473053892216,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1395,
+      "step": 9194
+    },
+    {
+      "epoch": 13.76497005988024,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1366,
+      "step": 9195
+    },
+    {
+      "epoch": 13.766467065868264,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1384,
+      "step": 9196
+    },
+    {
+      "epoch": 13.767964071856287,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 9197
+    },
+    {
+      "epoch": 13.769461077844312,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 9198
+    },
+    {
+      "epoch": 13.770958083832335,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 9199
+    },
+    {
+      "epoch": 13.77245508982036,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 9200
+    },
+    {
+      "epoch": 13.773952095808383,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1395,
+      "step": 9201
+    },
+    {
+      "epoch": 13.775449101796408,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 9202
+    },
+    {
+      "epoch": 13.77694610778443,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 9203
+    },
+    {
+      "epoch": 13.778443113772456,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 9204
+    },
+    {
+      "epoch": 13.779940119760479,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 9205
+    },
+    {
+      "epoch": 13.781437125748504,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1476,
+      "step": 9206
+    },
+    {
+      "epoch": 13.782934131736527,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 9207
+    },
+    {
+      "epoch": 13.784431137724551,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1356,
+      "step": 9208
+    },
+    {
+      "epoch": 13.785928143712574,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1387,
+      "step": 9209
+    },
+    {
+      "epoch": 13.7874251497006,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 9210
+    },
+    {
+      "epoch": 13.788922155688622,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 9211
+    },
+    {
+      "epoch": 13.790419161676647,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1362,
+      "step": 9212
+    },
+    {
+      "epoch": 13.79191616766467,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 9213
+    },
+    {
+      "epoch": 13.793413173652695,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1373,
+      "step": 9214
+    },
+    {
+      "epoch": 13.794910179640718,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1326,
+      "step": 9215
+    },
+    {
+      "epoch": 13.796407185628743,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 9216
+    },
+    {
+      "epoch": 13.797904191616766,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 9217
+    },
+    {
+      "epoch": 13.79940119760479,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 9218
+    },
+    {
+      "epoch": 13.800898203592814,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1354,
+      "step": 9219
+    },
+    {
+      "epoch": 13.802395209580839,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1359,
+      "step": 9220
+    },
+    {
+      "epoch": 13.803892215568862,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1429,
+      "step": 9221
+    },
+    {
+      "epoch": 13.805389221556887,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1326,
+      "step": 9222
+    },
+    {
+      "epoch": 13.80688622754491,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 9223
+    },
+    {
+      "epoch": 13.808383233532934,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 9224
+    },
+    {
+      "epoch": 13.809880239520957,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 9225
+    },
+    {
+      "epoch": 13.811377245508982,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 9226
+    },
+    {
+      "epoch": 13.812874251497005,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1319,
+      "step": 9227
+    },
+    {
+      "epoch": 13.81437125748503,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 9228
+    },
+    {
+      "epoch": 13.815868263473053,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 9229
+    },
+    {
+      "epoch": 13.817365269461078,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 9230
+    },
+    {
+      "epoch": 13.818862275449101,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 9231
+    },
+    {
+      "epoch": 13.820359281437126,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1344,
+      "step": 9232
+    },
+    {
+      "epoch": 13.821856287425149,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 9233
+    },
+    {
+      "epoch": 13.823353293413174,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1416,
+      "step": 9234
+    },
+    {
+      "epoch": 13.824850299401197,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.146,
+      "step": 9235
+    },
+    {
+      "epoch": 13.826347305389222,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1356,
+      "step": 9236
+    },
+    {
+      "epoch": 13.827844311377245,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1454,
+      "step": 9237
+    },
+    {
+      "epoch": 13.82934131736527,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 9238
+    },
+    {
+      "epoch": 13.830838323353294,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 9239
+    },
+    {
+      "epoch": 13.832335329341317,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1395,
+      "step": 9240
+    },
+    {
+      "epoch": 13.83383233532934,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 9241
+    },
+    {
+      "epoch": 13.835329341317365,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 9242
+    },
+    {
+      "epoch": 13.83682634730539,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 9243
+    },
+    {
+      "epoch": 13.838323353293413,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 9244
+    },
+    {
+      "epoch": 13.839820359281438,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 9245
+    },
+    {
+      "epoch": 13.841317365269461,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 9246
+    },
+    {
+      "epoch": 13.842814371257486,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 9247
+    },
+    {
+      "epoch": 13.844311377245509,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 9248
+    },
+    {
+      "epoch": 13.845808383233534,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1346,
+      "step": 9249
+    },
+    {
+      "epoch": 13.847305389221557,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 9250
+    },
+    {
+      "epoch": 13.848802395209582,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 9251
+    },
+    {
+      "epoch": 13.850299401197605,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1362,
+      "step": 9252
+    },
+    {
+      "epoch": 13.85179640718563,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 9253
+    },
+    {
+      "epoch": 13.853293413173652,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1324,
+      "step": 9254
+    },
+    {
+      "epoch": 13.854790419161677,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1349,
+      "step": 9255
+    },
+    {
+      "epoch": 13.8562874251497,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1375,
+      "step": 9256
+    },
+    {
+      "epoch": 13.857784431137725,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 9257
+    },
+    {
+      "epoch": 13.859281437125748,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1379,
+      "step": 9258
+    },
+    {
+      "epoch": 13.860778443113773,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1471,
+      "step": 9259
+    },
+    {
+      "epoch": 13.862275449101796,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1367,
+      "step": 9260
+    },
+    {
+      "epoch": 13.863772455089821,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1434,
+      "step": 9261
+    },
+    {
+      "epoch": 13.865269461077844,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 9262
+    },
+    {
+      "epoch": 13.866766467065869,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.137,
+      "step": 9263
+    },
+    {
+      "epoch": 13.868263473053892,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 9264
+    },
+    {
+      "epoch": 13.869760479041917,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1379,
+      "step": 9265
+    },
+    {
+      "epoch": 13.87125748502994,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 9266
+    },
+    {
+      "epoch": 13.872754491017965,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1368,
+      "step": 9267
+    },
+    {
+      "epoch": 13.874251497005988,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1338,
+      "step": 9268
+    },
+    {
+      "epoch": 13.875748502994012,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1381,
+      "step": 9269
+    },
+    {
+      "epoch": 13.877245508982035,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 9270
+    },
+    {
+      "epoch": 13.87874251497006,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1376,
+      "step": 9271
+    },
+    {
+      "epoch": 13.880239520958083,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.137,
+      "step": 9272
+    },
+    {
+      "epoch": 13.881736526946108,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1384,
+      "step": 9273
+    },
+    {
+      "epoch": 13.883233532934131,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1363,
+      "step": 9274
+    },
+    {
+      "epoch": 13.884730538922156,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1368,
+      "step": 9275
+    },
+    {
+      "epoch": 13.886227544910179,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 9276
+    },
+    {
+      "epoch": 13.887724550898204,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1352,
+      "step": 9277
+    },
+    {
+      "epoch": 13.889221556886227,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 9278
+    },
+    {
+      "epoch": 13.890718562874252,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1379,
+      "step": 9279
+    },
+    {
+      "epoch": 13.892215568862275,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1351,
+      "step": 9280
+    },
+    {
+      "epoch": 13.8937125748503,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 9281
+    },
+    {
+      "epoch": 13.895209580838323,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1361,
+      "step": 9282
+    },
+    {
+      "epoch": 13.896706586826348,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1371,
+      "step": 9283
+    },
+    {
+      "epoch": 13.89820359281437,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 9284
+    },
+    {
+      "epoch": 13.899700598802395,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 9285
+    },
+    {
+      "epoch": 13.901197604790418,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1354,
+      "step": 9286
+    },
+    {
+      "epoch": 13.902694610778443,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1333,
+      "step": 9287
+    },
+    {
+      "epoch": 13.904191616766466,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 9288
+    },
+    {
+      "epoch": 13.905688622754491,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 9289
+    },
+    {
+      "epoch": 13.907185628742514,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1354,
+      "step": 9290
+    },
+    {
+      "epoch": 13.908682634730539,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1381,
+      "step": 9291
+    },
+    {
+      "epoch": 13.910179640718562,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 9292
+    },
+    {
+      "epoch": 13.911676646706587,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 9293
+    },
+    {
+      "epoch": 13.91317365269461,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 9294
+    },
+    {
+      "epoch": 13.914670658682635,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 9295
+    },
+    {
+      "epoch": 13.91616766467066,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 9296
+    },
+    {
+      "epoch": 13.917664670658683,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 9297
+    },
+    {
+      "epoch": 13.919161676646706,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 9298
+    },
+    {
+      "epoch": 13.92065868263473,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 9299
+    },
+    {
+      "epoch": 13.922155688622755,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1393,
+      "step": 9300
+    },
+    {
+      "epoch": 13.923652694610778,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 9301
+    },
+    {
+      "epoch": 13.925149700598803,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1355,
+      "step": 9302
+    },
+    {
+      "epoch": 13.926646706586826,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1317,
+      "step": 9303
+    },
+    {
+      "epoch": 13.928143712574851,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1378,
+      "step": 9304
+    },
+    {
+      "epoch": 13.929640718562874,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1366,
+      "step": 9305
+    },
+    {
+      "epoch": 13.931137724550899,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1416,
+      "step": 9306
+    },
+    {
+      "epoch": 13.932634730538922,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 9307
+    },
+    {
+      "epoch": 13.934131736526947,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 9308
+    },
+    {
+      "epoch": 13.93562874251497,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1476,
+      "step": 9309
+    },
+    {
+      "epoch": 13.937125748502995,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 9310
+    },
+    {
+      "epoch": 13.938622754491018,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 9311
+    },
+    {
+      "epoch": 13.940119760479043,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 9312
+    },
+    {
+      "epoch": 13.941616766467066,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1334,
+      "step": 9313
+    },
+    {
+      "epoch": 13.94311377245509,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1416,
+      "step": 9314
+    },
+    {
+      "epoch": 13.944610778443113,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 9315
+    },
+    {
+      "epoch": 13.946107784431138,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1352,
+      "step": 9316
+    },
+    {
+      "epoch": 13.947604790419161,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 9317
+    },
+    {
+      "epoch": 13.949101796407186,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 9318
+    },
+    {
+      "epoch": 13.95059880239521,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1363,
+      "step": 9319
+    },
+    {
+      "epoch": 13.952095808383234,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 9320
+    },
+    {
+      "epoch": 13.953592814371257,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1354,
+      "step": 9321
+    },
+    {
+      "epoch": 13.955089820359282,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1367,
+      "step": 9322
+    },
+    {
+      "epoch": 13.956586826347305,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 9323
+    },
+    {
+      "epoch": 13.95808383233533,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 9324
+    },
+    {
+      "epoch": 13.959580838323353,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 9325
+    },
+    {
+      "epoch": 13.961077844311378,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1381,
+      "step": 9326
+    },
+    {
+      "epoch": 13.9625748502994,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 9327
+    },
+    {
+      "epoch": 13.964071856287426,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 9328
+    },
+    {
+      "epoch": 13.965568862275449,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1303,
+      "step": 9329
+    },
+    {
+      "epoch": 13.967065868263473,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.131,
+      "step": 9330
+    },
+    {
+      "epoch": 13.968562874251496,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1366,
+      "step": 9331
+    },
+    {
+      "epoch": 13.970059880239521,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 9332
+    },
+    {
+      "epoch": 13.971556886227544,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1364,
+      "step": 9333
+    },
+    {
+      "epoch": 13.97305389221557,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 9334
+    },
+    {
+      "epoch": 13.974550898203592,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 9335
+    },
+    {
+      "epoch": 13.976047904191617,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 9336
+    },
+    {
+      "epoch": 13.97754491017964,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 9337
+    },
+    {
+      "epoch": 13.979041916167665,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1342,
+      "step": 9338
+    },
+    {
+      "epoch": 13.980538922155688,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1392,
+      "step": 9339
+    },
+    {
+      "epoch": 13.982035928143713,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1346,
+      "step": 9340
+    },
+    {
+      "epoch": 13.983532934131736,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1364,
+      "step": 9341
+    },
+    {
+      "epoch": 13.98502994011976,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 9342
+    },
+    {
+      "epoch": 13.986526946107784,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 9343
+    },
+    {
+      "epoch": 13.988023952095809,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 9344
+    },
+    {
+      "epoch": 13.989520958083832,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1386,
+      "step": 9345
+    },
+    {
+      "epoch": 13.991017964071856,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 9346
+    },
+    {
+      "epoch": 13.99251497005988,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1476,
+      "step": 9347
+    },
+    {
+      "epoch": 13.994011976047904,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 9348
+    },
+    {
+      "epoch": 13.995508982035929,
+      "grad_norm": 0.05078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 9349
+    },
+    {
+      "epoch": 13.997005988023952,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 9350
+    },
+    {
+      "epoch": 13.998502994011975,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 9351
+    },
+    {
+      "epoch": 14.0,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1443,
+      "step": 9352
+    },
+    {
+      "epoch": 14.001497005988025,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1446,
+      "step": 9353
+    },
+    {
+      "epoch": 14.002994011976048,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1345,
+      "step": 9354
+    },
+    {
+      "epoch": 14.004491017964073,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1434,
+      "step": 9355
+    },
+    {
+      "epoch": 14.005988023952096,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1307,
+      "step": 9356
+    },
+    {
+      "epoch": 14.00748502994012,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 9357
+    },
+    {
+      "epoch": 14.008982035928144,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1299,
+      "step": 9358
+    },
+    {
+      "epoch": 14.010479041916168,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1316,
+      "step": 9359
+    },
+    {
+      "epoch": 14.011976047904191,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1351,
+      "step": 9360
+    },
+    {
+      "epoch": 14.013473053892216,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.146,
+      "step": 9361
+    },
+    {
+      "epoch": 14.01497005988024,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 9362
+    },
+    {
+      "epoch": 14.016467065868264,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 9363
+    },
+    {
+      "epoch": 14.017964071856287,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 9364
+    },
+    {
+      "epoch": 14.019461077844312,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 9365
+    },
+    {
+      "epoch": 14.020958083832335,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1395,
+      "step": 9366
+    },
+    {
+      "epoch": 14.02245508982036,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1459,
+      "step": 9367
+    },
+    {
+      "epoch": 14.023952095808383,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1393,
+      "step": 9368
+    },
+    {
+      "epoch": 14.025449101796408,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 9369
+    },
+    {
+      "epoch": 14.02694610778443,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1353,
+      "step": 9370
+    },
+    {
+      "epoch": 14.028443113772456,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 9371
+    },
+    {
+      "epoch": 14.029940119760479,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.136,
+      "step": 9372
+    },
+    {
+      "epoch": 14.031437125748504,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1351,
+      "step": 9373
+    },
+    {
+      "epoch": 14.032934131736527,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1334,
+      "step": 9374
+    },
+    {
+      "epoch": 14.034431137724551,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1345,
+      "step": 9375
+    },
+    {
+      "epoch": 14.035928143712574,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1366,
+      "step": 9376
+    },
+    {
+      "epoch": 14.0374251497006,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 9377
+    },
+    {
+      "epoch": 14.038922155688622,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1481,
+      "step": 9378
+    },
+    {
+      "epoch": 14.040419161676647,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1352,
+      "step": 9379
+    },
+    {
+      "epoch": 14.04191616766467,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.137,
+      "step": 9380
+    },
+    {
+      "epoch": 14.043413173652695,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1364,
+      "step": 9381
+    },
+    {
+      "epoch": 14.044910179640718,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1347,
+      "step": 9382
+    },
+    {
+      "epoch": 14.046407185628743,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 9383
+    },
+    {
+      "epoch": 14.047904191616766,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1443,
+      "step": 9384
+    },
+    {
+      "epoch": 14.04940119760479,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1371,
+      "step": 9385
+    },
+    {
+      "epoch": 14.050898203592814,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 9386
+    },
+    {
+      "epoch": 14.052395209580839,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 9387
+    },
+    {
+      "epoch": 14.053892215568862,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1349,
+      "step": 9388
+    },
+    {
+      "epoch": 14.055389221556887,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1346,
+      "step": 9389
+    },
+    {
+      "epoch": 14.05688622754491,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 9390
+    },
+    {
+      "epoch": 14.058383233532934,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 9391
+    },
+    {
+      "epoch": 14.059880239520957,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1346,
+      "step": 9392
+    },
+    {
+      "epoch": 14.061377245508982,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 9393
+    },
+    {
+      "epoch": 14.062874251497005,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1393,
+      "step": 9394
+    },
+    {
+      "epoch": 14.06437125748503,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.132,
+      "step": 9395
+    },
+    {
+      "epoch": 14.065868263473053,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 9396
+    },
+    {
+      "epoch": 14.067365269461078,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1372,
+      "step": 9397
+    },
+    {
+      "epoch": 14.068862275449101,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1393,
+      "step": 9398
+    },
+    {
+      "epoch": 14.070359281437126,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1343,
+      "step": 9399
+    },
+    {
+      "epoch": 14.071856287425149,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1429,
+      "step": 9400
+    },
+    {
+      "epoch": 14.073353293413174,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1356,
+      "step": 9401
+    },
+    {
+      "epoch": 14.074850299401197,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 9402
+    },
+    {
+      "epoch": 14.076347305389222,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 9403
+    },
+    {
+      "epoch": 14.077844311377245,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 9404
+    },
+    {
+      "epoch": 14.07934131736527,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 9405
+    },
+    {
+      "epoch": 14.080838323353293,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1329,
+      "step": 9406
+    },
+    {
+      "epoch": 14.082335329341317,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 9407
+    },
+    {
+      "epoch": 14.08383233532934,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1353,
+      "step": 9408
+    },
+    {
+      "epoch": 14.085329341317365,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1465,
+      "step": 9409
+    },
+    {
+      "epoch": 14.08682634730539,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1302,
+      "step": 9410
+    },
+    {
+      "epoch": 14.088323353293413,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 9411
+    },
+    {
+      "epoch": 14.089820359281438,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 9412
+    },
+    {
+      "epoch": 14.091317365269461,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1413,
+      "step": 9413
+    },
+    {
+      "epoch": 14.092814371257486,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 9414
+    },
+    {
+      "epoch": 14.094311377245509,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 9415
+    },
+    {
+      "epoch": 14.095808383233534,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 9416
+    },
+    {
+      "epoch": 14.097305389221557,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 9417
+    },
+    {
+      "epoch": 14.098802395209582,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 9418
+    },
+    {
+      "epoch": 14.100299401197605,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1459,
+      "step": 9419
+    },
+    {
+      "epoch": 14.10179640718563,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1376,
+      "step": 9420
+    },
+    {
+      "epoch": 14.103293413173652,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1393,
+      "step": 9421
+    },
+    {
+      "epoch": 14.104790419161677,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 9422
+    },
+    {
+      "epoch": 14.1062874251497,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1363,
+      "step": 9423
+    },
+    {
+      "epoch": 14.107784431137725,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 9424
+    },
+    {
+      "epoch": 14.109281437125748,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 9425
+    },
+    {
+      "epoch": 14.110778443113773,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1368,
+      "step": 9426
+    },
+    {
+      "epoch": 14.112275449101796,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1358,
+      "step": 9427
+    },
+    {
+      "epoch": 14.113772455089821,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 9428
+    },
+    {
+      "epoch": 14.115269461077844,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1377,
+      "step": 9429
+    },
+    {
+      "epoch": 14.116766467065869,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 9430
+    },
+    {
+      "epoch": 14.118263473053892,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1316,
+      "step": 9431
+    },
+    {
+      "epoch": 14.119760479041917,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 9432
+    },
+    {
+      "epoch": 14.12125748502994,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 9433
+    },
+    {
+      "epoch": 14.122754491017965,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0008,
+      "loss": 1.134,
+      "step": 9434
+    },
+    {
+      "epoch": 14.124251497005988,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 9435
+    },
+    {
+      "epoch": 14.125748502994012,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1376,
+      "step": 9436
+    },
+    {
+      "epoch": 14.127245508982035,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 9437
+    },
+    {
+      "epoch": 14.12874251497006,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 9438
+    },
+    {
+      "epoch": 14.130239520958083,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1434,
+      "step": 9439
+    },
+    {
+      "epoch": 14.131736526946108,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1336,
+      "step": 9440
+    },
+    {
+      "epoch": 14.133233532934131,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 9441
+    },
+    {
+      "epoch": 14.134730538922156,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 9442
+    },
+    {
+      "epoch": 14.136227544910179,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 9443
+    },
+    {
+      "epoch": 14.137724550898204,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1355,
+      "step": 9444
+    },
+    {
+      "epoch": 14.139221556886227,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 9445
+    },
+    {
+      "epoch": 14.140718562874252,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.134,
+      "step": 9446
+    },
+    {
+      "epoch": 14.142215568862275,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 9447
+    },
+    {
+      "epoch": 14.1437125748503,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 9448
+    },
+    {
+      "epoch": 14.145209580838323,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1377,
+      "step": 9449
+    },
+    {
+      "epoch": 14.146706586826348,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 9450
+    },
+    {
+      "epoch": 14.14820359281437,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1324,
+      "step": 9451
+    },
+    {
+      "epoch": 14.149700598802395,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1476,
+      "step": 9452
+    },
+    {
+      "epoch": 14.151197604790418,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1304,
+      "step": 9453
+    },
+    {
+      "epoch": 14.152694610778443,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1375,
+      "step": 9454
+    },
+    {
+      "epoch": 14.154191616766466,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1378,
+      "step": 9455
+    },
+    {
+      "epoch": 14.155688622754491,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 9456
+    },
+    {
+      "epoch": 14.157185628742514,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1343,
+      "step": 9457
+    },
+    {
+      "epoch": 14.158682634730539,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 9458
+    },
+    {
+      "epoch": 14.160179640718562,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1384,
+      "step": 9459
+    },
+    {
+      "epoch": 14.161676646706587,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1363,
+      "step": 9460
+    },
+    {
+      "epoch": 14.16317365269461,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.135,
+      "step": 9461
+    },
+    {
+      "epoch": 14.164670658682635,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 9462
+    },
+    {
+      "epoch": 14.16616766467066,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1367,
+      "step": 9463
+    },
+    {
+      "epoch": 14.167664670658683,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1416,
+      "step": 9464
+    },
+    {
+      "epoch": 14.169161676646707,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1359,
+      "step": 9465
+    },
+    {
+      "epoch": 14.17065868263473,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.136,
+      "step": 9466
+    },
+    {
+      "epoch": 14.172155688622755,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1359,
+      "step": 9467
+    },
+    {
+      "epoch": 14.173652694610778,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1371,
+      "step": 9468
+    },
+    {
+      "epoch": 14.175149700598803,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 9469
+    },
+    {
+      "epoch": 14.176646706586826,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1369,
+      "step": 9470
+    },
+    {
+      "epoch": 14.178143712574851,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1362,
+      "step": 9471
+    },
+    {
+      "epoch": 14.179640718562874,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1377,
+      "step": 9472
+    },
+    {
+      "epoch": 14.181137724550899,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 9473
+    },
+    {
+      "epoch": 14.182634730538922,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 9474
+    },
+    {
+      "epoch": 14.184131736526947,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 9475
+    },
+    {
+      "epoch": 14.18562874251497,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 9476
+    },
+    {
+      "epoch": 14.187125748502995,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1366,
+      "step": 9477
+    },
+    {
+      "epoch": 14.188622754491018,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1338,
+      "step": 9478
+    },
+    {
+      "epoch": 14.190119760479043,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1355,
+      "step": 9479
+    },
+    {
+      "epoch": 14.191616766467066,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1355,
+      "step": 9480
+    },
+    {
+      "epoch": 14.19311377245509,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1355,
+      "step": 9481
+    },
+    {
+      "epoch": 14.194610778443113,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1369,
+      "step": 9482
+    },
+    {
+      "epoch": 14.196107784431138,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1369,
+      "step": 9483
+    },
+    {
+      "epoch": 14.197604790419161,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1334,
+      "step": 9484
+    },
+    {
+      "epoch": 14.199101796407186,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1369,
+      "step": 9485
+    },
+    {
+      "epoch": 14.20059880239521,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1366,
+      "step": 9486
+    },
+    {
+      "epoch": 14.202095808383234,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 9487
+    },
+    {
+      "epoch": 14.203592814371257,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1392,
+      "step": 9488
+    },
+    {
+      "epoch": 14.205089820359282,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1368,
+      "step": 9489
+    },
+    {
+      "epoch": 14.206586826347305,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 9490
+    },
+    {
+      "epoch": 14.20808383233533,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1434,
+      "step": 9491
+    },
+    {
+      "epoch": 14.209580838323353,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1362,
+      "step": 9492
+    },
+    {
+      "epoch": 14.211077844311378,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 9493
+    },
+    {
+      "epoch": 14.2125748502994,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1352,
+      "step": 9494
+    },
+    {
+      "epoch": 14.214071856287426,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.136,
+      "step": 9495
+    },
+    {
+      "epoch": 14.215568862275449,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 9496
+    },
+    {
+      "epoch": 14.217065868263473,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1337,
+      "step": 9497
+    },
+    {
+      "epoch": 14.218562874251496,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1379,
+      "step": 9498
+    },
+    {
+      "epoch": 14.220059880239521,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 9499
+    },
+    {
+      "epoch": 14.221556886227544,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1276,
+      "step": 9500
+    },
+    {
+      "epoch": 14.22305389221557,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1372,
+      "step": 9501
+    },
+    {
+      "epoch": 14.224550898203592,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1443,
+      "step": 9502
+    },
+    {
+      "epoch": 14.226047904191617,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.137,
+      "step": 9503
+    },
+    {
+      "epoch": 14.22754491017964,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0008,
+      "loss": 1.1429,
+      "step": 9504
+    },
+    {
+      "epoch": 14.229041916167665,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 9505
+    },
+    {
+      "epoch": 14.230538922155688,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 9506
+    },
+    {
+      "epoch": 14.232035928143713,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1337,
+      "step": 9507
+    },
+    {
+      "epoch": 14.233532934131736,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 9508
+    },
+    {
+      "epoch": 14.23502994011976,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1341,
+      "step": 9509
+    },
+    {
+      "epoch": 14.236526946107784,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1339,
+      "step": 9510
+    },
+    {
+      "epoch": 14.238023952095809,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1353,
+      "step": 9511
+    },
+    {
+      "epoch": 14.239520958083832,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 9512
+    },
+    {
+      "epoch": 14.241017964071856,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1336,
+      "step": 9513
+    },
+    {
+      "epoch": 14.24251497005988,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1348,
+      "step": 9514
+    },
+    {
+      "epoch": 14.244011976047904,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1326,
+      "step": 9515
+    },
+    {
+      "epoch": 14.245508982035927,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 9516
+    },
+    {
+      "epoch": 14.247005988023952,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 9517
+    },
+    {
+      "epoch": 14.248502994011975,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1319,
+      "step": 9518
+    },
+    {
+      "epoch": 14.25,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 9519
+    },
+    {
+      "epoch": 14.251497005988025,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1352,
+      "step": 9520
+    },
+    {
+      "epoch": 14.252994011976048,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1383,
+      "step": 9521
+    },
+    {
+      "epoch": 14.254491017964073,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1357,
+      "step": 9522
+    },
+    {
+      "epoch": 14.255988023952096,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 9523
+    },
+    {
+      "epoch": 14.25748502994012,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 9524
+    },
+    {
+      "epoch": 14.258982035928144,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 9525
+    },
+    {
+      "epoch": 14.260479041916168,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1341,
+      "step": 9526
+    },
+    {
+      "epoch": 14.261976047904191,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 9527
+    },
+    {
+      "epoch": 14.263473053892216,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1392,
+      "step": 9528
+    },
+    {
+      "epoch": 14.26497005988024,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 9529
+    },
+    {
+      "epoch": 14.266467065868264,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 9530
+    },
+    {
+      "epoch": 14.267964071856287,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1355,
+      "step": 9531
+    },
+    {
+      "epoch": 14.269461077844312,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1383,
+      "step": 9532
+    },
+    {
+      "epoch": 14.270958083832335,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 9533
+    },
+    {
+      "epoch": 14.27245508982036,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1347,
+      "step": 9534
+    },
+    {
+      "epoch": 14.273952095808383,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1364,
+      "step": 9535
+    },
+    {
+      "epoch": 14.275449101796408,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1376,
+      "step": 9536
+    },
+    {
+      "epoch": 14.27694610778443,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 9537
+    },
+    {
+      "epoch": 14.278443113772456,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.137,
+      "step": 9538
+    },
+    {
+      "epoch": 14.279940119760479,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1341,
+      "step": 9539
+    },
+    {
+      "epoch": 14.281437125748504,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 9540
+    },
+    {
+      "epoch": 14.282934131736527,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1381,
+      "step": 9541
+    },
+    {
+      "epoch": 14.284431137724551,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 9542
+    },
+    {
+      "epoch": 14.285928143712574,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1296,
+      "step": 9543
+    },
+    {
+      "epoch": 14.2874251497006,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 9544
+    },
+    {
+      "epoch": 14.288922155688622,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 9545
+    },
+    {
+      "epoch": 14.290419161676647,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1343,
+      "step": 9546
+    },
+    {
+      "epoch": 14.29191616766467,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 9547
+    },
+    {
+      "epoch": 14.293413173652695,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1362,
+      "step": 9548
+    },
+    {
+      "epoch": 14.294910179640718,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1367,
+      "step": 9549
+    },
+    {
+      "epoch": 14.296407185628743,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 9550
+    },
+    {
+      "epoch": 14.297904191616766,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1333,
+      "step": 9551
+    },
+    {
+      "epoch": 14.29940119760479,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1347,
+      "step": 9552
+    },
+    {
+      "epoch": 14.300898203592814,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1379,
+      "step": 9553
+    },
+    {
+      "epoch": 14.302395209580839,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 9554
+    },
+    {
+      "epoch": 14.303892215568862,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1353,
+      "step": 9555
+    },
+    {
+      "epoch": 14.305389221556887,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 9556
+    },
+    {
+      "epoch": 14.30688622754491,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.14,
+      "step": 9557
+    },
+    {
+      "epoch": 14.308383233532934,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1351,
+      "step": 9558
+    },
+    {
+      "epoch": 14.309880239520957,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 9559
+    },
+    {
+      "epoch": 14.311377245508982,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 9560
+    },
+    {
+      "epoch": 14.312874251497005,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 9561
+    },
+    {
+      "epoch": 14.31437125748503,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1486,
+      "step": 9562
+    },
+    {
+      "epoch": 14.315868263473053,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1305,
+      "step": 9563
+    },
+    {
+      "epoch": 14.317365269461078,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1371,
+      "step": 9564
+    },
+    {
+      "epoch": 14.318862275449101,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1348,
+      "step": 9565
+    },
+    {
+      "epoch": 14.320359281437126,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1356,
+      "step": 9566
+    },
+    {
+      "epoch": 14.321856287425149,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1429,
+      "step": 9567
+    },
+    {
+      "epoch": 14.323353293413174,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1344,
+      "step": 9568
+    },
+    {
+      "epoch": 14.324850299401197,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1307,
+      "step": 9569
+    },
+    {
+      "epoch": 14.326347305389222,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0008,
+      "loss": 1.137,
+      "step": 9570
+    },
+    {
+      "epoch": 14.327844311377245,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1341,
+      "step": 9571
+    },
+    {
+      "epoch": 14.32934131736527,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 9572
+    },
+    {
+      "epoch": 14.330838323353294,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1395,
+      "step": 9573
+    },
+    {
+      "epoch": 14.332335329341317,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 9574
+    },
+    {
+      "epoch": 14.33383233532934,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1346,
+      "step": 9575
+    },
+    {
+      "epoch": 14.335329341317365,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1341,
+      "step": 9576
+    },
+    {
+      "epoch": 14.33682634730539,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 9577
+    },
+    {
+      "epoch": 14.338323353293413,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 9578
+    },
+    {
+      "epoch": 14.339820359281438,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 9579
+    },
+    {
+      "epoch": 14.341317365269461,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.134,
+      "step": 9580
+    },
+    {
+      "epoch": 14.342814371257486,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 9581
+    },
+    {
+      "epoch": 14.344311377245509,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1429,
+      "step": 9582
+    },
+    {
+      "epoch": 14.345808383233534,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 9583
+    },
+    {
+      "epoch": 14.347305389221557,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 9584
+    },
+    {
+      "epoch": 14.348802395209582,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1325,
+      "step": 9585
+    },
+    {
+      "epoch": 14.350299401197605,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1371,
+      "step": 9586
+    },
+    {
+      "epoch": 14.35179640718563,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1365,
+      "step": 9587
+    },
+    {
+      "epoch": 14.353293413173652,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1343,
+      "step": 9588
+    },
+    {
+      "epoch": 14.354790419161677,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1344,
+      "step": 9589
+    },
+    {
+      "epoch": 14.3562874251497,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1335,
+      "step": 9590
+    },
+    {
+      "epoch": 14.357784431137725,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 9591
+    },
+    {
+      "epoch": 14.359281437125748,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1369,
+      "step": 9592
+    },
+    {
+      "epoch": 14.360778443113773,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 9593
+    },
+    {
+      "epoch": 14.362275449101796,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1318,
+      "step": 9594
+    },
+    {
+      "epoch": 14.363772455089821,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1364,
+      "step": 9595
+    },
+    {
+      "epoch": 14.365269461077844,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1387,
+      "step": 9596
+    },
+    {
+      "epoch": 14.366766467065869,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 9597
+    },
+    {
+      "epoch": 14.368263473053892,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1355,
+      "step": 9598
+    },
+    {
+      "epoch": 14.369760479041917,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1375,
+      "step": 9599
+    },
+    {
+      "epoch": 14.37125748502994,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1352,
+      "step": 9600
+    },
+    {
+      "epoch": 14.372754491017965,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 9601
+    },
+    {
+      "epoch": 14.374251497005988,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1313,
+      "step": 9602
+    },
+    {
+      "epoch": 14.375748502994012,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 9603
+    },
+    {
+      "epoch": 14.377245508982035,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.135,
+      "step": 9604
+    },
+    {
+      "epoch": 14.37874251497006,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 9605
+    },
+    {
+      "epoch": 14.380239520958083,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 9606
+    },
+    {
+      "epoch": 14.381736526946108,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1368,
+      "step": 9607
+    },
+    {
+      "epoch": 14.383233532934131,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 9608
+    },
+    {
+      "epoch": 14.384730538922156,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1325,
+      "step": 9609
+    },
+    {
+      "epoch": 14.386227544910179,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1373,
+      "step": 9610
+    },
+    {
+      "epoch": 14.387724550898204,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1372,
+      "step": 9611
+    },
+    {
+      "epoch": 14.389221556886227,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1356,
+      "step": 9612
+    },
+    {
+      "epoch": 14.390718562874252,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 9613
+    },
+    {
+      "epoch": 14.392215568862275,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1368,
+      "step": 9614
+    },
+    {
+      "epoch": 14.3937125748503,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 9615
+    },
+    {
+      "epoch": 14.395209580838323,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 9616
+    },
+    {
+      "epoch": 14.396706586826348,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.131,
+      "step": 9617
+    },
+    {
+      "epoch": 14.39820359281437,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1369,
+      "step": 9618
+    },
+    {
+      "epoch": 14.399700598802395,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.129,
+      "step": 9619
+    },
+    {
+      "epoch": 14.401197604790418,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 9620
+    },
+    {
+      "epoch": 14.402694610778443,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 9621
+    },
+    {
+      "epoch": 14.404191616766466,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 9622
+    },
+    {
+      "epoch": 14.405688622754491,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1344,
+      "step": 9623
+    },
+    {
+      "epoch": 14.407185628742514,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 9624
+    },
+    {
+      "epoch": 14.408682634730539,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1352,
+      "step": 9625
+    },
+    {
+      "epoch": 14.410179640718562,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 9626
+    },
+    {
+      "epoch": 14.411676646706587,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1379,
+      "step": 9627
+    },
+    {
+      "epoch": 14.41317365269461,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1376,
+      "step": 9628
+    },
+    {
+      "epoch": 14.414670658682635,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1367,
+      "step": 9629
+    },
+    {
+      "epoch": 14.41616766467066,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1344,
+      "step": 9630
+    },
+    {
+      "epoch": 14.417664670658683,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 9631
+    },
+    {
+      "epoch": 14.419161676646706,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1352,
+      "step": 9632
+    },
+    {
+      "epoch": 14.42065868263473,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1367,
+      "step": 9633
+    },
+    {
+      "epoch": 14.422155688622755,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 9634
+    },
+    {
+      "epoch": 14.423652694610778,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1324,
+      "step": 9635
+    },
+    {
+      "epoch": 14.425149700598803,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1359,
+      "step": 9636
+    },
+    {
+      "epoch": 14.426646706586826,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1383,
+      "step": 9637
+    },
+    {
+      "epoch": 14.428143712574851,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 9638
+    },
+    {
+      "epoch": 14.429640718562874,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1387,
+      "step": 9639
+    },
+    {
+      "epoch": 14.431137724550899,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1371,
+      "step": 9640
+    },
+    {
+      "epoch": 14.432634730538922,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1343,
+      "step": 9641
+    },
+    {
+      "epoch": 14.434131736526947,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1361,
+      "step": 9642
+    },
+    {
+      "epoch": 14.43562874251497,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1369,
+      "step": 9643
+    },
+    {
+      "epoch": 14.437125748502995,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1481,
+      "step": 9644
+    },
+    {
+      "epoch": 14.438622754491018,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 9645
+    },
+    {
+      "epoch": 14.440119760479043,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1365,
+      "step": 9646
+    },
+    {
+      "epoch": 14.441616766467066,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1355,
+      "step": 9647
+    },
+    {
+      "epoch": 14.44311377245509,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1336,
+      "step": 9648
+    },
+    {
+      "epoch": 14.444610778443113,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1476,
+      "step": 9649
+    },
+    {
+      "epoch": 14.446107784431138,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1347,
+      "step": 9650
+    },
+    {
+      "epoch": 14.447604790419161,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1278,
+      "step": 9651
+    },
+    {
+      "epoch": 14.449101796407186,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 9652
+    },
+    {
+      "epoch": 14.45059880239521,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1304,
+      "step": 9653
+    },
+    {
+      "epoch": 14.452095808383234,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1384,
+      "step": 9654
+    },
+    {
+      "epoch": 14.453592814371257,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 9655
+    },
+    {
+      "epoch": 14.455089820359282,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1363,
+      "step": 9656
+    },
+    {
+      "epoch": 14.456586826347305,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 9657
+    },
+    {
+      "epoch": 14.45808383233533,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1352,
+      "step": 9658
+    },
+    {
+      "epoch": 14.459580838323353,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 9659
+    },
+    {
+      "epoch": 14.461077844311378,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1463,
+      "step": 9660
+    },
+    {
+      "epoch": 14.4625748502994,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1366,
+      "step": 9661
+    },
+    {
+      "epoch": 14.464071856287426,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1338,
+      "step": 9662
+    },
+    {
+      "epoch": 14.465568862275449,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 9663
+    },
+    {
+      "epoch": 14.467065868263473,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1323,
+      "step": 9664
+    },
+    {
+      "epoch": 14.468562874251496,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1367,
+      "step": 9665
+    },
+    {
+      "epoch": 14.470059880239521,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 9666
+    },
+    {
+      "epoch": 14.471556886227544,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 9667
+    },
+    {
+      "epoch": 14.47305389221557,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 9668
+    },
+    {
+      "epoch": 14.474550898203592,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 9669
+    },
+    {
+      "epoch": 14.476047904191617,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1376,
+      "step": 9670
+    },
+    {
+      "epoch": 14.47754491017964,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1337,
+      "step": 9671
+    },
+    {
+      "epoch": 14.479041916167665,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1434,
+      "step": 9672
+    },
+    {
+      "epoch": 14.480538922155688,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1352,
+      "step": 9673
+    },
+    {
+      "epoch": 14.482035928143713,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1369,
+      "step": 9674
+    },
+    {
+      "epoch": 14.483532934131736,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1345,
+      "step": 9675
+    },
+    {
+      "epoch": 14.48502994011976,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1413,
+      "step": 9676
+    },
+    {
+      "epoch": 14.486526946107784,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1346,
+      "step": 9677
+    },
+    {
+      "epoch": 14.488023952095809,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1328,
+      "step": 9678
+    },
+    {
+      "epoch": 14.489520958083832,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 9679
+    },
+    {
+      "epoch": 14.491017964071856,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 9680
+    },
+    {
+      "epoch": 14.49251497005988,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1356,
+      "step": 9681
+    },
+    {
+      "epoch": 14.494011976047904,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 9682
+    },
+    {
+      "epoch": 14.495508982035927,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 9683
+    },
+    {
+      "epoch": 14.497005988023952,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.134,
+      "step": 9684
+    },
+    {
+      "epoch": 14.498502994011975,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1372,
+      "step": 9685
+    },
+    {
+      "epoch": 14.5,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1395,
+      "step": 9686
+    },
+    {
+      "epoch": 14.501497005988025,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 9687
+    },
+    {
+      "epoch": 14.502994011976048,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1307,
+      "step": 9688
+    },
+    {
+      "epoch": 14.504491017964071,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1376,
+      "step": 9689
+    },
+    {
+      "epoch": 14.505988023952096,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 9690
+    },
+    {
+      "epoch": 14.50748502994012,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 9691
+    },
+    {
+      "epoch": 14.508982035928144,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1352,
+      "step": 9692
+    },
+    {
+      "epoch": 14.510479041916168,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1319,
+      "step": 9693
+    },
+    {
+      "epoch": 14.511976047904191,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1311,
+      "step": 9694
+    },
+    {
+      "epoch": 14.513473053892216,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1314,
+      "step": 9695
+    },
+    {
+      "epoch": 14.51497005988024,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1384,
+      "step": 9696
+    },
+    {
+      "epoch": 14.516467065868264,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1357,
+      "step": 9697
+    },
+    {
+      "epoch": 14.517964071856287,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1379,
+      "step": 9698
+    },
+    {
+      "epoch": 14.519461077844312,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1339,
+      "step": 9699
+    },
+    {
+      "epoch": 14.520958083832335,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 9700
+    },
+    {
+      "epoch": 14.52245508982036,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1328,
+      "step": 9701
+    },
+    {
+      "epoch": 14.523952095808383,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 9702
+    },
+    {
+      "epoch": 14.525449101796408,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1371,
+      "step": 9703
+    },
+    {
+      "epoch": 14.52694610778443,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1366,
+      "step": 9704
+    },
+    {
+      "epoch": 14.528443113772456,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.133,
+      "step": 9705
+    },
+    {
+      "epoch": 14.529940119760479,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.134,
+      "step": 9706
+    },
+    {
+      "epoch": 14.531437125748504,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1328,
+      "step": 9707
+    },
+    {
+      "epoch": 14.532934131736527,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 9708
+    },
+    {
+      "epoch": 14.534431137724551,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1428,
+      "step": 9709
+    },
+    {
+      "epoch": 14.535928143712574,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1375,
+      "step": 9710
+    },
+    {
+      "epoch": 14.5374251497006,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 9711
+    },
+    {
+      "epoch": 14.538922155688622,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.14,
+      "step": 9712
+    },
+    {
+      "epoch": 14.540419161676647,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 9713
+    },
+    {
+      "epoch": 14.54191616766467,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 9714
+    },
+    {
+      "epoch": 14.543413173652695,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.132,
+      "step": 9715
+    },
+    {
+      "epoch": 14.544910179640718,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1378,
+      "step": 9716
+    },
+    {
+      "epoch": 14.546407185628743,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1349,
+      "step": 9717
+    },
+    {
+      "epoch": 14.547904191616766,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1378,
+      "step": 9718
+    },
+    {
+      "epoch": 14.54940119760479,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 9719
+    },
+    {
+      "epoch": 14.550898203592814,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 9720
+    },
+    {
+      "epoch": 14.552395209580839,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1376,
+      "step": 9721
+    },
+    {
+      "epoch": 14.553892215568862,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 9722
+    },
+    {
+      "epoch": 14.555389221556887,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1371,
+      "step": 9723
+    },
+    {
+      "epoch": 14.55688622754491,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1344,
+      "step": 9724
+    },
+    {
+      "epoch": 14.558383233532934,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1443,
+      "step": 9725
+    },
+    {
+      "epoch": 14.559880239520957,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1381,
+      "step": 9726
+    },
+    {
+      "epoch": 14.561377245508982,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 9727
+    },
+    {
+      "epoch": 14.562874251497005,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1369,
+      "step": 9728
+    },
+    {
+      "epoch": 14.56437125748503,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1342,
+      "step": 9729
+    },
+    {
+      "epoch": 14.565868263473053,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1305,
+      "step": 9730
+    },
+    {
+      "epoch": 14.567365269461078,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.14,
+      "step": 9731
+    },
+    {
+      "epoch": 14.568862275449101,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1326,
+      "step": 9732
+    },
+    {
+      "epoch": 14.570359281437126,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1363,
+      "step": 9733
+    },
+    {
+      "epoch": 14.571856287425149,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 9734
+    },
+    {
+      "epoch": 14.573353293413174,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1333,
+      "step": 9735
+    },
+    {
+      "epoch": 14.574850299401197,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 9736
+    },
+    {
+      "epoch": 14.576347305389222,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1364,
+      "step": 9737
+    },
+    {
+      "epoch": 14.577844311377245,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.136,
+      "step": 9738
+    },
+    {
+      "epoch": 14.57934131736527,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 9739
+    },
+    {
+      "epoch": 14.580838323353294,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 9740
+    },
+    {
+      "epoch": 14.582335329341317,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 9741
+    },
+    {
+      "epoch": 14.58383233532934,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 9742
+    },
+    {
+      "epoch": 14.585329341317365,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.144,
+      "step": 9743
+    },
+    {
+      "epoch": 14.58682634730539,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1329,
+      "step": 9744
+    },
+    {
+      "epoch": 14.588323353293413,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1327,
+      "step": 9745
+    },
+    {
+      "epoch": 14.589820359281438,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1416,
+      "step": 9746
+    },
+    {
+      "epoch": 14.591317365269461,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 9747
+    },
+    {
+      "epoch": 14.592814371257486,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1352,
+      "step": 9748
+    },
+    {
+      "epoch": 14.594311377245509,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1378,
+      "step": 9749
+    },
+    {
+      "epoch": 14.595808383233534,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 9750
+    },
+    {
+      "epoch": 14.597305389221557,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.135,
+      "step": 9751
+    },
+    {
+      "epoch": 14.598802395209582,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1386,
+      "step": 9752
+    },
+    {
+      "epoch": 14.600299401197605,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1342,
+      "step": 9753
+    },
+    {
+      "epoch": 14.60179640718563,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1416,
+      "step": 9754
+    },
+    {
+      "epoch": 14.603293413173652,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 9755
+    },
+    {
+      "epoch": 14.604790419161677,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1328,
+      "step": 9756
+    },
+    {
+      "epoch": 14.6062874251497,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1301,
+      "step": 9757
+    },
+    {
+      "epoch": 14.607784431137725,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 9758
+    },
+    {
+      "epoch": 14.609281437125748,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1352,
+      "step": 9759
+    },
+    {
+      "epoch": 14.610778443113773,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1331,
+      "step": 9760
+    },
+    {
+      "epoch": 14.612275449101796,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1367,
+      "step": 9761
+    },
+    {
+      "epoch": 14.613772455089821,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1329,
+      "step": 9762
+    },
+    {
+      "epoch": 14.615269461077844,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1332,
+      "step": 9763
+    },
+    {
+      "epoch": 14.616766467065869,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.133,
+      "step": 9764
+    },
+    {
+      "epoch": 14.618263473053892,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1363,
+      "step": 9765
+    },
+    {
+      "epoch": 14.619760479041917,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 9766
+    },
+    {
+      "epoch": 14.62125748502994,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1327,
+      "step": 9767
+    },
+    {
+      "epoch": 14.622754491017965,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1349,
+      "step": 9768
+    },
+    {
+      "epoch": 14.624251497005988,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1345,
+      "step": 9769
+    },
+    {
+      "epoch": 14.625748502994012,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 9770
+    },
+    {
+      "epoch": 14.627245508982035,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1351,
+      "step": 9771
+    },
+    {
+      "epoch": 14.62874251497006,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 9772
+    },
+    {
+      "epoch": 14.630239520958083,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 9773
+    },
+    {
+      "epoch": 14.631736526946108,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1457,
+      "step": 9774
+    },
+    {
+      "epoch": 14.633233532934131,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1362,
+      "step": 9775
+    },
+    {
+      "epoch": 14.634730538922156,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 9776
+    },
+    {
+      "epoch": 14.636227544910179,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1376,
+      "step": 9777
+    },
+    {
+      "epoch": 14.637724550898204,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1354,
+      "step": 9778
+    },
+    {
+      "epoch": 14.639221556886227,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 9779
+    },
+    {
+      "epoch": 14.640718562874252,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1353,
+      "step": 9780
+    },
+    {
+      "epoch": 14.642215568862275,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1327,
+      "step": 9781
+    },
+    {
+      "epoch": 14.6437125748503,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1378,
+      "step": 9782
+    },
+    {
+      "epoch": 14.645209580838323,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1333,
+      "step": 9783
+    },
+    {
+      "epoch": 14.646706586826348,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 9784
+    },
+    {
+      "epoch": 14.64820359281437,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1393,
+      "step": 9785
+    },
+    {
+      "epoch": 14.649700598802395,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1387,
+      "step": 9786
+    },
+    {
+      "epoch": 14.651197604790418,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 9787
+    },
+    {
+      "epoch": 14.652694610778443,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 9788
+    },
+    {
+      "epoch": 14.654191616766466,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 9789
+    },
+    {
+      "epoch": 14.655688622754491,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 9790
+    },
+    {
+      "epoch": 14.657185628742514,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1357,
+      "step": 9791
+    },
+    {
+      "epoch": 14.658682634730539,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1379,
+      "step": 9792
+    },
+    {
+      "epoch": 14.660179640718562,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1362,
+      "step": 9793
+    },
+    {
+      "epoch": 14.661676646706587,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1324,
+      "step": 9794
+    },
+    {
+      "epoch": 14.66317365269461,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1335,
+      "step": 9795
+    },
+    {
+      "epoch": 14.664670658682635,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 9796
+    },
+    {
+      "epoch": 14.66616766467066,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1344,
+      "step": 9797
+    },
+    {
+      "epoch": 14.667664670658683,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1329,
+      "step": 9798
+    },
+    {
+      "epoch": 14.669161676646706,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 9799
+    },
+    {
+      "epoch": 14.67065868263473,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 9800
+    },
+    {
+      "epoch": 14.672155688622755,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 9801
+    },
+    {
+      "epoch": 14.673652694610778,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.132,
+      "step": 9802
+    },
+    {
+      "epoch": 14.675149700598803,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.134,
+      "step": 9803
+    },
+    {
+      "epoch": 14.676646706586826,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 9804
+    },
+    {
+      "epoch": 14.678143712574851,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.134,
+      "step": 9805
+    },
+    {
+      "epoch": 14.679640718562874,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1377,
+      "step": 9806
+    },
+    {
+      "epoch": 14.681137724550899,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1472,
+      "step": 9807
+    },
+    {
+      "epoch": 14.682634730538922,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 9808
+    },
+    {
+      "epoch": 14.684131736526947,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1365,
+      "step": 9809
+    },
+    {
+      "epoch": 14.68562874251497,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1357,
+      "step": 9810
+    },
+    {
+      "epoch": 14.687125748502995,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1298,
+      "step": 9811
+    },
+    {
+      "epoch": 14.688622754491018,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1348,
+      "step": 9812
+    },
+    {
+      "epoch": 14.690119760479043,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1369,
+      "step": 9813
+    },
+    {
+      "epoch": 14.691616766467066,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1315,
+      "step": 9814
+    },
+    {
+      "epoch": 14.69311377245509,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 9815
+    },
+    {
+      "epoch": 14.694610778443113,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1383,
+      "step": 9816
+    },
+    {
+      "epoch": 14.696107784431138,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 9817
+    },
+    {
+      "epoch": 14.697604790419161,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1381,
+      "step": 9818
+    },
+    {
+      "epoch": 14.699101796407186,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1377,
+      "step": 9819
+    },
+    {
+      "epoch": 14.70059880239521,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1384,
+      "step": 9820
+    },
+    {
+      "epoch": 14.702095808383234,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1301,
+      "step": 9821
+    },
+    {
+      "epoch": 14.703592814371257,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1443,
+      "step": 9822
+    },
+    {
+      "epoch": 14.705089820359282,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 9823
+    },
+    {
+      "epoch": 14.706586826347305,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1364,
+      "step": 9824
+    },
+    {
+      "epoch": 14.70808383233533,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1434,
+      "step": 9825
+    },
+    {
+      "epoch": 14.709580838323353,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 9826
+    },
+    {
+      "epoch": 14.711077844311378,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 9827
+    },
+    {
+      "epoch": 14.7125748502994,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1339,
+      "step": 9828
+    },
+    {
+      "epoch": 14.714071856287426,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 9829
+    },
+    {
+      "epoch": 14.715568862275449,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1369,
+      "step": 9830
+    },
+    {
+      "epoch": 14.717065868263473,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 9831
+    },
+    {
+      "epoch": 14.718562874251496,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.146,
+      "step": 9832
+    },
+    {
+      "epoch": 14.720059880239521,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 9833
+    },
+    {
+      "epoch": 14.721556886227544,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 9834
+    },
+    {
+      "epoch": 14.72305389221557,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 9835
+    },
+    {
+      "epoch": 14.724550898203592,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 9836
+    },
+    {
+      "epoch": 14.726047904191617,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1368,
+      "step": 9837
+    },
+    {
+      "epoch": 14.72754491017964,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1375,
+      "step": 9838
+    },
+    {
+      "epoch": 14.729041916167665,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1313,
+      "step": 9839
+    },
+    {
+      "epoch": 14.730538922155688,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 9840
+    },
+    {
+      "epoch": 14.732035928143713,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 9841
+    },
+    {
+      "epoch": 14.733532934131736,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.131,
+      "step": 9842
+    },
+    {
+      "epoch": 14.73502994011976,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1366,
+      "step": 9843
+    },
+    {
+      "epoch": 14.736526946107784,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 9844
+    },
+    {
+      "epoch": 14.738023952095809,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 9845
+    },
+    {
+      "epoch": 14.739520958083832,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.135,
+      "step": 9846
+    },
+    {
+      "epoch": 14.741017964071856,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 9847
+    },
+    {
+      "epoch": 14.74251497005988,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1279,
+      "step": 9848
+    },
+    {
+      "epoch": 14.744011976047904,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1379,
+      "step": 9849
+    },
+    {
+      "epoch": 14.745508982035929,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1362,
+      "step": 9850
+    },
+    {
+      "epoch": 14.747005988023952,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1342,
+      "step": 9851
+    },
+    {
+      "epoch": 14.748502994011975,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 9852
+    },
+    {
+      "epoch": 14.75,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1375,
+      "step": 9853
+    },
+    {
+      "epoch": 14.751497005988025,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1305,
+      "step": 9854
+    },
+    {
+      "epoch": 14.752994011976048,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 9855
+    },
+    {
+      "epoch": 14.754491017964071,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 9856
+    },
+    {
+      "epoch": 14.755988023952096,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 9857
+    },
+    {
+      "epoch": 14.75748502994012,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1392,
+      "step": 9858
+    },
+    {
+      "epoch": 14.758982035928144,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1314,
+      "step": 9859
+    },
+    {
+      "epoch": 14.760479041916168,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1369,
+      "step": 9860
+    },
+    {
+      "epoch": 14.761976047904191,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1375,
+      "step": 9861
+    },
+    {
+      "epoch": 14.763473053892216,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 9862
+    },
+    {
+      "epoch": 14.76497005988024,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 9863
+    },
+    {
+      "epoch": 14.766467065868264,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 9864
+    },
+    {
+      "epoch": 14.767964071856287,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 9865
+    },
+    {
+      "epoch": 14.769461077844312,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 9866
+    },
+    {
+      "epoch": 14.770958083832335,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 9867
+    },
+    {
+      "epoch": 14.77245508982036,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 9868
+    },
+    {
+      "epoch": 14.773952095808383,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1321,
+      "step": 9869
+    },
+    {
+      "epoch": 14.775449101796408,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1379,
+      "step": 9870
+    },
+    {
+      "epoch": 14.77694610778443,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 9871
+    },
+    {
+      "epoch": 14.778443113772456,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1429,
+      "step": 9872
+    },
+    {
+      "epoch": 14.779940119760479,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1346,
+      "step": 9873
+    },
+    {
+      "epoch": 14.781437125748504,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1366,
+      "step": 9874
+    },
+    {
+      "epoch": 14.782934131736527,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 9875
+    },
+    {
+      "epoch": 14.784431137724551,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1383,
+      "step": 9876
+    },
+    {
+      "epoch": 14.785928143712574,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 9877
+    },
+    {
+      "epoch": 14.7874251497006,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1459,
+      "step": 9878
+    },
+    {
+      "epoch": 14.788922155688622,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1348,
+      "step": 9879
+    },
+    {
+      "epoch": 14.790419161676647,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1329,
+      "step": 9880
+    },
+    {
+      "epoch": 14.79191616766467,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 9881
+    },
+    {
+      "epoch": 14.793413173652695,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 9882
+    },
+    {
+      "epoch": 14.794910179640718,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1356,
+      "step": 9883
+    },
+    {
+      "epoch": 14.796407185628743,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1429,
+      "step": 9884
+    },
+    {
+      "epoch": 14.797904191616766,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1395,
+      "step": 9885
+    },
+    {
+      "epoch": 14.79940119760479,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 9886
+    },
+    {
+      "epoch": 14.800898203592814,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1333,
+      "step": 9887
+    },
+    {
+      "epoch": 14.802395209580839,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 9888
+    },
+    {
+      "epoch": 14.803892215568862,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1326,
+      "step": 9889
+    },
+    {
+      "epoch": 14.805389221556887,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1314,
+      "step": 9890
+    },
+    {
+      "epoch": 14.80688622754491,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1359,
+      "step": 9891
+    },
+    {
+      "epoch": 14.808383233532934,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1428,
+      "step": 9892
+    },
+    {
+      "epoch": 14.809880239520957,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.132,
+      "step": 9893
+    },
+    {
+      "epoch": 14.811377245508982,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 9894
+    },
+    {
+      "epoch": 14.812874251497005,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1373,
+      "step": 9895
+    },
+    {
+      "epoch": 14.81437125748503,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1366,
+      "step": 9896
+    },
+    {
+      "epoch": 14.815868263473053,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 9897
+    },
+    {
+      "epoch": 14.817365269461078,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1381,
+      "step": 9898
+    },
+    {
+      "epoch": 14.818862275449101,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1361,
+      "step": 9899
+    },
+    {
+      "epoch": 14.820359281437126,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1361,
+      "step": 9900
+    },
+    {
+      "epoch": 14.821856287425149,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 9901
+    },
+    {
+      "epoch": 14.823353293413174,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 9902
+    },
+    {
+      "epoch": 14.824850299401197,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 9903
+    },
+    {
+      "epoch": 14.826347305389222,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 9904
+    },
+    {
+      "epoch": 14.827844311377245,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 9905
+    },
+    {
+      "epoch": 14.82934131736527,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1392,
+      "step": 9906
+    },
+    {
+      "epoch": 14.830838323353294,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 9907
+    },
+    {
+      "epoch": 14.832335329341317,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1321,
+      "step": 9908
+    },
+    {
+      "epoch": 14.83383233532934,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 9909
+    },
+    {
+      "epoch": 14.835329341317365,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 9910
+    },
+    {
+      "epoch": 14.83682634730539,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1463,
+      "step": 9911
+    },
+    {
+      "epoch": 14.838323353293413,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1368,
+      "step": 9912
+    },
+    {
+      "epoch": 14.839820359281438,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1373,
+      "step": 9913
+    },
+    {
+      "epoch": 14.841317365269461,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 9914
+    },
+    {
+      "epoch": 14.842814371257486,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1325,
+      "step": 9915
+    },
+    {
+      "epoch": 14.844311377245509,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0008,
+      "loss": 1.146,
+      "step": 9916
+    },
+    {
+      "epoch": 14.845808383233534,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1378,
+      "step": 9917
+    },
+    {
+      "epoch": 14.847305389221557,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1317,
+      "step": 9918
+    },
+    {
+      "epoch": 14.848802395209582,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 9919
+    },
+    {
+      "epoch": 14.850299401197605,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1357,
+      "step": 9920
+    },
+    {
+      "epoch": 14.85179640718563,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1375,
+      "step": 9921
+    },
+    {
+      "epoch": 14.853293413173652,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1386,
+      "step": 9922
+    },
+    {
+      "epoch": 14.854790419161677,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 9923
+    },
+    {
+      "epoch": 14.8562874251497,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1376,
+      "step": 9924
+    },
+    {
+      "epoch": 14.857784431137725,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1383,
+      "step": 9925
+    },
+    {
+      "epoch": 14.859281437125748,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 9926
+    },
+    {
+      "epoch": 14.860778443113773,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.144,
+      "step": 9927
+    },
+    {
+      "epoch": 14.862275449101796,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1363,
+      "step": 9928
+    },
+    {
+      "epoch": 14.863772455089821,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 9929
+    },
+    {
+      "epoch": 14.865269461077844,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 9930
+    },
+    {
+      "epoch": 14.866766467065869,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1278,
+      "step": 9931
+    },
+    {
+      "epoch": 14.868263473053892,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1459,
+      "step": 9932
+    },
+    {
+      "epoch": 14.869760479041917,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1328,
+      "step": 9933
+    },
+    {
+      "epoch": 14.87125748502994,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1413,
+      "step": 9934
+    },
+    {
+      "epoch": 14.872754491017965,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1349,
+      "step": 9935
+    },
+    {
+      "epoch": 14.874251497005988,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1413,
+      "step": 9936
+    },
+    {
+      "epoch": 14.875748502994012,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1395,
+      "step": 9937
+    },
+    {
+      "epoch": 14.877245508982035,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 9938
+    },
+    {
+      "epoch": 14.87874251497006,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 9939
+    },
+    {
+      "epoch": 14.880239520958083,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 9940
+    },
+    {
+      "epoch": 14.881736526946108,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 9941
+    },
+    {
+      "epoch": 14.883233532934131,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1375,
+      "step": 9942
+    },
+    {
+      "epoch": 14.884730538922156,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1416,
+      "step": 9943
+    },
+    {
+      "epoch": 14.886227544910179,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 9944
+    },
+    {
+      "epoch": 14.887724550898204,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1334,
+      "step": 9945
+    },
+    {
+      "epoch": 14.889221556886227,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1376,
+      "step": 9946
+    },
+    {
+      "epoch": 14.890718562874252,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.136,
+      "step": 9947
+    },
+    {
+      "epoch": 14.892215568862275,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1352,
+      "step": 9948
+    },
+    {
+      "epoch": 14.8937125748503,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1379,
+      "step": 9949
+    },
+    {
+      "epoch": 14.895209580838323,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 9950
+    },
+    {
+      "epoch": 14.896706586826348,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 9951
+    },
+    {
+      "epoch": 14.89820359281437,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 9952
+    },
+    {
+      "epoch": 14.899700598802395,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 9953
+    },
+    {
+      "epoch": 14.901197604790418,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1354,
+      "step": 9954
+    },
+    {
+      "epoch": 14.902694610778443,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1372,
+      "step": 9955
+    },
+    {
+      "epoch": 14.904191616766466,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 9956
+    },
+    {
+      "epoch": 14.905688622754491,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 9957
+    },
+    {
+      "epoch": 14.907185628742514,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 9958
+    },
+    {
+      "epoch": 14.908682634730539,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 9959
+    },
+    {
+      "epoch": 14.910179640718562,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1325,
+      "step": 9960
+    },
+    {
+      "epoch": 14.911676646706587,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1393,
+      "step": 9961
+    },
+    {
+      "epoch": 14.91317365269461,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1316,
+      "step": 9962
+    },
+    {
+      "epoch": 14.914670658682635,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 9963
+    },
+    {
+      "epoch": 14.91616766467066,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1381,
+      "step": 9964
+    },
+    {
+      "epoch": 14.917664670658683,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1369,
+      "step": 9965
+    },
+    {
+      "epoch": 14.919161676646706,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1353,
+      "step": 9966
+    },
+    {
+      "epoch": 14.92065868263473,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 9967
+    },
+    {
+      "epoch": 14.922155688622755,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1313,
+      "step": 9968
+    },
+    {
+      "epoch": 14.923652694610778,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 9969
+    },
+    {
+      "epoch": 14.925149700598803,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1323,
+      "step": 9970
+    },
+    {
+      "epoch": 14.926646706586826,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.13,
+      "step": 9971
+    },
+    {
+      "epoch": 14.928143712574851,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 9972
+    },
+    {
+      "epoch": 14.929640718562874,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 9973
+    },
+    {
+      "epoch": 14.931137724550899,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1328,
+      "step": 9974
+    },
+    {
+      "epoch": 14.932634730538922,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 9975
+    },
+    {
+      "epoch": 14.934131736526947,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1343,
+      "step": 9976
+    },
+    {
+      "epoch": 14.93562874251497,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.137,
+      "step": 9977
+    },
+    {
+      "epoch": 14.937125748502995,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1317,
+      "step": 9978
+    },
+    {
+      "epoch": 14.938622754491018,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1367,
+      "step": 9979
+    },
+    {
+      "epoch": 14.940119760479043,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 9980
+    },
+    {
+      "epoch": 14.941616766467066,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.134,
+      "step": 9981
+    },
+    {
+      "epoch": 14.94311377245509,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1383,
+      "step": 9982
+    },
+    {
+      "epoch": 14.944610778443113,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1364,
+      "step": 9983
+    },
+    {
+      "epoch": 14.946107784431138,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1376,
+      "step": 9984
+    },
+    {
+      "epoch": 14.947604790419161,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 9985
+    },
+    {
+      "epoch": 14.949101796407186,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1327,
+      "step": 9986
+    },
+    {
+      "epoch": 14.95059880239521,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1366,
+      "step": 9987
+    },
+    {
+      "epoch": 14.952095808383234,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 9988
+    },
+    {
+      "epoch": 14.953592814371257,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1348,
+      "step": 9989
+    },
+    {
+      "epoch": 14.955089820359282,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 9990
+    },
+    {
+      "epoch": 14.956586826347305,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1381,
+      "step": 9991
+    },
+    {
+      "epoch": 14.95808383233533,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 9992
+    },
+    {
+      "epoch": 14.959580838323353,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1352,
+      "step": 9993
+    },
+    {
+      "epoch": 14.961077844311378,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 9994
+    },
+    {
+      "epoch": 14.9625748502994,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1307,
+      "step": 9995
+    },
+    {
+      "epoch": 14.964071856287426,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.134,
+      "step": 9996
+    },
+    {
+      "epoch": 14.965568862275449,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 9997
+    },
+    {
+      "epoch": 14.967065868263473,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1395,
+      "step": 9998
+    },
+    {
+      "epoch": 14.968562874251496,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1343,
+      "step": 9999
+    },
+    {
+      "epoch": 14.970059880239521,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1363,
+      "step": 10000
+    },
+    {
+      "epoch": 14.971556886227544,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 10001
+    },
+    {
+      "epoch": 14.97305389221557,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1953,
+      "step": 10002
+    },
+    {
+      "epoch": 14.974550898203592,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0008,
+      "loss": 1.197,
+      "step": 10003
+    },
+    {
+      "epoch": 14.976047904191617,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.2462,
+      "step": 10004
+    },
+    {
+      "epoch": 14.97754491017964,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.0008,
+      "loss": 1.322,
+      "step": 10005
+    },
+    {
+      "epoch": 14.979041916167665,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.0008,
+      "loss": 1.3121,
+      "step": 10006
+    },
+    {
+      "epoch": 14.980538922155688,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0008,
+      "loss": 1.4386,
+      "step": 10007
+    },
+    {
+      "epoch": 14.982035928143713,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.0008,
+      "loss": 1.415,
+      "step": 10008
+    },
+    {
+      "epoch": 14.983532934131736,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.0008,
+      "loss": 1.4384,
+      "step": 10009
+    },
+    {
+      "epoch": 14.98502994011976,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.4024,
+      "step": 10010
+    },
+    {
+      "epoch": 14.986526946107784,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.0008,
+      "loss": 1.4154,
+      "step": 10011
+    },
+    {
+      "epoch": 14.988023952095809,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0008,
+      "loss": 1.3349,
+      "step": 10012
+    },
+    {
+      "epoch": 14.989520958083832,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0008,
+      "loss": 1.3392,
+      "step": 10013
+    },
+    {
+      "epoch": 14.991017964071856,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0008,
+      "loss": 1.3292,
+      "step": 10014
+    },
+    {
+      "epoch": 14.99251497005988,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0008,
+      "loss": 1.33,
+      "step": 10015
+    },
+    {
+      "epoch": 14.994011976047904,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0008,
+      "loss": 1.3175,
+      "step": 10016
+    },
+    {
+      "epoch": 14.995508982035929,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0008,
+      "loss": 1.3063,
+      "step": 10017
+    },
+    {
+      "epoch": 14.997005988023952,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0008,
+      "loss": 1.2948,
+      "step": 10018
+    },
+    {
+      "epoch": 14.998502994011975,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0008,
+      "loss": 1.2935,
+      "step": 10019
+    },
+    {
+      "epoch": 15.0,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0008,
+      "loss": 1.3024,
+      "step": 10020
+    },
+    {
+      "epoch": 15.001497005988025,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0008,
+      "loss": 1.2768,
+      "step": 10021
+    },
+    {
+      "epoch": 15.002994011976048,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0008,
+      "loss": 1.2798,
+      "step": 10022
+    },
+    {
+      "epoch": 15.004491017964073,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0008,
+      "loss": 1.2622,
+      "step": 10023
+    },
+    {
+      "epoch": 15.005988023952096,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0008,
+      "loss": 1.2656,
+      "step": 10024
+    },
+    {
+      "epoch": 15.00748502994012,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0008,
+      "loss": 1.2513,
+      "step": 10025
+    },
+    {
+      "epoch": 15.008982035928144,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0008,
+      "loss": 1.2537,
+      "step": 10026
+    },
+    {
+      "epoch": 15.010479041916168,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0008,
+      "loss": 1.2473,
+      "step": 10027
+    },
+    {
+      "epoch": 15.011976047904191,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.2435,
+      "step": 10028
+    },
+    {
+      "epoch": 15.013473053892216,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0008,
+      "loss": 1.2405,
+      "step": 10029
+    },
+    {
+      "epoch": 15.01497005988024,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 0.0008,
+      "loss": 1.2356,
+      "step": 10030
+    },
+    {
+      "epoch": 15.016467065868264,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0008,
+      "loss": 1.2365,
+      "step": 10031
+    },
+    {
+      "epoch": 15.017964071856287,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0008,
+      "loss": 1.2272,
+      "step": 10032
+    },
+    {
+      "epoch": 15.019461077844312,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0008,
+      "loss": 1.2254,
+      "step": 10033
+    },
+    {
+      "epoch": 15.020958083832335,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.2223,
+      "step": 10034
+    },
+    {
+      "epoch": 15.02245508982036,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 0.0008,
+      "loss": 1.2255,
+      "step": 10035
+    },
+    {
+      "epoch": 15.023952095808383,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0008,
+      "loss": 1.2276,
+      "step": 10036
+    },
+    {
+      "epoch": 15.025449101796408,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0008,
+      "loss": 1.2096,
+      "step": 10037
+    },
+    {
+      "epoch": 15.02694610778443,
+      "grad_norm": 0.220703125,
+      "learning_rate": 0.0008,
+      "loss": 1.2236,
+      "step": 10038
+    },
+    {
+      "epoch": 15.028443113772456,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0008,
+      "loss": 1.2168,
+      "step": 10039
+    },
+    {
+      "epoch": 15.029940119760479,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.2155,
+      "step": 10040
+    },
+    {
+      "epoch": 15.031437125748504,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0008,
+      "loss": 1.2094,
+      "step": 10041
+    },
+    {
+      "epoch": 15.032934131736527,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0008,
+      "loss": 1.2182,
+      "step": 10042
+    },
+    {
+      "epoch": 15.034431137724551,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0008,
+      "loss": 1.2101,
+      "step": 10043
+    },
+    {
+      "epoch": 15.035928143712574,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.2025,
+      "step": 10044
+    },
+    {
+      "epoch": 15.0374251497006,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.2118,
+      "step": 10045
+    },
+    {
+      "epoch": 15.038922155688622,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.2121,
+      "step": 10046
+    },
+    {
+      "epoch": 15.040419161676647,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0008,
+      "loss": 1.2078,
+      "step": 10047
+    },
+    {
+      "epoch": 15.04191616766467,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1975,
+      "step": 10048
+    },
+    {
+      "epoch": 15.043413173652695,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0008,
+      "loss": 1.2134,
+      "step": 10049
+    },
+    {
+      "epoch": 15.044910179640718,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.2077,
+      "step": 10050
+    },
+    {
+      "epoch": 15.046407185628743,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0008,
+      "loss": 1.2121,
+      "step": 10051
+    },
+    {
+      "epoch": 15.047904191616766,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0008,
+      "loss": 1.2069,
+      "step": 10052
+    },
+    {
+      "epoch": 15.04940119760479,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0008,
+      "loss": 1.2011,
+      "step": 10053
+    },
+    {
+      "epoch": 15.050898203592814,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0008,
+      "loss": 1.2021,
+      "step": 10054
+    },
+    {
+      "epoch": 15.052395209580839,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0008,
+      "loss": 1.2066,
+      "step": 10055
+    },
+    {
+      "epoch": 15.053892215568862,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.2001,
+      "step": 10056
+    },
+    {
+      "epoch": 15.055389221556887,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0008,
+      "loss": 1.2066,
+      "step": 10057
+    },
+    {
+      "epoch": 15.05688622754491,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1997,
+      "step": 10058
+    },
+    {
+      "epoch": 15.058383233532934,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0008,
+      "loss": 1.2026,
+      "step": 10059
+    },
+    {
+      "epoch": 15.059880239520957,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1989,
+      "step": 10060
+    },
+    {
+      "epoch": 15.061377245508982,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1982,
+      "step": 10061
+    },
+    {
+      "epoch": 15.062874251497005,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.195,
+      "step": 10062
+    },
+    {
+      "epoch": 15.06437125748503,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1939,
+      "step": 10063
+    },
+    {
+      "epoch": 15.065868263473053,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1967,
+      "step": 10064
+    },
+    {
+      "epoch": 15.067365269461078,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1991,
+      "step": 10065
+    },
+    {
+      "epoch": 15.068862275449101,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1929,
+      "step": 10066
+    },
+    {
+      "epoch": 15.070359281437126,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1991,
+      "step": 10067
+    },
+    {
+      "epoch": 15.071856287425149,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1991,
+      "step": 10068
+    },
+    {
+      "epoch": 15.073353293413174,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0008,
+      "loss": 1.2034,
+      "step": 10069
+    },
+    {
+      "epoch": 15.074850299401197,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1983,
+      "step": 10070
+    },
+    {
+      "epoch": 15.076347305389222,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.2024,
+      "step": 10071
+    },
+    {
+      "epoch": 15.077844311377245,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1969,
+      "step": 10072
+    },
+    {
+      "epoch": 15.07934131736527,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1962,
+      "step": 10073
+    },
+    {
+      "epoch": 15.080838323353293,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0008,
+      "loss": 1.2029,
+      "step": 10074
+    },
+    {
+      "epoch": 15.082335329341317,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0008,
+      "loss": 1.2002,
+      "step": 10075
+    },
+    {
+      "epoch": 15.08383233532934,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0008,
+      "loss": 1.2045,
+      "step": 10076
+    },
+    {
+      "epoch": 15.085329341317365,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1967,
+      "step": 10077
+    },
+    {
+      "epoch": 15.08682634730539,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1967,
+      "step": 10078
+    },
+    {
+      "epoch": 15.088323353293413,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1964,
+      "step": 10079
+    },
+    {
+      "epoch": 15.089820359281438,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1886,
+      "step": 10080
+    },
+    {
+      "epoch": 15.091317365269461,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.2011,
+      "step": 10081
+    },
+    {
+      "epoch": 15.092814371257486,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0008,
+      "loss": 1.1915,
+      "step": 10082
+    },
+    {
+      "epoch": 15.094311377245509,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0008,
+      "loss": 1.1937,
+      "step": 10083
+    },
+    {
+      "epoch": 15.095808383233534,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1951,
+      "step": 10084
+    },
+    {
+      "epoch": 15.097305389221557,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1879,
+      "step": 10085
+    },
+    {
+      "epoch": 15.098802395209582,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1895,
+      "step": 10086
+    },
+    {
+      "epoch": 15.100299401197605,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1964,
+      "step": 10087
+    },
+    {
+      "epoch": 15.10179640718563,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0008,
+      "loss": 1.1939,
+      "step": 10088
+    },
+    {
+      "epoch": 15.103293413173652,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0008,
+      "loss": 1.2019,
+      "step": 10089
+    },
+    {
+      "epoch": 15.104790419161677,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1935,
+      "step": 10090
+    },
+    {
+      "epoch": 15.1062874251497,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1927,
+      "step": 10091
+    },
+    {
+      "epoch": 15.107784431137725,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1872,
+      "step": 10092
+    },
+    {
+      "epoch": 15.109281437125748,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1908,
+      "step": 10093
+    },
+    {
+      "epoch": 15.110778443113773,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0008,
+      "loss": 1.185,
+      "step": 10094
+    },
+    {
+      "epoch": 15.112275449101796,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1925,
+      "step": 10095
+    },
+    {
+      "epoch": 15.113772455089821,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0008,
+      "loss": 1.187,
+      "step": 10096
+    },
+    {
+      "epoch": 15.115269461077844,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1898,
+      "step": 10097
+    },
+    {
+      "epoch": 15.116766467065869,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1908,
+      "step": 10098
+    },
+    {
+      "epoch": 15.118263473053892,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1927,
+      "step": 10099
+    },
+    {
+      "epoch": 15.119760479041917,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1933,
+      "step": 10100
+    },
+    {
+      "epoch": 15.12125748502994,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1867,
+      "step": 10101
+    },
+    {
+      "epoch": 15.122754491017965,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1823,
+      "step": 10102
+    },
+    {
+      "epoch": 15.124251497005988,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1814,
+      "step": 10103
+    },
+    {
+      "epoch": 15.125748502994012,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1911,
+      "step": 10104
+    },
+    {
+      "epoch": 15.127245508982035,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1813,
+      "step": 10105
+    },
+    {
+      "epoch": 15.12874251497006,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1893,
+      "step": 10106
+    },
+    {
+      "epoch": 15.130239520958083,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1838,
+      "step": 10107
+    },
+    {
+      "epoch": 15.131736526946108,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1872,
+      "step": 10108
+    },
+    {
+      "epoch": 15.133233532934131,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1911,
+      "step": 10109
+    },
+    {
+      "epoch": 15.134730538922156,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1896,
+      "step": 10110
+    },
+    {
+      "epoch": 15.136227544910179,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1874,
+      "step": 10111
+    },
+    {
+      "epoch": 15.137724550898204,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1867,
+      "step": 10112
+    },
+    {
+      "epoch": 15.139221556886227,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1852,
+      "step": 10113
+    },
+    {
+      "epoch": 15.140718562874252,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1812,
+      "step": 10114
+    },
+    {
+      "epoch": 15.142215568862275,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1879,
+      "step": 10115
+    },
+    {
+      "epoch": 15.1437125748503,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1872,
+      "step": 10116
+    },
+    {
+      "epoch": 15.145209580838323,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1864,
+      "step": 10117
+    },
+    {
+      "epoch": 15.146706586826348,
+      "grad_norm": 0.23828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1896,
+      "step": 10118
+    },
+    {
+      "epoch": 15.14820359281437,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1956,
+      "step": 10119
+    },
+    {
+      "epoch": 15.149700598802395,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1972,
+      "step": 10120
+    },
+    {
+      "epoch": 15.151197604790418,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1886,
+      "step": 10121
+    },
+    {
+      "epoch": 15.152694610778443,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1964,
+      "step": 10122
+    },
+    {
+      "epoch": 15.154191616766466,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1911,
+      "step": 10123
+    },
+    {
+      "epoch": 15.155688622754491,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1974,
+      "step": 10124
+    },
+    {
+      "epoch": 15.157185628742514,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1901,
+      "step": 10125
+    },
+    {
+      "epoch": 15.158682634730539,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1833,
+      "step": 10126
+    },
+    {
+      "epoch": 15.160179640718562,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0008,
+      "loss": 1.184,
+      "step": 10127
+    },
+    {
+      "epoch": 15.161676646706587,
+      "grad_norm": 0.23046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1825,
+      "step": 10128
+    },
+    {
+      "epoch": 15.16317365269461,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1908,
+      "step": 10129
+    },
+    {
+      "epoch": 15.164670658682635,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0008,
+      "loss": 1.186,
+      "step": 10130
+    },
+    {
+      "epoch": 15.16616766467066,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0008,
+      "loss": 1.1848,
+      "step": 10131
+    },
+    {
+      "epoch": 15.167664670658683,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1879,
+      "step": 10132
+    },
+    {
+      "epoch": 15.169161676646707,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1837,
+      "step": 10133
+    },
+    {
+      "epoch": 15.17065868263473,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.192,
+      "step": 10134
+    },
+    {
+      "epoch": 15.172155688622755,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1921,
+      "step": 10135
+    },
+    {
+      "epoch": 15.173652694610778,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1854,
+      "step": 10136
+    },
+    {
+      "epoch": 15.175149700598803,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1867,
+      "step": 10137
+    },
+    {
+      "epoch": 15.176646706586826,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0008,
+      "loss": 1.187,
+      "step": 10138
+    },
+    {
+      "epoch": 15.178143712574851,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1886,
+      "step": 10139
+    },
+    {
+      "epoch": 15.179640718562874,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1865,
+      "step": 10140
+    },
+    {
+      "epoch": 15.181137724550899,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1871,
+      "step": 10141
+    },
+    {
+      "epoch": 15.182634730538922,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1883,
+      "step": 10142
+    },
+    {
+      "epoch": 15.184131736526947,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1778,
+      "step": 10143
+    },
+    {
+      "epoch": 15.18562874251497,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.182,
+      "step": 10144
+    },
+    {
+      "epoch": 15.187125748502995,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1829,
+      "step": 10145
+    },
+    {
+      "epoch": 15.188622754491018,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1779,
+      "step": 10146
+    },
+    {
+      "epoch": 15.190119760479043,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1845,
+      "step": 10147
+    },
+    {
+      "epoch": 15.191616766467066,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1856,
+      "step": 10148
+    },
+    {
+      "epoch": 15.19311377245509,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1764,
+      "step": 10149
+    },
+    {
+      "epoch": 15.194610778443113,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1791,
+      "step": 10150
+    },
+    {
+      "epoch": 15.196107784431138,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1829,
+      "step": 10151
+    },
+    {
+      "epoch": 15.197604790419161,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1779,
+      "step": 10152
+    },
+    {
+      "epoch": 15.199101796407186,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1816,
+      "step": 10153
+    },
+    {
+      "epoch": 15.20059880239521,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1851,
+      "step": 10154
+    },
+    {
+      "epoch": 15.202095808383234,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1788,
+      "step": 10155
+    },
+    {
+      "epoch": 15.203592814371257,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1736,
+      "step": 10156
+    },
+    {
+      "epoch": 15.205089820359282,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1862,
+      "step": 10157
+    },
+    {
+      "epoch": 15.206586826347305,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.179,
+      "step": 10158
+    },
+    {
+      "epoch": 15.20808383233533,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1776,
+      "step": 10159
+    },
+    {
+      "epoch": 15.209580838323353,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.172,
+      "step": 10160
+    },
+    {
+      "epoch": 15.211077844311378,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.176,
+      "step": 10161
+    },
+    {
+      "epoch": 15.2125748502994,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1837,
+      "step": 10162
+    },
+    {
+      "epoch": 15.214071856287426,
+      "grad_norm": 0.21484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1855,
+      "step": 10163
+    },
+    {
+      "epoch": 15.215568862275449,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1963,
+      "step": 10164
+    },
+    {
+      "epoch": 15.217065868263473,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1825,
+      "step": 10165
+    },
+    {
+      "epoch": 15.218562874251496,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1807,
+      "step": 10166
+    },
+    {
+      "epoch": 15.220059880239521,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1775,
+      "step": 10167
+    },
+    {
+      "epoch": 15.221556886227544,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1853,
+      "step": 10168
+    },
+    {
+      "epoch": 15.22305389221557,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1803,
+      "step": 10169
+    },
+    {
+      "epoch": 15.224550898203592,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.188,
+      "step": 10170
+    },
+    {
+      "epoch": 15.226047904191617,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1799,
+      "step": 10171
+    },
+    {
+      "epoch": 15.22754491017964,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.175,
+      "step": 10172
+    },
+    {
+      "epoch": 15.229041916167665,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1806,
+      "step": 10173
+    },
+    {
+      "epoch": 15.230538922155688,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0008,
+      "loss": 1.1758,
+      "step": 10174
+    },
+    {
+      "epoch": 15.232035928143713,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1775,
+      "step": 10175
+    },
+    {
+      "epoch": 15.233532934131736,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1863,
+      "step": 10176
+    },
+    {
+      "epoch": 15.23502994011976,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1771,
+      "step": 10177
+    },
+    {
+      "epoch": 15.236526946107784,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1763,
+      "step": 10178
+    },
+    {
+      "epoch": 15.238023952095809,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1832,
+      "step": 10179
+    },
+    {
+      "epoch": 15.239520958083832,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.183,
+      "step": 10180
+    },
+    {
+      "epoch": 15.241017964071856,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1787,
+      "step": 10181
+    },
+    {
+      "epoch": 15.24251497005988,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1767,
+      "step": 10182
+    },
+    {
+      "epoch": 15.244011976047904,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1749,
+      "step": 10183
+    },
+    {
+      "epoch": 15.245508982035927,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1788,
+      "step": 10184
+    },
+    {
+      "epoch": 15.247005988023952,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1799,
+      "step": 10185
+    },
+    {
+      "epoch": 15.248502994011975,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1837,
+      "step": 10186
+    },
+    {
+      "epoch": 15.25,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1717,
+      "step": 10187
+    },
+    {
+      "epoch": 15.251497005988025,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1763,
+      "step": 10188
+    },
+    {
+      "epoch": 15.252994011976048,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1751,
+      "step": 10189
+    },
+    {
+      "epoch": 15.254491017964073,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1748,
+      "step": 10190
+    },
+    {
+      "epoch": 15.255988023952096,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1754,
+      "step": 10191
+    },
+    {
+      "epoch": 15.25748502994012,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1802,
+      "step": 10192
+    },
+    {
+      "epoch": 15.258982035928144,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1813,
+      "step": 10193
+    },
+    {
+      "epoch": 15.260479041916168,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1771,
+      "step": 10194
+    },
+    {
+      "epoch": 15.261976047904191,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1715,
+      "step": 10195
+    },
+    {
+      "epoch": 15.263473053892216,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1703,
+      "step": 10196
+    },
+    {
+      "epoch": 15.26497005988024,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.166,
+      "step": 10197
+    },
+    {
+      "epoch": 15.266467065868264,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1747,
+      "step": 10198
+    },
+    {
+      "epoch": 15.267964071856287,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1742,
+      "step": 10199
+    },
+    {
+      "epoch": 15.269461077844312,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1746,
+      "step": 10200
+    },
+    {
+      "epoch": 15.270958083832335,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1698,
+      "step": 10201
+    },
+    {
+      "epoch": 15.27245508982036,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1697,
+      "step": 10202
+    },
+    {
+      "epoch": 15.273952095808383,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1774,
+      "step": 10203
+    },
+    {
+      "epoch": 15.275449101796408,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1756,
+      "step": 10204
+    },
+    {
+      "epoch": 15.27694610778443,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1756,
+      "step": 10205
+    },
+    {
+      "epoch": 15.278443113772456,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0008,
+      "loss": 1.1811,
+      "step": 10206
+    },
+    {
+      "epoch": 15.279940119760479,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1842,
+      "step": 10207
+    },
+    {
+      "epoch": 15.281437125748504,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.2017,
+      "step": 10208
+    },
+    {
+      "epoch": 15.282934131736527,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.0008,
+      "loss": 1.211,
+      "step": 10209
+    },
+    {
+      "epoch": 15.284431137724551,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1962,
+      "step": 10210
+    },
+    {
+      "epoch": 15.285928143712574,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0008,
+      "loss": 1.2019,
+      "step": 10211
+    },
+    {
+      "epoch": 15.2874251497006,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0008,
+      "loss": 1.19,
+      "step": 10212
+    },
+    {
+      "epoch": 15.288922155688622,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1935,
+      "step": 10213
+    },
+    {
+      "epoch": 15.290419161676647,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1886,
+      "step": 10214
+    },
+    {
+      "epoch": 15.29191616766467,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1925,
+      "step": 10215
+    },
+    {
+      "epoch": 15.293413173652695,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1823,
+      "step": 10216
+    },
+    {
+      "epoch": 15.294910179640718,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1803,
+      "step": 10217
+    },
+    {
+      "epoch": 15.296407185628743,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1814,
+      "step": 10218
+    },
+    {
+      "epoch": 15.297904191616766,
+      "grad_norm": 0.216796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1821,
+      "step": 10219
+    },
+    {
+      "epoch": 15.29940119760479,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1714,
+      "step": 10220
+    },
+    {
+      "epoch": 15.300898203592814,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1753,
+      "step": 10221
+    },
+    {
+      "epoch": 15.302395209580839,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1812,
+      "step": 10222
+    },
+    {
+      "epoch": 15.303892215568862,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1746,
+      "step": 10223
+    },
+    {
+      "epoch": 15.305389221556887,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1741,
+      "step": 10224
+    },
+    {
+      "epoch": 15.30688622754491,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1798,
+      "step": 10225
+    },
+    {
+      "epoch": 15.308383233532934,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1718,
+      "step": 10226
+    },
+    {
+      "epoch": 15.309880239520957,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1733,
+      "step": 10227
+    },
+    {
+      "epoch": 15.311377245508982,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0008,
+      "loss": 1.17,
+      "step": 10228
+    },
+    {
+      "epoch": 15.312874251497005,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1695,
+      "step": 10229
+    },
+    {
+      "epoch": 15.31437125748503,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1696,
+      "step": 10230
+    },
+    {
+      "epoch": 15.315868263473053,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1637,
+      "step": 10231
+    },
+    {
+      "epoch": 15.317365269461078,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1708,
+      "step": 10232
+    },
+    {
+      "epoch": 15.318862275449101,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.172,
+      "step": 10233
+    },
+    {
+      "epoch": 15.320359281437126,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1734,
+      "step": 10234
+    },
+    {
+      "epoch": 15.321856287425149,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1715,
+      "step": 10235
+    },
+    {
+      "epoch": 15.323353293413174,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1648,
+      "step": 10236
+    },
+    {
+      "epoch": 15.324850299401197,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1736,
+      "step": 10237
+    },
+    {
+      "epoch": 15.326347305389222,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1684,
+      "step": 10238
+    },
+    {
+      "epoch": 15.327844311377245,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1694,
+      "step": 10239
+    },
+    {
+      "epoch": 15.32934131736527,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1621,
+      "step": 10240
+    },
+    {
+      "epoch": 15.330838323353294,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1726,
+      "step": 10241
+    },
+    {
+      "epoch": 15.332335329341317,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1757,
+      "step": 10242
+    },
+    {
+      "epoch": 15.33383233532934,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1782,
+      "step": 10243
+    },
+    {
+      "epoch": 15.335329341317365,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1713,
+      "step": 10244
+    },
+    {
+      "epoch": 15.33682634730539,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1693,
+      "step": 10245
+    },
+    {
+      "epoch": 15.338323353293413,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1737,
+      "step": 10246
+    },
+    {
+      "epoch": 15.339820359281438,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1747,
+      "step": 10247
+    },
+    {
+      "epoch": 15.341317365269461,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1707,
+      "step": 10248
+    },
+    {
+      "epoch": 15.342814371257486,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.164,
+      "step": 10249
+    },
+    {
+      "epoch": 15.344311377245509,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1663,
+      "step": 10250
+    },
+    {
+      "epoch": 15.345808383233534,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1603,
+      "step": 10251
+    },
+    {
+      "epoch": 15.347305389221557,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1664,
+      "step": 10252
+    },
+    {
+      "epoch": 15.348802395209582,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1656,
+      "step": 10253
+    },
+    {
+      "epoch": 15.350299401197605,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1663,
+      "step": 10254
+    },
+    {
+      "epoch": 15.35179640718563,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1658,
+      "step": 10255
+    },
+    {
+      "epoch": 15.353293413173652,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1655,
+      "step": 10256
+    },
+    {
+      "epoch": 15.354790419161677,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1631,
+      "step": 10257
+    },
+    {
+      "epoch": 15.3562874251497,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1724,
+      "step": 10258
+    },
+    {
+      "epoch": 15.357784431137725,
+      "grad_norm": 0.18359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1674,
+      "step": 10259
+    },
+    {
+      "epoch": 15.359281437125748,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1656,
+      "step": 10260
+    },
+    {
+      "epoch": 15.360778443113773,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1653,
+      "step": 10261
+    },
+    {
+      "epoch": 15.362275449101796,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1645,
+      "step": 10262
+    },
+    {
+      "epoch": 15.363772455089821,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1746,
+      "step": 10263
+    },
+    {
+      "epoch": 15.365269461077844,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1789,
+      "step": 10264
+    },
+    {
+      "epoch": 15.366766467065869,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1737,
+      "step": 10265
+    },
+    {
+      "epoch": 15.368263473053892,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1677,
+      "step": 10266
+    },
+    {
+      "epoch": 15.369760479041917,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1637,
+      "step": 10267
+    },
+    {
+      "epoch": 15.37125748502994,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1624,
+      "step": 10268
+    },
+    {
+      "epoch": 15.372754491017965,
+      "grad_norm": 0.21875,
+      "learning_rate": 0.0008,
+      "loss": 1.1657,
+      "step": 10269
+    },
+    {
+      "epoch": 15.374251497005988,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0008,
+      "loss": 1.164,
+      "step": 10270
+    },
+    {
+      "epoch": 15.375748502994012,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1686,
+      "step": 10271
+    },
+    {
+      "epoch": 15.377245508982035,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1666,
+      "step": 10272
+    },
+    {
+      "epoch": 15.37874251497006,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1666,
+      "step": 10273
+    },
+    {
+      "epoch": 15.380239520958083,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.159,
+      "step": 10274
+    },
+    {
+      "epoch": 15.381736526946108,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.163,
+      "step": 10275
+    },
+    {
+      "epoch": 15.383233532934131,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1669,
+      "step": 10276
+    },
+    {
+      "epoch": 15.384730538922156,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1725,
+      "step": 10277
+    },
+    {
+      "epoch": 15.386227544910179,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1594,
+      "step": 10278
+    },
+    {
+      "epoch": 15.387724550898204,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1586,
+      "step": 10279
+    },
+    {
+      "epoch": 15.389221556886227,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1613,
+      "step": 10280
+    },
+    {
+      "epoch": 15.390718562874252,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1625,
+      "step": 10281
+    },
+    {
+      "epoch": 15.392215568862275,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1702,
+      "step": 10282
+    },
+    {
+      "epoch": 15.3937125748503,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1576,
+      "step": 10283
+    },
+    {
+      "epoch": 15.395209580838323,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1593,
+      "step": 10284
+    },
+    {
+      "epoch": 15.396706586826348,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1568,
+      "step": 10285
+    },
+    {
+      "epoch": 15.39820359281437,
+      "grad_norm": 0.1875,
+      "learning_rate": 0.0008,
+      "loss": 1.1603,
+      "step": 10286
+    },
+    {
+      "epoch": 15.399700598802395,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1577,
+      "step": 10287
+    },
+    {
+      "epoch": 15.401197604790418,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1656,
+      "step": 10288
+    },
+    {
+      "epoch": 15.402694610778443,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1624,
+      "step": 10289
+    },
+    {
+      "epoch": 15.404191616766466,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1657,
+      "step": 10290
+    },
+    {
+      "epoch": 15.405688622754491,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1584,
+      "step": 10291
+    },
+    {
+      "epoch": 15.407185628742514,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1592,
+      "step": 10292
+    },
+    {
+      "epoch": 15.408682634730539,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1533,
+      "step": 10293
+    },
+    {
+      "epoch": 15.410179640718562,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1605,
+      "step": 10294
+    },
+    {
+      "epoch": 15.411676646706587,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1563,
+      "step": 10295
+    },
+    {
+      "epoch": 15.41317365269461,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1627,
+      "step": 10296
+    },
+    {
+      "epoch": 15.414670658682635,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1555,
+      "step": 10297
+    },
+    {
+      "epoch": 15.41616766467066,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.157,
+      "step": 10298
+    },
+    {
+      "epoch": 15.417664670658683,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1598,
+      "step": 10299
+    },
+    {
+      "epoch": 15.419161676646706,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1579,
+      "step": 10300
+    },
+    {
+      "epoch": 15.42065868263473,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1524,
+      "step": 10301
+    },
+    {
+      "epoch": 15.422155688622755,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1597,
+      "step": 10302
+    },
+    {
+      "epoch": 15.423652694610778,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1601,
+      "step": 10303
+    },
+    {
+      "epoch": 15.425149700598803,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1565,
+      "step": 10304
+    },
+    {
+      "epoch": 15.426646706586826,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1585,
+      "step": 10305
+    },
+    {
+      "epoch": 15.428143712574851,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1529,
+      "step": 10306
+    },
+    {
+      "epoch": 15.429640718562874,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1569,
+      "step": 10307
+    },
+    {
+      "epoch": 15.431137724550899,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1521,
+      "step": 10308
+    },
+    {
+      "epoch": 15.432634730538922,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1514,
+      "step": 10309
+    },
+    {
+      "epoch": 15.434131736526947,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1517,
+      "step": 10310
+    },
+    {
+      "epoch": 15.43562874251497,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1582,
+      "step": 10311
+    },
+    {
+      "epoch": 15.437125748502995,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1616,
+      "step": 10312
+    },
+    {
+      "epoch": 15.438622754491018,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1554,
+      "step": 10313
+    },
+    {
+      "epoch": 15.440119760479043,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.152,
+      "step": 10314
+    },
+    {
+      "epoch": 15.441616766467066,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1562,
+      "step": 10315
+    },
+    {
+      "epoch": 15.44311377245509,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1528,
+      "step": 10316
+    },
+    {
+      "epoch": 15.444610778443113,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1558,
+      "step": 10317
+    },
+    {
+      "epoch": 15.446107784431138,
+      "grad_norm": 0.24609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1567,
+      "step": 10318
+    },
+    {
+      "epoch": 15.447604790419161,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1573,
+      "step": 10319
+    },
+    {
+      "epoch": 15.449101796407186,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1663,
+      "step": 10320
+    },
+    {
+      "epoch": 15.45059880239521,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1592,
+      "step": 10321
+    },
+    {
+      "epoch": 15.452095808383234,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1534,
+      "step": 10322
+    },
+    {
+      "epoch": 15.453592814371257,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1607,
+      "step": 10323
+    },
+    {
+      "epoch": 15.455089820359282,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0008,
+      "loss": 1.1545,
+      "step": 10324
+    },
+    {
+      "epoch": 15.456586826347305,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1568,
+      "step": 10325
+    },
+    {
+      "epoch": 15.45808383233533,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1555,
+      "step": 10326
+    },
+    {
+      "epoch": 15.459580838323353,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1584,
+      "step": 10327
+    },
+    {
+      "epoch": 15.461077844311378,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0008,
+      "loss": 1.16,
+      "step": 10328
+    },
+    {
+      "epoch": 15.4625748502994,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1597,
+      "step": 10329
+    },
+    {
+      "epoch": 15.464071856287426,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1532,
+      "step": 10330
+    },
+    {
+      "epoch": 15.465568862275449,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0008,
+      "loss": 1.1508,
+      "step": 10331
+    },
+    {
+      "epoch": 15.467065868263473,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1585,
+      "step": 10332
+    },
+    {
+      "epoch": 15.468562874251496,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1617,
+      "step": 10333
+    },
+    {
+      "epoch": 15.470059880239521,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1609,
+      "step": 10334
+    },
+    {
+      "epoch": 15.471556886227544,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1608,
+      "step": 10335
+    },
+    {
+      "epoch": 15.47305389221557,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1592,
+      "step": 10336
+    },
+    {
+      "epoch": 15.474550898203592,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0008,
+      "loss": 1.1598,
+      "step": 10337
+    },
+    {
+      "epoch": 15.476047904191617,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.156,
+      "step": 10338
+    },
+    {
+      "epoch": 15.47754491017964,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0008,
+      "loss": 1.16,
+      "step": 10339
+    },
+    {
+      "epoch": 15.479041916167665,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0008,
+      "loss": 1.161,
+      "step": 10340
+    },
+    {
+      "epoch": 15.480538922155688,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1555,
+      "step": 10341
+    },
+    {
+      "epoch": 15.482035928143713,
+      "grad_norm": 0.212890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1576,
+      "step": 10342
+    },
+    {
+      "epoch": 15.483532934131736,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1636,
+      "step": 10343
+    },
+    {
+      "epoch": 15.48502994011976,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1559,
+      "step": 10344
+    },
+    {
+      "epoch": 15.486526946107784,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1585,
+      "step": 10345
+    },
+    {
+      "epoch": 15.488023952095809,
+      "grad_norm": 0.20703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1547,
+      "step": 10346
+    },
+    {
+      "epoch": 15.489520958083832,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1529,
+      "step": 10347
+    },
+    {
+      "epoch": 15.491017964071856,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1527,
+      "step": 10348
+    },
+    {
+      "epoch": 15.49251497005988,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1539,
+      "step": 10349
+    },
+    {
+      "epoch": 15.494011976047904,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1589,
+      "step": 10350
+    },
+    {
+      "epoch": 15.495508982035927,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0008,
+      "loss": 1.161,
+      "step": 10351
+    },
+    {
+      "epoch": 15.497005988023952,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0008,
+      "loss": 1.151,
+      "step": 10352
+    },
+    {
+      "epoch": 15.498502994011975,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1538,
+      "step": 10353
+    },
+    {
+      "epoch": 15.5,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1566,
+      "step": 10354
+    },
+    {
+      "epoch": 15.501497005988025,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.154,
+      "step": 10355
+    },
+    {
+      "epoch": 15.502994011976048,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0008,
+      "loss": 1.155,
+      "step": 10356
+    },
+    {
+      "epoch": 15.504491017964071,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1457,
+      "step": 10357
+    },
+    {
+      "epoch": 15.505988023952096,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1486,
+      "step": 10358
+    },
+    {
+      "epoch": 15.50748502994012,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1554,
+      "step": 10359
+    },
+    {
+      "epoch": 15.508982035928144,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1525,
+      "step": 10360
+    },
+    {
+      "epoch": 15.510479041916168,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1507,
+      "step": 10361
+    },
+    {
+      "epoch": 15.511976047904191,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1575,
+      "step": 10362
+    },
+    {
+      "epoch": 15.513473053892216,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.155,
+      "step": 10363
+    },
+    {
+      "epoch": 15.51497005988024,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1564,
+      "step": 10364
+    },
+    {
+      "epoch": 15.516467065868264,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.152,
+      "step": 10365
+    },
+    {
+      "epoch": 15.517964071856287,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1519,
+      "step": 10366
+    },
+    {
+      "epoch": 15.519461077844312,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1526,
+      "step": 10367
+    },
+    {
+      "epoch": 15.520958083832335,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1488,
+      "step": 10368
+    },
+    {
+      "epoch": 15.52245508982036,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1573,
+      "step": 10369
+    },
+    {
+      "epoch": 15.523952095808383,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1499,
+      "step": 10370
+    },
+    {
+      "epoch": 15.525449101796408,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1482,
+      "step": 10371
+    },
+    {
+      "epoch": 15.52694610778443,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1481,
+      "step": 10372
+    },
+    {
+      "epoch": 15.528443113772456,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1562,
+      "step": 10373
+    },
+    {
+      "epoch": 15.529940119760479,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1533,
+      "step": 10374
+    },
+    {
+      "epoch": 15.531437125748504,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 10375
+    },
+    {
+      "epoch": 15.532934131736527,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1512,
+      "step": 10376
+    },
+    {
+      "epoch": 15.534431137724551,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1476,
+      "step": 10377
+    },
+    {
+      "epoch": 15.535928143712574,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1492,
+      "step": 10378
+    },
+    {
+      "epoch": 15.5374251497006,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1475,
+      "step": 10379
+    },
+    {
+      "epoch": 15.538922155688622,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1484,
+      "step": 10380
+    },
+    {
+      "epoch": 15.540419161676647,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1562,
+      "step": 10381
+    },
+    {
+      "epoch": 15.54191616766467,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1464,
+      "step": 10382
+    },
+    {
+      "epoch": 15.543413173652695,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 10383
+    },
+    {
+      "epoch": 15.544910179640718,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1546,
+      "step": 10384
+    },
+    {
+      "epoch": 15.546407185628743,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 10385
+    },
+    {
+      "epoch": 15.547904191616766,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 10386
+    },
+    {
+      "epoch": 15.54940119760479,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1559,
+      "step": 10387
+    },
+    {
+      "epoch": 15.550898203592814,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 10388
+    },
+    {
+      "epoch": 15.552395209580839,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1505,
+      "step": 10389
+    },
+    {
+      "epoch": 15.553892215568862,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1525,
+      "step": 10390
+    },
+    {
+      "epoch": 15.555389221556887,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1492,
+      "step": 10391
+    },
+    {
+      "epoch": 15.55688622754491,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1506,
+      "step": 10392
+    },
+    {
+      "epoch": 15.558383233532934,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1434,
+      "step": 10393
+    },
+    {
+      "epoch": 15.559880239520957,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1532,
+      "step": 10394
+    },
+    {
+      "epoch": 15.561377245508982,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1472,
+      "step": 10395
+    },
+    {
+      "epoch": 15.562874251497005,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1519,
+      "step": 10396
+    },
+    {
+      "epoch": 15.56437125748503,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1481,
+      "step": 10397
+    },
+    {
+      "epoch": 15.565868263473053,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0008,
+      "loss": 1.153,
+      "step": 10398
+    },
+    {
+      "epoch": 15.567365269461078,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 10399
+    },
+    {
+      "epoch": 15.568862275449101,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1464,
+      "step": 10400
+    },
+    {
+      "epoch": 15.570359281437126,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1493,
+      "step": 10401
+    },
+    {
+      "epoch": 15.571856287425149,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 10402
+    },
+    {
+      "epoch": 15.573353293413174,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1519,
+      "step": 10403
+    },
+    {
+      "epoch": 15.574850299401197,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1465,
+      "step": 10404
+    },
+    {
+      "epoch": 15.576347305389222,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0008,
+      "loss": 1.148,
+      "step": 10405
+    },
+    {
+      "epoch": 15.577844311377245,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1472,
+      "step": 10406
+    },
+    {
+      "epoch": 15.57934131736527,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 10407
+    },
+    {
+      "epoch": 15.580838323353294,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 10408
+    },
+    {
+      "epoch": 15.582335329341317,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1511,
+      "step": 10409
+    },
+    {
+      "epoch": 15.58383233532934,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1476,
+      "step": 10410
+    },
+    {
+      "epoch": 15.585329341317365,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1513,
+      "step": 10411
+    },
+    {
+      "epoch": 15.58682634730539,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1509,
+      "step": 10412
+    },
+    {
+      "epoch": 15.588323353293413,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1504,
+      "step": 10413
+    },
+    {
+      "epoch": 15.589820359281438,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1544,
+      "step": 10414
+    },
+    {
+      "epoch": 15.591317365269461,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.151,
+      "step": 10415
+    },
+    {
+      "epoch": 15.592814371257486,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1552,
+      "step": 10416
+    },
+    {
+      "epoch": 15.594311377245509,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 10417
+    },
+    {
+      "epoch": 15.595808383233534,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1487,
+      "step": 10418
+    },
+    {
+      "epoch": 15.597305389221557,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 10419
+    },
+    {
+      "epoch": 15.598802395209582,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1476,
+      "step": 10420
+    },
+    {
+      "epoch": 15.600299401197605,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 10421
+    },
+    {
+      "epoch": 15.60179640718563,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1488,
+      "step": 10422
+    },
+    {
+      "epoch": 15.603293413173652,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.153,
+      "step": 10423
+    },
+    {
+      "epoch": 15.604790419161677,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1501,
+      "step": 10424
+    },
+    {
+      "epoch": 15.6062874251497,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 10425
+    },
+    {
+      "epoch": 15.607784431137725,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 10426
+    },
+    {
+      "epoch": 15.609281437125748,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1471,
+      "step": 10427
+    },
+    {
+      "epoch": 15.610778443113773,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1495,
+      "step": 10428
+    },
+    {
+      "epoch": 15.612275449101796,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 10429
+    },
+    {
+      "epoch": 15.613772455089821,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.157,
+      "step": 10430
+    },
+    {
+      "epoch": 15.615269461077844,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1513,
+      "step": 10431
+    },
+    {
+      "epoch": 15.616766467065869,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.151,
+      "step": 10432
+    },
+    {
+      "epoch": 15.618263473053892,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 10433
+    },
+    {
+      "epoch": 15.619760479041917,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1483,
+      "step": 10434
+    },
+    {
+      "epoch": 15.62125748502994,
+      "grad_norm": 0.19921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1463,
+      "step": 10435
+    },
+    {
+      "epoch": 15.622754491017965,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1538,
+      "step": 10436
+    },
+    {
+      "epoch": 15.624251497005988,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1529,
+      "step": 10437
+    },
+    {
+      "epoch": 15.625748502994012,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 10438
+    },
+    {
+      "epoch": 15.627245508982035,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.144,
+      "step": 10439
+    },
+    {
+      "epoch": 15.62874251497006,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1475,
+      "step": 10440
+    },
+    {
+      "epoch": 15.630239520958083,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 10441
+    },
+    {
+      "epoch": 15.631736526946108,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.148,
+      "step": 10442
+    },
+    {
+      "epoch": 15.633233532934131,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1465,
+      "step": 10443
+    },
+    {
+      "epoch": 15.634730538922156,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 10444
+    },
+    {
+      "epoch": 15.636227544910179,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 10445
+    },
+    {
+      "epoch": 15.637724550898204,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 10446
+    },
+    {
+      "epoch": 15.639221556886227,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1459,
+      "step": 10447
+    },
+    {
+      "epoch": 15.640718562874252,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1487,
+      "step": 10448
+    },
+    {
+      "epoch": 15.642215568862275,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1483,
+      "step": 10449
+    },
+    {
+      "epoch": 15.6437125748503,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1457,
+      "step": 10450
+    },
+    {
+      "epoch": 15.645209580838323,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 10451
+    },
+    {
+      "epoch": 15.646706586826348,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1379,
+      "step": 10452
+    },
+    {
+      "epoch": 15.64820359281437,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1488,
+      "step": 10453
+    },
+    {
+      "epoch": 15.649700598802395,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1457,
+      "step": 10454
+    },
+    {
+      "epoch": 15.651197604790418,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 10455
+    },
+    {
+      "epoch": 15.652694610778443,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 10456
+    },
+    {
+      "epoch": 15.654191616766466,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1507,
+      "step": 10457
+    },
+    {
+      "epoch": 15.655688622754491,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 10458
+    },
+    {
+      "epoch": 15.657185628742514,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1534,
+      "step": 10459
+    },
+    {
+      "epoch": 15.658682634730539,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1493,
+      "step": 10460
+    },
+    {
+      "epoch": 15.660179640718562,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 10461
+    },
+    {
+      "epoch": 15.661676646706587,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1468,
+      "step": 10462
+    },
+    {
+      "epoch": 15.66317365269461,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 10463
+    },
+    {
+      "epoch": 15.664670658682635,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1491,
+      "step": 10464
+    },
+    {
+      "epoch": 15.66616766467066,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1479,
+      "step": 10465
+    },
+    {
+      "epoch": 15.667664670658683,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 10466
+    },
+    {
+      "epoch": 15.669161676646706,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1496,
+      "step": 10467
+    },
+    {
+      "epoch": 15.67065868263473,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 10468
+    },
+    {
+      "epoch": 15.672155688622755,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 10469
+    },
+    {
+      "epoch": 15.673652694610778,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 10470
+    },
+    {
+      "epoch": 15.675149700598803,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 10471
+    },
+    {
+      "epoch": 15.676646706586826,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.151,
+      "step": 10472
+    },
+    {
+      "epoch": 15.678143712574851,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 10473
+    },
+    {
+      "epoch": 15.679640718562874,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1443,
+      "step": 10474
+    },
+    {
+      "epoch": 15.681137724550899,
+      "grad_norm": 0.138671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 10475
+    },
+    {
+      "epoch": 15.682634730538922,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1457,
+      "step": 10476
+    },
+    {
+      "epoch": 15.684131736526947,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1479,
+      "step": 10477
+    },
+    {
+      "epoch": 15.68562874251497,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 10478
+    },
+    {
+      "epoch": 15.687125748502995,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1446,
+      "step": 10479
+    },
+    {
+      "epoch": 15.688622754491018,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1518,
+      "step": 10480
+    },
+    {
+      "epoch": 15.690119760479043,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1494,
+      "step": 10481
+    },
+    {
+      "epoch": 15.691616766467066,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 10482
+    },
+    {
+      "epoch": 15.69311377245509,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1471,
+      "step": 10483
+    },
+    {
+      "epoch": 15.694610778443113,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 10484
+    },
+    {
+      "epoch": 15.696107784431138,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1413,
+      "step": 10485
+    },
+    {
+      "epoch": 15.697604790419161,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 10486
+    },
+    {
+      "epoch": 15.699101796407186,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 10487
+    },
+    {
+      "epoch": 15.70059880239521,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1536,
+      "step": 10488
+    },
+    {
+      "epoch": 15.702095808383234,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0008,
+      "loss": 1.144,
+      "step": 10489
+    },
+    {
+      "epoch": 15.703592814371257,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0008,
+      "loss": 1.15,
+      "step": 10490
+    },
+    {
+      "epoch": 15.705089820359282,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 10491
+    },
+    {
+      "epoch": 15.706586826347305,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1479,
+      "step": 10492
+    },
+    {
+      "epoch": 15.70808383233533,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 10493
+    },
+    {
+      "epoch": 15.709580838323353,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1487,
+      "step": 10494
+    },
+    {
+      "epoch": 15.711077844311378,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1413,
+      "step": 10495
+    },
+    {
+      "epoch": 15.7125748502994,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.14,
+      "step": 10496
+    },
+    {
+      "epoch": 15.714071856287426,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 10497
+    },
+    {
+      "epoch": 15.715568862275449,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 10498
+    },
+    {
+      "epoch": 15.717065868263473,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1502,
+      "step": 10499
+    },
+    {
+      "epoch": 15.718562874251496,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1477,
+      "step": 10500
+    },
+    {
+      "epoch": 15.720059880239521,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 10501
+    },
+    {
+      "epoch": 15.721556886227544,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 10502
+    },
+    {
+      "epoch": 15.72305389221557,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1473,
+      "step": 10503
+    },
+    {
+      "epoch": 15.724550898203592,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 10504
+    },
+    {
+      "epoch": 15.726047904191617,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 10505
+    },
+    {
+      "epoch": 15.72754491017964,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 10506
+    },
+    {
+      "epoch": 15.729041916167665,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 10507
+    },
+    {
+      "epoch": 15.730538922155688,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 10508
+    },
+    {
+      "epoch": 15.732035928143713,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1428,
+      "step": 10509
+    },
+    {
+      "epoch": 15.733532934131736,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1477,
+      "step": 10510
+    },
+    {
+      "epoch": 15.73502994011976,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1471,
+      "step": 10511
+    },
+    {
+      "epoch": 15.736526946107784,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1534,
+      "step": 10512
+    },
+    {
+      "epoch": 15.738023952095809,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1514,
+      "step": 10513
+    },
+    {
+      "epoch": 15.739520958083832,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 10514
+    },
+    {
+      "epoch": 15.741017964071856,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 10515
+    },
+    {
+      "epoch": 15.74251497005988,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1485,
+      "step": 10516
+    },
+    {
+      "epoch": 15.744011976047904,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 10517
+    },
+    {
+      "epoch": 15.745508982035929,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 10518
+    },
+    {
+      "epoch": 15.747005988023952,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 10519
+    },
+    {
+      "epoch": 15.748502994011975,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 10520
+    },
+    {
+      "epoch": 15.75,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 10521
+    },
+    {
+      "epoch": 15.751497005988025,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 10522
+    },
+    {
+      "epoch": 15.752994011976048,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1486,
+      "step": 10523
+    },
+    {
+      "epoch": 15.754491017964071,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 10524
+    },
+    {
+      "epoch": 15.755988023952096,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1365,
+      "step": 10525
+    },
+    {
+      "epoch": 15.75748502994012,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1507,
+      "step": 10526
+    },
+    {
+      "epoch": 15.758982035928144,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 10527
+    },
+    {
+      "epoch": 15.760479041916168,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1464,
+      "step": 10528
+    },
+    {
+      "epoch": 15.761976047904191,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1475,
+      "step": 10529
+    },
+    {
+      "epoch": 15.763473053892216,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1485,
+      "step": 10530
+    },
+    {
+      "epoch": 15.76497005988024,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 10531
+    },
+    {
+      "epoch": 15.766467065868264,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1464,
+      "step": 10532
+    },
+    {
+      "epoch": 15.767964071856287,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 10533
+    },
+    {
+      "epoch": 15.769461077844312,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1496,
+      "step": 10534
+    },
+    {
+      "epoch": 15.770958083832335,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 10535
+    },
+    {
+      "epoch": 15.77245508982036,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1477,
+      "step": 10536
+    },
+    {
+      "epoch": 15.773952095808383,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.144,
+      "step": 10537
+    },
+    {
+      "epoch": 15.775449101796408,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1485,
+      "step": 10538
+    },
+    {
+      "epoch": 15.77694610778443,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 10539
+    },
+    {
+      "epoch": 15.778443113772456,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1416,
+      "step": 10540
+    },
+    {
+      "epoch": 15.779940119760479,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 10541
+    },
+    {
+      "epoch": 15.781437125748504,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 10542
+    },
+    {
+      "epoch": 15.782934131736527,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1454,
+      "step": 10543
+    },
+    {
+      "epoch": 15.784431137724551,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1494,
+      "step": 10544
+    },
+    {
+      "epoch": 15.785928143712574,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 10545
+    },
+    {
+      "epoch": 15.7874251497006,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 10546
+    },
+    {
+      "epoch": 15.788922155688622,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1499,
+      "step": 10547
+    },
+    {
+      "epoch": 15.790419161676647,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 10548
+    },
+    {
+      "epoch": 15.79191616766467,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.146,
+      "step": 10549
+    },
+    {
+      "epoch": 15.793413173652695,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 10550
+    },
+    {
+      "epoch": 15.794910179640718,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1454,
+      "step": 10551
+    },
+    {
+      "epoch": 15.796407185628743,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 10552
+    },
+    {
+      "epoch": 15.797904191616766,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1505,
+      "step": 10553
+    },
+    {
+      "epoch": 15.79940119760479,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 10554
+    },
+    {
+      "epoch": 15.800898203592814,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1446,
+      "step": 10555
+    },
+    {
+      "epoch": 15.802395209580839,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1507,
+      "step": 10556
+    },
+    {
+      "epoch": 15.803892215568862,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 10557
+    },
+    {
+      "epoch": 15.805389221556887,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 10558
+    },
+    {
+      "epoch": 15.80688622754491,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 10559
+    },
+    {
+      "epoch": 15.808383233532934,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.148,
+      "step": 10560
+    },
+    {
+      "epoch": 15.809880239520957,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1465,
+      "step": 10561
+    },
+    {
+      "epoch": 15.811377245508982,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 10562
+    },
+    {
+      "epoch": 15.812874251497005,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 10563
+    },
+    {
+      "epoch": 15.81437125748503,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 10564
+    },
+    {
+      "epoch": 15.815868263473053,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1533,
+      "step": 10565
+    },
+    {
+      "epoch": 15.817365269461078,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1393,
+      "step": 10566
+    },
+    {
+      "epoch": 15.818862275449101,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 10567
+    },
+    {
+      "epoch": 15.820359281437126,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 10568
+    },
+    {
+      "epoch": 15.821856287425149,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 10569
+    },
+    {
+      "epoch": 15.823353293413174,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 10570
+    },
+    {
+      "epoch": 15.824850299401197,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 10571
+    },
+    {
+      "epoch": 15.826347305389222,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1359,
+      "step": 10572
+    },
+    {
+      "epoch": 15.827844311377245,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 10573
+    },
+    {
+      "epoch": 15.82934131736527,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1375,
+      "step": 10574
+    },
+    {
+      "epoch": 15.830838323353294,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.144,
+      "step": 10575
+    },
+    {
+      "epoch": 15.832335329341317,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1383,
+      "step": 10576
+    },
+    {
+      "epoch": 15.83383233532934,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 10577
+    },
+    {
+      "epoch": 15.835329341317365,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 10578
+    },
+    {
+      "epoch": 15.83682634730539,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1489,
+      "step": 10579
+    },
+    {
+      "epoch": 15.838323353293413,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1457,
+      "step": 10580
+    },
+    {
+      "epoch": 15.839820359281438,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 10581
+    },
+    {
+      "epoch": 15.841317365269461,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.14,
+      "step": 10582
+    },
+    {
+      "epoch": 15.842814371257486,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1501,
+      "step": 10583
+    },
+    {
+      "epoch": 15.844311377245509,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 10584
+    },
+    {
+      "epoch": 15.845808383233534,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 10585
+    },
+    {
+      "epoch": 15.847305389221557,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 10586
+    },
+    {
+      "epoch": 15.848802395209582,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 10587
+    },
+    {
+      "epoch": 15.850299401197605,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1506,
+      "step": 10588
+    },
+    {
+      "epoch": 15.85179640718563,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1476,
+      "step": 10589
+    },
+    {
+      "epoch": 15.853293413173652,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 10590
+    },
+    {
+      "epoch": 15.854790419161677,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 10591
+    },
+    {
+      "epoch": 15.8562874251497,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 10592
+    },
+    {
+      "epoch": 15.857784431137725,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.149,
+      "step": 10593
+    },
+    {
+      "epoch": 15.859281437125748,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 10594
+    },
+    {
+      "epoch": 15.860778443113773,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 10595
+    },
+    {
+      "epoch": 15.862275449101796,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1472,
+      "step": 10596
+    },
+    {
+      "epoch": 15.863772455089821,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 10597
+    },
+    {
+      "epoch": 15.865269461077844,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1476,
+      "step": 10598
+    },
+    {
+      "epoch": 15.866766467065869,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 10599
+    },
+    {
+      "epoch": 15.868263473053892,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 10600
+    },
+    {
+      "epoch": 15.869760479041917,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1487,
+      "step": 10601
+    },
+    {
+      "epoch": 15.87125748502994,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 10602
+    },
+    {
+      "epoch": 15.872754491017965,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1502,
+      "step": 10603
+    },
+    {
+      "epoch": 15.874251497005988,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 10604
+    },
+    {
+      "epoch": 15.875748502994012,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1485,
+      "step": 10605
+    },
+    {
+      "epoch": 15.877245508982035,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1384,
+      "step": 10606
+    },
+    {
+      "epoch": 15.87874251497006,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 10607
+    },
+    {
+      "epoch": 15.880239520958083,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 10608
+    },
+    {
+      "epoch": 15.881736526946108,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1429,
+      "step": 10609
+    },
+    {
+      "epoch": 15.883233532934131,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 10610
+    },
+    {
+      "epoch": 15.884730538922156,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 10611
+    },
+    {
+      "epoch": 15.886227544910179,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 10612
+    },
+    {
+      "epoch": 15.887724550898204,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1476,
+      "step": 10613
+    },
+    {
+      "epoch": 15.889221556886227,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 10614
+    },
+    {
+      "epoch": 15.890718562874252,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 10615
+    },
+    {
+      "epoch": 15.892215568862275,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 10616
+    },
+    {
+      "epoch": 15.8937125748503,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 10617
+    },
+    {
+      "epoch": 15.895209580838323,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 10618
+    },
+    {
+      "epoch": 15.896706586826348,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1463,
+      "step": 10619
+    },
+    {
+      "epoch": 15.89820359281437,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 10620
+    },
+    {
+      "epoch": 15.899700598802395,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1395,
+      "step": 10621
+    },
+    {
+      "epoch": 15.901197604790418,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 10622
+    },
+    {
+      "epoch": 15.902694610778443,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1541,
+      "step": 10623
+    },
+    {
+      "epoch": 15.904191616766466,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 10624
+    },
+    {
+      "epoch": 15.905688622754491,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 10625
+    },
+    {
+      "epoch": 15.907185628742514,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1463,
+      "step": 10626
+    },
+    {
+      "epoch": 15.908682634730539,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 10627
+    },
+    {
+      "epoch": 15.910179640718562,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 10628
+    },
+    {
+      "epoch": 15.911676646706587,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 10629
+    },
+    {
+      "epoch": 15.91317365269461,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 10630
+    },
+    {
+      "epoch": 15.914670658682635,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 10631
+    },
+    {
+      "epoch": 15.91616766467066,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 10632
+    },
+    {
+      "epoch": 15.917664670658683,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1464,
+      "step": 10633
+    },
+    {
+      "epoch": 15.919161676646706,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1459,
+      "step": 10634
+    },
+    {
+      "epoch": 15.92065868263473,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1429,
+      "step": 10635
+    },
+    {
+      "epoch": 15.922155688622755,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 10636
+    },
+    {
+      "epoch": 15.923652694610778,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 10637
+    },
+    {
+      "epoch": 15.925149700598803,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 10638
+    },
+    {
+      "epoch": 15.926646706586826,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 10639
+    },
+    {
+      "epoch": 15.928143712574851,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 10640
+    },
+    {
+      "epoch": 15.929640718562874,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 10641
+    },
+    {
+      "epoch": 15.931137724550899,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1486,
+      "step": 10642
+    },
+    {
+      "epoch": 15.932634730538922,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 10643
+    },
+    {
+      "epoch": 15.934131736526947,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 10644
+    },
+    {
+      "epoch": 15.93562874251497,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 10645
+    },
+    {
+      "epoch": 15.937125748502995,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1392,
+      "step": 10646
+    },
+    {
+      "epoch": 15.938622754491018,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 10647
+    },
+    {
+      "epoch": 15.940119760479043,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 10648
+    },
+    {
+      "epoch": 15.941616766467066,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.146,
+      "step": 10649
+    },
+    {
+      "epoch": 15.94311377245509,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 10650
+    },
+    {
+      "epoch": 15.944610778443113,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 10651
+    },
+    {
+      "epoch": 15.946107784431138,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 10652
+    },
+    {
+      "epoch": 15.947604790419161,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1378,
+      "step": 10653
+    },
+    {
+      "epoch": 15.949101796407186,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 10654
+    },
+    {
+      "epoch": 15.95059880239521,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 10655
+    },
+    {
+      "epoch": 15.952095808383234,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 10656
+    },
+    {
+      "epoch": 15.953592814371257,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1413,
+      "step": 10657
+    },
+    {
+      "epoch": 15.955089820359282,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1416,
+      "step": 10658
+    },
+    {
+      "epoch": 15.956586826347305,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 10659
+    },
+    {
+      "epoch": 15.95808383233533,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 10660
+    },
+    {
+      "epoch": 15.959580838323353,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1465,
+      "step": 10661
+    },
+    {
+      "epoch": 15.961077844311378,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1386,
+      "step": 10662
+    },
+    {
+      "epoch": 15.9625748502994,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 10663
+    },
+    {
+      "epoch": 15.964071856287426,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1465,
+      "step": 10664
+    },
+    {
+      "epoch": 15.965568862275449,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.15,
+      "step": 10665
+    },
+    {
+      "epoch": 15.967065868263473,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 10666
+    },
+    {
+      "epoch": 15.968562874251496,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1533,
+      "step": 10667
+    },
+    {
+      "epoch": 15.970059880239521,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 10668
+    },
+    {
+      "epoch": 15.971556886227544,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 10669
+    },
+    {
+      "epoch": 15.97305389221557,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 10670
+    },
+    {
+      "epoch": 15.974550898203592,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 10671
+    },
+    {
+      "epoch": 15.976047904191617,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 10672
+    },
+    {
+      "epoch": 15.97754491017964,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 10673
+    },
+    {
+      "epoch": 15.979041916167665,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 10674
+    },
+    {
+      "epoch": 15.980538922155688,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 10675
+    },
+    {
+      "epoch": 15.982035928143713,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1429,
+      "step": 10676
+    },
+    {
+      "epoch": 15.983532934131736,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1468,
+      "step": 10677
+    },
+    {
+      "epoch": 15.98502994011976,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1336,
+      "step": 10678
+    },
+    {
+      "epoch": 15.986526946107784,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 10679
+    },
+    {
+      "epoch": 15.988023952095809,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 10680
+    },
+    {
+      "epoch": 15.989520958083832,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1386,
+      "step": 10681
+    },
+    {
+      "epoch": 15.991017964071856,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 10682
+    },
+    {
+      "epoch": 15.99251497005988,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1413,
+      "step": 10683
+    },
+    {
+      "epoch": 15.994011976047904,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 10684
+    },
+    {
+      "epoch": 15.995508982035929,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1387,
+      "step": 10685
+    },
+    {
+      "epoch": 15.997005988023952,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 10686
+    },
+    {
+      "epoch": 15.998502994011975,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 10687
+    },
+    {
+      "epoch": 16.0,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1321,
+      "step": 10688
+    },
+    {
+      "epoch": 16.001497005988025,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 10689
+    },
+    {
+      "epoch": 16.00299401197605,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1479,
+      "step": 10690
+    },
+    {
+      "epoch": 16.00449101796407,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 10691
+    },
+    {
+      "epoch": 16.005988023952096,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 10692
+    },
+    {
+      "epoch": 16.00748502994012,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1416,
+      "step": 10693
+    },
+    {
+      "epoch": 16.008982035928145,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 10694
+    },
+    {
+      "epoch": 16.010479041916167,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1349,
+      "step": 10695
+    },
+    {
+      "epoch": 16.01197604790419,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 10696
+    },
+    {
+      "epoch": 16.013473053892216,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 10697
+    },
+    {
+      "epoch": 16.01497005988024,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 10698
+    },
+    {
+      "epoch": 16.016467065868262,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 10699
+    },
+    {
+      "epoch": 16.017964071856287,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 10700
+    },
+    {
+      "epoch": 16.019461077844312,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1326,
+      "step": 10701
+    },
+    {
+      "epoch": 16.020958083832337,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 10702
+    },
+    {
+      "epoch": 16.022455089820358,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 10703
+    },
+    {
+      "epoch": 16.023952095808383,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 10704
+    },
+    {
+      "epoch": 16.025449101796408,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 10705
+    },
+    {
+      "epoch": 16.026946107784433,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1381,
+      "step": 10706
+    },
+    {
+      "epoch": 16.028443113772454,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 10707
+    },
+    {
+      "epoch": 16.02994011976048,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1443,
+      "step": 10708
+    },
+    {
+      "epoch": 16.031437125748504,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1365,
+      "step": 10709
+    },
+    {
+      "epoch": 16.03293413173653,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 10710
+    },
+    {
+      "epoch": 16.03443113772455,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 10711
+    },
+    {
+      "epoch": 16.035928143712574,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1368,
+      "step": 10712
+    },
+    {
+      "epoch": 16.0374251497006,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1381,
+      "step": 10713
+    },
+    {
+      "epoch": 16.038922155688624,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1377,
+      "step": 10714
+    },
+    {
+      "epoch": 16.040419161676645,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 10715
+    },
+    {
+      "epoch": 16.04191616766467,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 10716
+    },
+    {
+      "epoch": 16.043413173652695,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1367,
+      "step": 10717
+    },
+    {
+      "epoch": 16.04491017964072,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1365,
+      "step": 10718
+    },
+    {
+      "epoch": 16.04640718562874,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 10719
+    },
+    {
+      "epoch": 16.047904191616766,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1429,
+      "step": 10720
+    },
+    {
+      "epoch": 16.04940119760479,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1515,
+      "step": 10721
+    },
+    {
+      "epoch": 16.050898203592816,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1311,
+      "step": 10722
+    },
+    {
+      "epoch": 16.052395209580837,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 10723
+    },
+    {
+      "epoch": 16.05389221556886,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.144,
+      "step": 10724
+    },
+    {
+      "epoch": 16.055389221556887,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1446,
+      "step": 10725
+    },
+    {
+      "epoch": 16.05688622754491,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.14,
+      "step": 10726
+    },
+    {
+      "epoch": 16.058383233532933,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1355,
+      "step": 10727
+    },
+    {
+      "epoch": 16.059880239520957,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1506,
+      "step": 10728
+    },
+    {
+      "epoch": 16.061377245508982,
+      "grad_norm": 0.0625,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 10729
+    },
+    {
+      "epoch": 16.062874251497007,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 10730
+    },
+    {
+      "epoch": 16.06437125748503,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 10731
+    },
+    {
+      "epoch": 16.065868263473053,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 10732
+    },
+    {
+      "epoch": 16.067365269461078,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 10733
+    },
+    {
+      "epoch": 16.068862275449103,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 10734
+    },
+    {
+      "epoch": 16.070359281437124,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 10735
+    },
+    {
+      "epoch": 16.07185628742515,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1468,
+      "step": 10736
+    },
+    {
+      "epoch": 16.073353293413174,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 10737
+    },
+    {
+      "epoch": 16.0748502994012,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1392,
+      "step": 10738
+    },
+    {
+      "epoch": 16.07634730538922,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 10739
+    },
+    {
+      "epoch": 16.077844311377245,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1392,
+      "step": 10740
+    },
+    {
+      "epoch": 16.07934131736527,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1496,
+      "step": 10741
+    },
+    {
+      "epoch": 16.080838323353294,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 10742
+    },
+    {
+      "epoch": 16.082335329341316,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 10743
+    },
+    {
+      "epoch": 16.08383233532934,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 10744
+    },
+    {
+      "epoch": 16.085329341317365,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1489,
+      "step": 10745
+    },
+    {
+      "epoch": 16.08682634730539,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 10746
+    },
+    {
+      "epoch": 16.088323353293415,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 10747
+    },
+    {
+      "epoch": 16.089820359281436,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 10748
+    },
+    {
+      "epoch": 16.09131736526946,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1386,
+      "step": 10749
+    },
+    {
+      "epoch": 16.092814371257486,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.144,
+      "step": 10750
+    },
+    {
+      "epoch": 16.09431137724551,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 10751
+    },
+    {
+      "epoch": 16.095808383233532,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 10752
+    },
+    {
+      "epoch": 16.097305389221557,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1472,
+      "step": 10753
+    },
+    {
+      "epoch": 16.09880239520958,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 10754
+    },
+    {
+      "epoch": 16.100299401197606,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1392,
+      "step": 10755
+    },
+    {
+      "epoch": 16.101796407185628,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 10756
+    },
+    {
+      "epoch": 16.103293413173652,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 10757
+    },
+    {
+      "epoch": 16.104790419161677,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1428,
+      "step": 10758
+    },
+    {
+      "epoch": 16.106287425149702,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1381,
+      "step": 10759
+    },
+    {
+      "epoch": 16.107784431137723,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 10760
+    },
+    {
+      "epoch": 16.10928143712575,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1377,
+      "step": 10761
+    },
+    {
+      "epoch": 16.110778443113773,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1336,
+      "step": 10762
+    },
+    {
+      "epoch": 16.112275449101798,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1359,
+      "step": 10763
+    },
+    {
+      "epoch": 16.11377245508982,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 10764
+    },
+    {
+      "epoch": 16.115269461077844,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1373,
+      "step": 10765
+    },
+    {
+      "epoch": 16.11676646706587,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1375,
+      "step": 10766
+    },
+    {
+      "epoch": 16.118263473053894,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1413,
+      "step": 10767
+    },
+    {
+      "epoch": 16.119760479041915,
+      "grad_norm": 0.142578125,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 10768
+    },
+    {
+      "epoch": 16.12125748502994,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 10769
+    },
+    {
+      "epoch": 16.122754491017965,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 10770
+    },
+    {
+      "epoch": 16.12425149700599,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0008,
+      "loss": 1.144,
+      "step": 10771
+    },
+    {
+      "epoch": 16.12574850299401,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 10772
+    },
+    {
+      "epoch": 16.127245508982035,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.147,
+      "step": 10773
+    },
+    {
+      "epoch": 16.12874251497006,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 10774
+    },
+    {
+      "epoch": 16.130239520958085,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1498,
+      "step": 10775
+    },
+    {
+      "epoch": 16.131736526946106,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 10776
+    },
+    {
+      "epoch": 16.13323353293413,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 10777
+    },
+    {
+      "epoch": 16.134730538922156,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1375,
+      "step": 10778
+    },
+    {
+      "epoch": 16.13622754491018,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 10779
+    },
+    {
+      "epoch": 16.137724550898202,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1387,
+      "step": 10780
+    },
+    {
+      "epoch": 16.139221556886227,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 10781
+    },
+    {
+      "epoch": 16.14071856287425,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 10782
+    },
+    {
+      "epoch": 16.142215568862277,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1365,
+      "step": 10783
+    },
+    {
+      "epoch": 16.143712574850298,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 10784
+    },
+    {
+      "epoch": 16.145209580838323,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 10785
+    },
+    {
+      "epoch": 16.146706586826348,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 10786
+    },
+    {
+      "epoch": 16.148203592814372,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1434,
+      "step": 10787
+    },
+    {
+      "epoch": 16.149700598802394,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 10788
+    },
+    {
+      "epoch": 16.15119760479042,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 10789
+    },
+    {
+      "epoch": 16.152694610778443,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1359,
+      "step": 10790
+    },
+    {
+      "epoch": 16.154191616766468,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 10791
+    },
+    {
+      "epoch": 16.15568862275449,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.149,
+      "step": 10792
+    },
+    {
+      "epoch": 16.157185628742514,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 10793
+    },
+    {
+      "epoch": 16.15868263473054,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1379,
+      "step": 10794
+    },
+    {
+      "epoch": 16.160179640718564,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1367,
+      "step": 10795
+    },
+    {
+      "epoch": 16.161676646706585,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.137,
+      "step": 10796
+    },
+    {
+      "epoch": 16.16317365269461,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1392,
+      "step": 10797
+    },
+    {
+      "epoch": 16.164670658682635,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 10798
+    },
+    {
+      "epoch": 16.16616766467066,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 10799
+    },
+    {
+      "epoch": 16.16766467065868,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.144,
+      "step": 10800
+    },
+    {
+      "epoch": 16.169161676646706,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 10801
+    },
+    {
+      "epoch": 16.17065868263473,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1462,
+      "step": 10802
+    },
+    {
+      "epoch": 16.172155688622755,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 10803
+    },
+    {
+      "epoch": 16.17365269461078,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 10804
+    },
+    {
+      "epoch": 16.1751497005988,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1355,
+      "step": 10805
+    },
+    {
+      "epoch": 16.176646706586826,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 10806
+    },
+    {
+      "epoch": 16.17814371257485,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 10807
+    },
+    {
+      "epoch": 16.179640718562876,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1462,
+      "step": 10808
+    },
+    {
+      "epoch": 16.181137724550897,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1365,
+      "step": 10809
+    },
+    {
+      "epoch": 16.182634730538922,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 10810
+    },
+    {
+      "epoch": 16.184131736526947,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 10811
+    },
+    {
+      "epoch": 16.18562874251497,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1328,
+      "step": 10812
+    },
+    {
+      "epoch": 16.187125748502993,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 10813
+    },
+    {
+      "epoch": 16.188622754491018,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1459,
+      "step": 10814
+    },
+    {
+      "epoch": 16.190119760479043,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1377,
+      "step": 10815
+    },
+    {
+      "epoch": 16.191616766467067,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1434,
+      "step": 10816
+    },
+    {
+      "epoch": 16.19311377245509,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 10817
+    },
+    {
+      "epoch": 16.194610778443113,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 10818
+    },
+    {
+      "epoch": 16.19610778443114,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 10819
+    },
+    {
+      "epoch": 16.197604790419163,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1334,
+      "step": 10820
+    },
+    {
+      "epoch": 16.199101796407184,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1375,
+      "step": 10821
+    },
+    {
+      "epoch": 16.20059880239521,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1469,
+      "step": 10822
+    },
+    {
+      "epoch": 16.202095808383234,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 10823
+    },
+    {
+      "epoch": 16.20359281437126,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1413,
+      "step": 10824
+    },
+    {
+      "epoch": 16.20508982035928,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1357,
+      "step": 10825
+    },
+    {
+      "epoch": 16.206586826347305,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 10826
+    },
+    {
+      "epoch": 16.20808383233533,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1381,
+      "step": 10827
+    },
+    {
+      "epoch": 16.209580838323355,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 10828
+    },
+    {
+      "epoch": 16.211077844311376,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 10829
+    },
+    {
+      "epoch": 16.2125748502994,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 10830
+    },
+    {
+      "epoch": 16.214071856287426,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1378,
+      "step": 10831
+    },
+    {
+      "epoch": 16.21556886227545,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1392,
+      "step": 10832
+    },
+    {
+      "epoch": 16.21706586826347,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 10833
+    },
+    {
+      "epoch": 16.218562874251496,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1386,
+      "step": 10834
+    },
+    {
+      "epoch": 16.22005988023952,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 10835
+    },
+    {
+      "epoch": 16.221556886227546,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1364,
+      "step": 10836
+    },
+    {
+      "epoch": 16.223053892215567,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1362,
+      "step": 10837
+    },
+    {
+      "epoch": 16.224550898203592,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 10838
+    },
+    {
+      "epoch": 16.226047904191617,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 10839
+    },
+    {
+      "epoch": 16.227544910179642,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 10840
+    },
+    {
+      "epoch": 16.229041916167663,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 10841
+    },
+    {
+      "epoch": 16.230538922155688,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 10842
+    },
+    {
+      "epoch": 16.232035928143713,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 10843
+    },
+    {
+      "epoch": 16.233532934131738,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1428,
+      "step": 10844
+    },
+    {
+      "epoch": 16.23502994011976,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1428,
+      "step": 10845
+    },
+    {
+      "epoch": 16.236526946107784,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 10846
+    },
+    {
+      "epoch": 16.23802395209581,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 10847
+    },
+    {
+      "epoch": 16.239520958083833,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 10848
+    },
+    {
+      "epoch": 16.241017964071855,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1383,
+      "step": 10849
+    },
+    {
+      "epoch": 16.24251497005988,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1428,
+      "step": 10850
+    },
+    {
+      "epoch": 16.244011976047904,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 10851
+    },
+    {
+      "epoch": 16.24550898203593,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1366,
+      "step": 10852
+    },
+    {
+      "epoch": 16.24700598802395,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1395,
+      "step": 10853
+    },
+    {
+      "epoch": 16.248502994011975,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 10854
+    },
+    {
+      "epoch": 16.25,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 10855
+    },
+    {
+      "epoch": 16.251497005988025,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 10856
+    },
+    {
+      "epoch": 16.25299401197605,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 10857
+    },
+    {
+      "epoch": 16.25449101796407,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 10858
+    },
+    {
+      "epoch": 16.255988023952096,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 10859
+    },
+    {
+      "epoch": 16.25748502994012,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 10860
+    },
+    {
+      "epoch": 16.258982035928145,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 10861
+    },
+    {
+      "epoch": 16.260479041916167,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 10862
+    },
+    {
+      "epoch": 16.26197604790419,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 10863
+    },
+    {
+      "epoch": 16.263473053892216,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1428,
+      "step": 10864
+    },
+    {
+      "epoch": 16.26497005988024,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 10865
+    },
+    {
+      "epoch": 16.266467065868262,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 10866
+    },
+    {
+      "epoch": 16.267964071856287,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 10867
+    },
+    {
+      "epoch": 16.269461077844312,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1384,
+      "step": 10868
+    },
+    {
+      "epoch": 16.270958083832337,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1386,
+      "step": 10869
+    },
+    {
+      "epoch": 16.272455089820358,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 10870
+    },
+    {
+      "epoch": 16.273952095808383,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1323,
+      "step": 10871
+    },
+    {
+      "epoch": 16.275449101796408,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 10872
+    },
+    {
+      "epoch": 16.276946107784433,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1387,
+      "step": 10873
+    },
+    {
+      "epoch": 16.278443113772454,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 10874
+    },
+    {
+      "epoch": 16.27994011976048,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 10875
+    },
+    {
+      "epoch": 16.281437125748504,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1375,
+      "step": 10876
+    },
+    {
+      "epoch": 16.28293413173653,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 10877
+    },
+    {
+      "epoch": 16.28443113772455,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1443,
+      "step": 10878
+    },
+    {
+      "epoch": 16.285928143712574,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1379,
+      "step": 10879
+    },
+    {
+      "epoch": 16.2874251497006,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1429,
+      "step": 10880
+    },
+    {
+      "epoch": 16.288922155688624,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1369,
+      "step": 10881
+    },
+    {
+      "epoch": 16.290419161676645,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 10882
+    },
+    {
+      "epoch": 16.29191616766467,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 10883
+    },
+    {
+      "epoch": 16.293413173652695,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1387,
+      "step": 10884
+    },
+    {
+      "epoch": 16.29491017964072,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1463,
+      "step": 10885
+    },
+    {
+      "epoch": 16.29640718562874,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 10886
+    },
+    {
+      "epoch": 16.297904191616766,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 10887
+    },
+    {
+      "epoch": 16.29940119760479,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 10888
+    },
+    {
+      "epoch": 16.300898203592816,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 10889
+    },
+    {
+      "epoch": 16.302395209580837,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 10890
+    },
+    {
+      "epoch": 16.30389221556886,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.136,
+      "step": 10891
+    },
+    {
+      "epoch": 16.305389221556887,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1428,
+      "step": 10892
+    },
+    {
+      "epoch": 16.30688622754491,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 10893
+    },
+    {
+      "epoch": 16.308383233532933,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1366,
+      "step": 10894
+    },
+    {
+      "epoch": 16.309880239520957,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 10895
+    },
+    {
+      "epoch": 16.311377245508982,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1383,
+      "step": 10896
+    },
+    {
+      "epoch": 16.312874251497007,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 10897
+    },
+    {
+      "epoch": 16.31437125748503,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1361,
+      "step": 10898
+    },
+    {
+      "epoch": 16.315868263473053,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1465,
+      "step": 10899
+    },
+    {
+      "epoch": 16.317365269461078,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1322,
+      "step": 10900
+    },
+    {
+      "epoch": 16.318862275449103,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 10901
+    },
+    {
+      "epoch": 16.320359281437124,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 10902
+    },
+    {
+      "epoch": 16.32185628742515,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 10903
+    },
+    {
+      "epoch": 16.323353293413174,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1386,
+      "step": 10904
+    },
+    {
+      "epoch": 16.3248502994012,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 10905
+    },
+    {
+      "epoch": 16.32634730538922,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1446,
+      "step": 10906
+    },
+    {
+      "epoch": 16.327844311377245,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1373,
+      "step": 10907
+    },
+    {
+      "epoch": 16.32934131736527,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 10908
+    },
+    {
+      "epoch": 16.330838323353294,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 10909
+    },
+    {
+      "epoch": 16.33233532934132,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 10910
+    },
+    {
+      "epoch": 16.33383233532934,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 10911
+    },
+    {
+      "epoch": 16.335329341317365,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 10912
+    },
+    {
+      "epoch": 16.33682634730539,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1366,
+      "step": 10913
+    },
+    {
+      "epoch": 16.338323353293415,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1353,
+      "step": 10914
+    },
+    {
+      "epoch": 16.339820359281436,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 10915
+    },
+    {
+      "epoch": 16.34131736526946,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1375,
+      "step": 10916
+    },
+    {
+      "epoch": 16.342814371257486,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 10917
+    },
+    {
+      "epoch": 16.34431137724551,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 10918
+    },
+    {
+      "epoch": 16.345808383233532,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 10919
+    },
+    {
+      "epoch": 16.347305389221557,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1371,
+      "step": 10920
+    },
+    {
+      "epoch": 16.34880239520958,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 10921
+    },
+    {
+      "epoch": 16.350299401197606,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1459,
+      "step": 10922
+    },
+    {
+      "epoch": 16.351796407185628,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 10923
+    },
+    {
+      "epoch": 16.353293413173652,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1375,
+      "step": 10924
+    },
+    {
+      "epoch": 16.354790419161677,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 10925
+    },
+    {
+      "epoch": 16.356287425149702,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 10926
+    },
+    {
+      "epoch": 16.357784431137723,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1357,
+      "step": 10927
+    },
+    {
+      "epoch": 16.35928143712575,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1507,
+      "step": 10928
+    },
+    {
+      "epoch": 16.360778443113773,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1341,
+      "step": 10929
+    },
+    {
+      "epoch": 16.362275449101798,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 10930
+    },
+    {
+      "epoch": 16.36377245508982,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1373,
+      "step": 10931
+    },
+    {
+      "epoch": 16.365269461077844,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 10932
+    },
+    {
+      "epoch": 16.36676646706587,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 10933
+    },
+    {
+      "epoch": 16.368263473053894,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1324,
+      "step": 10934
+    },
+    {
+      "epoch": 16.369760479041915,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 10935
+    },
+    {
+      "epoch": 16.37125748502994,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 10936
+    },
+    {
+      "epoch": 16.372754491017965,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 10937
+    },
+    {
+      "epoch": 16.37425149700599,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1475,
+      "step": 10938
+    },
+    {
+      "epoch": 16.37574850299401,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1335,
+      "step": 10939
+    },
+    {
+      "epoch": 16.377245508982035,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1383,
+      "step": 10940
+    },
+    {
+      "epoch": 16.37874251497006,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 10941
+    },
+    {
+      "epoch": 16.380239520958085,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 10942
+    },
+    {
+      "epoch": 16.381736526946106,
+      "grad_norm": 0.062255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 10943
+    },
+    {
+      "epoch": 16.38323353293413,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 10944
+    },
+    {
+      "epoch": 16.384730538922156,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.137,
+      "step": 10945
+    },
+    {
+      "epoch": 16.38622754491018,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 10946
+    },
+    {
+      "epoch": 16.387724550898202,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1342,
+      "step": 10947
+    },
+    {
+      "epoch": 16.389221556886227,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 10948
+    },
+    {
+      "epoch": 16.39071856287425,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 10949
+    },
+    {
+      "epoch": 16.392215568862277,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 10950
+    },
+    {
+      "epoch": 16.393712574850298,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.137,
+      "step": 10951
+    },
+    {
+      "epoch": 16.395209580838323,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 10952
+    },
+    {
+      "epoch": 16.396706586826348,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.135,
+      "step": 10953
+    },
+    {
+      "epoch": 16.398203592814372,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1416,
+      "step": 10954
+    },
+    {
+      "epoch": 16.399700598802394,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1368,
+      "step": 10955
+    },
+    {
+      "epoch": 16.40119760479042,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 10956
+    },
+    {
+      "epoch": 16.402694610778443,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1473,
+      "step": 10957
+    },
+    {
+      "epoch": 16.404191616766468,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 10958
+    },
+    {
+      "epoch": 16.40568862275449,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 10959
+    },
+    {
+      "epoch": 16.407185628742514,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1343,
+      "step": 10960
+    },
+    {
+      "epoch": 16.40868263473054,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 10961
+    },
+    {
+      "epoch": 16.410179640718564,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1428,
+      "step": 10962
+    },
+    {
+      "epoch": 16.411676646706585,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 10963
+    },
+    {
+      "epoch": 16.41317365269461,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1457,
+      "step": 10964
+    },
+    {
+      "epoch": 16.414670658682635,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.136,
+      "step": 10965
+    },
+    {
+      "epoch": 16.41616766467066,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 10966
+    },
+    {
+      "epoch": 16.41766467065868,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 10967
+    },
+    {
+      "epoch": 16.419161676646706,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 10968
+    },
+    {
+      "epoch": 16.42065868263473,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1359,
+      "step": 10969
+    },
+    {
+      "epoch": 16.422155688622755,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.148,
+      "step": 10970
+    },
+    {
+      "epoch": 16.42365269461078,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1379,
+      "step": 10971
+    },
+    {
+      "epoch": 16.4251497005988,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.135,
+      "step": 10972
+    },
+    {
+      "epoch": 16.426646706586826,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1339,
+      "step": 10973
+    },
+    {
+      "epoch": 16.42814371257485,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1383,
+      "step": 10974
+    },
+    {
+      "epoch": 16.429640718562876,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1362,
+      "step": 10975
+    },
+    {
+      "epoch": 16.431137724550897,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 10976
+    },
+    {
+      "epoch": 16.432634730538922,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 10977
+    },
+    {
+      "epoch": 16.434131736526947,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1464,
+      "step": 10978
+    },
+    {
+      "epoch": 16.43562874251497,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 10979
+    },
+    {
+      "epoch": 16.437125748502993,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1469,
+      "step": 10980
+    },
+    {
+      "epoch": 16.438622754491018,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 10981
+    },
+    {
+      "epoch": 16.440119760479043,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 10982
+    },
+    {
+      "epoch": 16.441616766467067,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 10983
+    },
+    {
+      "epoch": 16.44311377245509,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1384,
+      "step": 10984
+    },
+    {
+      "epoch": 16.444610778443113,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.14,
+      "step": 10985
+    },
+    {
+      "epoch": 16.44610778443114,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 10986
+    },
+    {
+      "epoch": 16.447604790419163,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1361,
+      "step": 10987
+    },
+    {
+      "epoch": 16.449101796407184,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1381,
+      "step": 10988
+    },
+    {
+      "epoch": 16.45059880239521,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1428,
+      "step": 10989
+    },
+    {
+      "epoch": 16.452095808383234,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1384,
+      "step": 10990
+    },
+    {
+      "epoch": 16.45359281437126,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 10991
+    },
+    {
+      "epoch": 16.45508982035928,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 10992
+    },
+    {
+      "epoch": 16.456586826347305,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1333,
+      "step": 10993
+    },
+    {
+      "epoch": 16.45808383233533,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1336,
+      "step": 10994
+    },
+    {
+      "epoch": 16.459580838323355,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1362,
+      "step": 10995
+    },
+    {
+      "epoch": 16.461077844311376,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 10996
+    },
+    {
+      "epoch": 16.4625748502994,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1343,
+      "step": 10997
+    },
+    {
+      "epoch": 16.464071856287426,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 10998
+    },
+    {
+      "epoch": 16.46556886227545,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 10999
+    },
+    {
+      "epoch": 16.46706586826347,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1373,
+      "step": 11000
+    },
+    {
+      "epoch": 16.468562874251496,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1537,
+      "step": 11001
+    },
+    {
+      "epoch": 16.47005988023952,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0008,
+      "loss": 1.2339,
+      "step": 11002
+    },
+    {
+      "epoch": 16.471556886227546,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.0008,
+      "loss": 1.2478,
+      "step": 11003
+    },
+    {
+      "epoch": 16.473053892215567,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0008,
+      "loss": 1.2196,
+      "step": 11004
+    },
+    {
+      "epoch": 16.474550898203592,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0008,
+      "loss": 1.2444,
+      "step": 11005
+    },
+    {
+      "epoch": 16.476047904191617,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.0008,
+      "loss": 1.2377,
+      "step": 11006
+    },
+    {
+      "epoch": 16.477544910179642,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.0008,
+      "loss": 1.2631,
+      "step": 11007
+    },
+    {
+      "epoch": 16.479041916167663,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0008,
+      "loss": 1.2588,
+      "step": 11008
+    },
+    {
+      "epoch": 16.480538922155688,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.0008,
+      "loss": 1.283,
+      "step": 11009
+    },
+    {
+      "epoch": 16.482035928143713,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.0008,
+      "loss": 1.2862,
+      "step": 11010
+    },
+    {
+      "epoch": 16.483532934131738,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0008,
+      "loss": 1.2847,
+      "step": 11011
+    },
+    {
+      "epoch": 16.48502994011976,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0008,
+      "loss": 1.2827,
+      "step": 11012
+    },
+    {
+      "epoch": 16.486526946107784,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.0008,
+      "loss": 1.2914,
+      "step": 11013
+    },
+    {
+      "epoch": 16.48802395209581,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.0008,
+      "loss": 1.2901,
+      "step": 11014
+    },
+    {
+      "epoch": 16.489520958083833,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0008,
+      "loss": 1.2746,
+      "step": 11015
+    },
+    {
+      "epoch": 16.491017964071855,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0008,
+      "loss": 1.2701,
+      "step": 11016
+    },
+    {
+      "epoch": 16.49251497005988,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.0008,
+      "loss": 1.2886,
+      "step": 11017
+    },
+    {
+      "epoch": 16.494011976047904,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.0008,
+      "loss": 1.3552,
+      "step": 11018
+    },
+    {
+      "epoch": 16.49550898203593,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0008,
+      "loss": 1.2829,
+      "step": 11019
+    },
+    {
+      "epoch": 16.49700598802395,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0008,
+      "loss": 1.2804,
+      "step": 11020
+    },
+    {
+      "epoch": 16.498502994011975,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0008,
+      "loss": 1.276,
+      "step": 11021
+    },
+    {
+      "epoch": 16.5,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0008,
+      "loss": 1.2676,
+      "step": 11022
+    },
+    {
+      "epoch": 16.501497005988025,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0008,
+      "loss": 1.2626,
+      "step": 11023
+    },
+    {
+      "epoch": 16.50299401197605,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0008,
+      "loss": 1.2453,
+      "step": 11024
+    },
+    {
+      "epoch": 16.50449101796407,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0008,
+      "loss": 1.248,
+      "step": 11025
+    },
+    {
+      "epoch": 16.505988023952096,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0008,
+      "loss": 1.2363,
+      "step": 11026
+    },
+    {
+      "epoch": 16.50748502994012,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0008,
+      "loss": 1.236,
+      "step": 11027
+    },
+    {
+      "epoch": 16.508982035928145,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0008,
+      "loss": 1.228,
+      "step": 11028
+    },
+    {
+      "epoch": 16.510479041916167,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0008,
+      "loss": 1.2281,
+      "step": 11029
+    },
+    {
+      "epoch": 16.51197604790419,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0008,
+      "loss": 1.2243,
+      "step": 11030
+    },
+    {
+      "epoch": 16.513473053892216,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0008,
+      "loss": 1.2231,
+      "step": 11031
+    },
+    {
+      "epoch": 16.51497005988024,
+      "grad_norm": 0.203125,
+      "learning_rate": 0.0008,
+      "loss": 1.2208,
+      "step": 11032
+    },
+    {
+      "epoch": 16.516467065868262,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0008,
+      "loss": 1.2153,
+      "step": 11033
+    },
+    {
+      "epoch": 16.517964071856287,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.2161,
+      "step": 11034
+    },
+    {
+      "epoch": 16.519461077844312,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0008,
+      "loss": 1.2174,
+      "step": 11035
+    },
+    {
+      "epoch": 16.520958083832337,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.0008,
+      "loss": 1.2215,
+      "step": 11036
+    },
+    {
+      "epoch": 16.522455089820358,
+      "grad_norm": 0.921875,
+      "learning_rate": 0.0008,
+      "loss": 1.2288,
+      "step": 11037
+    },
+    {
+      "epoch": 16.523952095808383,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0008,
+      "loss": 1.2081,
+      "step": 11038
+    },
+    {
+      "epoch": 16.525449101796408,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0008,
+      "loss": 1.2025,
+      "step": 11039
+    },
+    {
+      "epoch": 16.526946107784433,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0008,
+      "loss": 1.2166,
+      "step": 11040
+    },
+    {
+      "epoch": 16.528443113772454,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0008,
+      "loss": 1.2124,
+      "step": 11041
+    },
+    {
+      "epoch": 16.52994011976048,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0008,
+      "loss": 1.2155,
+      "step": 11042
+    },
+    {
+      "epoch": 16.531437125748504,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0008,
+      "loss": 1.2082,
+      "step": 11043
+    },
+    {
+      "epoch": 16.53293413173653,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1996,
+      "step": 11044
+    },
+    {
+      "epoch": 16.53443113772455,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0008,
+      "loss": 1.191,
+      "step": 11045
+    },
+    {
+      "epoch": 16.535928143712574,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1928,
+      "step": 11046
+    },
+    {
+      "epoch": 16.5374251497006,
+      "grad_norm": 0.197265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1906,
+      "step": 11047
+    },
+    {
+      "epoch": 16.538922155688624,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1906,
+      "step": 11048
+    },
+    {
+      "epoch": 16.540419161676645,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1896,
+      "step": 11049
+    },
+    {
+      "epoch": 16.54191616766467,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1849,
+      "step": 11050
+    },
+    {
+      "epoch": 16.543413173652695,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1869,
+      "step": 11051
+    },
+    {
+      "epoch": 16.54491017964072,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1876,
+      "step": 11052
+    },
+    {
+      "epoch": 16.54640718562874,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1822,
+      "step": 11053
+    },
+    {
+      "epoch": 16.547904191616766,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1865,
+      "step": 11054
+    },
+    {
+      "epoch": 16.54940119760479,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1717,
+      "step": 11055
+    },
+    {
+      "epoch": 16.550898203592816,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1808,
+      "step": 11056
+    },
+    {
+      "epoch": 16.552395209580837,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.175,
+      "step": 11057
+    },
+    {
+      "epoch": 16.55389221556886,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1804,
+      "step": 11058
+    },
+    {
+      "epoch": 16.555389221556887,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1745,
+      "step": 11059
+    },
+    {
+      "epoch": 16.55688622754491,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1694,
+      "step": 11060
+    },
+    {
+      "epoch": 16.558383233532933,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1768,
+      "step": 11061
+    },
+    {
+      "epoch": 16.559880239520957,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.171,
+      "step": 11062
+    },
+    {
+      "epoch": 16.561377245508982,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1725,
+      "step": 11063
+    },
+    {
+      "epoch": 16.562874251497007,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1779,
+      "step": 11064
+    },
+    {
+      "epoch": 16.56437125748503,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1795,
+      "step": 11065
+    },
+    {
+      "epoch": 16.565868263473053,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1998,
+      "step": 11066
+    },
+    {
+      "epoch": 16.567365269461078,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1738,
+      "step": 11067
+    },
+    {
+      "epoch": 16.568862275449103,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0008,
+      "loss": 1.1762,
+      "step": 11068
+    },
+    {
+      "epoch": 16.570359281437124,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1845,
+      "step": 11069
+    },
+    {
+      "epoch": 16.57185628742515,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1764,
+      "step": 11070
+    },
+    {
+      "epoch": 16.573353293413174,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1823,
+      "step": 11071
+    },
+    {
+      "epoch": 16.5748502994012,
+      "grad_norm": 0.208984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1664,
+      "step": 11072
+    },
+    {
+      "epoch": 16.57634730538922,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1715,
+      "step": 11073
+    },
+    {
+      "epoch": 16.577844311377245,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1694,
+      "step": 11074
+    },
+    {
+      "epoch": 16.57934131736527,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1633,
+      "step": 11075
+    },
+    {
+      "epoch": 16.580838323353294,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1594,
+      "step": 11076
+    },
+    {
+      "epoch": 16.58233532934132,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1658,
+      "step": 11077
+    },
+    {
+      "epoch": 16.58383233532934,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0008,
+      "loss": 1.1665,
+      "step": 11078
+    },
+    {
+      "epoch": 16.585329341317365,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1682,
+      "step": 11079
+    },
+    {
+      "epoch": 16.58682634730539,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1667,
+      "step": 11080
+    },
+    {
+      "epoch": 16.58832335329341,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1605,
+      "step": 11081
+    },
+    {
+      "epoch": 16.589820359281436,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1633,
+      "step": 11082
+    },
+    {
+      "epoch": 16.59131736526946,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1626,
+      "step": 11083
+    },
+    {
+      "epoch": 16.592814371257486,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1626,
+      "step": 11084
+    },
+    {
+      "epoch": 16.59431137724551,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1606,
+      "step": 11085
+    },
+    {
+      "epoch": 16.595808383233532,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1572,
+      "step": 11086
+    },
+    {
+      "epoch": 16.597305389221557,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1589,
+      "step": 11087
+    },
+    {
+      "epoch": 16.59880239520958,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1535,
+      "step": 11088
+    },
+    {
+      "epoch": 16.600299401197606,
+      "grad_norm": 0.140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1563,
+      "step": 11089
+    },
+    {
+      "epoch": 16.601796407185628,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1593,
+      "step": 11090
+    },
+    {
+      "epoch": 16.603293413173652,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1585,
+      "step": 11091
+    },
+    {
+      "epoch": 16.604790419161677,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0008,
+      "loss": 1.1593,
+      "step": 11092
+    },
+    {
+      "epoch": 16.606287425149702,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.163,
+      "step": 11093
+    },
+    {
+      "epoch": 16.607784431137723,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1582,
+      "step": 11094
+    },
+    {
+      "epoch": 16.60928143712575,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1586,
+      "step": 11095
+    },
+    {
+      "epoch": 16.610778443113773,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1556,
+      "step": 11096
+    },
+    {
+      "epoch": 16.612275449101798,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.155,
+      "step": 11097
+    },
+    {
+      "epoch": 16.61377245508982,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1638,
+      "step": 11098
+    },
+    {
+      "epoch": 16.615269461077844,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1563,
+      "step": 11099
+    },
+    {
+      "epoch": 16.61676646706587,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1597,
+      "step": 11100
+    },
+    {
+      "epoch": 16.618263473053894,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1577,
+      "step": 11101
+    },
+    {
+      "epoch": 16.619760479041915,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1585,
+      "step": 11102
+    },
+    {
+      "epoch": 16.62125748502994,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1597,
+      "step": 11103
+    },
+    {
+      "epoch": 16.622754491017965,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1537,
+      "step": 11104
+    },
+    {
+      "epoch": 16.62425149700599,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1552,
+      "step": 11105
+    },
+    {
+      "epoch": 16.62574850299401,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1551,
+      "step": 11106
+    },
+    {
+      "epoch": 16.627245508982035,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1477,
+      "step": 11107
+    },
+    {
+      "epoch": 16.62874251497006,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1582,
+      "step": 11108
+    },
+    {
+      "epoch": 16.630239520958085,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.157,
+      "step": 11109
+    },
+    {
+      "epoch": 16.631736526946106,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1535,
+      "step": 11110
+    },
+    {
+      "epoch": 16.63323353293413,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1575,
+      "step": 11111
+    },
+    {
+      "epoch": 16.634730538922156,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1559,
+      "step": 11112
+    },
+    {
+      "epoch": 16.63622754491018,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1557,
+      "step": 11113
+    },
+    {
+      "epoch": 16.637724550898202,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1539,
+      "step": 11114
+    },
+    {
+      "epoch": 16.639221556886227,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1605,
+      "step": 11115
+    },
+    {
+      "epoch": 16.64071856287425,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1561,
+      "step": 11116
+    },
+    {
+      "epoch": 16.642215568862277,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1545,
+      "step": 11117
+    },
+    {
+      "epoch": 16.643712574850298,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1518,
+      "step": 11118
+    },
+    {
+      "epoch": 16.645209580838323,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1506,
+      "step": 11119
+    },
+    {
+      "epoch": 16.646706586826348,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1522,
+      "step": 11120
+    },
+    {
+      "epoch": 16.648203592814372,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1531,
+      "step": 11121
+    },
+    {
+      "epoch": 16.649700598802394,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1532,
+      "step": 11122
+    },
+    {
+      "epoch": 16.65119760479042,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1592,
+      "step": 11123
+    },
+    {
+      "epoch": 16.652694610778443,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1598,
+      "step": 11124
+    },
+    {
+      "epoch": 16.654191616766468,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1529,
+      "step": 11125
+    },
+    {
+      "epoch": 16.65568862275449,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1574,
+      "step": 11126
+    },
+    {
+      "epoch": 16.657185628742514,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1462,
+      "step": 11127
+    },
+    {
+      "epoch": 16.65868263473054,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1524,
+      "step": 11128
+    },
+    {
+      "epoch": 16.660179640718564,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1567,
+      "step": 11129
+    },
+    {
+      "epoch": 16.66167664670659,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1529,
+      "step": 11130
+    },
+    {
+      "epoch": 16.66317365269461,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1506,
+      "step": 11131
+    },
+    {
+      "epoch": 16.664670658682635,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 11132
+    },
+    {
+      "epoch": 16.66616766467066,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1503,
+      "step": 11133
+    },
+    {
+      "epoch": 16.66766467065868,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 11134
+    },
+    {
+      "epoch": 16.669161676646706,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1542,
+      "step": 11135
+    },
+    {
+      "epoch": 16.67065868263473,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1519,
+      "step": 11136
+    },
+    {
+      "epoch": 16.672155688622755,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1602,
+      "step": 11137
+    },
+    {
+      "epoch": 16.67365269461078,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1489,
+      "step": 11138
+    },
+    {
+      "epoch": 16.6751497005988,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1504,
+      "step": 11139
+    },
+    {
+      "epoch": 16.676646706586826,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1483,
+      "step": 11140
+    },
+    {
+      "epoch": 16.67814371257485,
+      "grad_norm": 0.115234375,
+      "learning_rate": 0.0008,
+      "loss": 1.153,
+      "step": 11141
+    },
+    {
+      "epoch": 16.679640718562876,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1476,
+      "step": 11142
+    },
+    {
+      "epoch": 16.681137724550897,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1546,
+      "step": 11143
+    },
+    {
+      "epoch": 16.682634730538922,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1521,
+      "step": 11144
+    },
+    {
+      "epoch": 16.684131736526947,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1565,
+      "step": 11145
+    },
+    {
+      "epoch": 16.68562874251497,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1555,
+      "step": 11146
+    },
+    {
+      "epoch": 16.687125748502993,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1517,
+      "step": 11147
+    },
+    {
+      "epoch": 16.688622754491018,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1492,
+      "step": 11148
+    },
+    {
+      "epoch": 16.690119760479043,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1494,
+      "step": 11149
+    },
+    {
+      "epoch": 16.691616766467067,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1535,
+      "step": 11150
+    },
+    {
+      "epoch": 16.69311377245509,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1471,
+      "step": 11151
+    },
+    {
+      "epoch": 16.694610778443113,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1564,
+      "step": 11152
+    },
+    {
+      "epoch": 16.69610778443114,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1487,
+      "step": 11153
+    },
+    {
+      "epoch": 16.697604790419163,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1528,
+      "step": 11154
+    },
+    {
+      "epoch": 16.699101796407184,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1518,
+      "step": 11155
+    },
+    {
+      "epoch": 16.70059880239521,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 11156
+    },
+    {
+      "epoch": 16.702095808383234,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.149,
+      "step": 11157
+    },
+    {
+      "epoch": 16.70359281437126,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1488,
+      "step": 11158
+    },
+    {
+      "epoch": 16.70508982035928,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1471,
+      "step": 11159
+    },
+    {
+      "epoch": 16.706586826347305,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1465,
+      "step": 11160
+    },
+    {
+      "epoch": 16.70808383233533,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1531,
+      "step": 11161
+    },
+    {
+      "epoch": 16.709580838323355,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1481,
+      "step": 11162
+    },
+    {
+      "epoch": 16.711077844311376,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.147,
+      "step": 11163
+    },
+    {
+      "epoch": 16.7125748502994,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1505,
+      "step": 11164
+    },
+    {
+      "epoch": 16.714071856287426,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 11165
+    },
+    {
+      "epoch": 16.71556886227545,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1524,
+      "step": 11166
+    },
+    {
+      "epoch": 16.71706586826347,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1476,
+      "step": 11167
+    },
+    {
+      "epoch": 16.718562874251496,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1575,
+      "step": 11168
+    },
+    {
+      "epoch": 16.72005988023952,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1608,
+      "step": 11169
+    },
+    {
+      "epoch": 16.721556886227546,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1537,
+      "step": 11170
+    },
+    {
+      "epoch": 16.723053892215567,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 11171
+    },
+    {
+      "epoch": 16.724550898203592,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 11172
+    },
+    {
+      "epoch": 16.726047904191617,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 11173
+    },
+    {
+      "epoch": 16.727544910179642,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1502,
+      "step": 11174
+    },
+    {
+      "epoch": 16.729041916167663,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1529,
+      "step": 11175
+    },
+    {
+      "epoch": 16.730538922155688,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1506,
+      "step": 11176
+    },
+    {
+      "epoch": 16.732035928143713,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1526,
+      "step": 11177
+    },
+    {
+      "epoch": 16.733532934131738,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1471,
+      "step": 11178
+    },
+    {
+      "epoch": 16.73502994011976,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.153,
+      "step": 11179
+    },
+    {
+      "epoch": 16.736526946107784,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1528,
+      "step": 11180
+    },
+    {
+      "epoch": 16.73802395209581,
+      "grad_norm": 0.125,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 11181
+    },
+    {
+      "epoch": 16.739520958083833,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1463,
+      "step": 11182
+    },
+    {
+      "epoch": 16.741017964071855,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1509,
+      "step": 11183
+    },
+    {
+      "epoch": 16.74251497005988,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1508,
+      "step": 11184
+    },
+    {
+      "epoch": 16.744011976047904,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1459,
+      "step": 11185
+    },
+    {
+      "epoch": 16.74550898203593,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 11186
+    },
+    {
+      "epoch": 16.74700598802395,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1492,
+      "step": 11187
+    },
+    {
+      "epoch": 16.748502994011975,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 11188
+    },
+    {
+      "epoch": 16.75,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1528,
+      "step": 11189
+    },
+    {
+      "epoch": 16.751497005988025,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1529,
+      "step": 11190
+    },
+    {
+      "epoch": 16.75299401197605,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1525,
+      "step": 11191
+    },
+    {
+      "epoch": 16.75449101796407,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 11192
+    },
+    {
+      "epoch": 16.755988023952096,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0008,
+      "loss": 1.147,
+      "step": 11193
+    },
+    {
+      "epoch": 16.75748502994012,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1531,
+      "step": 11194
+    },
+    {
+      "epoch": 16.758982035928145,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1559,
+      "step": 11195
+    },
+    {
+      "epoch": 16.760479041916167,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1571,
+      "step": 11196
+    },
+    {
+      "epoch": 16.76197604790419,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1498,
+      "step": 11197
+    },
+    {
+      "epoch": 16.763473053892216,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1517,
+      "step": 11198
+    },
+    {
+      "epoch": 16.76497005988024,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 11199
+    },
+    {
+      "epoch": 16.766467065868262,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 11200
+    },
+    {
+      "epoch": 16.767964071856287,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1509,
+      "step": 11201
+    },
+    {
+      "epoch": 16.769461077844312,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1484,
+      "step": 11202
+    },
+    {
+      "epoch": 16.770958083832337,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.152,
+      "step": 11203
+    },
+    {
+      "epoch": 16.772455089820358,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1491,
+      "step": 11204
+    },
+    {
+      "epoch": 16.773952095808383,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1484,
+      "step": 11205
+    },
+    {
+      "epoch": 16.775449101796408,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1542,
+      "step": 11206
+    },
+    {
+      "epoch": 16.776946107784433,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 11207
+    },
+    {
+      "epoch": 16.778443113772454,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1429,
+      "step": 11208
+    },
+    {
+      "epoch": 16.77994011976048,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 11209
+    },
+    {
+      "epoch": 16.781437125748504,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1484,
+      "step": 11210
+    },
+    {
+      "epoch": 16.78293413173653,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1487,
+      "step": 11211
+    },
+    {
+      "epoch": 16.78443113772455,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 11212
+    },
+    {
+      "epoch": 16.785928143712574,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 11213
+    },
+    {
+      "epoch": 16.7874251497006,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 11214
+    },
+    {
+      "epoch": 16.788922155688624,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1531,
+      "step": 11215
+    },
+    {
+      "epoch": 16.790419161676645,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 11216
+    },
+    {
+      "epoch": 16.79191616766467,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0008,
+      "loss": 1.149,
+      "step": 11217
+    },
+    {
+      "epoch": 16.793413173652695,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 11218
+    },
+    {
+      "epoch": 16.79491017964072,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1505,
+      "step": 11219
+    },
+    {
+      "epoch": 16.79640718562874,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 11220
+    },
+    {
+      "epoch": 16.797904191616766,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1485,
+      "step": 11221
+    },
+    {
+      "epoch": 16.79940119760479,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 11222
+    },
+    {
+      "epoch": 16.800898203592816,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1495,
+      "step": 11223
+    },
+    {
+      "epoch": 16.802395209580837,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1493,
+      "step": 11224
+    },
+    {
+      "epoch": 16.80389221556886,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1457,
+      "step": 11225
+    },
+    {
+      "epoch": 16.805389221556887,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1476,
+      "step": 11226
+    },
+    {
+      "epoch": 16.80688622754491,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 11227
+    },
+    {
+      "epoch": 16.808383233532933,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1507,
+      "step": 11228
+    },
+    {
+      "epoch": 16.809880239520957,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1493,
+      "step": 11229
+    },
+    {
+      "epoch": 16.811377245508982,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1504,
+      "step": 11230
+    },
+    {
+      "epoch": 16.812874251497007,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 11231
+    },
+    {
+      "epoch": 16.81437125748503,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1539,
+      "step": 11232
+    },
+    {
+      "epoch": 16.815868263473053,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1446,
+      "step": 11233
+    },
+    {
+      "epoch": 16.817365269461078,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 11234
+    },
+    {
+      "epoch": 16.818862275449103,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 11235
+    },
+    {
+      "epoch": 16.820359281437124,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 11236
+    },
+    {
+      "epoch": 16.82185628742515,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1512,
+      "step": 11237
+    },
+    {
+      "epoch": 16.823353293413174,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 11238
+    },
+    {
+      "epoch": 16.8248502994012,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 11239
+    },
+    {
+      "epoch": 16.82634730538922,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 11240
+    },
+    {
+      "epoch": 16.827844311377245,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 11241
+    },
+    {
+      "epoch": 16.82934131736527,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1468,
+      "step": 11242
+    },
+    {
+      "epoch": 16.830838323353294,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 11243
+    },
+    {
+      "epoch": 16.83233532934132,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1524,
+      "step": 11244
+    },
+    {
+      "epoch": 16.83383233532934,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1468,
+      "step": 11245
+    },
+    {
+      "epoch": 16.835329341317365,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1479,
+      "step": 11246
+    },
+    {
+      "epoch": 16.83682634730539,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 11247
+    },
+    {
+      "epoch": 16.83832335329341,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 11248
+    },
+    {
+      "epoch": 16.839820359281436,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1446,
+      "step": 11249
+    },
+    {
+      "epoch": 16.84131736526946,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1484,
+      "step": 11250
+    },
+    {
+      "epoch": 16.842814371257486,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1464,
+      "step": 11251
+    },
+    {
+      "epoch": 16.84431137724551,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1446,
+      "step": 11252
+    },
+    {
+      "epoch": 16.845808383233532,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 11253
+    },
+    {
+      "epoch": 16.847305389221557,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 0.0008,
+      "loss": 1.149,
+      "step": 11254
+    },
+    {
+      "epoch": 16.84880239520958,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 11255
+    },
+    {
+      "epoch": 16.850299401197606,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 11256
+    },
+    {
+      "epoch": 16.851796407185628,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 11257
+    },
+    {
+      "epoch": 16.853293413173652,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1494,
+      "step": 11258
+    },
+    {
+      "epoch": 16.854790419161677,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 11259
+    },
+    {
+      "epoch": 16.856287425149702,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 11260
+    },
+    {
+      "epoch": 16.857784431137723,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1465,
+      "step": 11261
+    },
+    {
+      "epoch": 16.85928143712575,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 11262
+    },
+    {
+      "epoch": 16.860778443113773,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1486,
+      "step": 11263
+    },
+    {
+      "epoch": 16.862275449101798,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1518,
+      "step": 11264
+    },
+    {
+      "epoch": 16.86377245508982,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1393,
+      "step": 11265
+    },
+    {
+      "epoch": 16.865269461077844,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0008,
+      "loss": 1.148,
+      "step": 11266
+    },
+    {
+      "epoch": 16.86676646706587,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1459,
+      "step": 11267
+    },
+    {
+      "epoch": 16.868263473053894,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1498,
+      "step": 11268
+    },
+    {
+      "epoch": 16.869760479041915,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 11269
+    },
+    {
+      "epoch": 16.87125748502994,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 11270
+    },
+    {
+      "epoch": 16.872754491017965,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1509,
+      "step": 11271
+    },
+    {
+      "epoch": 16.87425149700599,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 11272
+    },
+    {
+      "epoch": 16.87574850299401,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 11273
+    },
+    {
+      "epoch": 16.877245508982035,
+      "grad_norm": 0.1953125,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 11274
+    },
+    {
+      "epoch": 16.87874251497006,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1429,
+      "step": 11275
+    },
+    {
+      "epoch": 16.880239520958085,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 11276
+    },
+    {
+      "epoch": 16.881736526946106,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1496,
+      "step": 11277
+    },
+    {
+      "epoch": 16.88323353293413,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1468,
+      "step": 11278
+    },
+    {
+      "epoch": 16.884730538922156,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 11279
+    },
+    {
+      "epoch": 16.88622754491018,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1487,
+      "step": 11280
+    },
+    {
+      "epoch": 16.887724550898202,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 11281
+    },
+    {
+      "epoch": 16.889221556886227,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1482,
+      "step": 11282
+    },
+    {
+      "epoch": 16.89071856287425,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 11283
+    },
+    {
+      "epoch": 16.892215568862277,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 11284
+    },
+    {
+      "epoch": 16.893712574850298,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1462,
+      "step": 11285
+    },
+    {
+      "epoch": 16.895209580838323,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1491,
+      "step": 11286
+    },
+    {
+      "epoch": 16.896706586826348,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1524,
+      "step": 11287
+    },
+    {
+      "epoch": 16.898203592814372,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1512,
+      "step": 11288
+    },
+    {
+      "epoch": 16.899700598802394,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 11289
+    },
+    {
+      "epoch": 16.90119760479042,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1506,
+      "step": 11290
+    },
+    {
+      "epoch": 16.902694610778443,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 11291
+    },
+    {
+      "epoch": 16.904191616766468,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 11292
+    },
+    {
+      "epoch": 16.90568862275449,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 11293
+    },
+    {
+      "epoch": 16.907185628742514,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1508,
+      "step": 11294
+    },
+    {
+      "epoch": 16.90868263473054,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1524,
+      "step": 11295
+    },
+    {
+      "epoch": 16.910179640718564,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1459,
+      "step": 11296
+    },
+    {
+      "epoch": 16.91167664670659,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1465,
+      "step": 11297
+    },
+    {
+      "epoch": 16.91317365269461,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 11298
+    },
+    {
+      "epoch": 16.914670658682635,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 11299
+    },
+    {
+      "epoch": 16.91616766467066,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1471,
+      "step": 11300
+    },
+    {
+      "epoch": 16.91766467065868,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 11301
+    },
+    {
+      "epoch": 16.919161676646706,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 11302
+    },
+    {
+      "epoch": 16.92065868263473,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 11303
+    },
+    {
+      "epoch": 16.922155688622755,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1416,
+      "step": 11304
+    },
+    {
+      "epoch": 16.92365269461078,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 11305
+    },
+    {
+      "epoch": 16.9251497005988,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1488,
+      "step": 11306
+    },
+    {
+      "epoch": 16.926646706586826,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 11307
+    },
+    {
+      "epoch": 16.92814371257485,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 11308
+    },
+    {
+      "epoch": 16.929640718562876,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1462,
+      "step": 11309
+    },
+    {
+      "epoch": 16.931137724550897,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1495,
+      "step": 11310
+    },
+    {
+      "epoch": 16.932634730538922,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1491,
+      "step": 11311
+    },
+    {
+      "epoch": 16.934131736526947,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1457,
+      "step": 11312
+    },
+    {
+      "epoch": 16.93562874251497,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 11313
+    },
+    {
+      "epoch": 16.937125748502993,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 11314
+    },
+    {
+      "epoch": 16.938622754491018,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 11315
+    },
+    {
+      "epoch": 16.940119760479043,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 11316
+    },
+    {
+      "epoch": 16.941616766467067,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1533,
+      "step": 11317
+    },
+    {
+      "epoch": 16.94311377245509,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 11318
+    },
+    {
+      "epoch": 16.944610778443113,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1473,
+      "step": 11319
+    },
+    {
+      "epoch": 16.94610778443114,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1475,
+      "step": 11320
+    },
+    {
+      "epoch": 16.947604790419163,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.147,
+      "step": 11321
+    },
+    {
+      "epoch": 16.949101796407184,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1459,
+      "step": 11322
+    },
+    {
+      "epoch": 16.95059880239521,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 11323
+    },
+    {
+      "epoch": 16.952095808383234,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1378,
+      "step": 11324
+    },
+    {
+      "epoch": 16.95359281437126,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1416,
+      "step": 11325
+    },
+    {
+      "epoch": 16.95508982035928,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 11326
+    },
+    {
+      "epoch": 16.956586826347305,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 11327
+    },
+    {
+      "epoch": 16.95808383233533,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 11328
+    },
+    {
+      "epoch": 16.959580838323355,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 11329
+    },
+    {
+      "epoch": 16.961077844311376,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 11330
+    },
+    {
+      "epoch": 16.9625748502994,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.149,
+      "step": 11331
+    },
+    {
+      "epoch": 16.964071856287426,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 11332
+    },
+    {
+      "epoch": 16.96556886227545,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 11333
+    },
+    {
+      "epoch": 16.96706586826347,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1507,
+      "step": 11334
+    },
+    {
+      "epoch": 16.968562874251496,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 11335
+    },
+    {
+      "epoch": 16.97005988023952,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1459,
+      "step": 11336
+    },
+    {
+      "epoch": 16.971556886227546,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 11337
+    },
+    {
+      "epoch": 16.973053892215567,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1386,
+      "step": 11338
+    },
+    {
+      "epoch": 16.974550898203592,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1506,
+      "step": 11339
+    },
+    {
+      "epoch": 16.976047904191617,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 11340
+    },
+    {
+      "epoch": 16.977544910179642,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 11341
+    },
+    {
+      "epoch": 16.979041916167663,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1459,
+      "step": 11342
+    },
+    {
+      "epoch": 16.980538922155688,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.144,
+      "step": 11343
+    },
+    {
+      "epoch": 16.982035928143713,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 11344
+    },
+    {
+      "epoch": 16.983532934131738,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1483,
+      "step": 11345
+    },
+    {
+      "epoch": 16.98502994011976,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 11346
+    },
+    {
+      "epoch": 16.986526946107784,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1493,
+      "step": 11347
+    },
+    {
+      "epoch": 16.98802395209581,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1338,
+      "step": 11348
+    },
+    {
+      "epoch": 16.989520958083833,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 11349
+    },
+    {
+      "epoch": 16.991017964071855,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 11350
+    },
+    {
+      "epoch": 16.99251497005988,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 11351
+    },
+    {
+      "epoch": 16.994011976047904,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1516,
+      "step": 11352
+    },
+    {
+      "epoch": 16.99550898203593,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 11353
+    },
+    {
+      "epoch": 16.99700598802395,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 11354
+    },
+    {
+      "epoch": 16.998502994011975,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 11355
+    },
+    {
+      "epoch": 17.0,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 11356
+    },
+    {
+      "epoch": 17.001497005988025,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1538,
+      "step": 11357
+    },
+    {
+      "epoch": 17.00299401197605,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 11358
+    },
+    {
+      "epoch": 17.00449101796407,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 11359
+    },
+    {
+      "epoch": 17.005988023952096,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1519,
+      "step": 11360
+    },
+    {
+      "epoch": 17.00748502994012,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 11361
+    },
+    {
+      "epoch": 17.008982035928145,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1473,
+      "step": 11362
+    },
+    {
+      "epoch": 17.010479041916167,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 11363
+    },
+    {
+      "epoch": 17.01197604790419,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1489,
+      "step": 11364
+    },
+    {
+      "epoch": 17.013473053892216,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1472,
+      "step": 11365
+    },
+    {
+      "epoch": 17.01497005988024,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 11366
+    },
+    {
+      "epoch": 17.016467065868262,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 11367
+    },
+    {
+      "epoch": 17.017964071856287,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1392,
+      "step": 11368
+    },
+    {
+      "epoch": 17.019461077844312,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1482,
+      "step": 11369
+    },
+    {
+      "epoch": 17.020958083832337,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1505,
+      "step": 11370
+    },
+    {
+      "epoch": 17.022455089820358,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1381,
+      "step": 11371
+    },
+    {
+      "epoch": 17.023952095808383,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 11372
+    },
+    {
+      "epoch": 17.025449101796408,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 11373
+    },
+    {
+      "epoch": 17.026946107784433,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1469,
+      "step": 11374
+    },
+    {
+      "epoch": 17.028443113772454,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 11375
+    },
+    {
+      "epoch": 17.02994011976048,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1483,
+      "step": 11376
+    },
+    {
+      "epoch": 17.031437125748504,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 11377
+    },
+    {
+      "epoch": 17.03293413173653,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 11378
+    },
+    {
+      "epoch": 17.03443113772455,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1482,
+      "step": 11379
+    },
+    {
+      "epoch": 17.035928143712574,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 11380
+    },
+    {
+      "epoch": 17.0374251497006,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0008,
+      "loss": 1.146,
+      "step": 11381
+    },
+    {
+      "epoch": 17.038922155688624,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 11382
+    },
+    {
+      "epoch": 17.040419161676645,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 11383
+    },
+    {
+      "epoch": 17.04191616766467,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 11384
+    },
+    {
+      "epoch": 17.043413173652695,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1484,
+      "step": 11385
+    },
+    {
+      "epoch": 17.04491017964072,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 11386
+    },
+    {
+      "epoch": 17.04640718562874,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 11387
+    },
+    {
+      "epoch": 17.047904191616766,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1446,
+      "step": 11388
+    },
+    {
+      "epoch": 17.04940119760479,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 11389
+    },
+    {
+      "epoch": 17.050898203592816,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 11390
+    },
+    {
+      "epoch": 17.052395209580837,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1481,
+      "step": 11391
+    },
+    {
+      "epoch": 17.05389221556886,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1428,
+      "step": 11392
+    },
+    {
+      "epoch": 17.055389221556887,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1457,
+      "step": 11393
+    },
+    {
+      "epoch": 17.05688622754491,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 11394
+    },
+    {
+      "epoch": 17.058383233532933,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1416,
+      "step": 11395
+    },
+    {
+      "epoch": 17.059880239520957,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1469,
+      "step": 11396
+    },
+    {
+      "epoch": 17.061377245508982,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1495,
+      "step": 11397
+    },
+    {
+      "epoch": 17.062874251497007,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 11398
+    },
+    {
+      "epoch": 17.06437125748503,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 11399
+    },
+    {
+      "epoch": 17.065868263473053,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.144,
+      "step": 11400
+    },
+    {
+      "epoch": 17.067365269461078,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 11401
+    },
+    {
+      "epoch": 17.068862275449103,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1446,
+      "step": 11402
+    },
+    {
+      "epoch": 17.070359281437124,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 11403
+    },
+    {
+      "epoch": 17.07185628742515,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 11404
+    },
+    {
+      "epoch": 17.073353293413174,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 11405
+    },
+    {
+      "epoch": 17.0748502994012,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 11406
+    },
+    {
+      "epoch": 17.07634730538922,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1341,
+      "step": 11407
+    },
+    {
+      "epoch": 17.077844311377245,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0008,
+      "loss": 1.136,
+      "step": 11408
+    },
+    {
+      "epoch": 17.07934131736527,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1488,
+      "step": 11409
+    },
+    {
+      "epoch": 17.080838323353294,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1499,
+      "step": 11410
+    },
+    {
+      "epoch": 17.082335329341316,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 11411
+    },
+    {
+      "epoch": 17.08383233532934,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1475,
+      "step": 11412
+    },
+    {
+      "epoch": 17.085329341317365,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 11413
+    },
+    {
+      "epoch": 17.08682634730539,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 11414
+    },
+    {
+      "epoch": 17.088323353293415,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.147,
+      "step": 11415
+    },
+    {
+      "epoch": 17.089820359281436,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 11416
+    },
+    {
+      "epoch": 17.09131736526946,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1469,
+      "step": 11417
+    },
+    {
+      "epoch": 17.092814371257486,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 11418
+    },
+    {
+      "epoch": 17.09431137724551,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 11419
+    },
+    {
+      "epoch": 17.095808383233532,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 11420
+    },
+    {
+      "epoch": 17.097305389221557,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 11421
+    },
+    {
+      "epoch": 17.09880239520958,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1336,
+      "step": 11422
+    },
+    {
+      "epoch": 17.100299401197606,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1369,
+      "step": 11423
+    },
+    {
+      "epoch": 17.101796407185628,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 11424
+    },
+    {
+      "epoch": 17.103293413173652,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.15,
+      "step": 11425
+    },
+    {
+      "epoch": 17.104790419161677,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 11426
+    },
+    {
+      "epoch": 17.106287425149702,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1381,
+      "step": 11427
+    },
+    {
+      "epoch": 17.107784431137723,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 11428
+    },
+    {
+      "epoch": 17.10928143712575,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 11429
+    },
+    {
+      "epoch": 17.110778443113773,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 11430
+    },
+    {
+      "epoch": 17.112275449101798,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 11431
+    },
+    {
+      "epoch": 17.11377245508982,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 11432
+    },
+    {
+      "epoch": 17.115269461077844,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 11433
+    },
+    {
+      "epoch": 17.11676646706587,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.15,
+      "step": 11434
+    },
+    {
+      "epoch": 17.118263473053894,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1393,
+      "step": 11435
+    },
+    {
+      "epoch": 17.119760479041915,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 11436
+    },
+    {
+      "epoch": 17.12125748502994,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 11437
+    },
+    {
+      "epoch": 17.122754491017965,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 11438
+    },
+    {
+      "epoch": 17.12425149700599,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1395,
+      "step": 11439
+    },
+    {
+      "epoch": 17.12574850299401,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 11440
+    },
+    {
+      "epoch": 17.127245508982035,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 11441
+    },
+    {
+      "epoch": 17.12874251497006,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 11442
+    },
+    {
+      "epoch": 17.130239520958085,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.15,
+      "step": 11443
+    },
+    {
+      "epoch": 17.131736526946106,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 11444
+    },
+    {
+      "epoch": 17.13323353293413,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 11445
+    },
+    {
+      "epoch": 17.134730538922156,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 11446
+    },
+    {
+      "epoch": 17.13622754491018,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 11447
+    },
+    {
+      "epoch": 17.137724550898202,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 11448
+    },
+    {
+      "epoch": 17.139221556886227,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1443,
+      "step": 11449
+    },
+    {
+      "epoch": 17.14071856287425,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1463,
+      "step": 11450
+    },
+    {
+      "epoch": 17.142215568862277,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 11451
+    },
+    {
+      "epoch": 17.143712574850298,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 11452
+    },
+    {
+      "epoch": 17.145209580838323,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 11453
+    },
+    {
+      "epoch": 17.146706586826348,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 11454
+    },
+    {
+      "epoch": 17.148203592814372,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 11455
+    },
+    {
+      "epoch": 17.149700598802394,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 11456
+    },
+    {
+      "epoch": 17.15119760479042,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1384,
+      "step": 11457
+    },
+    {
+      "epoch": 17.152694610778443,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1361,
+      "step": 11458
+    },
+    {
+      "epoch": 17.154191616766468,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1387,
+      "step": 11459
+    },
+    {
+      "epoch": 17.15568862275449,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 11460
+    },
+    {
+      "epoch": 17.157185628742514,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1358,
+      "step": 11461
+    },
+    {
+      "epoch": 17.15868263473054,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1484,
+      "step": 11462
+    },
+    {
+      "epoch": 17.160179640718564,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1515,
+      "step": 11463
+    },
+    {
+      "epoch": 17.161676646706585,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1465,
+      "step": 11464
+    },
+    {
+      "epoch": 17.16317365269461,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 11465
+    },
+    {
+      "epoch": 17.164670658682635,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 11466
+    },
+    {
+      "epoch": 17.16616766467066,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1379,
+      "step": 11467
+    },
+    {
+      "epoch": 17.16766467065868,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1491,
+      "step": 11468
+    },
+    {
+      "epoch": 17.169161676646706,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 11469
+    },
+    {
+      "epoch": 17.17065868263473,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 11470
+    },
+    {
+      "epoch": 17.172155688622755,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 11471
+    },
+    {
+      "epoch": 17.17365269461078,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1416,
+      "step": 11472
+    },
+    {
+      "epoch": 17.1751497005988,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.14,
+      "step": 11473
+    },
+    {
+      "epoch": 17.176646706586826,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 11474
+    },
+    {
+      "epoch": 17.17814371257485,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1348,
+      "step": 11475
+    },
+    {
+      "epoch": 17.179640718562876,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 11476
+    },
+    {
+      "epoch": 17.181137724550897,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1356,
+      "step": 11477
+    },
+    {
+      "epoch": 17.182634730538922,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 11478
+    },
+    {
+      "epoch": 17.184131736526947,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 11479
+    },
+    {
+      "epoch": 17.18562874251497,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 11480
+    },
+    {
+      "epoch": 17.187125748502993,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1387,
+      "step": 11481
+    },
+    {
+      "epoch": 17.188622754491018,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 11482
+    },
+    {
+      "epoch": 17.190119760479043,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1462,
+      "step": 11483
+    },
+    {
+      "epoch": 17.191616766467067,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 11484
+    },
+    {
+      "epoch": 17.19311377245509,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 11485
+    },
+    {
+      "epoch": 17.194610778443113,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 11486
+    },
+    {
+      "epoch": 17.19610778443114,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 11487
+    },
+    {
+      "epoch": 17.197604790419163,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1372,
+      "step": 11488
+    },
+    {
+      "epoch": 17.199101796407184,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 11489
+    },
+    {
+      "epoch": 17.20059880239521,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 11490
+    },
+    {
+      "epoch": 17.202095808383234,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 11491
+    },
+    {
+      "epoch": 17.20359281437126,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1372,
+      "step": 11492
+    },
+    {
+      "epoch": 17.20508982035928,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1336,
+      "step": 11493
+    },
+    {
+      "epoch": 17.206586826347305,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1468,
+      "step": 11494
+    },
+    {
+      "epoch": 17.20808383233533,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 11495
+    },
+    {
+      "epoch": 17.209580838323355,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 11496
+    },
+    {
+      "epoch": 17.211077844311376,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1386,
+      "step": 11497
+    },
+    {
+      "epoch": 17.2125748502994,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 11498
+    },
+    {
+      "epoch": 17.214071856287426,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 11499
+    },
+    {
+      "epoch": 17.21556886227545,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1481,
+      "step": 11500
+    },
+    {
+      "epoch": 17.21706586826347,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 11501
+    },
+    {
+      "epoch": 17.218562874251496,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1463,
+      "step": 11502
+    },
+    {
+      "epoch": 17.22005988023952,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 11503
+    },
+    {
+      "epoch": 17.221556886227546,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 11504
+    },
+    {
+      "epoch": 17.223053892215567,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 11505
+    },
+    {
+      "epoch": 17.224550898203592,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 11506
+    },
+    {
+      "epoch": 17.226047904191617,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 11507
+    },
+    {
+      "epoch": 17.227544910179642,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 11508
+    },
+    {
+      "epoch": 17.229041916167663,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1476,
+      "step": 11509
+    },
+    {
+      "epoch": 17.230538922155688,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 11510
+    },
+    {
+      "epoch": 17.232035928143713,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1468,
+      "step": 11511
+    },
+    {
+      "epoch": 17.233532934131738,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 11512
+    },
+    {
+      "epoch": 17.23502994011976,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1322,
+      "step": 11513
+    },
+    {
+      "epoch": 17.236526946107784,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 11514
+    },
+    {
+      "epoch": 17.23802395209581,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 11515
+    },
+    {
+      "epoch": 17.239520958083833,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 11516
+    },
+    {
+      "epoch": 17.241017964071855,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.134,
+      "step": 11517
+    },
+    {
+      "epoch": 17.24251497005988,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 11518
+    },
+    {
+      "epoch": 17.244011976047904,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 11519
+    },
+    {
+      "epoch": 17.24550898203593,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 11520
+    },
+    {
+      "epoch": 17.24700598802395,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 11521
+    },
+    {
+      "epoch": 17.248502994011975,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1369,
+      "step": 11522
+    },
+    {
+      "epoch": 17.25,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 11523
+    },
+    {
+      "epoch": 17.251497005988025,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 11524
+    },
+    {
+      "epoch": 17.25299401197605,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1386,
+      "step": 11525
+    },
+    {
+      "epoch": 17.25449101796407,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1386,
+      "step": 11526
+    },
+    {
+      "epoch": 17.255988023952096,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 11527
+    },
+    {
+      "epoch": 17.25748502994012,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 11528
+    },
+    {
+      "epoch": 17.258982035928145,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 11529
+    },
+    {
+      "epoch": 17.260479041916167,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 11530
+    },
+    {
+      "epoch": 17.26197604790419,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 11531
+    },
+    {
+      "epoch": 17.263473053892216,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1443,
+      "step": 11532
+    },
+    {
+      "epoch": 17.26497005988024,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 11533
+    },
+    {
+      "epoch": 17.266467065868262,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 11534
+    },
+    {
+      "epoch": 17.267964071856287,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1428,
+      "step": 11535
+    },
+    {
+      "epoch": 17.269461077844312,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 11536
+    },
+    {
+      "epoch": 17.270958083832337,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 11537
+    },
+    {
+      "epoch": 17.272455089820358,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1463,
+      "step": 11538
+    },
+    {
+      "epoch": 17.273952095808383,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 11539
+    },
+    {
+      "epoch": 17.275449101796408,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 11540
+    },
+    {
+      "epoch": 17.276946107784433,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 11541
+    },
+    {
+      "epoch": 17.278443113772454,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 11542
+    },
+    {
+      "epoch": 17.27994011976048,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1364,
+      "step": 11543
+    },
+    {
+      "epoch": 17.281437125748504,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 11544
+    },
+    {
+      "epoch": 17.28293413173653,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 11545
+    },
+    {
+      "epoch": 17.28443113772455,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.137,
+      "step": 11546
+    },
+    {
+      "epoch": 17.285928143712574,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1472,
+      "step": 11547
+    },
+    {
+      "epoch": 17.2874251497006,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 11548
+    },
+    {
+      "epoch": 17.288922155688624,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 11549
+    },
+    {
+      "epoch": 17.290419161676645,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1319,
+      "step": 11550
+    },
+    {
+      "epoch": 17.29191616766467,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1334,
+      "step": 11551
+    },
+    {
+      "epoch": 17.293413173652695,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 11552
+    },
+    {
+      "epoch": 17.29491017964072,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 11553
+    },
+    {
+      "epoch": 17.29640718562874,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 11554
+    },
+    {
+      "epoch": 17.297904191616766,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 11555
+    },
+    {
+      "epoch": 17.29940119760479,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 11556
+    },
+    {
+      "epoch": 17.300898203592816,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 11557
+    },
+    {
+      "epoch": 17.302395209580837,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 11558
+    },
+    {
+      "epoch": 17.30389221556886,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 11559
+    },
+    {
+      "epoch": 17.305389221556887,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1357,
+      "step": 11560
+    },
+    {
+      "epoch": 17.30688622754491,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 11561
+    },
+    {
+      "epoch": 17.308383233532933,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 11562
+    },
+    {
+      "epoch": 17.309880239520957,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 11563
+    },
+    {
+      "epoch": 17.311377245508982,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 11564
+    },
+    {
+      "epoch": 17.312874251497007,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 11565
+    },
+    {
+      "epoch": 17.31437125748503,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1393,
+      "step": 11566
+    },
+    {
+      "epoch": 17.315868263473053,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 11567
+    },
+    {
+      "epoch": 17.317365269461078,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.14,
+      "step": 11568
+    },
+    {
+      "epoch": 17.318862275449103,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1464,
+      "step": 11569
+    },
+    {
+      "epoch": 17.320359281437124,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 11570
+    },
+    {
+      "epoch": 17.32185628742515,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 11571
+    },
+    {
+      "epoch": 17.323353293413174,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 11572
+    },
+    {
+      "epoch": 17.3248502994012,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1494,
+      "step": 11573
+    },
+    {
+      "epoch": 17.32634730538922,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1481,
+      "step": 11574
+    },
+    {
+      "epoch": 17.327844311377245,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 11575
+    },
+    {
+      "epoch": 17.32934131736527,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 11576
+    },
+    {
+      "epoch": 17.330838323353294,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1459,
+      "step": 11577
+    },
+    {
+      "epoch": 17.33233532934132,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1462,
+      "step": 11578
+    },
+    {
+      "epoch": 17.33383233532934,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 11579
+    },
+    {
+      "epoch": 17.335329341317365,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 11580
+    },
+    {
+      "epoch": 17.33682634730539,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 11581
+    },
+    {
+      "epoch": 17.338323353293415,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 11582
+    },
+    {
+      "epoch": 17.339820359281436,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 11583
+    },
+    {
+      "epoch": 17.34131736526946,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 11584
+    },
+    {
+      "epoch": 17.342814371257486,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 11585
+    },
+    {
+      "epoch": 17.34431137724551,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1328,
+      "step": 11586
+    },
+    {
+      "epoch": 17.345808383233532,
+      "grad_norm": 0.119140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 11587
+    },
+    {
+      "epoch": 17.347305389221557,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1384,
+      "step": 11588
+    },
+    {
+      "epoch": 17.34880239520958,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1465,
+      "step": 11589
+    },
+    {
+      "epoch": 17.350299401197606,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 11590
+    },
+    {
+      "epoch": 17.351796407185628,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 11591
+    },
+    {
+      "epoch": 17.353293413173652,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1463,
+      "step": 11592
+    },
+    {
+      "epoch": 17.354790419161677,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1454,
+      "step": 11593
+    },
+    {
+      "epoch": 17.356287425149702,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 11594
+    },
+    {
+      "epoch": 17.357784431137723,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1446,
+      "step": 11595
+    },
+    {
+      "epoch": 17.35928143712575,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 11596
+    },
+    {
+      "epoch": 17.360778443113773,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 11597
+    },
+    {
+      "epoch": 17.362275449101798,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 11598
+    },
+    {
+      "epoch": 17.36377245508982,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 11599
+    },
+    {
+      "epoch": 17.365269461077844,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 11600
+    },
+    {
+      "epoch": 17.36676646706587,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1327,
+      "step": 11601
+    },
+    {
+      "epoch": 17.368263473053894,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1371,
+      "step": 11602
+    },
+    {
+      "epoch": 17.369760479041915,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 11603
+    },
+    {
+      "epoch": 17.37125748502994,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 11604
+    },
+    {
+      "epoch": 17.372754491017965,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 11605
+    },
+    {
+      "epoch": 17.37425149700599,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1492,
+      "step": 11606
+    },
+    {
+      "epoch": 17.37574850299401,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 11607
+    },
+    {
+      "epoch": 17.377245508982035,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1457,
+      "step": 11608
+    },
+    {
+      "epoch": 17.37874251497006,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1371,
+      "step": 11609
+    },
+    {
+      "epoch": 17.380239520958085,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1463,
+      "step": 11610
+    },
+    {
+      "epoch": 17.381736526946106,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 11611
+    },
+    {
+      "epoch": 17.38323353293413,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1362,
+      "step": 11612
+    },
+    {
+      "epoch": 17.384730538922156,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 11613
+    },
+    {
+      "epoch": 17.38622754491018,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 11614
+    },
+    {
+      "epoch": 17.387724550898202,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 11615
+    },
+    {
+      "epoch": 17.389221556886227,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 11616
+    },
+    {
+      "epoch": 17.39071856287425,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 11617
+    },
+    {
+      "epoch": 17.392215568862277,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 11618
+    },
+    {
+      "epoch": 17.393712574850298,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 11619
+    },
+    {
+      "epoch": 17.395209580838323,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1471,
+      "step": 11620
+    },
+    {
+      "epoch": 17.396706586826348,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1375,
+      "step": 11621
+    },
+    {
+      "epoch": 17.398203592814372,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 11622
+    },
+    {
+      "epoch": 17.399700598802394,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 11623
+    },
+    {
+      "epoch": 17.40119760479042,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 11624
+    },
+    {
+      "epoch": 17.402694610778443,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 11625
+    },
+    {
+      "epoch": 17.404191616766468,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 11626
+    },
+    {
+      "epoch": 17.40568862275449,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 11627
+    },
+    {
+      "epoch": 17.407185628742514,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1416,
+      "step": 11628
+    },
+    {
+      "epoch": 17.40868263473054,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1344,
+      "step": 11629
+    },
+    {
+      "epoch": 17.410179640718564,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 11630
+    },
+    {
+      "epoch": 17.411676646706585,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 11631
+    },
+    {
+      "epoch": 17.41317365269461,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1413,
+      "step": 11632
+    },
+    {
+      "epoch": 17.414670658682635,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1462,
+      "step": 11633
+    },
+    {
+      "epoch": 17.41616766467066,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 11634
+    },
+    {
+      "epoch": 17.41766467065868,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1318,
+      "step": 11635
+    },
+    {
+      "epoch": 17.419161676646706,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 11636
+    },
+    {
+      "epoch": 17.42065868263473,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1468,
+      "step": 11637
+    },
+    {
+      "epoch": 17.422155688622755,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 11638
+    },
+    {
+      "epoch": 17.42365269461078,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 11639
+    },
+    {
+      "epoch": 17.4251497005988,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 11640
+    },
+    {
+      "epoch": 17.426646706586826,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 11641
+    },
+    {
+      "epoch": 17.42814371257485,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 11642
+    },
+    {
+      "epoch": 17.429640718562876,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 11643
+    },
+    {
+      "epoch": 17.431137724550897,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 11644
+    },
+    {
+      "epoch": 17.432634730538922,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1479,
+      "step": 11645
+    },
+    {
+      "epoch": 17.434131736526947,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1357,
+      "step": 11646
+    },
+    {
+      "epoch": 17.43562874251497,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 11647
+    },
+    {
+      "epoch": 17.437125748502993,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1476,
+      "step": 11648
+    },
+    {
+      "epoch": 17.438622754491018,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 11649
+    },
+    {
+      "epoch": 17.440119760479043,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1485,
+      "step": 11650
+    },
+    {
+      "epoch": 17.441616766467067,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1331,
+      "step": 11651
+    },
+    {
+      "epoch": 17.44311377245509,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 11652
+    },
+    {
+      "epoch": 17.444610778443113,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1416,
+      "step": 11653
+    },
+    {
+      "epoch": 17.44610778443114,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 11654
+    },
+    {
+      "epoch": 17.447604790419163,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1459,
+      "step": 11655
+    },
+    {
+      "epoch": 17.449101796407184,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1379,
+      "step": 11656
+    },
+    {
+      "epoch": 17.45059880239521,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 11657
+    },
+    {
+      "epoch": 17.452095808383234,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 11658
+    },
+    {
+      "epoch": 17.45359281437126,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1378,
+      "step": 11659
+    },
+    {
+      "epoch": 17.45508982035928,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1371,
+      "step": 11660
+    },
+    {
+      "epoch": 17.456586826347305,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1462,
+      "step": 11661
+    },
+    {
+      "epoch": 17.45808383233533,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1395,
+      "step": 11662
+    },
+    {
+      "epoch": 17.459580838323355,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.137,
+      "step": 11663
+    },
+    {
+      "epoch": 17.461077844311376,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.144,
+      "step": 11664
+    },
+    {
+      "epoch": 17.4625748502994,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 11665
+    },
+    {
+      "epoch": 17.464071856287426,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 11666
+    },
+    {
+      "epoch": 17.46556886227545,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.133,
+      "step": 11667
+    },
+    {
+      "epoch": 17.46706586826347,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 11668
+    },
+    {
+      "epoch": 17.468562874251496,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 11669
+    },
+    {
+      "epoch": 17.47005988023952,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 11670
+    },
+    {
+      "epoch": 17.471556886227546,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1379,
+      "step": 11671
+    },
+    {
+      "epoch": 17.473053892215567,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1359,
+      "step": 11672
+    },
+    {
+      "epoch": 17.474550898203592,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 11673
+    },
+    {
+      "epoch": 17.476047904191617,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1443,
+      "step": 11674
+    },
+    {
+      "epoch": 17.477544910179642,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1387,
+      "step": 11675
+    },
+    {
+      "epoch": 17.479041916167663,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 11676
+    },
+    {
+      "epoch": 17.480538922155688,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 11677
+    },
+    {
+      "epoch": 17.482035928143713,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1383,
+      "step": 11678
+    },
+    {
+      "epoch": 17.483532934131738,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1392,
+      "step": 11679
+    },
+    {
+      "epoch": 17.48502994011976,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 11680
+    },
+    {
+      "epoch": 17.486526946107784,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 11681
+    },
+    {
+      "epoch": 17.48802395209581,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 11682
+    },
+    {
+      "epoch": 17.489520958083833,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 11683
+    },
+    {
+      "epoch": 17.491017964071855,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1393,
+      "step": 11684
+    },
+    {
+      "epoch": 17.49251497005988,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1357,
+      "step": 11685
+    },
+    {
+      "epoch": 17.494011976047904,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.144,
+      "step": 11686
+    },
+    {
+      "epoch": 17.49550898203593,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 11687
+    },
+    {
+      "epoch": 17.49700598802395,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 11688
+    },
+    {
+      "epoch": 17.498502994011975,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 11689
+    },
+    {
+      "epoch": 17.5,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1373,
+      "step": 11690
+    },
+    {
+      "epoch": 17.501497005988025,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 11691
+    },
+    {
+      "epoch": 17.50299401197605,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1384,
+      "step": 11692
+    },
+    {
+      "epoch": 17.50449101796407,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 11693
+    },
+    {
+      "epoch": 17.505988023952096,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 11694
+    },
+    {
+      "epoch": 17.50748502994012,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 11695
+    },
+    {
+      "epoch": 17.508982035928145,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 11696
+    },
+    {
+      "epoch": 17.510479041916167,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 11697
+    },
+    {
+      "epoch": 17.51197604790419,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1378,
+      "step": 11698
+    },
+    {
+      "epoch": 17.513473053892216,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 11699
+    },
+    {
+      "epoch": 17.51497005988024,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 11700
+    },
+    {
+      "epoch": 17.516467065868262,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 11701
+    },
+    {
+      "epoch": 17.517964071856287,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1341,
+      "step": 11702
+    },
+    {
+      "epoch": 17.519461077844312,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1485,
+      "step": 11703
+    },
+    {
+      "epoch": 17.520958083832337,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 11704
+    },
+    {
+      "epoch": 17.522455089820358,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1368,
+      "step": 11705
+    },
+    {
+      "epoch": 17.523952095808383,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 11706
+    },
+    {
+      "epoch": 17.525449101796408,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 11707
+    },
+    {
+      "epoch": 17.526946107784433,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 11708
+    },
+    {
+      "epoch": 17.528443113772454,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1384,
+      "step": 11709
+    },
+    {
+      "epoch": 17.52994011976048,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.14,
+      "step": 11710
+    },
+    {
+      "epoch": 17.531437125748504,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1357,
+      "step": 11711
+    },
+    {
+      "epoch": 17.53293413173653,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 11712
+    },
+    {
+      "epoch": 17.53443113772455,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1429,
+      "step": 11713
+    },
+    {
+      "epoch": 17.535928143712574,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 11714
+    },
+    {
+      "epoch": 17.5374251497006,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1443,
+      "step": 11715
+    },
+    {
+      "epoch": 17.538922155688624,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 11716
+    },
+    {
+      "epoch": 17.540419161676645,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 11717
+    },
+    {
+      "epoch": 17.54191616766467,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 11718
+    },
+    {
+      "epoch": 17.543413173652695,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 11719
+    },
+    {
+      "epoch": 17.54491017964072,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 11720
+    },
+    {
+      "epoch": 17.54640718562874,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1525,
+      "step": 11721
+    },
+    {
+      "epoch": 17.547904191616766,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1354,
+      "step": 11722
+    },
+    {
+      "epoch": 17.54940119760479,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 11723
+    },
+    {
+      "epoch": 17.550898203592816,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 11724
+    },
+    {
+      "epoch": 17.552395209580837,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 11725
+    },
+    {
+      "epoch": 17.55389221556886,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 11726
+    },
+    {
+      "epoch": 17.555389221556887,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1335,
+      "step": 11727
+    },
+    {
+      "epoch": 17.55688622754491,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.134,
+      "step": 11728
+    },
+    {
+      "epoch": 17.558383233532933,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 11729
+    },
+    {
+      "epoch": 17.559880239520957,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1323,
+      "step": 11730
+    },
+    {
+      "epoch": 17.561377245508982,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1483,
+      "step": 11731
+    },
+    {
+      "epoch": 17.562874251497007,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 11732
+    },
+    {
+      "epoch": 17.56437125748503,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 11733
+    },
+    {
+      "epoch": 17.565868263473053,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 11734
+    },
+    {
+      "epoch": 17.567365269461078,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 11735
+    },
+    {
+      "epoch": 17.568862275449103,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1377,
+      "step": 11736
+    },
+    {
+      "epoch": 17.570359281437124,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1472,
+      "step": 11737
+    },
+    {
+      "epoch": 17.57185628742515,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1457,
+      "step": 11738
+    },
+    {
+      "epoch": 17.573353293413174,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 11739
+    },
+    {
+      "epoch": 17.5748502994012,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1395,
+      "step": 11740
+    },
+    {
+      "epoch": 17.57634730538922,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1363,
+      "step": 11741
+    },
+    {
+      "epoch": 17.577844311377245,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 11742
+    },
+    {
+      "epoch": 17.57934131736527,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1367,
+      "step": 11743
+    },
+    {
+      "epoch": 17.580838323353294,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 11744
+    },
+    {
+      "epoch": 17.58233532934132,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1393,
+      "step": 11745
+    },
+    {
+      "epoch": 17.58383233532934,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 11746
+    },
+    {
+      "epoch": 17.585329341317365,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 11747
+    },
+    {
+      "epoch": 17.58682634730539,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 11748
+    },
+    {
+      "epoch": 17.58832335329341,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.144,
+      "step": 11749
+    },
+    {
+      "epoch": 17.589820359281436,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1365,
+      "step": 11750
+    },
+    {
+      "epoch": 17.59131736526946,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 11751
+    },
+    {
+      "epoch": 17.592814371257486,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 11752
+    },
+    {
+      "epoch": 17.59431137724551,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 11753
+    },
+    {
+      "epoch": 17.595808383233532,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 11754
+    },
+    {
+      "epoch": 17.597305389221557,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1377,
+      "step": 11755
+    },
+    {
+      "epoch": 17.59880239520958,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 11756
+    },
+    {
+      "epoch": 17.600299401197606,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 11757
+    },
+    {
+      "epoch": 17.601796407185628,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 11758
+    },
+    {
+      "epoch": 17.603293413173652,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 11759
+    },
+    {
+      "epoch": 17.604790419161677,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 11760
+    },
+    {
+      "epoch": 17.606287425149702,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 11761
+    },
+    {
+      "epoch": 17.607784431137723,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 11762
+    },
+    {
+      "epoch": 17.60928143712575,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 11763
+    },
+    {
+      "epoch": 17.610778443113773,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 11764
+    },
+    {
+      "epoch": 17.612275449101798,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 11765
+    },
+    {
+      "epoch": 17.61377245508982,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 11766
+    },
+    {
+      "epoch": 17.615269461077844,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1462,
+      "step": 11767
+    },
+    {
+      "epoch": 17.61676646706587,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 11768
+    },
+    {
+      "epoch": 17.618263473053894,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 11769
+    },
+    {
+      "epoch": 17.619760479041915,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 11770
+    },
+    {
+      "epoch": 17.62125748502994,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1379,
+      "step": 11771
+    },
+    {
+      "epoch": 17.622754491017965,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 11772
+    },
+    {
+      "epoch": 17.62425149700599,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 11773
+    },
+    {
+      "epoch": 17.62574850299401,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1484,
+      "step": 11774
+    },
+    {
+      "epoch": 17.627245508982035,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1365,
+      "step": 11775
+    },
+    {
+      "epoch": 17.62874251497006,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 11776
+    },
+    {
+      "epoch": 17.630239520958085,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 11777
+    },
+    {
+      "epoch": 17.631736526946106,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 11778
+    },
+    {
+      "epoch": 17.63323353293413,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1347,
+      "step": 11779
+    },
+    {
+      "epoch": 17.634730538922156,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 11780
+    },
+    {
+      "epoch": 17.63622754491018,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 11781
+    },
+    {
+      "epoch": 17.637724550898202,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 11782
+    },
+    {
+      "epoch": 17.639221556886227,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1345,
+      "step": 11783
+    },
+    {
+      "epoch": 17.64071856287425,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1316,
+      "step": 11784
+    },
+    {
+      "epoch": 17.642215568862277,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1351,
+      "step": 11785
+    },
+    {
+      "epoch": 17.643712574850298,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1361,
+      "step": 11786
+    },
+    {
+      "epoch": 17.645209580838323,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 11787
+    },
+    {
+      "epoch": 17.646706586826348,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1386,
+      "step": 11788
+    },
+    {
+      "epoch": 17.648203592814372,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 11789
+    },
+    {
+      "epoch": 17.649700598802394,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 11790
+    },
+    {
+      "epoch": 17.65119760479042,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1479,
+      "step": 11791
+    },
+    {
+      "epoch": 17.652694610778443,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 11792
+    },
+    {
+      "epoch": 17.654191616766468,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.136,
+      "step": 11793
+    },
+    {
+      "epoch": 17.65568862275449,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.133,
+      "step": 11794
+    },
+    {
+      "epoch": 17.657185628742514,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1366,
+      "step": 11795
+    },
+    {
+      "epoch": 17.65868263473054,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 11796
+    },
+    {
+      "epoch": 17.660179640718564,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 11797
+    },
+    {
+      "epoch": 17.66167664670659,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1343,
+      "step": 11798
+    },
+    {
+      "epoch": 17.66317365269461,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 11799
+    },
+    {
+      "epoch": 17.664670658682635,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1377,
+      "step": 11800
+    },
+    {
+      "epoch": 17.66616766467066,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 11801
+    },
+    {
+      "epoch": 17.66766467065868,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 11802
+    },
+    {
+      "epoch": 17.669161676646706,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 11803
+    },
+    {
+      "epoch": 17.67065868263473,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1429,
+      "step": 11804
+    },
+    {
+      "epoch": 17.672155688622755,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 11805
+    },
+    {
+      "epoch": 17.67365269461078,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 11806
+    },
+    {
+      "epoch": 17.6751497005988,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1367,
+      "step": 11807
+    },
+    {
+      "epoch": 17.676646706586826,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 11808
+    },
+    {
+      "epoch": 17.67814371257485,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1485,
+      "step": 11809
+    },
+    {
+      "epoch": 17.679640718562876,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1377,
+      "step": 11810
+    },
+    {
+      "epoch": 17.681137724550897,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 11811
+    },
+    {
+      "epoch": 17.682634730538922,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1413,
+      "step": 11812
+    },
+    {
+      "epoch": 17.684131736526947,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 11813
+    },
+    {
+      "epoch": 17.68562874251497,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 11814
+    },
+    {
+      "epoch": 17.687125748502993,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 11815
+    },
+    {
+      "epoch": 17.688622754491018,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1371,
+      "step": 11816
+    },
+    {
+      "epoch": 17.690119760479043,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1413,
+      "step": 11817
+    },
+    {
+      "epoch": 17.691616766467067,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 11818
+    },
+    {
+      "epoch": 17.69311377245509,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1386,
+      "step": 11819
+    },
+    {
+      "epoch": 17.694610778443113,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1365,
+      "step": 11820
+    },
+    {
+      "epoch": 17.69610778443114,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 11821
+    },
+    {
+      "epoch": 17.697604790419163,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1483,
+      "step": 11822
+    },
+    {
+      "epoch": 17.699101796407184,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1472,
+      "step": 11823
+    },
+    {
+      "epoch": 17.70059880239521,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 11824
+    },
+    {
+      "epoch": 17.702095808383234,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 11825
+    },
+    {
+      "epoch": 17.70359281437126,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 11826
+    },
+    {
+      "epoch": 17.70508982035928,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1363,
+      "step": 11827
+    },
+    {
+      "epoch": 17.706586826347305,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 11828
+    },
+    {
+      "epoch": 17.70808383233533,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 11829
+    },
+    {
+      "epoch": 17.709580838323355,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 11830
+    },
+    {
+      "epoch": 17.711077844311376,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 11831
+    },
+    {
+      "epoch": 17.7125748502994,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 11832
+    },
+    {
+      "epoch": 17.714071856287426,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 11833
+    },
+    {
+      "epoch": 17.71556886227545,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 11834
+    },
+    {
+      "epoch": 17.71706586826347,
+      "grad_norm": 0.12109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 11835
+    },
+    {
+      "epoch": 17.718562874251496,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 11836
+    },
+    {
+      "epoch": 17.72005988023952,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1428,
+      "step": 11837
+    },
+    {
+      "epoch": 17.721556886227546,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 11838
+    },
+    {
+      "epoch": 17.723053892215567,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1523,
+      "step": 11839
+    },
+    {
+      "epoch": 17.724550898203592,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1335,
+      "step": 11840
+    },
+    {
+      "epoch": 17.726047904191617,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 11841
+    },
+    {
+      "epoch": 17.727544910179642,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1323,
+      "step": 11842
+    },
+    {
+      "epoch": 17.729041916167663,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1363,
+      "step": 11843
+    },
+    {
+      "epoch": 17.730538922155688,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 11844
+    },
+    {
+      "epoch": 17.732035928143713,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 11845
+    },
+    {
+      "epoch": 17.733532934131738,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 11846
+    },
+    {
+      "epoch": 17.73502994011976,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 11847
+    },
+    {
+      "epoch": 17.736526946107784,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1331,
+      "step": 11848
+    },
+    {
+      "epoch": 17.73802395209581,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 11849
+    },
+    {
+      "epoch": 17.739520958083833,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 11850
+    },
+    {
+      "epoch": 17.741017964071855,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 11851
+    },
+    {
+      "epoch": 17.74251497005988,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1386,
+      "step": 11852
+    },
+    {
+      "epoch": 17.744011976047904,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1383,
+      "step": 11853
+    },
+    {
+      "epoch": 17.74550898203593,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 11854
+    },
+    {
+      "epoch": 17.74700598802395,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 11855
+    },
+    {
+      "epoch": 17.748502994011975,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 11856
+    },
+    {
+      "epoch": 17.75,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 11857
+    },
+    {
+      "epoch": 17.751497005988025,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1376,
+      "step": 11858
+    },
+    {
+      "epoch": 17.75299401197605,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 11859
+    },
+    {
+      "epoch": 17.75449101796407,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1482,
+      "step": 11860
+    },
+    {
+      "epoch": 17.755988023952096,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1351,
+      "step": 11861
+    },
+    {
+      "epoch": 17.75748502994012,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1365,
+      "step": 11862
+    },
+    {
+      "epoch": 17.758982035928145,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1381,
+      "step": 11863
+    },
+    {
+      "epoch": 17.760479041916167,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 11864
+    },
+    {
+      "epoch": 17.76197604790419,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 11865
+    },
+    {
+      "epoch": 17.763473053892216,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1351,
+      "step": 11866
+    },
+    {
+      "epoch": 17.76497005988024,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 11867
+    },
+    {
+      "epoch": 17.766467065868262,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1434,
+      "step": 11868
+    },
+    {
+      "epoch": 17.767964071856287,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 11869
+    },
+    {
+      "epoch": 17.769461077844312,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1479,
+      "step": 11870
+    },
+    {
+      "epoch": 17.770958083832337,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 11871
+    },
+    {
+      "epoch": 17.772455089820358,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 11872
+    },
+    {
+      "epoch": 17.773952095808383,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1383,
+      "step": 11873
+    },
+    {
+      "epoch": 17.775449101796408,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 11874
+    },
+    {
+      "epoch": 17.776946107784433,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1344,
+      "step": 11875
+    },
+    {
+      "epoch": 17.778443113772454,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 11876
+    },
+    {
+      "epoch": 17.77994011976048,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1359,
+      "step": 11877
+    },
+    {
+      "epoch": 17.781437125748504,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1454,
+      "step": 11878
+    },
+    {
+      "epoch": 17.78293413173653,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1428,
+      "step": 11879
+    },
+    {
+      "epoch": 17.78443113772455,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 11880
+    },
+    {
+      "epoch": 17.785928143712574,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1443,
+      "step": 11881
+    },
+    {
+      "epoch": 17.7874251497006,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 11882
+    },
+    {
+      "epoch": 17.788922155688624,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 11883
+    },
+    {
+      "epoch": 17.790419161676645,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.136,
+      "step": 11884
+    },
+    {
+      "epoch": 17.79191616766467,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 11885
+    },
+    {
+      "epoch": 17.793413173652695,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.14,
+      "step": 11886
+    },
+    {
+      "epoch": 17.79491017964072,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1357,
+      "step": 11887
+    },
+    {
+      "epoch": 17.79640718562874,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 11888
+    },
+    {
+      "epoch": 17.797904191616766,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1375,
+      "step": 11889
+    },
+    {
+      "epoch": 17.79940119760479,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 11890
+    },
+    {
+      "epoch": 17.800898203592816,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 11891
+    },
+    {
+      "epoch": 17.802395209580837,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1361,
+      "step": 11892
+    },
+    {
+      "epoch": 17.80389221556886,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1367,
+      "step": 11893
+    },
+    {
+      "epoch": 17.805389221556887,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 11894
+    },
+    {
+      "epoch": 17.80688622754491,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1443,
+      "step": 11895
+    },
+    {
+      "epoch": 17.808383233532933,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 11896
+    },
+    {
+      "epoch": 17.809880239520957,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 11897
+    },
+    {
+      "epoch": 17.811377245508982,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1393,
+      "step": 11898
+    },
+    {
+      "epoch": 17.812874251497007,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.132,
+      "step": 11899
+    },
+    {
+      "epoch": 17.81437125748503,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 11900
+    },
+    {
+      "epoch": 17.815868263473053,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 11901
+    },
+    {
+      "epoch": 17.817365269461078,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 11902
+    },
+    {
+      "epoch": 17.818862275449103,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 11903
+    },
+    {
+      "epoch": 17.820359281437124,
+      "grad_norm": 0.123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 11904
+    },
+    {
+      "epoch": 17.82185628742515,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1504,
+      "step": 11905
+    },
+    {
+      "epoch": 17.823353293413174,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1364,
+      "step": 11906
+    },
+    {
+      "epoch": 17.8248502994012,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 11907
+    },
+    {
+      "epoch": 17.82634730538922,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1454,
+      "step": 11908
+    },
+    {
+      "epoch": 17.827844311377245,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 11909
+    },
+    {
+      "epoch": 17.82934131736527,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1475,
+      "step": 11910
+    },
+    {
+      "epoch": 17.830838323353294,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1416,
+      "step": 11911
+    },
+    {
+      "epoch": 17.83233532934132,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1377,
+      "step": 11912
+    },
+    {
+      "epoch": 17.83383233532934,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1416,
+      "step": 11913
+    },
+    {
+      "epoch": 17.835329341317365,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 11914
+    },
+    {
+      "epoch": 17.83682634730539,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 11915
+    },
+    {
+      "epoch": 17.83832335329341,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1429,
+      "step": 11916
+    },
+    {
+      "epoch": 17.839820359281436,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 11917
+    },
+    {
+      "epoch": 17.84131736526946,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1372,
+      "step": 11918
+    },
+    {
+      "epoch": 17.842814371257486,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 11919
+    },
+    {
+      "epoch": 17.84431137724551,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 11920
+    },
+    {
+      "epoch": 17.845808383233532,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 11921
+    },
+    {
+      "epoch": 17.847305389221557,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 11922
+    },
+    {
+      "epoch": 17.84880239520958,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1345,
+      "step": 11923
+    },
+    {
+      "epoch": 17.850299401197606,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1328,
+      "step": 11924
+    },
+    {
+      "epoch": 17.851796407185628,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1372,
+      "step": 11925
+    },
+    {
+      "epoch": 17.853293413173652,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 11926
+    },
+    {
+      "epoch": 17.854790419161677,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 11927
+    },
+    {
+      "epoch": 17.856287425149702,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1355,
+      "step": 11928
+    },
+    {
+      "epoch": 17.857784431137723,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 11929
+    },
+    {
+      "epoch": 17.85928143712575,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.137,
+      "step": 11930
+    },
+    {
+      "epoch": 17.860778443113773,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 11931
+    },
+    {
+      "epoch": 17.862275449101798,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 11932
+    },
+    {
+      "epoch": 17.86377245508982,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 11933
+    },
+    {
+      "epoch": 17.865269461077844,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1341,
+      "step": 11934
+    },
+    {
+      "epoch": 17.86676646706587,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1477,
+      "step": 11935
+    },
+    {
+      "epoch": 17.868263473053894,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 11936
+    },
+    {
+      "epoch": 17.869760479041915,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1375,
+      "step": 11937
+    },
+    {
+      "epoch": 17.87125748502994,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 11938
+    },
+    {
+      "epoch": 17.872754491017965,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 11939
+    },
+    {
+      "epoch": 17.87425149700599,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1345,
+      "step": 11940
+    },
+    {
+      "epoch": 17.87574850299401,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 11941
+    },
+    {
+      "epoch": 17.877245508982035,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 11942
+    },
+    {
+      "epoch": 17.87874251497006,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 11943
+    },
+    {
+      "epoch": 17.880239520958085,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.151,
+      "step": 11944
+    },
+    {
+      "epoch": 17.881736526946106,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 11945
+    },
+    {
+      "epoch": 17.88323353293413,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1365,
+      "step": 11946
+    },
+    {
+      "epoch": 17.884730538922156,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 11947
+    },
+    {
+      "epoch": 17.88622754491018,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 11948
+    },
+    {
+      "epoch": 17.887724550898202,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1369,
+      "step": 11949
+    },
+    {
+      "epoch": 17.889221556886227,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 11950
+    },
+    {
+      "epoch": 17.89071856287425,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 11951
+    },
+    {
+      "epoch": 17.892215568862277,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 11952
+    },
+    {
+      "epoch": 17.893712574850298,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 11953
+    },
+    {
+      "epoch": 17.895209580838323,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1475,
+      "step": 11954
+    },
+    {
+      "epoch": 17.896706586826348,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 11955
+    },
+    {
+      "epoch": 17.898203592814372,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 11956
+    },
+    {
+      "epoch": 17.899700598802394,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1386,
+      "step": 11957
+    },
+    {
+      "epoch": 17.90119760479042,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1386,
+      "step": 11958
+    },
+    {
+      "epoch": 17.902694610778443,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1453,
+      "step": 11959
+    },
+    {
+      "epoch": 17.904191616766468,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.137,
+      "step": 11960
+    },
+    {
+      "epoch": 17.90568862275449,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 11961
+    },
+    {
+      "epoch": 17.907185628742514,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 11962
+    },
+    {
+      "epoch": 17.90868263473054,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 11963
+    },
+    {
+      "epoch": 17.910179640718564,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1325,
+      "step": 11964
+    },
+    {
+      "epoch": 17.91167664670659,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 11965
+    },
+    {
+      "epoch": 17.91317365269461,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1376,
+      "step": 11966
+    },
+    {
+      "epoch": 17.914670658682635,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 11967
+    },
+    {
+      "epoch": 17.91616766467066,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 11968
+    },
+    {
+      "epoch": 17.91766467065868,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.144,
+      "step": 11969
+    },
+    {
+      "epoch": 17.919161676646706,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 11970
+    },
+    {
+      "epoch": 17.92065868263473,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1355,
+      "step": 11971
+    },
+    {
+      "epoch": 17.922155688622755,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 11972
+    },
+    {
+      "epoch": 17.92365269461078,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 11973
+    },
+    {
+      "epoch": 17.9251497005988,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 11974
+    },
+    {
+      "epoch": 17.926646706586826,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1364,
+      "step": 11975
+    },
+    {
+      "epoch": 17.92814371257485,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 11976
+    },
+    {
+      "epoch": 17.929640718562876,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 11977
+    },
+    {
+      "epoch": 17.931137724550897,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1497,
+      "step": 11978
+    },
+    {
+      "epoch": 17.932634730538922,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1381,
+      "step": 11979
+    },
+    {
+      "epoch": 17.934131736526947,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 11980
+    },
+    {
+      "epoch": 17.93562874251497,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1367,
+      "step": 11981
+    },
+    {
+      "epoch": 17.937125748502993,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 11982
+    },
+    {
+      "epoch": 17.938622754491018,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1413,
+      "step": 11983
+    },
+    {
+      "epoch": 17.940119760479043,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 11984
+    },
+    {
+      "epoch": 17.941616766467067,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 11985
+    },
+    {
+      "epoch": 17.94311377245509,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1368,
+      "step": 11986
+    },
+    {
+      "epoch": 17.944610778443113,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1354,
+      "step": 11987
+    },
+    {
+      "epoch": 17.94610778443114,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 11988
+    },
+    {
+      "epoch": 17.947604790419163,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1361,
+      "step": 11989
+    },
+    {
+      "epoch": 17.949101796407184,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 11990
+    },
+    {
+      "epoch": 17.95059880239521,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1428,
+      "step": 11991
+    },
+    {
+      "epoch": 17.952095808383234,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1367,
+      "step": 11992
+    },
+    {
+      "epoch": 17.95359281437126,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1364,
+      "step": 11993
+    },
+    {
+      "epoch": 17.95508982035928,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 11994
+    },
+    {
+      "epoch": 17.956586826347305,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 11995
+    },
+    {
+      "epoch": 17.95808383233533,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0008,
+      "loss": 1.147,
+      "step": 11996
+    },
+    {
+      "epoch": 17.959580838323355,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 11997
+    },
+    {
+      "epoch": 17.961077844311376,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 11998
+    },
+    {
+      "epoch": 17.9625748502994,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 11999
+    },
+    {
+      "epoch": 17.964071856287426,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 12000
+    },
+    {
+      "epoch": 17.96556886227545,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1469,
+      "step": 12001
+    },
+    {
+      "epoch": 17.96706586826347,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1497,
+      "step": 12002
+    },
+    {
+      "epoch": 17.968562874251496,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1484,
+      "step": 12003
+    },
+    {
+      "epoch": 17.97005988023952,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1496,
+      "step": 12004
+    },
+    {
+      "epoch": 17.971556886227546,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1488,
+      "step": 12005
+    },
+    {
+      "epoch": 17.973053892215567,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 12006
+    },
+    {
+      "epoch": 17.974550898203592,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1473,
+      "step": 12007
+    },
+    {
+      "epoch": 17.976047904191617,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 12008
+    },
+    {
+      "epoch": 17.977544910179642,
+      "grad_norm": 0.2265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 12009
+    },
+    {
+      "epoch": 17.979041916167663,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 12010
+    },
+    {
+      "epoch": 17.980538922155688,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1466,
+      "step": 12011
+    },
+    {
+      "epoch": 17.982035928143713,
+      "grad_norm": 0.201171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 12012
+    },
+    {
+      "epoch": 17.983532934131738,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1367,
+      "step": 12013
+    },
+    {
+      "epoch": 17.98502994011976,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 12014
+    },
+    {
+      "epoch": 17.986526946107784,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 12015
+    },
+    {
+      "epoch": 17.98802395209581,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 12016
+    },
+    {
+      "epoch": 17.989520958083833,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 12017
+    },
+    {
+      "epoch": 17.991017964071855,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 12018
+    },
+    {
+      "epoch": 17.99251497005988,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1355,
+      "step": 12019
+    },
+    {
+      "epoch": 17.994011976047904,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 12020
+    },
+    {
+      "epoch": 17.99550898203593,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1451,
+      "step": 12021
+    },
+    {
+      "epoch": 17.99700598802395,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1428,
+      "step": 12022
+    },
+    {
+      "epoch": 17.998502994011975,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 12023
+    },
+    {
+      "epoch": 18.0,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 12024
+    },
+    {
+      "epoch": 18.001497005988025,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1368,
+      "step": 12025
+    },
+    {
+      "epoch": 18.00299401197605,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1372,
+      "step": 12026
+    },
+    {
+      "epoch": 18.00449101796407,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1339,
+      "step": 12027
+    },
+    {
+      "epoch": 18.005988023952096,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 12028
+    },
+    {
+      "epoch": 18.00748502994012,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 12029
+    },
+    {
+      "epoch": 18.008982035928145,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1355,
+      "step": 12030
+    },
+    {
+      "epoch": 18.010479041916167,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 12031
+    },
+    {
+      "epoch": 18.01197604790419,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.14,
+      "step": 12032
+    },
+    {
+      "epoch": 18.013473053892216,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1356,
+      "step": 12033
+    },
+    {
+      "epoch": 18.01497005988024,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 12034
+    },
+    {
+      "epoch": 18.016467065868262,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 12035
+    },
+    {
+      "epoch": 18.017964071856287,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1364,
+      "step": 12036
+    },
+    {
+      "epoch": 18.019461077844312,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 12037
+    },
+    {
+      "epoch": 18.020958083832337,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1366,
+      "step": 12038
+    },
+    {
+      "epoch": 18.022455089820358,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 12039
+    },
+    {
+      "epoch": 18.023952095808383,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 12040
+    },
+    {
+      "epoch": 18.025449101796408,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.137,
+      "step": 12041
+    },
+    {
+      "epoch": 18.026946107784433,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 12042
+    },
+    {
+      "epoch": 18.028443113772454,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1476,
+      "step": 12043
+    },
+    {
+      "epoch": 18.02994011976048,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1429,
+      "step": 12044
+    },
+    {
+      "epoch": 18.031437125748504,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1356,
+      "step": 12045
+    },
+    {
+      "epoch": 18.03293413173653,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1427,
+      "step": 12046
+    },
+    {
+      "epoch": 18.03443113772455,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1393,
+      "step": 12047
+    },
+    {
+      "epoch": 18.035928143712574,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1379,
+      "step": 12048
+    },
+    {
+      "epoch": 18.0374251497006,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1355,
+      "step": 12049
+    },
+    {
+      "epoch": 18.038922155688624,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1358,
+      "step": 12050
+    },
+    {
+      "epoch": 18.040419161676645,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 12051
+    },
+    {
+      "epoch": 18.04191616766467,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 12052
+    },
+    {
+      "epoch": 18.043413173652695,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 12053
+    },
+    {
+      "epoch": 18.04491017964072,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 12054
+    },
+    {
+      "epoch": 18.04640718562874,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 12055
+    },
+    {
+      "epoch": 18.047904191616766,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 12056
+    },
+    {
+      "epoch": 18.04940119760479,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 12057
+    },
+    {
+      "epoch": 18.050898203592816,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1454,
+      "step": 12058
+    },
+    {
+      "epoch": 18.052395209580837,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 12059
+    },
+    {
+      "epoch": 18.05389221556886,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 12060
+    },
+    {
+      "epoch": 18.055389221556887,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 12061
+    },
+    {
+      "epoch": 18.05688622754491,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 12062
+    },
+    {
+      "epoch": 18.058383233532933,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 12063
+    },
+    {
+      "epoch": 18.059880239520957,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1348,
+      "step": 12064
+    },
+    {
+      "epoch": 18.061377245508982,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1368,
+      "step": 12065
+    },
+    {
+      "epoch": 18.062874251497007,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 12066
+    },
+    {
+      "epoch": 18.06437125748503,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 12067
+    },
+    {
+      "epoch": 18.065868263473053,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1358,
+      "step": 12068
+    },
+    {
+      "epoch": 18.067365269461078,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 12069
+    },
+    {
+      "epoch": 18.068862275449103,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1376,
+      "step": 12070
+    },
+    {
+      "epoch": 18.070359281437124,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1371,
+      "step": 12071
+    },
+    {
+      "epoch": 18.07185628742515,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1449,
+      "step": 12072
+    },
+    {
+      "epoch": 18.073353293413174,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1457,
+      "step": 12073
+    },
+    {
+      "epoch": 18.0748502994012,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1335,
+      "step": 12074
+    },
+    {
+      "epoch": 18.07634730538922,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1328,
+      "step": 12075
+    },
+    {
+      "epoch": 18.077844311377245,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1344,
+      "step": 12076
+    },
+    {
+      "epoch": 18.07934131736527,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 12077
+    },
+    {
+      "epoch": 18.080838323353294,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1464,
+      "step": 12078
+    },
+    {
+      "epoch": 18.082335329341316,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 12079
+    },
+    {
+      "epoch": 18.08383233532934,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 12080
+    },
+    {
+      "epoch": 18.085329341317365,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 12081
+    },
+    {
+      "epoch": 18.08682634730539,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 12082
+    },
+    {
+      "epoch": 18.088323353293415,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 12083
+    },
+    {
+      "epoch": 18.089820359281436,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 12084
+    },
+    {
+      "epoch": 18.09131736526946,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1364,
+      "step": 12085
+    },
+    {
+      "epoch": 18.092814371257486,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 12086
+    },
+    {
+      "epoch": 18.09431137724551,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 12087
+    },
+    {
+      "epoch": 18.095808383233532,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1455,
+      "step": 12088
+    },
+    {
+      "epoch": 18.097305389221557,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 12089
+    },
+    {
+      "epoch": 18.09880239520958,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1504,
+      "step": 12090
+    },
+    {
+      "epoch": 18.100299401197606,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1392,
+      "step": 12091
+    },
+    {
+      "epoch": 18.101796407185628,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1395,
+      "step": 12092
+    },
+    {
+      "epoch": 18.103293413173652,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1332,
+      "step": 12093
+    },
+    {
+      "epoch": 18.104790419161677,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1369,
+      "step": 12094
+    },
+    {
+      "epoch": 18.106287425149702,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 12095
+    },
+    {
+      "epoch": 18.107784431137723,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1369,
+      "step": 12096
+    },
+    {
+      "epoch": 18.10928143712575,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1362,
+      "step": 12097
+    },
+    {
+      "epoch": 18.110778443113773,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1383,
+      "step": 12098
+    },
+    {
+      "epoch": 18.112275449101798,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1367,
+      "step": 12099
+    },
+    {
+      "epoch": 18.11377245508982,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 12100
+    },
+    {
+      "epoch": 18.115269461077844,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1378,
+      "step": 12101
+    },
+    {
+      "epoch": 18.11676646706587,
+      "grad_norm": 0.109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1348,
+      "step": 12102
+    },
+    {
+      "epoch": 18.118263473053894,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1358,
+      "step": 12103
+    },
+    {
+      "epoch": 18.119760479041915,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 12104
+    },
+    {
+      "epoch": 18.12125748502994,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 12105
+    },
+    {
+      "epoch": 18.122754491017965,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.134,
+      "step": 12106
+    },
+    {
+      "epoch": 18.12425149700599,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1381,
+      "step": 12107
+    },
+    {
+      "epoch": 18.12574850299401,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 12108
+    },
+    {
+      "epoch": 18.127245508982035,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1359,
+      "step": 12109
+    },
+    {
+      "epoch": 18.12874251497006,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1359,
+      "step": 12110
+    },
+    {
+      "epoch": 18.130239520958085,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.14,
+      "step": 12111
+    },
+    {
+      "epoch": 18.131736526946106,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 12112
+    },
+    {
+      "epoch": 18.13323353293413,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.14,
+      "step": 12113
+    },
+    {
+      "epoch": 18.134730538922156,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 12114
+    },
+    {
+      "epoch": 18.13622754491018,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 12115
+    },
+    {
+      "epoch": 18.137724550898202,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 12116
+    },
+    {
+      "epoch": 18.139221556886227,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 12117
+    },
+    {
+      "epoch": 18.14071856287425,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 12118
+    },
+    {
+      "epoch": 18.142215568862277,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.136,
+      "step": 12119
+    },
+    {
+      "epoch": 18.143712574850298,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 12120
+    },
+    {
+      "epoch": 18.145209580838323,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 12121
+    },
+    {
+      "epoch": 18.146706586826348,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 12122
+    },
+    {
+      "epoch": 18.148203592814372,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1395,
+      "step": 12123
+    },
+    {
+      "epoch": 18.149700598802394,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1356,
+      "step": 12124
+    },
+    {
+      "epoch": 18.15119760479042,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1362,
+      "step": 12125
+    },
+    {
+      "epoch": 18.152694610778443,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.137,
+      "step": 12126
+    },
+    {
+      "epoch": 18.154191616766468,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1366,
+      "step": 12127
+    },
+    {
+      "epoch": 18.15568862275449,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 12128
+    },
+    {
+      "epoch": 18.157185628742514,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1381,
+      "step": 12129
+    },
+    {
+      "epoch": 18.15868263473054,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1337,
+      "step": 12130
+    },
+    {
+      "epoch": 18.160179640718564,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 12131
+    },
+    {
+      "epoch": 18.161676646706585,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.137,
+      "step": 12132
+    },
+    {
+      "epoch": 18.16317365269461,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1381,
+      "step": 12133
+    },
+    {
+      "epoch": 18.164670658682635,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1362,
+      "step": 12134
+    },
+    {
+      "epoch": 18.16616766467066,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1467,
+      "step": 12135
+    },
+    {
+      "epoch": 18.16766467065868,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 12136
+    },
+    {
+      "epoch": 18.169161676646706,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1384,
+      "step": 12137
+    },
+    {
+      "epoch": 18.17065868263473,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1383,
+      "step": 12138
+    },
+    {
+      "epoch": 18.172155688622755,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1386,
+      "step": 12139
+    },
+    {
+      "epoch": 18.17365269461078,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1381,
+      "step": 12140
+    },
+    {
+      "epoch": 18.1751497005988,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1378,
+      "step": 12141
+    },
+    {
+      "epoch": 18.176646706586826,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 12142
+    },
+    {
+      "epoch": 18.17814371257485,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 12143
+    },
+    {
+      "epoch": 18.179640718562876,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 12144
+    },
+    {
+      "epoch": 18.181137724550897,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 12145
+    },
+    {
+      "epoch": 18.182634730538922,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 12146
+    },
+    {
+      "epoch": 18.184131736526947,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 12147
+    },
+    {
+      "epoch": 18.18562874251497,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1359,
+      "step": 12148
+    },
+    {
+      "epoch": 18.187125748502993,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1369,
+      "step": 12149
+    },
+    {
+      "epoch": 18.188622754491018,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1357,
+      "step": 12150
+    },
+    {
+      "epoch": 18.190119760479043,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1323,
+      "step": 12151
+    },
+    {
+      "epoch": 18.191616766467067,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 12152
+    },
+    {
+      "epoch": 18.19311377245509,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 12153
+    },
+    {
+      "epoch": 18.194610778443113,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 12154
+    },
+    {
+      "epoch": 18.19610778443114,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1355,
+      "step": 12155
+    },
+    {
+      "epoch": 18.197604790419163,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 12156
+    },
+    {
+      "epoch": 18.199101796407184,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 12157
+    },
+    {
+      "epoch": 18.20059880239521,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 12158
+    },
+    {
+      "epoch": 18.202095808383234,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1346,
+      "step": 12159
+    },
+    {
+      "epoch": 18.20359281437126,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 12160
+    },
+    {
+      "epoch": 18.20508982035928,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 12161
+    },
+    {
+      "epoch": 18.206586826347305,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1482,
+      "step": 12162
+    },
+    {
+      "epoch": 18.20808383233533,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 12163
+    },
+    {
+      "epoch": 18.209580838323355,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1387,
+      "step": 12164
+    },
+    {
+      "epoch": 18.211077844311376,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 12165
+    },
+    {
+      "epoch": 18.2125748502994,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 12166
+    },
+    {
+      "epoch": 18.214071856287426,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1341,
+      "step": 12167
+    },
+    {
+      "epoch": 18.21556886227545,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1302,
+      "step": 12168
+    },
+    {
+      "epoch": 18.21706586826347,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 12169
+    },
+    {
+      "epoch": 18.218562874251496,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1416,
+      "step": 12170
+    },
+    {
+      "epoch": 18.22005988023952,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 12171
+    },
+    {
+      "epoch": 18.221556886227546,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 12172
+    },
+    {
+      "epoch": 18.223053892215567,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.149,
+      "step": 12173
+    },
+    {
+      "epoch": 18.224550898203592,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 12174
+    },
+    {
+      "epoch": 18.226047904191617,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 12175
+    },
+    {
+      "epoch": 18.227544910179642,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 12176
+    },
+    {
+      "epoch": 18.229041916167663,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 12177
+    },
+    {
+      "epoch": 18.230538922155688,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 12178
+    },
+    {
+      "epoch": 18.232035928143713,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.137,
+      "step": 12179
+    },
+    {
+      "epoch": 18.233532934131738,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 12180
+    },
+    {
+      "epoch": 18.23502994011976,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1486,
+      "step": 12181
+    },
+    {
+      "epoch": 18.236526946107784,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1332,
+      "step": 12182
+    },
+    {
+      "epoch": 18.23802395209581,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 12183
+    },
+    {
+      "epoch": 18.239520958083833,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 12184
+    },
+    {
+      "epoch": 18.241017964071855,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1369,
+      "step": 12185
+    },
+    {
+      "epoch": 18.24251497005988,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 12186
+    },
+    {
+      "epoch": 18.244011976047904,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.14,
+      "step": 12187
+    },
+    {
+      "epoch": 18.24550898203593,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 12188
+    },
+    {
+      "epoch": 18.24700598802395,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 12189
+    },
+    {
+      "epoch": 18.248502994011975,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1315,
+      "step": 12190
+    },
+    {
+      "epoch": 18.25,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 12191
+    },
+    {
+      "epoch": 18.251497005988025,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.136,
+      "step": 12192
+    },
+    {
+      "epoch": 18.25299401197605,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1454,
+      "step": 12193
+    },
+    {
+      "epoch": 18.25449101796407,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1307,
+      "step": 12194
+    },
+    {
+      "epoch": 18.255988023952096,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1358,
+      "step": 12195
+    },
+    {
+      "epoch": 18.25748502994012,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 12196
+    },
+    {
+      "epoch": 18.258982035928145,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 12197
+    },
+    {
+      "epoch": 18.260479041916167,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 12198
+    },
+    {
+      "epoch": 18.26197604790419,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1491,
+      "step": 12199
+    },
+    {
+      "epoch": 18.263473053892216,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 12200
+    },
+    {
+      "epoch": 18.26497005988024,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 12201
+    },
+    {
+      "epoch": 18.266467065868262,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 12202
+    },
+    {
+      "epoch": 18.267964071856287,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1387,
+      "step": 12203
+    },
+    {
+      "epoch": 18.269461077844312,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 12204
+    },
+    {
+      "epoch": 18.270958083832337,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 12205
+    },
+    {
+      "epoch": 18.272455089820358,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1366,
+      "step": 12206
+    },
+    {
+      "epoch": 18.273952095808383,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1428,
+      "step": 12207
+    },
+    {
+      "epoch": 18.275449101796408,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 12208
+    },
+    {
+      "epoch": 18.276946107784433,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1429,
+      "step": 12209
+    },
+    {
+      "epoch": 18.278443113772454,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 12210
+    },
+    {
+      "epoch": 18.27994011976048,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1305,
+      "step": 12211
+    },
+    {
+      "epoch": 18.281437125748504,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1359,
+      "step": 12212
+    },
+    {
+      "epoch": 18.28293413173653,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1344,
+      "step": 12213
+    },
+    {
+      "epoch": 18.28443113772455,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 12214
+    },
+    {
+      "epoch": 18.285928143712574,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1363,
+      "step": 12215
+    },
+    {
+      "epoch": 18.2874251497006,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1416,
+      "step": 12216
+    },
+    {
+      "epoch": 18.288922155688624,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 12217
+    },
+    {
+      "epoch": 18.290419161676645,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1369,
+      "step": 12218
+    },
+    {
+      "epoch": 18.29191616766467,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1392,
+      "step": 12219
+    },
+    {
+      "epoch": 18.293413173652695,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 12220
+    },
+    {
+      "epoch": 18.29491017964072,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 12221
+    },
+    {
+      "epoch": 18.29640718562874,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 12222
+    },
+    {
+      "epoch": 18.297904191616766,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 12223
+    },
+    {
+      "epoch": 18.29940119760479,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1375,
+      "step": 12224
+    },
+    {
+      "epoch": 18.300898203592816,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1429,
+      "step": 12225
+    },
+    {
+      "epoch": 18.302395209580837,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1361,
+      "step": 12226
+    },
+    {
+      "epoch": 18.30389221556886,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1462,
+      "step": 12227
+    },
+    {
+      "epoch": 18.305389221556887,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.137,
+      "step": 12228
+    },
+    {
+      "epoch": 18.30688622754491,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 12229
+    },
+    {
+      "epoch": 18.308383233532933,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1348,
+      "step": 12230
+    },
+    {
+      "epoch": 18.309880239520957,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1347,
+      "step": 12231
+    },
+    {
+      "epoch": 18.311377245508982,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1361,
+      "step": 12232
+    },
+    {
+      "epoch": 18.312874251497007,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 12233
+    },
+    {
+      "epoch": 18.31437125748503,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1372,
+      "step": 12234
+    },
+    {
+      "epoch": 18.315868263473053,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 12235
+    },
+    {
+      "epoch": 18.317365269461078,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1355,
+      "step": 12236
+    },
+    {
+      "epoch": 18.318862275449103,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1356,
+      "step": 12237
+    },
+    {
+      "epoch": 18.320359281437124,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 12238
+    },
+    {
+      "epoch": 18.32185628742515,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 12239
+    },
+    {
+      "epoch": 18.323353293413174,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1376,
+      "step": 12240
+    },
+    {
+      "epoch": 18.3248502994012,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 12241
+    },
+    {
+      "epoch": 18.32634730538922,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 12242
+    },
+    {
+      "epoch": 18.327844311377245,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.136,
+      "step": 12243
+    },
+    {
+      "epoch": 18.32934131736527,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.146,
+      "step": 12244
+    },
+    {
+      "epoch": 18.330838323353294,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 12245
+    },
+    {
+      "epoch": 18.33233532934132,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1413,
+      "step": 12246
+    },
+    {
+      "epoch": 18.33383233532934,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1393,
+      "step": 12247
+    },
+    {
+      "epoch": 18.335329341317365,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1331,
+      "step": 12248
+    },
+    {
+      "epoch": 18.33682634730539,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 12249
+    },
+    {
+      "epoch": 18.338323353293415,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 12250
+    },
+    {
+      "epoch": 18.339820359281436,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 12251
+    },
+    {
+      "epoch": 18.34131736526946,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1368,
+      "step": 12252
+    },
+    {
+      "epoch": 18.342814371257486,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1351,
+      "step": 12253
+    },
+    {
+      "epoch": 18.34431137724551,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1359,
+      "step": 12254
+    },
+    {
+      "epoch": 18.345808383233532,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1381,
+      "step": 12255
+    },
+    {
+      "epoch": 18.347305389221557,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 12256
+    },
+    {
+      "epoch": 18.34880239520958,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 12257
+    },
+    {
+      "epoch": 18.350299401197606,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1333,
+      "step": 12258
+    },
+    {
+      "epoch": 18.351796407185628,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1356,
+      "step": 12259
+    },
+    {
+      "epoch": 18.353293413173652,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 12260
+    },
+    {
+      "epoch": 18.354790419161677,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 12261
+    },
+    {
+      "epoch": 18.356287425149702,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1344,
+      "step": 12262
+    },
+    {
+      "epoch": 18.357784431137723,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 12263
+    },
+    {
+      "epoch": 18.35928143712575,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 12264
+    },
+    {
+      "epoch": 18.360778443113773,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1372,
+      "step": 12265
+    },
+    {
+      "epoch": 18.362275449101798,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 12266
+    },
+    {
+      "epoch": 18.36377245508982,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 12267
+    },
+    {
+      "epoch": 18.365269461077844,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 12268
+    },
+    {
+      "epoch": 18.36676646706587,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1353,
+      "step": 12269
+    },
+    {
+      "epoch": 18.368263473053894,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1341,
+      "step": 12270
+    },
+    {
+      "epoch": 18.369760479041915,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1368,
+      "step": 12271
+    },
+    {
+      "epoch": 18.37125748502994,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 12272
+    },
+    {
+      "epoch": 18.372754491017965,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1442,
+      "step": 12273
+    },
+    {
+      "epoch": 18.37425149700599,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.132,
+      "step": 12274
+    },
+    {
+      "epoch": 18.37574850299401,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1354,
+      "step": 12275
+    },
+    {
+      "epoch": 18.377245508982035,
+      "grad_norm": 0.10546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1351,
+      "step": 12276
+    },
+    {
+      "epoch": 18.37874251497006,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1424,
+      "step": 12277
+    },
+    {
+      "epoch": 18.380239520958085,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 12278
+    },
+    {
+      "epoch": 18.381736526946106,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1345,
+      "step": 12279
+    },
+    {
+      "epoch": 18.38323353293413,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 12280
+    },
+    {
+      "epoch": 18.384730538922156,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1347,
+      "step": 12281
+    },
+    {
+      "epoch": 18.38622754491018,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 12282
+    },
+    {
+      "epoch": 18.387724550898202,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 12283
+    },
+    {
+      "epoch": 18.389221556886227,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1306,
+      "step": 12284
+    },
+    {
+      "epoch": 18.39071856287425,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 12285
+    },
+    {
+      "epoch": 18.392215568862277,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 12286
+    },
+    {
+      "epoch": 18.393712574850298,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1379,
+      "step": 12287
+    },
+    {
+      "epoch": 18.395209580838323,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1376,
+      "step": 12288
+    },
+    {
+      "epoch": 18.396706586826348,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 12289
+    },
+    {
+      "epoch": 18.398203592814372,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 12290
+    },
+    {
+      "epoch": 18.399700598802394,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 12291
+    },
+    {
+      "epoch": 18.40119760479042,
+      "grad_norm": 0.130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 12292
+    },
+    {
+      "epoch": 18.402694610778443,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 12293
+    },
+    {
+      "epoch": 18.404191616766468,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 12294
+    },
+    {
+      "epoch": 18.40568862275449,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 12295
+    },
+    {
+      "epoch": 18.407185628742514,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1342,
+      "step": 12296
+    },
+    {
+      "epoch": 18.40868263473054,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 12297
+    },
+    {
+      "epoch": 18.410179640718564,
+      "grad_norm": 0.1328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 12298
+    },
+    {
+      "epoch": 18.411676646706585,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 12299
+    },
+    {
+      "epoch": 18.41317365269461,
+      "grad_norm": 0.111328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1361,
+      "step": 12300
+    },
+    {
+      "epoch": 18.414670658682635,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1332,
+      "step": 12301
+    },
+    {
+      "epoch": 18.41616766467066,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1384,
+      "step": 12302
+    },
+    {
+      "epoch": 18.41766467065868,
+      "grad_norm": 0.107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 12303
+    },
+    {
+      "epoch": 18.419161676646706,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 0.0008,
+      "loss": 1.136,
+      "step": 12304
+    },
+    {
+      "epoch": 18.42065868263473,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 12305
+    },
+    {
+      "epoch": 18.422155688622755,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 12306
+    },
+    {
+      "epoch": 18.42365269461078,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1339,
+      "step": 12307
+    },
+    {
+      "epoch": 18.4251497005988,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1395,
+      "step": 12308
+    },
+    {
+      "epoch": 18.426646706586826,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 12309
+    },
+    {
+      "epoch": 18.42814371257485,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1365,
+      "step": 12310
+    },
+    {
+      "epoch": 18.429640718562876,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 12311
+    },
+    {
+      "epoch": 18.431137724550897,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1375,
+      "step": 12312
+    },
+    {
+      "epoch": 18.432634730538922,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1393,
+      "step": 12313
+    },
+    {
+      "epoch": 18.434131736526947,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 12314
+    },
+    {
+      "epoch": 18.43562874251497,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1334,
+      "step": 12315
+    },
+    {
+      "epoch": 18.437125748502993,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1314,
+      "step": 12316
+    },
+    {
+      "epoch": 18.438622754491018,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 12317
+    },
+    {
+      "epoch": 18.440119760479043,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 12318
+    },
+    {
+      "epoch": 18.441616766467067,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1376,
+      "step": 12319
+    },
+    {
+      "epoch": 18.44311377245509,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.14,
+      "step": 12320
+    },
+    {
+      "epoch": 18.444610778443113,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1371,
+      "step": 12321
+    },
+    {
+      "epoch": 18.44610778443114,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1313,
+      "step": 12322
+    },
+    {
+      "epoch": 18.447604790419163,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1365,
+      "step": 12323
+    },
+    {
+      "epoch": 18.449101796407184,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1386,
+      "step": 12324
+    },
+    {
+      "epoch": 18.45059880239521,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1348,
+      "step": 12325
+    },
+    {
+      "epoch": 18.452095808383234,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1323,
+      "step": 12326
+    },
+    {
+      "epoch": 18.45359281437126,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 12327
+    },
+    {
+      "epoch": 18.45508982035928,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1334,
+      "step": 12328
+    },
+    {
+      "epoch": 18.456586826347305,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1384,
+      "step": 12329
+    },
+    {
+      "epoch": 18.45808383233533,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1387,
+      "step": 12330
+    },
+    {
+      "epoch": 18.459580838323355,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1429,
+      "step": 12331
+    },
+    {
+      "epoch": 18.461077844311376,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1315,
+      "step": 12332
+    },
+    {
+      "epoch": 18.4625748502994,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.133,
+      "step": 12333
+    },
+    {
+      "epoch": 18.464071856287426,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.137,
+      "step": 12334
+    },
+    {
+      "epoch": 18.46556886227545,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 12335
+    },
+    {
+      "epoch": 18.46706586826347,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 12336
+    },
+    {
+      "epoch": 18.468562874251496,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1393,
+      "step": 12337
+    },
+    {
+      "epoch": 18.47005988023952,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1481,
+      "step": 12338
+    },
+    {
+      "epoch": 18.471556886227546,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1356,
+      "step": 12339
+    },
+    {
+      "epoch": 18.473053892215567,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1347,
+      "step": 12340
+    },
+    {
+      "epoch": 18.474550898203592,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1446,
+      "step": 12341
+    },
+    {
+      "epoch": 18.476047904191617,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 12342
+    },
+    {
+      "epoch": 18.477544910179642,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1351,
+      "step": 12343
+    },
+    {
+      "epoch": 18.479041916167663,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1333,
+      "step": 12344
+    },
+    {
+      "epoch": 18.480538922155688,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 12345
+    },
+    {
+      "epoch": 18.482035928143713,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1368,
+      "step": 12346
+    },
+    {
+      "epoch": 18.483532934131738,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1384,
+      "step": 12347
+    },
+    {
+      "epoch": 18.48502994011976,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 12348
+    },
+    {
+      "epoch": 18.486526946107784,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1377,
+      "step": 12349
+    },
+    {
+      "epoch": 18.48802395209581,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1351,
+      "step": 12350
+    },
+    {
+      "epoch": 18.489520958083833,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1387,
+      "step": 12351
+    },
+    {
+      "epoch": 18.491017964071855,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 12352
+    },
+    {
+      "epoch": 18.49251497005988,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 12353
+    },
+    {
+      "epoch": 18.494011976047904,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 12354
+    },
+    {
+      "epoch": 18.49550898203593,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.137,
+      "step": 12355
+    },
+    {
+      "epoch": 18.49700598802395,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 12356
+    },
+    {
+      "epoch": 18.498502994011975,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 12357
+    },
+    {
+      "epoch": 18.5,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1447,
+      "step": 12358
+    },
+    {
+      "epoch": 18.501497005988025,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 12359
+    },
+    {
+      "epoch": 18.50299401197605,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1506,
+      "step": 12360
+    },
+    {
+      "epoch": 18.50449101796407,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1373,
+      "step": 12361
+    },
+    {
+      "epoch": 18.505988023952096,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1368,
+      "step": 12362
+    },
+    {
+      "epoch": 18.50748502994012,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 12363
+    },
+    {
+      "epoch": 18.508982035928145,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 12364
+    },
+    {
+      "epoch": 18.510479041916167,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.134,
+      "step": 12365
+    },
+    {
+      "epoch": 18.51197604790419,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 12366
+    },
+    {
+      "epoch": 18.513473053892216,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 12367
+    },
+    {
+      "epoch": 18.51497005988024,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1355,
+      "step": 12368
+    },
+    {
+      "epoch": 18.516467065868262,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 12369
+    },
+    {
+      "epoch": 18.517964071856287,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 12370
+    },
+    {
+      "epoch": 18.519461077844312,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1357,
+      "step": 12371
+    },
+    {
+      "epoch": 18.520958083832337,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 12372
+    },
+    {
+      "epoch": 18.522455089820358,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 12373
+    },
+    {
+      "epoch": 18.523952095808383,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1357,
+      "step": 12374
+    },
+    {
+      "epoch": 18.525449101796408,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 12375
+    },
+    {
+      "epoch": 18.526946107784433,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 12376
+    },
+    {
+      "epoch": 18.528443113772454,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1371,
+      "step": 12377
+    },
+    {
+      "epoch": 18.52994011976048,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1381,
+      "step": 12378
+    },
+    {
+      "epoch": 18.531437125748504,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1358,
+      "step": 12379
+    },
+    {
+      "epoch": 18.53293413173653,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1362,
+      "step": 12380
+    },
+    {
+      "epoch": 18.53443113772455,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.14,
+      "step": 12381
+    },
+    {
+      "epoch": 18.535928143712574,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1373,
+      "step": 12382
+    },
+    {
+      "epoch": 18.5374251497006,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 12383
+    },
+    {
+      "epoch": 18.538922155688624,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1377,
+      "step": 12384
+    },
+    {
+      "epoch": 18.540419161676645,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1343,
+      "step": 12385
+    },
+    {
+      "epoch": 18.54191616766467,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 12386
+    },
+    {
+      "epoch": 18.543413173652695,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.136,
+      "step": 12387
+    },
+    {
+      "epoch": 18.54491017964072,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1373,
+      "step": 12388
+    },
+    {
+      "epoch": 18.54640718562874,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1329,
+      "step": 12389
+    },
+    {
+      "epoch": 18.547904191616766,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 12390
+    },
+    {
+      "epoch": 18.54940119760479,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1395,
+      "step": 12391
+    },
+    {
+      "epoch": 18.550898203592816,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1347,
+      "step": 12392
+    },
+    {
+      "epoch": 18.552395209580837,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.137,
+      "step": 12393
+    },
+    {
+      "epoch": 18.55389221556886,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1462,
+      "step": 12394
+    },
+    {
+      "epoch": 18.555389221556887,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 12395
+    },
+    {
+      "epoch": 18.55688622754491,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1419,
+      "step": 12396
+    },
+    {
+      "epoch": 18.558383233532933,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1436,
+      "step": 12397
+    },
+    {
+      "epoch": 18.559880239520957,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 12398
+    },
+    {
+      "epoch": 18.561377245508982,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1377,
+      "step": 12399
+    },
+    {
+      "epoch": 18.562874251497007,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1364,
+      "step": 12400
+    },
+    {
+      "epoch": 18.56437125748503,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1379,
+      "step": 12401
+    },
+    {
+      "epoch": 18.565868263473053,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 12402
+    },
+    {
+      "epoch": 18.567365269461078,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 12403
+    },
+    {
+      "epoch": 18.568862275449103,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1378,
+      "step": 12404
+    },
+    {
+      "epoch": 18.570359281437124,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1354,
+      "step": 12405
+    },
+    {
+      "epoch": 18.57185628742515,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1381,
+      "step": 12406
+    },
+    {
+      "epoch": 18.573353293413174,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1465,
+      "step": 12407
+    },
+    {
+      "epoch": 18.5748502994012,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1293,
+      "step": 12408
+    },
+    {
+      "epoch": 18.57634730538922,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.137,
+      "step": 12409
+    },
+    {
+      "epoch": 18.577844311377245,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 12410
+    },
+    {
+      "epoch": 18.57934131736527,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1351,
+      "step": 12411
+    },
+    {
+      "epoch": 18.580838323353294,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 12412
+    },
+    {
+      "epoch": 18.58233532934132,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 12413
+    },
+    {
+      "epoch": 18.58383233532934,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 12414
+    },
+    {
+      "epoch": 18.585329341317365,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.133,
+      "step": 12415
+    },
+    {
+      "epoch": 18.58682634730539,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1381,
+      "step": 12416
+    },
+    {
+      "epoch": 18.58832335329341,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 12417
+    },
+    {
+      "epoch": 18.589820359281436,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1353,
+      "step": 12418
+    },
+    {
+      "epoch": 18.59131736526946,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 12419
+    },
+    {
+      "epoch": 18.592814371257486,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 12420
+    },
+    {
+      "epoch": 18.59431137724551,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 12421
+    },
+    {
+      "epoch": 18.595808383233532,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 12422
+    },
+    {
+      "epoch": 18.597305389221557,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1323,
+      "step": 12423
+    },
+    {
+      "epoch": 18.59880239520958,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1358,
+      "step": 12424
+    },
+    {
+      "epoch": 18.600299401197606,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 12425
+    },
+    {
+      "epoch": 18.601796407185628,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 12426
+    },
+    {
+      "epoch": 18.603293413173652,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 12427
+    },
+    {
+      "epoch": 18.604790419161677,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1319,
+      "step": 12428
+    },
+    {
+      "epoch": 18.606287425149702,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1383,
+      "step": 12429
+    },
+    {
+      "epoch": 18.607784431137723,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 12430
+    },
+    {
+      "epoch": 18.60928143712575,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1355,
+      "step": 12431
+    },
+    {
+      "epoch": 18.610778443113773,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1307,
+      "step": 12432
+    },
+    {
+      "epoch": 18.612275449101798,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 12433
+    },
+    {
+      "epoch": 18.61377245508982,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 12434
+    },
+    {
+      "epoch": 18.615269461077844,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1443,
+      "step": 12435
+    },
+    {
+      "epoch": 18.61676646706587,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1384,
+      "step": 12436
+    },
+    {
+      "epoch": 18.618263473053894,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1384,
+      "step": 12437
+    },
+    {
+      "epoch": 18.619760479041915,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 12438
+    },
+    {
+      "epoch": 18.62125748502994,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 12439
+    },
+    {
+      "epoch": 18.622754491017965,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1334,
+      "step": 12440
+    },
+    {
+      "epoch": 18.62425149700599,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0008,
+      "loss": 1.136,
+      "step": 12441
+    },
+    {
+      "epoch": 18.62574850299401,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 12442
+    },
+    {
+      "epoch": 18.627245508982035,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 12443
+    },
+    {
+      "epoch": 18.62874251497006,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 12444
+    },
+    {
+      "epoch": 18.630239520958085,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 12445
+    },
+    {
+      "epoch": 18.631736526946106,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1332,
+      "step": 12446
+    },
+    {
+      "epoch": 18.63323353293413,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1333,
+      "step": 12447
+    },
+    {
+      "epoch": 18.634730538922156,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 12448
+    },
+    {
+      "epoch": 18.63622754491018,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1355,
+      "step": 12449
+    },
+    {
+      "epoch": 18.637724550898202,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 12450
+    },
+    {
+      "epoch": 18.639221556886227,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1432,
+      "step": 12451
+    },
+    {
+      "epoch": 18.64071856287425,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 12452
+    },
+    {
+      "epoch": 18.642215568862277,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1366,
+      "step": 12453
+    },
+    {
+      "epoch": 18.643712574850298,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1429,
+      "step": 12454
+    },
+    {
+      "epoch": 18.645209580838323,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 12455
+    },
+    {
+      "epoch": 18.646706586826348,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 12456
+    },
+    {
+      "epoch": 18.648203592814372,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1387,
+      "step": 12457
+    },
+    {
+      "epoch": 18.649700598802394,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1345,
+      "step": 12458
+    },
+    {
+      "epoch": 18.65119760479042,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1365,
+      "step": 12459
+    },
+    {
+      "epoch": 18.652694610778443,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1364,
+      "step": 12460
+    },
+    {
+      "epoch": 18.654191616766468,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 12461
+    },
+    {
+      "epoch": 18.65568862275449,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 12462
+    },
+    {
+      "epoch": 18.657185628742514,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1325,
+      "step": 12463
+    },
+    {
+      "epoch": 18.65868263473054,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1365,
+      "step": 12464
+    },
+    {
+      "epoch": 18.660179640718564,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1378,
+      "step": 12465
+    },
+    {
+      "epoch": 18.66167664670659,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1362,
+      "step": 12466
+    },
+    {
+      "epoch": 18.66317365269461,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1363,
+      "step": 12467
+    },
+    {
+      "epoch": 18.664670658682635,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1317,
+      "step": 12468
+    },
+    {
+      "epoch": 18.66616766467066,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 12469
+    },
+    {
+      "epoch": 18.66766467065868,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 12470
+    },
+    {
+      "epoch": 18.669161676646706,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 12471
+    },
+    {
+      "epoch": 18.67065868263473,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 12472
+    },
+    {
+      "epoch": 18.672155688622755,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.134,
+      "step": 12473
+    },
+    {
+      "epoch": 18.67365269461078,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1393,
+      "step": 12474
+    },
+    {
+      "epoch": 18.6751497005988,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1383,
+      "step": 12475
+    },
+    {
+      "epoch": 18.676646706586826,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1358,
+      "step": 12476
+    },
+    {
+      "epoch": 18.67814371257485,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1314,
+      "step": 12477
+    },
+    {
+      "epoch": 18.679640718562876,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 12478
+    },
+    {
+      "epoch": 18.681137724550897,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 12479
+    },
+    {
+      "epoch": 18.682634730538922,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1294,
+      "step": 12480
+    },
+    {
+      "epoch": 18.684131736526947,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 12481
+    },
+    {
+      "epoch": 18.68562874251497,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 12482
+    },
+    {
+      "epoch": 18.687125748502993,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1393,
+      "step": 12483
+    },
+    {
+      "epoch": 18.688622754491018,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1392,
+      "step": 12484
+    },
+    {
+      "epoch": 18.690119760479043,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1333,
+      "step": 12485
+    },
+    {
+      "epoch": 18.691616766467067,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 12486
+    },
+    {
+      "epoch": 18.69311377245509,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1368,
+      "step": 12487
+    },
+    {
+      "epoch": 18.694610778443113,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1369,
+      "step": 12488
+    },
+    {
+      "epoch": 18.69610778443114,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 12489
+    },
+    {
+      "epoch": 18.697604790419163,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1375,
+      "step": 12490
+    },
+    {
+      "epoch": 18.699101796407184,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1333,
+      "step": 12491
+    },
+    {
+      "epoch": 18.70059880239521,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1364,
+      "step": 12492
+    },
+    {
+      "epoch": 18.702095808383234,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1363,
+      "step": 12493
+    },
+    {
+      "epoch": 18.70359281437126,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1358,
+      "step": 12494
+    },
+    {
+      "epoch": 18.70508982035928,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 12495
+    },
+    {
+      "epoch": 18.706586826347305,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 12496
+    },
+    {
+      "epoch": 18.70808383233533,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1386,
+      "step": 12497
+    },
+    {
+      "epoch": 18.709580838323355,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 12498
+    },
+    {
+      "epoch": 18.711077844311376,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1372,
+      "step": 12499
+    },
+    {
+      "epoch": 18.7125748502994,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 12500
+    },
+    {
+      "epoch": 18.714071856287426,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1319,
+      "step": 12501
+    },
+    {
+      "epoch": 18.71556886227545,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1358,
+      "step": 12502
+    },
+    {
+      "epoch": 18.71706586826347,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 12503
+    },
+    {
+      "epoch": 18.718562874251496,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1378,
+      "step": 12504
+    },
+    {
+      "epoch": 18.72005988023952,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 12505
+    },
+    {
+      "epoch": 18.721556886227546,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1495,
+      "step": 12506
+    },
+    {
+      "epoch": 18.723053892215567,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 12507
+    },
+    {
+      "epoch": 18.724550898203592,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1364,
+      "step": 12508
+    },
+    {
+      "epoch": 18.726047904191617,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 12509
+    },
+    {
+      "epoch": 18.727544910179642,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 12510
+    },
+    {
+      "epoch": 18.729041916167663,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 12511
+    },
+    {
+      "epoch": 18.730538922155688,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 12512
+    },
+    {
+      "epoch": 18.732035928143713,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 12513
+    },
+    {
+      "epoch": 18.733532934131738,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1359,
+      "step": 12514
+    },
+    {
+      "epoch": 18.73502994011976,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.14,
+      "step": 12515
+    },
+    {
+      "epoch": 18.736526946107784,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1386,
+      "step": 12516
+    },
+    {
+      "epoch": 18.73802395209581,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1329,
+      "step": 12517
+    },
+    {
+      "epoch": 18.739520958083833,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 12518
+    },
+    {
+      "epoch": 18.741017964071855,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 12519
+    },
+    {
+      "epoch": 18.74251497005988,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 12520
+    },
+    {
+      "epoch": 18.744011976047904,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 12521
+    },
+    {
+      "epoch": 18.74550898203593,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1392,
+      "step": 12522
+    },
+    {
+      "epoch": 18.74700598802395,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1343,
+      "step": 12523
+    },
+    {
+      "epoch": 18.748502994011975,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 12524
+    },
+    {
+      "epoch": 18.75,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1365,
+      "step": 12525
+    },
+    {
+      "epoch": 18.751497005988025,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1433,
+      "step": 12526
+    },
+    {
+      "epoch": 18.75299401197605,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 12527
+    },
+    {
+      "epoch": 18.75449101796407,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1375,
+      "step": 12528
+    },
+    {
+      "epoch": 18.755988023952096,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 12529
+    },
+    {
+      "epoch": 18.75748502994012,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1375,
+      "step": 12530
+    },
+    {
+      "epoch": 18.758982035928145,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 12531
+    },
+    {
+      "epoch": 18.760479041916167,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1347,
+      "step": 12532
+    },
+    {
+      "epoch": 18.76197604790419,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 12533
+    },
+    {
+      "epoch": 18.763473053892216,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 12534
+    },
+    {
+      "epoch": 18.76497005988024,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 12535
+    },
+    {
+      "epoch": 18.766467065868262,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1347,
+      "step": 12536
+    },
+    {
+      "epoch": 18.767964071856287,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1387,
+      "step": 12537
+    },
+    {
+      "epoch": 18.769461077844312,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1353,
+      "step": 12538
+    },
+    {
+      "epoch": 18.770958083832337,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 12539
+    },
+    {
+      "epoch": 18.772455089820358,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1359,
+      "step": 12540
+    },
+    {
+      "epoch": 18.773952095808383,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 12541
+    },
+    {
+      "epoch": 18.775449101796408,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1371,
+      "step": 12542
+    },
+    {
+      "epoch": 18.776946107784433,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.143,
+      "step": 12543
+    },
+    {
+      "epoch": 18.778443113772454,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 12544
+    },
+    {
+      "epoch": 18.77994011976048,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 12545
+    },
+    {
+      "epoch": 18.781437125748504,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1366,
+      "step": 12546
+    },
+    {
+      "epoch": 18.78293413173653,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1395,
+      "step": 12547
+    },
+    {
+      "epoch": 18.78443113772455,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.14,
+      "step": 12548
+    },
+    {
+      "epoch": 18.785928143712574,
+      "grad_norm": 0.05859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 12549
+    },
+    {
+      "epoch": 18.7874251497006,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1478,
+      "step": 12550
+    },
+    {
+      "epoch": 18.788922155688624,
+      "grad_norm": 0.060546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 12551
+    },
+    {
+      "epoch": 18.790419161676645,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1365,
+      "step": 12552
+    },
+    {
+      "epoch": 18.79191616766467,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1336,
+      "step": 12553
+    },
+    {
+      "epoch": 18.793413173652695,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 12554
+    },
+    {
+      "epoch": 18.79491017964072,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1373,
+      "step": 12555
+    },
+    {
+      "epoch": 18.79640718562874,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1368,
+      "step": 12556
+    },
+    {
+      "epoch": 18.797904191616766,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1387,
+      "step": 12557
+    },
+    {
+      "epoch": 18.79940119760479,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1384,
+      "step": 12558
+    },
+    {
+      "epoch": 18.800898203592816,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 12559
+    },
+    {
+      "epoch": 18.802395209580837,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 12560
+    },
+    {
+      "epoch": 18.80389221556886,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1441,
+      "step": 12561
+    },
+    {
+      "epoch": 18.805389221556887,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 12562
+    },
+    {
+      "epoch": 18.80688622754491,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1362,
+      "step": 12563
+    },
+    {
+      "epoch": 18.808383233532933,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1352,
+      "step": 12564
+    },
+    {
+      "epoch": 18.809880239520957,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1488,
+      "step": 12565
+    },
+    {
+      "epoch": 18.811377245508982,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 12566
+    },
+    {
+      "epoch": 18.812874251497007,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1359,
+      "step": 12567
+    },
+    {
+      "epoch": 18.81437125748503,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1345,
+      "step": 12568
+    },
+    {
+      "epoch": 18.815868263473053,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1317,
+      "step": 12569
+    },
+    {
+      "epoch": 18.817365269461078,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 12570
+    },
+    {
+      "epoch": 18.818862275449103,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1334,
+      "step": 12571
+    },
+    {
+      "epoch": 18.820359281437124,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 12572
+    },
+    {
+      "epoch": 18.82185628742515,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1361,
+      "step": 12573
+    },
+    {
+      "epoch": 18.823353293413174,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1304,
+      "step": 12574
+    },
+    {
+      "epoch": 18.8248502994012,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1392,
+      "step": 12575
+    },
+    {
+      "epoch": 18.82634730538922,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1364,
+      "step": 12576
+    },
+    {
+      "epoch": 18.827844311377245,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 12577
+    },
+    {
+      "epoch": 18.82934131736527,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1369,
+      "step": 12578
+    },
+    {
+      "epoch": 18.830838323353294,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 12579
+    },
+    {
+      "epoch": 18.83233532934132,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 12580
+    },
+    {
+      "epoch": 18.83383233532934,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1459,
+      "step": 12581
+    },
+    {
+      "epoch": 18.835329341317365,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1379,
+      "step": 12582
+    },
+    {
+      "epoch": 18.83682634730539,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1346,
+      "step": 12583
+    },
+    {
+      "epoch": 18.83832335329341,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1383,
+      "step": 12584
+    },
+    {
+      "epoch": 18.839820359281436,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1327,
+      "step": 12585
+    },
+    {
+      "epoch": 18.84131736526946,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1327,
+      "step": 12586
+    },
+    {
+      "epoch": 18.842814371257486,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1444,
+      "step": 12587
+    },
+    {
+      "epoch": 18.84431137724551,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1335,
+      "step": 12588
+    },
+    {
+      "epoch": 18.845808383233532,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 12589
+    },
+    {
+      "epoch": 18.847305389221557,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1292,
+      "step": 12590
+    },
+    {
+      "epoch": 18.84880239520958,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1429,
+      "step": 12591
+    },
+    {
+      "epoch": 18.850299401197606,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1395,
+      "step": 12592
+    },
+    {
+      "epoch": 18.851796407185628,
+      "grad_norm": 0.056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1348,
+      "step": 12593
+    },
+    {
+      "epoch": 18.853293413173652,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1337,
+      "step": 12594
+    },
+    {
+      "epoch": 18.854790419161677,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1456,
+      "step": 12595
+    },
+    {
+      "epoch": 18.856287425149702,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1366,
+      "step": 12596
+    },
+    {
+      "epoch": 18.857784431137723,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 12597
+    },
+    {
+      "epoch": 18.85928143712575,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 12598
+    },
+    {
+      "epoch": 18.860778443113773,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 12599
+    },
+    {
+      "epoch": 18.862275449101798,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1336,
+      "step": 12600
+    },
+    {
+      "epoch": 18.86377245508982,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 12601
+    },
+    {
+      "epoch": 18.865269461077844,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1313,
+      "step": 12602
+    },
+    {
+      "epoch": 18.86676646706587,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1413,
+      "step": 12603
+    },
+    {
+      "epoch": 18.868263473053894,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1344,
+      "step": 12604
+    },
+    {
+      "epoch": 18.869760479041915,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.132,
+      "step": 12605
+    },
+    {
+      "epoch": 18.87125748502994,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1423,
+      "step": 12606
+    },
+    {
+      "epoch": 18.872754491017965,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 12607
+    },
+    {
+      "epoch": 18.87425149700599,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 12608
+    },
+    {
+      "epoch": 18.87574850299401,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1354,
+      "step": 12609
+    },
+    {
+      "epoch": 18.877245508982035,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1359,
+      "step": 12610
+    },
+    {
+      "epoch": 18.87874251497006,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1353,
+      "step": 12611
+    },
+    {
+      "epoch": 18.880239520958085,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 12612
+    },
+    {
+      "epoch": 18.881736526946106,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1357,
+      "step": 12613
+    },
+    {
+      "epoch": 18.88323353293413,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 12614
+    },
+    {
+      "epoch": 18.884730538922156,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1348,
+      "step": 12615
+    },
+    {
+      "epoch": 18.88622754491018,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1371,
+      "step": 12616
+    },
+    {
+      "epoch": 18.887724550898202,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 12617
+    },
+    {
+      "epoch": 18.889221556886227,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.133,
+      "step": 12618
+    },
+    {
+      "epoch": 18.89071856287425,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1378,
+      "step": 12619
+    },
+    {
+      "epoch": 18.892215568862277,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 12620
+    },
+    {
+      "epoch": 18.893712574850298,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 12621
+    },
+    {
+      "epoch": 18.895209580838323,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1378,
+      "step": 12622
+    },
+    {
+      "epoch": 18.896706586826348,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 12623
+    },
+    {
+      "epoch": 18.898203592814372,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1362,
+      "step": 12624
+    },
+    {
+      "epoch": 18.899700598802394,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1354,
+      "step": 12625
+    },
+    {
+      "epoch": 18.90119760479042,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 12626
+    },
+    {
+      "epoch": 18.902694610778443,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 12627
+    },
+    {
+      "epoch": 18.904191616766468,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1296,
+      "step": 12628
+    },
+    {
+      "epoch": 18.90568862275449,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1343,
+      "step": 12629
+    },
+    {
+      "epoch": 18.907185628742514,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1392,
+      "step": 12630
+    },
+    {
+      "epoch": 18.90868263473054,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1379,
+      "step": 12631
+    },
+    {
+      "epoch": 18.910179640718564,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1377,
+      "step": 12632
+    },
+    {
+      "epoch": 18.91167664670659,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 12633
+    },
+    {
+      "epoch": 18.91317365269461,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1373,
+      "step": 12634
+    },
+    {
+      "epoch": 18.914670658682635,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1351,
+      "step": 12635
+    },
+    {
+      "epoch": 18.91616766467066,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1368,
+      "step": 12636
+    },
+    {
+      "epoch": 18.91766467065868,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1377,
+      "step": 12637
+    },
+    {
+      "epoch": 18.919161676646706,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 12638
+    },
+    {
+      "epoch": 18.92065868263473,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 12639
+    },
+    {
+      "epoch": 18.922155688622755,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1383,
+      "step": 12640
+    },
+    {
+      "epoch": 18.92365269461078,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1384,
+      "step": 12641
+    },
+    {
+      "epoch": 18.9251497005988,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 12642
+    },
+    {
+      "epoch": 18.926646706586826,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1379,
+      "step": 12643
+    },
+    {
+      "epoch": 18.92814371257485,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 12644
+    },
+    {
+      "epoch": 18.929640718562876,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 12645
+    },
+    {
+      "epoch": 18.931137724550897,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1368,
+      "step": 12646
+    },
+    {
+      "epoch": 18.932634730538922,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1345,
+      "step": 12647
+    },
+    {
+      "epoch": 18.934131736526947,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.144,
+      "step": 12648
+    },
+    {
+      "epoch": 18.93562874251497,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1379,
+      "step": 12649
+    },
+    {
+      "epoch": 18.937125748502993,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1399,
+      "step": 12650
+    },
+    {
+      "epoch": 18.938622754491018,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 12651
+    },
+    {
+      "epoch": 18.940119760479043,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1366,
+      "step": 12652
+    },
+    {
+      "epoch": 18.941616766467067,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 12653
+    },
+    {
+      "epoch": 18.94311377245509,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 12654
+    },
+    {
+      "epoch": 18.944610778443113,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.135,
+      "step": 12655
+    },
+    {
+      "epoch": 18.94610778443114,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1381,
+      "step": 12656
+    },
+    {
+      "epoch": 18.947604790419163,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 12657
+    },
+    {
+      "epoch": 18.949101796407184,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1348,
+      "step": 12658
+    },
+    {
+      "epoch": 18.95059880239521,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1378,
+      "step": 12659
+    },
+    {
+      "epoch": 18.952095808383234,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 12660
+    },
+    {
+      "epoch": 18.95359281437126,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1392,
+      "step": 12661
+    },
+    {
+      "epoch": 18.95508982035928,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1376,
+      "step": 12662
+    },
+    {
+      "epoch": 18.956586826347305,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 12663
+    },
+    {
+      "epoch": 18.95808383233533,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 12664
+    },
+    {
+      "epoch": 18.959580838323355,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1342,
+      "step": 12665
+    },
+    {
+      "epoch": 18.961077844311376,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.137,
+      "step": 12666
+    },
+    {
+      "epoch": 18.9625748502994,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1344,
+      "step": 12667
+    },
+    {
+      "epoch": 18.964071856287426,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1369,
+      "step": 12668
+    },
+    {
+      "epoch": 18.96556886227545,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1376,
+      "step": 12669
+    },
+    {
+      "epoch": 18.96706586826347,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 12670
+    },
+    {
+      "epoch": 18.968562874251496,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.14,
+      "step": 12671
+    },
+    {
+      "epoch": 18.97005988023952,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1435,
+      "step": 12672
+    },
+    {
+      "epoch": 18.971556886227546,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1317,
+      "step": 12673
+    },
+    {
+      "epoch": 18.973053892215567,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 12674
+    },
+    {
+      "epoch": 18.974550898203592,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1361,
+      "step": 12675
+    },
+    {
+      "epoch": 18.976047904191617,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1359,
+      "step": 12676
+    },
+    {
+      "epoch": 18.977544910179642,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1315,
+      "step": 12677
+    },
+    {
+      "epoch": 18.979041916167663,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 12678
+    },
+    {
+      "epoch": 18.980538922155688,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1362,
+      "step": 12679
+    },
+    {
+      "epoch": 18.982035928143713,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 12680
+    },
+    {
+      "epoch": 18.983532934131738,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1338,
+      "step": 12681
+    },
+    {
+      "epoch": 18.98502994011976,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 12682
+    },
+    {
+      "epoch": 18.986526946107784,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 12683
+    },
+    {
+      "epoch": 18.98802395209581,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1378,
+      "step": 12684
+    },
+    {
+      "epoch": 18.989520958083833,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.14,
+      "step": 12685
+    },
+    {
+      "epoch": 18.991017964071855,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1371,
+      "step": 12686
+    },
+    {
+      "epoch": 18.99251497005988,
+      "grad_norm": 0.08056640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 12687
+    },
+    {
+      "epoch": 18.994011976047904,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1363,
+      "step": 12688
+    },
+    {
+      "epoch": 18.99550898203593,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1381,
+      "step": 12689
+    },
+    {
+      "epoch": 18.99700598802395,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1354,
+      "step": 12690
+    },
+    {
+      "epoch": 18.998502994011975,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1448,
+      "step": 12691
+    },
+    {
+      "epoch": 19.0,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.136,
+      "step": 12692
+    },
+    {
+      "epoch": 19.001497005988025,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1364,
+      "step": 12693
+    },
+    {
+      "epoch": 19.00299401197605,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 12694
+    },
+    {
+      "epoch": 19.00449101796407,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1391,
+      "step": 12695
+    },
+    {
+      "epoch": 19.005988023952096,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1348,
+      "step": 12696
+    },
+    {
+      "epoch": 19.00748502994012,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 12697
+    },
+    {
+      "epoch": 19.008982035928145,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1413,
+      "step": 12698
+    },
+    {
+      "epoch": 19.010479041916167,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1412,
+      "step": 12699
+    },
+    {
+      "epoch": 19.01197604790419,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1359,
+      "step": 12700
+    },
+    {
+      "epoch": 19.013473053892216,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 12701
+    },
+    {
+      "epoch": 19.01497005988024,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1365,
+      "step": 12702
+    },
+    {
+      "epoch": 19.016467065868262,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1345,
+      "step": 12703
+    },
+    {
+      "epoch": 19.017964071856287,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1309,
+      "step": 12704
+    },
+    {
+      "epoch": 19.019461077844312,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1364,
+      "step": 12705
+    },
+    {
+      "epoch": 19.020958083832337,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1362,
+      "step": 12706
+    },
+    {
+      "epoch": 19.022455089820358,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1342,
+      "step": 12707
+    },
+    {
+      "epoch": 19.023952095808383,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 12708
+    },
+    {
+      "epoch": 19.025449101796408,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 12709
+    },
+    {
+      "epoch": 19.026946107784433,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1393,
+      "step": 12710
+    },
+    {
+      "epoch": 19.028443113772454,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1334,
+      "step": 12711
+    },
+    {
+      "epoch": 19.02994011976048,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1373,
+      "step": 12712
+    },
+    {
+      "epoch": 19.031437125748504,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1351,
+      "step": 12713
+    },
+    {
+      "epoch": 19.03293413173653,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1329,
+      "step": 12714
+    },
+    {
+      "epoch": 19.03443113772455,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 12715
+    },
+    {
+      "epoch": 19.035928143712574,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1331,
+      "step": 12716
+    },
+    {
+      "epoch": 19.0374251497006,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1321,
+      "step": 12717
+    },
+    {
+      "epoch": 19.038922155688624,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 12718
+    },
+    {
+      "epoch": 19.040419161676645,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 12719
+    },
+    {
+      "epoch": 19.04191616766467,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1347,
+      "step": 12720
+    },
+    {
+      "epoch": 19.043413173652695,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1321,
+      "step": 12721
+    },
+    {
+      "epoch": 19.04491017964072,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 12722
+    },
+    {
+      "epoch": 19.04640718562874,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1429,
+      "step": 12723
+    },
+    {
+      "epoch": 19.047904191616766,
+      "grad_norm": 0.099609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1372,
+      "step": 12724
+    },
+    {
+      "epoch": 19.04940119760479,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1381,
+      "step": 12725
+    },
+    {
+      "epoch": 19.050898203592816,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1353,
+      "step": 12726
+    },
+    {
+      "epoch": 19.052395209580837,
+      "grad_norm": 0.1171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1349,
+      "step": 12727
+    },
+    {
+      "epoch": 19.05389221556886,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 12728
+    },
+    {
+      "epoch": 19.055389221556887,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1365,
+      "step": 12729
+    },
+    {
+      "epoch": 19.05688622754491,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1351,
+      "step": 12730
+    },
+    {
+      "epoch": 19.058383233532933,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1392,
+      "step": 12731
+    },
+    {
+      "epoch": 19.059880239520957,
+      "grad_norm": 0.12890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1373,
+      "step": 12732
+    },
+    {
+      "epoch": 19.061377245508982,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.137,
+      "step": 12733
+    },
+    {
+      "epoch": 19.062874251497007,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 12734
+    },
+    {
+      "epoch": 19.06437125748503,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 12735
+    },
+    {
+      "epoch": 19.065868263473053,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1313,
+      "step": 12736
+    },
+    {
+      "epoch": 19.067365269461078,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 12737
+    },
+    {
+      "epoch": 19.068862275449103,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 12738
+    },
+    {
+      "epoch": 19.070359281437124,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1472,
+      "step": 12739
+    },
+    {
+      "epoch": 19.07185628742515,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1315,
+      "step": 12740
+    },
+    {
+      "epoch": 19.073353293413174,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1376,
+      "step": 12741
+    },
+    {
+      "epoch": 19.0748502994012,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1362,
+      "step": 12742
+    },
+    {
+      "epoch": 19.07634730538922,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1367,
+      "step": 12743
+    },
+    {
+      "epoch": 19.077844311377245,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.14,
+      "step": 12744
+    },
+    {
+      "epoch": 19.07934131736527,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.142,
+      "step": 12745
+    },
+    {
+      "epoch": 19.080838323353294,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.138,
+      "step": 12746
+    },
+    {
+      "epoch": 19.082335329341316,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1347,
+      "step": 12747
+    },
+    {
+      "epoch": 19.08383233532934,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1349,
+      "step": 12748
+    },
+    {
+      "epoch": 19.085329341317365,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1446,
+      "step": 12749
+    },
+    {
+      "epoch": 19.08682634730539,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 12750
+    },
+    {
+      "epoch": 19.088323353293415,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 12751
+    },
+    {
+      "epoch": 19.089820359281436,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 12752
+    },
+    {
+      "epoch": 19.09131736526946,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 12753
+    },
+    {
+      "epoch": 19.092814371257486,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 12754
+    },
+    {
+      "epoch": 19.09431137724551,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1368,
+      "step": 12755
+    },
+    {
+      "epoch": 19.095808383233532,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 12756
+    },
+    {
+      "epoch": 19.097305389221557,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1296,
+      "step": 12757
+    },
+    {
+      "epoch": 19.09880239520958,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1474,
+      "step": 12758
+    },
+    {
+      "epoch": 19.100299401197606,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1361,
+      "step": 12759
+    },
+    {
+      "epoch": 19.101796407185628,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1438,
+      "step": 12760
+    },
+    {
+      "epoch": 19.103293413173652,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 12761
+    },
+    {
+      "epoch": 19.104790419161677,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 12762
+    },
+    {
+      "epoch": 19.106287425149702,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1375,
+      "step": 12763
+    },
+    {
+      "epoch": 19.107784431137723,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.132,
+      "step": 12764
+    },
+    {
+      "epoch": 19.10928143712575,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1426,
+      "step": 12765
+    },
+    {
+      "epoch": 19.110778443113773,
+      "grad_norm": 0.095703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 12766
+    },
+    {
+      "epoch": 19.112275449101798,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1346,
+      "step": 12767
+    },
+    {
+      "epoch": 19.11377245508982,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1383,
+      "step": 12768
+    },
+    {
+      "epoch": 19.115269461077844,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.132,
+      "step": 12769
+    },
+    {
+      "epoch": 19.11676646706587,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1365,
+      "step": 12770
+    },
+    {
+      "epoch": 19.118263473053894,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1334,
+      "step": 12771
+    },
+    {
+      "epoch": 19.119760479041915,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 12772
+    },
+    {
+      "epoch": 19.12125748502994,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 12773
+    },
+    {
+      "epoch": 19.122754491017965,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1378,
+      "step": 12774
+    },
+    {
+      "epoch": 19.12425149700599,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1337,
+      "step": 12775
+    },
+    {
+      "epoch": 19.12574850299401,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1386,
+      "step": 12776
+    },
+    {
+      "epoch": 19.127245508982035,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1387,
+      "step": 12777
+    },
+    {
+      "epoch": 19.12874251497006,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1334,
+      "step": 12778
+    },
+    {
+      "epoch": 19.130239520958085,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.137,
+      "step": 12779
+    },
+    {
+      "epoch": 19.131736526946106,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1375,
+      "step": 12780
+    },
+    {
+      "epoch": 19.13323353293413,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1366,
+      "step": 12781
+    },
+    {
+      "epoch": 19.134730538922156,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.132,
+      "step": 12782
+    },
+    {
+      "epoch": 19.13622754491018,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1378,
+      "step": 12783
+    },
+    {
+      "epoch": 19.137724550898202,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1291,
+      "step": 12784
+    },
+    {
+      "epoch": 19.139221556886227,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1356,
+      "step": 12785
+    },
+    {
+      "epoch": 19.14071856287425,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1371,
+      "step": 12786
+    },
+    {
+      "epoch": 19.142215568862277,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1386,
+      "step": 12787
+    },
+    {
+      "epoch": 19.143712574850298,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1375,
+      "step": 12788
+    },
+    {
+      "epoch": 19.145209580838323,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1373,
+      "step": 12789
+    },
+    {
+      "epoch": 19.146706586826348,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1349,
+      "step": 12790
+    },
+    {
+      "epoch": 19.148203592814372,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 12791
+    },
+    {
+      "epoch": 19.149700598802394,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 12792
+    },
+    {
+      "epoch": 19.15119760479042,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1336,
+      "step": 12793
+    },
+    {
+      "epoch": 19.152694610778443,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1286,
+      "step": 12794
+    },
+    {
+      "epoch": 19.154191616766468,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 12795
+    },
+    {
+      "epoch": 19.15568862275449,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1355,
+      "step": 12796
+    },
+    {
+      "epoch": 19.157185628742514,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1358,
+      "step": 12797
+    },
+    {
+      "epoch": 19.15868263473054,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 12798
+    },
+    {
+      "epoch": 19.160179640718564,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1429,
+      "step": 12799
+    },
+    {
+      "epoch": 19.161676646706585,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1361,
+      "step": 12800
+    },
+    {
+      "epoch": 19.16317365269461,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1369,
+      "step": 12801
+    },
+    {
+      "epoch": 19.164670658682635,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.127,
+      "step": 12802
+    },
+    {
+      "epoch": 19.16616766467066,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1312,
+      "step": 12803
+    },
+    {
+      "epoch": 19.16766467065868,
+      "grad_norm": 0.06494140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1376,
+      "step": 12804
+    },
+    {
+      "epoch": 19.169161676646706,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1325,
+      "step": 12805
+    },
+    {
+      "epoch": 19.17065868263473,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1364,
+      "step": 12806
+    },
+    {
+      "epoch": 19.172155688622755,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.136,
+      "step": 12807
+    },
+    {
+      "epoch": 19.17365269461078,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1336,
+      "step": 12808
+    },
+    {
+      "epoch": 19.1751497005988,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1363,
+      "step": 12809
+    },
+    {
+      "epoch": 19.176646706586826,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 12810
+    },
+    {
+      "epoch": 19.17814371257485,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1329,
+      "step": 12811
+    },
+    {
+      "epoch": 19.179640718562876,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1385,
+      "step": 12812
+    },
+    {
+      "epoch": 19.181137724550897,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1409,
+      "step": 12813
+    },
+    {
+      "epoch": 19.182634730538922,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 12814
+    },
+    {
+      "epoch": 19.184131736526947,
+      "grad_norm": 0.060302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1379,
+      "step": 12815
+    },
+    {
+      "epoch": 19.18562874251497,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1342,
+      "step": 12816
+    },
+    {
+      "epoch": 19.187125748502993,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1403,
+      "step": 12817
+    },
+    {
+      "epoch": 19.188622754491018,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 12818
+    },
+    {
+      "epoch": 19.190119760479043,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 12819
+    },
+    {
+      "epoch": 19.191616766467067,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1407,
+      "step": 12820
+    },
+    {
+      "epoch": 19.19311377245509,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 12821
+    },
+    {
+      "epoch": 19.194610778443113,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1308,
+      "step": 12822
+    },
+    {
+      "epoch": 19.19610778443114,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1353,
+      "step": 12823
+    },
+    {
+      "epoch": 19.197604790419163,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1387,
+      "step": 12824
+    },
+    {
+      "epoch": 19.199101796407184,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 12825
+    },
+    {
+      "epoch": 19.20059880239521,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1452,
+      "step": 12826
+    },
+    {
+      "epoch": 19.202095808383234,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1378,
+      "step": 12827
+    },
+    {
+      "epoch": 19.20359281437126,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1342,
+      "step": 12828
+    },
+    {
+      "epoch": 19.20508982035928,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1352,
+      "step": 12829
+    },
+    {
+      "epoch": 19.206586826347305,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 12830
+    },
+    {
+      "epoch": 19.20808383233533,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1445,
+      "step": 12831
+    },
+    {
+      "epoch": 19.209580838323355,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1319,
+      "step": 12832
+    },
+    {
+      "epoch": 19.211077844311376,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 12833
+    },
+    {
+      "epoch": 19.2125748502994,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1439,
+      "step": 12834
+    },
+    {
+      "epoch": 19.214071856287426,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1367,
+      "step": 12835
+    },
+    {
+      "epoch": 19.21556886227545,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1304,
+      "step": 12836
+    },
+    {
+      "epoch": 19.21706586826347,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1363,
+      "step": 12837
+    },
+    {
+      "epoch": 19.218562874251496,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 12838
+    },
+    {
+      "epoch": 19.22005988023952,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1371,
+      "step": 12839
+    },
+    {
+      "epoch": 19.221556886227546,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1364,
+      "step": 12840
+    },
+    {
+      "epoch": 19.223053892215567,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1379,
+      "step": 12841
+    },
+    {
+      "epoch": 19.224550898203592,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1376,
+      "step": 12842
+    },
+    {
+      "epoch": 19.226047904191617,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1392,
+      "step": 12843
+    },
+    {
+      "epoch": 19.227544910179642,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1373,
+      "step": 12844
+    },
+    {
+      "epoch": 19.229041916167663,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1289,
+      "step": 12845
+    },
+    {
+      "epoch": 19.230538922155688,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1312,
+      "step": 12846
+    },
+    {
+      "epoch": 19.232035928143713,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1374,
+      "step": 12847
+    },
+    {
+      "epoch": 19.233532934131738,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1357,
+      "step": 12848
+    },
+    {
+      "epoch": 19.23502994011976,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1346,
+      "step": 12849
+    },
+    {
+      "epoch": 19.236526946107784,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1372,
+      "step": 12850
+    },
+    {
+      "epoch": 19.23802395209581,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 12851
+    },
+    {
+      "epoch": 19.239520958083833,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1332,
+      "step": 12852
+    },
+    {
+      "epoch": 19.241017964071855,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.135,
+      "step": 12853
+    },
+    {
+      "epoch": 19.24251497005988,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1342,
+      "step": 12854
+    },
+    {
+      "epoch": 19.244011976047904,
+      "grad_norm": 0.06787109375,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 12855
+    },
+    {
+      "epoch": 19.24550898203593,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1384,
+      "step": 12856
+    },
+    {
+      "epoch": 19.24700598802395,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1262,
+      "step": 12857
+    },
+    {
+      "epoch": 19.248502994011975,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1312,
+      "step": 12858
+    },
+    {
+      "epoch": 19.25,
+      "grad_norm": 0.083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1383,
+      "step": 12859
+    },
+    {
+      "epoch": 19.251497005988025,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1367,
+      "step": 12860
+    },
+    {
+      "epoch": 19.25299401197605,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1422,
+      "step": 12861
+    },
+    {
+      "epoch": 19.25449101796407,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1341,
+      "step": 12862
+    },
+    {
+      "epoch": 19.255988023952096,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1373,
+      "step": 12863
+    },
+    {
+      "epoch": 19.25748502994012,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1369,
+      "step": 12864
+    },
+    {
+      "epoch": 19.258982035928145,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1335,
+      "step": 12865
+    },
+    {
+      "epoch": 19.260479041916167,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.134,
+      "step": 12866
+    },
+    {
+      "epoch": 19.26197604790419,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1322,
+      "step": 12867
+    },
+    {
+      "epoch": 19.263473053892216,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1383,
+      "step": 12868
+    },
+    {
+      "epoch": 19.26497005988024,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 12869
+    },
+    {
+      "epoch": 19.266467065868262,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1364,
+      "step": 12870
+    },
+    {
+      "epoch": 19.267964071856287,
+      "grad_norm": 0.13671875,
+      "learning_rate": 0.0008,
+      "loss": 1.145,
+      "step": 12871
+    },
+    {
+      "epoch": 19.269461077844312,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1376,
+      "step": 12872
+    },
+    {
+      "epoch": 19.270958083832337,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.129,
+      "step": 12873
+    },
+    {
+      "epoch": 19.272455089820358,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1341,
+      "step": 12874
+    },
+    {
+      "epoch": 19.273952095808383,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1406,
+      "step": 12875
+    },
+    {
+      "epoch": 19.275449101796408,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1277,
+      "step": 12876
+    },
+    {
+      "epoch": 19.276946107784433,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1333,
+      "step": 12877
+    },
+    {
+      "epoch": 19.278443113772454,
+      "grad_norm": 0.134765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1434,
+      "step": 12878
+    },
+    {
+      "epoch": 19.27994011976048,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1386,
+      "step": 12879
+    },
+    {
+      "epoch": 19.281437125748504,
+      "grad_norm": 0.126953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1316,
+      "step": 12880
+    },
+    {
+      "epoch": 19.28293413173653,
+      "grad_norm": 0.1015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1361,
+      "step": 12881
+    },
+    {
+      "epoch": 19.28443113772455,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1371,
+      "step": 12882
+    },
+    {
+      "epoch": 19.285928143712574,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1359,
+      "step": 12883
+    },
+    {
+      "epoch": 19.2874251497006,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1382,
+      "step": 12884
+    },
+    {
+      "epoch": 19.288922155688624,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1421,
+      "step": 12885
+    },
+    {
+      "epoch": 19.290419161676645,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1377,
+      "step": 12886
+    },
+    {
+      "epoch": 19.29191616766467,
+      "grad_norm": 0.07958984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1357,
+      "step": 12887
+    },
+    {
+      "epoch": 19.293413173652695,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1362,
+      "step": 12888
+    },
+    {
+      "epoch": 19.29491017964072,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1331,
+      "step": 12889
+    },
+    {
+      "epoch": 19.29640718562874,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1463,
+      "step": 12890
+    },
+    {
+      "epoch": 19.297904191616766,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1434,
+      "step": 12891
+    },
+    {
+      "epoch": 19.29940119760479,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 12892
+    },
+    {
+      "epoch": 19.300898203592816,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1392,
+      "step": 12893
+    },
+    {
+      "epoch": 19.302395209580837,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1415,
+      "step": 12894
+    },
+    {
+      "epoch": 19.30389221556886,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1362,
+      "step": 12895
+    },
+    {
+      "epoch": 19.305389221556887,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 12896
+    },
+    {
+      "epoch": 19.30688622754491,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.134,
+      "step": 12897
+    },
+    {
+      "epoch": 19.308383233532933,
+      "grad_norm": 0.08203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1454,
+      "step": 12898
+    },
+    {
+      "epoch": 19.309880239520957,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.134,
+      "step": 12899
+    },
+    {
+      "epoch": 19.311377245508982,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1376,
+      "step": 12900
+    },
+    {
+      "epoch": 19.312874251497007,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1446,
+      "step": 12901
+    },
+    {
+      "epoch": 19.31437125748503,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1372,
+      "step": 12902
+    },
+    {
+      "epoch": 19.315868263473053,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1331,
+      "step": 12903
+    },
+    {
+      "epoch": 19.317365269461078,
+      "grad_norm": 0.0712890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 12904
+    },
+    {
+      "epoch": 19.318862275449103,
+      "grad_norm": 0.09375,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 12905
+    },
+    {
+      "epoch": 19.320359281437124,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1394,
+      "step": 12906
+    },
+    {
+      "epoch": 19.32185628742515,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1396,
+      "step": 12907
+    },
+    {
+      "epoch": 19.323353293413174,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1411,
+      "step": 12908
+    },
+    {
+      "epoch": 19.3248502994012,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.136,
+      "step": 12909
+    },
+    {
+      "epoch": 19.32634730538922,
+      "grad_norm": 0.11328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1378,
+      "step": 12910
+    },
+    {
+      "epoch": 19.327844311377245,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.133,
+      "step": 12911
+    },
+    {
+      "epoch": 19.32934131736527,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1387,
+      "step": 12912
+    },
+    {
+      "epoch": 19.330838323353294,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1266,
+      "step": 12913
+    },
+    {
+      "epoch": 19.33233532934132,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 12914
+    },
+    {
+      "epoch": 19.33383233532934,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1359,
+      "step": 12915
+    },
+    {
+      "epoch": 19.335329341317365,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 12916
+    },
+    {
+      "epoch": 19.33682634730539,
+      "grad_norm": 0.087890625,
+      "learning_rate": 0.0008,
+      "loss": 1.1341,
+      "step": 12917
+    },
+    {
+      "epoch": 19.338323353293415,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1312,
+      "step": 12918
+    },
+    {
+      "epoch": 19.339820359281436,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1379,
+      "step": 12919
+    },
+    {
+      "epoch": 19.34131736526946,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.137,
+      "step": 12920
+    },
+    {
+      "epoch": 19.342814371257486,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1371,
+      "step": 12921
+    },
+    {
+      "epoch": 19.34431137724551,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1387,
+      "step": 12922
+    },
+    {
+      "epoch": 19.345808383233532,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1388,
+      "step": 12923
+    },
+    {
+      "epoch": 19.347305389221557,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1377,
+      "step": 12924
+    },
+    {
+      "epoch": 19.34880239520958,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 12925
+    },
+    {
+      "epoch": 19.350299401197606,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 12926
+    },
+    {
+      "epoch": 19.351796407185628,
+      "grad_norm": 0.078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1364,
+      "step": 12927
+    },
+    {
+      "epoch": 19.353293413173652,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 12928
+    },
+    {
+      "epoch": 19.354790419161677,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1349,
+      "step": 12929
+    },
+    {
+      "epoch": 19.356287425149702,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1401,
+      "step": 12930
+    },
+    {
+      "epoch": 19.357784431137723,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1373,
+      "step": 12931
+    },
+    {
+      "epoch": 19.35928143712575,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 12932
+    },
+    {
+      "epoch": 19.360778443113773,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1371,
+      "step": 12933
+    },
+    {
+      "epoch": 19.362275449101798,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1369,
+      "step": 12934
+    },
+    {
+      "epoch": 19.36377245508982,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1368,
+      "step": 12935
+    },
+    {
+      "epoch": 19.365269461077844,
+      "grad_norm": 0.06982421875,
+      "learning_rate": 0.0008,
+      "loss": 1.1336,
+      "step": 12936
+    },
+    {
+      "epoch": 19.36676646706587,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1336,
+      "step": 12937
+    },
+    {
+      "epoch": 19.368263473053894,
+      "grad_norm": 0.080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 12938
+    },
+    {
+      "epoch": 19.369760479041915,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1378,
+      "step": 12939
+    },
+    {
+      "epoch": 19.37125748502994,
+      "grad_norm": 0.076171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1369,
+      "step": 12940
+    },
+    {
+      "epoch": 19.372754491017965,
+      "grad_norm": 0.0732421875,
+      "learning_rate": 0.0008,
+      "loss": 1.139,
+      "step": 12941
+    },
+    {
+      "epoch": 19.37425149700599,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1417,
+      "step": 12942
+    },
+    {
+      "epoch": 19.37574850299401,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1356,
+      "step": 12943
+    },
+    {
+      "epoch": 19.377245508982035,
+      "grad_norm": 0.0771484375,
+      "learning_rate": 0.0008,
+      "loss": 1.135,
+      "step": 12944
+    },
+    {
+      "epoch": 19.37874251497006,
+      "grad_norm": 0.06396484375,
+      "learning_rate": 0.0008,
+      "loss": 1.1343,
+      "step": 12945
+    },
+    {
+      "epoch": 19.380239520958085,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 0.0008,
+      "loss": 1.1367,
+      "step": 12946
+    },
+    {
+      "epoch": 19.381736526946106,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.141,
+      "step": 12947
+    },
+    {
+      "epoch": 19.38323353293413,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 0.0008,
+      "loss": 1.1461,
+      "step": 12948
+    },
+    {
+      "epoch": 19.384730538922156,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 12949
+    },
+    {
+      "epoch": 19.38622754491018,
+      "grad_norm": 0.08154296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1311,
+      "step": 12950
+    },
+    {
+      "epoch": 19.387724550898202,
+      "grad_norm": 0.103515625,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 12951
+    },
+    {
+      "epoch": 19.389221556886227,
+      "grad_norm": 0.0673828125,
+      "learning_rate": 0.0008,
+      "loss": 1.1384,
+      "step": 12952
+    },
+    {
+      "epoch": 19.39071856287425,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 0.0008,
+      "loss": 1.1408,
+      "step": 12953
+    },
+    {
+      "epoch": 19.392215568862277,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1393,
+      "step": 12954
+    },
+    {
+      "epoch": 19.393712574850298,
+      "grad_norm": 0.0703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 12955
+    },
+    {
+      "epoch": 19.395209580838323,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1389,
+      "step": 12956
+    },
+    {
+      "epoch": 19.396706586826348,
+      "grad_norm": 0.0693359375,
+      "learning_rate": 0.0008,
+      "loss": 1.134,
+      "step": 12957
+    },
+    {
+      "epoch": 19.398203592814372,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 0.0008,
+      "loss": 1.1331,
+      "step": 12958
+    },
+    {
+      "epoch": 19.399700598802394,
+      "grad_norm": 0.091796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1368,
+      "step": 12959
+    },
+    {
+      "epoch": 19.40119760479042,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1328,
+      "step": 12960
+    },
+    {
+      "epoch": 19.402694610778443,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1458,
+      "step": 12961
+    },
+    {
+      "epoch": 19.404191616766468,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1418,
+      "step": 12962
+    },
+    {
+      "epoch": 19.40568862275449,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 0.0008,
+      "loss": 1.1361,
+      "step": 12963
+    },
+    {
+      "epoch": 19.407185628742514,
+      "grad_norm": 0.08984375,
+      "learning_rate": 0.0008,
+      "loss": 1.1343,
+      "step": 12964
+    },
+    {
+      "epoch": 19.40868263473054,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 0.0008,
+      "loss": 1.1425,
+      "step": 12965
+    },
+    {
+      "epoch": 19.410179640718564,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1323,
+      "step": 12966
+    },
+    {
+      "epoch": 19.411676646706585,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1365,
+      "step": 12967
+    },
+    {
+      "epoch": 19.41317365269461,
+      "grad_norm": 0.09765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1372,
+      "step": 12968
+    },
+    {
+      "epoch": 19.414670658682635,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 0.0008,
+      "loss": 1.1469,
+      "step": 12969
+    },
+    {
+      "epoch": 19.41616766467066,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1364,
+      "step": 12970
+    },
+    {
+      "epoch": 19.41766467065868,
+      "grad_norm": 0.072265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1402,
+      "step": 12971
+    },
+    {
+      "epoch": 19.419161676646706,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.137,
+      "step": 12972
+    },
+    {
+      "epoch": 19.42065868263473,
+      "grad_norm": 0.06884765625,
+      "learning_rate": 0.0008,
+      "loss": 1.1302,
+      "step": 12973
+    },
+    {
+      "epoch": 19.422155688622755,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1397,
+      "step": 12974
+    },
+    {
+      "epoch": 19.42365269461078,
+      "grad_norm": 0.068359375,
+      "learning_rate": 0.0008,
+      "loss": 1.1333,
+      "step": 12975
+    },
+    {
+      "epoch": 19.4251497005988,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 12976
+    },
+    {
+      "epoch": 19.426646706586826,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1344,
+      "step": 12977
+    },
+    {
+      "epoch": 19.42814371257485,
+      "grad_norm": 0.0654296875,
+      "learning_rate": 0.0008,
+      "loss": 1.1354,
+      "step": 12978
+    },
+    {
+      "epoch": 19.429640718562876,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.1338,
+      "step": 12979
+    },
+    {
+      "epoch": 19.431137724550897,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1398,
+      "step": 12980
+    },
+    {
+      "epoch": 19.432634730538922,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1393,
+      "step": 12981
+    },
+    {
+      "epoch": 19.434131736526947,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 0.0008,
+      "loss": 1.1395,
+      "step": 12982
+    },
+    {
+      "epoch": 19.43562874251497,
+      "grad_norm": 0.06689453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1431,
+      "step": 12983
+    },
+    {
+      "epoch": 19.437125748502993,
+      "grad_norm": 0.07421875,
+      "learning_rate": 0.0008,
+      "loss": 1.14,
+      "step": 12984
+    },
+    {
+      "epoch": 19.438622754491018,
+      "grad_norm": 0.064453125,
+      "learning_rate": 0.0008,
+      "loss": 1.1405,
+      "step": 12985
+    },
+    {
+      "epoch": 19.440119760479043,
+      "grad_norm": 0.06640625,
+      "learning_rate": 0.0008,
+      "loss": 1.1347,
+      "step": 12986
+    },
+    {
+      "epoch": 19.441616766467067,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 0.0008,
+      "loss": 1.1354,
+      "step": 12987
+    },
+    {
+      "epoch": 19.44311377245509,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1404,
+      "step": 12988
+    },
+    {
+      "epoch": 19.444610778443113,
+      "grad_norm": 0.0859375,
+      "learning_rate": 0.0008,
+      "loss": 1.1392,
+      "step": 12989
+    },
+    {
+      "epoch": 19.44610778443114,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 0.0008,
+      "loss": 1.1336,
+      "step": 12990
+    },
+    {
+      "epoch": 19.447604790419163,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 0.0008,
+      "loss": 1.131,
+      "step": 12991
+    },
+    {
+      "epoch": 19.449101796407184,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 0.0008,
+      "loss": 1.1363,
+      "step": 12992
+    },
+    {
+      "epoch": 19.45059880239521,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1356,
+      "step": 12993
+    },
+    {
+      "epoch": 19.452095808383234,
+      "grad_norm": 0.07470703125,
+      "learning_rate": 0.0008,
+      "loss": 1.1414,
+      "step": 12994
+    },
+    {
+      "epoch": 19.45359281437126,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 0.0008,
+      "loss": 1.1437,
+      "step": 12995
+    },
+    {
+      "epoch": 19.45508982035928,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 0.0008,
+      "loss": 1.1347,
+      "step": 12996
+    },
+    {
+      "epoch": 19.456586826347305,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1349,
+      "step": 12997
+    },
+    {
+      "epoch": 19.45808383233533,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 0.0008,
+      "loss": 1.1314,
+      "step": 12998
+    },
+    {
+      "epoch": 19.459580838323355,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 0.0008,
+      "loss": 1.1362,
+      "step": 12999
+    },
+    {
+      "epoch": 19.461077844311376,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 0.0008,
+      "loss": 1.1371,
+      "step": 13000
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1002000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1500,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 7.753739279037104e+18,
+  "train_batch_size": 120,
+  "trial_name": null,
+  "trial_params": null
+}