diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,15549 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.997971602434077,
+  "eval_steps": 500,
+  "global_step": 2217,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.1875,
+      "learning_rate": 7.462686567164179e-08,
+      "loss": 0.4942,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.1875,
+      "learning_rate": 1.4925373134328358e-07,
+      "loss": 0.5581,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.9375,
+      "learning_rate": 2.2388059701492537e-07,
+      "loss": 0.5681,
+      "step": 3
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.25,
+      "learning_rate": 2.9850746268656716e-07,
+      "loss": 0.5357,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.4375,
+      "learning_rate": 3.7313432835820895e-07,
+      "loss": 0.598,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.4375,
+      "learning_rate": 4.4776119402985074e-07,
+      "loss": 0.5585,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.84375,
+      "learning_rate": 5.223880597014925e-07,
+      "loss": 0.5029,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.375,
+      "learning_rate": 5.970149253731343e-07,
+      "loss": 0.5416,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.0,
+      "learning_rate": 6.716417910447762e-07,
+      "loss": 0.5441,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.4375,
+      "learning_rate": 7.462686567164179e-07,
+      "loss": 0.5095,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.65625,
+      "learning_rate": 8.208955223880598e-07,
+      "loss": 0.5805,
+      "step": 11
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.375,
+      "learning_rate": 8.955223880597015e-07,
+      "loss": 0.523,
+      "step": 12
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.0625,
+      "learning_rate": 9.701492537313434e-07,
+      "loss": 0.5647,
+      "step": 13
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.40625,
+      "learning_rate": 1.044776119402985e-06,
+      "loss": 0.4831,
+      "step": 14
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.71875,
+      "learning_rate": 1.119402985074627e-06,
+      "loss": 0.5457,
+      "step": 15
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.03125,
+      "learning_rate": 1.1940298507462686e-06,
+      "loss": 0.5563,
+      "step": 16
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.8125,
+      "learning_rate": 1.2686567164179105e-06,
+      "loss": 0.4933,
+      "step": 17
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.21875,
+      "learning_rate": 1.3432835820895524e-06,
+      "loss": 0.5024,
+      "step": 18
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.375,
+      "learning_rate": 1.417910447761194e-06,
+      "loss": 0.5899,
+      "step": 19
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.9375,
+      "learning_rate": 1.4925373134328358e-06,
+      "loss": 0.5974,
+      "step": 20
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.84375,
+      "learning_rate": 1.5671641791044779e-06,
+      "loss": 0.5633,
+      "step": 21
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.53125,
+      "learning_rate": 1.6417910447761196e-06,
+      "loss": 0.5673,
+      "step": 22
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.921875,
+      "learning_rate": 1.7164179104477613e-06,
+      "loss": 0.4681,
+      "step": 23
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.34375,
+      "learning_rate": 1.791044776119403e-06,
+      "loss": 0.5022,
+      "step": 24
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.21875,
+      "learning_rate": 1.865671641791045e-06,
+      "loss": 0.5051,
+      "step": 25
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.9402985074626867e-06,
+      "loss": 0.5223,
+      "step": 26
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.875,
+      "learning_rate": 2.0149253731343284e-06,
+      "loss": 0.5641,
+      "step": 27
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.96875,
+      "learning_rate": 2.08955223880597e-06,
+      "loss": 0.5895,
+      "step": 28
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.5,
+      "learning_rate": 2.1641791044776118e-06,
+      "loss": 0.5748,
+      "step": 29
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.0625,
+      "learning_rate": 2.238805970149254e-06,
+      "loss": 0.4378,
+      "step": 30
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.75,
+      "learning_rate": 2.3134328358208956e-06,
+      "loss": 0.5653,
+      "step": 31
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.28125,
+      "learning_rate": 2.3880597014925373e-06,
+      "loss": 0.5975,
+      "step": 32
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.0625,
+      "learning_rate": 2.4626865671641794e-06,
+      "loss": 0.5144,
+      "step": 33
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.9375,
+      "learning_rate": 2.537313432835821e-06,
+      "loss": 0.521,
+      "step": 34
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.03125,
+      "learning_rate": 2.6119402985074627e-06,
+      "loss": 0.4909,
+      "step": 35
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.0625,
+      "learning_rate": 2.686567164179105e-06,
+      "loss": 0.5472,
+      "step": 36
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.921875,
+      "learning_rate": 2.7611940298507465e-06,
+      "loss": 0.5314,
+      "step": 37
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.125,
+      "learning_rate": 2.835820895522388e-06,
+      "loss": 0.5836,
+      "step": 38
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.796875,
+      "learning_rate": 2.9104477611940303e-06,
+      "loss": 0.4819,
+      "step": 39
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.953125,
+      "learning_rate": 2.9850746268656716e-06,
+      "loss": 0.54,
+      "step": 40
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.859375,
+      "learning_rate": 3.0597014925373137e-06,
+      "loss": 0.5376,
+      "step": 41
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.3125,
+      "learning_rate": 3.1343283582089558e-06,
+      "loss": 0.5207,
+      "step": 42
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.125,
+      "learning_rate": 3.208955223880597e-06,
+      "loss": 0.5241,
+      "step": 43
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.0625,
+      "learning_rate": 3.283582089552239e-06,
+      "loss": 0.5604,
+      "step": 44
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.0625,
+      "learning_rate": 3.3582089552238813e-06,
+      "loss": 0.5929,
+      "step": 45
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.25,
+      "learning_rate": 3.4328358208955225e-06,
+      "loss": 0.5044,
+      "step": 46
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.09375,
+      "learning_rate": 3.5074626865671646e-06,
+      "loss": 0.508,
+      "step": 47
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.03125,
+      "learning_rate": 3.582089552238806e-06,
+      "loss": 0.5145,
+      "step": 48
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.890625,
+      "learning_rate": 3.656716417910448e-06,
+      "loss": 0.4583,
+      "step": 49
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.890625,
+      "learning_rate": 3.73134328358209e-06,
+      "loss": 0.4811,
+      "step": 50
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.875,
+      "learning_rate": 3.8059701492537314e-06,
+      "loss": 0.5278,
+      "step": 51
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.984375,
+      "learning_rate": 3.8805970149253735e-06,
+      "loss": 0.5448,
+      "step": 52
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.859375,
+      "learning_rate": 3.955223880597015e-06,
+      "loss": 0.5159,
+      "step": 53
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.75,
+      "learning_rate": 4.029850746268657e-06,
+      "loss": 0.5025,
+      "step": 54
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.104477611940299e-06,
+      "loss": 0.5253,
+      "step": 55
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.953125,
+      "learning_rate": 4.17910447761194e-06,
+      "loss": 0.4619,
+      "step": 56
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.253731343283583e-06,
+      "loss": 0.5459,
+      "step": 57
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.75,
+      "learning_rate": 4.3283582089552236e-06,
+      "loss": 0.5172,
+      "step": 58
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.21875,
+      "learning_rate": 4.402985074626866e-06,
+      "loss": 0.5367,
+      "step": 59
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.477611940298508e-06,
+      "loss": 0.5485,
+      "step": 60
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.0,
+      "learning_rate": 4.5522388059701495e-06,
+      "loss": 0.5143,
+      "step": 61
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.626865671641791e-06,
+      "loss": 0.5031,
+      "step": 62
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.701492537313434e-06,
+      "loss": 0.5039,
+      "step": 63
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.7761194029850745e-06,
+      "loss": 0.577,
+      "step": 64
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.850746268656717e-06,
+      "loss": 0.503,
+      "step": 65
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.625,
+      "learning_rate": 4.925373134328359e-06,
+      "loss": 0.5953,
+      "step": 66
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.84375,
+      "learning_rate": 5e-06,
+      "loss": 0.5618,
+      "step": 67
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.859375,
+      "learning_rate": 4.999997331097176e-06,
+      "loss": 0.5523,
+      "step": 68
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.999989324394401e-06,
+      "loss": 0.5126,
+      "step": 69
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.34375,
+      "learning_rate": 4.99997597990877e-06,
+      "loss": 0.4954,
+      "step": 70
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.999957297668777e-06,
+      "loss": 0.5012,
+      "step": 71
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.0,
+      "learning_rate": 4.999933277714309e-06,
+      "loss": 0.494,
+      "step": 72
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.953125,
+      "learning_rate": 4.999903920096652e-06,
+      "loss": 0.5878,
+      "step": 73
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.875,
+      "learning_rate": 4.999869224878489e-06,
+      "loss": 0.5563,
+      "step": 74
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.999829192133897e-06,
+      "loss": 0.5665,
+      "step": 75
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.953125,
+      "learning_rate": 4.999783821948353e-06,
+      "loss": 0.5347,
+      "step": 76
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.999733114418726e-06,
+      "loss": 0.5604,
+      "step": 77
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.999677069653283e-06,
+      "loss": 0.5279,
+      "step": 78
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.999615687771687e-06,
+      "loss": 0.5789,
+      "step": 79
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.890625,
+      "learning_rate": 4.9995489689049945e-06,
+      "loss": 0.5111,
+      "step": 80
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.999476913195661e-06,
+      "loss": 0.5252,
+      "step": 81
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.578125,
+      "learning_rate": 4.999399520797532e-06,
+      "loss": 0.4542,
+      "step": 82
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.999316791875851e-06,
+      "loss": 0.5271,
+      "step": 83
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.999228726607254e-06,
+      "loss": 0.4904,
+      "step": 84
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.4375,
+      "learning_rate": 4.99913532517977e-06,
+      "loss": 0.589,
+      "step": 85
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.999036587792824e-06,
+      "loss": 0.5267,
+      "step": 86
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.953125,
+      "learning_rate": 4.998932514657232e-06,
+      "loss": 0.4813,
+      "step": 87
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.998823105995203e-06,
+      "loss": 0.4574,
+      "step": 88
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.75,
+      "learning_rate": 4.998708362040336e-06,
+      "loss": 0.5086,
+      "step": 89
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.21875,
+      "learning_rate": 4.998588283037627e-06,
+      "loss": 0.5696,
+      "step": 90
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.998462869243456e-06,
+      "loss": 0.4705,
+      "step": 91
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.4375,
+      "learning_rate": 4.998332120925598e-06,
+      "loss": 0.5489,
+      "step": 92
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.859375,
+      "learning_rate": 4.998196038363218e-06,
+      "loss": 0.5079,
+      "step": 93
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.90625,
+      "learning_rate": 4.998054621846867e-06,
+      "loss": 0.5419,
+      "step": 94
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.997907871678486e-06,
+      "loss": 0.5058,
+      "step": 95
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.997755788171408e-06,
+      "loss": 0.5138,
+      "step": 96
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.375,
+      "learning_rate": 4.997598371650346e-06,
+      "loss": 0.5563,
+      "step": 97
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.9974356224514055e-06,
+      "loss": 0.5368,
+      "step": 98
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.15625,
+      "learning_rate": 4.997267540922076e-06,
+      "loss": 0.5293,
+      "step": 99
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.997094127421231e-06,
+      "loss": 0.5481,
+      "step": 100
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.996915382319131e-06,
+      "loss": 0.5812,
+      "step": 101
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.996731305997416e-06,
+      "loss": 0.5915,
+      "step": 102
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.9965418988491155e-06,
+      "loss": 0.5004,
+      "step": 103
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.25,
+      "learning_rate": 4.996347161278634e-06,
+      "loss": 0.5518,
+      "step": 104
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.996147093701761e-06,
+      "loss": 0.5681,
+      "step": 105
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.21875,
+      "learning_rate": 4.995941696545664e-06,
+      "loss": 0.591,
+      "step": 106
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.859375,
+      "learning_rate": 4.995730970248893e-06,
+      "loss": 0.5224,
+      "step": 107
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.921875,
+      "learning_rate": 4.995514915261373e-06,
+      "loss": 0.4931,
+      "step": 108
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.828125,
+      "learning_rate": 4.995293532044408e-06,
+      "loss": 0.5272,
+      "step": 109
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.9950668210706795e-06,
+      "loss": 0.5117,
+      "step": 110
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.994834782824242e-06,
+      "loss": 0.5456,
+      "step": 111
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.994597417800524e-06,
+      "loss": 0.5047,
+      "step": 112
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.953125,
+      "learning_rate": 4.994354726506332e-06,
+      "loss": 0.4668,
+      "step": 113
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.21875,
+      "learning_rate": 4.99410670945984e-06,
+      "loss": 0.5171,
+      "step": 114
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.859375,
+      "learning_rate": 4.993853367190594e-06,
+      "loss": 0.534,
+      "step": 115
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.75,
+      "learning_rate": 4.993594700239513e-06,
+      "loss": 0.5669,
+      "step": 116
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.993330709158879e-06,
+      "loss": 0.5304,
+      "step": 117
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.34375,
+      "learning_rate": 4.993061394512349e-06,
+      "loss": 0.5318,
+      "step": 118
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.992786756874941e-06,
+      "loss": 0.5175,
+      "step": 119
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.992506796833039e-06,
+      "loss": 0.5755,
+      "step": 120
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.28125,
+      "learning_rate": 4.992221514984393e-06,
+      "loss": 0.5288,
+      "step": 121
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.859375,
+      "learning_rate": 4.991930911938116e-06,
+      "loss": 0.4815,
+      "step": 122
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.890625,
+      "learning_rate": 4.991634988314678e-06,
+      "loss": 0.536,
+      "step": 123
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.991333744745913e-06,
+      "loss": 0.5209,
+      "step": 124
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.96875,
+      "learning_rate": 4.991027181875014e-06,
+      "loss": 0.6006,
+      "step": 125
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.8125,
+      "learning_rate": 4.990715300356531e-06,
+      "loss": 0.5317,
+      "step": 126
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.990398100856367e-06,
+      "loss": 0.5373,
+      "step": 127
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.990075584051782e-06,
+      "loss": 0.5577,
+      "step": 128
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.953125,
+      "learning_rate": 4.989747750631391e-06,
+      "loss": 0.5615,
+      "step": 129
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.15625,
+      "learning_rate": 4.989414601295155e-06,
+      "loss": 0.5521,
+      "step": 130
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.21875,
+      "learning_rate": 4.989076136754391e-06,
+      "loss": 0.5626,
+      "step": 131
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.9887323577317625e-06,
+      "loss": 0.4784,
+      "step": 132
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.953125,
+      "learning_rate": 4.988383264961278e-06,
+      "loss": 0.5678,
+      "step": 133
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.988028859188294e-06,
+      "loss": 0.5094,
+      "step": 134
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.98766914116951e-06,
+      "loss": 0.4817,
+      "step": 135
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.987304111672969e-06,
+      "loss": 0.4925,
+      "step": 136
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.986933771478052e-06,
+      "loss": 0.5424,
+      "step": 137
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.0,
+      "learning_rate": 4.986558121375482e-06,
+      "loss": 0.5655,
+      "step": 138
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.96875,
+      "learning_rate": 4.986177162167317e-06,
+      "loss": 0.4402,
+      "step": 139
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.985790894666951e-06,
+      "loss": 0.505,
+      "step": 140
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.90625,
+      "learning_rate": 4.985399319699114e-06,
+      "loss": 0.5064,
+      "step": 141
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.21875,
+      "learning_rate": 4.9850024380998655e-06,
+      "loss": 0.542,
+      "step": 142
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.984600250716596e-06,
+      "loss": 0.5066,
+      "step": 143
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.34375,
+      "learning_rate": 4.984192758408025e-06,
+      "loss": 0.5865,
+      "step": 144
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.983779962044198e-06,
+      "loss": 0.5004,
+      "step": 145
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.96875,
+      "learning_rate": 4.983361862506486e-06,
+      "loss": 0.5478,
+      "step": 146
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.982938460687583e-06,
+      "loss": 0.5142,
+      "step": 147
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.9825097574915034e-06,
+      "loss": 0.4945,
+      "step": 148
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.982075753833581e-06,
+      "loss": 0.5701,
+      "step": 149
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.953125,
+      "learning_rate": 4.981636450640467e-06,
+      "loss": 0.5201,
+      "step": 150
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.8125,
+      "learning_rate": 4.981191848850126e-06,
+      "loss": 0.4863,
+      "step": 151
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.96875,
+      "learning_rate": 4.980741949411839e-06,
+      "loss": 0.5533,
+      "step": 152
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.980286753286196e-06,
+      "loss": 0.5055,
+      "step": 153
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.28125,
+      "learning_rate": 4.979826261445094e-06,
+      "loss": 0.5153,
+      "step": 154
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.15625,
+      "learning_rate": 4.979360474871742e-06,
+      "loss": 0.5173,
+      "step": 155
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.978889394560651e-06,
+      "loss": 0.5037,
+      "step": 156
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.978413021517634e-06,
+      "loss": 0.4972,
+      "step": 157
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.977931356759806e-06,
+      "loss": 0.5616,
+      "step": 158
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.977444401315582e-06,
+      "loss": 0.5372,
+      "step": 159
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.953125,
+      "learning_rate": 4.976952156224668e-06,
+      "loss": 0.4757,
+      "step": 160
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.96875,
+      "learning_rate": 4.9764546225380705e-06,
+      "loss": 0.5059,
+      "step": 161
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.671875,
+      "learning_rate": 4.975951801318083e-06,
+      "loss": 0.4689,
+      "step": 162
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.859375,
+      "learning_rate": 4.975443693638291e-06,
+      "loss": 0.4836,
+      "step": 163
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.9749303005835665e-06,
+      "loss": 0.437,
+      "step": 164
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.974411623250066e-06,
+      "loss": 0.5327,
+      "step": 165
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.973887662745229e-06,
+      "loss": 0.5484,
+      "step": 166
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.973358420187776e-06,
+      "loss": 0.5706,
+      "step": 167
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.972823896707704e-06,
+      "loss": 0.5354,
+      "step": 168
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.972284093446286e-06,
+      "loss": 0.5472,
+      "step": 169
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.8125,
+      "learning_rate": 4.971739011556067e-06,
+      "loss": 0.5017,
+      "step": 170
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.859375,
+      "learning_rate": 4.971188652200865e-06,
+      "loss": 0.5389,
+      "step": 171
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.970633016555765e-06,
+      "loss": 0.504,
+      "step": 172
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.970072105807116e-06,
+      "loss": 0.4988,
+      "step": 173
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.875,
+      "learning_rate": 4.9695059211525305e-06,
+      "loss": 0.5645,
+      "step": 174
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.125,
+      "learning_rate": 4.9689344638008825e-06,
+      "loss": 0.4706,
+      "step": 175
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.828125,
+      "learning_rate": 4.968357734972303e-06,
+      "loss": 0.5677,
+      "step": 176
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.828125,
+      "learning_rate": 4.967775735898179e-06,
+      "loss": 0.5555,
+      "step": 177
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.96718846782115e-06,
+      "loss": 0.5455,
+      "step": 178
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.966595931995105e-06,
+      "loss": 0.5248,
+      "step": 179
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.34375,
+      "learning_rate": 4.965998129685179e-06,
+      "loss": 0.539,
+      "step": 180
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.875,
+      "learning_rate": 4.965395062167756e-06,
+      "loss": 0.5264,
+      "step": 181
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.125,
+      "learning_rate": 4.964786730730455e-06,
+      "loss": 0.4891,
+      "step": 182
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.964173136672142e-06,
+      "loss": 0.4629,
+      "step": 183
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.8125,
+      "learning_rate": 4.9635542813029115e-06,
+      "loss": 0.5626,
+      "step": 184
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.125,
+      "learning_rate": 4.962930165944098e-06,
+      "loss": 0.4849,
+      "step": 185
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.962300791928263e-06,
+      "loss": 0.5057,
+      "step": 186
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.890625,
+      "learning_rate": 4.961666160599198e-06,
+      "loss": 0.5327,
+      "step": 187
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.921875,
+      "learning_rate": 4.961026273311918e-06,
+      "loss": 0.5671,
+      "step": 188
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.703125,
+      "learning_rate": 4.96038113143266e-06,
+      "loss": 0.528,
+      "step": 189
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.921875,
+      "learning_rate": 4.95973073633888e-06,
+      "loss": 0.5385,
+      "step": 190
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.9590750894192545e-06,
+      "loss": 0.4935,
+      "step": 191
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.958414192073665e-06,
+      "loss": 0.5222,
+      "step": 192
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.875,
+      "learning_rate": 4.9577480457132124e-06,
+      "loss": 0.5057,
+      "step": 193
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.15625,
+      "learning_rate": 4.957076651760199e-06,
+      "loss": 0.4899,
+      "step": 194
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.953125,
+      "learning_rate": 4.956400011648132e-06,
+      "loss": 0.539,
+      "step": 195
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.9557181268217225e-06,
+      "loss": 0.4896,
+      "step": 196
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.828125,
+      "learning_rate": 4.955030998736876e-06,
+      "loss": 0.4911,
+      "step": 197
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.546875,
+      "learning_rate": 4.9543386288606965e-06,
+      "loss": 0.5125,
+      "step": 198
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.953641018671478e-06,
+      "loss": 0.543,
+      "step": 199
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.859375,
+      "learning_rate": 4.952938169658704e-06,
+      "loss": 0.5118,
+      "step": 200
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.952230083323042e-06,
+      "loss": 0.5169,
+      "step": 201
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.951516761176344e-06,
+      "loss": 0.4962,
+      "step": 202
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.950798204741638e-06,
+      "loss": 0.5818,
+      "step": 203
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.8125,
+      "learning_rate": 4.950074415553133e-06,
+      "loss": 0.4994,
+      "step": 204
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.9493453951562045e-06,
+      "loss": 0.535,
+      "step": 205
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.859375,
+      "learning_rate": 4.948611145107402e-06,
+      "loss": 0.5219,
+      "step": 206
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.890625,
+      "learning_rate": 4.947871666974438e-06,
+      "loss": 0.5195,
+      "step": 207
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.125,
+      "learning_rate": 4.94712696233619e-06,
+      "loss": 0.5293,
+      "step": 208
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.3125,
+      "learning_rate": 4.946377032782691e-06,
+      "loss": 0.5578,
+      "step": 209
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.125,
+      "learning_rate": 4.945621879915135e-06,
+      "loss": 0.5218,
+      "step": 210
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.944861505345865e-06,
+      "loss": 0.5103,
+      "step": 211
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.0,
+      "learning_rate": 4.944095910698372e-06,
+      "loss": 0.5631,
+      "step": 212
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.943325097607297e-06,
+      "loss": 0.5044,
+      "step": 213
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.942549067718418e-06,
+      "loss": 0.4896,
+      "step": 214
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.941767822688653e-06,
+      "loss": 0.5014,
+      "step": 215
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.59375,
+      "learning_rate": 4.94098136418606e-06,
+      "loss": 0.4664,
+      "step": 216
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.90625,
+      "learning_rate": 4.940189693889819e-06,
+      "loss": 0.4594,
+      "step": 217
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.939392813490245e-06,
+      "loss": 0.5002,
+      "step": 218
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.25,
+      "learning_rate": 4.938590724688774e-06,
+      "loss": 0.5415,
+      "step": 219
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.3125,
+      "learning_rate": 4.937783429197966e-06,
+      "loss": 0.5,
+      "step": 220
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.34375,
+      "learning_rate": 4.936970928741494e-06,
+      "loss": 0.5639,
+      "step": 221
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.75,
+      "learning_rate": 4.936153225054147e-06,
+      "loss": 0.4749,
+      "step": 222
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.93533031988182e-06,
+      "loss": 0.5596,
+      "step": 223
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.934502214981518e-06,
+      "loss": 0.5568,
+      "step": 224
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.703125,
+      "learning_rate": 4.933668912121346e-06,
+      "loss": 0.5592,
+      "step": 225
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.932830413080507e-06,
+      "loss": 0.5039,
+      "step": 226
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.75,
+      "learning_rate": 4.931986719649298e-06,
+      "loss": 0.436,
+      "step": 227
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.931137833629111e-06,
+      "loss": 0.5774,
+      "step": 228
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.930283756832418e-06,
+      "loss": 0.5137,
+      "step": 229
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.9294244910827795e-06,
+      "loss": 0.4677,
+      "step": 230
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.921875,
+      "learning_rate": 4.928560038214831e-06,
+      "loss": 0.4982,
+      "step": 231
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.927690400074287e-06,
+      "loss": 0.4999,
+      "step": 232
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.92681557851793e-06,
+      "loss": 0.4383,
+      "step": 233
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.890625,
+      "learning_rate": 4.925935575413611e-06,
+      "loss": 0.5272,
+      "step": 234
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.925050392640245e-06,
+      "loss": 0.5277,
+      "step": 235
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.924160032087804e-06,
+      "loss": 0.5273,
+      "step": 236
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.875,
+      "learning_rate": 4.923264495657319e-06,
+      "loss": 0.5184,
+      "step": 237
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.922363785260868e-06,
+      "loss": 0.5405,
+      "step": 238
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.96875,
+      "learning_rate": 4.921457902821578e-06,
+      "loss": 0.5141,
+      "step": 239
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.859375,
+      "learning_rate": 4.920546850273619e-06,
+      "loss": 0.482,
+      "step": 240
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.828125,
+      "learning_rate": 4.9196306295622e-06,
+      "loss": 0.485,
+      "step": 241
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.918709242643563e-06,
+      "loss": 0.5076,
+      "step": 242
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.917782691484984e-06,
+      "loss": 0.5465,
+      "step": 243
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.75,
+      "learning_rate": 4.91685097806476e-06,
+      "loss": 0.4651,
+      "step": 244
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.953125,
+      "learning_rate": 4.915914104372215e-06,
+      "loss": 0.5495,
+      "step": 245
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.0,
+      "learning_rate": 4.914972072407689e-06,
+      "loss": 0.5026,
+      "step": 246
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.914024884182535e-06,
+      "loss": 0.4651,
+      "step": 247
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.71875,
+      "learning_rate": 4.913072541719115e-06,
+      "loss": 0.489,
+      "step": 248
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.912115047050796e-06,
+      "loss": 0.4921,
+      "step": 249
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.911152402221949e-06,
+      "loss": 0.5128,
+      "step": 250
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.0,
+      "learning_rate": 4.910184609287936e-06,
+      "loss": 0.5323,
+      "step": 251
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.21875,
+      "learning_rate": 4.909211670315115e-06,
+      "loss": 0.5185,
+      "step": 252
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.921875,
+      "learning_rate": 4.9082335873808275e-06,
+      "loss": 0.5404,
+      "step": 253
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.907250362573401e-06,
+      "loss": 0.5331,
+      "step": 254
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.96875,
+      "learning_rate": 4.906261997992141e-06,
+      "loss": 0.537,
+      "step": 255
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.703125,
+      "learning_rate": 4.905268495747327e-06,
+      "loss": 0.5155,
+      "step": 256
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.904269857960208e-06,
+      "loss": 0.4854,
+      "step": 257
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.903266086762996e-06,
+      "loss": 0.4975,
+      "step": 258
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.828125,
+      "learning_rate": 4.902257184298867e-06,
+      "loss": 0.538,
+      "step": 259
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.901243152721952e-06,
+      "loss": 0.4607,
+      "step": 260
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.0,
+      "learning_rate": 4.900223994197329e-06,
+      "loss": 0.5058,
+      "step": 261
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.53125,
+      "learning_rate": 4.899199710901028e-06,
+      "loss": 0.5602,
+      "step": 262
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.21875,
+      "learning_rate": 4.89817030502002e-06,
+      "loss": 0.5226,
+      "step": 263
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.89713577875221e-06,
+      "loss": 0.5302,
+      "step": 264
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.96875,
+      "learning_rate": 4.896096134306441e-06,
+      "loss": 0.5005,
+      "step": 265
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.75,
+      "learning_rate": 4.8950513739024785e-06,
+      "loss": 0.5418,
+      "step": 266
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.96875,
+      "learning_rate": 4.894001499771015e-06,
+      "loss": 0.5332,
+      "step": 267
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.96875,
+      "learning_rate": 4.89294651415366e-06,
+      "loss": 0.6134,
+      "step": 268
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.859375,
+      "learning_rate": 4.891886419302937e-06,
+      "loss": 0.4403,
+      "step": 269
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.125,
+      "learning_rate": 4.890821217482277e-06,
+      "loss": 0.5088,
+      "step": 270
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.889750910966017e-06,
+      "loss": 0.522,
+      "step": 271
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.888675502039392e-06,
+      "loss": 0.4872,
+      "step": 272
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.921875,
+      "learning_rate": 4.887594992998531e-06,
+      "loss": 0.5219,
+      "step": 273
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.890625,
+      "learning_rate": 4.886509386150454e-06,
+      "loss": 0.5542,
+      "step": 274
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.859375,
+      "learning_rate": 4.885418683813064e-06,
+      "loss": 0.4613,
+      "step": 275
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.90625,
+      "learning_rate": 4.8843228883151445e-06,
+      "loss": 0.5691,
+      "step": 276
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.883222001996352e-06,
+      "loss": 0.4812,
+      "step": 277
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.882116027207212e-06,
+      "loss": 0.4851,
+      "step": 278
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.875,
+      "learning_rate": 4.881004966309119e-06,
+      "loss": 0.4338,
+      "step": 279
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.879888821674321e-06,
+      "loss": 0.4589,
+      "step": 280
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.878767595685925e-06,
+      "loss": 0.4923,
+      "step": 281
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.0,
+      "learning_rate": 4.8776412907378845e-06,
+      "loss": 0.4878,
+      "step": 282
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.75,
+      "learning_rate": 4.876509909234999e-06,
+      "loss": 0.5518,
+      "step": 283
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.15625,
+      "learning_rate": 4.875373453592906e-06,
+      "loss": 0.5756,
+      "step": 284
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.0,
+      "learning_rate": 4.874231926238078e-06,
+      "loss": 0.4641,
+      "step": 285
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.875,
+      "learning_rate": 4.873085329607814e-06,
+      "loss": 0.5353,
+      "step": 286
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.96875,
+      "learning_rate": 4.871933666150239e-06,
+      "loss": 0.5336,
+      "step": 287
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.870776938324296e-06,
+      "loss": 0.5476,
+      "step": 288
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.125,
+      "learning_rate": 4.8696151485997385e-06,
+      "loss": 0.5649,
+      "step": 289
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.868448299457132e-06,
+      "loss": 0.6093,
+      "step": 290
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.867276393387839e-06,
+      "loss": 0.5126,
+      "step": 291
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.866099432894023e-06,
+      "loss": 0.453,
+      "step": 292
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.0,
+      "learning_rate": 4.864917420488642e-06,
+      "loss": 0.4362,
+      "step": 293
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.375,
+      "learning_rate": 4.863730358695433e-06,
+      "loss": 0.5463,
+      "step": 294
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.86253825004892e-06,
+      "loss": 0.5472,
+      "step": 295
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.861341097094398e-06,
+      "loss": 0.5867,
+      "step": 296
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.8601389023879395e-06,
+      "loss": 0.5247,
+      "step": 297
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.921875,
+      "learning_rate": 4.858931668496374e-06,
+      "loss": 0.5014,
+      "step": 298
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.857719397997293e-06,
+      "loss": 0.5465,
+      "step": 299
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.856502093479046e-06,
+      "loss": 0.5226,
+      "step": 300
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.921875,
+      "learning_rate": 4.855279757540723e-06,
+      "loss": 0.5286,
+      "step": 301
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.854052392792162e-06,
+      "loss": 0.4559,
+      "step": 302
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.40625,
+      "learning_rate": 4.852820001853937e-06,
+      "loss": 0.6101,
+      "step": 303
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.851582587357353e-06,
+      "loss": 0.5135,
+      "step": 304
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.828125,
+      "learning_rate": 4.850340151944441e-06,
+      "loss": 0.5706,
+      "step": 305
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.849092698267954e-06,
+      "loss": 0.4993,
+      "step": 306
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.8478402289913566e-06,
+      "loss": 0.4933,
+      "step": 307
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.15625,
+      "learning_rate": 4.846582746788824e-06,
+      "loss": 0.5252,
+      "step": 308
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.5625,
+      "learning_rate": 4.845320254345235e-06,
+      "loss": 0.5304,
+      "step": 309
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.125,
+      "learning_rate": 4.844052754356166e-06,
+      "loss": 0.4609,
+      "step": 310
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.842780249527882e-06,
+      "loss": 0.5093,
+      "step": 311
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.841502742577339e-06,
+      "loss": 0.5772,
+      "step": 312
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.840220236232168e-06,
+      "loss": 0.4823,
+      "step": 313
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.0,
+      "learning_rate": 4.838932733230679e-06,
+      "loss": 0.5399,
+      "step": 314
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.837640236321848e-06,
+      "loss": 0.4958,
+      "step": 315
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.836342748265314e-06,
+      "loss": 0.3872,
+      "step": 316
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.835040271831371e-06,
+      "loss": 0.494,
+      "step": 317
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.25,
+      "learning_rate": 4.833732809800966e-06,
+      "loss": 0.4818,
+      "step": 318
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.832420364965692e-06,
+      "loss": 0.503,
+      "step": 319
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.90625,
+      "learning_rate": 4.831102940127778e-06,
+      "loss": 0.5298,
+      "step": 320
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.875,
+      "learning_rate": 4.8297805381000865e-06,
+      "loss": 0.5445,
+      "step": 321
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.859375,
+      "learning_rate": 4.828453161706108e-06,
+      "loss": 0.4851,
+      "step": 322
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.827120813779954e-06,
+      "loss": 0.4849,
+      "step": 323
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.921875,
+      "learning_rate": 4.825783497166349e-06,
+      "loss": 0.5092,
+      "step": 324
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.824441214720629e-06,
+      "loss": 0.4534,
+      "step": 325
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.82309396930873e-06,
+      "loss": 0.6579,
+      "step": 326
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.821741763807186e-06,
+      "loss": 0.5193,
+      "step": 327
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.820384601103121e-06,
+      "loss": 0.5145,
+      "step": 328
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.890625,
+      "learning_rate": 4.819022484094243e-06,
+      "loss": 0.5975,
+      "step": 329
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.81765541568884e-06,
+      "loss": 0.4511,
+      "step": 330
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.816283398805767e-06,
+      "loss": 0.5174,
+      "step": 331
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.625,
+      "learning_rate": 4.814906436374451e-06,
+      "loss": 0.445,
+      "step": 332
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.75,
+      "learning_rate": 4.8135245313348735e-06,
+      "loss": 0.519,
+      "step": 333
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.812137686637571e-06,
+      "loss": 0.5243,
+      "step": 334
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.810745905243626e-06,
+      "loss": 0.5719,
+      "step": 335
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.3125,
+      "learning_rate": 4.809349190124663e-06,
+      "loss": 0.4999,
+      "step": 336
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.375,
+      "learning_rate": 4.807947544262838e-06,
+      "loss": 0.5104,
+      "step": 337
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.96875,
+      "learning_rate": 4.806540970650839e-06,
+      "loss": 0.4198,
+      "step": 338
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.80512947229187e-06,
+      "loss": 0.4948,
+      "step": 339
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.803713052199653e-06,
+      "loss": 0.5553,
+      "step": 340
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.96875,
+      "learning_rate": 4.8022917133984185e-06,
+      "loss": 0.4681,
+      "step": 341
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.800865458922899e-06,
+      "loss": 0.5126,
+      "step": 342
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.875,
+      "learning_rate": 4.799434291818322e-06,
+      "loss": 0.5351,
+      "step": 343
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.90625,
+      "learning_rate": 4.797998215140404e-06,
+      "loss": 0.5508,
+      "step": 344
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.859375,
+      "learning_rate": 4.796557231955343e-06,
+      "loss": 0.4793,
+      "step": 345
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.795111345339818e-06,
+      "loss": 0.4273,
+      "step": 346
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.859375,
+      "learning_rate": 4.793660558380969e-06,
+      "loss": 0.5596,
+      "step": 347
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.7922048741764075e-06,
+      "loss": 0.5531,
+      "step": 348
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.59375,
+      "learning_rate": 4.7907442958341945e-06,
+      "loss": 0.5224,
+      "step": 349
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.15625,
+      "learning_rate": 4.789278826472845e-06,
+      "loss": 0.5544,
+      "step": 350
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.90625,
+      "learning_rate": 4.787808469221315e-06,
+      "loss": 0.4534,
+      "step": 351
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.786333227218996e-06,
+      "loss": 0.4953,
+      "step": 352
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.953125,
+      "learning_rate": 4.78485310361571e-06,
+      "loss": 0.507,
+      "step": 353
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.625,
+      "learning_rate": 4.783368101571704e-06,
+      "loss": 0.5068,
+      "step": 354
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.921875,
+      "learning_rate": 4.781878224257637e-06,
+      "loss": 0.5119,
+      "step": 355
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.780383474854578e-06,
+      "loss": 0.5965,
+      "step": 356
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.7788838565540044e-06,
+      "loss": 0.4577,
+      "step": 357
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.777379372557779e-06,
+      "loss": 0.5231,
+      "step": 358
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.875,
+      "learning_rate": 4.775870026078163e-06,
+      "loss": 0.5825,
+      "step": 359
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.774355820337794e-06,
+      "loss": 0.4585,
+      "step": 360
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.75,
+      "learning_rate": 4.772836758569688e-06,
+      "loss": 0.5317,
+      "step": 361
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.921875,
+      "learning_rate": 4.771312844017224e-06,
+      "loss": 0.4843,
+      "step": 362
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.76978407993415e-06,
+      "loss": 0.5051,
+      "step": 363
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.768250469584562e-06,
+      "loss": 0.5145,
+      "step": 364
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.766712016242906e-06,
+      "loss": 0.5806,
+      "step": 365
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.765168723193968e-06,
+      "loss": 0.5373,
+      "step": 366
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.763620593732867e-06,
+      "loss": 0.5075,
+      "step": 367
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.762067631165049e-06,
+      "loss": 0.5802,
+      "step": 368
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.953125,
+      "learning_rate": 4.7605098388062795e-06,
+      "loss": 0.4953,
+      "step": 369
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.5,
+      "learning_rate": 4.758947219982635e-06,
+      "loss": 0.5063,
+      "step": 370
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.757379778030498e-06,
+      "loss": 0.4432,
+      "step": 371
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.875,
+      "learning_rate": 4.755807516296548e-06,
+      "loss": 0.4938,
+      "step": 372
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.754230438137757e-06,
+      "loss": 0.5108,
+      "step": 373
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.752648546921379e-06,
+      "loss": 0.4701,
+      "step": 374
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.0,
+      "learning_rate": 4.751061846024944e-06,
+      "loss": 0.5183,
+      "step": 375
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.546875,
+      "learning_rate": 4.749470338836256e-06,
+      "loss": 0.5374,
+      "step": 376
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.747874028753375e-06,
+      "loss": 0.4938,
+      "step": 377
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.746272919184618e-06,
+      "loss": 0.5103,
+      "step": 378
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.921875,
+      "learning_rate": 4.74466701354855e-06,
+      "loss": 0.5428,
+      "step": 379
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.890625,
+      "learning_rate": 4.743056315273976e-06,
+      "loss": 0.4903,
+      "step": 380
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.741440827799934e-06,
+      "loss": 0.5513,
+      "step": 381
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.7398205545756865e-06,
+      "loss": 0.5065,
+      "step": 382
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.738195499060715e-06,
+      "loss": 0.415,
+      "step": 383
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.921875,
+      "learning_rate": 4.736565664724713e-06,
+      "loss": 0.4985,
+      "step": 384
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.734931055047575e-06,
+      "loss": 0.4934,
+      "step": 385
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.15625,
+      "learning_rate": 4.733291673519392e-06,
+      "loss": 0.519,
+      "step": 386
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.875,
+      "learning_rate": 4.731647523640446e-06,
+      "loss": 0.5192,
+      "step": 387
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.15625,
+      "learning_rate": 4.729998608921195e-06,
+      "loss": 0.5716,
+      "step": 388
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.8125,
+      "learning_rate": 4.728344932882277e-06,
+      "loss": 0.5372,
+      "step": 389
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.96875,
+      "learning_rate": 4.72668649905449e-06,
+      "loss": 0.5741,
+      "step": 390
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.725023310978794e-06,
+      "loss": 0.4895,
+      "step": 391
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.953125,
+      "learning_rate": 4.723355372206297e-06,
+      "loss": 0.5254,
+      "step": 392
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.96875,
+      "learning_rate": 4.721682686298256e-06,
+      "loss": 0.5022,
+      "step": 393
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.828125,
+      "learning_rate": 4.720005256826056e-06,
+      "loss": 0.5052,
+      "step": 394
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.703125,
+      "learning_rate": 4.718323087371216e-06,
+      "loss": 0.5315,
+      "step": 395
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.75,
+      "learning_rate": 4.7166361815253744e-06,
+      "loss": 0.5386,
+      "step": 396
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.7149445428902794e-06,
+      "loss": 0.4749,
+      "step": 397
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.713248175077787e-06,
+      "loss": 0.4057,
+      "step": 398
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.71154708170985e-06,
+      "loss": 0.4556,
+      "step": 399
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.890625,
+      "learning_rate": 4.70984126641851e-06,
+      "loss": 0.5298,
+      "step": 400
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.90625,
+      "learning_rate": 4.708130732845893e-06,
+      "loss": 0.4741,
+      "step": 401
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.71875,
+      "learning_rate": 4.706415484644196e-06,
+      "loss": 0.5325,
+      "step": 402
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.704695525475683e-06,
+      "loss": 0.5799,
+      "step": 403
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.75,
+      "learning_rate": 4.702970859012679e-06,
+      "loss": 0.4813,
+      "step": 404
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.701241488937556e-06,
+      "loss": 0.4488,
+      "step": 405
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.0,
+      "learning_rate": 4.699507418942731e-06,
+      "loss": 0.5165,
+      "step": 406
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.71875,
+      "learning_rate": 4.697768652730656e-06,
+      "loss": 0.5,
+      "step": 407
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.696025194013809e-06,
+      "loss": 0.4896,
+      "step": 408
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.90625,
+      "learning_rate": 4.694277046514688e-06,
+      "loss": 0.4714,
+      "step": 409
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.671875,
+      "learning_rate": 4.6925242139658005e-06,
+      "loss": 0.5049,
+      "step": 410
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.828125,
+      "learning_rate": 4.690766700109659e-06,
+      "loss": 0.5042,
+      "step": 411
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.689004508698771e-06,
+      "loss": 0.5135,
+      "step": 412
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.71875,
+      "learning_rate": 4.68723764349563e-06,
+      "loss": 0.4952,
+      "step": 413
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.53125,
+      "learning_rate": 4.68546610827271e-06,
+      "loss": 0.548,
+      "step": 414
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.683689906812454e-06,
+      "loss": 0.5052,
+      "step": 415
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.681909042907269e-06,
+      "loss": 0.5376,
+      "step": 416
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.68012352035952e-06,
+      "loss": 0.5452,
+      "step": 417
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.859375,
+      "learning_rate": 4.678333342981513e-06,
+      "loss": 0.5261,
+      "step": 418
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.676538514595495e-06,
+      "loss": 0.5492,
+      "step": 419
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.674739039033648e-06,
+      "loss": 0.6001,
+      "step": 420
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.75,
+      "learning_rate": 4.672934920138069e-06,
+      "loss": 0.4281,
+      "step": 421
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.671126161760773e-06,
+      "loss": 0.5506,
+      "step": 422
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.15625,
+      "learning_rate": 4.669312767763681e-06,
+      "loss": 0.4882,
+      "step": 423
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.5,
+      "learning_rate": 4.6674947420186106e-06,
+      "loss": 0.4854,
+      "step": 424
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.665672088407268e-06,
+      "loss": 0.5522,
+      "step": 425
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.663844810821244e-06,
+      "loss": 0.5045,
+      "step": 426
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.703125,
+      "learning_rate": 4.662012913161998e-06,
+      "loss": 0.4777,
+      "step": 427
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.660176399340854e-06,
+      "loss": 0.535,
+      "step": 428
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.658335273278997e-06,
+      "loss": 0.5198,
+      "step": 429
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.656489538907453e-06,
+      "loss": 0.4893,
+      "step": 430
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.654639200167092e-06,
+      "loss": 0.5119,
+      "step": 431
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.652784261008614e-06,
+      "loss": 0.4639,
+      "step": 432
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.859375,
+      "learning_rate": 4.650924725392538e-06,
+      "loss": 0.4373,
+      "step": 433
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.625,
+      "learning_rate": 4.649060597289204e-06,
+      "loss": 0.5383,
+      "step": 434
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.875,
+      "learning_rate": 4.6471918806787516e-06,
+      "loss": 0.5234,
+      "step": 435
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.828125,
+      "learning_rate": 4.645318579551119e-06,
+      "loss": 0.5013,
+      "step": 436
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.53125,
+      "learning_rate": 4.643440697906033e-06,
+      "loss": 0.517,
+      "step": 437
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.671875,
+      "learning_rate": 4.641558239753001e-06,
+      "loss": 0.4662,
+      "step": 438
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.639671209111301e-06,
+      "loss": 0.5311,
+      "step": 439
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.0,
+      "learning_rate": 4.637779610009974e-06,
+      "loss": 0.4719,
+      "step": 440
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.635883446487816e-06,
+      "loss": 0.5679,
+      "step": 441
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.633982722593367e-06,
+      "loss": 0.5536,
+      "step": 442
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.8125,
+      "learning_rate": 4.6320774423849055e-06,
+      "loss": 0.516,
+      "step": 443
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.630167609930437e-06,
+      "loss": 0.4598,
+      "step": 444
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.25,
+      "learning_rate": 4.628253229307689e-06,
+      "loss": 0.5591,
+      "step": 445
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.75,
+      "learning_rate": 4.6263343046040954e-06,
+      "loss": 0.5138,
+      "step": 446
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.624410839916798e-06,
+      "loss": 0.4872,
+      "step": 447
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.622482839352628e-06,
+      "loss": 0.5569,
+      "step": 448
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.620550307028101e-06,
+      "loss": 0.5407,
+      "step": 449
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.8125,
+      "learning_rate": 4.618613247069411e-06,
+      "loss": 0.5128,
+      "step": 450
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.616671663612418e-06,
+      "loss": 0.5323,
+      "step": 451
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.921875,
+      "learning_rate": 4.6147255608026395e-06,
+      "loss": 0.5824,
+      "step": 452
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.612774942795244e-06,
+      "loss": 0.4799,
+      "step": 453
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.8125,
+      "learning_rate": 4.610819813755038e-06,
+      "loss": 0.5403,
+      "step": 454
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.5,
+      "learning_rate": 4.608860177856462e-06,
+      "loss": 0.5352,
+      "step": 455
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.875,
+      "learning_rate": 4.606896039283577e-06,
+      "loss": 0.5168,
+      "step": 456
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.0,
+      "learning_rate": 4.604927402230061e-06,
+      "loss": 0.5088,
+      "step": 457
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.6029542708991925e-06,
+      "loss": 0.4717,
+      "step": 458
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.60097664950385e-06,
+      "loss": 0.5052,
+      "step": 459
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.598994542266496e-06,
+      "loss": 0.5658,
+      "step": 460
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.0,
+      "learning_rate": 4.597007953419173e-06,
+      "loss": 0.4909,
+      "step": 461
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.28125,
+      "learning_rate": 4.595016887203489e-06,
+      "loss": 0.522,
+      "step": 462
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.96875,
+      "learning_rate": 4.593021347870614e-06,
+      "loss": 0.5103,
+      "step": 463
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.5910213396812705e-06,
+      "loss": 0.5174,
+      "step": 464
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.5,
+      "learning_rate": 4.58901686690572e-06,
+      "loss": 0.4901,
+      "step": 465
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.587007933823755e-06,
+      "loss": 0.5575,
+      "step": 466
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.584994544724695e-06,
+      "loss": 0.4458,
+      "step": 467
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.582976703907371e-06,
+      "loss": 0.5132,
+      "step": 468
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.71875,
+      "learning_rate": 4.580954415680121e-06,
+      "loss": 0.5243,
+      "step": 469
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.828125,
+      "learning_rate": 4.578927684360777e-06,
+      "loss": 0.4474,
+      "step": 470
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.576896514276657e-06,
+      "loss": 0.4832,
+      "step": 471
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.5748609097645595e-06,
+      "loss": 0.4726,
+      "step": 472
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.5728208751707475e-06,
+      "loss": 0.4582,
+      "step": 473
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.0,
+      "learning_rate": 4.570776414850945e-06,
+      "loss": 0.5126,
+      "step": 474
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.875,
+      "learning_rate": 4.568727533170324e-06,
+      "loss": 0.4849,
+      "step": 475
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.566674234503498e-06,
+      "loss": 0.5691,
+      "step": 476
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.5,
+      "learning_rate": 4.564616523234511e-06,
+      "loss": 0.4939,
+      "step": 477
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.875,
+      "learning_rate": 4.562554403756827e-06,
+      "loss": 0.4586,
+      "step": 478
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.560487880473324e-06,
+      "loss": 0.4203,
+      "step": 479
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.53125,
+      "learning_rate": 4.558416957796282e-06,
+      "loss": 0.5099,
+      "step": 480
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.556341640147375e-06,
+      "loss": 0.5674,
+      "step": 481
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.554261931957658e-06,
+      "loss": 0.4808,
+      "step": 482
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.0,
+      "learning_rate": 4.552177837667562e-06,
+      "loss": 0.4432,
+      "step": 483
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.75,
+      "learning_rate": 4.550089361726885e-06,
+      "loss": 0.5151,
+      "step": 484
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.875,
+      "learning_rate": 4.547996508594777e-06,
+      "loss": 0.5181,
+      "step": 485
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.5458992827397364e-06,
+      "loss": 0.536,
+      "step": 486
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.890625,
+      "learning_rate": 4.543797688639596e-06,
+      "loss": 0.5202,
+      "step": 487
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.541691730781517e-06,
+      "loss": 0.4415,
+      "step": 488
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.75,
+      "learning_rate": 4.539581413661976e-06,
+      "loss": 0.4711,
+      "step": 489
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.8125,
+      "learning_rate": 4.537466741786759e-06,
+      "loss": 0.4893,
+      "step": 490
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.535347719670947e-06,
+      "loss": 0.4926,
+      "step": 491
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.533224351838914e-06,
+      "loss": 0.494,
+      "step": 492
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.859375,
+      "learning_rate": 4.531096642824308e-06,
+      "loss": 0.5893,
+      "step": 493
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.528964597170048e-06,
+      "loss": 0.4723,
+      "step": 494
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.625,
+      "learning_rate": 4.5268282194283145e-06,
+      "loss": 0.4578,
+      "step": 495
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.524687514160532e-06,
+      "loss": 0.542,
+      "step": 496
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.522542485937369e-06,
+      "loss": 0.5218,
+      "step": 497
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.859375,
+      "learning_rate": 4.520393139338723e-06,
+      "loss": 0.5305,
+      "step": 498
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.518239478953713e-06,
+      "loss": 0.5403,
+      "step": 499
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.859375,
+      "learning_rate": 4.516081509380666e-06,
+      "loss": 0.4893,
+      "step": 500
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.28125,
+      "learning_rate": 4.513919235227111e-06,
+      "loss": 0.5554,
+      "step": 501
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.671875,
+      "learning_rate": 4.5117526611097685e-06,
+      "loss": 0.5084,
+      "step": 502
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.921875,
+      "learning_rate": 4.5095817916545385e-06,
+      "loss": 0.4831,
+      "step": 503
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.507406631496491e-06,
+      "loss": 0.4704,
+      "step": 504
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.0,
+      "learning_rate": 4.505227185279862e-06,
+      "loss": 0.5716,
+      "step": 505
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.503043457658033e-06,
+      "loss": 0.4562,
+      "step": 506
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.375,
+      "learning_rate": 4.500855453293532e-06,
+      "loss": 0.57,
+      "step": 507
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.71875,
+      "learning_rate": 4.498663176858014e-06,
+      "loss": 0.5835,
+      "step": 508
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.0,
+      "learning_rate": 4.496466633032257e-06,
+      "loss": 0.5323,
+      "step": 509
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.494265826506152e-06,
+      "loss": 0.4424,
+      "step": 510
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.671875,
+      "learning_rate": 4.492060761978688e-06,
+      "loss": 0.496,
+      "step": 511
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.578125,
+      "learning_rate": 4.48985144415795e-06,
+      "loss": 0.4816,
+      "step": 512
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.4876378777611e-06,
+      "loss": 0.5532,
+      "step": 513
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.421875,
+      "learning_rate": 4.485420067514373e-06,
+      "loss": 0.4505,
+      "step": 514
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.483198018153065e-06,
+      "loss": 0.5747,
+      "step": 515
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.4809717344215235e-06,
+      "loss": 0.5647,
+      "step": 516
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.515625,
+      "learning_rate": 4.478741221073136e-06,
+      "loss": 0.4387,
+      "step": 517
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.476506482870321e-06,
+      "loss": 0.4948,
+      "step": 518
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.474267524584519e-06,
+      "loss": 0.5218,
+      "step": 519
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.472024350996178e-06,
+      "loss": 0.4733,
+      "step": 520
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.125,
+      "learning_rate": 4.469776966894749e-06,
+      "loss": 0.518,
+      "step": 521
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.75,
+      "learning_rate": 4.467525377078672e-06,
+      "loss": 0.4933,
+      "step": 522
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.828125,
+      "learning_rate": 4.465269586355365e-06,
+      "loss": 0.5198,
+      "step": 523
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.75,
+      "learning_rate": 4.463009599541219e-06,
+      "loss": 0.5213,
+      "step": 524
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.460745421461582e-06,
+      "loss": 0.5154,
+      "step": 525
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.8125,
+      "learning_rate": 4.458477056950748e-06,
+      "loss": 0.4632,
+      "step": 526
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.0,
+      "learning_rate": 4.456204510851957e-06,
+      "loss": 0.5119,
+      "step": 527
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.515625,
+      "learning_rate": 4.4539277880173695e-06,
+      "loss": 0.4784,
+      "step": 528
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.451646893308068e-06,
+      "loss": 0.5135,
+      "step": 529
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.34375,
+      "learning_rate": 4.449361831594041e-06,
+      "loss": 0.6156,
+      "step": 530
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.447072607754177e-06,
+      "loss": 0.4689,
+      "step": 531
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.4447792266762466e-06,
+      "loss": 0.5554,
+      "step": 532
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.96875,
+      "learning_rate": 4.442481693256899e-06,
+      "loss": 0.5223,
+      "step": 533
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.8125,
+      "learning_rate": 4.440180012401649e-06,
+      "loss": 0.4859,
+      "step": 534
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.437874189024867e-06,
+      "loss": 0.4811,
+      "step": 535
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.953125,
+      "learning_rate": 4.435564228049768e-06,
+      "loss": 0.5529,
+      "step": 536
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.433250134408401e-06,
+      "loss": 0.5041,
+      "step": 537
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.430931913041639e-06,
+      "loss": 0.4458,
+      "step": 538
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.4286095688991655e-06,
+      "loss": 0.4998,
+      "step": 539
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.426283106939474e-06,
+      "loss": 0.4814,
+      "step": 540
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.423952532129841e-06,
+      "loss": 0.5255,
+      "step": 541
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.953125,
+      "learning_rate": 4.4216178494463305e-06,
+      "loss": 0.5096,
+      "step": 542
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.890625,
+      "learning_rate": 4.419279063873776e-06,
+      "loss": 0.5004,
+      "step": 543
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.25,
+      "learning_rate": 4.416936180405769e-06,
+      "loss": 0.5353,
+      "step": 544
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.4145892040446525e-06,
+      "loss": 0.5355,
+      "step": 545
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.412238139801509e-06,
+      "loss": 0.3986,
+      "step": 546
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.4098829926961485e-06,
+      "loss": 0.4783,
+      "step": 547
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.407523767757096e-06,
+      "loss": 0.4791,
+      "step": 548
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.125,
+      "learning_rate": 4.405160470021586e-06,
+      "loss": 0.5744,
+      "step": 549
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.671875,
+      "learning_rate": 4.402793104535548e-06,
+      "loss": 0.5085,
+      "step": 550
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.400421676353599e-06,
+      "loss": 0.5143,
+      "step": 551
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.953125,
+      "learning_rate": 4.398046190539025e-06,
+      "loss": 0.5363,
+      "step": 552
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.39566665216378e-06,
+      "loss": 0.4665,
+      "step": 553
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.578125,
+      "learning_rate": 4.393283066308469e-06,
+      "loss": 0.4841,
+      "step": 554
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.3908954380623394e-06,
+      "loss": 0.4201,
+      "step": 555
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.388503772523269e-06,
+      "loss": 0.4805,
+      "step": 556
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.75,
+      "learning_rate": 4.386108074797757e-06,
+      "loss": 0.4868,
+      "step": 557
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.125,
+      "learning_rate": 4.383708350000911e-06,
+      "loss": 0.5354,
+      "step": 558
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.34375,
+      "learning_rate": 4.381304603256436e-06,
+      "loss": 0.5532,
+      "step": 559
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.0,
+      "learning_rate": 4.378896839696625e-06,
+      "loss": 0.5394,
+      "step": 560
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.890625,
+      "learning_rate": 4.376485064462348e-06,
+      "loss": 0.4706,
+      "step": 561
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.0,
+      "learning_rate": 4.37406928270304e-06,
+      "loss": 0.5667,
+      "step": 562
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.371649499576691e-06,
+      "loss": 0.5342,
+      "step": 563
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.875,
+      "learning_rate": 4.369225720249834e-06,
+      "loss": 0.5272,
+      "step": 564
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.671875,
+      "learning_rate": 4.3667979498975326e-06,
+      "loss": 0.5386,
+      "step": 565
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.0,
+      "learning_rate": 4.364366193703373e-06,
+      "loss": 0.542,
+      "step": 566
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.59375,
+      "learning_rate": 4.361930456859455e-06,
+      "loss": 0.5111,
+      "step": 567
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.359490744566372e-06,
+      "loss": 0.4392,
+      "step": 568
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.0,
+      "learning_rate": 4.3570470620332086e-06,
+      "loss": 0.4685,
+      "step": 569
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.703125,
+      "learning_rate": 4.354599414477526e-06,
+      "loss": 0.4973,
+      "step": 570
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.0,
+      "learning_rate": 4.352147807125351e-06,
+      "loss": 0.6541,
+      "step": 571
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.349692245211166e-06,
+      "loss": 0.5162,
+      "step": 572
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.96875,
+      "learning_rate": 4.347232733977893e-06,
+      "loss": 0.5296,
+      "step": 573
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.344769278676893e-06,
+      "loss": 0.4881,
+      "step": 574
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.75,
+      "learning_rate": 4.34230188456794e-06,
+      "loss": 0.5219,
+      "step": 575
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.339830556919225e-06,
+      "loss": 0.5144,
+      "step": 576
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.75,
+      "learning_rate": 4.337355301007336e-06,
+      "loss": 0.5526,
+      "step": 577
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.334876122117242e-06,
+      "loss": 0.5431,
+      "step": 578
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.859375,
+      "learning_rate": 4.332393025542298e-06,
+      "loss": 0.4611,
+      "step": 579
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.5,
+      "learning_rate": 4.329906016584217e-06,
+      "loss": 0.5104,
+      "step": 580
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.59375,
+      "learning_rate": 4.327415100553066e-06,
+      "loss": 0.4021,
+      "step": 581
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.324920282767256e-06,
+      "loss": 0.4773,
+      "step": 582
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.828125,
+      "learning_rate": 4.322421568553529e-06,
+      "loss": 0.5395,
+      "step": 583
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.319918963246944e-06,
+      "loss": 0.4883,
+      "step": 584
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.317412472190871e-06,
+      "loss": 0.4271,
+      "step": 585
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.8125,
+      "learning_rate": 4.314902100736973e-06,
+      "loss": 0.4897,
+      "step": 586
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.312387854245201e-06,
+      "loss": 0.4699,
+      "step": 587
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.3098697380837785e-06,
+      "loss": 0.534,
+      "step": 588
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.671875,
+      "learning_rate": 4.30734775762919e-06,
+      "loss": 0.5035,
+      "step": 589
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.828125,
+      "learning_rate": 4.304821918266175e-06,
+      "loss": 0.5323,
+      "step": 590
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.302292225387707e-06,
+      "loss": 0.5466,
+      "step": 591
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.2997586843949905e-06,
+      "loss": 0.5769,
+      "step": 592
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.671875,
+      "learning_rate": 4.297221300697444e-06,
+      "loss": 0.526,
+      "step": 593
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.890625,
+      "learning_rate": 4.294680079712694e-06,
+      "loss": 0.4931,
+      "step": 594
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.292135026866556e-06,
+      "loss": 0.4981,
+      "step": 595
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.625,
+      "learning_rate": 4.289586147593031e-06,
+      "loss": 0.4877,
+      "step": 596
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.287033447334286e-06,
+      "loss": 0.5371,
+      "step": 597
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.28447693154065e-06,
+      "loss": 0.4978,
+      "step": 598
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.921875,
+      "learning_rate": 4.281916605670595e-06,
+      "loss": 0.4601,
+      "step": 599
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.27935247519073e-06,
+      "loss": 0.5274,
+      "step": 600
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.828125,
+      "learning_rate": 4.276784545575788e-06,
+      "loss": 0.5157,
+      "step": 601
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.274212822308612e-06,
+      "loss": 0.5252,
+      "step": 602
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.271637310880147e-06,
+      "loss": 0.5092,
+      "step": 603
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.2690580167894225e-06,
+      "loss": 0.5195,
+      "step": 604
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.266474945543548e-06,
+      "loss": 0.5291,
+      "step": 605
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.263888102657696e-06,
+      "loss": 0.5233,
+      "step": 606
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.546875,
+      "learning_rate": 4.261297493655092e-06,
+      "loss": 0.4955,
+      "step": 607
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.258703124067005e-06,
+      "loss": 0.4833,
+      "step": 608
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.75,
+      "learning_rate": 4.256104999432729e-06,
+      "loss": 0.4543,
+      "step": 609
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.253503125299579e-06,
+      "loss": 0.4964,
+      "step": 610
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.250897507222873e-06,
+      "loss": 0.5416,
+      "step": 611
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.248288150765925e-06,
+      "loss": 0.4889,
+      "step": 612
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.2456750615000306e-06,
+      "loss": 0.5405,
+      "step": 613
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.243058245004455e-06,
+      "loss": 0.5238,
+      "step": 614
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.24043770686642e-06,
+      "loss": 0.5527,
+      "step": 615
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.828125,
+      "learning_rate": 4.237813452681095e-06,
+      "loss": 0.5281,
+      "step": 616
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.703125,
+      "learning_rate": 4.2351854880515856e-06,
+      "loss": 0.4338,
+      "step": 617
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.5,
+      "learning_rate": 4.232553818588915e-06,
+      "loss": 0.4748,
+      "step": 618
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.71875,
+      "learning_rate": 4.22991844991202e-06,
+      "loss": 0.4588,
+      "step": 619
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.890625,
+      "learning_rate": 4.227279387647736e-06,
+      "loss": 0.5539,
+      "step": 620
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.71875,
+      "learning_rate": 4.224636637430783e-06,
+      "loss": 0.4833,
+      "step": 621
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.221990204903756e-06,
+      "loss": 0.5088,
+      "step": 622
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.71875,
+      "learning_rate": 4.219340095717111e-06,
+      "loss": 0.5624,
+      "step": 623
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.216686315529156e-06,
+      "loss": 0.479,
+      "step": 624
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.90625,
+      "learning_rate": 4.2140288700060365e-06,
+      "loss": 0.4578,
+      "step": 625
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.125,
+      "learning_rate": 4.211367764821722e-06,
+      "loss": 0.5681,
+      "step": 626
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.208703005658e-06,
+      "loss": 0.5444,
+      "step": 627
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.921875,
+      "learning_rate": 4.206034598204454e-06,
+      "loss": 0.4983,
+      "step": 628
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.546875,
+      "learning_rate": 4.203362548158462e-06,
+      "loss": 0.4936,
+      "step": 629
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.200686861225177e-06,
+      "loss": 0.5009,
+      "step": 630
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.75,
+      "learning_rate": 4.1980075431175184e-06,
+      "loss": 0.489,
+      "step": 631
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.96875,
+      "learning_rate": 4.195324599556158e-06,
+      "loss": 0.5307,
+      "step": 632
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.192638036269507e-06,
+      "loss": 0.5674,
+      "step": 633
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.189947858993708e-06,
+      "loss": 0.5353,
+      "step": 634
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.671875,
+      "learning_rate": 4.1872540734726176e-06,
+      "loss": 0.6288,
+      "step": 635
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.184556685457798e-06,
+      "loss": 0.499,
+      "step": 636
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.0,
+      "learning_rate": 4.1818557007085e-06,
+      "loss": 0.5046,
+      "step": 637
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.515625,
+      "learning_rate": 4.179151124991659e-06,
+      "loss": 0.3983,
+      "step": 638
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.53125,
+      "learning_rate": 4.176442964081874e-06,
+      "loss": 0.4948,
+      "step": 639
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.859375,
+      "learning_rate": 4.1737312237613994e-06,
+      "loss": 0.5184,
+      "step": 640
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.625,
+      "learning_rate": 4.1710159098201316e-06,
+      "loss": 0.4978,
+      "step": 641
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.828125,
+      "learning_rate": 4.168297028055599e-06,
+      "loss": 0.4792,
+      "step": 642
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.921875,
+      "learning_rate": 4.165574584272946e-06,
+      "loss": 0.5216,
+      "step": 643
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.671875,
+      "learning_rate": 4.1628485842849235e-06,
+      "loss": 0.4444,
+      "step": 644
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.160119033911874e-06,
+      "loss": 0.46,
+      "step": 645
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.828125,
+      "learning_rate": 4.157385938981722e-06,
+      "loss": 0.5138,
+      "step": 646
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.890625,
+      "learning_rate": 4.154649305329959e-06,
+      "loss": 0.5109,
+      "step": 647
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.151909138799631e-06,
+      "loss": 0.5211,
+      "step": 648
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.149165445241331e-06,
+      "loss": 0.5669,
+      "step": 649
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.921875,
+      "learning_rate": 4.146418230513179e-06,
+      "loss": 0.5181,
+      "step": 650
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.578125,
+      "learning_rate": 4.143667500480815e-06,
+      "loss": 0.5239,
+      "step": 651
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.53125,
+      "learning_rate": 4.140913261017382e-06,
+      "loss": 0.4881,
+      "step": 652
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.59375,
+      "learning_rate": 4.1381555180035205e-06,
+      "loss": 0.4726,
+      "step": 653
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.125,
+      "learning_rate": 4.1353942773273475e-06,
+      "loss": 0.4569,
+      "step": 654
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.13262954488445e-06,
+      "loss": 0.5122,
+      "step": 655
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.8125,
+      "learning_rate": 4.1298613265778685e-06,
+      "loss": 0.4878,
+      "step": 656
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.875,
+      "learning_rate": 4.12708962831809e-06,
+      "loss": 0.4864,
+      "step": 657
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.921875,
+      "learning_rate": 4.124314456023026e-06,
+      "loss": 0.5686,
+      "step": 658
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.121535815618012e-06,
+      "loss": 0.4882,
+      "step": 659
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.828125,
+      "learning_rate": 4.1187537130357825e-06,
+      "loss": 0.477,
+      "step": 660
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.115968154216467e-06,
+      "loss": 0.4888,
+      "step": 661
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.113179145107576e-06,
+      "loss": 0.4567,
+      "step": 662
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.110386691663982e-06,
+      "loss": 0.4888,
+      "step": 663
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.53125,
+      "learning_rate": 4.107590799847917e-06,
+      "loss": 0.4802,
+      "step": 664
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.1047914756289505e-06,
+      "loss": 0.5147,
+      "step": 665
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.101988724983983e-06,
+      "loss": 0.4245,
+      "step": 666
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.671875,
+      "learning_rate": 4.099182553897228e-06,
+      "loss": 0.4655,
+      "step": 667
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.096372968360206e-06,
+      "loss": 0.4621,
+      "step": 668
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.093559974371725e-06,
+      "loss": 0.5014,
+      "step": 669
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.59375,
+      "learning_rate": 4.090743577937869e-06,
+      "loss": 0.5232,
+      "step": 670
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.8125,
+      "learning_rate": 4.087923785071992e-06,
+      "loss": 0.4738,
+      "step": 671
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.71875,
+      "learning_rate": 4.085100601794695e-06,
+      "loss": 0.4916,
+      "step": 672
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.875,
+      "learning_rate": 4.082274034133819e-06,
+      "loss": 0.5435,
+      "step": 673
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.079444088124431e-06,
+      "loss": 0.4313,
+      "step": 674
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.076610769808814e-06,
+      "loss": 0.5122,
+      "step": 675
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.073774085236447e-06,
+      "loss": 0.5177,
+      "step": 676
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.070934040463999e-06,
+      "loss": 0.4815,
+      "step": 677
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.0680906415553126e-06,
+      "loss": 0.469,
+      "step": 678
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.859375,
+      "learning_rate": 4.065243894581392e-06,
+      "loss": 0.5778,
+      "step": 679
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.062393805620391e-06,
+      "loss": 0.5396,
+      "step": 680
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.059540380757598e-06,
+      "loss": 0.5081,
+      "step": 681
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.515625,
+      "learning_rate": 4.056683626085423e-06,
+      "loss": 0.5294,
+      "step": 682
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.053823547703386e-06,
+      "loss": 0.4541,
+      "step": 683
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.75,
+      "learning_rate": 4.050960151718105e-06,
+      "loss": 0.4372,
+      "step": 684
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.875,
+      "learning_rate": 4.04809344424328e-06,
+      "loss": 0.5431,
+      "step": 685
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.045223431399683e-06,
+      "loss": 0.5235,
+      "step": 686
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.042350119315142e-06,
+      "loss": 0.4716,
+      "step": 687
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.578125,
+      "learning_rate": 4.0394735141245285e-06,
+      "loss": 0.469,
+      "step": 688
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.921875,
+      "learning_rate": 4.036593621969747e-06,
+      "loss": 0.516,
+      "step": 689
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.033710448999719e-06,
+      "loss": 0.507,
+      "step": 690
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.90625,
+      "learning_rate": 4.030824001370372e-06,
+      "loss": 0.5095,
+      "step": 691
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.027934285244624e-06,
+      "loss": 0.4877,
+      "step": 692
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.875,
+      "learning_rate": 4.025041306792372e-06,
+      "loss": 0.4956,
+      "step": 693
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.5,
+      "learning_rate": 4.022145072190479e-06,
+      "loss": 0.4405,
+      "step": 694
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.019245587622761e-06,
+      "loss": 0.4561,
+      "step": 695
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.0163428592799695e-06,
+      "loss": 0.5106,
+      "step": 696
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.828125,
+      "learning_rate": 4.013436893359787e-06,
+      "loss": 0.4677,
+      "step": 697
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.703125,
+      "learning_rate": 4.010527696066805e-06,
+      "loss": 0.5545,
+      "step": 698
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.90625,
+      "learning_rate": 4.007615273612515e-06,
+      "loss": 0.5496,
+      "step": 699
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.004699632215296e-06,
+      "loss": 0.4975,
+      "step": 700
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.001780778100398e-06,
+      "loss": 0.469,
+      "step": 701
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.4375,
+      "learning_rate": 3.998858717499931e-06,
+      "loss": 0.5949,
+      "step": 702
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.609375,
+      "learning_rate": 3.995933456652853e-06,
+      "loss": 0.488,
+      "step": 703
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.75,
+      "learning_rate": 3.993005001804953e-06,
+      "loss": 0.5561,
+      "step": 704
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.90625,
+      "learning_rate": 3.990073359208838e-06,
+      "loss": 0.4881,
+      "step": 705
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.21875,
+      "learning_rate": 3.987138535123927e-06,
+      "loss": 0.5503,
+      "step": 706
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.734375,
+      "learning_rate": 3.984200535816427e-06,
+      "loss": 0.4776,
+      "step": 707
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.921875,
+      "learning_rate": 3.981259367559325e-06,
+      "loss": 0.5134,
+      "step": 708
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.484375,
+      "learning_rate": 3.978315036632373e-06,
+      "loss": 0.438,
+      "step": 709
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.46875,
+      "learning_rate": 3.975367549322083e-06,
+      "loss": 0.4759,
+      "step": 710
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.578125,
+      "learning_rate": 3.972416911921695e-06,
+      "loss": 0.5444,
+      "step": 711
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.796875,
+      "learning_rate": 3.969463130731183e-06,
+      "loss": 0.5427,
+      "step": 712
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.78125,
+      "learning_rate": 3.966506212057232e-06,
+      "loss": 0.5113,
+      "step": 713
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.59375,
+      "learning_rate": 3.963546162213223e-06,
+      "loss": 0.4529,
+      "step": 714
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.984375,
+      "learning_rate": 3.960582987519225e-06,
+      "loss": 0.444,
+      "step": 715
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.921875,
+      "learning_rate": 3.957616694301979e-06,
+      "loss": 0.4905,
+      "step": 716
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.15625,
+      "learning_rate": 3.9546472888948825e-06,
+      "loss": 0.565,
+      "step": 717
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.6875,
+      "learning_rate": 3.951674777637979e-06,
+      "loss": 0.4918,
+      "step": 718
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.84375,
+      "learning_rate": 3.948699166877946e-06,
+      "loss": 0.4855,
+      "step": 719
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.6875,
+      "learning_rate": 3.945720462968073e-06,
+      "loss": 0.6208,
+      "step": 720
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.96875,
+      "learning_rate": 3.942738672268258e-06,
+      "loss": 0.5875,
+      "step": 721
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.78125,
+      "learning_rate": 3.9397538011449896e-06,
+      "loss": 0.4687,
+      "step": 722
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.6875,
+      "learning_rate": 3.936765855971332e-06,
+      "loss": 0.5216,
+      "step": 723
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.703125,
+      "learning_rate": 3.933774843126914e-06,
+      "loss": 0.5253,
+      "step": 724
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.703125,
+      "learning_rate": 3.930780768997913e-06,
+      "loss": 0.5201,
+      "step": 725
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.609375,
+      "learning_rate": 3.927783639977043e-06,
+      "loss": 0.4889,
+      "step": 726
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.75,
+      "learning_rate": 3.924783462463541e-06,
+      "loss": 0.4704,
+      "step": 727
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.84375,
+      "learning_rate": 3.921780242863153e-06,
+      "loss": 0.4776,
+      "step": 728
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.75,
+      "learning_rate": 3.9187739875881215e-06,
+      "loss": 0.5636,
+      "step": 729
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.953125,
+      "learning_rate": 3.915764703057167e-06,
+      "loss": 0.5393,
+      "step": 730
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.890625,
+      "learning_rate": 3.912752395695481e-06,
+      "loss": 0.5358,
+      "step": 731
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.96875,
+      "learning_rate": 3.909737071934707e-06,
+      "loss": 0.5129,
+      "step": 732
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.921875,
+      "learning_rate": 3.906718738212932e-06,
+      "loss": 0.5271,
+      "step": 733
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.6875,
+      "learning_rate": 3.903697400974665e-06,
+      "loss": 0.5384,
+      "step": 734
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.828125,
+      "learning_rate": 3.9006730666708316e-06,
+      "loss": 0.5528,
+      "step": 735
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.484375,
+      "learning_rate": 3.897645741758755e-06,
+      "loss": 0.3948,
+      "step": 736
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.640625,
+      "learning_rate": 3.894615432702144e-06,
+      "loss": 0.4721,
+      "step": 737
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.8125,
+      "learning_rate": 3.891582145971079e-06,
+      "loss": 0.5235,
+      "step": 738
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.796875,
+      "learning_rate": 3.888545888041998e-06,
+      "loss": 0.5312,
+      "step": 739
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.78125,
+      "learning_rate": 3.885506665397683e-06,
+      "loss": 0.4589,
+      "step": 740
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.59375,
+      "learning_rate": 3.882464484527247e-06,
+      "loss": 0.4226,
+      "step": 741
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.25,
+      "learning_rate": 3.879419351926115e-06,
+      "loss": 0.3831,
+      "step": 742
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.75,
+      "learning_rate": 3.876371274096021e-06,
+      "loss": 0.4699,
+      "step": 743
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.375,
+      "learning_rate": 3.873320257544981e-06,
+      "loss": 0.3985,
+      "step": 744
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.90625,
+      "learning_rate": 3.87026630878729e-06,
+      "loss": 0.4686,
+      "step": 745
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.53125,
+      "learning_rate": 3.867209434343502e-06,
+      "loss": 0.3904,
+      "step": 746
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.09375,
+      "learning_rate": 3.8641496407404165e-06,
+      "loss": 0.4212,
+      "step": 747
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.546875,
+      "learning_rate": 3.861086934511069e-06,
+      "loss": 0.47,
+      "step": 748
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.96875,
+      "learning_rate": 3.8580213221947085e-06,
+      "loss": 0.4427,
+      "step": 749
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.53125,
+      "learning_rate": 3.854952810336795e-06,
+      "loss": 0.3536,
+      "step": 750
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.96875,
+      "learning_rate": 3.851881405488975e-06,
+      "loss": 0.4397,
+      "step": 751
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.828125,
+      "learning_rate": 3.8488071142090745e-06,
+      "loss": 0.4282,
+      "step": 752
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.640625,
+      "learning_rate": 3.84572994306108e-06,
+      "loss": 0.3905,
+      "step": 753
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.390625,
+      "learning_rate": 3.842649898615129e-06,
+      "loss": 0.537,
+      "step": 754
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.40625,
+      "learning_rate": 3.839566987447492e-06,
+      "loss": 0.37,
+      "step": 755
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.90625,
+      "learning_rate": 3.836481216140563e-06,
+      "loss": 0.4553,
+      "step": 756
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.09375,
+      "learning_rate": 3.833392591282838e-06,
+      "loss": 0.4038,
+      "step": 757
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.59375,
+      "learning_rate": 3.830301119468912e-06,
+      "loss": 0.3641,
+      "step": 758
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.78125,
+      "learning_rate": 3.827206807299453e-06,
+      "loss": 0.4331,
+      "step": 759
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.640625,
+      "learning_rate": 3.824109661381198e-06,
+      "loss": 0.4372,
+      "step": 760
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.75,
+      "learning_rate": 3.821009688326929e-06,
+      "loss": 0.4691,
+      "step": 761
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.78125,
+      "learning_rate": 3.817906894755471e-06,
+      "loss": 0.3706,
+      "step": 762
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.671875,
+      "learning_rate": 3.814801287291665e-06,
+      "loss": 0.4186,
+      "step": 763
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.890625,
+      "learning_rate": 3.8116928725663638e-06,
+      "loss": 0.4558,
+      "step": 764
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.546875,
+      "learning_rate": 3.8085816572164124e-06,
+      "loss": 0.3606,
+      "step": 765
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.828125,
+      "learning_rate": 3.8054676478846364e-06,
+      "loss": 0.4882,
+      "step": 766
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.5,
+      "learning_rate": 3.802350851219826e-06,
+      "loss": 0.4236,
+      "step": 767
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.5625,
+      "learning_rate": 3.7992312738767234e-06,
+      "loss": 0.5069,
+      "step": 768
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.8125,
+      "learning_rate": 3.796108922516008e-06,
+      "loss": 0.4273,
+      "step": 769
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.546875,
+      "learning_rate": 3.792983803804281e-06,
+      "loss": 0.395,
+      "step": 770
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.6875,
+      "learning_rate": 3.789855924414053e-06,
+      "loss": 0.3807,
+      "step": 771
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.15625,
+      "learning_rate": 3.7867252910237284e-06,
+      "loss": 0.4703,
+      "step": 772
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.515625,
+      "learning_rate": 3.7835919103175943e-06,
+      "loss": 0.3674,
+      "step": 773
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.734375,
+      "learning_rate": 3.7804557889857996e-06,
+      "loss": 0.4198,
+      "step": 774
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.5,
+      "learning_rate": 3.7773169337243476e-06,
+      "loss": 0.439,
+      "step": 775
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.75,
+      "learning_rate": 3.7741753512350778e-06,
+      "loss": 0.4329,
+      "step": 776
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.625,
+      "learning_rate": 3.771031048225653e-06,
+      "loss": 0.4229,
+      "step": 777
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.796875,
+      "learning_rate": 3.7678840314095444e-06,
+      "loss": 0.3659,
+      "step": 778
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.828125,
+      "learning_rate": 3.7647343075060172e-06,
+      "loss": 0.4797,
+      "step": 779
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.609375,
+      "learning_rate": 3.761581883240118e-06,
+      "loss": 0.472,
+      "step": 780
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.4375,
+      "learning_rate": 3.7584267653426577e-06,
+      "loss": 0.4667,
+      "step": 781
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.765625,
+      "learning_rate": 3.755268960550199e-06,
+      "loss": 0.422,
+      "step": 782
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 4.03125,
+      "learning_rate": 3.75210847560504e-06,
+      "loss": 0.4097,
+      "step": 783
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 4.1875,
+      "learning_rate": 3.7489453172552036e-06,
+      "loss": 0.4119,
+      "step": 784
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.84375,
+      "learning_rate": 3.745779492254419e-06,
+      "loss": 0.4469,
+      "step": 785
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.765625,
+      "learning_rate": 3.742611007362111e-06,
+      "loss": 0.3989,
+      "step": 786
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.609375,
+      "learning_rate": 3.7394398693433798e-06,
+      "loss": 0.4095,
+      "step": 787
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 4.0,
+      "learning_rate": 3.7362660849689954e-06,
+      "loss": 0.427,
+      "step": 788
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.671875,
+      "learning_rate": 3.733089661015373e-06,
+      "loss": 0.3935,
+      "step": 789
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.71875,
+      "learning_rate": 3.729910604264568e-06,
+      "loss": 0.4244,
+      "step": 790
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.765625,
+      "learning_rate": 3.7267289215042544e-06,
+      "loss": 0.4751,
+      "step": 791
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.453125,
+      "learning_rate": 3.723544619527714e-06,
+      "loss": 0.4108,
+      "step": 792
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.390625,
+      "learning_rate": 3.7203577051338212e-06,
+      "loss": 0.3609,
+      "step": 793
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.5625,
+      "learning_rate": 3.7171681851270264e-06,
+      "loss": 0.4022,
+      "step": 794
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.84375,
+      "learning_rate": 3.713976066317347e-06,
+      "loss": 0.3798,
+      "step": 795
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 4.5,
+      "learning_rate": 3.7107813555203454e-06,
+      "loss": 0.4541,
+      "step": 796
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.53125,
+      "learning_rate": 3.70758405955712e-06,
+      "loss": 0.3719,
+      "step": 797
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.65625,
+      "learning_rate": 3.7043841852542884e-06,
+      "loss": 0.3691,
+      "step": 798
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.875,
+      "learning_rate": 3.701181739443974e-06,
+      "loss": 0.3977,
+      "step": 799
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 4.0625,
+      "learning_rate": 3.6979767289637904e-06,
+      "loss": 0.485,
+      "step": 800
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.75,
+      "learning_rate": 3.694769160656826e-06,
+      "loss": 0.344,
+      "step": 801
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.78125,
+      "learning_rate": 3.6915590413716314e-06,
+      "loss": 0.4373,
+      "step": 802
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.921875,
+      "learning_rate": 3.6883463779622035e-06,
+      "loss": 0.4541,
+      "step": 803
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.5625,
+      "learning_rate": 3.6851311772879726e-06,
+      "loss": 0.4041,
+      "step": 804
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.59375,
+      "learning_rate": 3.681913446213784e-06,
+      "loss": 0.4229,
+      "step": 805
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.703125,
+      "learning_rate": 3.6786931916098883e-06,
+      "loss": 0.3604,
+      "step": 806
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.5625,
+      "learning_rate": 3.675470420351921e-06,
+      "loss": 0.399,
+      "step": 807
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.9375,
+      "learning_rate": 3.672245139320894e-06,
+      "loss": 0.4473,
+      "step": 808
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.453125,
+      "learning_rate": 3.6690173554031766e-06,
+      "loss": 0.4033,
+      "step": 809
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.8125,
+      "learning_rate": 3.6657870754904816e-06,
+      "loss": 0.4285,
+      "step": 810
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.578125,
+      "learning_rate": 3.6625543064798517e-06,
+      "loss": 0.5205,
+      "step": 811
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.765625,
+      "learning_rate": 3.659319055273644e-06,
+      "loss": 0.3827,
+      "step": 812
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.65625,
+      "learning_rate": 3.6560813287795156e-06,
+      "loss": 0.3985,
+      "step": 813
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 4.21875,
+      "learning_rate": 3.652841133910409e-06,
+      "loss": 0.4901,
+      "step": 814
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 4.03125,
+      "learning_rate": 3.649598477584534e-06,
+      "loss": 0.5103,
+      "step": 815
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.75,
+      "learning_rate": 3.6463533667253604e-06,
+      "loss": 0.3683,
+      "step": 816
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.78125,
+      "learning_rate": 3.6431058082615966e-06,
+      "loss": 0.4859,
+      "step": 817
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.703125,
+      "learning_rate": 3.6398558091271763e-06,
+      "loss": 0.4269,
+      "step": 818
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.53125,
+      "learning_rate": 3.6366033762612457e-06,
+      "loss": 0.4095,
+      "step": 819
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.59375,
+      "learning_rate": 3.633348516608145e-06,
+      "loss": 0.398,
+      "step": 820
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.71875,
+      "learning_rate": 3.6300912371173993e-06,
+      "loss": 0.3974,
+      "step": 821
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.640625,
+      "learning_rate": 3.6268315447436973e-06,
+      "loss": 0.4393,
+      "step": 822
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 4.03125,
+      "learning_rate": 3.6235694464468826e-06,
+      "loss": 0.4335,
+      "step": 823
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.765625,
+      "learning_rate": 3.6203049491919313e-06,
+      "loss": 0.4206,
+      "step": 824
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.734375,
+      "learning_rate": 3.6170380599489458e-06,
+      "loss": 0.4327,
+      "step": 825
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.8125,
+      "learning_rate": 3.6137687856931336e-06,
+      "loss": 0.3635,
+      "step": 826
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 4.0625,
+      "learning_rate": 3.6104971334047954e-06,
+      "loss": 0.432,
+      "step": 827
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.359375,
+      "learning_rate": 3.6072231100693078e-06,
+      "loss": 0.3745,
+      "step": 828
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.625,
+      "learning_rate": 3.603946722677112e-06,
+      "loss": 0.4152,
+      "step": 829
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.84375,
+      "learning_rate": 3.6006679782236952e-06,
+      "loss": 0.4958,
+      "step": 830
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.6875,
+      "learning_rate": 3.5973868837095777e-06,
+      "loss": 0.3912,
+      "step": 831
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.671875,
+      "learning_rate": 3.5941034461402975e-06,
+      "loss": 0.4448,
+      "step": 832
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.4375,
+      "learning_rate": 3.5908176725263943e-06,
+      "loss": 0.4041,
+      "step": 833
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.734375,
+      "learning_rate": 3.5875295698833977e-06,
+      "loss": 0.4205,
+      "step": 834
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.375,
+      "learning_rate": 3.5842391452318087e-06,
+      "loss": 0.3695,
+      "step": 835
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 4.1875,
+      "learning_rate": 3.5809464055970866e-06,
+      "loss": 0.4478,
+      "step": 836
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.84375,
+      "learning_rate": 3.5776513580096316e-06,
+      "loss": 0.4544,
+      "step": 837
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.671875,
+      "learning_rate": 3.574354009504774e-06,
+      "loss": 0.4032,
+      "step": 838
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.5625,
+      "learning_rate": 3.571054367122756e-06,
+      "loss": 0.4231,
+      "step": 839
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.96875,
+      "learning_rate": 3.5677524379087182e-06,
+      "loss": 0.4561,
+      "step": 840
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.625,
+      "learning_rate": 3.564448228912682e-06,
+      "loss": 0.3935,
+      "step": 841
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.71875,
+      "learning_rate": 3.561141747189538e-06,
+      "loss": 0.4151,
+      "step": 842
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 4.28125,
+      "learning_rate": 3.557832999799029e-06,
+      "loss": 0.3755,
+      "step": 843
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.90625,
+      "learning_rate": 3.554521993805735e-06,
+      "loss": 0.4534,
+      "step": 844
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.34375,
+      "learning_rate": 3.55120873627906e-06,
+      "loss": 0.3448,
+      "step": 845
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.84375,
+      "learning_rate": 3.5478932342932116e-06,
+      "loss": 0.3996,
+      "step": 846
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.8125,
+      "learning_rate": 3.5445754949271925e-06,
+      "loss": 0.4114,
+      "step": 847
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.625,
+      "learning_rate": 3.5412555252647827e-06,
+      "loss": 0.3773,
+      "step": 848
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.78125,
+      "learning_rate": 3.5379333323945236e-06,
+      "loss": 0.4175,
+      "step": 849
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.625,
+      "learning_rate": 3.534608923409702e-06,
+      "loss": 0.4295,
+      "step": 850
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.765625,
+      "learning_rate": 3.531282305408339e-06,
+      "loss": 0.4373,
+      "step": 851
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.828125,
+      "learning_rate": 3.527953485493168e-06,
+      "loss": 0.397,
+      "step": 852
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.828125,
+      "learning_rate": 3.5246224707716293e-06,
+      "loss": 0.4403,
+      "step": 853
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.921875,
+      "learning_rate": 3.5212892683558454e-06,
+      "loss": 0.414,
+      "step": 854
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.59375,
+      "learning_rate": 3.5179538853626118e-06,
+      "loss": 0.4144,
+      "step": 855
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.671875,
+      "learning_rate": 3.5146163289133774e-06,
+      "loss": 0.3821,
+      "step": 856
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.71875,
+      "learning_rate": 3.5112766061342346e-06,
+      "loss": 0.4003,
+      "step": 857
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.40625,
+      "learning_rate": 3.5079347241559e-06,
+      "loss": 0.3585,
+      "step": 858
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.703125,
+      "learning_rate": 3.5045906901136983e-06,
+      "loss": 0.4354,
+      "step": 859
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.75,
+      "learning_rate": 3.501244511147553e-06,
+      "loss": 0.3731,
+      "step": 860
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.734375,
+      "learning_rate": 3.4978961944019645e-06,
+      "loss": 0.4459,
+      "step": 861
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.640625,
+      "learning_rate": 3.494545747025999e-06,
+      "loss": 0.3837,
+      "step": 862
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.375,
+      "learning_rate": 3.4911931761732703e-06,
+      "loss": 0.4063,
+      "step": 863
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.734375,
+      "learning_rate": 3.4878384890019283e-06,
+      "loss": 0.4982,
+      "step": 864
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 4.46875,
+      "learning_rate": 3.4844816926746384e-06,
+      "loss": 0.4559,
+      "step": 865
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.59375,
+      "learning_rate": 3.4811227943585735e-06,
+      "loss": 0.4212,
+      "step": 866
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.484375,
+      "learning_rate": 3.47776180122539e-06,
+      "loss": 0.4384,
+      "step": 867
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.296875,
+      "learning_rate": 3.474398720451221e-06,
+      "loss": 0.378,
+      "step": 868
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.953125,
+      "learning_rate": 3.471033559216654e-06,
+      "loss": 0.4854,
+      "step": 869
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.8125,
+      "learning_rate": 3.4676663247067195e-06,
+      "loss": 0.4243,
+      "step": 870
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.90625,
+      "learning_rate": 3.464297024110875e-06,
+      "loss": 0.4587,
+      "step": 871
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.734375,
+      "learning_rate": 3.4609256646229903e-06,
+      "loss": 0.4223,
+      "step": 872
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.625,
+      "learning_rate": 3.4575522534413292e-06,
+      "loss": 0.4282,
+      "step": 873
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.90625,
+      "learning_rate": 3.4541767977685358e-06,
+      "loss": 0.4219,
+      "step": 874
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.53125,
+      "learning_rate": 3.4507993048116232e-06,
+      "loss": 0.3998,
+      "step": 875
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 4.375,
+      "learning_rate": 3.4474197817819498e-06,
+      "loss": 0.3882,
+      "step": 876
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.6875,
+      "learning_rate": 3.4440382358952116e-06,
+      "loss": 0.4086,
+      "step": 877
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.796875,
+      "learning_rate": 3.4406546743714215e-06,
+      "loss": 0.4492,
+      "step": 878
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.65625,
+      "learning_rate": 3.4372691044348983e-06,
+      "loss": 0.3635,
+      "step": 879
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.59375,
+      "learning_rate": 3.433881533314246e-06,
+      "loss": 0.3732,
+      "step": 880
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.59375,
+      "learning_rate": 3.4304919682423447e-06,
+      "loss": 0.45,
+      "step": 881
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.328125,
+      "learning_rate": 3.4271004164563297e-06,
+      "loss": 0.3655,
+      "step": 882
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.765625,
+      "learning_rate": 3.4237068851975787e-06,
+      "loss": 0.4153,
+      "step": 883
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.828125,
+      "learning_rate": 3.4203113817116955e-06,
+      "loss": 0.44,
+      "step": 884
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.71875,
+      "learning_rate": 3.416913913248495e-06,
+      "loss": 0.4674,
+      "step": 885
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.6875,
+      "learning_rate": 3.4135144870619894e-06,
+      "loss": 0.4149,
+      "step": 886
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 4.03125,
+      "learning_rate": 3.4101131104103664e-06,
+      "loss": 0.3748,
+      "step": 887
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.375,
+      "learning_rate": 3.4067097905559836e-06,
+      "loss": 0.3601,
+      "step": 888
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.921875,
+      "learning_rate": 3.4033045347653436e-06,
+      "loss": 0.3995,
+      "step": 889
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.65625,
+      "learning_rate": 3.399897350309085e-06,
+      "loss": 0.3941,
+      "step": 890
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.71875,
+      "learning_rate": 3.3964882444619617e-06,
+      "loss": 0.3717,
+      "step": 891
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.484375,
+      "learning_rate": 3.393077224502832e-06,
+      "loss": 0.3984,
+      "step": 892
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.625,
+      "learning_rate": 3.389664297714641e-06,
+      "loss": 0.4254,
+      "step": 893
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.828125,
+      "learning_rate": 3.386249471384405e-06,
+      "loss": 0.3885,
+      "step": 894
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.546875,
+      "learning_rate": 3.3828327528031946e-06,
+      "loss": 0.3643,
+      "step": 895
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.90625,
+      "learning_rate": 3.3794141492661214e-06,
+      "loss": 0.4155,
+      "step": 896
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.859375,
+      "learning_rate": 3.3759936680723238e-06,
+      "loss": 0.4391,
+      "step": 897
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.640625,
+      "learning_rate": 3.3725713165249455e-06,
+      "loss": 0.4113,
+      "step": 898
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.53125,
+      "learning_rate": 3.3691471019311274e-06,
+      "loss": 0.4386,
+      "step": 899
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.484375,
+      "learning_rate": 3.3657210316019833e-06,
+      "loss": 0.4539,
+      "step": 900
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.65625,
+      "learning_rate": 3.3622931128525954e-06,
+      "loss": 0.474,
+      "step": 901
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.703125,
+      "learning_rate": 3.3588633530019872e-06,
+      "loss": 0.3805,
+      "step": 902
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.515625,
+      "learning_rate": 3.3554317593731165e-06,
+      "loss": 0.4863,
+      "step": 903
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.484375,
+      "learning_rate": 3.3519983392928547e-06,
+      "loss": 0.4385,
+      "step": 904
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 4.0,
+      "learning_rate": 3.3485631000919725e-06,
+      "loss": 0.447,
+      "step": 905
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 4.09375,
+      "learning_rate": 3.345126049105128e-06,
+      "loss": 0.4457,
+      "step": 906
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 3.828125,
+      "learning_rate": 3.341687193670844e-06,
+      "loss": 0.3966,
+      "step": 907
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 3.6875,
+      "learning_rate": 3.338246541131496e-06,
+      "loss": 0.4577,
+      "step": 908
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 3.4375,
+      "learning_rate": 3.3348040988332997e-06,
+      "loss": 0.4007,
+      "step": 909
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 3.703125,
+      "learning_rate": 3.33135987412629e-06,
+      "loss": 0.3996,
+      "step": 910
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 3.75,
+      "learning_rate": 3.3279138743643064e-06,
+      "loss": 0.4179,
+      "step": 911
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 3.390625,
+      "learning_rate": 3.324466106904981e-06,
+      "loss": 0.3758,
+      "step": 912
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 3.625,
+      "learning_rate": 3.321016579109718e-06,
+      "loss": 0.3933,
+      "step": 913
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 3.390625,
+      "learning_rate": 3.3175652983436823e-06,
+      "loss": 0.3735,
+      "step": 914
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 3.765625,
+      "learning_rate": 3.314112271975779e-06,
+      "loss": 0.4038,
+      "step": 915
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 3.78125,
+      "learning_rate": 3.310657507378642e-06,
+      "loss": 0.4403,
+      "step": 916
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 3.859375,
+      "learning_rate": 3.3072010119286156e-06,
+      "loss": 0.3859,
+      "step": 917
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 3.8125,
+      "learning_rate": 3.3037427930057415e-06,
+      "loss": 0.4546,
+      "step": 918
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 3.828125,
+      "learning_rate": 3.3002828579937386e-06,
+      "loss": 0.3353,
+      "step": 919
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 3.921875,
+      "learning_rate": 3.2968212142799914e-06,
+      "loss": 0.4433,
+      "step": 920
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 3.390625,
+      "learning_rate": 3.2933578692555334e-06,
+      "loss": 0.3985,
+      "step": 921
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 3.421875,
+      "learning_rate": 3.2898928303150286e-06,
+      "loss": 0.3764,
+      "step": 922
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 3.625,
+      "learning_rate": 3.2864261048567592e-06,
+      "loss": 0.4195,
+      "step": 923
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 3.796875,
+      "learning_rate": 3.2829577002826075e-06,
+      "loss": 0.4397,
+      "step": 924
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 3.671875,
+      "learning_rate": 3.279487623998041e-06,
+      "loss": 0.4316,
+      "step": 925
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 3.640625,
+      "learning_rate": 3.2760158834120987e-06,
+      "loss": 0.4133,
+      "step": 926
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 4.0,
+      "learning_rate": 3.272542485937369e-06,
+      "loss": 0.4414,
+      "step": 927
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 3.734375,
+      "learning_rate": 3.269067438989981e-06,
+      "loss": 0.4449,
+      "step": 928
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 3.78125,
+      "learning_rate": 3.2655907499895855e-06,
+      "loss": 0.4379,
+      "step": 929
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 3.8125,
+      "learning_rate": 3.262112426359338e-06,
+      "loss": 0.3901,
+      "step": 930
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 3.796875,
+      "learning_rate": 3.258632475525884e-06,
+      "loss": 0.4661,
+      "step": 931
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 3.71875,
+      "learning_rate": 3.2551509049193445e-06,
+      "loss": 0.4043,
+      "step": 932
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 3.796875,
+      "learning_rate": 3.2516677219732996e-06,
+      "loss": 0.4439,
+      "step": 933
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 3.453125,
+      "learning_rate": 3.2481829341247695e-06,
+      "loss": 0.3864,
+      "step": 934
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 3.703125,
+      "learning_rate": 3.244696548814202e-06,
+      "loss": 0.3988,
+      "step": 935
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 3.75,
+      "learning_rate": 3.2412085734854567e-06,
+      "loss": 0.3973,
+      "step": 936
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 3.5,
+      "learning_rate": 3.237719015585787e-06,
+      "loss": 0.4248,
+      "step": 937
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 4.0,
+      "learning_rate": 3.2342278825658256e-06,
+      "loss": 0.4152,
+      "step": 938
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 3.65625,
+      "learning_rate": 3.230735181879569e-06,
+      "loss": 0.463,
+      "step": 939
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 3.75,
+      "learning_rate": 3.227240920984359e-06,
+      "loss": 0.3756,
+      "step": 940
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 3.546875,
+      "learning_rate": 3.223745107340871e-06,
+      "loss": 0.4168,
+      "step": 941
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 3.59375,
+      "learning_rate": 3.2202477484130947e-06,
+      "loss": 0.4594,
+      "step": 942
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 3.625,
+      "learning_rate": 3.216748851668317e-06,
+      "loss": 0.3625,
+      "step": 943
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 3.890625,
+      "learning_rate": 3.2132484245771127e-06,
+      "loss": 0.3584,
+      "step": 944
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 3.65625,
+      "learning_rate": 3.2097464746133194e-06,
+      "loss": 0.4511,
+      "step": 945
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 3.515625,
+      "learning_rate": 3.2062430092540302e-06,
+      "loss": 0.3626,
+      "step": 946
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 3.546875,
+      "learning_rate": 3.202738035979571e-06,
+      "loss": 0.4182,
+      "step": 947
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 3.453125,
+      "learning_rate": 3.1992315622734893e-06,
+      "loss": 0.4046,
+      "step": 948
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 4.0625,
+      "learning_rate": 3.195723595622533e-06,
+      "loss": 0.4426,
+      "step": 949
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 3.8125,
+      "learning_rate": 3.192214143516642e-06,
+      "loss": 0.4081,
+      "step": 950
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 3.75,
+      "learning_rate": 3.1887032134489243e-06,
+      "loss": 0.413,
+      "step": 951
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 3.78125,
+      "learning_rate": 3.185190812915646e-06,
+      "loss": 0.3925,
+      "step": 952
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 3.515625,
+      "learning_rate": 3.1816769494162105e-06,
+      "loss": 0.3603,
+      "step": 953
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 3.625,
+      "learning_rate": 3.178161630453146e-06,
+      "loss": 0.4031,
+      "step": 954
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 3.859375,
+      "learning_rate": 3.17464486353209e-06,
+      "loss": 0.4356,
+      "step": 955
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 3.953125,
+      "learning_rate": 3.171126656161767e-06,
+      "loss": 0.4382,
+      "step": 956
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 3.96875,
+      "learning_rate": 3.167607015853983e-06,
+      "loss": 0.3697,
+      "step": 957
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 3.765625,
+      "learning_rate": 3.164085950123597e-06,
+      "loss": 0.4666,
+      "step": 958
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 3.8125,
+      "learning_rate": 3.1605634664885175e-06,
+      "loss": 0.4582,
+      "step": 959
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 3.828125,
+      "learning_rate": 3.1570395724696772e-06,
+      "loss": 0.4404,
+      "step": 960
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 3.921875,
+      "learning_rate": 3.1535142755910204e-06,
+      "loss": 0.4418,
+      "step": 961
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 3.703125,
+      "learning_rate": 3.149987583379486e-06,
+      "loss": 0.4748,
+      "step": 962
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 3.921875,
+      "learning_rate": 3.146459503364994e-06,
+      "loss": 0.4077,
+      "step": 963
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 3.40625,
+      "learning_rate": 3.1429300430804265e-06,
+      "loss": 0.4008,
+      "step": 964
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 3.796875,
+      "learning_rate": 3.1393992100616134e-06,
+      "loss": 0.4063,
+      "step": 965
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 3.515625,
+      "learning_rate": 3.1358670118473135e-06,
+      "loss": 0.3416,
+      "step": 966
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 3.8125,
+      "learning_rate": 3.132333455979202e-06,
+      "loss": 0.4299,
+      "step": 967
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 3.6875,
+      "learning_rate": 3.128798550001854e-06,
+      "loss": 0.5074,
+      "step": 968
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 4.25,
+      "learning_rate": 3.1252623014627247e-06,
+      "loss": 0.4462,
+      "step": 969
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 4.0625,
+      "learning_rate": 3.121724717912138e-06,
+      "loss": 0.4629,
+      "step": 970
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 3.84375,
+      "learning_rate": 3.118185806903265e-06,
+      "loss": 0.3862,
+      "step": 971
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 3.671875,
+      "learning_rate": 3.1146455759921166e-06,
+      "loss": 0.441,
+      "step": 972
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 3.921875,
+      "learning_rate": 3.111104032737517e-06,
+      "loss": 0.4289,
+      "step": 973
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 3.28125,
+      "learning_rate": 3.1075611847010946e-06,
+      "loss": 0.418,
+      "step": 974
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 3.84375,
+      "learning_rate": 3.1040170394472617e-06,
+      "loss": 0.4934,
+      "step": 975
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 3.65625,
+      "learning_rate": 3.1004716045432028e-06,
+      "loss": 0.3398,
+      "step": 976
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 3.609375,
+      "learning_rate": 3.0969248875588547e-06,
+      "loss": 0.4128,
+      "step": 977
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 4.09375,
+      "learning_rate": 3.093376896066893e-06,
+      "loss": 0.4488,
+      "step": 978
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 3.59375,
+      "learning_rate": 3.089827637642711e-06,
+      "loss": 0.3887,
+      "step": 979
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 3.75,
+      "learning_rate": 3.0862771198644108e-06,
+      "loss": 0.4198,
+      "step": 980
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 4.09375,
+      "learning_rate": 3.0827253503127826e-06,
+      "loss": 0.5276,
+      "step": 981
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 3.78125,
+      "learning_rate": 3.0791723365712866e-06,
+      "loss": 0.3945,
+      "step": 982
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 3.625,
+      "learning_rate": 3.075618086226043e-06,
+      "loss": 0.4192,
+      "step": 983
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 4.0625,
+      "learning_rate": 3.0720626068658107e-06,
+      "loss": 0.4923,
+      "step": 984
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 3.59375,
+      "learning_rate": 3.0685059060819716e-06,
+      "loss": 0.4582,
+      "step": 985
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 3.796875,
+      "learning_rate": 3.0649479914685187e-06,
+      "loss": 0.4293,
+      "step": 986
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 3.5625,
+      "learning_rate": 3.0613888706220336e-06,
+      "loss": 0.3386,
+      "step": 987
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 3.5,
+      "learning_rate": 3.0578285511416744e-06,
+      "loss": 0.3811,
+      "step": 988
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 3.4375,
+      "learning_rate": 3.0542670406291587e-06,
+      "loss": 0.418,
+      "step": 989
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 3.859375,
+      "learning_rate": 3.050704346688746e-06,
+      "loss": 0.4246,
+      "step": 990
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 3.609375,
+      "learning_rate": 3.0471404769272257e-06,
+      "loss": 0.5103,
+      "step": 991
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 3.875,
+      "learning_rate": 3.043575438953893e-06,
+      "loss": 0.4198,
+      "step": 992
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 3.765625,
+      "learning_rate": 3.0400092403805407e-06,
+      "loss": 0.4479,
+      "step": 993
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 3.453125,
+      "learning_rate": 3.036441888821439e-06,
+      "loss": 0.2901,
+      "step": 994
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 3.96875,
+      "learning_rate": 3.0328733918933194e-06,
+      "loss": 0.501,
+      "step": 995
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 4.0625,
+      "learning_rate": 3.0293037572153606e-06,
+      "loss": 0.4567,
+      "step": 996
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 3.515625,
+      "learning_rate": 3.025732992409166e-06,
+      "loss": 0.4077,
+      "step": 997
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 3.703125,
+      "learning_rate": 3.022161105098757e-06,
+      "loss": 0.4727,
+      "step": 998
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 3.6875,
+      "learning_rate": 3.0185881029105494e-06,
+      "loss": 0.3883,
+      "step": 999
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 3.390625,
+      "learning_rate": 3.015013993473341e-06,
+      "loss": 0.3605,
+      "step": 1000
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 3.859375,
+      "learning_rate": 3.0114387844182896e-06,
+      "loss": 0.4482,
+      "step": 1001
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 3.46875,
+      "learning_rate": 3.007862483378906e-06,
+      "loss": 0.4399,
+      "step": 1002
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 3.65625,
+      "learning_rate": 3.0042850979910284e-06,
+      "loss": 0.3854,
+      "step": 1003
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 3.828125,
+      "learning_rate": 3.0007066358928138e-06,
+      "loss": 0.4292,
+      "step": 1004
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.9971271047247147e-06,
+      "loss": 0.4137,
+      "step": 1005
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 3.859375,
+      "learning_rate": 2.9935465121294694e-06,
+      "loss": 0.4635,
+      "step": 1006
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 3.578125,
+      "learning_rate": 2.989964865752079e-06,
+      "loss": 0.382,
+      "step": 1007
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 3.8125,
+      "learning_rate": 2.9863821732397975e-06,
+      "loss": 0.4048,
+      "step": 1008
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.9827984422421118e-06,
+      "loss": 0.3655,
+      "step": 1009
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.979213680410724e-06,
+      "loss": 0.3849,
+      "step": 1010
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 3.59375,
+      "learning_rate": 2.975627895399541e-06,
+      "loss": 0.381,
+      "step": 1011
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.9720410948646507e-06,
+      "loss": 0.4208,
+      "step": 1012
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 3.671875,
+      "learning_rate": 2.9684532864643123e-06,
+      "loss": 0.4016,
+      "step": 1013
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 3.875,
+      "learning_rate": 2.964864477858933e-06,
+      "loss": 0.3976,
+      "step": 1014
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 3.921875,
+      "learning_rate": 2.9612746767110604e-06,
+      "loss": 0.3896,
+      "step": 1015
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.9576838906853567e-06,
+      "loss": 0.4096,
+      "step": 1016
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 3.625,
+      "learning_rate": 2.9540921274485913e-06,
+      "loss": 0.3737,
+      "step": 1017
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 3.859375,
+      "learning_rate": 2.950499394669616e-06,
+      "loss": 0.371,
+      "step": 1018
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 3.703125,
+      "learning_rate": 2.9469057000193556e-06,
+      "loss": 0.3946,
+      "step": 1019
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 4.0,
+      "learning_rate": 2.9433110511707875e-06,
+      "loss": 0.4255,
+      "step": 1020
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 3.625,
+      "learning_rate": 2.939715455798925e-06,
+      "loss": 0.3464,
+      "step": 1021
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 4.0625,
+      "learning_rate": 2.9361189215808063e-06,
+      "loss": 0.3853,
+      "step": 1022
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 3.796875,
+      "learning_rate": 2.932521456195469e-06,
+      "loss": 0.4557,
+      "step": 1023
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 3.59375,
+      "learning_rate": 2.9289230673239435e-06,
+      "loss": 0.4551,
+      "step": 1024
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 3.734375,
+      "learning_rate": 2.9253237626492285e-06,
+      "loss": 0.4551,
+      "step": 1025
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 3.765625,
+      "learning_rate": 2.9217235498562814e-06,
+      "loss": 0.4127,
+      "step": 1026
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 3.78125,
+      "learning_rate": 2.9181224366319947e-06,
+      "loss": 0.4128,
+      "step": 1027
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 3.625,
+      "learning_rate": 2.9145204306651877e-06,
+      "loss": 0.3925,
+      "step": 1028
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 3.703125,
+      "learning_rate": 2.9109175396465798e-06,
+      "loss": 0.5079,
+      "step": 1029
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 3.609375,
+      "learning_rate": 2.9073137712687883e-06,
+      "loss": 0.3844,
+      "step": 1030
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 3.765625,
+      "learning_rate": 2.9037091332262957e-06,
+      "loss": 0.4274,
+      "step": 1031
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 3.8125,
+      "learning_rate": 2.9001036332154474e-06,
+      "loss": 0.4373,
+      "step": 1032
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.8964972789344254e-06,
+      "loss": 0.3966,
+      "step": 1033
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 3.765625,
+      "learning_rate": 2.892890078083237e-06,
+      "loss": 0.4511,
+      "step": 1034
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 3.9375,
+      "learning_rate": 2.8892820383636976e-06,
+      "loss": 0.4629,
+      "step": 1035
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 4.09375,
+      "learning_rate": 2.8856731674794125e-06,
+      "loss": 0.4442,
+      "step": 1036
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 3.671875,
+      "learning_rate": 2.882063473135763e-06,
+      "loss": 0.4067,
+      "step": 1037
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 4.03125,
+      "learning_rate": 2.8784529630398866e-06,
+      "loss": 0.4317,
+      "step": 1038
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 3.75,
+      "learning_rate": 2.8748416449006657e-06,
+      "loss": 0.3959,
+      "step": 1039
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 3.734375,
+      "learning_rate": 2.8712295264287036e-06,
+      "loss": 0.4341,
+      "step": 1040
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 4.125,
+      "learning_rate": 2.867616615336318e-06,
+      "loss": 0.4192,
+      "step": 1041
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 3.921875,
+      "learning_rate": 2.864002919337513e-06,
+      "loss": 0.4681,
+      "step": 1042
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 3.671875,
+      "learning_rate": 2.860388446147972e-06,
+      "loss": 0.4151,
+      "step": 1043
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.8567732034850375e-06,
+      "loss": 0.4655,
+      "step": 1044
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 3.609375,
+      "learning_rate": 2.853157199067695e-06,
+      "loss": 0.3759,
+      "step": 1045
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 3.796875,
+      "learning_rate": 2.8495404406165563e-06,
+      "loss": 0.4299,
+      "step": 1046
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 3.828125,
+      "learning_rate": 2.845922935853841e-06,
+      "loss": 0.3753,
+      "step": 1047
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 3.71875,
+      "learning_rate": 2.842304692503366e-06,
+      "loss": 0.409,
+      "step": 1048
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 3.796875,
+      "learning_rate": 2.838685718290522e-06,
+      "loss": 0.4547,
+      "step": 1049
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 4.125,
+      "learning_rate": 2.8350660209422625e-06,
+      "loss": 0.3884,
+      "step": 1050
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 3.734375,
+      "learning_rate": 2.8314456081870824e-06,
+      "loss": 0.4177,
+      "step": 1051
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 3.875,
+      "learning_rate": 2.827824487755007e-06,
+      "loss": 0.4385,
+      "step": 1052
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 3.859375,
+      "learning_rate": 2.82420266737757e-06,
+      "loss": 0.429,
+      "step": 1053
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.8205801547878024e-06,
+      "loss": 0.453,
+      "step": 1054
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 3.75,
+      "learning_rate": 2.8169569577202096e-06,
+      "loss": 0.3798,
+      "step": 1055
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 3.8125,
+      "learning_rate": 2.813333083910761e-06,
+      "loss": 0.4599,
+      "step": 1056
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 3.8125,
+      "learning_rate": 2.80970854109687e-06,
+      "loss": 0.3539,
+      "step": 1057
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 3.640625,
+      "learning_rate": 2.8060833370173796e-06,
+      "loss": 0.4328,
+      "step": 1058
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.8024574794125426e-06,
+      "loss": 0.3851,
+      "step": 1059
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 3.765625,
+      "learning_rate": 2.7988309760240086e-06,
+      "loss": 0.4148,
+      "step": 1060
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 3.734375,
+      "learning_rate": 2.795203834594806e-06,
+      "loss": 0.4064,
+      "step": 1061
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 3.625,
+      "learning_rate": 2.7915760628693254e-06,
+      "loss": 0.428,
+      "step": 1062
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 3.75,
+      "learning_rate": 2.7879476685933027e-06,
+      "loss": 0.4328,
+      "step": 1063
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 3.578125,
+      "learning_rate": 2.7843186595138023e-06,
+      "loss": 0.3538,
+      "step": 1064
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.780689043379203e-06,
+      "loss": 0.4083,
+      "step": 1065
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 3.828125,
+      "learning_rate": 2.7770588279391803e-06,
+      "loss": 0.4019,
+      "step": 1066
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 3.578125,
+      "learning_rate": 2.773428020944687e-06,
+      "loss": 0.4044,
+      "step": 1067
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.7697966301479385e-06,
+      "loss": 0.4021,
+      "step": 1068
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.7661646633023998e-06,
+      "loss": 0.3805,
+      "step": 1069
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 3.796875,
+      "learning_rate": 2.7625321281627636e-06,
+      "loss": 0.3948,
+      "step": 1070
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 3.71875,
+      "learning_rate": 2.7588990324849372e-06,
+      "loss": 0.4081,
+      "step": 1071
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 3.828125,
+      "learning_rate": 2.7552653840260233e-06,
+      "loss": 0.3842,
+      "step": 1072
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 3.84375,
+      "learning_rate": 2.7516311905443067e-06,
+      "loss": 0.4635,
+      "step": 1073
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 3.765625,
+      "learning_rate": 2.747996459799233e-06,
+      "loss": 0.4186,
+      "step": 1074
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 3.921875,
+      "learning_rate": 2.744361199551398e-06,
+      "loss": 0.3961,
+      "step": 1075
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 4.25,
+      "learning_rate": 2.7407254175625265e-06,
+      "loss": 0.4396,
+      "step": 1076
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 4.28125,
+      "learning_rate": 2.7370891215954572e-06,
+      "loss": 0.4753,
+      "step": 1077
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 3.578125,
+      "learning_rate": 2.7334523194141272e-06,
+      "loss": 0.3693,
+      "step": 1078
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.7298150187835526e-06,
+      "loss": 0.3853,
+      "step": 1079
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 3.875,
+      "learning_rate": 2.7261772274698173e-06,
+      "loss": 0.4045,
+      "step": 1080
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.7225389532400474e-06,
+      "loss": 0.3673,
+      "step": 1081
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 3.578125,
+      "learning_rate": 2.718900203862406e-06,
+      "loss": 0.4042,
+      "step": 1082
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 3.78125,
+      "learning_rate": 2.715260987106066e-06,
+      "loss": 0.4502,
+      "step": 1083
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 3.59375,
+      "learning_rate": 2.7116213107412017e-06,
+      "loss": 0.4084,
+      "step": 1084
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.7079811825389652e-06,
+      "loss": 0.3601,
+      "step": 1085
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 3.703125,
+      "learning_rate": 2.7043406102714777e-06,
+      "loss": 0.3875,
+      "step": 1086
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 3.984375,
+      "learning_rate": 2.7006996017118033e-06,
+      "loss": 0.4499,
+      "step": 1087
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.6970581646339417e-06,
+      "loss": 0.4054,
+      "step": 1088
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 3.71875,
+      "learning_rate": 2.6934163068128073e-06,
+      "loss": 0.4478,
+      "step": 1089
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 3.578125,
+      "learning_rate": 2.6897740360242102e-06,
+      "loss": 0.3883,
+      "step": 1090
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 3.9375,
+      "learning_rate": 2.6861313600448447e-06,
+      "loss": 0.4619,
+      "step": 1091
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 3.734375,
+      "learning_rate": 2.682488286652269e-06,
+      "loss": 0.4204,
+      "step": 1092
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 4.09375,
+      "learning_rate": 2.678844823624891e-06,
+      "loss": 0.4267,
+      "step": 1093
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 3.953125,
+      "learning_rate": 2.675200978741948e-06,
+      "loss": 0.4731,
+      "step": 1094
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 3.703125,
+      "learning_rate": 2.671556759783497e-06,
+      "loss": 0.3908,
+      "step": 1095
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 3.78125,
+      "learning_rate": 2.6679121745303873e-06,
+      "loss": 0.3827,
+      "step": 1096
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 3.703125,
+      "learning_rate": 2.6642672307642575e-06,
+      "loss": 0.4393,
+      "step": 1097
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 3.5,
+      "learning_rate": 2.660621936267506e-06,
+      "loss": 0.4241,
+      "step": 1098
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 3.375,
+      "learning_rate": 2.6569762988232838e-06,
+      "loss": 0.3555,
+      "step": 1099
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 3.8125,
+      "learning_rate": 2.6533303262154714e-06,
+      "loss": 0.3598,
+      "step": 1100
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 3.640625,
+      "learning_rate": 2.649684026228666e-06,
+      "loss": 0.4442,
+      "step": 1101
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 3.578125,
+      "learning_rate": 2.6460374066481654e-06,
+      "loss": 0.4024,
+      "step": 1102
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 3.5,
+      "learning_rate": 2.642390475259946e-06,
+      "loss": 0.4043,
+      "step": 1103
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 3.671875,
+      "learning_rate": 2.638743239850654e-06,
+      "loss": 0.4602,
+      "step": 1104
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 3.953125,
+      "learning_rate": 2.6350957082075816e-06,
+      "loss": 0.4289,
+      "step": 1105
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 3.640625,
+      "learning_rate": 2.631447888118656e-06,
+      "loss": 0.3833,
+      "step": 1106
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 3.25,
+      "learning_rate": 2.627799787372418e-06,
+      "loss": 0.3533,
+      "step": 1107
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 3.828125,
+      "learning_rate": 2.6241514137580103e-06,
+      "loss": 0.4018,
+      "step": 1108
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 3.828125,
+      "learning_rate": 2.620502775065154e-06,
+      "loss": 0.4066,
+      "step": 1109
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 3.90625,
+      "learning_rate": 2.6168538790841413e-06,
+      "loss": 0.4175,
+      "step": 1110
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 4.0,
+      "learning_rate": 2.6132047336058098e-06,
+      "loss": 0.4562,
+      "step": 1111
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.6095553464215323e-06,
+      "loss": 0.376,
+      "step": 1112
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.605905725323196e-06,
+      "loss": 0.3866,
+      "step": 1113
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 3.71875,
+      "learning_rate": 2.602255878103187e-06,
+      "loss": 0.3946,
+      "step": 1114
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 4.09375,
+      "learning_rate": 2.5986058125543775e-06,
+      "loss": 0.4103,
+      "step": 1115
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.5949555364701022e-06,
+      "loss": 0.4149,
+      "step": 1116
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 3.765625,
+      "learning_rate": 2.591305057644148e-06,
+      "loss": 0.3931,
+      "step": 1117
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 3.9375,
+      "learning_rate": 2.5876543838707313e-06,
+      "loss": 0.432,
+      "step": 1118
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 4.21875,
+      "learning_rate": 2.58400352294449e-06,
+      "loss": 0.5089,
+      "step": 1119
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.5803524826604563e-06,
+      "loss": 0.395,
+      "step": 1120
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 3.734375,
+      "learning_rate": 2.5767012708140486e-06,
+      "loss": 0.3844,
+      "step": 1121
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.57304989520105e-06,
+      "loss": 0.3658,
+      "step": 1122
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 3.84375,
+      "learning_rate": 2.569398363617594e-06,
+      "loss": 0.4758,
+      "step": 1123
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 3.59375,
+      "learning_rate": 2.565746683860147e-06,
+      "loss": 0.4085,
+      "step": 1124
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 3.640625,
+      "learning_rate": 2.5620948637254923e-06,
+      "loss": 0.4027,
+      "step": 1125
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 3.84375,
+      "learning_rate": 2.558442911010711e-06,
+      "loss": 0.4365,
+      "step": 1126
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.5547908335131704e-06,
+      "loss": 0.3968,
+      "step": 1127
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 3.765625,
+      "learning_rate": 2.5511386390305013e-06,
+      "loss": 0.4416,
+      "step": 1128
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.5474863353605856e-06,
+      "loss": 0.3724,
+      "step": 1129
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 3.71875,
+      "learning_rate": 2.543833930301538e-06,
+      "loss": 0.4622,
+      "step": 1130
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 3.875,
+      "learning_rate": 2.5401814316516897e-06,
+      "loss": 0.4184,
+      "step": 1131
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 3.609375,
+      "learning_rate": 2.5365288472095732e-06,
+      "loss": 0.3729,
+      "step": 1132
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.5328761847739013e-06,
+      "loss": 0.3951,
+      "step": 1133
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.5292234521435566e-06,
+      "loss": 0.3684,
+      "step": 1134
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 3.640625,
+      "learning_rate": 2.5255706571175675e-06,
+      "loss": 0.3841,
+      "step": 1135
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 3.578125,
+      "learning_rate": 2.5219178074950995e-06,
+      "loss": 0.4267,
+      "step": 1136
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.5182649110754325e-06,
+      "loss": 0.4105,
+      "step": 1137
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 3.859375,
+      "learning_rate": 2.514611975657948e-06,
+      "loss": 0.4291,
+      "step": 1138
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 3.640625,
+      "learning_rate": 2.5109590090421086e-06,
+      "loss": 0.3812,
+      "step": 1139
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 3.625,
+      "learning_rate": 2.507306019027445e-06,
+      "loss": 0.4083,
+      "step": 1140
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 3.671875,
+      "learning_rate": 2.503653013413538e-06,
+      "loss": 0.4294,
+      "step": 1141
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 3.8125,
+      "learning_rate": 2.5e-06,
+      "loss": 0.437,
+      "step": 1142
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 3.6875,
+      "learning_rate": 2.4963469865864624e-06,
+      "loss": 0.4448,
+      "step": 1143
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 3.765625,
+      "learning_rate": 2.4926939809725557e-06,
+      "loss": 0.4042,
+      "step": 1144
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 3.625,
+      "learning_rate": 2.489040990957892e-06,
+      "loss": 0.3957,
+      "step": 1145
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 3.78125,
+      "learning_rate": 2.4853880243420532e-06,
+      "loss": 0.4185,
+      "step": 1146
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 3.8125,
+      "learning_rate": 2.4817350889245675e-06,
+      "loss": 0.4249,
+      "step": 1147
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.4780821925049005e-06,
+      "loss": 0.3755,
+      "step": 1148
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.4744293428824333e-06,
+      "loss": 0.4354,
+      "step": 1149
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 3.8125,
+      "learning_rate": 2.470776547856444e-06,
+      "loss": 0.3774,
+      "step": 1150
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 3.9375,
+      "learning_rate": 2.467123815226099e-06,
+      "loss": 0.4242,
+      "step": 1151
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 3.828125,
+      "learning_rate": 2.463471152790427e-06,
+      "loss": 0.4791,
+      "step": 1152
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 4.03125,
+      "learning_rate": 2.4598185683483107e-06,
+      "loss": 0.4927,
+      "step": 1153
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 3.734375,
+      "learning_rate": 2.456166069698463e-06,
+      "loss": 0.4365,
+      "step": 1154
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 3.859375,
+      "learning_rate": 2.4525136646394157e-06,
+      "loss": 0.4706,
+      "step": 1155
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 3.75,
+      "learning_rate": 2.4488613609694995e-06,
+      "loss": 0.4274,
+      "step": 1156
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.44520916648683e-06,
+      "loss": 0.3889,
+      "step": 1157
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.4415570889892894e-06,
+      "loss": 0.3786,
+      "step": 1158
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 3.671875,
+      "learning_rate": 2.4379051362745085e-06,
+      "loss": 0.4329,
+      "step": 1159
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 3.765625,
+      "learning_rate": 2.434253316139854e-06,
+      "loss": 0.4383,
+      "step": 1160
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 3.90625,
+      "learning_rate": 2.4306016363824067e-06,
+      "loss": 0.42,
+      "step": 1161
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 3.6875,
+      "learning_rate": 2.426950104798951e-06,
+      "loss": 0.4029,
+      "step": 1162
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.4232987291859526e-06,
+      "loss": 0.3738,
+      "step": 1163
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 3.921875,
+      "learning_rate": 2.419647517339545e-06,
+      "loss": 0.4427,
+      "step": 1164
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 3.96875,
+      "learning_rate": 2.415996477055511e-06,
+      "loss": 0.4347,
+      "step": 1165
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 3.90625,
+      "learning_rate": 2.4123456161292687e-06,
+      "loss": 0.4017,
+      "step": 1166
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 3.609375,
+      "learning_rate": 2.408694942355853e-06,
+      "loss": 0.4197,
+      "step": 1167
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 3.828125,
+      "learning_rate": 2.4050444635298986e-06,
+      "loss": 0.4628,
+      "step": 1168
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 3.875,
+      "learning_rate": 2.401394187445623e-06,
+      "loss": 0.3869,
+      "step": 1169
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.397744121896814e-06,
+      "loss": 0.399,
+      "step": 1170
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 4.03125,
+      "learning_rate": 2.3940942746768054e-06,
+      "loss": 0.4387,
+      "step": 1171
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 3.796875,
+      "learning_rate": 2.390444653578469e-06,
+      "loss": 0.4369,
+      "step": 1172
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 3.921875,
+      "learning_rate": 2.386795266394191e-06,
+      "loss": 0.4919,
+      "step": 1173
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 4.0625,
+      "learning_rate": 2.3831461209158587e-06,
+      "loss": 0.4353,
+      "step": 1174
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.3794972249348465e-06,
+      "loss": 0.3441,
+      "step": 1175
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 3.8125,
+      "learning_rate": 2.3758485862419906e-06,
+      "loss": 0.4671,
+      "step": 1176
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 3.8125,
+      "learning_rate": 2.3722002126275826e-06,
+      "loss": 0.4519,
+      "step": 1177
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 4.09375,
+      "learning_rate": 2.3685521118813447e-06,
+      "loss": 0.3527,
+      "step": 1178
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 3.640625,
+      "learning_rate": 2.3649042917924192e-06,
+      "loss": 0.4612,
+      "step": 1179
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 3.734375,
+      "learning_rate": 2.3612567601493467e-06,
+      "loss": 0.4659,
+      "step": 1180
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 3.859375,
+      "learning_rate": 2.357609524740055e-06,
+      "loss": 0.3832,
+      "step": 1181
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.3539625933518354e-06,
+      "loss": 0.3603,
+      "step": 1182
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.3503159737713336e-06,
+      "loss": 0.4216,
+      "step": 1183
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.3466696737845294e-06,
+      "loss": 0.3837,
+      "step": 1184
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 3.96875,
+      "learning_rate": 2.3430237011767166e-06,
+      "loss": 0.4539,
+      "step": 1185
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 3.953125,
+      "learning_rate": 2.3393780637324943e-06,
+      "loss": 0.4181,
+      "step": 1186
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 3.96875,
+      "learning_rate": 2.3357327692357434e-06,
+      "loss": 0.506,
+      "step": 1187
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.3320878254696136e-06,
+      "loss": 0.4011,
+      "step": 1188
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 3.890625,
+      "learning_rate": 2.3284432402165048e-06,
+      "loss": 0.4328,
+      "step": 1189
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.3247990212580527e-06,
+      "loss": 0.4044,
+      "step": 1190
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.32115517637511e-06,
+      "loss": 0.4307,
+      "step": 1191
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 3.375,
+      "learning_rate": 2.317511713347731e-06,
+      "loss": 0.4218,
+      "step": 1192
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 3.78125,
+      "learning_rate": 2.3138686399551557e-06,
+      "loss": 0.4107,
+      "step": 1193
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 3.6875,
+      "learning_rate": 2.3102259639757906e-06,
+      "loss": 0.4379,
+      "step": 1194
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 3.734375,
+      "learning_rate": 2.306583693187193e-06,
+      "loss": 0.4369,
+      "step": 1195
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 3.75,
+      "learning_rate": 2.3029418353660587e-06,
+      "loss": 0.4238,
+      "step": 1196
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.2993003982881976e-06,
+      "loss": 0.3822,
+      "step": 1197
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 3.796875,
+      "learning_rate": 2.295659389728524e-06,
+      "loss": 0.3967,
+      "step": 1198
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 3.640625,
+      "learning_rate": 2.292018817461035e-06,
+      "loss": 0.3902,
+      "step": 1199
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.2883786892587987e-06,
+      "loss": 0.4249,
+      "step": 1200
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.2847390128939344e-06,
+      "loss": 0.4335,
+      "step": 1201
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 3.984375,
+      "learning_rate": 2.2810997961375943e-06,
+      "loss": 0.47,
+      "step": 1202
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.277461046759953e-06,
+      "loss": 0.3836,
+      "step": 1203
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 3.828125,
+      "learning_rate": 2.2738227725301835e-06,
+      "loss": 0.3891,
+      "step": 1204
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.2701849812164478e-06,
+      "loss": 0.3929,
+      "step": 1205
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.266547680585874e-06,
+      "loss": 0.4039,
+      "step": 1206
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.262910878404544e-06,
+      "loss": 0.4027,
+      "step": 1207
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 3.6875,
+      "learning_rate": 2.2592745824374748e-06,
+      "loss": 0.3795,
+      "step": 1208
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 3.8125,
+      "learning_rate": 2.2556388004486024e-06,
+      "loss": 0.4151,
+      "step": 1209
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.252003540200768e-06,
+      "loss": 0.3552,
+      "step": 1210
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 3.671875,
+      "learning_rate": 2.248368809455694e-06,
+      "loss": 0.3923,
+      "step": 1211
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 3.734375,
+      "learning_rate": 2.244734615973977e-06,
+      "loss": 0.3581,
+      "step": 1212
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 3.953125,
+      "learning_rate": 2.241100967515063e-06,
+      "loss": 0.4634,
+      "step": 1213
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.237467871837237e-06,
+      "loss": 0.3752,
+      "step": 1214
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 3.671875,
+      "learning_rate": 2.233835336697601e-06,
+      "loss": 0.4124,
+      "step": 1215
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 3.6875,
+      "learning_rate": 2.2302033698520627e-06,
+      "loss": 0.4057,
+      "step": 1216
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 3.859375,
+      "learning_rate": 2.2265719790553147e-06,
+      "loss": 0.422,
+      "step": 1217
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 3.625,
+      "learning_rate": 2.22294117206082e-06,
+      "loss": 0.423,
+      "step": 1218
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.2193109566207964e-06,
+      "loss": 0.4163,
+      "step": 1219
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 3.75,
+      "learning_rate": 2.215681340486198e-06,
+      "loss": 0.4991,
+      "step": 1220
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.212052331406698e-06,
+      "loss": 0.4234,
+      "step": 1221
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 3.71875,
+      "learning_rate": 2.2084239371306754e-06,
+      "loss": 0.39,
+      "step": 1222
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 3.78125,
+      "learning_rate": 2.2047961654051943e-06,
+      "loss": 0.3502,
+      "step": 1223
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 3.625,
+      "learning_rate": 2.201169023975992e-06,
+      "loss": 0.4126,
+      "step": 1224
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.1975425205874582e-06,
+      "loss": 0.3649,
+      "step": 1225
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 3.703125,
+      "learning_rate": 2.1939166629826217e-06,
+      "loss": 0.4573,
+      "step": 1226
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.19029145890313e-06,
+      "loss": 0.3746,
+      "step": 1227
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.186666916089239e-06,
+      "loss": 0.4124,
+      "step": 1228
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.183043042279791e-06,
+      "loss": 0.3371,
+      "step": 1229
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 3.9375,
+      "learning_rate": 2.179419845212198e-06,
+      "loss": 0.4155,
+      "step": 1230
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 3.609375,
+      "learning_rate": 2.1757973326224304e-06,
+      "loss": 0.428,
+      "step": 1231
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 3.828125,
+      "learning_rate": 2.1721755122449935e-06,
+      "loss": 0.4391,
+      "step": 1232
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 3.5,
+      "learning_rate": 2.1685543918129184e-06,
+      "loss": 0.4053,
+      "step": 1233
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 3.578125,
+      "learning_rate": 2.1649339790577388e-06,
+      "loss": 0.4053,
+      "step": 1234
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 3.671875,
+      "learning_rate": 2.161314281709479e-06,
+      "loss": 0.3834,
+      "step": 1235
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.1576953074966346e-06,
+      "loss": 0.4008,
+      "step": 1236
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 3.9375,
+      "learning_rate": 2.154077064146159e-06,
+      "loss": 0.4228,
+      "step": 1237
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 3.8125,
+      "learning_rate": 2.150459559383445e-06,
+      "loss": 0.4081,
+      "step": 1238
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 4.0625,
+      "learning_rate": 2.146842800932305e-06,
+      "loss": 0.4169,
+      "step": 1239
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 4.09375,
+      "learning_rate": 2.143226796514963e-06,
+      "loss": 0.4597,
+      "step": 1240
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 3.8125,
+      "learning_rate": 2.139611553852029e-06,
+      "loss": 0.3724,
+      "step": 1241
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.1359970806624886e-06,
+      "loss": 0.3996,
+      "step": 1242
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 4.15625,
+      "learning_rate": 2.1323833846636833e-06,
+      "loss": 0.4375,
+      "step": 1243
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 3.671875,
+      "learning_rate": 2.1287704735712972e-06,
+      "loss": 0.4405,
+      "step": 1244
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 3.953125,
+      "learning_rate": 2.1251583550993348e-06,
+      "loss": 0.414,
+      "step": 1245
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.121547036960114e-06,
+      "loss": 0.3974,
+      "step": 1246
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 3.953125,
+      "learning_rate": 2.1179365268642377e-06,
+      "loss": 0.4287,
+      "step": 1247
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.1143268325205883e-06,
+      "loss": 0.4153,
+      "step": 1248
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 3.59375,
+      "learning_rate": 2.110717961636303e-06,
+      "loss": 0.4174,
+      "step": 1249
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 3.765625,
+      "learning_rate": 2.107109921916764e-06,
+      "loss": 0.3653,
+      "step": 1250
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 3.828125,
+      "learning_rate": 2.1035027210655755e-06,
+      "loss": 0.454,
+      "step": 1251
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.099896366784554e-06,
+      "loss": 0.3824,
+      "step": 1252
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.0962908667737047e-06,
+      "loss": 0.397,
+      "step": 1253
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 3.6875,
+      "learning_rate": 2.092686228731212e-06,
+      "loss": 0.4191,
+      "step": 1254
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 3.703125,
+      "learning_rate": 2.08908246035342e-06,
+      "loss": 0.3889,
+      "step": 1255
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.0854795693348136e-06,
+      "loss": 0.42,
+      "step": 1256
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 3.703125,
+      "learning_rate": 2.0818775633680057e-06,
+      "loss": 0.4495,
+      "step": 1257
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 3.796875,
+      "learning_rate": 2.0782764501437195e-06,
+      "loss": 0.4581,
+      "step": 1258
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.0746762373507723e-06,
+      "loss": 0.3772,
+      "step": 1259
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.0710769326760578e-06,
+      "loss": 0.4038,
+      "step": 1260
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.0674785438045325e-06,
+      "loss": 0.3933,
+      "step": 1261
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.063881078419195e-06,
+      "loss": 0.3476,
+      "step": 1262
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 3.984375,
+      "learning_rate": 2.0602845442010753e-06,
+      "loss": 0.3748,
+      "step": 1263
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 3.84375,
+      "learning_rate": 2.0566889488292134e-06,
+      "loss": 0.4145,
+      "step": 1264
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 4.21875,
+      "learning_rate": 2.053094299980645e-06,
+      "loss": 0.3841,
+      "step": 1265
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 4.125,
+      "learning_rate": 2.0495006053303844e-06,
+      "loss": 0.4159,
+      "step": 1266
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 3.828125,
+      "learning_rate": 2.045907872551409e-06,
+      "loss": 0.4812,
+      "step": 1267
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 3.734375,
+      "learning_rate": 2.0423161093146437e-06,
+      "loss": 0.4934,
+      "step": 1268
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 3.609375,
+      "learning_rate": 2.0387253232889404e-06,
+      "loss": 0.4393,
+      "step": 1269
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 3.59375,
+      "learning_rate": 2.035135522141068e-06,
+      "loss": 0.4276,
+      "step": 1270
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 3.71875,
+      "learning_rate": 2.031546713535688e-06,
+      "loss": 0.4138,
+      "step": 1271
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 3.703125,
+      "learning_rate": 2.0279589051353493e-06,
+      "loss": 0.4144,
+      "step": 1272
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.024372104600459e-06,
+      "loss": 0.3446,
+      "step": 1273
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.0207863195892764e-06,
+      "loss": 0.4119,
+      "step": 1274
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.017201557757889e-06,
+      "loss": 0.3964,
+      "step": 1275
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.0136178267602033e-06,
+      "loss": 0.4212,
+      "step": 1276
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 3.75,
+      "learning_rate": 2.010035134247922e-06,
+      "loss": 0.414,
+      "step": 1277
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.0064534878705323e-06,
+      "loss": 0.3647,
+      "step": 1278
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 4.21875,
+      "learning_rate": 2.002872895275286e-06,
+      "loss": 0.4262,
+      "step": 1279
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 3.875,
+      "learning_rate": 1.9992933641071867e-06,
+      "loss": 0.4255,
+      "step": 1280
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 4.03125,
+      "learning_rate": 1.995714902008972e-06,
+      "loss": 0.4635,
+      "step": 1281
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.992137516621095e-06,
+      "loss": 0.4332,
+      "step": 1282
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.9885612155817112e-06,
+      "loss": 0.3941,
+      "step": 1283
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.98498600652666e-06,
+      "loss": 0.3375,
+      "step": 1284
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.981411897089451e-06,
+      "loss": 0.4404,
+      "step": 1285
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.9778388949012437e-06,
+      "loss": 0.4267,
+      "step": 1286
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.9742670075908353e-06,
+      "loss": 0.5038,
+      "step": 1287
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.970696242784641e-06,
+      "loss": 0.4085,
+      "step": 1288
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.96712660810668e-06,
+      "loss": 0.3811,
+      "step": 1289
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.9635581111785612e-06,
+      "loss": 0.4124,
+      "step": 1290
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.9599907596194597e-06,
+      "loss": 0.3543,
+      "step": 1291
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 3.859375,
+      "learning_rate": 1.956424561046108e-06,
+      "loss": 0.4977,
+      "step": 1292
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.9528595230727756e-06,
+      "loss": 0.4144,
+      "step": 1293
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 3.75,
+      "learning_rate": 1.9492956533112545e-06,
+      "loss": 0.4412,
+      "step": 1294
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.9457329593708425e-06,
+      "loss": 0.4021,
+      "step": 1295
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.9421714488583273e-06,
+      "loss": 0.3843,
+      "step": 1296
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 4.09375,
+      "learning_rate": 1.9386111293779673e-06,
+      "loss": 0.4281,
+      "step": 1297
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.9350520085314817e-06,
+      "loss": 0.3656,
+      "step": 1298
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 3.625,
+      "learning_rate": 1.9314940939180284e-06,
+      "loss": 0.4113,
+      "step": 1299
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 3.5,
+      "learning_rate": 1.92793739313419e-06,
+      "loss": 0.3952,
+      "step": 1300
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 5.65625,
+      "learning_rate": 1.9243819137739574e-06,
+      "loss": 0.396,
+      "step": 1301
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 3.375,
+      "learning_rate": 1.9208276634287143e-06,
+      "loss": 0.3775,
+      "step": 1302
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.9172746496872187e-06,
+      "loss": 0.4132,
+      "step": 1303
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 3.75,
+      "learning_rate": 1.9137228801355896e-06,
+      "loss": 0.4321,
+      "step": 1304
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.9101723623572898e-06,
+      "loss": 0.3413,
+      "step": 1305
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.9066231039331084e-06,
+      "loss": 0.4138,
+      "step": 1306
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.903075112441145e-06,
+      "loss": 0.3904,
+      "step": 1307
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.8995283954567972e-06,
+      "loss": 0.3761,
+      "step": 1308
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.8959829605527391e-06,
+      "loss": 0.4002,
+      "step": 1309
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.8924388152989065e-06,
+      "loss": 0.4054,
+      "step": 1310
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.8888959672624838e-06,
+      "loss": 0.3851,
+      "step": 1311
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.8853544240078838e-06,
+      "loss": 0.3959,
+      "step": 1312
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 3.84375,
+      "learning_rate": 1.8818141930967355e-06,
+      "loss": 0.3713,
+      "step": 1313
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.8782752820878636e-06,
+      "loss": 0.4222,
+      "step": 1314
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.8747376985372768e-06,
+      "loss": 0.3605,
+      "step": 1315
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.8712014499981468e-06,
+      "loss": 0.4102,
+      "step": 1316
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 4.15625,
+      "learning_rate": 1.8676665440207982e-06,
+      "loss": 0.4472,
+      "step": 1317
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.8641329881526875e-06,
+      "loss": 0.3928,
+      "step": 1318
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 3.828125,
+      "learning_rate": 1.8606007899383876e-06,
+      "loss": 0.4112,
+      "step": 1319
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 3.90625,
+      "learning_rate": 1.8570699569195742e-06,
+      "loss": 0.4549,
+      "step": 1320
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.8535404966350067e-06,
+      "loss": 0.4042,
+      "step": 1321
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.8500124166205152e-06,
+      "loss": 0.3693,
+      "step": 1322
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.8464857244089809e-06,
+      "loss": 0.4034,
+      "step": 1323
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.8429604275303236e-06,
+      "loss": 0.3277,
+      "step": 1324
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.8394365335114823e-06,
+      "loss": 0.4522,
+      "step": 1325
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.8359140498764033e-06,
+      "loss": 0.4111,
+      "step": 1326
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.832392984146018e-06,
+      "loss": 0.4613,
+      "step": 1327
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.8288733438382333e-06,
+      "loss": 0.3672,
+      "step": 1328
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.8253551364679112e-06,
+      "loss": 0.4131,
+      "step": 1329
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 4.0625,
+      "learning_rate": 1.8218383695468544e-06,
+      "loss": 0.4408,
+      "step": 1330
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.8183230505837903e-06,
+      "loss": 0.3611,
+      "step": 1331
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.8148091870843553e-06,
+      "loss": 0.4338,
+      "step": 1332
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.8112967865510766e-06,
+      "loss": 0.4891,
+      "step": 1333
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.8077858564833583e-06,
+      "loss": 0.4636,
+      "step": 1334
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 3.625,
+      "learning_rate": 1.8042764043774675e-06,
+      "loss": 0.3963,
+      "step": 1335
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.8007684377265117e-06,
+      "loss": 0.3821,
+      "step": 1336
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.7972619640204298e-06,
+      "loss": 0.3987,
+      "step": 1337
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.7937569907459704e-06,
+      "loss": 0.4076,
+      "step": 1338
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.7902535253866815e-06,
+      "loss": 0.4041,
+      "step": 1339
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.7867515754228884e-06,
+      "loss": 0.3735,
+      "step": 1340
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 4.0,
+      "learning_rate": 1.7832511483316839e-06,
+      "loss": 0.4577,
+      "step": 1341
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 4.0,
+      "learning_rate": 1.7797522515869062e-06,
+      "loss": 0.477,
+      "step": 1342
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 3.625,
+      "learning_rate": 1.776254892659129e-06,
+      "loss": 0.3445,
+      "step": 1343
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.7727590790156412e-06,
+      "loss": 0.3493,
+      "step": 1344
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.7692648181204316e-06,
+      "loss": 0.4463,
+      "step": 1345
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 4.15625,
+      "learning_rate": 1.7657721174341752e-06,
+      "loss": 0.4711,
+      "step": 1346
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 4.0,
+      "learning_rate": 1.7622809844142138e-06,
+      "loss": 0.4332,
+      "step": 1347
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.758791426514544e-06,
+      "loss": 0.4154,
+      "step": 1348
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.755303451185799e-06,
+      "loss": 0.4437,
+      "step": 1349
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.7518170658752314e-06,
+      "loss": 0.3922,
+      "step": 1350
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.7483322780267004e-06,
+      "loss": 0.3833,
+      "step": 1351
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.7448490950806552e-06,
+      "loss": 0.3669,
+      "step": 1352
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.7413675244741162e-06,
+      "loss": 0.4091,
+      "step": 1353
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.7378875736406632e-06,
+      "loss": 0.3936,
+      "step": 1354
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.7344092500104149e-06,
+      "loss": 0.3923,
+      "step": 1355
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 3.828125,
+      "learning_rate": 1.7309325610100196e-06,
+      "loss": 0.4324,
+      "step": 1356
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 3.875,
+      "learning_rate": 1.7274575140626318e-06,
+      "loss": 0.46,
+      "step": 1357
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.7239841165879028e-06,
+      "loss": 0.429,
+      "step": 1358
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.7205123760019595e-06,
+      "loss": 0.4183,
+      "step": 1359
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.717042299717393e-06,
+      "loss": 0.3425,
+      "step": 1360
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 3.921875,
+      "learning_rate": 1.7135738951432418e-06,
+      "loss": 0.4878,
+      "step": 1361
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.7101071696849718e-06,
+      "loss": 0.3872,
+      "step": 1362
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 3.875,
+      "learning_rate": 1.7066421307444675e-06,
+      "loss": 0.4188,
+      "step": 1363
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.7031787857200088e-06,
+      "loss": 0.3823,
+      "step": 1364
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.6997171420062625e-06,
+      "loss": 0.4429,
+      "step": 1365
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 3.953125,
+      "learning_rate": 1.6962572069942595e-06,
+      "loss": 0.4052,
+      "step": 1366
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 3.5,
+      "learning_rate": 1.6927989880713852e-06,
+      "loss": 0.4164,
+      "step": 1367
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.6893424926213593e-06,
+      "loss": 0.3827,
+      "step": 1368
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.6858877280242213e-06,
+      "loss": 0.4855,
+      "step": 1369
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 3.90625,
+      "learning_rate": 1.6824347016563186e-06,
+      "loss": 0.4322,
+      "step": 1370
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.6789834208902822e-06,
+      "loss": 0.4429,
+      "step": 1371
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 3.953125,
+      "learning_rate": 1.6755338930950194e-06,
+      "loss": 0.4295,
+      "step": 1372
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.6720861256356946e-06,
+      "loss": 0.3633,
+      "step": 1373
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 3.375,
+      "learning_rate": 1.668640125873711e-06,
+      "loss": 0.3742,
+      "step": 1374
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.6651959011667012e-06,
+      "loss": 0.4378,
+      "step": 1375
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.6617534588685047e-06,
+      "loss": 0.3964,
+      "step": 1376
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.6583128063291576e-06,
+      "loss": 0.3854,
+      "step": 1377
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.6548739508948723e-06,
+      "loss": 0.3886,
+      "step": 1378
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 3.875,
+      "learning_rate": 1.651436899908027e-06,
+      "loss": 0.4351,
+      "step": 1379
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.6480016607071463e-06,
+      "loss": 0.3781,
+      "step": 1380
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.6445682406268843e-06,
+      "loss": 0.4185,
+      "step": 1381
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.6411366469980136e-06,
+      "loss": 0.3805,
+      "step": 1382
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.6377068871474057e-06,
+      "loss": 0.3852,
+      "step": 1383
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.6342789683980175e-06,
+      "loss": 0.4123,
+      "step": 1384
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.6308528980688743e-06,
+      "loss": 0.4121,
+      "step": 1385
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.6274286834750555e-06,
+      "loss": 0.4021,
+      "step": 1386
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.6240063319276767e-06,
+      "loss": 0.4298,
+      "step": 1387
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.6205858507338784e-06,
+      "loss": 0.3985,
+      "step": 1388
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.6171672471968065e-06,
+      "loss": 0.3987,
+      "step": 1389
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 3.9375,
+      "learning_rate": 1.6137505286155958e-06,
+      "loss": 0.4484,
+      "step": 1390
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 4.25,
+      "learning_rate": 1.6103357022853594e-06,
+      "loss": 0.4618,
+      "step": 1391
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 3.5,
+      "learning_rate": 1.6069227754971683e-06,
+      "loss": 0.3491,
+      "step": 1392
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.6035117555380395e-06,
+      "loss": 0.448,
+      "step": 1393
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 4.125,
+      "learning_rate": 1.6001026496909162e-06,
+      "loss": 0.4013,
+      "step": 1394
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.5966954652346572e-06,
+      "loss": 0.3798,
+      "step": 1395
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.5932902094440166e-06,
+      "loss": 0.396,
+      "step": 1396
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.5898868895896336e-06,
+      "loss": 0.4705,
+      "step": 1397
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.5864855129380114e-06,
+      "loss": 0.4266,
+      "step": 1398
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.5830860867515053e-06,
+      "loss": 0.4258,
+      "step": 1399
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 3.890625,
+      "learning_rate": 1.5796886182883053e-06,
+      "loss": 0.4868,
+      "step": 1400
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 3.875,
+      "learning_rate": 1.5762931148024223e-06,
+      "loss": 0.4304,
+      "step": 1401
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 3.375,
+      "learning_rate": 1.5728995835436711e-06,
+      "loss": 0.4231,
+      "step": 1402
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 3.5,
+      "learning_rate": 1.5695080317576561e-06,
+      "loss": 0.4124,
+      "step": 1403
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.5661184666857545e-06,
+      "loss": 0.4292,
+      "step": 1404
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 3.625,
+      "learning_rate": 1.562730895565102e-06,
+      "loss": 0.4084,
+      "step": 1405
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.559345325628579e-06,
+      "loss": 0.3964,
+      "step": 1406
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 3.625,
+      "learning_rate": 1.5559617641047886e-06,
+      "loss": 0.3852,
+      "step": 1407
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 3.625,
+      "learning_rate": 1.5525802182180508e-06,
+      "loss": 0.4844,
+      "step": 1408
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 3.375,
+      "learning_rate": 1.5492006951883776e-06,
+      "loss": 0.346,
+      "step": 1409
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 3.625,
+      "learning_rate": 1.5458232022314644e-06,
+      "loss": 0.3859,
+      "step": 1410
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.5424477465586718e-06,
+      "loss": 0.4181,
+      "step": 1411
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.539074335377011e-06,
+      "loss": 0.3582,
+      "step": 1412
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 3.84375,
+      "learning_rate": 1.535702975889125e-06,
+      "loss": 0.4832,
+      "step": 1413
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.5323336752932811e-06,
+      "loss": 0.4457,
+      "step": 1414
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.5289664407833468e-06,
+      "loss": 0.3654,
+      "step": 1415
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 4.125,
+      "learning_rate": 1.5256012795487797e-06,
+      "loss": 0.4592,
+      "step": 1416
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.5222381987746104e-06,
+      "loss": 0.3998,
+      "step": 1417
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 3.859375,
+      "learning_rate": 1.5188772056414271e-06,
+      "loss": 0.4665,
+      "step": 1418
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 4.125,
+      "learning_rate": 1.515518307325362e-06,
+      "loss": 0.3665,
+      "step": 1419
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.5121615109980727e-06,
+      "loss": 0.3742,
+      "step": 1420
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.5088068238267306e-06,
+      "loss": 0.4016,
+      "step": 1421
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.505454252974001e-06,
+      "loss": 0.468,
+      "step": 1422
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 3.96875,
+      "learning_rate": 1.5021038055980359e-06,
+      "loss": 0.4283,
+      "step": 1423
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.498755488852447e-06,
+      "loss": 0.3642,
+      "step": 1424
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 3.5,
+      "learning_rate": 1.4954093098863021e-06,
+      "loss": 0.4267,
+      "step": 1425
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.4920652758441012e-06,
+      "loss": 0.3781,
+      "step": 1426
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.488723393865766e-06,
+      "loss": 0.4283,
+      "step": 1427
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.4853836710866233e-06,
+      "loss": 0.3629,
+      "step": 1428
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.4820461146373893e-06,
+      "loss": 0.449,
+      "step": 1429
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.4787107316441552e-06,
+      "loss": 0.433,
+      "step": 1430
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 3.625,
+      "learning_rate": 1.4753775292283707e-06,
+      "loss": 0.4169,
+      "step": 1431
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.4720465145068321e-06,
+      "loss": 0.4426,
+      "step": 1432
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 3.5,
+      "learning_rate": 1.468717694591662e-06,
+      "loss": 0.376,
+      "step": 1433
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 3.90625,
+      "learning_rate": 1.4653910765902985e-06,
+      "loss": 0.3959,
+      "step": 1434
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.4620666676054768e-06,
+      "loss": 0.4617,
+      "step": 1435
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.458744474735218e-06,
+      "loss": 0.436,
+      "step": 1436
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 4.0,
+      "learning_rate": 1.4554245050728085e-06,
+      "loss": 0.387,
+      "step": 1437
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 3.90625,
+      "learning_rate": 1.4521067657067894e-06,
+      "loss": 0.4118,
+      "step": 1438
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 3.984375,
+      "learning_rate": 1.4487912637209415e-06,
+      "loss": 0.4833,
+      "step": 1439
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 3.5,
+      "learning_rate": 1.4454780061942647e-06,
+      "loss": 0.3602,
+      "step": 1440
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.4421670002009714e-06,
+      "loss": 0.4177,
+      "step": 1441
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.4388582528104628e-06,
+      "loss": 0.3515,
+      "step": 1442
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.4355517710873184e-06,
+      "loss": 0.414,
+      "step": 1443
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 3.875,
+      "learning_rate": 1.4322475620912824e-06,
+      "loss": 0.4104,
+      "step": 1444
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.4289456328772444e-06,
+      "loss": 0.3821,
+      "step": 1445
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.4256459904952273e-06,
+      "loss": 0.4136,
+      "step": 1446
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 3.625,
+      "learning_rate": 1.4223486419903692e-06,
+      "loss": 0.4188,
+      "step": 1447
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 3.75,
+      "learning_rate": 1.4190535944029148e-06,
+      "loss": 0.4012,
+      "step": 1448
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.415760854768191e-06,
+      "loss": 0.413,
+      "step": 1449
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 4.0,
+      "learning_rate": 1.4124704301166025e-06,
+      "loss": 0.4543,
+      "step": 1450
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.4091823274736055e-06,
+      "loss": 0.4092,
+      "step": 1451
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 3.75,
+      "learning_rate": 1.4058965538597033e-06,
+      "loss": 0.4655,
+      "step": 1452
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.4026131162904233e-06,
+      "loss": 0.4228,
+      "step": 1453
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.399332021776306e-06,
+      "loss": 0.4073,
+      "step": 1454
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.3960532773228886e-06,
+      "loss": 0.3469,
+      "step": 1455
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.3927768899306929e-06,
+      "loss": 0.3828,
+      "step": 1456
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.389502866595206e-06,
+      "loss": 0.3752,
+      "step": 1457
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.3862312143068668e-06,
+      "loss": 0.4742,
+      "step": 1458
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 4.09375,
+      "learning_rate": 1.3829619400510553e-06,
+      "loss": 0.4533,
+      "step": 1459
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 3.953125,
+      "learning_rate": 1.379695050808069e-06,
+      "loss": 0.447,
+      "step": 1460
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.3764305535531183e-06,
+      "loss": 0.3421,
+      "step": 1461
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.373168455256303e-06,
+      "loss": 0.3931,
+      "step": 1462
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.3699087628826018e-06,
+      "loss": 0.4535,
+      "step": 1463
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.3666514833918553e-06,
+      "loss": 0.4222,
+      "step": 1464
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 3.890625,
+      "learning_rate": 1.3633966237387557e-06,
+      "loss": 0.3838,
+      "step": 1465
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 3.625,
+      "learning_rate": 1.3601441908728246e-06,
+      "loss": 0.3443,
+      "step": 1466
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 3.375,
+      "learning_rate": 1.3568941917384038e-06,
+      "loss": 0.3906,
+      "step": 1467
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 3.828125,
+      "learning_rate": 1.3536466332746402e-06,
+      "loss": 0.4028,
+      "step": 1468
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 3.625,
+      "learning_rate": 1.3504015224154659e-06,
+      "loss": 0.3774,
+      "step": 1469
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 3.25,
+      "learning_rate": 1.3471588660895924e-06,
+      "loss": 0.3528,
+      "step": 1470
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.3439186712204852e-06,
+      "loss": 0.3787,
+      "step": 1471
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.3406809447263568e-06,
+      "loss": 0.3809,
+      "step": 1472
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 4.21875,
+      "learning_rate": 1.337445693520149e-06,
+      "loss": 0.371,
+      "step": 1473
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.3342129245095193e-06,
+      "loss": 0.3736,
+      "step": 1474
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.330982644596825e-06,
+      "loss": 0.3996,
+      "step": 1475
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.3277548606791063e-06,
+      "loss": 0.4094,
+      "step": 1476
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 3.859375,
+      "learning_rate": 1.324529579648079e-06,
+      "loss": 0.4403,
+      "step": 1477
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.3213068083901123e-06,
+      "loss": 0.3783,
+      "step": 1478
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 4.125,
+      "learning_rate": 1.3180865537862164e-06,
+      "loss": 0.4529,
+      "step": 1479
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.3148688227120285e-06,
+      "loss": 0.3785,
+      "step": 1480
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 3.90625,
+      "learning_rate": 1.3116536220377967e-06,
+      "loss": 0.3879,
+      "step": 1481
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.3084409586283697e-06,
+      "loss": 0.4051,
+      "step": 1482
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.3052308393431754e-06,
+      "loss": 0.3878,
+      "step": 1483
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.3020232710362113e-06,
+      "loss": 0.4208,
+      "step": 1484
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.2988182605560264e-06,
+      "loss": 0.3545,
+      "step": 1485
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.2956158147457116e-06,
+      "loss": 0.3805,
+      "step": 1486
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.2924159404428804e-06,
+      "loss": 0.3974,
+      "step": 1487
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.2892186444796557e-06,
+      "loss": 0.3445,
+      "step": 1488
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 3.890625,
+      "learning_rate": 1.286023933682654e-06,
+      "loss": 0.4193,
+      "step": 1489
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.2828318148729736e-06,
+      "loss": 0.3466,
+      "step": 1490
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.2796422948661798e-06,
+      "loss": 0.356,
+      "step": 1491
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.2764553804722868e-06,
+      "loss": 0.3906,
+      "step": 1492
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.2732710784957458e-06,
+      "loss": 0.4156,
+      "step": 1493
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 3.375,
+      "learning_rate": 1.2700893957354326e-06,
+      "loss": 0.356,
+      "step": 1494
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.2669103389846268e-06,
+      "loss": 0.4296,
+      "step": 1495
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.2637339150310052e-06,
+      "loss": 0.3567,
+      "step": 1496
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.2605601306566206e-06,
+      "loss": 0.3681,
+      "step": 1497
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 4.0,
+      "learning_rate": 1.2573889926378905e-06,
+      "loss": 0.4172,
+      "step": 1498
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.254220507745581e-06,
+      "loss": 0.3533,
+      "step": 1499
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.2510546827447972e-06,
+      "loss": 0.3596,
+      "step": 1500
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.2478915243949612e-06,
+      "loss": 0.4189,
+      "step": 1501
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.244731039449802e-06,
+      "loss": 0.49,
+      "step": 1502
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 3.75,
+      "learning_rate": 1.2415732346573422e-06,
+      "loss": 0.4445,
+      "step": 1503
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.238418116759882e-06,
+      "loss": 0.3813,
+      "step": 1504
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.2352656924939832e-06,
+      "loss": 0.3848,
+      "step": 1505
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.2321159685904569e-06,
+      "loss": 0.4179,
+      "step": 1506
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.2289689517743475e-06,
+      "loss": 0.3462,
+      "step": 1507
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 3.75,
+      "learning_rate": 1.2258246487649229e-06,
+      "loss": 0.4033,
+      "step": 1508
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 3.09375,
+      "learning_rate": 1.2226830662756533e-06,
+      "loss": 0.3189,
+      "step": 1509
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.2195442110142017e-06,
+      "loss": 0.3486,
+      "step": 1510
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.2164080896824063e-06,
+      "loss": 0.3731,
+      "step": 1511
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.2132747089762712e-06,
+      "loss": 0.3829,
+      "step": 1512
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.2101440755859475e-06,
+      "loss": 0.3516,
+      "step": 1513
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.20701619619572e-06,
+      "loss": 0.3984,
+      "step": 1514
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 3.75,
+      "learning_rate": 1.203891077483993e-06,
+      "loss": 0.4093,
+      "step": 1515
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.2007687261232768e-06,
+      "loss": 0.3261,
+      "step": 1516
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.1976491487801747e-06,
+      "loss": 0.3833,
+      "step": 1517
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.1945323521153646e-06,
+      "loss": 0.4144,
+      "step": 1518
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 3.625,
+      "learning_rate": 1.191418342783589e-06,
+      "loss": 0.3674,
+      "step": 1519
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.1883071274336373e-06,
+      "loss": 0.3696,
+      "step": 1520
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.1851987127083352e-06,
+      "loss": 0.4526,
+      "step": 1521
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.1820931052445297e-06,
+      "loss": 0.3665,
+      "step": 1522
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.1789903116730714e-06,
+      "loss": 0.3431,
+      "step": 1523
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.1758903386188037e-06,
+      "loss": 0.3569,
+      "step": 1524
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.172793192700547e-06,
+      "loss": 0.3754,
+      "step": 1525
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.1696988805310888e-06,
+      "loss": 0.365,
+      "step": 1526
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 3.828125,
+      "learning_rate": 1.1666074087171628e-06,
+      "loss": 0.4785,
+      "step": 1527
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 3.375,
+      "learning_rate": 1.1635187838594384e-06,
+      "loss": 0.3779,
+      "step": 1528
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 3.5,
+      "learning_rate": 1.160433012552508e-06,
+      "loss": 0.3619,
+      "step": 1529
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.921875,
+      "learning_rate": 1.1573501013848717e-06,
+      "loss": 0.2947,
+      "step": 1530
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.1542700569389207e-06,
+      "loss": 0.3748,
+      "step": 1531
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 3.625,
+      "learning_rate": 1.1511928857909266e-06,
+      "loss": 0.4562,
+      "step": 1532
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.1481185945110251e-06,
+      "loss": 0.3702,
+      "step": 1533
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.1450471896632056e-06,
+      "loss": 0.3833,
+      "step": 1534
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.141978677805292e-06,
+      "loss": 0.3575,
+      "step": 1535
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.138913065488933e-06,
+      "loss": 0.449,
+      "step": 1536
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.1358503592595837e-06,
+      "loss": 0.4174,
+      "step": 1537
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.1327905656564981e-06,
+      "loss": 0.3675,
+      "step": 1538
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.12973369121271e-06,
+      "loss": 0.4058,
+      "step": 1539
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.1266797424550195e-06,
+      "loss": 0.3959,
+      "step": 1540
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.1236287259039805e-06,
+      "loss": 0.3687,
+      "step": 1541
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.1205806480738852e-06,
+      "loss": 0.3545,
+      "step": 1542
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 3.375,
+      "learning_rate": 1.1175355154727546e-06,
+      "loss": 0.3791,
+      "step": 1543
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 3.5,
+      "learning_rate": 1.114493334602318e-06,
+      "loss": 0.3701,
+      "step": 1544
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.1114541119580033e-06,
+      "loss": 0.4256,
+      "step": 1545
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.108417854028922e-06,
+      "loss": 0.3796,
+      "step": 1546
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.1053845672978567e-06,
+      "loss": 0.3958,
+      "step": 1547
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.1023542582412461e-06,
+      "loss": 0.4231,
+      "step": 1548
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.0993269333291695e-06,
+      "loss": 0.4101,
+      "step": 1549
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 3.625,
+      "learning_rate": 1.0963025990253364e-06,
+      "loss": 0.4389,
+      "step": 1550
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.0932812617870688e-06,
+      "loss": 0.3625,
+      "step": 1551
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.0902629280652933e-06,
+      "loss": 0.4064,
+      "step": 1552
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.0872476043045198e-06,
+      "loss": 0.3989,
+      "step": 1553
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.084235296942833e-06,
+      "loss": 0.3886,
+      "step": 1554
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.081226012411879e-06,
+      "loss": 0.3104,
+      "step": 1555
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 4.0,
+      "learning_rate": 1.0782197571368463e-06,
+      "loss": 0.4517,
+      "step": 1556
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.0752165375364593e-06,
+      "loss": 0.4479,
+      "step": 1557
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.0722163600229581e-06,
+      "loss": 0.4605,
+      "step": 1558
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.0692192310020878e-06,
+      "loss": 0.3438,
+      "step": 1559
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.0662251568730867e-06,
+      "loss": 0.3794,
+      "step": 1560
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.0632341440286687e-06,
+      "loss": 0.4225,
+      "step": 1561
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.060246198855011e-06,
+      "loss": 0.3351,
+      "step": 1562
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.0572613277317423e-06,
+      "loss": 0.3861,
+      "step": 1563
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 3.15625,
+      "learning_rate": 1.054279537031928e-06,
+      "loss": 0.3991,
+      "step": 1564
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.0513008331220549e-06,
+      "loss": 0.3682,
+      "step": 1565
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.0483252223620208e-06,
+      "loss": 0.3757,
+      "step": 1566
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 3.125,
+      "learning_rate": 1.0453527111051183e-06,
+      "loss": 0.3347,
+      "step": 1567
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.0423833056980215e-06,
+      "loss": 0.3704,
+      "step": 1568
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.0394170124807753e-06,
+      "loss": 0.3902,
+      "step": 1569
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.036453837786778e-06,
+      "loss": 0.4423,
+      "step": 1570
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.0334937879427694e-06,
+      "loss": 0.3658,
+      "step": 1571
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.0305368692688175e-06,
+      "loss": 0.3204,
+      "step": 1572
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.0275830880783052e-06,
+      "loss": 0.2994,
+      "step": 1573
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.0246324506779182e-06,
+      "loss": 0.3505,
+      "step": 1574
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.0216849633676268e-06,
+      "loss": 0.385,
+      "step": 1575
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.0187406324406768e-06,
+      "loss": 0.4414,
+      "step": 1576
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.0157994641835737e-06,
+      "loss": 0.387,
+      "step": 1577
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.0128614648760735e-06,
+      "loss": 0.3766,
+      "step": 1578
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.0099266407911624e-06,
+      "loss": 0.3633,
+      "step": 1579
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.0069949981950483e-06,
+      "loss": 0.375,
+      "step": 1580
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.0040665433471481e-06,
+      "loss": 0.3271,
+      "step": 1581
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.0011412825000693e-06,
+      "loss": 0.3672,
+      "step": 1582
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 3.640625,
+      "learning_rate": 9.982192218996028e-07,
+      "loss": 0.4213,
+      "step": 1583
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 3.75,
+      "learning_rate": 9.95300367784705e-07,
+      "loss": 0.422,
+      "step": 1584
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 3.3125,
+      "learning_rate": 9.92384726387485e-07,
+      "loss": 0.3418,
+      "step": 1585
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 3.703125,
+      "learning_rate": 9.894723039331955e-07,
+      "loss": 0.4724,
+      "step": 1586
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 3.640625,
+      "learning_rate": 9.865631066402138e-07,
+      "loss": 0.3874,
+      "step": 1587
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.836571407200313e-07,
+      "loss": 0.3305,
+      "step": 1588
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.807544123772401e-07,
+      "loss": 0.3173,
+      "step": 1589
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 3.6875,
+      "learning_rate": 9.778549278095217e-07,
+      "loss": 0.4212,
+      "step": 1590
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 3.5625,
+      "learning_rate": 9.749586932076283e-07,
+      "loss": 0.3473,
+      "step": 1591
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 3.5,
+      "learning_rate": 9.72065714755377e-07,
+      "loss": 0.3548,
+      "step": 1592
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 3.78125,
+      "learning_rate": 9.69175998629629e-07,
+      "loss": 0.3973,
+      "step": 1593
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 3.5625,
+      "learning_rate": 9.662895510002813e-07,
+      "loss": 0.3538,
+      "step": 1594
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.63406378030254e-07,
+      "loss": 0.3311,
+      "step": 1595
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 3.46875,
+      "learning_rate": 9.605264858754723e-07,
+      "loss": 0.3717,
+      "step": 1596
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 3.734375,
+      "learning_rate": 9.576498806848592e-07,
+      "loss": 0.3943,
+      "step": 1597
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 3.5,
+      "learning_rate": 9.547765686003169e-07,
+      "loss": 0.3768,
+      "step": 1598
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 3.546875,
+      "learning_rate": 9.519065557567201e-07,
+      "loss": 0.4423,
+      "step": 1599
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.49039848281895e-07,
+      "loss": 0.3358,
+      "step": 1600
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 3.71875,
+      "learning_rate": 9.461764522966146e-07,
+      "loss": 0.4161,
+      "step": 1601
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 3.578125,
+      "learning_rate": 9.433163739145773e-07,
+      "loss": 0.442,
+      "step": 1602
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 3.34375,
+      "learning_rate": 9.404596192424023e-07,
+      "loss": 0.3474,
+      "step": 1603
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 3.546875,
+      "learning_rate": 9.376061943796092e-07,
+      "loss": 0.4123,
+      "step": 1604
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.347561054186083e-07,
+      "loss": 0.3717,
+      "step": 1605
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 3.28125,
+      "learning_rate": 9.319093584446878e-07,
+      "loss": 0.3627,
+      "step": 1606
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 3.578125,
+      "learning_rate": 9.290659595360019e-07,
+      "loss": 0.4002,
+      "step": 1607
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 3.53125,
+      "learning_rate": 9.262259147635544e-07,
+      "loss": 0.4009,
+      "step": 1608
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 3.5625,
+      "learning_rate": 9.233892301911868e-07,
+      "loss": 0.3833,
+      "step": 1609
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 3.25,
+      "learning_rate": 9.205559118755695e-07,
+      "loss": 0.3614,
+      "step": 1610
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 3.40625,
+      "learning_rate": 9.177259658661816e-07,
+      "loss": 0.39,
+      "step": 1611
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 3.484375,
+      "learning_rate": 9.148993982053059e-07,
+      "loss": 0.3913,
+      "step": 1612
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 3.46875,
+      "learning_rate": 9.120762149280085e-07,
+      "loss": 0.361,
+      "step": 1613
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 3.46875,
+      "learning_rate": 9.092564220621314e-07,
+      "loss": 0.3489,
+      "step": 1614
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 3.296875,
+      "learning_rate": 9.064400256282757e-07,
+      "loss": 0.3704,
+      "step": 1615
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 3.328125,
+      "learning_rate": 9.036270316397944e-07,
+      "loss": 0.3411,
+      "step": 1616
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 3.375,
+      "learning_rate": 9.008174461027724e-07,
+      "loss": 0.3962,
+      "step": 1617
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 3.5625,
+      "learning_rate": 8.980112750160178e-07,
+      "loss": 0.3285,
+      "step": 1618
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 3.40625,
+      "learning_rate": 8.952085243710501e-07,
+      "loss": 0.4118,
+      "step": 1619
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 3.5625,
+      "learning_rate": 8.924092001520831e-07,
+      "loss": 0.3352,
+      "step": 1620
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 3.578125,
+      "learning_rate": 8.896133083360184e-07,
+      "loss": 0.3851,
+      "step": 1621
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 3.65625,
+      "learning_rate": 8.868208548924254e-07,
+      "loss": 0.355,
+      "step": 1622
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 3.703125,
+      "learning_rate": 8.840318457835337e-07,
+      "loss": 0.4127,
+      "step": 1623
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 3.765625,
+      "learning_rate": 8.812462869642183e-07,
+      "loss": 0.4202,
+      "step": 1624
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 3.46875,
+      "learning_rate": 8.784641843819891e-07,
+      "loss": 0.4238,
+      "step": 1625
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 3.390625,
+      "learning_rate": 8.756855439769749e-07,
+      "loss": 0.3629,
+      "step": 1626
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 3.3125,
+      "learning_rate": 8.729103716819113e-07,
+      "loss": 0.3626,
+      "step": 1627
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 3.390625,
+      "learning_rate": 8.701386734221312e-07,
+      "loss": 0.3832,
+      "step": 1628
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 3.546875,
+      "learning_rate": 8.673704551155507e-07,
+      "loss": 0.3504,
+      "step": 1629
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 3.5,
+      "learning_rate": 8.646057226726532e-07,
+      "loss": 0.3742,
+      "step": 1630
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.125,
+      "learning_rate": 8.618444819964802e-07,
+      "loss": 0.3572,
+      "step": 1631
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.8125,
+      "learning_rate": 8.59086738982618e-07,
+      "loss": 0.4097,
+      "step": 1632
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.5,
+      "learning_rate": 8.56332499519186e-07,
+      "loss": 0.4249,
+      "step": 1633
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.859375,
+      "learning_rate": 8.535817694868217e-07,
+      "loss": 0.4312,
+      "step": 1634
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.515625,
+      "learning_rate": 8.508345547586699e-07,
+      "loss": 0.3497,
+      "step": 1635
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.71875,
+      "learning_rate": 8.480908612003694e-07,
+      "loss": 0.4075,
+      "step": 1636
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.75,
+      "learning_rate": 8.453506946700419e-07,
+      "loss": 0.4115,
+      "step": 1637
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 3.84375,
+      "learning_rate": 8.426140610182787e-07,
+      "loss": 0.3918,
+      "step": 1638
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 3.375,
+      "learning_rate": 8.398809660881266e-07,
+      "loss": 0.4016,
+      "step": 1639
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 3.640625,
+      "learning_rate": 8.371514157150775e-07,
+      "loss": 0.3916,
+      "step": 1640
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 3.53125,
+      "learning_rate": 8.344254157270543e-07,
+      "loss": 0.3778,
+      "step": 1641
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 3.78125,
+      "learning_rate": 8.317029719444017e-07,
+      "loss": 0.4009,
+      "step": 1642
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 3.59375,
+      "learning_rate": 8.289840901798696e-07,
+      "loss": 0.4369,
+      "step": 1643
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 3.71875,
+      "learning_rate": 8.262687762386015e-07,
+      "loss": 0.416,
+      "step": 1644
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 3.0625,
+      "learning_rate": 8.235570359181269e-07,
+      "loss": 0.2966,
+      "step": 1645
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 3.375,
+      "learning_rate": 8.208488750083409e-07,
+      "loss": 0.3459,
+      "step": 1646
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 3.53125,
+      "learning_rate": 8.181442992915001e-07,
+      "loss": 0.3355,
+      "step": 1647
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 3.296875,
+      "learning_rate": 8.154433145422033e-07,
+      "loss": 0.3867,
+      "step": 1648
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 3.765625,
+      "learning_rate": 8.127459265273832e-07,
+      "loss": 0.3631,
+      "step": 1649
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 3.703125,
+      "learning_rate": 8.100521410062923e-07,
+      "loss": 0.4557,
+      "step": 1650
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 3.609375,
+      "learning_rate": 8.073619637304933e-07,
+      "loss": 0.442,
+      "step": 1651
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 3.421875,
+      "learning_rate": 8.046754004438428e-07,
+      "loss": 0.3819,
+      "step": 1652
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 3.578125,
+      "learning_rate": 8.019924568824817e-07,
+      "loss": 0.3294,
+      "step": 1653
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 3.171875,
+      "learning_rate": 7.993131387748227e-07,
+      "loss": 0.3656,
+      "step": 1654
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 3.9375,
+      "learning_rate": 7.966374518415385e-07,
+      "loss": 0.4466,
+      "step": 1655
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 3.578125,
+      "learning_rate": 7.93965401795547e-07,
+      "loss": 0.4354,
+      "step": 1656
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 3.3125,
+      "learning_rate": 7.912969943420018e-07,
+      "loss": 0.3513,
+      "step": 1657
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 3.578125,
+      "learning_rate": 7.886322351782782e-07,
+      "loss": 0.3817,
+      "step": 1658
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 3.421875,
+      "learning_rate": 7.859711299939645e-07,
+      "loss": 0.378,
+      "step": 1659
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 3.328125,
+      "learning_rate": 7.833136844708447e-07,
+      "loss": 0.3341,
+      "step": 1660
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3.234375,
+      "learning_rate": 7.806599042828902e-07,
+      "loss": 0.3482,
+      "step": 1661
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3.5625,
+      "learning_rate": 7.780097950962448e-07,
+      "loss": 0.342,
+      "step": 1662
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3.640625,
+      "learning_rate": 7.753633625692167e-07,
+      "loss": 0.3946,
+      "step": 1663
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3.515625,
+      "learning_rate": 7.72720612352264e-07,
+      "loss": 0.3584,
+      "step": 1664
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3.3125,
+      "learning_rate": 7.700815500879802e-07,
+      "loss": 0.338,
+      "step": 1665
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3.84375,
+      "learning_rate": 7.674461814110861e-07,
+      "loss": 0.4917,
+      "step": 1666
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3.71875,
+      "learning_rate": 7.648145119484152e-07,
+      "loss": 0.4232,
+      "step": 1667
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 3.640625,
+      "learning_rate": 7.621865473189052e-07,
+      "loss": 0.4347,
+      "step": 1668
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 3.515625,
+      "learning_rate": 7.595622931335813e-07,
+      "loss": 0.3468,
+      "step": 1669
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 3.765625,
+      "learning_rate": 7.569417549955466e-07,
+      "loss": 0.3808,
+      "step": 1670
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 3.6875,
+      "learning_rate": 7.543249384999698e-07,
+      "loss": 0.3776,
+      "step": 1671
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 3.4375,
+      "learning_rate": 7.517118492340749e-07,
+      "loss": 0.3842,
+      "step": 1672
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 3.546875,
+      "learning_rate": 7.491024927771276e-07,
+      "loss": 0.3926,
+      "step": 1673
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 3.40625,
+      "learning_rate": 7.464968747004223e-07,
+      "loss": 0.3593,
+      "step": 1674
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 3.71875,
+      "learning_rate": 7.438950005672721e-07,
+      "loss": 0.4482,
+      "step": 1675
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 3.875,
+      "learning_rate": 7.412968759329956e-07,
+      "loss": 0.4581,
+      "step": 1676
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 3.5,
+      "learning_rate": 7.387025063449082e-07,
+      "loss": 0.3529,
+      "step": 1677
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 3.734375,
+      "learning_rate": 7.361118973423051e-07,
+      "loss": 0.4199,
+      "step": 1678
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 3.71875,
+      "learning_rate": 7.335250544564538e-07,
+      "loss": 0.4056,
+      "step": 1679
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 3.390625,
+      "learning_rate": 7.30941983210578e-07,
+      "loss": 0.32,
+      "step": 1680
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 3.453125,
+      "learning_rate": 7.283626891198536e-07,
+      "loss": 0.3632,
+      "step": 1681
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 3.578125,
+      "learning_rate": 7.25787177691388e-07,
+      "loss": 0.3602,
+      "step": 1682
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.453125,
+      "learning_rate": 7.232154544242126e-07,
+      "loss": 0.3817,
+      "step": 1683
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.25,
+      "learning_rate": 7.2064752480927e-07,
+      "loss": 0.3533,
+      "step": 1684
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.453125,
+      "learning_rate": 7.180833943294058e-07,
+      "loss": 0.3331,
+      "step": 1685
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.546875,
+      "learning_rate": 7.155230684593512e-07,
+      "loss": 0.4516,
+      "step": 1686
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.671875,
+      "learning_rate": 7.129665526657145e-07,
+      "loss": 0.3638,
+      "step": 1687
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.375,
+      "learning_rate": 7.104138524069695e-07,
+      "loss": 0.3736,
+      "step": 1688
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.34375,
+      "learning_rate": 7.078649731334436e-07,
+      "loss": 0.3694,
+      "step": 1689
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.625,
+      "learning_rate": 7.053199202873063e-07,
+      "loss": 0.4359,
+      "step": 1690
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.59375,
+      "learning_rate": 7.027786993025562e-07,
+      "loss": 0.3536,
+      "step": 1691
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.609375,
+      "learning_rate": 7.002413156050109e-07,
+      "loss": 0.388,
+      "step": 1692
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.765625,
+      "learning_rate": 6.977077746122935e-07,
+      "loss": 0.3803,
+      "step": 1693
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.703125,
+      "learning_rate": 6.951780817338258e-07,
+      "loss": 0.4653,
+      "step": 1694
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.453125,
+      "learning_rate": 6.926522423708104e-07,
+      "loss": 0.3904,
+      "step": 1695
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.453125,
+      "learning_rate": 6.901302619162229e-07,
+      "loss": 0.3593,
+      "step": 1696
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.53125,
+      "learning_rate": 6.876121457547996e-07,
+      "loss": 0.3927,
+      "step": 1697
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 3.375,
+      "learning_rate": 6.850978992630269e-07,
+      "loss": 0.3865,
+      "step": 1698
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 3.5,
+      "learning_rate": 6.825875278091293e-07,
+      "loss": 0.3246,
+      "step": 1699
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 3.375,
+      "learning_rate": 6.800810367530561e-07,
+      "loss": 0.3155,
+      "step": 1700
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.125,
+      "learning_rate": 6.775784314464717e-07,
+      "loss": 0.4116,
+      "step": 1701
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 3.53125,
+      "learning_rate": 6.750797172327442e-07,
+      "loss": 0.3386,
+      "step": 1702
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 3.53125,
+      "learning_rate": 6.72584899446935e-07,
+      "loss": 0.3805,
+      "step": 1703
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 3.4375,
+      "learning_rate": 6.700939834157846e-07,
+      "loss": 0.3539,
+      "step": 1704
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.46875,
+      "learning_rate": 6.676069744577032e-07,
+      "loss": 0.3597,
+      "step": 1705
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.375,
+      "learning_rate": 6.651238778827581e-07,
+      "loss": 0.3478,
+      "step": 1706
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.265625,
+      "learning_rate": 6.626446989926652e-07,
+      "loss": 0.3144,
+      "step": 1707
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.453125,
+      "learning_rate": 6.601694430807748e-07,
+      "loss": 0.3963,
+      "step": 1708
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.625,
+      "learning_rate": 6.576981154320605e-07,
+      "loss": 0.3848,
+      "step": 1709
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.578125,
+      "learning_rate": 6.552307213231082e-07,
+      "loss": 0.4138,
+      "step": 1710
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.40625,
+      "learning_rate": 6.527672660221071e-07,
+      "loss": 0.3593,
+      "step": 1711
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.53125,
+      "learning_rate": 6.503077547888353e-07,
+      "loss": 0.3946,
+      "step": 1712
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.6875,
+      "learning_rate": 6.478521928746495e-07,
+      "loss": 0.3728,
+      "step": 1713
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.359375,
+      "learning_rate": 6.454005855224741e-07,
+      "loss": 0.3569,
+      "step": 1714
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.265625,
+      "learning_rate": 6.429529379667917e-07,
+      "loss": 0.3561,
+      "step": 1715
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.515625,
+      "learning_rate": 6.405092554336281e-07,
+      "loss": 0.4464,
+      "step": 1716
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.6875,
+      "learning_rate": 6.380695431405453e-07,
+      "loss": 0.3986,
+      "step": 1717
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.890625,
+      "learning_rate": 6.356338062966271e-07,
+      "loss": 0.3926,
+      "step": 1718
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.40625,
+      "learning_rate": 6.332020501024683e-07,
+      "loss": 0.3865,
+      "step": 1719
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 3.625,
+      "learning_rate": 6.30774279750167e-07,
+      "loss": 0.3636,
+      "step": 1720
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 3.5625,
+      "learning_rate": 6.283505004233095e-07,
+      "loss": 0.3675,
+      "step": 1721
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 3.6875,
+      "learning_rate": 6.259307172969606e-07,
+      "loss": 0.4441,
+      "step": 1722
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 3.546875,
+      "learning_rate": 6.235149355376524e-07,
+      "loss": 0.3778,
+      "step": 1723
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 3.21875,
+      "learning_rate": 6.211031603033751e-07,
+      "loss": 0.3722,
+      "step": 1724
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 3.65625,
+      "learning_rate": 6.186953967435646e-07,
+      "loss": 0.3796,
+      "step": 1725
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 3.328125,
+      "learning_rate": 6.162916499990898e-07,
+      "loss": 0.3528,
+      "step": 1726
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 3.40625,
+      "learning_rate": 6.138919252022435e-07,
+      "loss": 0.4287,
+      "step": 1727
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 3.53125,
+      "learning_rate": 6.114962274767311e-07,
+      "loss": 0.418,
+      "step": 1728
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 3.59375,
+      "learning_rate": 6.091045619376615e-07,
+      "loss": 0.4183,
+      "step": 1729
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 3.421875,
+      "learning_rate": 6.067169336915322e-07,
+      "loss": 0.3601,
+      "step": 1730
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 3.4375,
+      "learning_rate": 6.043333478362212e-07,
+      "loss": 0.3863,
+      "step": 1731
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 3.484375,
+      "learning_rate": 6.019538094609758e-07,
+      "loss": 0.464,
+      "step": 1732
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 3.296875,
+      "learning_rate": 5.995783236464014e-07,
+      "loss": 0.377,
+      "step": 1733
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 3.34375,
+      "learning_rate": 5.972068954644514e-07,
+      "loss": 0.3519,
+      "step": 1734
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.3125,
+      "learning_rate": 5.948395299784146e-07,
+      "loss": 0.353,
+      "step": 1735
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.40625,
+      "learning_rate": 5.924762322429045e-07,
+      "loss": 0.3422,
+      "step": 1736
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.53125,
+      "learning_rate": 5.901170073038523e-07,
+      "loss": 0.3746,
+      "step": 1737
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.484375,
+      "learning_rate": 5.877618601984913e-07,
+      "loss": 0.3488,
+      "step": 1738
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.4375,
+      "learning_rate": 5.854107959553482e-07,
+      "loss": 0.3551,
+      "step": 1739
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.59375,
+      "learning_rate": 5.830638195942317e-07,
+      "loss": 0.398,
+      "step": 1740
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.40625,
+      "learning_rate": 5.807209361262253e-07,
+      "loss": 0.3533,
+      "step": 1741
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 3.34375,
+      "learning_rate": 5.783821505536696e-07,
+      "loss": 0.3483,
+      "step": 1742
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 3.53125,
+      "learning_rate": 5.760474678701597e-07,
+      "loss": 0.3806,
+      "step": 1743
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 3.390625,
+      "learning_rate": 5.737168930605272e-07,
+      "loss": 0.3021,
+      "step": 1744
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 3.5,
+      "learning_rate": 5.713904311008341e-07,
+      "loss": 0.3766,
+      "step": 1745
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 3.671875,
+      "learning_rate": 5.690680869583623e-07,
+      "loss": 0.4608,
+      "step": 1746
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 3.390625,
+      "learning_rate": 5.667498655916002e-07,
+      "loss": 0.3706,
+      "step": 1747
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 3.6875,
+      "learning_rate": 5.64435771950233e-07,
+      "loss": 0.3892,
+      "step": 1748
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.640625,
+      "learning_rate": 5.621258109751331e-07,
+      "loss": 0.4237,
+      "step": 1749
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.03125,
+      "learning_rate": 5.598199875983518e-07,
+      "loss": 0.4187,
+      "step": 1750
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.75,
+      "learning_rate": 5.575183067431017e-07,
+      "loss": 0.3952,
+      "step": 1751
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.5,
+      "learning_rate": 5.552207733237544e-07,
+      "loss": 0.3712,
+      "step": 1752
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.515625,
+      "learning_rate": 5.52927392245824e-07,
+      "loss": 0.3883,
+      "step": 1753
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.90625,
+      "learning_rate": 5.506381684059586e-07,
+      "loss": 0.4205,
+      "step": 1754
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.109375,
+      "learning_rate": 5.483531066919329e-07,
+      "loss": 0.3172,
+      "step": 1755
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.625,
+      "learning_rate": 5.460722119826317e-07,
+      "loss": 0.329,
+      "step": 1756
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 3.4375,
+      "learning_rate": 5.437954891480443e-07,
+      "loss": 0.3465,
+      "step": 1757
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 3.546875,
+      "learning_rate": 5.415229430492519e-07,
+      "loss": 0.3789,
+      "step": 1758
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 3.46875,
+      "learning_rate": 5.392545785384193e-07,
+      "loss": 0.3249,
+      "step": 1759
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 3.328125,
+      "learning_rate": 5.369904004587809e-07,
+      "loss": 0.3836,
+      "step": 1760
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 3.5,
+      "learning_rate": 5.347304136446349e-07,
+      "loss": 0.4307,
+      "step": 1761
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 3.515625,
+      "learning_rate": 5.324746229213282e-07,
+      "loss": 0.3809,
+      "step": 1762
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 3.609375,
+      "learning_rate": 5.302230331052511e-07,
+      "loss": 0.3522,
+      "step": 1763
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.28125,
+      "learning_rate": 5.279756490038226e-07,
+      "loss": 0.3859,
+      "step": 1764
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.5625,
+      "learning_rate": 5.257324754154821e-07,
+      "loss": 0.3826,
+      "step": 1765
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.53125,
+      "learning_rate": 5.234935171296793e-07,
+      "loss": 0.4064,
+      "step": 1766
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.5,
+      "learning_rate": 5.21258778926865e-07,
+      "loss": 0.3687,
+      "step": 1767
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.28125,
+      "learning_rate": 5.190282655784778e-07,
+      "loss": 0.3591,
+      "step": 1768
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.75,
+      "learning_rate": 5.168019818469358e-07,
+      "loss": 0.3621,
+      "step": 1769
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.40625,
+      "learning_rate": 5.14579932485628e-07,
+      "loss": 0.3828,
+      "step": 1770
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.765625,
+      "learning_rate": 5.123621222389005e-07,
+      "loss": 0.381,
+      "step": 1771
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.421875,
+      "learning_rate": 5.101485558420504e-07,
+      "loss": 0.3742,
+      "step": 1772
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.34375,
+      "learning_rate": 5.079392380213122e-07,
+      "loss": 0.3786,
+      "step": 1773
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.515625,
+      "learning_rate": 5.057341734938493e-07,
+      "loss": 0.4111,
+      "step": 1774
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.875,
+      "learning_rate": 5.035333669677437e-07,
+      "loss": 0.4075,
+      "step": 1775
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.578125,
+      "learning_rate": 5.013368231419871e-07,
+      "loss": 0.3663,
+      "step": 1776
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.99144546706469e-07,
+      "loss": 0.3556,
+      "step": 1777
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.96956542341967e-07,
+      "loss": 0.3828,
+      "step": 1778
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.947728147201389e-07,
+      "loss": 0.3552,
+      "step": 1779
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.71875,
+      "learning_rate": 4.925933685035089e-07,
+      "loss": 0.3682,
+      "step": 1780
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.546875,
+      "learning_rate": 4.904182083454628e-07,
+      "loss": 0.4185,
+      "step": 1781
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.882473388902323e-07,
+      "loss": 0.4185,
+      "step": 1782
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.860807647728888e-07,
+      "loss": 0.352,
+      "step": 1783
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.839184906193342e-07,
+      "loss": 0.3516,
+      "step": 1784
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.671875,
+      "learning_rate": 4.817605210462875e-07,
+      "loss": 0.3806,
+      "step": 1785
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.796068606612775e-07,
+      "loss": 0.4353,
+      "step": 1786
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.625,
+      "learning_rate": 4.774575140626317e-07,
+      "loss": 0.366,
+      "step": 1787
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.921875,
+      "learning_rate": 4.753124858394686e-07,
+      "loss": 0.3662,
+      "step": 1788
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.731717805716865e-07,
+      "loss": 0.4369,
+      "step": 1789
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.7103540282995185e-07,
+      "loss": 0.353,
+      "step": 1790
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.53125,
+      "learning_rate": 4.689033571756929e-07,
+      "loss": 0.4282,
+      "step": 1791
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.667756481610866e-07,
+      "loss": 0.3463,
+      "step": 1792
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.703125,
+      "learning_rate": 4.646522803290535e-07,
+      "loss": 0.4466,
+      "step": 1793
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.90625,
+      "learning_rate": 4.6253325821324246e-07,
+      "loss": 0.394,
+      "step": 1794
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.59375,
+      "learning_rate": 4.604185863380245e-07,
+      "loss": 0.411,
+      "step": 1795
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.53125,
+      "learning_rate": 4.583082692184837e-07,
+      "loss": 0.3977,
+      "step": 1796
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.5620231136040414e-07,
+      "loss": 0.3627,
+      "step": 1797
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.5410071726026403e-07,
+      "loss": 0.3948,
+      "step": 1798
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.421875,
+      "learning_rate": 4.5200349140522377e-07,
+      "loss": 0.3745,
+      "step": 1799
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.4991063827311637e-07,
+      "loss": 0.3972,
+      "step": 1800
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.4782216233243857e-07,
+      "loss": 0.4185,
+      "step": 1801
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.515625,
+      "learning_rate": 4.4573806804234344e-07,
+      "loss": 0.4294,
+      "step": 1802
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.375,
+      "learning_rate": 4.4365835985262646e-07,
+      "loss": 0.3446,
+      "step": 1803
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.4158304220371787e-07,
+      "loss": 0.3719,
+      "step": 1804
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.3951211952667604e-07,
+      "loss": 0.3953,
+      "step": 1805
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.374455962431731e-07,
+      "loss": 0.3579,
+      "step": 1806
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.515625,
+      "learning_rate": 4.3538347676548965e-07,
+      "loss": 0.384,
+      "step": 1807
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.3332576549650257e-07,
+      "loss": 0.392,
+      "step": 1808
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.312724668296761e-07,
+      "loss": 0.4162,
+      "step": 1809
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.75,
+      "learning_rate": 4.2922358514905557e-07,
+      "loss": 0.4487,
+      "step": 1810
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.2717912482925313e-07,
+      "loss": 0.3357,
+      "step": 1811
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.5,
+      "learning_rate": 4.251390902354413e-07,
+      "loss": 0.3819,
+      "step": 1812
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.231034857233429e-07,
+      "loss": 0.4124,
+      "step": 1813
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.53125,
+      "learning_rate": 4.2107231563922324e-07,
+      "loss": 0.4346,
+      "step": 1814
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.1904558431987903e-07,
+      "loss": 0.3775,
+      "step": 1815
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.1702329609262877e-07,
+      "loss": 0.3347,
+      "step": 1816
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.150054552753055e-07,
+      "loss": 0.35,
+      "step": 1817
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.1299206617624513e-07,
+      "loss": 0.4274,
+      "step": 1818
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.1098313309428094e-07,
+      "loss": 0.345,
+      "step": 1819
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.089786603187298e-07,
+      "loss": 0.3508,
+      "step": 1820
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.578125,
+      "learning_rate": 4.0697865212938644e-07,
+      "loss": 0.424,
+      "step": 1821
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.375,
+      "learning_rate": 4.04983112796512e-07,
+      "loss": 0.4082,
+      "step": 1822
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.375,
+      "learning_rate": 4.029920465808279e-07,
+      "loss": 0.3361,
+      "step": 1823
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.010054577335043e-07,
+      "loss": 0.3786,
+      "step": 1824
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.578125,
+      "learning_rate": 3.9902335049615046e-07,
+      "loss": 0.3668,
+      "step": 1825
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.296875,
+      "learning_rate": 3.9704572910080787e-07,
+      "loss": 0.3015,
+      "step": 1826
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.765625,
+      "learning_rate": 3.950725977699396e-07,
+      "loss": 0.4104,
+      "step": 1827
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.25,
+      "learning_rate": 3.9310396071642334e-07,
+      "loss": 0.3607,
+      "step": 1828
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.390625,
+      "learning_rate": 3.911398221435392e-07,
+      "loss": 0.3589,
+      "step": 1829
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.53125,
+      "learning_rate": 3.891801862449629e-07,
+      "loss": 0.414,
+      "step": 1830
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.546875,
+      "learning_rate": 3.8722505720475674e-07,
+      "loss": 0.4149,
+      "step": 1831
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.296875,
+      "learning_rate": 3.8527443919736007e-07,
+      "loss": 0.3743,
+      "step": 1832
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.5,
+      "learning_rate": 3.8332833638758224e-07,
+      "loss": 0.3905,
+      "step": 1833
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.421875,
+      "learning_rate": 3.8138675293058954e-07,
+      "loss": 0.3458,
+      "step": 1834
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.609375,
+      "learning_rate": 3.794496929718994e-07,
+      "loss": 0.4465,
+      "step": 1835
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.53125,
+      "learning_rate": 3.775171606473729e-07,
+      "loss": 0.3758,
+      "step": 1836
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.59375,
+      "learning_rate": 3.7558916008320263e-07,
+      "loss": 0.3312,
+      "step": 1837
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.390625,
+      "learning_rate": 3.7366569539590504e-07,
+      "loss": 0.3866,
+      "step": 1838
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.5,
+      "learning_rate": 3.7174677069231223e-07,
+      "loss": 0.4056,
+      "step": 1839
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.46875,
+      "learning_rate": 3.69832390069563e-07,
+      "loss": 0.3766,
+      "step": 1840
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.25,
+      "learning_rate": 3.679225576150955e-07,
+      "loss": 0.3603,
+      "step": 1841
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.15625,
+      "learning_rate": 3.6601727740663396e-07,
+      "loss": 0.3422,
+      "step": 1842
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.25,
+      "learning_rate": 3.6411655351218515e-07,
+      "loss": 0.3195,
+      "step": 1843
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.3125,
+      "learning_rate": 3.6222038999002646e-07,
+      "loss": 0.3132,
+      "step": 1844
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.703125,
+      "learning_rate": 3.603287908886996e-07,
+      "loss": 0.3842,
+      "step": 1845
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.71875,
+      "learning_rate": 3.5844176024699973e-07,
+      "loss": 0.4294,
+      "step": 1846
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.4375,
+      "learning_rate": 3.5655930209396784e-07,
+      "loss": 0.3536,
+      "step": 1847
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.4375,
+      "learning_rate": 3.546814204488813e-07,
+      "loss": 0.335,
+      "step": 1848
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.53125,
+      "learning_rate": 3.528081193212482e-07,
+      "loss": 0.4257,
+      "step": 1849
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.609375,
+      "learning_rate": 3.509394027107957e-07,
+      "loss": 0.3792,
+      "step": 1850
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.953125,
+      "learning_rate": 3.490752746074616e-07,
+      "loss": 0.4395,
+      "step": 1851
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.4375,
+      "learning_rate": 3.4721573899138743e-07,
+      "loss": 0.3489,
+      "step": 1852
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.5625,
+      "learning_rate": 3.453607998329087e-07,
+      "loss": 0.3565,
+      "step": 1853
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.71875,
+      "learning_rate": 3.43510461092548e-07,
+      "loss": 0.4201,
+      "step": 1854
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.484375,
+      "learning_rate": 3.4166472672100466e-07,
+      "loss": 0.3439,
+      "step": 1855
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.515625,
+      "learning_rate": 3.3982360065914655e-07,
+      "loss": 0.3177,
+      "step": 1856
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.75,
+      "learning_rate": 3.379870868380031e-07,
+      "loss": 0.4215,
+      "step": 1857
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.703125,
+      "learning_rate": 3.361551891787562e-07,
+      "loss": 0.4171,
+      "step": 1858
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.75,
+      "learning_rate": 3.343279115927317e-07,
+      "loss": 0.3858,
+      "step": 1859
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.28125,
+      "learning_rate": 3.3250525798139006e-07,
+      "loss": 0.3718,
+      "step": 1860
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.875,
+      "learning_rate": 3.3068723223631914e-07,
+      "loss": 0.378,
+      "step": 1861
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.625,
+      "learning_rate": 3.288738382392273e-07,
+      "loss": 0.3899,
+      "step": 1862
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.046875,
+      "learning_rate": 3.270650798619318e-07,
+      "loss": 0.3352,
+      "step": 1863
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.40625,
+      "learning_rate": 3.252609609663529e-07,
+      "loss": 0.3882,
+      "step": 1864
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.40625,
+      "learning_rate": 3.2346148540450455e-07,
+      "loss": 0.3839,
+      "step": 1865
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.359375,
+      "learning_rate": 3.2166665701848844e-07,
+      "loss": 0.3968,
+      "step": 1866
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.40625,
+      "learning_rate": 3.1987647964048075e-07,
+      "loss": 0.3748,
+      "step": 1867
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 3.515625,
+      "learning_rate": 3.18090957092731e-07,
+      "loss": 0.3694,
+      "step": 1868
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 3.71875,
+      "learning_rate": 3.163100931875471e-07,
+      "loss": 0.4059,
+      "step": 1869
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 3.5625,
+      "learning_rate": 3.145338917272905e-07,
+      "loss": 0.4377,
+      "step": 1870
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 3.53125,
+      "learning_rate": 3.1276235650437024e-07,
+      "loss": 0.3519,
+      "step": 1871
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 3.359375,
+      "learning_rate": 3.1099549130122945e-07,
+      "loss": 0.3503,
+      "step": 1872
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 3.546875,
+      "learning_rate": 3.092332998903416e-07,
+      "loss": 0.3494,
+      "step": 1873
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 3.359375,
+      "learning_rate": 3.074757860342001e-07,
+      "loss": 0.3496,
+      "step": 1874
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3.5,
+      "learning_rate": 3.057229534853126e-07,
+      "loss": 0.3901,
+      "step": 1875
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3.265625,
+      "learning_rate": 3.0397480598619136e-07,
+      "loss": 0.3546,
+      "step": 1876
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3.75,
+      "learning_rate": 3.022313472693447e-07,
+      "loss": 0.376,
+      "step": 1877
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3.765625,
+      "learning_rate": 3.004925810572698e-07,
+      "loss": 0.4154,
+      "step": 1878
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.03125,
+      "learning_rate": 2.987585110624447e-07,
+      "loss": 0.5071,
+      "step": 1879
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3.875,
+      "learning_rate": 2.970291409873219e-07,
+      "loss": 0.4191,
+      "step": 1880
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3.78125,
+      "learning_rate": 2.953044745243175e-07,
+      "loss": 0.4351,
+      "step": 1881
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.9358451535580533e-07,
+      "loss": 0.3596,
+      "step": 1882
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.03125,
+      "learning_rate": 2.918692671541076e-07,
+      "loss": 0.3997,
+      "step": 1883
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.640625,
+      "learning_rate": 2.901587335814898e-07,
+      "loss": 0.378,
+      "step": 1884
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.884529182901508e-07,
+      "loss": 0.3634,
+      "step": 1885
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.6875,
+      "learning_rate": 2.867518249222137e-07,
+      "loss": 0.3674,
+      "step": 1886
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.850554571097211e-07,
+      "loss": 0.3961,
+      "step": 1887
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.578125,
+      "learning_rate": 2.8336381847462623e-07,
+      "loss": 0.3661,
+      "step": 1888
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.796875,
+      "learning_rate": 2.8167691262878383e-07,
+      "loss": 0.4172,
+      "step": 1889
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.875,
+      "learning_rate": 2.799947431739444e-07,
+      "loss": 0.4099,
+      "step": 1890
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.25,
+      "learning_rate": 2.7831731370174473e-07,
+      "loss": 0.3421,
+      "step": 1891
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.7664462779370293e-07,
+      "loss": 0.3392,
+      "step": 1892
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.84375,
+      "learning_rate": 2.749766890212069e-07,
+      "loss": 0.382,
+      "step": 1893
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.640625,
+      "learning_rate": 2.733135009455107e-07,
+      "loss": 0.4227,
+      "step": 1894
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.716550671177237e-07,
+      "loss": 0.3481,
+      "step": 1895
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.7000139107880437e-07,
+      "loss": 0.3627,
+      "step": 1896
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.703125,
+      "learning_rate": 2.6835247635955466e-07,
+      "loss": 0.373,
+      "step": 1897
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.667083264806081e-07,
+      "loss": 0.448,
+      "step": 1898
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.650689449524257e-07,
+      "loss": 0.3917,
+      "step": 1899
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.6343433527528716e-07,
+      "loss": 0.3204,
+      "step": 1900
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.6180450093928526e-07,
+      "loss": 0.3581,
+      "step": 1901
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.6017944542431394e-07,
+      "loss": 0.3937,
+      "step": 1902
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.585591722000669e-07,
+      "loss": 0.3491,
+      "step": 1903
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.569436847260248e-07,
+      "loss": 0.402,
+      "step": 1904
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.553329864514503e-07,
+      "loss": 0.3548,
+      "step": 1905
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.5372708081538275e-07,
+      "loss": 0.4006,
+      "step": 1906
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.521259712466256e-07,
+      "loss": 0.318,
+      "step": 1907
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.505296611637445e-07,
+      "loss": 0.3376,
+      "step": 1908
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.4893815397505544e-07,
+      "loss": 0.3698,
+      "step": 1909
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.640625,
+      "learning_rate": 2.4735145307862195e-07,
+      "loss": 0.3483,
+      "step": 1910
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.457695618622438e-07,
+      "loss": 0.4065,
+      "step": 1911
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.796875,
+      "learning_rate": 2.441924837034529e-07,
+      "loss": 0.4062,
+      "step": 1912
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.734375,
+      "learning_rate": 2.426202219695026e-07,
+      "loss": 0.4107,
+      "step": 1913
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.78125,
+      "learning_rate": 2.4105278001736564e-07,
+      "loss": 0.408,
+      "step": 1914
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.3949016119372113e-07,
+      "loss": 0.4065,
+      "step": 1915
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.5,
+      "learning_rate": 2.3793236883495164e-07,
+      "loss": 0.3934,
+      "step": 1916
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.3637940626713346e-07,
+      "loss": 0.3765,
+      "step": 1917
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.25,
+      "learning_rate": 2.3483127680603292e-07,
+      "loss": 0.3343,
+      "step": 1918
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.3328798375709505e-07,
+      "loss": 0.3357,
+      "step": 1919
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.3174953041543852e-07,
+      "loss": 0.4,
+      "step": 1920
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.625,
+      "learning_rate": 2.3021592006585046e-07,
+      "loss": 0.4734,
+      "step": 1921
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.703125,
+      "learning_rate": 2.286871559827758e-07,
+      "loss": 0.4624,
+      "step": 1922
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.921875,
+      "learning_rate": 2.271632414303132e-07,
+      "loss": 0.4075,
+      "step": 1923
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.640625,
+      "learning_rate": 2.256441796622061e-07,
+      "loss": 0.4176,
+      "step": 1924
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.241299739218375e-07,
+      "loss": 0.4798,
+      "step": 1925
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.984375,
+      "learning_rate": 2.2262062744222101e-07,
+      "loss": 0.4338,
+      "step": 1926
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.2111614344599686e-07,
+      "loss": 0.3128,
+      "step": 1927
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.1961652514542187e-07,
+      "loss": 0.3868,
+      "step": 1928
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.1812177574236398e-07,
+      "loss": 0.3874,
+      "step": 1929
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.1663189842829697e-07,
+      "loss": 0.3544,
+      "step": 1930
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.859375,
+      "learning_rate": 2.151468963842901e-07,
+      "loss": 0.4692,
+      "step": 1931
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.136667727810049e-07,
+      "loss": 0.3708,
+      "step": 1932
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.1219153077868622e-07,
+      "loss": 0.3404,
+      "step": 1933
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.1072117352715582e-07,
+      "loss": 0.3543,
+      "step": 1934
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.578125,
+      "learning_rate": 2.0925570416580582e-07,
+      "loss": 0.3258,
+      "step": 1935
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.0779512582359358e-07,
+      "loss": 0.3637,
+      "step": 1936
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.0633944161903147e-07,
+      "loss": 0.3811,
+      "step": 1937
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.71875,
+      "learning_rate": 2.0488865466018303e-07,
+      "loss": 0.4417,
+      "step": 1938
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.0344276804465657e-07,
+      "loss": 0.366,
+      "step": 1939
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.625,
+      "learning_rate": 2.0200178485959677e-07,
+      "loss": 0.3977,
+      "step": 1940
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.0056570818167864e-07,
+      "loss": 0.3983,
+      "step": 1941
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.859375,
+      "learning_rate": 1.9913454107710172e-07,
+      "loss": 0.3882,
+      "step": 1942
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.977082866015817e-07,
+      "loss": 0.4274,
+      "step": 1943
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.9628694780034791e-07,
+      "loss": 0.474,
+      "step": 1944
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.9487052770813114e-07,
+      "loss": 0.3395,
+      "step": 1945
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.9345902934916227e-07,
+      "loss": 0.3883,
+      "step": 1946
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.9205245573716196e-07,
+      "loss": 0.351,
+      "step": 1947
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.9065080987533764e-07,
+      "loss": 0.3829,
+      "step": 1948
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.8925409475637451e-07,
+      "loss": 0.3897,
+      "step": 1949
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.8786231336242993e-07,
+      "loss": 0.3384,
+      "step": 1950
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.864754686651274e-07,
+      "loss": 0.348,
+      "step": 1951
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.8509356362554964e-07,
+      "loss": 0.3679,
+      "step": 1952
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.837166011942332e-07,
+      "loss": 0.3236,
+      "step": 1953
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.5,
+      "learning_rate": 1.8234458431116124e-07,
+      "loss": 0.3896,
+      "step": 1954
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.8097751590575725e-07,
+      "loss": 0.3593,
+      "step": 1955
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.7961539889687962e-07,
+      "loss": 0.3802,
+      "step": 1956
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.7825823619281452e-07,
+      "loss": 0.3603,
+      "step": 1957
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.7690603069127067e-07,
+      "loss": 0.3927,
+      "step": 1958
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.7555878527937164e-07,
+      "loss": 0.3933,
+      "step": 1959
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.742165028336515e-07,
+      "loss": 0.3783,
+      "step": 1960
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.7287918622004683e-07,
+      "loss": 0.4024,
+      "step": 1961
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.7154683829389284e-07,
+      "loss": 0.4232,
+      "step": 1962
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.03125,
+      "learning_rate": 1.7021946189991478e-07,
+      "loss": 0.4014,
+      "step": 1963
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.688970598722231e-07,
+      "loss": 0.3474,
+      "step": 1964
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.675796350343084e-07,
+      "loss": 0.3719,
+      "step": 1965
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.90625,
+      "learning_rate": 1.6626719019903426e-07,
+      "loss": 0.3856,
+      "step": 1966
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.649597281686302e-07,
+      "loss": 0.3874,
+      "step": 1967
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.5,
+      "learning_rate": 1.6365725173468737e-07,
+      "loss": 0.3952,
+      "step": 1968
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.6235976367815236e-07,
+      "loss": 0.3277,
+      "step": 1969
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.610672667693211e-07,
+      "loss": 0.3421,
+      "step": 1970
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.75,
+      "learning_rate": 1.5977976376783222e-07,
+      "loss": 0.4059,
+      "step": 1971
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.5849725742266232e-07,
+      "loss": 0.3541,
+      "step": 1972
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.5721975047211845e-07,
+      "loss": 0.3714,
+      "step": 1973
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.890625,
+      "learning_rate": 1.5594724564383462e-07,
+      "loss": 0.3618,
+      "step": 1974
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.5467974565476495e-07,
+      "loss": 0.3378,
+      "step": 1975
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.5341725321117606e-07,
+      "loss": 0.3517,
+      "step": 1976
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.5215977100864394e-07,
+      "loss": 0.397,
+      "step": 1977
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.5090730173204616e-07,
+      "loss": 0.4013,
+      "step": 1978
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.4965984805555889e-07,
+      "loss": 0.3443,
+      "step": 1979
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.484174126426477e-07,
+      "loss": 0.3338,
+      "step": 1980
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.5,
+      "learning_rate": 1.4717999814606394e-07,
+      "loss": 0.4067,
+      "step": 1981
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.4594760720783863e-07,
+      "loss": 0.3855,
+      "step": 1982
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.447202424592775e-07,
+      "loss": 0.3706,
+      "step": 1983
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.4349790652095492e-07,
+      "loss": 0.3642,
+      "step": 1984
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.422806020027065e-07,
+      "loss": 0.3676,
+      "step": 1985
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.410683315036268e-07,
+      "loss": 0.3816,
+      "step": 1986
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.890625,
+      "learning_rate": 1.3986109761206097e-07,
+      "loss": 0.413,
+      "step": 1987
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.386589029056018e-07,
+      "loss": 0.3598,
+      "step": 1988
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.3746174995108163e-07,
+      "loss": 0.3709,
+      "step": 1989
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.375,
+      "learning_rate": 1.3626964130456804e-07,
+      "loss": 0.399,
+      "step": 1990
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.350825795113589e-07,
+      "loss": 0.3651,
+      "step": 1991
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.339005671059765e-07,
+      "loss": 0.3273,
+      "step": 1992
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.3272360661216206e-07,
+      "loss": 0.3709,
+      "step": 1993
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.315517005428696e-07,
+      "loss": 0.3781,
+      "step": 1994
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.3038485140026148e-07,
+      "loss": 0.4158,
+      "step": 1995
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.2922306167570452e-07,
+      "loss": 0.3223,
+      "step": 1996
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.2806633384976092e-07,
+      "loss": 0.3973,
+      "step": 1997
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.109375,
+      "learning_rate": 1.269146703921864e-07,
+      "loss": 0.3263,
+      "step": 1998
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.2576807376192267e-07,
+      "loss": 0.3934,
+      "step": 1999
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.2462654640709392e-07,
+      "loss": 0.478,
+      "step": 2000
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.5,
+      "learning_rate": 1.2349009076500134e-07,
+      "loss": 0.3904,
+      "step": 2001
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.223587092621162e-07,
+      "loss": 0.3712,
+      "step": 2002
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.2123240431407613e-07,
+      "loss": 0.399,
+      "step": 2003
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.2011117832567977e-07,
+      "loss": 0.3523,
+      "step": 2004
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.1899503369088211e-07,
+      "loss": 0.3919,
+      "step": 2005
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.1788397279278858e-07,
+      "loss": 0.3673,
+      "step": 2006
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.1677799800364958e-07,
+      "loss": 0.4644,
+      "step": 2007
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.1567711168485596e-07,
+      "loss": 0.374,
+      "step": 2008
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.921875,
+      "learning_rate": 1.1458131618693552e-07,
+      "loss": 0.3954,
+      "step": 2009
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.134906138495459e-07,
+      "loss": 0.3519,
+      "step": 2010
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.375,
+      "learning_rate": 1.1240500700146928e-07,
+      "loss": 0.3533,
+      "step": 2011
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.5,
+      "learning_rate": 1.1132449796060874e-07,
+      "loss": 0.3844,
+      "step": 2012
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.1024908903398385e-07,
+      "loss": 0.3742,
+      "step": 2013
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.0917878251772373e-07,
+      "loss": 0.419,
+      "step": 2014
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.0811358069706401e-07,
+      "loss": 0.3731,
+      "step": 2015
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.070534858463404e-07,
+      "loss": 0.3519,
+      "step": 2016
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.0599850022898539e-07,
+      "loss": 0.3617,
+      "step": 2017
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.049486260975216e-07,
+      "loss": 0.4191,
+      "step": 2018
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.0390386569355954e-07,
+      "loss": 0.3616,
+      "step": 2019
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.0286422124779011e-07,
+      "loss": 0.3708,
+      "step": 2020
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.0182969497998074e-07,
+      "loss": 0.4017,
+      "step": 2021
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.15625,
+      "learning_rate": 1.0080028909897232e-07,
+      "loss": 0.3307,
+      "step": 2022
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.828125,
+      "learning_rate": 9.977600580267171e-08,
+      "loss": 0.4018,
+      "step": 2023
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.875684727804924e-08,
+      "loss": 0.3768,
+      "step": 2024
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.453125,
+      "learning_rate": 9.774281570113259e-08,
+      "loss": 0.4237,
+      "step": 2025
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.328125,
+      "learning_rate": 9.673391323700348e-08,
+      "loss": 0.3051,
+      "step": 2026
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.573014203979241e-08,
+      "loss": 0.3399,
+      "step": 2027
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.3125,
+      "learning_rate": 9.47315042526728e-08,
+      "loss": 0.3895,
+      "step": 2028
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.28125,
+      "learning_rate": 9.373800200785905e-08,
+      "loss": 0.3583,
+      "step": 2029
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 3.484375,
+      "learning_rate": 9.274963742659932e-08,
+      "loss": 0.4077,
+      "step": 2030
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.176641261917336e-08,
+      "loss": 0.3695,
+      "step": 2031
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 3.46875,
+      "learning_rate": 9.078832968488632e-08,
+      "loss": 0.3638,
+      "step": 2032
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 3.734375,
+      "learning_rate": 8.981539071206441e-08,
+      "loss": 0.362,
+      "step": 2033
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 3.609375,
+      "learning_rate": 8.884759777805147e-08,
+      "loss": 0.401,
+      "step": 2034
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 3.359375,
+      "learning_rate": 8.788495294920374e-08,
+      "loss": 0.3644,
+      "step": 2035
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 3.5625,
+      "learning_rate": 8.6927458280886e-08,
+      "loss": 0.3954,
+      "step": 2036
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 3.5,
+      "learning_rate": 8.597511581746626e-08,
+      "loss": 0.363,
+      "step": 2037
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 3.578125,
+      "learning_rate": 8.50279275923116e-08,
+      "loss": 0.3688,
+      "step": 2038
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 3.421875,
+      "learning_rate": 8.408589562778513e-08,
+      "loss": 0.3648,
+      "step": 2039
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 3.6875,
+      "learning_rate": 8.314902193524044e-08,
+      "loss": 0.3962,
+      "step": 2040
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 3.453125,
+      "learning_rate": 8.221730851501713e-08,
+      "loss": 0.3571,
+      "step": 2041
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 3.890625,
+      "learning_rate": 8.129075735643698e-08,
+      "loss": 0.4143,
+      "step": 2042
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 3.5625,
+      "learning_rate": 8.036937043780057e-08,
+      "loss": 0.3664,
+      "step": 2043
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 3.734375,
+      "learning_rate": 7.94531497263809e-08,
+      "loss": 0.3808,
+      "step": 2044
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 3.484375,
+      "learning_rate": 7.854209717842231e-08,
+      "loss": 0.3637,
+      "step": 2045
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 3.265625,
+      "learning_rate": 7.763621473913241e-08,
+      "loss": 0.3345,
+      "step": 2046
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 3.59375,
+      "learning_rate": 7.673550434268123e-08,
+      "loss": 0.4071,
+      "step": 2047
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 3.859375,
+      "learning_rate": 7.583996791219572e-08,
+      "loss": 0.4611,
+      "step": 2048
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 3.390625,
+      "learning_rate": 7.494960735975581e-08,
+      "loss": 0.3463,
+      "step": 2049
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 3.34375,
+      "learning_rate": 7.406442458638974e-08,
+      "loss": 0.4814,
+      "step": 2050
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 3.59375,
+      "learning_rate": 7.318442148207066e-08,
+      "loss": 0.3924,
+      "step": 2051
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 3.546875,
+      "learning_rate": 7.230959992571368e-08,
+      "loss": 0.3939,
+      "step": 2052
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 3.4375,
+      "learning_rate": 7.14399617851691e-08,
+      "loss": 0.3103,
+      "step": 2053
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 3.46875,
+      "learning_rate": 7.057550891722109e-08,
+      "loss": 0.3991,
+      "step": 2054
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 3.5625,
+      "learning_rate": 6.971624316758213e-08,
+      "loss": 0.3205,
+      "step": 2055
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 3.5,
+      "learning_rate": 6.88621663708891e-08,
+      "loss": 0.3447,
+      "step": 2056
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 3.3125,
+      "learning_rate": 6.801328035070138e-08,
+      "loss": 0.3494,
+      "step": 2057
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 3.59375,
+      "learning_rate": 6.716958691949355e-08,
+      "loss": 0.4167,
+      "step": 2058
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 3.765625,
+      "learning_rate": 6.633108787865467e-08,
+      "loss": 0.4169,
+      "step": 2059
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 3.53125,
+      "learning_rate": 6.549778501848208e-08,
+      "loss": 0.3947,
+      "step": 2060
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 3.75,
+      "learning_rate": 6.466968011818037e-08,
+      "loss": 0.3295,
+      "step": 2061
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 3.28125,
+      "learning_rate": 6.38467749458535e-08,
+      "loss": 0.3355,
+      "step": 2062
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 3.5,
+      "learning_rate": 6.302907125850577e-08,
+      "loss": 0.3587,
+      "step": 2063
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 3.3125,
+      "learning_rate": 6.221657080203364e-08,
+      "loss": 0.3247,
+      "step": 2064
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 3.328125,
+      "learning_rate": 6.140927531122581e-08,
+      "loss": 0.3498,
+      "step": 2065
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 3.328125,
+      "learning_rate": 6.0607186509756e-08,
+      "loss": 0.311,
+      "step": 2066
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 3.21875,
+      "learning_rate": 5.981030611018235e-08,
+      "loss": 0.3356,
+      "step": 2067
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 3.671875,
+      "learning_rate": 5.901863581394135e-08,
+      "loss": 0.3714,
+      "step": 2068
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 3.375,
+      "learning_rate": 5.823217731134673e-08,
+      "loss": 0.3747,
+      "step": 2069
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 3.765625,
+      "learning_rate": 5.745093228158305e-08,
+      "loss": 0.3917,
+      "step": 2070
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 3.703125,
+      "learning_rate": 5.667490239270379e-08,
+      "loss": 0.4031,
+      "step": 2071
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 3.421875,
+      "learning_rate": 5.5904089301627997e-08,
+      "loss": 0.3354,
+      "step": 2072
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 3.546875,
+      "learning_rate": 5.5138494654135844e-08,
+      "loss": 0.392,
+      "step": 2073
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 3.609375,
+      "learning_rate": 5.4378120084865314e-08,
+      "loss": 0.3501,
+      "step": 2074
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 3.3125,
+      "learning_rate": 5.3622967217309415e-08,
+      "loss": 0.3372,
+      "step": 2075
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 3.234375,
+      "learning_rate": 5.2873037663811464e-08,
+      "loss": 0.3538,
+      "step": 2076
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 3.171875,
+      "learning_rate": 5.212833302556258e-08,
+      "loss": 0.4672,
+      "step": 2077
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 3.5625,
+      "learning_rate": 5.138885489259893e-08,
+      "loss": 0.3841,
+      "step": 2078
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 3.296875,
+      "learning_rate": 5.065460484379614e-08,
+      "loss": 0.3374,
+      "step": 2079
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.992558444686768e-08,
+      "loss": 0.3969,
+      "step": 2080
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 3.515625,
+      "learning_rate": 4.920179525836205e-08,
+      "loss": 0.3712,
+      "step": 2081
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.8483238823656685e-08,
+      "loss": 0.4375,
+      "step": 2082
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 3.515625,
+      "learning_rate": 4.776991667695824e-08,
+      "loss": 0.36,
+      "step": 2083
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 3.53125,
+      "learning_rate": 4.706183034129647e-08,
+      "loss": 0.358,
+      "step": 2084
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.6358981328522015e-08,
+      "loss": 0.3694,
+      "step": 2085
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 3.53125,
+      "learning_rate": 4.566137113930363e-08,
+      "loss": 0.426,
+      "step": 2086
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 3.75,
+      "learning_rate": 4.4969001263124314e-08,
+      "loss": 0.3986,
+      "step": 2087
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 3.625,
+      "learning_rate": 4.428187317827848e-08,
+      "loss": 0.3785,
+      "step": 2088
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.359998835186841e-08,
+      "loss": 0.419,
+      "step": 2089
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 3.578125,
+      "learning_rate": 4.292334823980143e-08,
+      "loss": 0.4271,
+      "step": 2090
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.2251954286787746e-08,
+      "loss": 0.4488,
+      "step": 2091
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.158580792633482e-08,
+      "loss": 0.3493,
+      "step": 2092
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.09249105807466e-08,
+      "loss": 0.3392,
+      "step": 2093
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.0269263661119885e-08,
+      "loss": 0.3659,
+      "step": 2094
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 3.296875,
+      "learning_rate": 3.9618868567340984e-08,
+      "loss": 0.3688,
+      "step": 2095
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 3.296875,
+      "learning_rate": 3.897372668808297e-08,
+      "loss": 0.3522,
+      "step": 2096
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 3.578125,
+      "learning_rate": 3.833383940080232e-08,
+      "loss": 0.3763,
+      "step": 2097
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 3.859375,
+      "learning_rate": 3.769920807173699e-08,
+      "loss": 0.4082,
+      "step": 2098
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 3.4375,
+      "learning_rate": 3.7069834055902256e-08,
+      "loss": 0.3676,
+      "step": 2099
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 3.578125,
+      "learning_rate": 3.6445718697089015e-08,
+      "loss": 0.4571,
+      "step": 2100
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 3.203125,
+      "learning_rate": 3.5826863327859395e-08,
+      "loss": 0.3437,
+      "step": 2101
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 3.4375,
+      "learning_rate": 3.521326926954533e-08,
+      "loss": 0.3477,
+      "step": 2102
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 3.484375,
+      "learning_rate": 3.4604937832244944e-08,
+      "loss": 0.367,
+      "step": 2103
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 3.3125,
+      "learning_rate": 3.400187031482094e-08,
+      "loss": 0.3726,
+      "step": 2104
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 3.28125,
+      "learning_rate": 3.340406800489582e-08,
+      "loss": 0.3476,
+      "step": 2105
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 3.53125,
+      "learning_rate": 3.2811532178850245e-08,
+      "loss": 0.3604,
+      "step": 2106
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 3.375,
+      "learning_rate": 3.222426410182111e-08,
+      "loss": 0.3432,
+      "step": 2107
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 3.203125,
+      "learning_rate": 3.164226502769707e-08,
+      "loss": 0.3728,
+      "step": 2108
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 3.578125,
+      "learning_rate": 3.106553619911801e-08,
+      "loss": 0.4147,
+      "step": 2109
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 3.375,
+      "learning_rate": 3.04940788474703e-08,
+      "loss": 0.3625,
+      "step": 2110
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.992789419288489e-08,
+      "loss": 0.3062,
+      "step": 2111
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.9366983444235054e-08,
+      "loss": 0.4303,
+      "step": 2112
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.881134779913447e-08,
+      "loss": 0.3692,
+      "step": 2113
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.8260988443933045e-08,
+      "loss": 0.3942,
+      "step": 2114
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 3.90625,
+      "learning_rate": 2.77159065537147e-08,
+      "loss": 0.4246,
+      "step": 2115
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 3.765625,
+      "learning_rate": 2.717610329229653e-08,
+      "loss": 0.3899,
+      "step": 2116
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 3.71875,
+      "learning_rate": 2.6641579812224373e-08,
+      "loss": 0.4065,
+      "step": 2117
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.6112337254771414e-08,
+      "loss": 0.3826,
+      "step": 2118
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.5588376749934584e-08,
+      "loss": 0.3983,
+      "step": 2119
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.5069699416434e-08,
+      "loss": 0.3464,
+      "step": 2120
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.455630636170936e-08,
+      "loss": 0.3704,
+      "step": 2121
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.4048198681917157e-08,
+      "loss": 0.4041,
+      "step": 2122
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.3545377461929862e-08,
+      "loss": 0.3542,
+      "step": 2123
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.304784377533176e-08,
+      "loss": 0.3402,
+      "step": 2124
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 3.890625,
+      "learning_rate": 2.2555598684418643e-08,
+      "loss": 0.3509,
+      "step": 2125
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.2068643240193698e-08,
+      "loss": 0.3439,
+      "step": 2126
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.1586978482366072e-08,
+      "loss": 0.3212,
+      "step": 2127
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 3.578125,
+      "learning_rate": 2.1110605439349507e-08,
+      "loss": 0.392,
+      "step": 2128
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 3.640625,
+      "learning_rate": 2.0639525128258175e-08,
+      "loss": 0.4408,
+      "step": 2129
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 3.75,
+      "learning_rate": 2.0173738554906396e-08,
+      "loss": 0.4066,
+      "step": 2130
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.9713246713805588e-08,
+      "loss": 0.3757,
+      "step": 2131
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.925805058816177e-08,
+      "loss": 0.3712,
+      "step": 2132
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.8808151149874166e-08,
+      "loss": 0.3917,
+      "step": 2133
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.836354935953355e-08,
+      "loss": 0.3942,
+      "step": 2134
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 3.375,
+      "learning_rate": 1.792424616641919e-08,
+      "loss": 0.3914,
+      "step": 2135
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.7490242508496623e-08,
+      "loss": 0.3646,
+      "step": 2136
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.7061539312417107e-08,
+      "loss": 0.3556,
+      "step": 2137
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.6638137493514284e-08,
+      "loss": 0.4125,
+      "step": 2138
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.6220037955802514e-08,
+      "loss": 0.4753,
+      "step": 2139
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.5807241591976052e-08,
+      "loss": 0.4341,
+      "step": 2140
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.5399749283404874e-08,
+      "loss": 0.3527,
+      "step": 2141
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.4997561900135238e-08,
+      "loss": 0.437,
+      "step": 2142
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.4600680300886349e-08,
+      "loss": 0.3211,
+      "step": 2143
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 3.109375,
+      "learning_rate": 1.4209105333048978e-08,
+      "loss": 0.3737,
+      "step": 2144
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.3822837832683789e-08,
+      "loss": 0.4385,
+      "step": 2145
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.3441878624518568e-08,
+      "loss": 0.3731,
+      "step": 2146
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.3066228521948221e-08,
+      "loss": 0.3446,
+      "step": 2147
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.2695888327031448e-08,
+      "loss": 0.4252,
+      "step": 2148
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.2330858830490178e-08,
+      "loss": 0.3521,
+      "step": 2149
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.197114081170625e-08,
+      "loss": 0.3432,
+      "step": 2150
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 3.125,
+      "learning_rate": 1.1616735038722515e-08,
+      "loss": 0.3268,
+      "step": 2151
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.1267642268238121e-08,
+      "loss": 0.2998,
+      "step": 2152
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.092386324560879e-08,
+      "loss": 0.4658,
+      "step": 2153
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.058539870484515e-08,
+      "loss": 0.3432,
+      "step": 2154
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.0252249368609968e-08,
+      "loss": 0.3675,
+      "step": 2155
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 3.5625,
+      "learning_rate": 9.924415948217859e-09,
+      "loss": 0.3939,
+      "step": 2156
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 3.78125,
+      "learning_rate": 9.60189914363363e-09,
+      "loss": 0.4498,
+      "step": 2157
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 3.59375,
+      "learning_rate": 9.284699643469508e-09,
+      "loss": 0.3729,
+      "step": 2158
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 3.40625,
+      "learning_rate": 8.972818124985682e-09,
+      "loss": 0.3724,
+      "step": 2159
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 3.640625,
+      "learning_rate": 8.666255254086986e-09,
+      "loss": 0.4208,
+      "step": 2160
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 3.53125,
+      "learning_rate": 8.365011685322887e-09,
+      "loss": 0.4249,
+      "step": 2161
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 3.515625,
+      "learning_rate": 8.069088061885277e-09,
+      "loss": 0.2972,
+      "step": 2162
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 3.71875,
+      "learning_rate": 7.778485015607073e-09,
+      "loss": 0.3632,
+      "step": 2163
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 3.671875,
+      "learning_rate": 7.493203166961394e-09,
+      "loss": 0.3605,
+      "step": 2164
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 3.53125,
+      "learning_rate": 7.213243125059888e-09,
+      "loss": 0.3448,
+      "step": 2165
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 3.265625,
+      "learning_rate": 6.938605487651351e-09,
+      "loss": 0.3185,
+      "step": 2166
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 3.34375,
+      "learning_rate": 6.66929084112089e-09,
+      "loss": 0.3296,
+      "step": 2167
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 3.453125,
+      "learning_rate": 6.4052997604879815e-09,
+      "loss": 0.39,
+      "step": 2168
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 3.40625,
+      "learning_rate": 6.146632809405917e-09,
+      "loss": 0.3462,
+      "step": 2169
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 3.625,
+      "learning_rate": 5.89329054016069e-09,
+      "loss": 0.3796,
+      "step": 2170
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 3.265625,
+      "learning_rate": 5.645273493668224e-09,
+      "loss": 0.3469,
+      "step": 2171
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 3.703125,
+      "learning_rate": 5.402582199476036e-09,
+      "loss": 0.413,
+      "step": 2172
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 3.21875,
+      "learning_rate": 5.165217175759074e-09,
+      "loss": 0.338,
+      "step": 2173
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 3.71875,
+      "learning_rate": 4.933178929321103e-09,
+      "loss": 0.3149,
+      "step": 2174
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.7064679555919316e-09,
+      "loss": 0.4198,
+      "step": 2175
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.4850847386274096e-09,
+      "loss": 0.3953,
+      "step": 2176
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 3.59375,
+      "learning_rate": 4.269029751107489e-09,
+      "loss": 0.4589,
+      "step": 2177
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.058303454336221e-09,
+      "loss": 0.3714,
+      "step": 2178
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 3.375,
+      "learning_rate": 3.8529062982398115e-09,
+      "loss": 0.3725,
+      "step": 2179
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 3.3125,
+      "learning_rate": 3.652838721366625e-09,
+      "loss": 0.3229,
+      "step": 2180
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 3.578125,
+      "learning_rate": 3.4581011508849627e-09,
+      "loss": 0.3989,
+      "step": 2181
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 3.5,
+      "learning_rate": 3.2686940025836166e-09,
+      "loss": 0.3743,
+      "step": 2182
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 3.34375,
+      "learning_rate": 3.084617680869928e-09,
+      "loss": 0.3258,
+      "step": 2183
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.9058725787695084e-09,
+      "loss": 0.342,
+      "step": 2184
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.732459077924854e-09,
+      "loss": 0.3348,
+      "step": 2185
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.5643775485947875e-09,
+      "loss": 0.3737,
+      "step": 2186
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 3.78125,
+      "learning_rate": 2.4016283496544614e-09,
+      "loss": 0.3851,
+      "step": 2187
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 3.71875,
+      "learning_rate": 2.244211828592857e-09,
+      "loss": 0.381,
+      "step": 2188
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.09212832151362e-09,
+      "loss": 0.3737,
+      "step": 2189
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.945378153133393e-09,
+      "loss": 0.4281,
+      "step": 2190
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 4.40625,
+      "learning_rate": 1.8039616367823699e-09,
+      "loss": 0.3725,
+      "step": 2191
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 3.125,
+      "learning_rate": 1.6678790744015238e-09,
+      "loss": 0.337,
+      "step": 2192
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 3.125,
+      "learning_rate": 1.537130756543992e-09,
+      "loss": 0.34,
+      "step": 2193
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 3.375,
+      "learning_rate": 1.4117169623734106e-09,
+      "loss": 0.3481,
+      "step": 2194
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.2916379596636385e-09,
+      "loss": 0.4671,
+      "step": 2195
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.1768940047976462e-09,
+      "loss": 0.3572,
+      "step": 2196
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 3.375,
+      "learning_rate": 1.0674853427683484e-09,
+      "loss": 0.3619,
+      "step": 2197
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 3.734375,
+      "learning_rate": 9.634122071761065e-10,
+      "loss": 0.4076,
+      "step": 2198
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 3.671875,
+      "learning_rate": 8.646748202301158e-10,
+      "loss": 0.4031,
+      "step": 2199
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 3.65625,
+      "learning_rate": 7.712733927467409e-10,
+      "loss": 0.3915,
+      "step": 2200
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 3.5,
+      "learning_rate": 6.832081241495148e-10,
+      "loss": 0.3545,
+      "step": 2201
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 3.625,
+      "learning_rate": 6.004792024680295e-10,
+      "loss": 0.4565,
+      "step": 2202
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 3.296875,
+      "learning_rate": 5.230868043393233e-10,
+      "loss": 0.3428,
+      "step": 2203
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.510310950053831e-10,
+      "loss": 0.403,
+      "step": 2204
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 3.296875,
+      "learning_rate": 3.8431222831369905e-10,
+      "loss": 0.3544,
+      "step": 2205
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 3.859375,
+      "learning_rate": 3.2293034671754265e-10,
+      "loss": 0.3461,
+      "step": 2206
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 3.25,
+      "learning_rate": 2.668855812748561e-10,
+      "loss": 0.3099,
+      "step": 2207
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.1617805164741988e-10,
+      "loss": 0.3837,
+      "step": 2208
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.7080786610279565e-10,
+      "loss": 0.4317,
+      "step": 2209
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 3.375,
+      "learning_rate": 1.3077512151127293e-10,
+      "loss": 0.3566,
+      "step": 2210
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 3.34375,
+      "learning_rate": 9.60799033480897e-11,
+      "loss": 0.3656,
+      "step": 2211
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 3.8125,
+      "learning_rate": 6.672228569148953e-11,
+      "loss": 0.4197,
+      "step": 2212
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.270233122355416e-11,
+      "loss": 0.4238,
+      "step": 2213
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.402009122992599e-11,
+      "loss": 0.4018,
+      "step": 2214
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.0675605599530514e-11,
+      "loss": 0.4284,
+      "step": 2215
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.668902824576325e-12,
+      "loss": 0.3844,
+      "step": 2216
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 3.65625,
+      "learning_rate": 0.0,
+      "loss": 0.3469,
+      "step": 2217
+    },
+    {
+      "epoch": 3.0,
+      "step": 2217,
+      "total_flos": 3.0993976382906696e+18,
+      "train_loss": 0.4364770612552137,
+      "train_runtime": 30542.8191,
+      "train_samples_per_second": 9.296,
+      "train_steps_per_second": 0.073
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 2217,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 10000.0,
+  "total_flos": 3.0993976382906696e+18,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}