diff --git "a/checkpoint-2492/trainer_state.json" "b/checkpoint-2492/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-2492/trainer_state.json"
@@ -0,0 +1,17489 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.5002509284352102,
+  "eval_steps": 1246,
+  "global_step": 2492,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.46875,
+      "learning_rate": 2e-05,
+      "loss": 1.7713,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "eval_loss": 1.6074846982955933,
+      "eval_runtime": 346.224,
+      "eval_samples_per_second": 5.566,
+      "eval_steps_per_second": 2.784,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.578125,
+      "learning_rate": 4e-05,
+      "loss": 1.6697,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.796875,
+      "learning_rate": 6e-05,
+      "loss": 1.7212,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.78125,
+      "learning_rate": 8e-05,
+      "loss": 1.6821,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.0001,
+      "loss": 1.6306,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00012,
+      "loss": 1.5827,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00014,
+      "loss": 1.5527,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00016,
+      "loss": 1.4578,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00018,
+      "loss": 1.6053,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.0002,
+      "loss": 1.6568,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00019999998002980944,
+      "loss": 1.5348,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00019999992011924576,
+      "loss": 1.616,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00019999982026833282,
+      "loss": 1.5713,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.0001999996804771106,
+      "loss": 1.7015,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00019999950074563487,
+      "loss": 1.5893,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.0001999992810739774,
+      "loss": 1.6496,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00019999902146222598,
+      "loss": 1.3536,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00019999872191048425,
+      "loss": 1.6131,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.0001999983824188719,
+      "loss": 1.4954,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00019999800298752449,
+      "loss": 1.783,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00019999758361659357,
+      "loss": 1.9696,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00019999712430624665,
+      "loss": 1.6409,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.0001999966250566672,
+      "loss": 1.4331,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.0001999960858680546,
+      "loss": 1.403,
+      "step": 24
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00019999550674062417,
+      "loss": 1.62,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00019999488767460726,
+      "loss": 1.5016,
+      "step": 26
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00019999422867025113,
+      "loss": 1.516,
+      "step": 27
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00019999352972781896,
+      "loss": 1.318,
+      "step": 28
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.0001999927908475899,
+      "loss": 1.5259,
+      "step": 29
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00019999201202985914,
+      "loss": 1.4007,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00019999119327493764,
+      "loss": 1.563,
+      "step": 31
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0001999903345831525,
+      "loss": 1.5174,
+      "step": 32
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00019998943595484663,
+      "loss": 1.6393,
+      "step": 33
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00019998849739037895,
+      "loss": 1.4087,
+      "step": 34
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00019998751889012436,
+      "loss": 1.3014,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00019998650045447366,
+      "loss": 1.4343,
+      "step": 36
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0001999854420838336,
+      "loss": 1.5195,
+      "step": 37
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0001999843437786269,
+      "loss": 1.6591,
+      "step": 38
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00019998320553929225,
+      "loss": 1.5985,
+      "step": 39
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00019998202736628426,
+      "loss": 1.4239,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0001999808092600735,
+      "loss": 1.5175,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00019997955122114648,
+      "loss": 1.2641,
+      "step": 42
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00019997825325000564,
+      "loss": 1.5261,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.375,
+      "learning_rate": 0.0001999769153471694,
+      "loss": 1.4452,
+      "step": 44
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00019997553751317218,
+      "loss": 1.8778,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00019997411974856423,
+      "loss": 1.4568,
+      "step": 46
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00019997266205391183,
+      "loss": 1.4988,
+      "step": 47
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0001999711644297972,
+      "loss": 1.5492,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00019996962687681853,
+      "loss": 1.5654,
+      "step": 49
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00019996804939558985,
+      "loss": 1.3653,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00019996643198674124,
+      "loss": 1.237,
+      "step": 51
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0001999647746509187,
+      "loss": 1.7542,
+      "step": 52
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.0001999630773887842,
+      "loss": 1.5381,
+      "step": 53
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.0001999613402010156,
+      "loss": 1.5043,
+      "step": 54
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.0001999595630883068,
+      "loss": 1.5275,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00019995774605136748,
+      "loss": 1.2261,
+      "step": 56
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00019995588909092345,
+      "loss": 1.2426,
+      "step": 57
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00019995399220771638,
+      "loss": 1.3828,
+      "step": 58
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00019995205540250389,
+      "loss": 1.5301,
+      "step": 59
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00019995007867605952,
+      "loss": 1.4278,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0001999480620291728,
+      "loss": 1.6096,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00019994600546264919,
+      "loss": 1.7133,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00019994390897731007,
+      "loss": 1.5319,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0001999417725739928,
+      "loss": 1.619,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.0001999395962535507,
+      "loss": 1.5917,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00019993738001685297,
+      "loss": 1.4376,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00019993512386478476,
+      "loss": 1.4585,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0001999328277982472,
+      "loss": 1.6296,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.0001999304918181574,
+      "loss": 1.586,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.0001999281159254483,
+      "loss": 1.9335,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.0001999257001210688,
+      "loss": 1.5699,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001999232444059839,
+      "loss": 1.41,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00019992074878117435,
+      "loss": 1.7548,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00019991821324763693,
+      "loss": 1.7909,
+      "step": 74
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0001999156378063843,
+      "loss": 1.4009,
+      "step": 75
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00019991302245844517,
+      "loss": 1.6494,
+      "step": 76
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00019991036720486407,
+      "loss": 1.519,
+      "step": 77
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0001999076720467015,
+      "loss": 1.587,
+      "step": 78
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.000199904936985034,
+      "loss": 1.6404,
+      "step": 79
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00019990216202095388,
+      "loss": 1.3156,
+      "step": 80
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00019989934715556955,
+      "loss": 1.4342,
+      "step": 81
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00019989649239000518,
+      "loss": 1.4304,
+      "step": 82
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00019989359772540105,
+      "loss": 1.7874,
+      "step": 83
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0001998906631629133,
+      "loss": 1.6451,
+      "step": 84
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00019988768870371395,
+      "loss": 1.553,
+      "step": 85
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00019988467434899106,
+      "loss": 1.527,
+      "step": 86
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00019988162009994853,
+      "loss": 1.5158,
+      "step": 87
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.0001998785259578063,
+      "loss": 1.5275,
+      "step": 88
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0001998753919238001,
+      "loss": 1.7051,
+      "step": 89
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00019987221799918178,
+      "loss": 1.5388,
+      "step": 90
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00019986900418521894,
+      "loss": 1.4644,
+      "step": 91
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.0001998657504831952,
+      "loss": 1.4787,
+      "step": 92
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00019986245689441011,
+      "loss": 1.4985,
+      "step": 93
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00019985912342017917,
+      "loss": 1.3945,
+      "step": 94
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00019985575006183375,
+      "loss": 1.4905,
+      "step": 95
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.703125,
+      "learning_rate": 0.00019985233682072117,
+      "loss": 1.5378,
+      "step": 96
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00019984888369820473,
+      "loss": 1.6182,
+      "step": 97
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.0001998453906956636,
+      "loss": 1.7501,
+      "step": 98
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00019984185781449286,
+      "loss": 1.5575,
+      "step": 99
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00019983828505610362,
+      "loss": 1.5915,
+      "step": 100
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00019983467242192283,
+      "loss": 1.5046,
+      "step": 101
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.0001998310199133934,
+      "loss": 1.5739,
+      "step": 102
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00019982732753197414,
+      "loss": 1.585,
+      "step": 103
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001998235952791398,
+      "loss": 1.3631,
+      "step": 104
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00019981982315638106,
+      "loss": 1.4332,
+      "step": 105
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.0001998160111652045,
+      "loss": 1.754,
+      "step": 106
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.0001998121593071327,
+      "loss": 1.2536,
+      "step": 107
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00019980826758370405,
+      "loss": 1.536,
+      "step": 108
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00019980433599647295,
+      "loss": 1.6197,
+      "step": 109
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0001998003645470097,
+      "loss": 1.6064,
+      "step": 110
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00019979635323690045,
+      "loss": 1.5005,
+      "step": 111
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.375,
+      "learning_rate": 0.0001997923020677474,
+      "loss": 1.7366,
+      "step": 112
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.0001997882110411686,
+      "loss": 1.5608,
+      "step": 113
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00019978408015879794,
+      "loss": 1.4183,
+      "step": 114
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00019977990942228542,
+      "loss": 1.2348,
+      "step": 115
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00019977569883329678,
+      "loss": 1.6173,
+      "step": 116
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00019977144839351376,
+      "loss": 1.4582,
+      "step": 117
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00019976715810463404,
+      "loss": 1.4523,
+      "step": 118
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.0001997628279683711,
+      "loss": 1.6971,
+      "step": 119
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.0001997584579864545,
+      "loss": 1.744,
+      "step": 120
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00019975404816062955,
+      "loss": 1.4053,
+      "step": 121
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.0001997495984926576,
+      "loss": 1.5215,
+      "step": 122
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00019974510898431585,
+      "loss": 1.6396,
+      "step": 123
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.625,
+      "learning_rate": 0.00019974057963739741,
+      "loss": 1.183,
+      "step": 124
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00019973601045371136,
+      "loss": 1.438,
+      "step": 125
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.25,
+      "learning_rate": 0.0001997314014350826,
+      "loss": 1.4428,
+      "step": 126
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00019972675258335203,
+      "loss": 1.6131,
+      "step": 127
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.0001997220639003764,
+      "loss": 1.6727,
+      "step": 128
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00019971733538802838,
+      "loss": 1.4967,
+      "step": 129
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00019971256704819657,
+      "loss": 1.7573,
+      "step": 130
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00019970775888278545,
+      "loss": 1.5187,
+      "step": 131
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00019970291089371544,
+      "loss": 1.4927,
+      "step": 132
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.00019969802308292286,
+      "loss": 1.5802,
+      "step": 133
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00019969309545235986,
+      "loss": 1.631,
+      "step": 134
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00019968812800399461,
+      "loss": 1.6134,
+      "step": 135
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00019968312073981109,
+      "loss": 1.3874,
+      "step": 136
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00019967807366180925,
+      "loss": 1.5713,
+      "step": 137
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00019967298677200487,
+      "loss": 1.4594,
+      "step": 138
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00019966786007242972,
+      "loss": 1.4545,
+      "step": 139
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00019966269356513143,
+      "loss": 1.3715,
+      "step": 140
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00019965748725217348,
+      "loss": 1.6846,
+      "step": 141
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.0001996522411356353,
+      "loss": 1.4046,
+      "step": 142
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00019964695521761225,
+      "loss": 1.5223,
+      "step": 143
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00019964162950021549,
+      "loss": 1.3543,
+      "step": 144
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00019963626398557217,
+      "loss": 1.4538,
+      "step": 145
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00019963085867582528,
+      "loss": 1.7505,
+      "step": 146
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00019962541357313368,
+      "loss": 1.5098,
+      "step": 147
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00019961992867967228,
+      "loss": 1.5073,
+      "step": 148
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00019961440399763166,
+      "loss": 1.1638,
+      "step": 149
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00019960883952921843,
+      "loss": 1.4793,
+      "step": 150
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00019960323527665508,
+      "loss": 1.5584,
+      "step": 151
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00019959759124217995,
+      "loss": 1.4312,
+      "step": 152
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.0001995919074280473,
+      "loss": 1.4258,
+      "step": 153
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00019958618383652724,
+      "loss": 1.4048,
+      "step": 154
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00019958042046990587,
+      "loss": 1.384,
+      "step": 155
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00019957461733048501,
+      "loss": 1.3096,
+      "step": 156
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0001995687744205825,
+      "loss": 1.5201,
+      "step": 157
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00019956289174253197,
+      "loss": 1.3611,
+      "step": 158
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00019955696929868308,
+      "loss": 1.4786,
+      "step": 159
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.0001995510070914012,
+      "loss": 1.57,
+      "step": 160
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00019954500512306772,
+      "loss": 1.4074,
+      "step": 161
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00019953896339607976,
+      "loss": 1.5458,
+      "step": 162
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0001995328819128505,
+      "loss": 1.5702,
+      "step": 163
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00019952676067580885,
+      "loss": 1.4379,
+      "step": 164
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00019952059968739966,
+      "loss": 1.4716,
+      "step": 165
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00019951439895008366,
+      "loss": 1.868,
+      "step": 166
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00019950815846633748,
+      "loss": 1.2896,
+      "step": 167
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00019950187823865355,
+      "loss": 1.513,
+      "step": 168
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00019949555826954022,
+      "loss": 1.5134,
+      "step": 169
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00019948919856152176,
+      "loss": 1.736,
+      "step": 170
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.0001994827991171382,
+      "loss": 1.4261,
+      "step": 171
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00019947635993894554,
+      "loss": 1.3708,
+      "step": 172
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00019946988102951562,
+      "loss": 1.4693,
+      "step": 173
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00019946336239143607,
+      "loss": 1.4593,
+      "step": 174
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00019945680402731053,
+      "loss": 1.3815,
+      "step": 175
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.0001994502059397584,
+      "loss": 1.5947,
+      "step": 176
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00019944356813141504,
+      "loss": 1.2579,
+      "step": 177
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00019943689060493154,
+      "loss": 1.4636,
+      "step": 178
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00019943017336297497,
+      "loss": 1.5428,
+      "step": 179
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0001994234164082282,
+      "loss": 1.387,
+      "step": 180
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00019941661974339,
+      "loss": 1.4689,
+      "step": 181
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.000199409783371175,
+      "loss": 1.2979,
+      "step": 182
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00019940290729431363,
+      "loss": 1.4404,
+      "step": 183
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00019939599151555225,
+      "loss": 1.4686,
+      "step": 184
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00019938903603765304,
+      "loss": 1.8386,
+      "step": 185
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00019938204086339406,
+      "loss": 1.4209,
+      "step": 186
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.0001993750059955692,
+      "loss": 1.4783,
+      "step": 187
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00019936793143698818,
+      "loss": 1.5969,
+      "step": 188
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.21875,
+      "learning_rate": 0.0001993608171904767,
+      "loss": 1.3878,
+      "step": 189
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.0001993536632588761,
+      "loss": 1.5059,
+      "step": 190
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0001993464696450438,
+      "loss": 1.4611,
+      "step": 191
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00019933923635185284,
+      "loss": 1.4915,
+      "step": 192
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00019933196338219226,
+      "loss": 1.6292,
+      "step": 193
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.000199324650738967,
+      "loss": 1.2962,
+      "step": 194
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00019931729842509767,
+      "loss": 1.7946,
+      "step": 195
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.0001993099064435208,
+      "loss": 1.322,
+      "step": 196
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00019930247479718884,
+      "loss": 1.3847,
+      "step": 197
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00019929500348906998,
+      "loss": 1.4842,
+      "step": 198
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00019928749252214831,
+      "loss": 1.3623,
+      "step": 199
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0001992799418994237,
+      "loss": 1.4669,
+      "step": 200
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00019927235162391197,
+      "loss": 1.5735,
+      "step": 201
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0001992647216986446,
+      "loss": 1.6234,
+      "step": 202
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.0001992570521266691,
+      "loss": 1.5895,
+      "step": 203
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00019924934291104868,
+      "loss": 1.6561,
+      "step": 204
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00019924159405486247,
+      "loss": 1.4892,
+      "step": 205
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00019923380556120536,
+      "loss": 1.3873,
+      "step": 206
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.0001992259774331881,
+      "loss": 1.6435,
+      "step": 207
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00019921810967393728,
+      "loss": 1.6225,
+      "step": 208
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00019921020228659534,
+      "loss": 1.432,
+      "step": 209
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.0001992022552743205,
+      "loss": 1.5165,
+      "step": 210
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.1875,
+      "learning_rate": 0.0001991942686402868,
+      "loss": 1.3685,
+      "step": 211
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00019918624238768418,
+      "loss": 1.5645,
+      "step": 212
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00019917817651971833,
+      "loss": 1.5071,
+      "step": 213
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.0001991700710396108,
+      "loss": 1.588,
+      "step": 214
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00019916192595059894,
+      "loss": 1.3804,
+      "step": 215
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00019915374125593593,
+      "loss": 1.3841,
+      "step": 216
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00019914551695889075,
+      "loss": 1.5071,
+      "step": 217
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00019913725306274828,
+      "loss": 1.667,
+      "step": 218
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00019912894957080908,
+      "loss": 1.6396,
+      "step": 219
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.00019912060648638962,
+      "loss": 1.7855,
+      "step": 220
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00019911222381282218,
+      "loss": 1.5712,
+      "step": 221
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.0001991038015534548,
+      "loss": 1.3365,
+      "step": 222
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.0001990953397116514,
+      "loss": 1.3764,
+      "step": 223
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.96875,
+      "learning_rate": 0.00019908683829079164,
+      "loss": 1.4361,
+      "step": 224
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.84375,
+      "learning_rate": 0.000199078297294271,
+      "loss": 1.7311,
+      "step": 225
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00019906971672550085,
+      "loss": 1.3825,
+      "step": 226
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00019906109658790827,
+      "loss": 1.6561,
+      "step": 227
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00019905243688493615,
+      "loss": 1.5624,
+      "step": 228
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00019904373762004326,
+      "loss": 1.4859,
+      "step": 229
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.0001990349987967041,
+      "loss": 1.4023,
+      "step": 230
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00019902622041840894,
+      "loss": 1.4844,
+      "step": 231
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00019901740248866396,
+      "loss": 1.5597,
+      "step": 232
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00019900854501099107,
+      "loss": 1.4096,
+      "step": 233
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00019899964798892795,
+      "loss": 1.7266,
+      "step": 234
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.0001989907114260281,
+      "loss": 1.3375,
+      "step": 235
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00019898173532586087,
+      "loss": 1.7208,
+      "step": 236
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00019897271969201131,
+      "loss": 1.4982,
+      "step": 237
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00019896366452808027,
+      "loss": 1.5062,
+      "step": 238
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5,
+      "learning_rate": 0.0001989545698376845,
+      "loss": 1.5933,
+      "step": 239
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00019894543562445636,
+      "loss": 1.6261,
+      "step": 240
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00019893626189204416,
+      "loss": 1.6287,
+      "step": 241
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00019892704864411188,
+      "loss": 1.365,
+      "step": 242
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00019891779588433934,
+      "loss": 1.3112,
+      "step": 243
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00019890850361642214,
+      "loss": 1.3559,
+      "step": 244
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00019889917184407165,
+      "loss": 1.5695,
+      "step": 245
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.000198889800571015,
+      "loss": 1.6937,
+      "step": 246
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00019888038980099513,
+      "loss": 1.3175,
+      "step": 247
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00019887093953777066,
+      "loss": 1.8866,
+      "step": 248
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00019886144978511618,
+      "loss": 1.4656,
+      "step": 249
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00019885192054682187,
+      "loss": 1.4383,
+      "step": 250
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00019884235182669375,
+      "loss": 1.5011,
+      "step": 251
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00019883274362855361,
+      "loss": 1.6484,
+      "step": 252
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.000198823095956239,
+      "loss": 1.6417,
+      "step": 253
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00019881340881360325,
+      "loss": 1.6253,
+      "step": 254
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.0001988036822045154,
+      "loss": 1.3941,
+      "step": 255
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00019879391613286028,
+      "loss": 1.4892,
+      "step": 256
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00019878411060253857,
+      "loss": 1.61,
+      "step": 257
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00019877426561746664,
+      "loss": 1.3753,
+      "step": 258
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00019876438118157657,
+      "loss": 1.3674,
+      "step": 259
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.0001987544572988162,
+      "loss": 1.4873,
+      "step": 260
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00019874449397314924,
+      "loss": 1.4838,
+      "step": 261
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00019873449120855506,
+      "loss": 1.5402,
+      "step": 262
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.0001987244490090288,
+      "loss": 1.5877,
+      "step": 263
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00019871436737858134,
+      "loss": 1.2991,
+      "step": 264
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00019870424632123933,
+      "loss": 1.5268,
+      "step": 265
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00019869408584104517,
+      "loss": 1.4216,
+      "step": 266
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00019868388594205698,
+      "loss": 1.5021,
+      "step": 267
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00019867364662834864,
+      "loss": 1.4991,
+      "step": 268
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0001986633679040098,
+      "loss": 1.6474,
+      "step": 269
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00019865304977314575,
+      "loss": 1.3488,
+      "step": 270
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00019864269223987768,
+      "loss": 1.5294,
+      "step": 271
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.0001986322953083424,
+      "loss": 1.3768,
+      "step": 272
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00019862185898269238,
+      "loss": 1.4052,
+      "step": 273
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.0001986113832670961,
+      "loss": 1.3252,
+      "step": 274
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.0001986008681657375,
+      "loss": 1.6497,
+      "step": 275
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00019859031368281636,
+      "loss": 1.3989,
+      "step": 276
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00019857971982254816,
+      "loss": 1.6634,
+      "step": 277
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00019856908658916423,
+      "loss": 1.3491,
+      "step": 278
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.0001985584139869114,
+      "loss": 1.4511,
+      "step": 279
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00019854770202005245,
+      "loss": 1.4468,
+      "step": 280
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00019853695069286568,
+      "loss": 1.4159,
+      "step": 281
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00019852616000964527,
+      "loss": 1.8567,
+      "step": 282
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.0001985153299747011,
+      "loss": 1.4583,
+      "step": 283
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00019850446059235861,
+      "loss": 1.5881,
+      "step": 284
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00019849355186695917,
+      "loss": 1.5484,
+      "step": 285
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00019848260380285975,
+      "loss": 1.5257,
+      "step": 286
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00019847161640443303,
+      "loss": 1.5502,
+      "step": 287
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00019846058967606744,
+      "loss": 1.55,
+      "step": 288
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00019844952362216707,
+      "loss": 1.4644,
+      "step": 289
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00019843841824715176,
+      "loss": 1.5144,
+      "step": 290
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.000198427273555457,
+      "loss": 1.5633,
+      "step": 291
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00019841608955153414,
+      "loss": 1.3997,
+      "step": 292
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00019840486623984998,
+      "loss": 1.6741,
+      "step": 293
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00019839360362488719,
+      "loss": 1.5441,
+      "step": 294
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00019838230171114415,
+      "loss": 1.5567,
+      "step": 295
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00019837096050313483,
+      "loss": 1.5753,
+      "step": 296
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00019835958000538902,
+      "loss": 1.4375,
+      "step": 297
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00019834816022245205,
+      "loss": 1.5508,
+      "step": 298
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00019833670115888507,
+      "loss": 1.8535,
+      "step": 299
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00019832520281926486,
+      "loss": 1.6437,
+      "step": 300
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00019831366520818396,
+      "loss": 1.6356,
+      "step": 301
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00019830208833025042,
+      "loss": 1.6086,
+      "step": 302
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00019829047219008822,
+      "loss": 1.504,
+      "step": 303
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00019827881679233678,
+      "loss": 1.3962,
+      "step": 304
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00019826712214165138,
+      "loss": 1.3045,
+      "step": 305
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00019825538824270287,
+      "loss": 1.5953,
+      "step": 306
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00019824361510017782,
+      "loss": 1.3438,
+      "step": 307
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0001982318027187785,
+      "loss": 1.3101,
+      "step": 308
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00019821995110322278,
+      "loss": 1.7562,
+      "step": 309
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00019820806025824426,
+      "loss": 1.3025,
+      "step": 310
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00019819613018859218,
+      "loss": 1.5367,
+      "step": 311
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0001981841608990315,
+      "loss": 1.5079,
+      "step": 312
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.0001981721523943427,
+      "loss": 1.512,
+      "step": 313
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0001981601046793221,
+      "loss": 1.6677,
+      "step": 314
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.0001981480177587816,
+      "loss": 1.459,
+      "step": 315
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00019813589163754876,
+      "loss": 1.4897,
+      "step": 316
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00019812372632046677,
+      "loss": 1.4271,
+      "step": 317
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.0001981115218123945,
+      "loss": 1.4733,
+      "step": 318
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00019809927811820655,
+      "loss": 1.3294,
+      "step": 319
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.000198086995242793,
+      "loss": 1.5159,
+      "step": 320
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00019807467319105974,
+      "loss": 1.5483,
+      "step": 321
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00019806231196792826,
+      "loss": 1.4135,
+      "step": 322
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.0001980499115783356,
+      "loss": 1.4924,
+      "step": 323
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00019803747202723463,
+      "loss": 1.4862,
+      "step": 324
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00019802499331959364,
+      "loss": 1.5085,
+      "step": 325
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00019801247546039674,
+      "loss": 1.5443,
+      "step": 326
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00019799991845464358,
+      "loss": 1.383,
+      "step": 327
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00019798732230734954,
+      "loss": 1.5014,
+      "step": 328
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.0001979746870235455,
+      "loss": 1.4689,
+      "step": 329
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00019796201260827804,
+      "loss": 1.4847,
+      "step": 330
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00019794929906660945,
+      "loss": 1.5658,
+      "step": 331
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00019793654640361748,
+      "loss": 1.3523,
+      "step": 332
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00019792375462439563,
+      "loss": 1.6846,
+      "step": 333
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.000197910923734053,
+      "loss": 1.2952,
+      "step": 334
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.0001978980537377142,
+      "loss": 1.5212,
+      "step": 335
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00019788514464051972,
+      "loss": 1.6539,
+      "step": 336
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00019787219644762534,
+      "loss": 1.3719,
+      "step": 337
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00019785920916420274,
+      "loss": 1.3044,
+      "step": 338
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00019784618279543903,
+      "loss": 1.3819,
+      "step": 339
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.000197833117346537,
+      "loss": 1.4986,
+      "step": 340
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00019782001282271506,
+      "loss": 1.6878,
+      "step": 341
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00019780686922920716,
+      "loss": 1.6453,
+      "step": 342
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00019779368657126293,
+      "loss": 1.5213,
+      "step": 343
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.0001977804648541476,
+      "loss": 1.6063,
+      "step": 344
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00019776720408314194,
+      "loss": 1.6762,
+      "step": 345
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00019775390426354236,
+      "loss": 1.8088,
+      "step": 346
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00019774056540066085,
+      "loss": 1.8082,
+      "step": 347
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00019772718749982505,
+      "loss": 1.6273,
+      "step": 348
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0001977137705663781,
+      "loss": 1.5376,
+      "step": 349
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00019770031460567876,
+      "loss": 1.4629,
+      "step": 350
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00019768681962310144,
+      "loss": 1.518,
+      "step": 351
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00019767328562403602,
+      "loss": 1.5077,
+      "step": 352
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.0001976597126138881,
+      "loss": 1.3512,
+      "step": 353
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00019764610059807875,
+      "loss": 1.4923,
+      "step": 354
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00019763244958204468,
+      "loss": 1.8882,
+      "step": 355
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00019761875957123817,
+      "loss": 1.7094,
+      "step": 356
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.375,
+      "learning_rate": 0.000197605030571127,
+      "loss": 1.3824,
+      "step": 357
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0001975912625871947,
+      "loss": 1.7543,
+      "step": 358
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00019757745562494012,
+      "loss": 1.641,
+      "step": 359
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.0001975636096898779,
+      "loss": 1.269,
+      "step": 360
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00019754972478753816,
+      "loss": 1.5104,
+      "step": 361
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00019753580092346656,
+      "loss": 1.2405,
+      "step": 362
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00019752183810322433,
+      "loss": 1.5498,
+      "step": 363
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0001975078363323883,
+      "loss": 1.5409,
+      "step": 364
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00019749379561655078,
+      "loss": 1.6554,
+      "step": 365
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.0001974797159613198,
+      "loss": 1.6083,
+      "step": 366
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00019746559737231868,
+      "loss": 1.4714,
+      "step": 367
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00019745143985518655,
+      "loss": 1.3656,
+      "step": 368
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00019743724341557794,
+      "loss": 1.5476,
+      "step": 369
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00019742300805916294,
+      "loss": 1.7405,
+      "step": 370
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00019740873379162725,
+      "loss": 1.6742,
+      "step": 371
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00019739442061867201,
+      "loss": 1.4479,
+      "step": 372
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00019738006854601404,
+      "loss": 1.6121,
+      "step": 373
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00019736567757938552,
+      "loss": 1.8205,
+      "step": 374
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00019735124772453431,
+      "loss": 1.5291,
+      "step": 375
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.375,
+      "learning_rate": 0.0001973367789872237,
+      "loss": 1.4695,
+      "step": 376
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00019732227137323266,
+      "loss": 1.4894,
+      "step": 377
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00019730772488835548,
+      "loss": 1.6287,
+      "step": 378
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00019729313953840215,
+      "loss": 1.4721,
+      "step": 379
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00019727851532919805,
+      "loss": 1.7247,
+      "step": 380
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00019726385226658422,
+      "loss": 1.4187,
+      "step": 381
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00019724915035641708,
+      "loss": 1.57,
+      "step": 382
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00019723440960456868,
+      "loss": 1.3531,
+      "step": 383
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00019721963001692648,
+      "loss": 1.782,
+      "step": 384
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00019720481159939354,
+      "loss": 1.4081,
+      "step": 385
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00019718995435788838,
+      "loss": 1.2604,
+      "step": 386
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00019717505829834506,
+      "loss": 1.5677,
+      "step": 387
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.0001971601234267131,
+      "loss": 1.4733,
+      "step": 388
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00019714514974895753,
+      "loss": 1.6811,
+      "step": 389
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00019713013727105895,
+      "loss": 1.6621,
+      "step": 390
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00019711508599901332,
+      "loss": 1.4454,
+      "step": 391
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00019709999593883225,
+      "loss": 1.7683,
+      "step": 392
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00019708486709654272,
+      "loss": 1.5736,
+      "step": 393
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00019706969947818728,
+      "loss": 1.2613,
+      "step": 394
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00019705449308982394,
+      "loss": 1.5227,
+      "step": 395
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00019703924793752612,
+      "loss": 1.6748,
+      "step": 396
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00019702396402738285,
+      "loss": 1.6829,
+      "step": 397
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.0001970086413654986,
+      "loss": 1.5011,
+      "step": 398
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00019699327995799327,
+      "loss": 1.3323,
+      "step": 399
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00019697787981100225,
+      "loss": 1.2017,
+      "step": 400
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00019696244093067643,
+      "loss": 1.3887,
+      "step": 401
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00019694696332318216,
+      "loss": 1.6484,
+      "step": 402
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00019693144699470128,
+      "loss": 1.9888,
+      "step": 403
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00019691589195143104,
+      "loss": 1.4005,
+      "step": 404
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00019690029819958418,
+      "loss": 1.6491,
+      "step": 405
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00019688466574538892,
+      "loss": 1.415,
+      "step": 406
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.0001968689945950889,
+      "loss": 1.6626,
+      "step": 407
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00019685328475494332,
+      "loss": 1.6892,
+      "step": 408
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00019683753623122663,
+      "loss": 1.481,
+      "step": 409
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00019682174903022892,
+      "loss": 1.3528,
+      "step": 410
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00019680592315825563,
+      "loss": 1.3676,
+      "step": 411
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0001967900586216277,
+      "loss": 1.7688,
+      "step": 412
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00019677415542668144,
+      "loss": 1.8405,
+      "step": 413
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00019675821357976873,
+      "loss": 1.3146,
+      "step": 414
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.0001967422330872567,
+      "loss": 1.813,
+      "step": 415
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00019672621395552814,
+      "loss": 1.5814,
+      "step": 416
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00019671015619098108,
+      "loss": 1.9684,
+      "step": 417
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.000196694059800029,
+      "loss": 1.5554,
+      "step": 418
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00019667792478910096,
+      "loss": 1.5912,
+      "step": 419
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.375,
+      "learning_rate": 0.0001966617511646413,
+      "loss": 1.5799,
+      "step": 420
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00019664553893310982,
+      "loss": 1.7361,
+      "step": 421
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00019662928810098177,
+      "loss": 1.6091,
+      "step": 422
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00019661299867474775,
+      "loss": 1.7153,
+      "step": 423
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0001965966706609139,
+      "loss": 1.7456,
+      "step": 424
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00019658030406600162,
+      "loss": 1.6354,
+      "step": 425
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.0001965638988965478,
+      "loss": 1.6233,
+      "step": 426
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00019654745515910476,
+      "loss": 1.4668,
+      "step": 427
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00019653097286024017,
+      "loss": 1.2301,
+      "step": 428
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00019651445200653708,
+      "loss": 1.5367,
+      "step": 429
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.375,
+      "learning_rate": 0.0001964978926045941,
+      "loss": 1.4191,
+      "step": 430
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00019648129466102497,
+      "loss": 1.3293,
+      "step": 431
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.0001964646581824591,
+      "loss": 1.704,
+      "step": 432
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00019644798317554107,
+      "loss": 1.4271,
+      "step": 433
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00019643126964693098,
+      "loss": 1.6204,
+      "step": 434
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00019641451760330433,
+      "loss": 1.4538,
+      "step": 435
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00019639772705135183,
+      "loss": 1.4909,
+      "step": 436
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.0001963808979977798,
+      "loss": 1.6014,
+      "step": 437
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00019636403044930977,
+      "loss": 1.4551,
+      "step": 438
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00019634712441267874,
+      "loss": 1.485,
+      "step": 439
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.000196330179894639,
+      "loss": 1.4513,
+      "step": 440
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.0001963131969019583,
+      "loss": 1.5908,
+      "step": 441
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00019629617544141966,
+      "loss": 1.5176,
+      "step": 442
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00019627911551982157,
+      "loss": 1.3734,
+      "step": 443
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00019626201714397778,
+      "loss": 1.4622,
+      "step": 444
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00019624488032071754,
+      "loss": 1.4094,
+      "step": 445
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.0001962277050568852,
+      "loss": 1.3841,
+      "step": 446
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.0001962104913593408,
+      "loss": 1.3523,
+      "step": 447
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00019619323923495942,
+      "loss": 1.5124,
+      "step": 448
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00019617594869063173,
+      "loss": 1.4121,
+      "step": 449
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00019615861973326358,
+      "loss": 1.3274,
+      "step": 450
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.0001961412523697762,
+      "loss": 1.3597,
+      "step": 451
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00019612384660710623,
+      "loss": 1.6987,
+      "step": 452
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.0001961064024522056,
+      "loss": 1.5839,
+      "step": 453
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00019608891991204153,
+      "loss": 1.6882,
+      "step": 454
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00019607139899359663,
+      "loss": 1.3767,
+      "step": 455
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00019605383970386884,
+      "loss": 1.5682,
+      "step": 456
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00019603624204987138,
+      "loss": 1.3616,
+      "step": 457
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00019601860603863286,
+      "loss": 1.4667,
+      "step": 458
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00019600093167719714,
+      "loss": 1.3562,
+      "step": 459
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00019598321897262342,
+      "loss": 1.5472,
+      "step": 460
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00019596546793198622,
+      "loss": 1.4332,
+      "step": 461
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00019594767856237542,
+      "loss": 1.5877,
+      "step": 462
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00019592985087089613,
+      "loss": 1.4411,
+      "step": 463
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00019591198486466877,
+      "loss": 1.5306,
+      "step": 464
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00019589408055082914,
+      "loss": 1.0967,
+      "step": 465
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00019587613793652827,
+      "loss": 1.5201,
+      "step": 466
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00019585815702893248,
+      "loss": 1.5792,
+      "step": 467
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00019584013783522348,
+      "loss": 1.5642,
+      "step": 468
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00019582208036259812,
+      "loss": 1.7622,
+      "step": 469
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.0001958039846182687,
+      "loss": 1.6048,
+      "step": 470
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00019578585060946272,
+      "loss": 1.3616,
+      "step": 471
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00019576767834342292,
+      "loss": 1.3354,
+      "step": 472
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.0001957494678274074,
+      "loss": 1.455,
+      "step": 473
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00019573121906868953,
+      "loss": 1.5178,
+      "step": 474
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00019571293207455787,
+      "loss": 1.5893,
+      "step": 475
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00019569460685231642,
+      "loss": 1.7788,
+      "step": 476
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.390625,
+      "learning_rate": 0.00019567624340928425,
+      "loss": 1.6235,
+      "step": 477
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00019565784175279584,
+      "loss": 1.4642,
+      "step": 478
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00019563940189020086,
+      "loss": 1.3778,
+      "step": 479
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00019562092382886427,
+      "loss": 1.5411,
+      "step": 480
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00019560240757616627,
+      "loss": 1.7177,
+      "step": 481
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00019558385313950234,
+      "loss": 1.5359,
+      "step": 482
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00019556526052628317,
+      "loss": 1.4694,
+      "step": 483
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00019554662974393473,
+      "loss": 1.3667,
+      "step": 484
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00019552796079989825,
+      "loss": 1.4745,
+      "step": 485
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00019550925370163009,
+      "loss": 1.3581,
+      "step": 486
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00019549050845660206,
+      "loss": 1.4969,
+      "step": 487
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00019547172507230101,
+      "loss": 1.5833,
+      "step": 488
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.0001954529035562291,
+      "loss": 1.556,
+      "step": 489
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0001954340439159037,
+      "loss": 1.4698,
+      "step": 490
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00019541514615885748,
+      "loss": 1.5273,
+      "step": 491
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00019539621029263823,
+      "loss": 1.7497,
+      "step": 492
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.000195377236324809,
+      "loss": 1.3911,
+      "step": 493
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.0001953582242629481,
+      "loss": 1.4641,
+      "step": 494
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.000195339174114649,
+      "loss": 1.5019,
+      "step": 495
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00019532008588752042,
+      "loss": 1.4851,
+      "step": 496
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00019530095958918625,
+      "loss": 1.3866,
+      "step": 497
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00019528179522728558,
+      "loss": 1.6702,
+      "step": 498
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.0001952625928094728,
+      "loss": 1.2754,
+      "step": 499
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00019524335234341736,
+      "loss": 1.8533,
+      "step": 500
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00019522407383680402,
+      "loss": 1.7172,
+      "step": 501
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.0001952047572973327,
+      "loss": 1.291,
+      "step": 502
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.0001951854027327184,
+      "loss": 1.156,
+      "step": 503
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00019516601015069155,
+      "loss": 1.4576,
+      "step": 504
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00019514657955899755,
+      "loss": 1.5379,
+      "step": 505
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00019512711096539702,
+      "loss": 1.5343,
+      "step": 506
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00019510760437766583,
+      "loss": 1.4725,
+      "step": 507
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00019508805980359502,
+      "loss": 1.641,
+      "step": 508
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0001950684772509907,
+      "loss": 1.4425,
+      "step": 509
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00019504885672767423,
+      "loss": 1.2601,
+      "step": 510
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00019502919824148217,
+      "loss": 1.5281,
+      "step": 511
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00019500950180026614,
+      "loss": 1.3606,
+      "step": 512
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00019498976741189297,
+      "loss": 1.9315,
+      "step": 513
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00019496999508424468,
+      "loss": 1.4164,
+      "step": 514
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00019495018482521844,
+      "loss": 1.3494,
+      "step": 515
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.0001949303366427265,
+      "loss": 1.6507,
+      "step": 516
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.0001949104505446963,
+      "loss": 1.4297,
+      "step": 517
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00019489052653907045,
+      "loss": 1.7807,
+      "step": 518
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00019487056463380663,
+      "loss": 1.6004,
+      "step": 519
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00019485056483687773,
+      "loss": 1.6449,
+      "step": 520
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00019483052715627176,
+      "loss": 1.3497,
+      "step": 521
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00019481045159999182,
+      "loss": 1.8284,
+      "step": 522
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00019479033817605617,
+      "loss": 1.6384,
+      "step": 523
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00019477018689249818,
+      "loss": 1.6449,
+      "step": 524
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.0001947499977573664,
+      "loss": 1.6561,
+      "step": 525
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00019472977077872436,
+      "loss": 1.5114,
+      "step": 526
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00019470950596465087,
+      "loss": 1.5778,
+      "step": 527
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00019468920332323973,
+      "loss": 1.4936,
+      "step": 528
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00019466886286259992,
+      "loss": 1.6421,
+      "step": 529
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00019464848459085546,
+      "loss": 1.6369,
+      "step": 530
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00019462806851614555,
+      "loss": 1.5903,
+      "step": 531
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00019460761464662444,
+      "loss": 1.506,
+      "step": 532
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00019458712299046146,
+      "loss": 1.4444,
+      "step": 533
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00019456659355584105,
+      "loss": 1.6022,
+      "step": 534
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00019454602635096278,
+      "loss": 1.7779,
+      "step": 535
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.0001945254213840413,
+      "loss": 1.6928,
+      "step": 536
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00019450477866330622,
+      "loss": 1.7402,
+      "step": 537
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00019448409819700237,
+      "loss": 1.5394,
+      "step": 538
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00019446337999338962,
+      "loss": 1.4911,
+      "step": 539
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00019444262406074288,
+      "loss": 1.1812,
+      "step": 540
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00019442183040735214,
+      "loss": 1.2605,
+      "step": 541
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00019440099904152248,
+      "loss": 1.4559,
+      "step": 542
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00019438012997157402,
+      "loss": 1.5438,
+      "step": 543
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00019435922320584198,
+      "loss": 1.471,
+      "step": 544
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00019433827875267655,
+      "loss": 1.541,
+      "step": 545
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00019431729662044305,
+      "loss": 1.4548,
+      "step": 546
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00019429627681752184,
+      "loss": 1.2798,
+      "step": 547
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00019427521935230826,
+      "loss": 1.7242,
+      "step": 548
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00019425412423321276,
+      "loss": 1.2921,
+      "step": 549
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00019423299146866088,
+      "loss": 1.4974,
+      "step": 550
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00019421182106709305,
+      "loss": 1.4301,
+      "step": 551
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.0001941906130369648,
+      "loss": 1.5389,
+      "step": 552
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0001941693673867467,
+      "loss": 1.6428,
+      "step": 553
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00019414808412492445,
+      "loss": 1.7382,
+      "step": 554
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00019412676325999856,
+      "loss": 1.6969,
+      "step": 555
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00019410540480048466,
+      "loss": 1.7304,
+      "step": 556
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00019408400875491343,
+      "loss": 1.3099,
+      "step": 557
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00019406257513183056,
+      "loss": 1.395,
+      "step": 558
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.0001940411039397967,
+      "loss": 1.3284,
+      "step": 559
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.0001940195951873875,
+      "loss": 1.6325,
+      "step": 560
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00019399804888319368,
+      "loss": 1.6357,
+      "step": 561
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.0001939764650358209,
+      "loss": 1.8126,
+      "step": 562
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00019395484365388978,
+      "loss": 1.4336,
+      "step": 563
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00019393318474603604,
+      "loss": 1.6064,
+      "step": 564
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00019391148832091033,
+      "loss": 1.1752,
+      "step": 565
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00019388975438717828,
+      "loss": 1.4882,
+      "step": 566
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0001938679829535205,
+      "loss": 1.6627,
+      "step": 567
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00019384617402863256,
+      "loss": 1.4215,
+      "step": 568
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00019382432762122504,
+      "loss": 1.3638,
+      "step": 569
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.0001938024437400235,
+      "loss": 1.6147,
+      "step": 570
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00019378052239376847,
+      "loss": 1.4148,
+      "step": 571
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00019375856359121535,
+      "loss": 1.6803,
+      "step": 572
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00019373656734113462,
+      "loss": 1.6736,
+      "step": 573
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00019371453365231163,
+      "loss": 1.4191,
+      "step": 574
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00019369246253354673,
+      "loss": 1.6516,
+      "step": 575
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.0001936703539936552,
+      "loss": 1.2879,
+      "step": 576
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00019364820804146734,
+      "loss": 1.3207,
+      "step": 577
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00019362602468582827,
+      "loss": 1.5588,
+      "step": 578
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00019360380393559806,
+      "loss": 1.4314,
+      "step": 579
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00019358154579965188,
+      "loss": 1.3858,
+      "step": 580
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00019355925028687965,
+      "loss": 1.432,
+      "step": 581
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00019353691740618624,
+      "loss": 1.8013,
+      "step": 582
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00019351454716649158,
+      "loss": 1.4057,
+      "step": 583
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.15625,
+      "learning_rate": 0.00019349213957673035,
+      "loss": 1.7005,
+      "step": 584
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.0001934696946458523,
+      "loss": 1.7312,
+      "step": 585
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00019344721238282193,
+      "loss": 1.6371,
+      "step": 586
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00019342469279661883,
+      "loss": 1.51,
+      "step": 587
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00019340213589623734,
+      "loss": 1.3423,
+      "step": 588
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0001933795416906868,
+      "loss": 1.6688,
+      "step": 589
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00019335691018899147,
+      "loss": 1.3107,
+      "step": 590
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00019333424140019037,
+      "loss": 1.5193,
+      "step": 591
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00019331153533333755,
+      "loss": 1.5691,
+      "step": 592
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.0001932887919975019,
+      "loss": 1.6378,
+      "step": 593
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00019326601140176716,
+      "loss": 1.4298,
+      "step": 594
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00019324319355523204,
+      "loss": 1.5007,
+      "step": 595
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00019322033846701003,
+      "loss": 1.7189,
+      "step": 596
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.0001931974461462296,
+      "loss": 1.2813,
+      "step": 597
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00019317451660203397,
+      "loss": 1.4609,
+      "step": 598
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00019315154984358128,
+      "loss": 1.4454,
+      "step": 599
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00019312854588004457,
+      "loss": 1.3361,
+      "step": 600
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00019310550472061172,
+      "loss": 1.3838,
+      "step": 601
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00019308242637448543,
+      "loss": 1.6563,
+      "step": 602
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00019305931085088332,
+      "loss": 1.4194,
+      "step": 603
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00019303615815903776,
+      "loss": 1.2311,
+      "step": 604
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00019301296830819608,
+      "loss": 1.2889,
+      "step": 605
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00019298974130762034,
+      "loss": 1.5105,
+      "step": 606
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00019296647716658756,
+      "loss": 1.2317,
+      "step": 607
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00019294317589438947,
+      "loss": 1.6059,
+      "step": 608
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.75,
+      "learning_rate": 0.0001929198375003327,
+      "loss": 1.4845,
+      "step": 609
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.0001928964619937387,
+      "loss": 1.5548,
+      "step": 610
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00019287304938394375,
+      "loss": 1.3618,
+      "step": 611
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.0001928495996802989,
+      "loss": 1.2621,
+      "step": 612
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00019282611289217009,
+      "loss": 1.455,
+      "step": 613
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00019280258902893799,
+      "loss": 1.3242,
+      "step": 614
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.0001927790280999982,
+      "loss": 1.4339,
+      "step": 615
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00019275543011476094,
+      "loss": 1.6739,
+      "step": 616
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0001927317950826514,
+      "loss": 1.8454,
+      "step": 617
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.0001927081230131095,
+      "loss": 1.4517,
+      "step": 618
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00019268441391558991,
+      "loss": 1.6463,
+      "step": 619
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.0001926606677995622,
+      "loss": 1.4474,
+      "step": 620
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00019263688467451059,
+      "loss": 1.6546,
+      "step": 621
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00019261306454993423,
+      "loss": 1.5611,
+      "step": 622
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.0001925892074353469,
+      "loss": 1.858,
+      "step": 623
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00019256531334027727,
+      "loss": 1.4844,
+      "step": 624
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00019254138227426865,
+      "loss": 1.3727,
+      "step": 625
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.0001925174142468793,
+      "loss": 1.6258,
+      "step": 626
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.0001924934092676821,
+      "loss": 1.3385,
+      "step": 627
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00019246936734626472,
+      "loss": 1.5926,
+      "step": 628
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00019244528849222964,
+      "loss": 1.5634,
+      "step": 629
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00019242117271519397,
+      "loss": 1.8999,
+      "step": 630
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001923970200247897,
+      "loss": 1.2813,
+      "step": 631
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.0001923728304306635,
+      "loss": 1.4927,
+      "step": 632
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00019234860394247677,
+      "loss": 1.6603,
+      "step": 633
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00019232434056990565,
+      "loss": 1.5586,
+      "step": 634
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00019230004032264107,
+      "loss": 1.6969,
+      "step": 635
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00019227570321038863,
+      "loss": 1.4574,
+      "step": 636
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00019225132924286862,
+      "loss": 1.3283,
+      "step": 637
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00019222691842981613,
+      "loss": 1.1849,
+      "step": 638
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.0001922024707809809,
+      "loss": 1.5663,
+      "step": 639
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00019217798630612748,
+      "loss": 1.6068,
+      "step": 640
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.000192153465015035,
+      "loss": 1.6785,
+      "step": 641
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00019212890691749741,
+      "loss": 1.6949,
+      "step": 642
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00019210431202332324,
+      "loss": 1.4438,
+      "step": 643
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00019207968034233586,
+      "loss": 1.8801,
+      "step": 644
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00019205501188437318,
+      "loss": 1.5233,
+      "step": 645
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00019203030665928793,
+      "loss": 1.5615,
+      "step": 646
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.75,
+      "learning_rate": 0.0001920055646769475,
+      "loss": 1.51,
+      "step": 647
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00019198078594723382,
+      "loss": 1.9251,
+      "step": 648
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.0001919559704800437,
+      "loss": 1.6288,
+      "step": 649
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.96875,
+      "learning_rate": 0.00019193111828528855,
+      "loss": 1.4211,
+      "step": 650
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00019190622937289436,
+      "loss": 1.6458,
+      "step": 651
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.0001918813037528019,
+      "loss": 1.6941,
+      "step": 652
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00019185634143496655,
+      "loss": 1.7412,
+      "step": 653
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00019183134242935833,
+      "loss": 1.5116,
+      "step": 654
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00019180630674596198,
+      "loss": 1.4274,
+      "step": 655
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.0001917812343947768,
+      "loss": 1.3951,
+      "step": 656
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00019175612538581684,
+      "loss": 1.5297,
+      "step": 657
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.0001917309797291107,
+      "loss": 1.8125,
+      "step": 658
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00019170579743470163,
+      "loss": 1.6461,
+      "step": 659
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00019168057851264754,
+      "loss": 1.202,
+      "step": 660
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00019165532297302104,
+      "loss": 1.5473,
+      "step": 661
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00019163003082590921,
+      "loss": 1.5754,
+      "step": 662
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00019160470208141384,
+      "loss": 1.5418,
+      "step": 663
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00019157933674965135,
+      "loss": 1.7034,
+      "step": 664
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00019155393484075273,
+      "loss": 1.4301,
+      "step": 665
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00019152849636486357,
+      "loss": 1.3163,
+      "step": 666
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.0001915030213321442,
+      "loss": 1.6925,
+      "step": 667
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00019147750975276933,
+      "loss": 1.3549,
+      "step": 668
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00019145196163692843,
+      "loss": 1.6268,
+      "step": 669
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00019142637699482554,
+      "loss": 1.5351,
+      "step": 670
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00019140075583667923,
+      "loss": 1.656,
+      "step": 671
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.0001913750981727227,
+      "loss": 1.4196,
+      "step": 672
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.0001913494040132037,
+      "loss": 1.3355,
+      "step": 673
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00019132367336838457,
+      "loss": 1.3584,
+      "step": 674
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00019129790624854227,
+      "loss": 1.6121,
+      "step": 675
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00019127210266396826,
+      "loss": 1.3768,
+      "step": 676
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.0001912462626249686,
+      "loss": 1.3339,
+      "step": 677
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00019122038614186387,
+      "loss": 1.7172,
+      "step": 678
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00019119447322498928,
+      "loss": 1.4404,
+      "step": 679
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00019116852388469452,
+      "loss": 1.7379,
+      "step": 680
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.00019114253813134387,
+      "loss": 1.2408,
+      "step": 681
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001911165159753161,
+      "loss": 1.5293,
+      "step": 682
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00019109045742700463,
+      "loss": 1.3647,
+      "step": 683
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00019106436249681727,
+      "loss": 1.5674,
+      "step": 684
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.0001910382311951765,
+      "loss": 1.3423,
+      "step": 685
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001910120635325192,
+      "loss": 1.6406,
+      "step": 686
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00019098585951929687,
+      "loss": 1.4642,
+      "step": 687
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.0001909596191659755,
+      "loss": 1.561,
+      "step": 688
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00019093334248303557,
+      "loss": 1.8744,
+      "step": 689
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00019090702948097206,
+      "loss": 1.3938,
+      "step": 690
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00019088068017029453,
+      "loss": 1.8793,
+      "step": 691
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.078125,
+      "learning_rate": 0.00019085429456152698,
+      "loss": 1.6933,
+      "step": 692
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.0001908278726652079,
+      "loss": 1.6767,
+      "step": 693
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00019080141449189037,
+      "loss": 1.5833,
+      "step": 694
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.0001907749200521418,
+      "loss": 1.671,
+      "step": 695
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.0001907483893565442,
+      "loss": 1.8103,
+      "step": 696
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00019072182241569406,
+      "loss": 1.547,
+      "step": 697
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00019069521924020225,
+      "loss": 1.5306,
+      "step": 698
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00019066857984069424,
+      "loss": 1.6267,
+      "step": 699
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.0001906419042278099,
+      "loss": 1.4324,
+      "step": 700
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00019061519241220354,
+      "loss": 1.687,
+      "step": 701
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.000190588444404544,
+      "loss": 1.6675,
+      "step": 702
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.0001905616602155145,
+      "loss": 1.7231,
+      "step": 703
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.0001905348398558128,
+      "loss": 1.4742,
+      "step": 704
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00019050798333615097,
+      "loss": 1.6397,
+      "step": 705
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00019048109066725565,
+      "loss": 1.4381,
+      "step": 706
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.0001904541618598679,
+      "loss": 1.0355,
+      "step": 707
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00019042719692474318,
+      "loss": 1.3453,
+      "step": 708
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00019040019587265137,
+      "loss": 1.6638,
+      "step": 709
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.0001903731587143768,
+      "loss": 1.7113,
+      "step": 710
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00019034608546071818,
+      "loss": 1.4676,
+      "step": 711
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00019031897612248875,
+      "loss": 1.6177,
+      "step": 712
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.000190291830710516,
+      "loss": 1.5941,
+      "step": 713
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00019026464923564197,
+      "loss": 1.3526,
+      "step": 714
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.000190237431708723,
+      "loss": 1.0494,
+      "step": 715
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.0001902101781406299,
+      "loss": 1.4564,
+      "step": 716
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00019018288854224787,
+      "loss": 1.593,
+      "step": 717
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.0001901555629244764,
+      "loss": 1.5527,
+      "step": 718
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00019012820129822954,
+      "loss": 1.6591,
+      "step": 719
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00019010080367443558,
+      "loss": 1.4237,
+      "step": 720
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00019007337006403727,
+      "loss": 1.5772,
+      "step": 721
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.0001900459004779916,
+      "loss": 1.6186,
+      "step": 722
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.0001900183949272702,
+      "loss": 1.423,
+      "step": 723
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.0001899908534228587,
+      "loss": 1.7121,
+      "step": 724
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00018996327597575743,
+      "loss": 1.5481,
+      "step": 725
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00018993566259698085,
+      "loss": 1.7453,
+      "step": 726
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00018990801329755788,
+      "loss": 1.5313,
+      "step": 727
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.0001898803280885317,
+      "loss": 1.6065,
+      "step": 728
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00018985260698096,
+      "loss": 1.6623,
+      "step": 729
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00018982484998591456,
+      "loss": 1.6831,
+      "step": 730
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00018979705711448172,
+      "loss": 1.5818,
+      "step": 731
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00018976922837776207,
+      "loss": 1.5864,
+      "step": 732
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00018974136378687044,
+      "loss": 1.852,
+      "step": 733
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00018971346335293614,
+      "loss": 1.5856,
+      "step": 734
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.0001896855270871026,
+      "loss": 1.431,
+      "step": 735
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00018965755500052777,
+      "loss": 1.3217,
+      "step": 736
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00018962954710438382,
+      "loss": 1.4862,
+      "step": 737
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00018960150340985708,
+      "loss": 1.6467,
+      "step": 738
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00018957342392814843,
+      "loss": 1.4346,
+      "step": 739
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.0001895453086704729,
+      "loss": 1.6372,
+      "step": 740
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.0001895171576480598,
+      "loss": 1.5022,
+      "step": 741
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00018948897087215275,
+      "loss": 1.5178,
+      "step": 742
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001894607483540097,
+      "loss": 1.5391,
+      "step": 743
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00018943249010490277,
+      "loss": 1.4266,
+      "step": 744
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.0001894041961361185,
+      "loss": 1.3838,
+      "step": 745
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00018937586645895754,
+      "loss": 1.6275,
+      "step": 746
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00018934750108473488,
+      "loss": 1.5298,
+      "step": 747
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00018931910002477977,
+      "loss": 1.4499,
+      "step": 748
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.609375,
+      "learning_rate": 0.0001892906632904357,
+      "loss": 1.1779,
+      "step": 749
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00018926219089306044,
+      "loss": 1.576,
+      "step": 750
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001892336828440259,
+      "loss": 1.5222,
+      "step": 751
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00018920513915471836,
+      "loss": 1.8117,
+      "step": 752
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00018917655983653829,
+      "loss": 1.571,
+      "step": 753
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00018914794490090033,
+      "loss": 1.6409,
+      "step": 754
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00018911929435923343,
+      "loss": 1.4455,
+      "step": 755
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.0001890906082229807,
+      "loss": 1.5237,
+      "step": 756
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00018906188650359953,
+      "loss": 1.6608,
+      "step": 757
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00018903312921256143,
+      "loss": 1.6695,
+      "step": 758
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00018900433636135223,
+      "loss": 1.4211,
+      "step": 759
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00018897550796147185,
+      "loss": 1.6753,
+      "step": 760
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.0001889466440244345,
+      "loss": 1.4361,
+      "step": 761
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00018891774456176856,
+      "loss": 1.6179,
+      "step": 762
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00018888880958501652,
+      "loss": 1.527,
+      "step": 763
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.0001888598391057352,
+      "loss": 1.5128,
+      "step": 764
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00018883083313549548,
+      "loss": 1.5217,
+      "step": 765
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00018880179168588243,
+      "loss": 1.3351,
+      "step": 766
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00018877271476849537,
+      "loss": 1.6041,
+      "step": 767
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00018874360239494768,
+      "loss": 1.4351,
+      "step": 768
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.000188714454576867,
+      "loss": 1.5158,
+      "step": 769
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.000188685271325895,
+      "loss": 1.6849,
+      "step": 770
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00018865605265368772,
+      "loss": 1.6695,
+      "step": 771
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00018862679857191506,
+      "loss": 1.7458,
+      "step": 772
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.0001885975090922613,
+      "loss": 1.4367,
+      "step": 773
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00018856818422642477,
+      "loss": 1.6887,
+      "step": 774
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.0001885388239861179,
+      "loss": 1.3208,
+      "step": 775
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00018850942838306728,
+      "loss": 1.572,
+      "step": 776
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00018847999742901365,
+      "loss": 1.7115,
+      "step": 777
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00018845053113571183,
+      "loss": 1.4649,
+      "step": 778
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00018842102951493076,
+      "loss": 1.7504,
+      "step": 779
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00018839149257845355,
+      "loss": 1.568,
+      "step": 780
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001883619203380773,
+      "loss": 1.615,
+      "step": 781
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00018833231280561332,
+      "loss": 1.7534,
+      "step": 782
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00018830266999288694,
+      "loss": 1.6697,
+      "step": 783
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00018827299191173763,
+      "loss": 1.5412,
+      "step": 784
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00018824327857401896,
+      "loss": 1.3499,
+      "step": 785
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.0001882135299915985,
+      "loss": 1.3235,
+      "step": 786
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00018818374617635795,
+      "loss": 1.5829,
+      "step": 787
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00018815392714019308,
+      "loss": 1.5437,
+      "step": 788
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00018812407289501377,
+      "loss": 1.3832,
+      "step": 789
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001880941834527439,
+      "loss": 1.4602,
+      "step": 790
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00018806425882532136,
+      "loss": 1.5213,
+      "step": 791
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.0001880342990246983,
+      "loss": 1.4885,
+      "step": 792
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00018800430406284063,
+      "loss": 1.381,
+      "step": 793
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00018797427395172856,
+      "loss": 1.71,
+      "step": 794
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00018794420870335614,
+      "loss": 1.6483,
+      "step": 795
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00018791410832973165,
+      "loss": 1.5048,
+      "step": 796
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001878839728428772,
+      "loss": 1.6351,
+      "step": 797
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00018785380225482906,
+      "loss": 1.2019,
+      "step": 798
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.0001878235965776375,
+      "loss": 1.5804,
+      "step": 799
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.0001877933558233667,
+      "loss": 1.5165,
+      "step": 800
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00018776308000409504,
+      "loss": 1.6228,
+      "step": 801
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00018773276913191476,
+      "loss": 1.4395,
+      "step": 802
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00018770242321893208,
+      "loss": 1.4832,
+      "step": 803
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00018767204227726733,
+      "loss": 1.6444,
+      "step": 804
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00018764162631905475,
+      "loss": 1.5267,
+      "step": 805
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00018761117535644263,
+      "loss": 1.9495,
+      "step": 806
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00018758068940159315,
+      "loss": 1.5612,
+      "step": 807
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00018755016846668256,
+      "loss": 1.4715,
+      "step": 808
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.000187519612563901,
+      "loss": 1.5836,
+      "step": 809
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00018748902170545264,
+      "loss": 1.4719,
+      "step": 810
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00018745839590355556,
+      "loss": 1.5917,
+      "step": 811
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00018742773517044182,
+      "loss": 1.6619,
+      "step": 812
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.0001873970395183575,
+      "loss": 1.4066,
+      "step": 813
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00018736630895956248,
+      "loss": 1.776,
+      "step": 814
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00018733554350633072,
+      "loss": 1.4729,
+      "step": 815
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00018730474317095,
+      "loss": 1.4835,
+      "step": 816
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00018727390796572215,
+      "loss": 1.5465,
+      "step": 817
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00018724303790296283,
+      "loss": 1.5244,
+      "step": 818
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00018721213299500167,
+      "loss": 1.5287,
+      "step": 819
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.0001871811932541822,
+      "loss": 1.7061,
+      "step": 820
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00018715021869286194,
+      "loss": 1.5598,
+      "step": 821
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00018711920932341211,
+      "loss": 1.8095,
+      "step": 822
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.0001870881651582181,
+      "loss": 1.8539,
+      "step": 823
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00018705708620967904,
+      "loss": 1.5093,
+      "step": 824
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001870259724902079,
+      "loss": 1.5852,
+      "step": 825
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00018699482401223172,
+      "loss": 1.3453,
+      "step": 826
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00018696364078819123,
+      "loss": 1.6365,
+      "step": 827
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00018693242283054123,
+      "loss": 1.51,
+      "step": 828
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5,
+      "learning_rate": 0.0001869011701517502,
+      "loss": 1.3306,
+      "step": 829
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00018686988276430064,
+      "loss": 1.2841,
+      "step": 830
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00018683856068068882,
+      "loss": 1.536,
+      "step": 831
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.0001868072039134249,
+      "loss": 1.8682,
+      "step": 832
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00018677581247503292,
+      "loss": 1.8636,
+      "step": 833
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.0001867443863780507,
+      "loss": 1.5747,
+      "step": 834
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00018671292563502995,
+      "loss": 1.4474,
+      "step": 835
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00018668143025853625,
+      "loss": 1.5834,
+      "step": 836
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00018664990026114893,
+      "loss": 1.6901,
+      "step": 837
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00018661833565546124,
+      "loss": 1.4891,
+      "step": 838
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00018658673645408012,
+      "loss": 1.4995,
+      "step": 839
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00018655510266962652,
+      "loss": 1.5482,
+      "step": 840
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.000186523434314735,
+      "loss": 1.7702,
+      "step": 841
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00018649173140205408,
+      "loss": 1.4902,
+      "step": 842
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.375,
+      "learning_rate": 0.000186459993944246,
+      "loss": 1.5365,
+      "step": 843
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.0001864282219539868,
+      "loss": 1.5006,
+      "step": 844
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.0001863964154439664,
+      "loss": 1.952,
+      "step": 845
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00018636457442688836,
+      "loss": 1.6504,
+      "step": 846
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00018633269891547017,
+      "loss": 1.6822,
+      "step": 847
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00018630078892244297,
+      "loss": 1.4663,
+      "step": 848
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00018626884446055182,
+      "loss": 1.8359,
+      "step": 849
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00018623686554255538,
+      "loss": 1.3008,
+      "step": 850
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00018620485218122617,
+      "loss": 1.4291,
+      "step": 851
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00018617280438935048,
+      "loss": 1.5415,
+      "step": 852
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00018614072217972827,
+      "loss": 1.4369,
+      "step": 853
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00018610860556517335,
+      "loss": 1.3853,
+      "step": 854
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.25,
+      "learning_rate": 0.0001860764545585132,
+      "loss": 1.5723,
+      "step": 855
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.000186044269172589,
+      "loss": 1.6622,
+      "step": 856
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.0001860120494202558,
+      "loss": 1.6586,
+      "step": 857
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.90625,
+      "learning_rate": 0.00018597979531438227,
+      "loss": 1.5474,
+      "step": 858
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00018594750686785075,
+      "loss": 1.4239,
+      "step": 859
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.0001859151840935575,
+      "loss": 1.2553,
+      "step": 860
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00018588282700441225,
+      "loss": 1.6793,
+      "step": 861
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00018585043561333858,
+      "loss": 1.5874,
+      "step": 862
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.15625,
+      "learning_rate": 0.00018581800993327372,
+      "loss": 1.682,
+      "step": 863
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00018578554997716865,
+      "loss": 1.4286,
+      "step": 864
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.1875,
+      "learning_rate": 0.00018575305575798797,
+      "loss": 1.7333,
+      "step": 865
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00018572052728870999,
+      "loss": 1.467,
+      "step": 866
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00018568796458232672,
+      "loss": 1.2283,
+      "step": 867
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00018565536765184382,
+      "loss": 1.506,
+      "step": 868
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00018562273651028064,
+      "loss": 1.524,
+      "step": 869
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00018559007117067018,
+      "loss": 1.4743,
+      "step": 870
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.0001855573716460591,
+      "loss": 1.4622,
+      "step": 871
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.0001855246379495077,
+      "loss": 1.3976,
+      "step": 872
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00018549187009408996,
+      "loss": 1.4853,
+      "step": 873
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00018545906809289345,
+      "loss": 1.7967,
+      "step": 874
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00018542623195901952,
+      "loss": 1.6207,
+      "step": 875
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.0001853933617055829,
+      "loss": 1.5086,
+      "step": 876
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.0001853604573457122,
+      "loss": 1.5612,
+      "step": 877
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001853275188925495,
+      "loss": 1.5157,
+      "step": 878
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00018529454635925053,
+      "loss": 1.5992,
+      "step": 879
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00018526153975898473,
+      "loss": 1.7258,
+      "step": 880
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.000185228499104935,
+      "loss": 1.5205,
+      "step": 881
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00018519542441029784,
+      "loss": 1.4877,
+      "step": 882
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00018516231568828354,
+      "loss": 1.3097,
+      "step": 883
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00018512917295211576,
+      "loss": 1.506,
+      "step": 884
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00018509599621503186,
+      "loss": 1.5352,
+      "step": 885
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00018506278549028276,
+      "loss": 1.6606,
+      "step": 886
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00018502954079113298,
+      "loss": 1.2368,
+      "step": 887
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001849962621308605,
+      "loss": 1.4378,
+      "step": 888
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.000184962949522757,
+      "loss": 1.6312,
+      "step": 889
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00018492960298012765,
+      "loss": 1.4748,
+      "step": 890
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.0001848962225162912,
+      "loss": 1.4986,
+      "step": 891
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00018486280814457992,
+      "loss": 1.6099,
+      "step": 892
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00018482935987833965,
+      "loss": 1.5754,
+      "step": 893
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00018479587773092974,
+      "loss": 1.2705,
+      "step": 894
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00018476236171572307,
+      "loss": 1.6246,
+      "step": 895
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00018472881184610613,
+      "loss": 1.5836,
+      "step": 896
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001846952281354788,
+      "loss": 1.505,
+      "step": 897
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00018466161059725456,
+      "loss": 1.4927,
+      "step": 898
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.0001846279592448604,
+      "loss": 1.7722,
+      "step": 899
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00018459427409173675,
+      "loss": 1.2283,
+      "step": 900
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00018456055515133767,
+      "loss": 1.5066,
+      "step": 901
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00018452680243713057,
+      "loss": 1.6962,
+      "step": 902
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00018449301596259644,
+      "loss": 1.372,
+      "step": 903
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.0001844591957412297,
+      "loss": 1.3442,
+      "step": 904
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.0001844253417865383,
+      "loss": 1.4349,
+      "step": 905
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00018439145411204363,
+      "loss": 1.514,
+      "step": 906
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00018435753273128057,
+      "loss": 1.4975,
+      "step": 907
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.0001843235776577974,
+      "loss": 1.5082,
+      "step": 908
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00018428958890515597,
+      "loss": 1.4875,
+      "step": 909
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.0001842555664869315,
+      "loss": 1.5782,
+      "step": 910
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00018422151041671265,
+      "loss": 1.3216,
+      "step": 911
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00018418742070810157,
+      "loss": 1.4147,
+      "step": 912
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00018415329737471377,
+      "loss": 1.4112,
+      "step": 913
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.0001841191404301783,
+      "loss": 1.4666,
+      "step": 914
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00018408494988813757,
+      "loss": 1.1704,
+      "step": 915
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00018405072576224737,
+      "loss": 1.4846,
+      "step": 916
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00018401646806617697,
+      "loss": 1.3152,
+      "step": 917
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00018398217681360902,
+      "loss": 1.3955,
+      "step": 918
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00018394785201823958,
+      "loss": 1.6733,
+      "step": 919
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00018391349369377812,
+      "loss": 1.6491,
+      "step": 920
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00018387910185394744,
+      "loss": 1.3877,
+      "step": 921
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00018384467651248385,
+      "loss": 1.4019,
+      "step": 922
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.0001838102176831369,
+      "loss": 1.5744,
+      "step": 923
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00018377572537966956,
+      "loss": 1.4349,
+      "step": 924
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00018374119961585823,
+      "loss": 1.5801,
+      "step": 925
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00018370664040549264,
+      "loss": 1.6586,
+      "step": 926
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.375,
+      "learning_rate": 0.00018367204776237585,
+      "loss": 1.4446,
+      "step": 927
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00018363742170032427,
+      "loss": 1.6023,
+      "step": 928
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00018360276223316775,
+      "loss": 1.6318,
+      "step": 929
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00018356806937474933,
+      "loss": 1.2924,
+      "step": 930
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00018353334313892555,
+      "loss": 1.4084,
+      "step": 931
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00018349858353956613,
+      "loss": 1.4841,
+      "step": 932
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.0001834637905905542,
+      "loss": 1.471,
+      "step": 933
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00018342896430578623,
+      "loss": 1.6634,
+      "step": 934
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00018339410469917197,
+      "loss": 1.2712,
+      "step": 935
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00018335921178463444,
+      "loss": 1.549,
+      "step": 936
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00018332428557611005,
+      "loss": 1.5683,
+      "step": 937
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00018328932608754843,
+      "loss": 1.407,
+      "step": 938
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00018325433333291256,
+      "loss": 1.4584,
+      "step": 939
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00018321930732617862,
+      "loss": 1.6057,
+      "step": 940
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.0001831842480813362,
+      "loss": 1.479,
+      "step": 941
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00018314915561238806,
+      "loss": 1.3984,
+      "step": 942
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00018311402993335024,
+      "loss": 1.321,
+      "step": 943
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00018307887105825215,
+      "loss": 1.4712,
+      "step": 944
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00018304367900113632,
+      "loss": 1.6424,
+      "step": 945
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.0001830084537760586,
+      "loss": 1.4925,
+      "step": 946
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.0001829731953970881,
+      "loss": 1.4025,
+      "step": 947
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.625,
+      "learning_rate": 0.0001829379038783071,
+      "loss": 1.4028,
+      "step": 948
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00018290257923381123,
+      "loss": 1.7876,
+      "step": 949
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00018286722147770922,
+      "loss": 1.8309,
+      "step": 950
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00018283183062412316,
+      "loss": 1.5782,
+      "step": 951
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00018279640668718828,
+      "loss": 1.6489,
+      "step": 952
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.000182760949681053,
+      "loss": 1.3551,
+      "step": 953
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.000182725459619879,
+      "loss": 1.5301,
+      "step": 954
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00018268993651784113,
+      "loss": 1.4967,
+      "step": 955
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00018265438038912748,
+      "loss": 1.6227,
+      "step": 956
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.0001826187912479393,
+      "loss": 1.6373,
+      "step": 957
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00018258316910849101,
+      "loss": 1.5191,
+      "step": 958
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00018254751398501026,
+      "loss": 1.4573,
+      "step": 959
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.0001825118258917378,
+      "loss": 1.5131,
+      "step": 960
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00018247610484292763,
+      "loss": 1.6604,
+      "step": 961
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00018244035085284685,
+      "loss": 1.5453,
+      "step": 962
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00018240456393577573,
+      "loss": 1.6494,
+      "step": 963
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00018236874410600774,
+      "loss": 1.3551,
+      "step": 964
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00018233289137784944,
+      "loss": 1.6789,
+      "step": 965
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.0001822970057656205,
+      "loss": 1.6761,
+      "step": 966
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00018226108728365382,
+      "loss": 1.3441,
+      "step": 967
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00018222513594629537,
+      "loss": 1.7487,
+      "step": 968
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.0001821891517679043,
+      "loss": 1.3922,
+      "step": 969
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.0001821531347628527,
+      "loss": 1.5552,
+      "step": 970
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00018211708494552604,
+      "loss": 1.7095,
+      "step": 971
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00018208100233032263,
+      "loss": 1.2853,
+      "step": 972
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00018204488693165412,
+      "loss": 1.5525,
+      "step": 973
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00018200873876394503,
+      "loss": 1.4181,
+      "step": 974
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00018197255784163318,
+      "loss": 1.6438,
+      "step": 975
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.0001819363441791693,
+      "loss": 1.2702,
+      "step": 976
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00018190009779101728,
+      "loss": 1.5192,
+      "step": 977
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00018186381869165406,
+      "loss": 1.7537,
+      "step": 978
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00018182750689556967,
+      "loss": 1.6456,
+      "step": 979
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00018179116241726716,
+      "loss": 1.7041,
+      "step": 980
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00018175478527126263,
+      "loss": 1.5447,
+      "step": 981
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00018171837547208532,
+      "loss": 1.6306,
+      "step": 982
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0001816819330342774,
+      "loss": 1.6013,
+      "step": 983
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.0001816454579723941,
+      "loss": 1.4876,
+      "step": 984
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00018160895030100375,
+      "loss": 1.5811,
+      "step": 985
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00018157241003468762,
+      "loss": 1.6152,
+      "step": 986
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00018153583718804,
+      "loss": 1.6625,
+      "step": 987
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.0001814992317756683,
+      "loss": 1.3254,
+      "step": 988
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00018146259381219282,
+      "loss": 1.6693,
+      "step": 989
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00018142592331224687,
+      "loss": 1.4896,
+      "step": 990
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00018138922029047687,
+      "loss": 1.75,
+      "step": 991
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00018135248476154207,
+      "loss": 1.5815,
+      "step": 992
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.0001813157167401148,
+      "loss": 1.5452,
+      "step": 993
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.0001812789162408804,
+      "loss": 1.3197,
+      "step": 994
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00018124208327853706,
+      "loss": 1.6673,
+      "step": 995
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.000181205217867796,
+      "loss": 1.1366,
+      "step": 996
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00018116832002338144,
+      "loss": 1.5062,
+      "step": 997
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00018113138976003055,
+      "loss": 1.7284,
+      "step": 998
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00018109442709249337,
+      "loss": 1.648,
+      "step": 999
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00018105743203553293,
+      "loss": 1.4417,
+      "step": 1000
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00018102040460392523,
+      "loss": 1.6242,
+      "step": 1001
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00018098334481245914,
+      "loss": 1.2839,
+      "step": 1002
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0001809462526759365,
+      "loss": 1.5728,
+      "step": 1003
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00018090912820917205,
+      "loss": 1.2653,
+      "step": 1004
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00018087197142699344,
+      "loss": 1.578,
+      "step": 1005
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.0001808347823442412,
+      "loss": 1.7955,
+      "step": 1006
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00018079756097576883,
+      "loss": 1.6464,
+      "step": 1007
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00018076030733644264,
+      "loss": 1.4645,
+      "step": 1008
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00018072302144114195,
+      "loss": 1.4759,
+      "step": 1009
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00018068570330475884,
+      "loss": 1.5497,
+      "step": 1010
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00018064835294219833,
+      "loss": 1.4924,
+      "step": 1011
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00018061097036837825,
+      "loss": 1.2192,
+      "step": 1012
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00018057355559822943,
+      "loss": 1.5788,
+      "step": 1013
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00018053610864669542,
+      "loss": 1.7215,
+      "step": 1014
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00018049862952873266,
+      "loss": 1.5349,
+      "step": 1015
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00018046111825931047,
+      "loss": 1.4214,
+      "step": 1016
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.000180423574853411,
+      "loss": 1.5651,
+      "step": 1017
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00018038599932602924,
+      "loss": 1.3974,
+      "step": 1018
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00018034839169217296,
+      "loss": 1.4884,
+      "step": 1019
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00018031075196686284,
+      "loss": 1.3973,
+      "step": 1020
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001802730801651323,
+      "loss": 1.4314,
+      "step": 1021
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.0001802353763020276,
+      "loss": 1.3261,
+      "step": 1022
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00018019764039260783,
+      "loss": 1.4372,
+      "step": 1023
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001801598724519448,
+      "loss": 1.353,
+      "step": 1024
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00018012207249512325,
+      "loss": 1.519,
+      "step": 1025
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00018008424053724058,
+      "loss": 1.5834,
+      "step": 1026
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00018004637659340702,
+      "loss": 1.6489,
+      "step": 1027
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.0001800084806787456,
+      "loss": 1.6221,
+      "step": 1028
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00017997055280839203,
+      "loss": 1.6973,
+      "step": 1029
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.0001799325929974949,
+      "loss": 1.3166,
+      "step": 1030
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00017989460126121548,
+      "loss": 1.6559,
+      "step": 1031
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00017985657761472788,
+      "loss": 1.7162,
+      "step": 1032
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.0001798185220732188,
+      "loss": 1.6973,
+      "step": 1033
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.0001797804346518878,
+      "loss": 1.5865,
+      "step": 1034
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00017974231536594713,
+      "loss": 1.5089,
+      "step": 1035
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.0001797041642306218,
+      "loss": 1.5918,
+      "step": 1036
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5,
+      "learning_rate": 0.0001796659812611495,
+      "loss": 1.5373,
+      "step": 1037
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.0001796277664727807,
+      "loss": 1.5773,
+      "step": 1038
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00017958951988077845,
+      "loss": 1.2739,
+      "step": 1039
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00017955124150041866,
+      "loss": 1.2865,
+      "step": 1040
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00017951293134698984,
+      "loss": 1.4585,
+      "step": 1041
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00017947458943579317,
+      "loss": 1.9428,
+      "step": 1042
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.625,
+      "learning_rate": 0.0001794362157821426,
+      "loss": 1.5048,
+      "step": 1043
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00017939781040136473,
+      "loss": 1.5543,
+      "step": 1044
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00017935937330879875,
+      "loss": 1.4926,
+      "step": 1045
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00017932090451979664,
+      "loss": 1.747,
+      "step": 1046
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00017928240404972297,
+      "loss": 1.3404,
+      "step": 1047
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00017924387191395493,
+      "loss": 1.6345,
+      "step": 1048
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00017920530812788246,
+      "loss": 1.5924,
+      "step": 1049
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00017916671270690803,
+      "loss": 1.6059,
+      "step": 1050
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00017912808566644686,
+      "loss": 1.4872,
+      "step": 1051
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.00017908942702192667,
+      "loss": 1.4223,
+      "step": 1052
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00017905073678878795,
+      "loss": 1.4989,
+      "step": 1053
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00017901201498248366,
+      "loss": 1.7193,
+      "step": 1054
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00017897326161847945,
+      "loss": 1.4194,
+      "step": 1055
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00017893447671225355,
+      "loss": 1.4403,
+      "step": 1056
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00017889566027929688,
+      "loss": 1.5915,
+      "step": 1057
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00017885681233511275,
+      "loss": 1.7215,
+      "step": 1058
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00017881793289521727,
+      "loss": 1.4276,
+      "step": 1059
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00017877902197513897,
+      "loss": 1.6739,
+      "step": 1060
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.0001787400795904191,
+      "loss": 1.425,
+      "step": 1061
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.0001787011057566113,
+      "loss": 1.5379,
+      "step": 1062
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00017866210048928197,
+      "loss": 1.4621,
+      "step": 1063
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.875,
+      "learning_rate": 0.0001786230638040099,
+      "loss": 1.2647,
+      "step": 1064
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00017858399571638647,
+      "loss": 1.6603,
+      "step": 1065
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00017854489624201568,
+      "loss": 1.7452,
+      "step": 1066
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00017850576539651396,
+      "loss": 1.6322,
+      "step": 1067
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00017846660319551037,
+      "loss": 1.6292,
+      "step": 1068
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00017842740965464638,
+      "loss": 1.7689,
+      "step": 1069
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00017838818478957608,
+      "loss": 1.3905,
+      "step": 1070
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00017834892861596604,
+      "loss": 1.4716,
+      "step": 1071
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00017830964114949531,
+      "loss": 1.506,
+      "step": 1072
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00017827032240585544,
+      "loss": 1.3947,
+      "step": 1073
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.0001782309724007505,
+      "loss": 1.5937,
+      "step": 1074
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00017819159114989706,
+      "loss": 1.4839,
+      "step": 1075
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00017815217866902407,
+      "loss": 1.3029,
+      "step": 1076
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.625,
+      "learning_rate": 0.0001781127349738731,
+      "loss": 1.8088,
+      "step": 1077
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00017807326008019807,
+      "loss": 1.6567,
+      "step": 1078
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.03125,
+      "learning_rate": 0.00017803375400376542,
+      "loss": 1.3267,
+      "step": 1079
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00017799421676035403,
+      "loss": 1.614,
+      "step": 1080
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.0001779546483657552,
+      "loss": 1.4496,
+      "step": 1081
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00017791504883577273,
+      "loss": 1.5657,
+      "step": 1082
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.00017787541818622277,
+      "loss": 1.5104,
+      "step": 1083
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00017783575643293405,
+      "loss": 1.2265,
+      "step": 1084
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00017779606359174749,
+      "loss": 1.3501,
+      "step": 1085
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.0001777563396785167,
+      "loss": 1.4533,
+      "step": 1086
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00017771658470910748,
+      "loss": 1.5147,
+      "step": 1087
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00017767679869939813,
+      "loss": 1.5183,
+      "step": 1088
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00017763698166527934,
+      "loss": 1.4717,
+      "step": 1089
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.0001775971336226542,
+      "loss": 1.6447,
+      "step": 1090
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00017755725458743813,
+      "loss": 1.8011,
+      "step": 1091
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00017751734457555903,
+      "loss": 1.4166,
+      "step": 1092
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00017747740360295704,
+      "loss": 1.2428,
+      "step": 1093
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00017743743168558478,
+      "loss": 1.152,
+      "step": 1094
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.0001773974288394072,
+      "loss": 1.2607,
+      "step": 1095
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00017735739508040156,
+      "loss": 1.4119,
+      "step": 1096
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.75,
+      "learning_rate": 0.0001773173304245575,
+      "loss": 1.2903,
+      "step": 1097
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.000177277234887877,
+      "loss": 1.3044,
+      "step": 1098
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00017723710848637433,
+      "loss": 1.5075,
+      "step": 1099
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.0001771969512360762,
+      "loss": 1.5784,
+      "step": 1100
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00017715676315302155,
+      "loss": 1.5476,
+      "step": 1101
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.0001771165442532616,
+      "loss": 1.5822,
+      "step": 1102
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00017707629455286,
+      "loss": 1.7717,
+      "step": 1103
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00017703601406789262,
+      "loss": 1.603,
+      "step": 1104
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00017699570281444759,
+      "loss": 1.6665,
+      "step": 1105
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00017695536080862542,
+      "loss": 1.2804,
+      "step": 1106
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00017691498806653885,
+      "loss": 1.27,
+      "step": 1107
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00017687458460431292,
+      "loss": 1.4948,
+      "step": 1108
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.0001768341504380849,
+      "loss": 1.5446,
+      "step": 1109
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.0001767936855840044,
+      "loss": 1.4639,
+      "step": 1110
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.00017675319005823318,
+      "loss": 1.5584,
+      "step": 1111
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00017671266387694537,
+      "loss": 1.4615,
+      "step": 1112
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.0001766721070563272,
+      "loss": 1.5837,
+      "step": 1113
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00017663151961257726,
+      "loss": 1.5212,
+      "step": 1114
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00017659090156190634,
+      "loss": 1.445,
+      "step": 1115
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00017655025292053744,
+      "loss": 1.4098,
+      "step": 1116
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00017650957370470577,
+      "loss": 1.869,
+      "step": 1117
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00017646886393065877,
+      "loss": 1.6403,
+      "step": 1118
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001764281236146561,
+      "loss": 1.8271,
+      "step": 1119
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00017638735277296957,
+      "loss": 1.7442,
+      "step": 1120
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00017634655142188322,
+      "loss": 1.4144,
+      "step": 1121
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00017630571957769325,
+      "loss": 1.4032,
+      "step": 1122
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00017626485725670806,
+      "loss": 1.648,
+      "step": 1123
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00017622396447524824,
+      "loss": 1.6452,
+      "step": 1124
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.0001761830412496465,
+      "loss": 1.704,
+      "step": 1125
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00017614208759624776,
+      "loss": 1.4751,
+      "step": 1126
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00017610110353140902,
+      "loss": 1.7576,
+      "step": 1127
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.375,
+      "learning_rate": 0.0001760600890714995,
+      "loss": 1.6408,
+      "step": 1128
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00017601904423290051,
+      "loss": 1.7851,
+      "step": 1129
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00017597796903200553,
+      "loss": 1.7309,
+      "step": 1130
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.0001759368634852202,
+      "loss": 1.5621,
+      "step": 1131
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00017589572760896214,
+      "loss": 1.5213,
+      "step": 1132
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00017585456141966125,
+      "loss": 1.5144,
+      "step": 1133
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001758133649337594,
+      "loss": 1.4575,
+      "step": 1134
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.0001757721381677107,
+      "loss": 1.5778,
+      "step": 1135
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.0001757308811379812,
+      "loss": 1.7309,
+      "step": 1136
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00017568959386104918,
+      "loss": 1.5829,
+      "step": 1137
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.0001756482763534049,
+      "loss": 1.3543,
+      "step": 1138
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00017560692863155072,
+      "loss": 1.3617,
+      "step": 1139
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00017556555071200112,
+      "loss": 1.9035,
+      "step": 1140
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00017552414261128255,
+      "loss": 1.5017,
+      "step": 1141
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.0001754827043459336,
+      "loss": 1.5304,
+      "step": 1142
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.90625,
+      "learning_rate": 0.00017544123593250485,
+      "loss": 1.5029,
+      "step": 1143
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00017539973738755894,
+      "loss": 1.3886,
+      "step": 1144
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00017535820872767056,
+      "loss": 1.5139,
+      "step": 1145
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.0001753166499694264,
+      "loss": 1.6079,
+      "step": 1146
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00017527506112942524,
+      "loss": 1.6762,
+      "step": 1147
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00017523344222427774,
+      "loss": 1.6132,
+      "step": 1148
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001751917932706067,
+      "loss": 1.5876,
+      "step": 1149
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00017515011428504687,
+      "loss": 1.5411,
+      "step": 1150
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00017510840528424496,
+      "loss": 1.5126,
+      "step": 1151
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00017506666628485973,
+      "loss": 1.5335,
+      "step": 1152
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00017502489730356193,
+      "loss": 1.4996,
+      "step": 1153
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001749830983570342,
+      "loss": 1.5372,
+      "step": 1154
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00017494126946197117,
+      "loss": 1.7249,
+      "step": 1155
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00017489941063507955,
+      "loss": 1.5656,
+      "step": 1156
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00017485752189307788,
+      "loss": 1.5422,
+      "step": 1157
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00017481560325269664,
+      "loss": 1.4538,
+      "step": 1158
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00017477365473067834,
+      "loss": 1.5257,
+      "step": 1159
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.0001747316763437773,
+      "loss": 1.4644,
+      "step": 1160
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00017468966810876003,
+      "loss": 1.4836,
+      "step": 1161
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00017464763004240456,
+      "loss": 1.6336,
+      "step": 1162
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.0001746055621615012,
+      "loss": 1.7933,
+      "step": 1163
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.000174563464482852,
+      "loss": 1.4033,
+      "step": 1164
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00017452133702327088,
+      "loss": 1.685,
+      "step": 1165
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00017447917979958376,
+      "loss": 1.5171,
+      "step": 1166
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00017443699282862835,
+      "loss": 1.363,
+      "step": 1167
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00017439477612725431,
+      "loss": 1.5724,
+      "step": 1168
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.890625,
+      "learning_rate": 0.0001743525297123232,
+      "loss": 1.519,
+      "step": 1169
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.0001743102536007083,
+      "loss": 1.3842,
+      "step": 1170
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00017426794780929495,
+      "loss": 1.7146,
+      "step": 1171
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00017422561235498017,
+      "loss": 1.4422,
+      "step": 1172
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.0001741832472546729,
+      "loss": 1.3318,
+      "step": 1173
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00017414085252529402,
+      "loss": 1.5796,
+      "step": 1174
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00017409842818377603,
+      "loss": 1.654,
+      "step": 1175
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00017405597424706345,
+      "loss": 1.5249,
+      "step": 1176
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.0001740134907321125,
+      "loss": 1.458,
+      "step": 1177
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00017397097765589126,
+      "loss": 1.787,
+      "step": 1178
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00017392843503537963,
+      "loss": 1.762,
+      "step": 1179
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00017388586288756932,
+      "loss": 1.4903,
+      "step": 1180
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00017384326122946378,
+      "loss": 1.3133,
+      "step": 1181
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00017380063007807825,
+      "loss": 1.4645,
+      "step": 1182
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00017375796945043982,
+      "loss": 1.633,
+      "step": 1183
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00017371527936358726,
+      "loss": 1.5572,
+      "step": 1184
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.0001736725598345712,
+      "loss": 1.6462,
+      "step": 1185
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00017362981088045397,
+      "loss": 1.5977,
+      "step": 1186
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00017358703251830964,
+      "loss": 1.4993,
+      "step": 1187
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.875,
+      "learning_rate": 0.0001735442247652241,
+      "loss": 1.6022,
+      "step": 1188
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00017350138763829487,
+      "loss": 1.5355,
+      "step": 1189
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.0001734585211546313,
+      "loss": 1.3328,
+      "step": 1190
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001734156253313544,
+      "loss": 1.5945,
+      "step": 1191
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00017337270018559698,
+      "loss": 1.5426,
+      "step": 1192
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00017332974573450348,
+      "loss": 1.43,
+      "step": 1193
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00017328676199523,
+      "loss": 1.813,
+      "step": 1194
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00017324374898494452,
+      "loss": 1.7011,
+      "step": 1195
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00017320070672082654,
+      "loss": 1.3096,
+      "step": 1196
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.0001731576352200673,
+      "loss": 1.7287,
+      "step": 1197
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00017311453449986975,
+      "loss": 1.3839,
+      "step": 1198
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00017307140457744846,
+      "loss": 1.319,
+      "step": 1199
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.0001730282454700297,
+      "loss": 1.5967,
+      "step": 1200
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00017298505719485138,
+      "loss": 1.738,
+      "step": 1201
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00017294183976916306,
+      "loss": 1.3976,
+      "step": 1202
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00017289859321022597,
+      "loss": 1.5968,
+      "step": 1203
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00017285531753531287,
+      "loss": 1.2877,
+      "step": 1204
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00017281201276170833,
+      "loss": 1.7717,
+      "step": 1205
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00017276867890670837,
+      "loss": 1.3279,
+      "step": 1206
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.0001727253159876207,
+      "loss": 1.5187,
+      "step": 1207
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00017268192402176467,
+      "loss": 1.5414,
+      "step": 1208
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00017263850302647117,
+      "loss": 1.5728,
+      "step": 1209
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00017259505301908277,
+      "loss": 1.4833,
+      "step": 1210
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00017255157401695347,
+      "loss": 1.6154,
+      "step": 1211
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00017250806603744904,
+      "loss": 1.4862,
+      "step": 1212
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00017246452909794668,
+      "loss": 1.6413,
+      "step": 1213
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00017242096321583522,
+      "loss": 1.2644,
+      "step": 1214
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00017237736840851505,
+      "loss": 1.4739,
+      "step": 1215
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00017233374469339808,
+      "loss": 1.6745,
+      "step": 1216
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00017229009208790785,
+      "loss": 1.7994,
+      "step": 1217
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00017224641060947928,
+      "loss": 1.7244,
+      "step": 1218
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.000172202700275559,
+      "loss": 1.4061,
+      "step": 1219
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00017215896110360501,
+      "loss": 1.6177,
+      "step": 1220
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.000172115193111087,
+      "loss": 1.7864,
+      "step": 1221
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00017207139631548598,
+      "loss": 1.34,
+      "step": 1222
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.0001720275707342946,
+      "loss": 1.4509,
+      "step": 1223
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00017198371638501696,
+      "loss": 1.5508,
+      "step": 1224
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00017193983328516867,
+      "loss": 1.1117,
+      "step": 1225
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00017189592145227677,
+      "loss": 1.6735,
+      "step": 1226
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00017185198090387987,
+      "loss": 1.3761,
+      "step": 1227
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.0001718080116575279,
+      "loss": 1.7723,
+      "step": 1228
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00017176401373078247,
+      "loss": 1.5255,
+      "step": 1229
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.0001717199871412164,
+      "loss": 1.3831,
+      "step": 1230
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00017167593190641416,
+      "loss": 1.4873,
+      "step": 1231
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.0001716318480439715,
+      "loss": 1.514,
+      "step": 1232
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00017158773557149576,
+      "loss": 1.3593,
+      "step": 1233
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00017154359450660561,
+      "loss": 1.4141,
+      "step": 1234
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00017149942486693114,
+      "loss": 1.4316,
+      "step": 1235
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00017145522667011388,
+      "loss": 1.5706,
+      "step": 1236
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00017141099993380675,
+      "loss": 1.6436,
+      "step": 1237
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.0001713667446756741,
+      "loss": 1.47,
+      "step": 1238
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00017132246091339167,
+      "loss": 1.4376,
+      "step": 1239
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00017127814866464645,
+      "loss": 1.1966,
+      "step": 1240
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00017123380794713708,
+      "loss": 1.3722,
+      "step": 1241
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.0001711894387785733,
+      "loss": 1.6136,
+      "step": 1242
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00017114504117667637,
+      "loss": 1.4858,
+      "step": 1243
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.0001711006151591788,
+      "loss": 1.3902,
+      "step": 1244
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.0001710561607438246,
+      "loss": 1.4368,
+      "step": 1245
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00017101167794836899,
+      "loss": 1.4023,
+      "step": 1246
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 1.2095195055007935,
+      "eval_runtime": 349.5174,
+      "eval_samples_per_second": 5.513,
+      "eval_steps_per_second": 2.758,
+      "step": 1246
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00017096716679057856,
+      "loss": 1.359,
+      "step": 1247
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00017092262728823126,
+      "loss": 1.4988,
+      "step": 1248
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001708780594591163,
+      "loss": 1.3656,
+      "step": 1249
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00017083346332103427,
+      "loss": 1.4709,
+      "step": 1250
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00017078883889179704,
+      "loss": 1.4535,
+      "step": 1251
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00017074418618922776,
+      "loss": 1.6865,
+      "step": 1252
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001706995052311609,
+      "loss": 1.5769,
+      "step": 1253
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001706547960354422,
+      "loss": 1.6101,
+      "step": 1254
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001706100586199287,
+      "loss": 1.1927,
+      "step": 1255
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00017056529300248866,
+      "loss": 1.3256,
+      "step": 1256
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00017052049920100165,
+      "loss": 1.3677,
+      "step": 1257
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00017047567723335854,
+      "loss": 1.6378,
+      "step": 1258
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00017043082711746133,
+      "loss": 1.6817,
+      "step": 1259
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00017038594887122335,
+      "loss": 1.3687,
+      "step": 1260
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.0001703410425125691,
+      "loss": 1.3779,
+      "step": 1261
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00017029610805943443,
+      "loss": 1.5712,
+      "step": 1262
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00017025114552976628,
+      "loss": 1.6453,
+      "step": 1263
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00017020615494152286,
+      "loss": 1.2682,
+      "step": 1264
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00017016113631267358,
+      "loss": 1.4084,
+      "step": 1265
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00017011608966119906,
+      "loss": 1.3628,
+      "step": 1266
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00017007101500509113,
+      "loss": 1.3959,
+      "step": 1267
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00017002591236235274,
+      "loss": 1.4389,
+      "step": 1268
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00016998078175099808,
+      "loss": 1.8289,
+      "step": 1269
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00016993562318905246,
+      "loss": 1.5421,
+      "step": 1270
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00016989043669455241,
+      "loss": 1.6205,
+      "step": 1271
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.0001698452222855456,
+      "loss": 1.5397,
+      "step": 1272
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.625,
+      "learning_rate": 0.0001697999799800908,
+      "loss": 1.3872,
+      "step": 1273
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00016975470979625797,
+      "loss": 1.35,
+      "step": 1274
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.0001697094117521282,
+      "loss": 1.6612,
+      "step": 1275
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00016966408586579373,
+      "loss": 1.6312,
+      "step": 1276
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00016961873215535784,
+      "loss": 1.5005,
+      "step": 1277
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.000169573350638935,
+      "loss": 1.5772,
+      "step": 1278
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00016952794133465076,
+      "loss": 1.4382,
+      "step": 1279
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00016948250426064178,
+      "loss": 1.6273,
+      "step": 1280
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.0001694370394350558,
+      "loss": 1.7781,
+      "step": 1281
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.0001693915468760516,
+      "loss": 1.5343,
+      "step": 1282
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.0001693460266017991,
+      "loss": 1.2473,
+      "step": 1283
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00016930047863047932,
+      "loss": 1.8205,
+      "step": 1284
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00016925490298028427,
+      "loss": 1.4092,
+      "step": 1285
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.000169209299669417,
+      "loss": 1.7339,
+      "step": 1286
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00016916366871609165,
+      "loss": 1.4143,
+      "step": 1287
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.0001691180101385335,
+      "loss": 1.3932,
+      "step": 1288
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.0001690723239549786,
+      "loss": 1.3746,
+      "step": 1289
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.0001690266101836743,
+      "loss": 1.52,
+      "step": 1290
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00016898086884287883,
+      "loss": 1.3779,
+      "step": 1291
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00016893509995086144,
+      "loss": 1.2575,
+      "step": 1292
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00016888930352590238,
+      "loss": 1.5925,
+      "step": 1293
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.000168843479586293,
+      "loss": 1.619,
+      "step": 1294
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00016879762815033545,
+      "loss": 1.5154,
+      "step": 1295
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00016875174923634302,
+      "loss": 1.8358,
+      "step": 1296
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00016870584286263993,
+      "loss": 1.4425,
+      "step": 1297
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.00016865990904756136,
+      "loss": 1.7059,
+      "step": 1298
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00016861394780945345,
+      "loss": 1.7842,
+      "step": 1299
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00016856795916667325,
+      "loss": 1.4706,
+      "step": 1300
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00016852194313758887,
+      "loss": 1.4855,
+      "step": 1301
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00016847589974057923,
+      "loss": 1.611,
+      "step": 1302
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016842982899403428,
+      "loss": 1.735,
+      "step": 1303
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016838373091635482,
+      "loss": 1.2943,
+      "step": 1304
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00016833760552595258,
+      "loss": 1.4719,
+      "step": 1305
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001682914528412503,
+      "loss": 1.6199,
+      "step": 1306
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00016824527288068146,
+      "loss": 1.6889,
+      "step": 1307
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00016819906566269057,
+      "loss": 1.7251,
+      "step": 1308
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00016815283120573288,
+      "loss": 1.6231,
+      "step": 1309
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00016810656952827467,
+      "loss": 1.4416,
+      "step": 1310
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00016806028064879306,
+      "loss": 1.4151,
+      "step": 1311
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00016801396458577595,
+      "loss": 1.5705,
+      "step": 1312
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001679676213577222,
+      "loss": 1.4478,
+      "step": 1313
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.0001679212509831414,
+      "loss": 1.3156,
+      "step": 1314
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00016787485348055412,
+      "loss": 1.4202,
+      "step": 1315
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00016782842886849168,
+      "loss": 1.4063,
+      "step": 1316
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00016778197716549624,
+      "loss": 1.2529,
+      "step": 1317
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001677354983901208,
+      "loss": 1.3859,
+      "step": 1318
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00016768899256092913,
+      "loss": 1.6049,
+      "step": 1319
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00016764245969649587,
+      "loss": 1.6805,
+      "step": 1320
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00016759589981540641,
+      "loss": 1.5533,
+      "step": 1321
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00016754931293625695,
+      "loss": 1.331,
+      "step": 1322
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00016750269907765444,
+      "loss": 1.4939,
+      "step": 1323
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00016745605825821668,
+      "loss": 1.5602,
+      "step": 1324
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00016740939049657215,
+      "loss": 1.6936,
+      "step": 1325
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016736269581136014,
+      "loss": 1.4709,
+      "step": 1326
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00016731597422123073,
+      "loss": 1.6383,
+      "step": 1327
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00016726922574484464,
+      "loss": 1.4453,
+      "step": 1328
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.0001672224504008734,
+      "loss": 1.5169,
+      "step": 1329
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00016717564820799928,
+      "loss": 1.5968,
+      "step": 1330
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00016712881918491524,
+      "loss": 1.4834,
+      "step": 1331
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00016708196335032502,
+      "loss": 1.4554,
+      "step": 1332
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00016703508072294293,
+      "loss": 1.4103,
+      "step": 1333
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016698817132149415,
+      "loss": 1.8153,
+      "step": 1334
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00016694123516471443,
+      "loss": 1.5874,
+      "step": 1335
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00016689427227135025,
+      "loss": 1.1395,
+      "step": 1336
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00016684728266015879,
+      "loss": 1.5701,
+      "step": 1337
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00016680026634990786,
+      "loss": 1.4721,
+      "step": 1338
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016675322335937595,
+      "loss": 1.3294,
+      "step": 1339
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00016670615370735227,
+      "loss": 1.4185,
+      "step": 1340
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00016665905741263654,
+      "loss": 1.657,
+      "step": 1341
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00016661193449403922,
+      "loss": 1.5933,
+      "step": 1342
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001665647849703814,
+      "loss": 1.7177,
+      "step": 1343
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.0001665176088604948,
+      "loss": 1.4478,
+      "step": 1344
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00016647040618322168,
+      "loss": 1.4425,
+      "step": 1345
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00016642317695741502,
+      "loss": 1.6084,
+      "step": 1346
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00016637592120193832,
+      "loss": 1.5035,
+      "step": 1347
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00016632863893566573,
+      "loss": 1.6858,
+      "step": 1348
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00016628133017748193,
+      "loss": 1.6528,
+      "step": 1349
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.0001662339949462823,
+      "loss": 1.3561,
+      "step": 1350
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00016618663326097263,
+      "loss": 1.7063,
+      "step": 1351
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00016613924514046938,
+      "loss": 1.6164,
+      "step": 1352
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.4375,
+      "learning_rate": 0.00016609183060369958,
+      "loss": 1.6558,
+      "step": 1353
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00016604438966960072,
+      "loss": 1.5309,
+      "step": 1354
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00016599692235712095,
+      "loss": 1.38,
+      "step": 1355
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00016594942868521884,
+      "loss": 1.7297,
+      "step": 1356
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00016590190867286358,
+      "loss": 1.6041,
+      "step": 1357
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.0001658543623390348,
+      "loss": 1.6355,
+      "step": 1358
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00016580678970272276,
+      "loss": 1.623,
+      "step": 1359
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.0001657591907829281,
+      "loss": 1.6328,
+      "step": 1360
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00016571156559866203,
+      "loss": 1.4916,
+      "step": 1361
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00016566391416894622,
+      "loss": 1.3785,
+      "step": 1362
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00016561623651281283,
+      "loss": 1.5126,
+      "step": 1363
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.0001655685326493045,
+      "loss": 1.6248,
+      "step": 1364
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00016552080259747432,
+      "loss": 1.5102,
+      "step": 1365
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00016547304637638593,
+      "loss": 1.4575,
+      "step": 1366
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00016542526400511323,
+      "loss": 1.5368,
+      "step": 1367
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00016537745550274075,
+      "loss": 1.4704,
+      "step": 1368
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.0001653296208883634,
+      "loss": 1.4314,
+      "step": 1369
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00016528176018108646,
+      "loss": 1.5088,
+      "step": 1370
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.0001652338734000257,
+      "loss": 1.3929,
+      "step": 1371
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.0001651859605643073,
+      "loss": 1.4381,
+      "step": 1372
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.4375,
+      "learning_rate": 0.00016513802169306784,
+      "loss": 1.4458,
+      "step": 1373
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00016509005680545423,
+      "loss": 1.5203,
+      "step": 1374
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00016504206592062387,
+      "loss": 1.5067,
+      "step": 1375
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.0001649940490577445,
+      "loss": 1.5376,
+      "step": 1376
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00016494600623599424,
+      "loss": 1.483,
+      "step": 1377
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00016489793747456155,
+      "loss": 1.4846,
+      "step": 1378
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.0001648498427926453,
+      "loss": 1.6633,
+      "step": 1379
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.0001648017222094547,
+      "loss": 1.6565,
+      "step": 1380
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.0001647535757442093,
+      "loss": 1.5577,
+      "step": 1381
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00016470540341613892,
+      "loss": 1.2857,
+      "step": 1382
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00016465720524448384,
+      "loss": 1.2383,
+      "step": 1383
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00016460898124849452,
+      "loss": 1.2613,
+      "step": 1384
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.0001645607314474319,
+      "loss": 1.3296,
+      "step": 1385
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00016451245586056708,
+      "loss": 1.7099,
+      "step": 1386
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00016446415450718148,
+      "loss": 1.4778,
+      "step": 1387
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00016441582740656693,
+      "loss": 1.6256,
+      "step": 1388
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001643674745780254,
+      "loss": 1.5746,
+      "step": 1389
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00016431909604086922,
+      "loss": 1.5293,
+      "step": 1390
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00016427069181442093,
+      "loss": 1.5275,
+      "step": 1391
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.0001642222619180134,
+      "loss": 1.468,
+      "step": 1392
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.0001641738063709897,
+      "loss": 1.789,
+      "step": 1393
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.421875,
+      "learning_rate": 0.00016412532519270316,
+      "loss": 1.5199,
+      "step": 1394
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001640768184025173,
+      "loss": 1.5576,
+      "step": 1395
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.000164028286019806,
+      "loss": 1.3096,
+      "step": 1396
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00016397972806395328,
+      "loss": 1.5624,
+      "step": 1397
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00016393114455435327,
+      "loss": 1.5281,
+      "step": 1398
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001638825355104105,
+      "loss": 1.7554,
+      "step": 1399
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00016383390095153958,
+      "loss": 1.4449,
+      "step": 1400
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001637852408971653,
+      "loss": 1.7424,
+      "step": 1401
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00016373655536672276,
+      "loss": 1.2594,
+      "step": 1402
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001636878443796571,
+      "loss": 1.5085,
+      "step": 1403
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00016363910795542364,
+      "loss": 1.5266,
+      "step": 1404
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00016359034611348794,
+      "loss": 1.7234,
+      "step": 1405
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00016354155887332564,
+      "loss": 1.3752,
+      "step": 1406
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00016349274625442254,
+      "loss": 1.6621,
+      "step": 1407
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00016344390827627465,
+      "loss": 1.6427,
+      "step": 1408
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00016339504495838797,
+      "loss": 1.6467,
+      "step": 1409
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00016334615632027872,
+      "loss": 1.5806,
+      "step": 1410
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.0001632972423814732,
+      "loss": 1.7138,
+      "step": 1411
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00016324830316150787,
+      "loss": 1.7285,
+      "step": 1412
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001631993386799292,
+      "loss": 1.2404,
+      "step": 1413
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00016315034895629377,
+      "loss": 1.3281,
+      "step": 1414
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00016310133401016834,
+      "loss": 1.6741,
+      "step": 1415
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00016305229386112959,
+      "loss": 1.5625,
+      "step": 1416
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00016300322852876435,
+      "loss": 1.4703,
+      "step": 1417
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016295413803266954,
+      "loss": 1.4291,
+      "step": 1418
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016290502239245207,
+      "loss": 1.5193,
+      "step": 1419
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00016285588162772894,
+      "loss": 1.4889,
+      "step": 1420
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.0001628067157581271,
+      "loss": 1.9033,
+      "step": 1421
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.0001627575248032836,
+      "loss": 1.3462,
+      "step": 1422
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00016270830878284555,
+      "loss": 1.459,
+      "step": 1423
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00016265906771646997,
+      "loss": 1.41,
+      "step": 1424
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00016260980162382396,
+      "loss": 1.511,
+      "step": 1425
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.0001625605105245845,
+      "loss": 1.739,
+      "step": 1426
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00016251119443843877,
+      "loss": 1.6373,
+      "step": 1427
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00016246185338508374,
+      "loss": 1.4634,
+      "step": 1428
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001624124873842264,
+      "loss": 1.6238,
+      "step": 1429
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001623630964555837,
+      "loss": 1.5034,
+      "step": 1430
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00016231368061888265,
+      "loss": 1.6165,
+      "step": 1431
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 13.125,
+      "learning_rate": 0.00016226423989386003,
+      "loss": 1.1504,
+      "step": 1432
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00016221477430026276,
+      "loss": 1.5782,
+      "step": 1433
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00016216528385784747,
+      "loss": 1.4364,
+      "step": 1434
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0001621157685863809,
+      "loss": 1.668,
+      "step": 1435
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001620662285056396,
+      "loss": 1.6605,
+      "step": 1436
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00016201666363541012,
+      "loss": 1.6093,
+      "step": 1437
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00016196707399548883,
+      "loss": 1.4887,
+      "step": 1438
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00016191745960568202,
+      "loss": 1.8229,
+      "step": 1439
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00016186782048580582,
+      "loss": 1.3811,
+      "step": 1440
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001618181566556864,
+      "loss": 1.5787,
+      "step": 1441
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016176846813515956,
+      "loss": 1.695,
+      "step": 1442
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00016171875494407116,
+      "loss": 1.4391,
+      "step": 1443
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00016166901710227682,
+      "loss": 1.5143,
+      "step": 1444
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00016161925462964197,
+      "loss": 1.395,
+      "step": 1445
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00016156946754604205,
+      "loss": 1.4713,
+      "step": 1446
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016151965587136207,
+      "loss": 1.5884,
+      "step": 1447
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.8125,
+      "learning_rate": 0.00016146981962549713,
+      "loss": 1.3529,
+      "step": 1448
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00016141995882835194,
+      "loss": 1.3144,
+      "step": 1449
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001613700734998411,
+      "loss": 1.6648,
+      "step": 1450
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00016132016365988903,
+      "loss": 1.4124,
+      "step": 1451
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00016127022932842987,
+      "loss": 1.387,
+      "step": 1452
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00016122027052540764,
+      "loss": 1.6801,
+      "step": 1453
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00016117028727077602,
+      "loss": 1.2619,
+      "step": 1454
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00016112027958449854,
+      "loss": 1.4094,
+      "step": 1455
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00016107024748654845,
+      "loss": 1.5227,
+      "step": 1456
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00016102019099690877,
+      "loss": 1.4407,
+      "step": 1457
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00016097011013557224,
+      "loss": 1.568,
+      "step": 1458
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00016092000492254135,
+      "loss": 1.7368,
+      "step": 1459
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00016086987537782834,
+      "loss": 1.5851,
+      "step": 1460
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00016081972152145508,
+      "loss": 1.6653,
+      "step": 1461
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00016076954337345328,
+      "loss": 1.728,
+      "step": 1462
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00016071934095386424,
+      "loss": 1.1396,
+      "step": 1463
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00016066911428273906,
+      "loss": 1.736,
+      "step": 1464
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00016061886338013834,
+      "loss": 1.5624,
+      "step": 1465
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.0001605685882661326,
+      "loss": 1.6319,
+      "step": 1466
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00016051828896080186,
+      "loss": 1.4108,
+      "step": 1467
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00016046796548423585,
+      "loss": 1.6487,
+      "step": 1468
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00016041761785653397,
+      "loss": 1.5497,
+      "step": 1469
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00016036724609780532,
+      "loss": 1.707,
+      "step": 1470
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016031685022816844,
+      "loss": 1.5924,
+      "step": 1471
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.0001602664302677517,
+      "loss": 1.9031,
+      "step": 1472
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00016021598623669307,
+      "loss": 1.7518,
+      "step": 1473
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00016016551815514003,
+      "loss": 1.4376,
+      "step": 1474
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00016011502604324973,
+      "loss": 1.6604,
+      "step": 1475
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00016006450992118892,
+      "loss": 1.6325,
+      "step": 1476
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00016001396980913393,
+      "loss": 1.5616,
+      "step": 1477
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00015996340572727065,
+      "loss": 1.3595,
+      "step": 1478
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00015991281769579466,
+      "loss": 1.7323,
+      "step": 1479
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001598622057349109,
+      "loss": 1.4547,
+      "step": 1480
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.000159811569864834,
+      "loss": 1.4278,
+      "step": 1481
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00015976091010578815,
+      "loss": 1.4119,
+      "step": 1482
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00015971022647800708,
+      "loss": 1.5258,
+      "step": 1483
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.000159659519001734,
+      "loss": 1.4221,
+      "step": 1484
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.0001596087876972216,
+      "loss": 1.3583,
+      "step": 1485
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00015955803258473224,
+      "loss": 1.8021,
+      "step": 1486
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.0001595072536845377,
+      "loss": 1.4168,
+      "step": 1487
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015945645101691925,
+      "loss": 1.8009,
+      "step": 1488
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00015940562460216764,
+      "loss": 1.7529,
+      "step": 1489
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00015935477446058318,
+      "loss": 1.5002,
+      "step": 1490
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00015930390061247556,
+      "loss": 1.5142,
+      "step": 1491
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00015925300307816404,
+      "loss": 1.6381,
+      "step": 1492
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00015920208187797727,
+      "loss": 1.3551,
+      "step": 1493
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00015915113703225336,
+      "loss": 1.5823,
+      "step": 1494
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.0001591001685613399,
+      "loss": 1.6135,
+      "step": 1495
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00015904917648559383,
+      "loss": 1.6297,
+      "step": 1496
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00015899816082538167,
+      "loss": 1.3485,
+      "step": 1497
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00015894712160107922,
+      "loss": 1.488,
+      "step": 1498
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00015889605883307172,
+      "loss": 1.4986,
+      "step": 1499
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.0001588449725417539,
+      "loss": 1.4793,
+      "step": 1500
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00015879386274752973,
+      "loss": 1.6877,
+      "step": 1501
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.875,
+      "learning_rate": 0.0001587427294708127,
+      "loss": 1.513,
+      "step": 1502
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.0001586915727320257,
+      "loss": 1.3583,
+      "step": 1503
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00015864039255160083,
+      "loss": 1.3952,
+      "step": 1504
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00015858918894997968,
+      "loss": 1.6764,
+      "step": 1505
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00015853796194761322,
+      "loss": 1.556,
+      "step": 1506
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00015848671156496163,
+      "loss": 1.4642,
+      "step": 1507
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.0001584354378224945,
+      "loss": 1.6729,
+      "step": 1508
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00015838414074069082,
+      "loss": 1.4679,
+      "step": 1509
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00015833282034003883,
+      "loss": 1.5998,
+      "step": 1510
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001582814766410361,
+      "loss": 1.5292,
+      "step": 1511
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00015823010966418944,
+      "loss": 1.4632,
+      "step": 1512
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00015817871943001505,
+      "loss": 1.6433,
+      "step": 1513
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00015812730595903842,
+      "loss": 1.7387,
+      "step": 1514
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00015807586927179426,
+      "loss": 1.7881,
+      "step": 1515
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00015802440938882657,
+      "loss": 1.6313,
+      "step": 1516
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.0001579729263306886,
+      "loss": 1.5077,
+      "step": 1517
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00015792142011794295,
+      "loss": 1.6805,
+      "step": 1518
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00015786989077116133,
+      "loss": 1.3844,
+      "step": 1519
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00015781833831092482,
+      "loss": 1.4221,
+      "step": 1520
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00015776676275782356,
+      "loss": 1.631,
+      "step": 1521
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00015771516413245714,
+      "loss": 1.3192,
+      "step": 1522
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00015766354245543422,
+      "loss": 1.4797,
+      "step": 1523
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00015761189774737266,
+      "loss": 1.2426,
+      "step": 1524
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00015756023002889958,
+      "loss": 1.7204,
+      "step": 1525
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015750853932065123,
+      "loss": 1.7015,
+      "step": 1526
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00015745682564327314,
+      "loss": 1.6864,
+      "step": 1527
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00015740508901741985,
+      "loss": 1.4752,
+      "step": 1528
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.0001573533294637553,
+      "loss": 1.4741,
+      "step": 1529
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001573015470029523,
+      "loss": 1.623,
+      "step": 1530
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00015724974165569312,
+      "loss": 1.6505,
+      "step": 1531
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.90625,
+      "learning_rate": 0.0001571979134426689,
+      "loss": 1.6948,
+      "step": 1532
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00015714606238458005,
+      "loss": 1.3901,
+      "step": 1533
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.0001570941885021361,
+      "loss": 1.7931,
+      "step": 1534
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00015704229181605567,
+      "loss": 1.2555,
+      "step": 1535
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00015699037234706648,
+      "loss": 1.5489,
+      "step": 1536
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.0001569384301159054,
+      "loss": 1.7698,
+      "step": 1537
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.0001568864651433183,
+      "loss": 1.5066,
+      "step": 1538
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001568344774500602,
+      "loss": 1.6526,
+      "step": 1539
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00015678246705689525,
+      "loss": 1.4895,
+      "step": 1540
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00015673043398459652,
+      "loss": 1.3507,
+      "step": 1541
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00015667837825394628,
+      "loss": 1.2462,
+      "step": 1542
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00015662629988573572,
+      "loss": 1.8138,
+      "step": 1543
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00015657419890076517,
+      "loss": 1.5165,
+      "step": 1544
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.000156522075319844,
+      "loss": 1.5379,
+      "step": 1545
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00015646992916379051,
+      "loss": 1.4833,
+      "step": 1546
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015641776045343208,
+      "loss": 1.697,
+      "step": 1547
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00015636556920960515,
+      "loss": 1.2866,
+      "step": 1548
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00015631335545315502,
+      "loss": 1.4057,
+      "step": 1549
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.0001562611192049361,
+      "loss": 1.5358,
+      "step": 1550
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00015620886048581175,
+      "loss": 1.4326,
+      "step": 1551
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00015615657931665434,
+      "loss": 1.711,
+      "step": 1552
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00015610427571834508,
+      "loss": 1.5432,
+      "step": 1553
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00015605194971177432,
+      "loss": 1.5876,
+      "step": 1554
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5,
+      "learning_rate": 0.0001559996013178412,
+      "loss": 1.7905,
+      "step": 1555
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00015594723055745388,
+      "loss": 1.6608,
+      "step": 1556
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00015589483745152946,
+      "loss": 1.5071,
+      "step": 1557
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00015584242202099395,
+      "loss": 1.8638,
+      "step": 1558
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015578998428678224,
+      "loss": 1.5616,
+      "step": 1559
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00015573752426983818,
+      "loss": 1.6827,
+      "step": 1560
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00015568504199111452,
+      "loss": 1.4485,
+      "step": 1561
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00015563253747157284,
+      "loss": 1.4452,
+      "step": 1562
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00015558001073218372,
+      "loss": 1.501,
+      "step": 1563
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00015552746179392643,
+      "loss": 1.7402,
+      "step": 1564
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.0001554748906777893,
+      "loss": 1.3976,
+      "step": 1565
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.0001554222974047694,
+      "loss": 1.4235,
+      "step": 1566
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.0001553696819958727,
+      "loss": 1.694,
+      "step": 1567
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00015531704447211398,
+      "loss": 1.4443,
+      "step": 1568
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.0001552643848545169,
+      "loss": 1.4187,
+      "step": 1569
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00015521170316411386,
+      "loss": 1.7061,
+      "step": 1570
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00015515899942194616,
+      "loss": 1.4792,
+      "step": 1571
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00015510627364906387,
+      "loss": 1.2383,
+      "step": 1572
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00015505352586652586,
+      "loss": 1.2174,
+      "step": 1573
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00015500075609539983,
+      "loss": 1.9427,
+      "step": 1574
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00015494796435676218,
+      "loss": 1.3856,
+      "step": 1575
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015489515067169815,
+      "loss": 1.5725,
+      "step": 1576
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.0001548423150613017,
+      "loss": 1.3577,
+      "step": 1577
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00015478945754667563,
+      "loss": 1.6514,
+      "step": 1578
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00015473657814893144,
+      "loss": 1.4926,
+      "step": 1579
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00015468367688918925,
+      "loss": 1.412,
+      "step": 1580
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00015463075378857816,
+      "loss": 1.54,
+      "step": 1581
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00015457780886823577,
+      "loss": 1.6603,
+      "step": 1582
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.0001545248421493085,
+      "loss": 1.5862,
+      "step": 1583
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00015447185365295151,
+      "loss": 1.5844,
+      "step": 1584
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00015441884340032855,
+      "loss": 1.3721,
+      "step": 1585
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00015436581141261213,
+      "loss": 1.6988,
+      "step": 1586
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00015431275771098346,
+      "loss": 1.5496,
+      "step": 1587
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00015425968231663235,
+      "loss": 1.5401,
+      "step": 1588
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00015420658525075733,
+      "loss": 1.7295,
+      "step": 1589
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00015415346653456556,
+      "loss": 1.4988,
+      "step": 1590
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5,
+      "learning_rate": 0.0001541003261892729,
+      "loss": 1.7145,
+      "step": 1591
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00015404716423610376,
+      "loss": 1.6639,
+      "step": 1592
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.0001539939806962912,
+      "loss": 1.4988,
+      "step": 1593
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00015394077559107702,
+      "loss": 1.2403,
+      "step": 1594
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00015388754894171146,
+      "loss": 1.534,
+      "step": 1595
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.0001538343007694535,
+      "loss": 1.4749,
+      "step": 1596
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00015378103109557064,
+      "loss": 1.564,
+      "step": 1597
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00015372773994133898,
+      "loss": 1.6196,
+      "step": 1598
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00015367442732804322,
+      "loss": 1.6138,
+      "step": 1599
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00015362109327697664,
+      "loss": 1.5592,
+      "step": 1600
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00015356773780944103,
+      "loss": 1.7264,
+      "step": 1601
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.0001535143609467468,
+      "loss": 1.4949,
+      "step": 1602
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00015346096271021283,
+      "loss": 1.8838,
+      "step": 1603
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.0001534075431211666,
+      "loss": 1.4638,
+      "step": 1604
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00015335410220094416,
+      "loss": 1.534,
+      "step": 1605
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00015330063997088993,
+      "loss": 1.4887,
+      "step": 1606
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00015324715645235695,
+      "loss": 1.5772,
+      "step": 1607
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00015319365166670674,
+      "loss": 1.3885,
+      "step": 1608
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00015314012563530933,
+      "loss": 1.7713,
+      "step": 1609
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.0001530865783795432,
+      "loss": 1.6883,
+      "step": 1610
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00015303300992079537,
+      "loss": 1.5665,
+      "step": 1611
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00015297942028046126,
+      "loss": 1.5031,
+      "step": 1612
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00015292580947994476,
+      "loss": 1.6448,
+      "step": 1613
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00015287217754065825,
+      "loss": 1.4017,
+      "step": 1614
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00015281852448402255,
+      "loss": 1.6806,
+      "step": 1615
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00015276485033146684,
+      "loss": 1.4499,
+      "step": 1616
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00015271115510442882,
+      "loss": 1.5097,
+      "step": 1617
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00015265743882435457,
+      "loss": 1.4011,
+      "step": 1618
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015260370151269856,
+      "loss": 1.3623,
+      "step": 1619
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00015254994319092367,
+      "loss": 1.5572,
+      "step": 1620
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00015249616388050122,
+      "loss": 1.3959,
+      "step": 1621
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00015244236360291083,
+      "loss": 1.7081,
+      "step": 1622
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00015238854237964054,
+      "loss": 1.3523,
+      "step": 1623
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00015233470023218675,
+      "loss": 1.4648,
+      "step": 1624
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00015228083718205428,
+      "loss": 1.4213,
+      "step": 1625
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00015222695325075612,
+      "loss": 1.3962,
+      "step": 1626
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00015217304845981385,
+      "loss": 1.6649,
+      "step": 1627
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00015211912283075716,
+      "loss": 1.6562,
+      "step": 1628
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00015206517638512417,
+      "loss": 1.4251,
+      "step": 1629
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.0001520112091444613,
+      "loss": 1.6415,
+      "step": 1630
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00015195722113032332,
+      "loss": 1.7768,
+      "step": 1631
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00015190321236427314,
+      "loss": 1.5941,
+      "step": 1632
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00015184918286788218,
+      "loss": 1.4251,
+      "step": 1633
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00015179513266272996,
+      "loss": 1.5457,
+      "step": 1634
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00015174106177040435,
+      "loss": 1.4616,
+      "step": 1635
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00015168697021250145,
+      "loss": 1.5062,
+      "step": 1636
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001516328580106257,
+      "loss": 1.1532,
+      "step": 1637
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00015157872518638968,
+      "loss": 1.6822,
+      "step": 1638
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00015152457176141422,
+      "loss": 1.6915,
+      "step": 1639
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00015147039775732842,
+      "loss": 1.4629,
+      "step": 1640
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00015141620319576965,
+      "loss": 1.3855,
+      "step": 1641
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00015136198809838328,
+      "loss": 1.6028,
+      "step": 1642
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00015130775248682317,
+      "loss": 1.5965,
+      "step": 1643
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00015125349638275116,
+      "loss": 1.54,
+      "step": 1644
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00015119921980783738,
+      "loss": 1.6053,
+      "step": 1645
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00015114492278376008,
+      "loss": 1.7972,
+      "step": 1646
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00015109060533220566,
+      "loss": 1.6394,
+      "step": 1647
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.0001510362674748688,
+      "loss": 1.5829,
+      "step": 1648
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00015098190923345216,
+      "loss": 1.655,
+      "step": 1649
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.0001509275306296667,
+      "loss": 1.5933,
+      "step": 1650
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.0001508731316852314,
+      "loss": 1.4282,
+      "step": 1651
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001508187124218734,
+      "loss": 1.566,
+      "step": 1652
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00015076427286132805,
+      "loss": 1.3975,
+      "step": 1653
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.0001507098130253386,
+      "loss": 1.3266,
+      "step": 1654
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00015065533293565656,
+      "loss": 1.6534,
+      "step": 1655
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.0001506008326140415,
+      "loss": 1.4807,
+      "step": 1656
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00015054631208226104,
+      "loss": 1.5987,
+      "step": 1657
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.0001504917713620909,
+      "loss": 1.5163,
+      "step": 1658
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.015625,
+      "learning_rate": 0.00015043721047531484,
+      "loss": 1.6453,
+      "step": 1659
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00015038262944372467,
+      "loss": 1.7021,
+      "step": 1660
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00015032802828912032,
+      "loss": 1.5494,
+      "step": 1661
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00015027340703330965,
+      "loss": 1.4722,
+      "step": 1662
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.875,
+      "learning_rate": 0.0001502187656981086,
+      "loss": 1.4593,
+      "step": 1663
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00015016410430534116,
+      "loss": 1.207,
+      "step": 1664
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00015010942287683925,
+      "loss": 1.444,
+      "step": 1665
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.0001500547214344429,
+      "loss": 1.6505,
+      "step": 1666
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 1.5071,
+      "step": 1667
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001499452585953666,
+      "loss": 1.5508,
+      "step": 1668
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001498904972424065,
+      "loss": 1.3227,
+      "step": 1669
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001498357159629917,
+      "loss": 1.6901,
+      "step": 1670
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.000149780914779002,
+      "loss": 1.5102,
+      "step": 1671
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.0001497260937123252,
+      "loss": 1.3672,
+      "step": 1672
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.0001496712527848571,
+      "loss": 1.3731,
+      "step": 1673
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.0001496163920185013,
+      "loss": 1.6289,
+      "step": 1674
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00014956151143516946,
+      "loss": 1.6253,
+      "step": 1675
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00014950661105678104,
+      "loss": 1.7225,
+      "step": 1676
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.0001494516909052635,
+      "loss": 1.7156,
+      "step": 1677
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00014939675100255213,
+      "loss": 1.4298,
+      "step": 1678
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00014934179137059012,
+      "loss": 1.5245,
+      "step": 1679
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00014928681203132865,
+      "loss": 1.369,
+      "step": 1680
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00014923181300672658,
+      "loss": 1.6884,
+      "step": 1681
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00014917679431875074,
+      "loss": 1.3943,
+      "step": 1682
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.0001491217559893759,
+      "loss": 1.3063,
+      "step": 1683
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00014906669804058443,
+      "loss": 1.2378,
+      "step": 1684
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.0001490116204943668,
+      "loss": 1.5259,
+      "step": 1685
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00014895652337272114,
+      "loss": 1.7351,
+      "step": 1686
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014890140669765348,
+      "loss": 1.4509,
+      "step": 1687
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.0001488462704911776,
+      "loss": 1.6986,
+      "step": 1688
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00014879111477531513,
+      "loss": 1.7428,
+      "step": 1689
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.0001487359395720955,
+      "loss": 1.3974,
+      "step": 1690
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.0001486807449035558,
+      "loss": 1.6137,
+      "step": 1691
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001486255307917411,
+      "loss": 1.1817,
+      "step": 1692
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00014857029725870406,
+      "loss": 1.4732,
+      "step": 1693
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001485150443265052,
+      "loss": 1.4249,
+      "step": 1694
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.0001484597720172127,
+      "loss": 1.8498,
+      "step": 1695
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00014840448035290262,
+      "loss": 1.4435,
+      "step": 1696
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00014834916935565858,
+      "loss": 1.85,
+      "step": 1697
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00014829383904757206,
+      "loss": 1.384,
+      "step": 1698
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00014823848945074212,
+      "loss": 1.4985,
+      "step": 1699
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.00014818312058727568,
+      "loss": 1.432,
+      "step": 1700
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00014812773247928725,
+      "loss": 1.4784,
+      "step": 1701
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014807232514889903,
+      "loss": 1.5964,
+      "step": 1702
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00014801689861824096,
+      "loss": 1.1325,
+      "step": 1703
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014796145290945053,
+      "loss": 1.475,
+      "step": 1704
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00014790598804467306,
+      "loss": 1.3034,
+      "step": 1705
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.0001478505040460614,
+      "loss": 1.5027,
+      "step": 1706
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.000147795000935776,
+      "loss": 1.3611,
+      "step": 1707
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.0001477394787359851,
+      "loss": 1.5016,
+      "step": 1708
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00014768393746886446,
+      "loss": 1.5017,
+      "step": 1709
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00014762837715659745,
+      "loss": 1.6795,
+      "step": 1710
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.0001475727978213751,
+      "loss": 1.5,
+      "step": 1711
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00014751719948539598,
+      "loss": 1.4545,
+      "step": 1712
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.0001474615821708663,
+      "loss": 1.7193,
+      "step": 1713
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00014740594589999982,
+      "loss": 1.2165,
+      "step": 1714
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00014735029069501784,
+      "loss": 1.7336,
+      "step": 1715
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00014729461657814934,
+      "loss": 1.4078,
+      "step": 1716
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.0001472389235716307,
+      "loss": 1.8306,
+      "step": 1717
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00014718321169770595,
+      "loss": 1.523,
+      "step": 1718
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00014712748097862665,
+      "loss": 1.3429,
+      "step": 1719
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00014707173143665184,
+      "loss": 1.8801,
+      "step": 1720
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00014701596309404805,
+      "loss": 1.4829,
+      "step": 1721
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00014696017597308947,
+      "loss": 1.54,
+      "step": 1722
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.0001469043700960576,
+      "loss": 1.7453,
+      "step": 1723
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014684854548524157,
+      "loss": 1.5021,
+      "step": 1724
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.0001467927021629379,
+      "loss": 1.7448,
+      "step": 1725
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00014673684015145067,
+      "loss": 1.5894,
+      "step": 1726
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00014668095947309136,
+      "loss": 1.6298,
+      "step": 1727
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00014662506015017892,
+      "loss": 1.7013,
+      "step": 1728
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00014656914220503978,
+      "loss": 1.5277,
+      "step": 1729
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.0001465132056600077,
+      "loss": 1.579,
+      "step": 1730
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00014645725053742407,
+      "loss": 1.56,
+      "step": 1731
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00014640127685963753,
+      "loss": 1.5736,
+      "step": 1732
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00014634528464900414,
+      "loss": 1.3496,
+      "step": 1733
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00014628927392788742,
+      "loss": 1.5339,
+      "step": 1734
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00014623324471865832,
+      "loss": 1.763,
+      "step": 1735
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00014617719704369505,
+      "loss": 1.6617,
+      "step": 1736
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00014612113092538328,
+      "loss": 1.4818,
+      "step": 1737
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00014606504638611605,
+      "loss": 1.5252,
+      "step": 1738
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00014600894344829373,
+      "loss": 1.4108,
+      "step": 1739
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00014595282213432405,
+      "loss": 1.4589,
+      "step": 1740
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.0001458966824666221,
+      "loss": 1.3894,
+      "step": 1741
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.0001458405244676102,
+      "loss": 1.444,
+      "step": 1742
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00014578434815971816,
+      "loss": 1.7341,
+      "step": 1743
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00014572815356538298,
+      "loss": 1.4364,
+      "step": 1744
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00014567194070704898,
+      "loss": 1.6184,
+      "step": 1745
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014561570960716777,
+      "loss": 1.6849,
+      "step": 1746
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.0001455594602881983,
+      "loss": 1.5138,
+      "step": 1747
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.0001455031927726068,
+      "loss": 1.7773,
+      "step": 1748
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00014544690708286667,
+      "loss": 1.4184,
+      "step": 1749
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014539060324145865,
+      "loss": 1.5095,
+      "step": 1750
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00014533428127087066,
+      "loss": 1.5153,
+      "step": 1751
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.000145277941193598,
+      "loss": 1.2706,
+      "step": 1752
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00014522158303214303,
+      "loss": 1.7005,
+      "step": 1753
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00014516520680901547,
+      "loss": 1.709,
+      "step": 1754
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00014510881254673218,
+      "loss": 1.3576,
+      "step": 1755
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00014505240026781727,
+      "loss": 1.7161,
+      "step": 1756
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00014499596999480193,
+      "loss": 1.6811,
+      "step": 1757
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00014493952175022475,
+      "loss": 1.56,
+      "step": 1758
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00014488305555663124,
+      "loss": 1.4501,
+      "step": 1759
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00014482657143657433,
+      "loss": 1.4765,
+      "step": 1760
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00014477006941261396,
+      "loss": 1.4413,
+      "step": 1761
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00014471354950731718,
+      "loss": 1.4961,
+      "step": 1762
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00014465701174325835,
+      "loss": 1.404,
+      "step": 1763
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00014460045614301886,
+      "loss": 1.637,
+      "step": 1764
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00014454388272918718,
+      "loss": 1.4834,
+      "step": 1765
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.000144487291524359,
+      "loss": 1.7757,
+      "step": 1766
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.000144430682551137,
+      "loss": 1.7076,
+      "step": 1767
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00014437405583213106,
+      "loss": 1.5343,
+      "step": 1768
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001443174113899581,
+      "loss": 1.384,
+      "step": 1769
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00014426074924724216,
+      "loss": 1.6481,
+      "step": 1770
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00014420406942661427,
+      "loss": 1.3969,
+      "step": 1771
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00014414737195071257,
+      "loss": 1.3362,
+      "step": 1772
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00014409065684218227,
+      "loss": 1.5292,
+      "step": 1773
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00014403392412367557,
+      "loss": 1.6671,
+      "step": 1774
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.0001439771738178518,
+      "loss": 1.4587,
+      "step": 1775
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00014392040594737719,
+      "loss": 1.8268,
+      "step": 1776
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014386362053492503,
+      "loss": 1.3468,
+      "step": 1777
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00014380681760317566,
+      "loss": 1.5829,
+      "step": 1778
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00014374999717481642,
+      "loss": 1.3879,
+      "step": 1779
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00014369315927254152,
+      "loss": 1.5198,
+      "step": 1780
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00014363630391905235,
+      "loss": 1.411,
+      "step": 1781
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00014357943113705703,
+      "loss": 1.6258,
+      "step": 1782
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00014352254094927082,
+      "loss": 1.4679,
+      "step": 1783
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00014346563337841591,
+      "loss": 1.3313,
+      "step": 1784
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00014340870844722139,
+      "loss": 1.6104,
+      "step": 1785
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00014335176617842325,
+      "loss": 1.3222,
+      "step": 1786
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00014329480659476448,
+      "loss": 1.5228,
+      "step": 1787
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00014323782971899495,
+      "loss": 1.7596,
+      "step": 1788
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00014318083557387148,
+      "loss": 1.4952,
+      "step": 1789
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00014312382418215768,
+      "loss": 1.5452,
+      "step": 1790
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014306679556662415,
+      "loss": 1.5727,
+      "step": 1791
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00014300974975004835,
+      "loss": 1.7836,
+      "step": 1792
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00014295268675521455,
+      "loss": 1.5429,
+      "step": 1793
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.000142895606604914,
+      "loss": 1.4211,
+      "step": 1794
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00014283850932194468,
+      "loss": 1.5792,
+      "step": 1795
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00014278139492911147,
+      "loss": 1.4404,
+      "step": 1796
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00014272426344922608,
+      "loss": 1.2278,
+      "step": 1797
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.000142667114905107,
+      "loss": 1.4741,
+      "step": 1798
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00014260994931957965,
+      "loss": 1.5911,
+      "step": 1799
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00014255276671547616,
+      "loss": 1.5588,
+      "step": 1800
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00014249556711563546,
+      "loss": 1.6719,
+      "step": 1801
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00014243835054290327,
+      "loss": 1.6026,
+      "step": 1802
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014238111702013215,
+      "loss": 1.5258,
+      "step": 1803
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00014232386657018136,
+      "loss": 1.6018,
+      "step": 1804
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00014226659921591698,
+      "loss": 1.6251,
+      "step": 1805
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00014220931498021177,
+      "loss": 1.6415,
+      "step": 1806
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00014215201388594534,
+      "loss": 1.2976,
+      "step": 1807
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00014209469595600384,
+      "loss": 1.4798,
+      "step": 1808
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00014203736121328036,
+      "loss": 1.499,
+      "step": 1809
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.34375,
+      "learning_rate": 0.00014198000968067466,
+      "loss": 1.3932,
+      "step": 1810
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00014192264138109307,
+      "loss": 1.4568,
+      "step": 1811
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.0001418652563374487,
+      "loss": 1.5905,
+      "step": 1812
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00014180785457266145,
+      "loss": 1.437,
+      "step": 1813
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001417504361096577,
+      "loss": 1.597,
+      "step": 1814
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5,
+      "learning_rate": 0.0001416930009713707,
+      "loss": 1.6899,
+      "step": 1815
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00014163554918074014,
+      "loss": 1.6524,
+      "step": 1816
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001415780807607126,
+      "loss": 1.4941,
+      "step": 1817
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00014152059573424114,
+      "loss": 1.4608,
+      "step": 1818
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.0001414630941242855,
+      "loss": 1.5521,
+      "step": 1819
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.0001414055759538121,
+      "loss": 1.6503,
+      "step": 1820
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014134804124579377,
+      "loss": 1.1515,
+      "step": 1821
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001412904900232102,
+      "loss": 1.3906,
+      "step": 1822
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.0001412329223090476,
+      "loss": 1.589,
+      "step": 1823
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00014117533812629866,
+      "loss": 1.6066,
+      "step": 1824
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00014111773749796272,
+      "loss": 1.4448,
+      "step": 1825
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00014106012044704576,
+      "loss": 1.7669,
+      "step": 1826
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00014100248699656017,
+      "loss": 1.6402,
+      "step": 1827
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014094483716952503,
+      "loss": 1.6175,
+      "step": 1828
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00014088717098896587,
+      "loss": 1.7148,
+      "step": 1829
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00014082948847791478,
+      "loss": 1.5332,
+      "step": 1830
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00014077178965941038,
+      "loss": 1.6742,
+      "step": 1831
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001407140745564978,
+      "loss": 1.8057,
+      "step": 1832
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.0001406563431922287,
+      "loss": 1.7565,
+      "step": 1833
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00014059859558966117,
+      "loss": 1.5585,
+      "step": 1834
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.0001405408317718598,
+      "loss": 1.496,
+      "step": 1835
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00014048305176189574,
+      "loss": 1.529,
+      "step": 1836
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00014042525558284654,
+      "loss": 1.3959,
+      "step": 1837
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00014036744325779616,
+      "loss": 1.6499,
+      "step": 1838
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00014030961480983512,
+      "loss": 1.5992,
+      "step": 1839
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00014025177026206028,
+      "loss": 1.6553,
+      "step": 1840
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00014019390963757503,
+      "loss": 1.683,
+      "step": 1841
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00014013603295948902,
+      "loss": 1.5113,
+      "step": 1842
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.0001400781402509185,
+      "loss": 1.6053,
+      "step": 1843
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00014002023153498604,
+      "loss": 1.4939,
+      "step": 1844
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00013996230683482059,
+      "loss": 1.7137,
+      "step": 1845
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00013990436617355745,
+      "loss": 1.665,
+      "step": 1846
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00013984640957433842,
+      "loss": 1.315,
+      "step": 1847
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.0001397884370603115,
+      "loss": 1.3715,
+      "step": 1848
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00013973044865463123,
+      "loss": 1.4179,
+      "step": 1849
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.0001396724443804583,
+      "loss": 1.6623,
+      "step": 1850
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.0001396144242609599,
+      "loss": 1.3618,
+      "step": 1851
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00013955638831930948,
+      "loss": 1.4905,
+      "step": 1852
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00013949833657868682,
+      "loss": 1.4006,
+      "step": 1853
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00013944026906227797,
+      "loss": 1.2619,
+      "step": 1854
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00013938218579327537,
+      "loss": 1.632,
+      "step": 1855
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00013932408679487765,
+      "loss": 1.7885,
+      "step": 1856
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00013926597209028982,
+      "loss": 1.739,
+      "step": 1857
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00013920784170272306,
+      "loss": 1.5529,
+      "step": 1858
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.0001391496956553949,
+      "loss": 1.4404,
+      "step": 1859
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001390915339715291,
+      "loss": 1.2866,
+      "step": 1860
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013903335667435565,
+      "loss": 1.6739,
+      "step": 1861
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013897516378711075,
+      "loss": 1.5655,
+      "step": 1862
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00013891695533303693,
+      "loss": 1.711,
+      "step": 1863
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.0001388587313353828,
+      "loss": 1.6114,
+      "step": 1864
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00013880049181740328,
+      "loss": 1.6813,
+      "step": 1865
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.00013874223680235948,
+      "loss": 1.4899,
+      "step": 1866
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00013868396631351866,
+      "loss": 1.6155,
+      "step": 1867
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00013862568037415423,
+      "loss": 1.6044,
+      "step": 1868
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00013856737900754584,
+      "loss": 1.8852,
+      "step": 1869
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.0001385090622369793,
+      "loss": 1.4336,
+      "step": 1870
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00013845073008574653,
+      "loss": 1.568,
+      "step": 1871
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00013839238257714568,
+      "loss": 1.659,
+      "step": 1872
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00013833401973448083,
+      "loss": 1.3567,
+      "step": 1873
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00013827564158106243,
+      "loss": 1.5852,
+      "step": 1874
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013821724814020694,
+      "loss": 1.4847,
+      "step": 1875
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00013815883943523683,
+      "loss": 1.4427,
+      "step": 1876
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00013810041548948084,
+      "loss": 1.4376,
+      "step": 1877
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00013804197632627375,
+      "loss": 1.5352,
+      "step": 1878
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00013798352196895628,
+      "loss": 1.5108,
+      "step": 1879
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00013792505244087538,
+      "loss": 1.5876,
+      "step": 1880
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.625,
+      "learning_rate": 0.000137866567765384,
+      "loss": 1.5049,
+      "step": 1881
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013780806796584113,
+      "loss": 1.4406,
+      "step": 1882
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00013774955306561183,
+      "loss": 1.7154,
+      "step": 1883
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00013769102308806716,
+      "loss": 1.6187,
+      "step": 1884
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013763247805658425,
+      "loss": 1.6808,
+      "step": 1885
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00013757391799454615,
+      "loss": 1.519,
+      "step": 1886
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.000137515342925342,
+      "loss": 1.561,
+      "step": 1887
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013745675287236695,
+      "loss": 1.8332,
+      "step": 1888
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.000137398147859022,
+      "loss": 1.7569,
+      "step": 1889
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.00013733952790871427,
+      "loss": 1.3986,
+      "step": 1890
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.0001372808930448568,
+      "loss": 1.811,
+      "step": 1891
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00013722224329086857,
+      "loss": 1.8258,
+      "step": 1892
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00013716357867017448,
+      "loss": 1.6126,
+      "step": 1893
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00013710489920620546,
+      "loss": 1.2489,
+      "step": 1894
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.0001370462049223983,
+      "loss": 1.4081,
+      "step": 1895
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00013698749584219565,
+      "loss": 1.4634,
+      "step": 1896
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00013692877198904623,
+      "loss": 1.5778,
+      "step": 1897
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00013687003338640454,
+      "loss": 1.5958,
+      "step": 1898
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.000136811280057731,
+      "loss": 1.6801,
+      "step": 1899
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.0001367525120264919,
+      "loss": 1.8658,
+      "step": 1900
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00013669372931615943,
+      "loss": 1.3611,
+      "step": 1901
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.0001366349319502116,
+      "loss": 1.6402,
+      "step": 1902
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00013657611995213236,
+      "loss": 1.4475,
+      "step": 1903
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00013651729334541142,
+      "loss": 1.4336,
+      "step": 1904
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00013645845215354432,
+      "loss": 1.5915,
+      "step": 1905
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00013639959640003252,
+      "loss": 1.5401,
+      "step": 1906
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013634072610838312,
+      "loss": 1.1419,
+      "step": 1907
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00013628184130210923,
+      "loss": 1.564,
+      "step": 1908
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013622294200472972,
+      "loss": 1.3806,
+      "step": 1909
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00013616402823976906,
+      "loss": 1.7523,
+      "step": 1910
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00013610510003075768,
+      "loss": 1.5352,
+      "step": 1911
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.625,
+      "learning_rate": 0.0001360461574012318,
+      "loss": 1.3683,
+      "step": 1912
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013598720037473323,
+      "loss": 1.4884,
+      "step": 1913
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001359282289748097,
+      "loss": 1.5263,
+      "step": 1914
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.0001358692432250146,
+      "loss": 1.5917,
+      "step": 1915
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.000135810243148907,
+      "loss": 1.6214,
+      "step": 1916
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013575122877005187,
+      "loss": 1.6158,
+      "step": 1917
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00013569220011201968,
+      "loss": 1.4017,
+      "step": 1918
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00013563315719838673,
+      "loss": 1.4559,
+      "step": 1919
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00013557410005273502,
+      "loss": 1.4683,
+      "step": 1920
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00013551502869865213,
+      "loss": 1.3481,
+      "step": 1921
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00013545594315973145,
+      "loss": 1.6084,
+      "step": 1922
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00013539684345957196,
+      "loss": 1.6497,
+      "step": 1923
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00013533772962177827,
+      "loss": 1.426,
+      "step": 1924
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00013527860166996067,
+      "loss": 1.3614,
+      "step": 1925
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00013521945962773516,
+      "loss": 1.5095,
+      "step": 1926
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00013516030351872323,
+      "loss": 1.4176,
+      "step": 1927
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00013510113336655203,
+      "loss": 1.4831,
+      "step": 1928
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.0001350419491948544,
+      "loss": 1.4125,
+      "step": 1929
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00013498275102726877,
+      "loss": 1.6879,
+      "step": 1930
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.75,
+      "learning_rate": 0.000134923538887439,
+      "loss": 1.5688,
+      "step": 1931
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00013486431279901472,
+      "loss": 1.7968,
+      "step": 1932
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.000134805072785651,
+      "loss": 1.5781,
+      "step": 1933
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00013474581887100862,
+      "loss": 1.4325,
+      "step": 1934
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00013468655107875375,
+      "loss": 1.3692,
+      "step": 1935
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00013462726943255814,
+      "loss": 1.5973,
+      "step": 1936
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00013456797395609916,
+      "loss": 1.2605,
+      "step": 1937
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00013450866467305968,
+      "loss": 1.8224,
+      "step": 1938
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00013444934160712797,
+      "loss": 1.7833,
+      "step": 1939
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00013439000478199795,
+      "loss": 1.6441,
+      "step": 1940
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00013433065422136892,
+      "loss": 1.5788,
+      "step": 1941
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00013427128994894578,
+      "loss": 1.3716,
+      "step": 1942
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00013421191198843883,
+      "loss": 1.4551,
+      "step": 1943
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00013415252036356383,
+      "loss": 1.7827,
+      "step": 1944
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00013409311509804205,
+      "loss": 1.3219,
+      "step": 1945
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00013403369621560013,
+      "loss": 1.7935,
+      "step": 1946
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00013397426373997027,
+      "loss": 1.8412,
+      "step": 1947
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00013391481769489,
+      "loss": 1.6615,
+      "step": 1948
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00013385535810410222,
+      "loss": 1.7211,
+      "step": 1949
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00013379588499135546,
+      "loss": 1.7281,
+      "step": 1950
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.0001337363983804034,
+      "loss": 1.2451,
+      "step": 1951
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00013367689829500525,
+      "loss": 1.4263,
+      "step": 1952
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00013361738475892558,
+      "loss": 1.6272,
+      "step": 1953
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.0001335578577959343,
+      "loss": 1.6434,
+      "step": 1954
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013349831742980674,
+      "loss": 1.5738,
+      "step": 1955
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.0001334387636843235,
+      "loss": 1.7193,
+      "step": 1956
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00013337919658327064,
+      "loss": 1.3047,
+      "step": 1957
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00013331961615043944,
+      "loss": 1.3131,
+      "step": 1958
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00013326002240962657,
+      "loss": 1.477,
+      "step": 1959
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00013320041538463398,
+      "loss": 1.7461,
+      "step": 1960
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00013314079509926897,
+      "loss": 1.6009,
+      "step": 1961
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00013308116157734407,
+      "loss": 1.3734,
+      "step": 1962
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00013302151484267718,
+      "loss": 1.8229,
+      "step": 1963
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.0001329618549190914,
+      "loss": 1.3454,
+      "step": 1964
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00013290218183041513,
+      "loss": 1.5958,
+      "step": 1965
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00013284249560048205,
+      "loss": 1.4851,
+      "step": 1966
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00013278279625313105,
+      "loss": 1.196,
+      "step": 1967
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00013272308381220632,
+      "loss": 1.4285,
+      "step": 1968
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00013266335830155717,
+      "loss": 1.5367,
+      "step": 1969
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0001326036197450382,
+      "loss": 1.5826,
+      "step": 1970
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00013254386816650933,
+      "loss": 1.3755,
+      "step": 1971
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.0001324841035898354,
+      "loss": 1.6814,
+      "step": 1972
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00013242432603888674,
+      "loss": 1.4604,
+      "step": 1973
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00013236453553753868,
+      "loss": 1.5471,
+      "step": 1974
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00013230473210967177,
+      "loss": 1.564,
+      "step": 1975
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00013224491577917176,
+      "loss": 1.637,
+      "step": 1976
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013218508656992947,
+      "loss": 1.4132,
+      "step": 1977
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00013212524450584095,
+      "loss": 1.7957,
+      "step": 1978
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013206538961080735,
+      "loss": 1.3189,
+      "step": 1979
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00013200552190873493,
+      "loss": 1.6487,
+      "step": 1980
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00013194564142353504,
+      "loss": 1.2718,
+      "step": 1981
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00013188574817912427,
+      "loss": 1.5104,
+      "step": 1982
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.0001318258421994241,
+      "loss": 1.5476,
+      "step": 1983
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001317659235083613,
+      "loss": 1.7173,
+      "step": 1984
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00013170599212986757,
+      "loss": 1.1228,
+      "step": 1985
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00013164604808787978,
+      "loss": 1.5552,
+      "step": 1986
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00013158609140633974,
+      "loss": 1.4771,
+      "step": 1987
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00013152612210919445,
+      "loss": 1.2489,
+      "step": 1988
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00013146614022039582,
+      "loss": 1.5803,
+      "step": 1989
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00013140614576390082,
+      "loss": 1.7512,
+      "step": 1990
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00013134613876367155,
+      "loss": 1.4342,
+      "step": 1991
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.796875,
+      "learning_rate": 0.000131286119243675,
+      "loss": 1.5986,
+      "step": 1992
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00013122608722788317,
+      "loss": 1.2769,
+      "step": 1993
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.0001311660427402731,
+      "loss": 1.5813,
+      "step": 1994
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00013110598580482678,
+      "loss": 1.3508,
+      "step": 1995
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00013104591644553116,
+      "loss": 1.5503,
+      "step": 1996
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00013098583468637825,
+      "loss": 1.6593,
+      "step": 1997
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00013092574055136483,
+      "loss": 1.6981,
+      "step": 1998
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00013086563406449275,
+      "loss": 1.2586,
+      "step": 1999
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00013080551524976883,
+      "loss": 1.4651,
+      "step": 2000
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00013074538413120466,
+      "loss": 1.5574,
+      "step": 2001
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00013068524073281697,
+      "loss": 1.4622,
+      "step": 2002
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00013062508507862713,
+      "loss": 1.5642,
+      "step": 2003
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.0001305649171926616,
+      "loss": 1.5768,
+      "step": 2004
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00013050473709895165,
+      "loss": 1.4917,
+      "step": 2005
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.00013044454482153344,
+      "loss": 1.5763,
+      "step": 2006
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00013038434038444808,
+      "loss": 1.4844,
+      "step": 2007
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.0001303241238117413,
+      "loss": 1.5353,
+      "step": 2008
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00013026389512746386,
+      "loss": 1.5092,
+      "step": 2009
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013020365435567144,
+      "loss": 1.6135,
+      "step": 2010
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00013014340152042432,
+      "loss": 1.6704,
+      "step": 2011
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00013008313664578775,
+      "loss": 1.575,
+      "step": 2012
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00013002285975583175,
+      "loss": 1.4537,
+      "step": 2013
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00012996257087463116,
+      "loss": 1.6171,
+      "step": 2014
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00012990227002626555,
+      "loss": 1.3887,
+      "step": 2015
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00012984195723481933,
+      "loss": 1.7258,
+      "step": 2016
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00012978163252438168,
+      "loss": 1.6075,
+      "step": 2017
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.0001297212959190465,
+      "loss": 1.2736,
+      "step": 2018
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00012966094744291243,
+      "loss": 1.5524,
+      "step": 2019
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00012960058712008294,
+      "loss": 1.3697,
+      "step": 2020
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00012954021497466612,
+      "loss": 1.3618,
+      "step": 2021
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00012947983103077485,
+      "loss": 1.7185,
+      "step": 2022
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00012941943531252675,
+      "loss": 1.2999,
+      "step": 2023
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.000129359027844044,
+      "loss": 1.3469,
+      "step": 2024
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001292986086494537,
+      "loss": 1.4561,
+      "step": 2025
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.0001292381777528874,
+      "loss": 1.8479,
+      "step": 2026
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.65625,
+      "learning_rate": 0.00012917773517848147,
+      "loss": 1.3257,
+      "step": 2027
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00012911728095037697,
+      "loss": 1.507,
+      "step": 2028
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00012905681509271943,
+      "loss": 1.7558,
+      "step": 2029
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.0001289963376296592,
+      "loss": 1.5343,
+      "step": 2030
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00012893584858535126,
+      "loss": 1.6736,
+      "step": 2031
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00012887534798395508,
+      "loss": 1.6803,
+      "step": 2032
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00012881483584963488,
+      "loss": 1.7301,
+      "step": 2033
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.0001287543122065594,
+      "loss": 1.5592,
+      "step": 2034
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00012869377707890208,
+      "loss": 1.4695,
+      "step": 2035
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00012863323049084078,
+      "loss": 1.5463,
+      "step": 2036
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00012857267246655813,
+      "loss": 1.3655,
+      "step": 2037
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001285121030302412,
+      "loss": 1.4866,
+      "step": 2038
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00012845152220608167,
+      "loss": 1.5475,
+      "step": 2039
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00012839093001827572,
+      "loss": 1.3771,
+      "step": 2040
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.0001283303264910241,
+      "loss": 1.4456,
+      "step": 2041
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00012826971164853215,
+      "loss": 1.4213,
+      "step": 2042
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00012820908551500958,
+      "loss": 1.4646,
+      "step": 2043
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00012814844811467077,
+      "loss": 1.5628,
+      "step": 2044
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00012808779947173448,
+      "loss": 1.5391,
+      "step": 2045
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00012802713961042406,
+      "loss": 1.4491,
+      "step": 2046
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00012796646855496726,
+      "loss": 1.4768,
+      "step": 2047
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00012790578632959634,
+      "loss": 1.332,
+      "step": 2048
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.000127845092958548,
+      "loss": 1.6062,
+      "step": 2049
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00012778438846606342,
+      "loss": 1.3506,
+      "step": 2050
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00012772367287638818,
+      "loss": 1.4466,
+      "step": 2051
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00012766294621377236,
+      "loss": 1.3982,
+      "step": 2052
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00012760220850247035,
+      "loss": 1.1691,
+      "step": 2053
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.0001275414597667411,
+      "loss": 1.7725,
+      "step": 2054
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00012748070003084784,
+      "loss": 1.5578,
+      "step": 2055
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00012741992931905827,
+      "loss": 1.4641,
+      "step": 2056
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00012735914765564442,
+      "loss": 1.3818,
+      "step": 2057
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00012729835506488271,
+      "loss": 1.4009,
+      "step": 2058
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00012723755157105398,
+      "loss": 1.4448,
+      "step": 2059
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00012717673719844333,
+      "loss": 1.4723,
+      "step": 2060
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00012711591197134026,
+      "loss": 1.7235,
+      "step": 2061
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00012705507591403861,
+      "loss": 1.4035,
+      "step": 2062
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00012699422905083652,
+      "loss": 1.4995,
+      "step": 2063
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00012693337140603646,
+      "loss": 1.4796,
+      "step": 2064
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00012687250300394526,
+      "loss": 1.6539,
+      "step": 2065
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.0001268116238688739,
+      "loss": 1.1813,
+      "step": 2066
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.0001267507340251378,
+      "loss": 1.5284,
+      "step": 2067
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001266898334970566,
+      "loss": 1.8136,
+      "step": 2068
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00012662892230895414,
+      "loss": 1.7365,
+      "step": 2069
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00012656800048515865,
+      "loss": 1.6054,
+      "step": 2070
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001265070680500025,
+      "loss": 1.4294,
+      "step": 2071
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00012644612502782233,
+      "loss": 1.4559,
+      "step": 2072
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00012638517144295908,
+      "loss": 1.6608,
+      "step": 2073
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00012632420731975775,
+      "loss": 1.4444,
+      "step": 2074
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001262632326825677,
+      "loss": 1.7222,
+      "step": 2075
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.0001262022475557424,
+      "loss": 1.5385,
+      "step": 2076
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00012614125196363962,
+      "loss": 1.8211,
+      "step": 2077
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00012608024593062115,
+      "loss": 1.6859,
+      "step": 2078
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00012601922948105306,
+      "loss": 1.4273,
+      "step": 2079
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00012595820263930558,
+      "loss": 1.6874,
+      "step": 2080
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00012589716542975302,
+      "loss": 1.2748,
+      "step": 2081
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00012583611787677385,
+      "loss": 1.5425,
+      "step": 2082
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.0001257750600047508,
+      "loss": 1.7337,
+      "step": 2083
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00012571399183807048,
+      "loss": 1.6346,
+      "step": 2084
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00012565291340112388,
+      "loss": 1.4884,
+      "step": 2085
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00012559182471830587,
+      "loss": 1.5148,
+      "step": 2086
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00012553072581401552,
+      "loss": 1.4582,
+      "step": 2087
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00012546961671265603,
+      "loss": 1.382,
+      "step": 2088
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001254084974386345,
+      "loss": 1.4628,
+      "step": 2089
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00012534736801636228,
+      "loss": 1.2909,
+      "step": 2090
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00012528622847025466,
+      "loss": 1.4144,
+      "step": 2091
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00012522507882473103,
+      "loss": 1.3585,
+      "step": 2092
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.0001251639191042148,
+      "loss": 1.7652,
+      "step": 2093
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00012510274933313334,
+      "loss": 1.4428,
+      "step": 2094
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00012504156953591813,
+      "loss": 1.6217,
+      "step": 2095
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00012498037973700463,
+      "loss": 1.5114,
+      "step": 2096
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00012491917996083223,
+      "loss": 1.3822,
+      "step": 2097
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001248579702318444,
+      "loss": 1.5282,
+      "step": 2098
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00012479675057448854,
+      "loss": 1.2801,
+      "step": 2099
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00012473552101321597,
+      "loss": 1.6763,
+      "step": 2100
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.546875,
+      "learning_rate": 0.00012467428157248204,
+      "loss": 1.7147,
+      "step": 2101
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00012461303227674604,
+      "loss": 1.3254,
+      "step": 2102
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00012455177315047112,
+      "loss": 1.5197,
+      "step": 2103
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00012449050421812442,
+      "loss": 1.689,
+      "step": 2104
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00012442922550417703,
+      "loss": 1.5422,
+      "step": 2105
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00012436793703310382,
+      "loss": 1.7926,
+      "step": 2106
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5,
+      "learning_rate": 0.0001243066388293837,
+      "loss": 1.6835,
+      "step": 2107
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00012424533091749942,
+      "loss": 1.8023,
+      "step": 2108
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00012418401332193758,
+      "loss": 1.4996,
+      "step": 2109
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00012412268606718863,
+      "loss": 1.4759,
+      "step": 2110
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00012406134917774693,
+      "loss": 1.2546,
+      "step": 2111
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00012400000267811064,
+      "loss": 1.5952,
+      "step": 2112
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00012393864659278184,
+      "loss": 1.6716,
+      "step": 2113
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00012387728094626638,
+      "loss": 1.4333,
+      "step": 2114
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00012381590576307388,
+      "loss": 1.2656,
+      "step": 2115
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00012375452106771784,
+      "loss": 1.4614,
+      "step": 2116
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.0001236931268847156,
+      "loss": 1.2243,
+      "step": 2117
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00012363172323858813,
+      "loss": 1.578,
+      "step": 2118
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00012357031015386037,
+      "loss": 1.6578,
+      "step": 2119
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.0001235088876550609,
+      "loss": 1.6687,
+      "step": 2120
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001234474557667221,
+      "loss": 1.3929,
+      "step": 2121
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.0001233860145133801,
+      "loss": 1.6473,
+      "step": 2122
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.0001233245639195748,
+      "loss": 1.5477,
+      "step": 2123
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00012326310400984975,
+      "loss": 1.5484,
+      "step": 2124
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00012320163480875228,
+      "loss": 1.5598,
+      "step": 2125
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00012314015634083347,
+      "loss": 1.3159,
+      "step": 2126
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00012307866863064804,
+      "loss": 1.7599,
+      "step": 2127
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00012301717170275437,
+      "loss": 1.7255,
+      "step": 2128
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.0001229556655817146,
+      "loss": 1.41,
+      "step": 2129
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00012289415029209454,
+      "loss": 1.6789,
+      "step": 2130
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00012283262585846357,
+      "loss": 1.6258,
+      "step": 2131
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.875,
+      "learning_rate": 0.00012277109230539483,
+      "loss": 1.545,
+      "step": 2132
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00012270954965746503,
+      "loss": 1.5176,
+      "step": 2133
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.390625,
+      "learning_rate": 0.00012264799793925454,
+      "loss": 1.6878,
+      "step": 2134
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00012258643717534738,
+      "loss": 1.6104,
+      "step": 2135
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.0001225248673903311,
+      "loss": 1.3761,
+      "step": 2136
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00012246328860879697,
+      "loss": 1.7188,
+      "step": 2137
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00012240170085533976,
+      "loss": 1.4239,
+      "step": 2138
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.0001223401041545578,
+      "loss": 1.2726,
+      "step": 2139
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.0001222784985310532,
+      "loss": 1.7411,
+      "step": 2140
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00012221688400943127,
+      "loss": 1.5383,
+      "step": 2141
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00012215526061430123,
+      "loss": 1.4139,
+      "step": 2142
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.0001220936283702757,
+      "loss": 1.6433,
+      "step": 2143
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00012203198730197077,
+      "loss": 1.4498,
+      "step": 2144
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00012197033743400614,
+      "loss": 1.3799,
+      "step": 2145
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.000121908678791005,
+      "loss": 1.5357,
+      "step": 2146
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00012184701139759406,
+      "loss": 1.4376,
+      "step": 2147
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00012178533527840352,
+      "loss": 1.4482,
+      "step": 2148
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00012172365045806702,
+      "loss": 1.7115,
+      "step": 2149
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00012166195696122176,
+      "loss": 1.6967,
+      "step": 2150
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00012160025481250832,
+      "loss": 1.567,
+      "step": 2151
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00012153854403657077,
+      "loss": 1.6934,
+      "step": 2152
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00012147682465805667,
+      "loss": 1.2456,
+      "step": 2153
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00012141509670161693,
+      "loss": 1.212,
+      "step": 2154
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00012135336019190595,
+      "loss": 1.729,
+      "step": 2155
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00012129161515358156,
+      "loss": 1.5271,
+      "step": 2156
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.0001212298616113049,
+      "loss": 1.5194,
+      "step": 2157
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00012116809958974062,
+      "loss": 1.5516,
+      "step": 2158
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00012110632911355665,
+      "loss": 1.5902,
+      "step": 2159
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00012104455020742442,
+      "loss": 1.6438,
+      "step": 2160
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00012098276289601867,
+      "loss": 1.53,
+      "step": 2161
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.875,
+      "learning_rate": 0.0001209209672040174,
+      "loss": 1.6106,
+      "step": 2162
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.0001208591631561021,
+      "loss": 1.7666,
+      "step": 2163
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00012079735077695753,
+      "loss": 1.496,
+      "step": 2164
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00012073553009127184,
+      "loss": 1.304,
+      "step": 2165
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00012067370112373637,
+      "loss": 1.4388,
+      "step": 2166
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00012061186389904589,
+      "loss": 1.5301,
+      "step": 2167
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00012055001844189839,
+      "loss": 1.4096,
+      "step": 2168
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00012048816477699525,
+      "loss": 1.6525,
+      "step": 2169
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00012042630292904098,
+      "loss": 1.5326,
+      "step": 2170
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00012036443292274351,
+      "loss": 1.4422,
+      "step": 2171
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00012030255478281387,
+      "loss": 1.7764,
+      "step": 2172
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.84375,
+      "learning_rate": 0.0001202406685339665,
+      "loss": 1.5154,
+      "step": 2173
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00012017877420091899,
+      "loss": 1.3034,
+      "step": 2174
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00012011687180839213,
+      "loss": 1.5046,
+      "step": 2175
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00012005496138111003,
+      "loss": 1.7145,
+      "step": 2176
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.0001199930429437999,
+      "loss": 1.6015,
+      "step": 2177
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.0001199311165211922,
+      "loss": 1.3642,
+      "step": 2178
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00011986918213802065,
+      "loss": 1.5887,
+      "step": 2179
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.000119807239819022,
+      "loss": 1.4361,
+      "step": 2180
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00011974528958893627,
+      "loss": 1.633,
+      "step": 2181
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00011968333147250665,
+      "loss": 1.5969,
+      "step": 2182
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00011962136549447942,
+      "loss": 1.7471,
+      "step": 2183
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00011955939167960402,
+      "loss": 1.3708,
+      "step": 2184
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00011949741005263304,
+      "loss": 1.435,
+      "step": 2185
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00011943542063832219,
+      "loss": 1.7387,
+      "step": 2186
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00011937342346143025,
+      "loss": 1.4209,
+      "step": 2187
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00011931141854671914,
+      "loss": 1.4455,
+      "step": 2188
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00011924940591895388,
+      "loss": 1.2709,
+      "step": 2189
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00011918738560290252,
+      "loss": 1.5354,
+      "step": 2190
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00011912535762333623,
+      "loss": 1.2983,
+      "step": 2191
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00011906332200502923,
+      "loss": 1.4244,
+      "step": 2192
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00011900127877275875,
+      "loss": 1.3919,
+      "step": 2193
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001189392279513051,
+      "loss": 1.3917,
+      "step": 2194
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00011887716956545163,
+      "loss": 1.3867,
+      "step": 2195
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00011881510363998469,
+      "loss": 1.2719,
+      "step": 2196
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00011875303019969368,
+      "loss": 1.606,
+      "step": 2197
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.0001186909492693709,
+      "loss": 1.614,
+      "step": 2198
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00011862886087381175,
+      "loss": 1.6384,
+      "step": 2199
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00011856676503781461,
+      "loss": 1.608,
+      "step": 2200
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00011850466178618071,
+      "loss": 1.6362,
+      "step": 2201
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00011844255114371441,
+      "loss": 1.2902,
+      "step": 2202
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00011838043313522287,
+      "loss": 1.5574,
+      "step": 2203
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00011831830778551627,
+      "loss": 1.7973,
+      "step": 2204
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00011825617511940775,
+      "loss": 1.6461,
+      "step": 2205
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.0001181940351617133,
+      "loss": 1.4618,
+      "step": 2206
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00011813188793725185,
+      "loss": 1.6983,
+      "step": 2207
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.0001180697334708453,
+      "loss": 1.5516,
+      "step": 2208
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.375,
+      "learning_rate": 0.0001180075717873183,
+      "loss": 1.7152,
+      "step": 2209
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00011794540291149851,
+      "loss": 1.3895,
+      "step": 2210
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00011788322686821637,
+      "loss": 1.4676,
+      "step": 2211
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00011782104368230529,
+      "loss": 1.4895,
+      "step": 2212
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00011775885337860146,
+      "loss": 1.452,
+      "step": 2213
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.0001176966559819439,
+      "loss": 1.3755,
+      "step": 2214
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.0001176344515171745,
+      "loss": 1.6941,
+      "step": 2215
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00011757224000913794,
+      "loss": 1.48,
+      "step": 2216
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00011751002148268175,
+      "loss": 1.5548,
+      "step": 2217
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00011744779596265628,
+      "loss": 1.6518,
+      "step": 2218
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00011738556347391456,
+      "loss": 1.8347,
+      "step": 2219
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00011732332404131252,
+      "loss": 1.5617,
+      "step": 2220
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00011726107768970889,
+      "loss": 1.4437,
+      "step": 2221
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00011719882444396501,
+      "loss": 1.4378,
+      "step": 2222
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.0001171365643289451,
+      "loss": 1.5058,
+      "step": 2223
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.00011707429736951612,
+      "loss": 1.5916,
+      "step": 2224
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.84375,
+      "learning_rate": 0.00011701202359054766,
+      "loss": 1.4509,
+      "step": 2225
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00011694974301691215,
+      "loss": 1.6898,
+      "step": 2226
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00011688745567348467,
+      "loss": 1.4438,
+      "step": 2227
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00011682516158514305,
+      "loss": 1.779,
+      "step": 2228
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00011676286077676777,
+      "loss": 1.5184,
+      "step": 2229
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.000116700553273242,
+      "loss": 1.3374,
+      "step": 2230
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00011663823909945161,
+      "loss": 1.5393,
+      "step": 2231
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00011657591828028512,
+      "loss": 1.6542,
+      "step": 2232
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00011651359084063369,
+      "loss": 1.3663,
+      "step": 2233
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00011645125680539115,
+      "loss": 1.2715,
+      "step": 2234
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00011638891619945393,
+      "loss": 1.534,
+      "step": 2235
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00011632656904772112,
+      "loss": 1.5097,
+      "step": 2236
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00011626421537509442,
+      "loss": 1.7112,
+      "step": 2237
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00011620185520647811,
+      "loss": 1.4953,
+      "step": 2238
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00011613948856677908,
+      "loss": 1.6048,
+      "step": 2239
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00011607711548090681,
+      "loss": 1.5341,
+      "step": 2240
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00011601473597377334,
+      "loss": 1.5615,
+      "step": 2241
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00011595235007029329,
+      "loss": 1.452,
+      "step": 2242
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.75,
+      "learning_rate": 0.0001158899577953838,
+      "loss": 1.627,
+      "step": 2243
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00011582755917396463,
+      "loss": 1.3372,
+      "step": 2244
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00011576515423095802,
+      "loss": 1.6788,
+      "step": 2245
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00011570274299128869,
+      "loss": 1.775,
+      "step": 2246
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.000115640325479884,
+      "loss": 1.4801,
+      "step": 2247
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00011557790172167368,
+      "loss": 1.5731,
+      "step": 2248
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00011551547174159005,
+      "loss": 1.609,
+      "step": 2249
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00011545303556456789,
+      "loss": 1.5472,
+      "step": 2250
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.0001153905932155444,
+      "loss": 1.6302,
+      "step": 2251
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00011532814471945936,
+      "loss": 1.4858,
+      "step": 2252
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00011526569010125487,
+      "loss": 1.3376,
+      "step": 2253
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00011520322938587557,
+      "loss": 1.3162,
+      "step": 2254
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00011514076259826851,
+      "loss": 1.5519,
+      "step": 2255
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.0001150782897633832,
+      "loss": 1.5024,
+      "step": 2256
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00011501581090617146,
+      "loss": 1.6532,
+      "step": 2257
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00011495332605158763,
+      "loss": 1.737,
+      "step": 2258
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00011489083522458836,
+      "loss": 1.3395,
+      "step": 2259
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.00011482833845013281,
+      "loss": 1.605,
+      "step": 2260
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00011476583575318232,
+      "loss": 1.3863,
+      "step": 2261
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00011470332715870078,
+      "loss": 1.4911,
+      "step": 2262
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00011464081269165432,
+      "loss": 1.5394,
+      "step": 2263
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00011457829237701148,
+      "loss": 1.7868,
+      "step": 2264
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001145157662397431,
+      "loss": 1.7667,
+      "step": 2265
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00011445323430482235,
+      "loss": 1.4804,
+      "step": 2266
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00011439069659722476,
+      "loss": 1.4839,
+      "step": 2267
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00011432815314192809,
+      "loss": 1.4968,
+      "step": 2268
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00011426560396391244,
+      "loss": 1.5237,
+      "step": 2269
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00011420304908816022,
+      "loss": 1.6464,
+      "step": 2270
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00011414048853965604,
+      "loss": 1.5512,
+      "step": 2271
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.00011407792234338686,
+      "loss": 1.223,
+      "step": 2272
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00011401535052434185,
+      "loss": 1.5734,
+      "step": 2273
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00011395277310751243,
+      "loss": 1.3719,
+      "step": 2274
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00011389019011789224,
+      "loss": 1.44,
+      "step": 2275
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00011382760158047716,
+      "loss": 1.3845,
+      "step": 2276
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00011376500752026533,
+      "loss": 1.3274,
+      "step": 2277
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00011370240796225704,
+      "loss": 1.314,
+      "step": 2278
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00011363980293145478,
+      "loss": 1.6629,
+      "step": 2279
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00011357719245286327,
+      "loss": 1.7201,
+      "step": 2280
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00011351457655148931,
+      "loss": 1.6326,
+      "step": 2281
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00011345195525234199,
+      "loss": 1.2593,
+      "step": 2282
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00011338932858043246,
+      "loss": 1.6433,
+      "step": 2283
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00011332669656077409,
+      "loss": 1.365,
+      "step": 2284
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.0001132640592183823,
+      "loss": 1.3814,
+      "step": 2285
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.0001132014165782747,
+      "loss": 1.6316,
+      "step": 2286
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00011313876866547102,
+      "loss": 1.6266,
+      "step": 2287
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00011307611550499309,
+      "loss": 1.4714,
+      "step": 2288
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00011301345712186476,
+      "loss": 1.5148,
+      "step": 2289
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00011295079354111205,
+      "loss": 1.8258,
+      "step": 2290
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00011288812478776304,
+      "loss": 1.3397,
+      "step": 2291
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00011282545088684791,
+      "loss": 1.3863,
+      "step": 2292
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00011276277186339879,
+      "loss": 1.6747,
+      "step": 2293
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00011270008774244996,
+      "loss": 1.5575,
+      "step": 2294
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00011263739854903766,
+      "loss": 1.6463,
+      "step": 2295
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00011257470430820023,
+      "loss": 1.4906,
+      "step": 2296
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00011251200504497797,
+      "loss": 1.5387,
+      "step": 2297
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.0001124493007844132,
+      "loss": 1.4031,
+      "step": 2298
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00011238659155155029,
+      "loss": 1.6981,
+      "step": 2299
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00011232387737143546,
+      "loss": 1.7024,
+      "step": 2300
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00011226115826911709,
+      "loss": 1.4376,
+      "step": 2301
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00011219843426964534,
+      "loss": 1.7225,
+      "step": 2302
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.09375,
+      "learning_rate": 0.00011213570539807246,
+      "loss": 1.1501,
+      "step": 2303
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00011207297167945259,
+      "loss": 1.6787,
+      "step": 2304
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00011201023313884183,
+      "loss": 1.5954,
+      "step": 2305
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00011194748980129818,
+      "loss": 1.2664,
+      "step": 2306
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00011188474169188156,
+      "loss": 1.6192,
+      "step": 2307
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5,
+      "learning_rate": 0.0001118219888356538,
+      "loss": 1.5462,
+      "step": 2308
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00011175923125767868,
+      "loss": 1.3433,
+      "step": 2309
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00011169646898302175,
+      "loss": 1.3757,
+      "step": 2310
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00011163370203675054,
+      "loss": 1.6943,
+      "step": 2311
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.0001115709304439344,
+      "loss": 1.3815,
+      "step": 2312
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00011150815422964454,
+      "loss": 1.1371,
+      "step": 2313
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00011144537341895403,
+      "loss": 1.5735,
+      "step": 2314
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00011138258803693773,
+      "loss": 1.4415,
+      "step": 2315
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00011131979810867241,
+      "loss": 1.1199,
+      "step": 2316
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00011125700365923659,
+      "loss": 1.4228,
+      "step": 2317
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00011119420471371058,
+      "loss": 1.4015,
+      "step": 2318
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00011113140129717656,
+      "loss": 1.5112,
+      "step": 2319
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00011106859343471845,
+      "loss": 1.6245,
+      "step": 2320
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.0001110057811514219,
+      "loss": 1.4159,
+      "step": 2321
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00011094296447237447,
+      "loss": 1.6261,
+      "step": 2322
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00011088014342266529,
+      "loss": 1.6812,
+      "step": 2323
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00011081731802738537,
+      "loss": 1.5678,
+      "step": 2324
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00011075448831162738,
+      "loss": 1.6353,
+      "step": 2325
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00011069165430048578,
+      "loss": 1.6525,
+      "step": 2326
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00011062881601905674,
+      "loss": 1.3345,
+      "step": 2327
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00011056597349243805,
+      "loss": 1.4905,
+      "step": 2328
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00011050312674572925,
+      "loss": 1.7742,
+      "step": 2329
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00011044027580403165,
+      "loss": 1.4913,
+      "step": 2330
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00011037742069244807,
+      "loss": 1.4412,
+      "step": 2331
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00011031456143608315,
+      "loss": 1.3885,
+      "step": 2332
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.00011025169806004302,
+      "loss": 1.4994,
+      "step": 2333
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.3125,
+      "learning_rate": 0.00011018883058943566,
+      "loss": 1.5274,
+      "step": 2334
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00011012595904937053,
+      "loss": 1.5813,
+      "step": 2335
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00011006308346495875,
+      "loss": 1.551,
+      "step": 2336
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00011000020386131309,
+      "loss": 1.5399,
+      "step": 2337
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.00010993732026354789,
+      "loss": 1.6653,
+      "step": 2338
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00010987443269677911,
+      "loss": 1.5154,
+      "step": 2339
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00010981154118612427,
+      "loss": 1.3334,
+      "step": 2340
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001097486457567025,
+      "loss": 1.6552,
+      "step": 2341
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00010968574643363443,
+      "loss": 1.4482,
+      "step": 2342
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00010962284324204236,
+      "loss": 1.6985,
+      "step": 2343
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00010955993620705001,
+      "loss": 1.4766,
+      "step": 2344
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00010949702535378268,
+      "loss": 1.3386,
+      "step": 2345
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00010943411070736725,
+      "loss": 1.2108,
+      "step": 2346
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00010937119229293207,
+      "loss": 1.5196,
+      "step": 2347
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00010930827013560694,
+      "loss": 1.1567,
+      "step": 2348
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00010924534426052325,
+      "loss": 1.5481,
+      "step": 2349
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00010918241469281383,
+      "loss": 1.3404,
+      "step": 2350
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00010911948145761299,
+      "loss": 1.6599,
+      "step": 2351
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00010905654458005647,
+      "loss": 1.3293,
+      "step": 2352
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00010899360408528158,
+      "loss": 1.2478,
+      "step": 2353
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00010893065999842691,
+      "loss": 1.7218,
+      "step": 2354
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.0001088677123446326,
+      "loss": 1.4803,
+      "step": 2355
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00010880476114904021,
+      "loss": 1.8836,
+      "step": 2356
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.00010874180643679264,
+      "loss": 1.5486,
+      "step": 2357
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00010867884823303427,
+      "loss": 1.4991,
+      "step": 2358
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00010861588656291082,
+      "loss": 1.3809,
+      "step": 2359
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00010855292145156945,
+      "loss": 1.4535,
+      "step": 2360
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00010848995292415866,
+      "loss": 1.6257,
+      "step": 2361
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00010842698100582834,
+      "loss": 1.4606,
+      "step": 2362
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00010836400572172967,
+      "loss": 1.1892,
+      "step": 2363
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00010830102709701522,
+      "loss": 1.7337,
+      "step": 2364
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00010823804515683891,
+      "loss": 1.5159,
+      "step": 2365
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00010817505992635599,
+      "loss": 1.3079,
+      "step": 2366
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00010811207143072297,
+      "loss": 1.8178,
+      "step": 2367
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.0001080490796950977,
+      "loss": 1.5584,
+      "step": 2368
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00010798608474463933,
+      "loss": 1.1467,
+      "step": 2369
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00010792308660450828,
+      "loss": 1.3158,
+      "step": 2370
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00010786008529986624,
+      "loss": 1.6642,
+      "step": 2371
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00010779708085587616,
+      "loss": 1.554,
+      "step": 2372
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00010773407329770228,
+      "loss": 1.3028,
+      "step": 2373
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00010767106265051006,
+      "loss": 1.3459,
+      "step": 2374
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00010760804893946619,
+      "loss": 1.3061,
+      "step": 2375
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00010754503218973855,
+      "loss": 1.0934,
+      "step": 2376
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00010748201242649629,
+      "loss": 1.5289,
+      "step": 2377
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00010741898967490978,
+      "loss": 1.7969,
+      "step": 2378
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.0001073559639601505,
+      "loss": 1.6482,
+      "step": 2379
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00010729293530739119,
+      "loss": 1.264,
+      "step": 2380
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00010722990374180573,
+      "loss": 1.6487,
+      "step": 2381
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00010716686928856916,
+      "loss": 1.6476,
+      "step": 2382
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00010710383197285767,
+      "loss": 1.7101,
+      "step": 2383
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00010704079181984865,
+      "loss": 1.5744,
+      "step": 2384
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00010697774885472049,
+      "loss": 1.2066,
+      "step": 2385
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00010691470310265288,
+      "loss": 1.4143,
+      "step": 2386
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00010685165458882649,
+      "loss": 1.5641,
+      "step": 2387
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00010678860333842313,
+      "loss": 1.655,
+      "step": 2388
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00010672554937662575,
+      "loss": 1.792,
+      "step": 2389
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.0001066624927286183,
+      "loss": 1.5206,
+      "step": 2390
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00010659943341958587,
+      "loss": 1.1943,
+      "step": 2391
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00010653637147471458,
+      "loss": 1.5255,
+      "step": 2392
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00010647330691919162,
+      "loss": 1.3434,
+      "step": 2393
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00010641023977820516,
+      "loss": 1.5061,
+      "step": 2394
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00010634717007694457,
+      "loss": 1.5378,
+      "step": 2395
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.0001062840978406,
+      "loss": 1.6465,
+      "step": 2396
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00010622102309436286,
+      "loss": 1.5147,
+      "step": 2397
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00010615794586342535,
+      "loss": 1.5373,
+      "step": 2398
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00010609486617298078,
+      "loss": 1.6422,
+      "step": 2399
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00010603178404822347,
+      "loss": 1.6051,
+      "step": 2400
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.0001059686995143486,
+      "loss": 1.5681,
+      "step": 2401
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00010590561259655243,
+      "loss": 1.5375,
+      "step": 2402
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00010584252332003204,
+      "loss": 1.4522,
+      "step": 2403
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00010577943170998557,
+      "loss": 1.5058,
+      "step": 2404
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00010571633779161205,
+      "loss": 1.566,
+      "step": 2405
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00010565324159011143,
+      "loss": 1.3351,
+      "step": 2406
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.0001055901431306846,
+      "loss": 1.494,
+      "step": 2407
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00010552704243853329,
+      "loss": 1.5319,
+      "step": 2408
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00010546393953886014,
+      "loss": 1.6249,
+      "step": 2409
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00010540083445686874,
+      "loss": 1.391,
+      "step": 2410
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00010533772721776348,
+      "loss": 1.4402,
+      "step": 2411
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00010527461784674959,
+      "loss": 1.4305,
+      "step": 2412
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00010521150636903327,
+      "loss": 1.5837,
+      "step": 2413
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001051483928098214,
+      "loss": 1.5243,
+      "step": 2414
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00010508527719432186,
+      "loss": 1.6573,
+      "step": 2415
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.34375,
+      "learning_rate": 0.00010502215954774322,
+      "loss": 1.5673,
+      "step": 2416
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.0001049590398952949,
+      "loss": 1.6679,
+      "step": 2417
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00010489591826218719,
+      "loss": 1.3325,
+      "step": 2418
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.000104832794673631,
+      "loss": 1.6615,
+      "step": 2419
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00010476966915483823,
+      "loss": 1.2863,
+      "step": 2420
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00010470654173102146,
+      "loss": 1.7735,
+      "step": 2421
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00010464341242739393,
+      "loss": 1.3264,
+      "step": 2422
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00010458028126916981,
+      "loss": 1.3915,
+      "step": 2423
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00010451714828156387,
+      "loss": 1.5982,
+      "step": 2424
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00010445401348979171,
+      "loss": 1.679,
+      "step": 2425
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.0001043908769190696,
+      "loss": 1.6166,
+      "step": 2426
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00010432773859461448,
+      "loss": 1.4861,
+      "step": 2427
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.0001042645985416441,
+      "loss": 1.6192,
+      "step": 2428
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00010420145678537678,
+      "loss": 1.6368,
+      "step": 2429
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00010413831335103163,
+      "loss": 1.3838,
+      "step": 2430
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00010407516826382838,
+      "loss": 1.4444,
+      "step": 2431
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00010401202154898735,
+      "loss": 1.8088,
+      "step": 2432
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00010394887323172962,
+      "loss": 1.4773,
+      "step": 2433
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.0001038857233372769,
+      "loss": 1.6061,
+      "step": 2434
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00010382257189085144,
+      "loss": 1.4206,
+      "step": 2435
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00010375941891767622,
+      "loss": 1.3997,
+      "step": 2436
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00010369626444297475,
+      "loss": 1.5606,
+      "step": 2437
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00010363310849197116,
+      "loss": 1.6184,
+      "step": 2438
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00010356995108989023,
+      "loss": 1.24,
+      "step": 2439
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001035067922619572,
+      "loss": 1.2681,
+      "step": 2440
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.000103443632033398,
+      "loss": 1.4559,
+      "step": 2441
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00010338047042943903,
+      "loss": 1.6672,
+      "step": 2442
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00010331730747530727,
+      "loss": 1.6735,
+      "step": 2443
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.0001032541431962303,
+      "loss": 1.6384,
+      "step": 2444
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00010319097761743611,
+      "loss": 1.4556,
+      "step": 2445
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00010312781076415327,
+      "loss": 1.4727,
+      "step": 2446
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00010306464266161093,
+      "loss": 1.2777,
+      "step": 2447
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00010300147333503862,
+      "loss": 1.4296,
+      "step": 2448
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00010293830280966643,
+      "loss": 1.63,
+      "step": 2449
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00010287513111072489,
+      "loss": 1.4,
+      "step": 2450
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00010281195826344501,
+      "loss": 1.5783,
+      "step": 2451
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00010274878429305833,
+      "loss": 1.591,
+      "step": 2452
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00010268560922479669,
+      "loss": 1.6555,
+      "step": 2453
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.0001026224330838925,
+      "loss": 1.6218,
+      "step": 2454
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00010255925589557854,
+      "loss": 1.6771,
+      "step": 2455
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00010249607768508799,
+      "loss": 1.3538,
+      "step": 2456
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00010243289847765457,
+      "loss": 1.6307,
+      "step": 2457
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00010236971829851216,
+      "loss": 1.8034,
+      "step": 2458
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00010230653717289524,
+      "loss": 1.7886,
+      "step": 2459
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00010224335512603862,
+      "loss": 1.6017,
+      "step": 2460
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.0001021801721831774,
+      "loss": 1.5403,
+      "step": 2461
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.0001021169883695471,
+      "loss": 1.4036,
+      "step": 2462
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00010205380371038356,
+      "loss": 1.4625,
+      "step": 2463
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.000101990618230923,
+      "loss": 1.4,
+      "step": 2464
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00010192743195640191,
+      "loss": 1.4642,
+      "step": 2465
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.0001018642449120572,
+      "loss": 1.4312,
+      "step": 2466
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00010180105712312593,
+      "loss": 1.7405,
+      "step": 2467
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00010173786861484558,
+      "loss": 1.2621,
+      "step": 2468
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00010167467941245387,
+      "loss": 1.5936,
+      "step": 2469
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00010161148954118886,
+      "loss": 1.4462,
+      "step": 2470
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00010154829902628877,
+      "loss": 1.354,
+      "step": 2471
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00010148510789299214,
+      "loss": 1.506,
+      "step": 2472
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00010142191616653778,
+      "loss": 1.2073,
+      "step": 2473
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00010135872387216466,
+      "loss": 1.4126,
+      "step": 2474
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00010129553103511207,
+      "loss": 1.4432,
+      "step": 2475
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00010123233768061941,
+      "loss": 1.2652,
+      "step": 2476
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.00010116914383392641,
+      "loss": 1.4566,
+      "step": 2477
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00010110594952027293,
+      "loss": 1.3762,
+      "step": 2478
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00010104275476489895,
+      "loss": 1.5277,
+      "step": 2479
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00010097955959304476,
+      "loss": 1.6115,
+      "step": 2480
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00010091636402995075,
+      "loss": 1.6221,
+      "step": 2481
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00010085316810085743,
+      "loss": 1.5007,
+      "step": 2482
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00010078997183100554,
+      "loss": 1.6768,
+      "step": 2483
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00010072677524563587,
+      "loss": 1.4063,
+      "step": 2484
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00010066357836998942,
+      "loss": 1.5115,
+      "step": 2485
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00010060038122930724,
+      "loss": 1.5206,
+      "step": 2486
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00010053718384883049,
+      "loss": 1.5898,
+      "step": 2487
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00010047398625380047,
+      "loss": 1.6385,
+      "step": 2488
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00010041078846945853,
+      "loss": 1.6463,
+      "step": 2489
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.0001003475905210461,
+      "loss": 1.559,
+      "step": 2490
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00010028439243380472,
+      "loss": 1.3598,
+      "step": 2491
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.0001002211942329759,
+      "loss": 1.2564,
+      "step": 2492
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 1.1769163608551025,
+      "eval_runtime": 349.3293,
+      "eval_samples_per_second": 5.516,
+      "eval_steps_per_second": 2.76,
+      "step": 2492
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 4981,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1246,
+  "total_flos": 1.762473646875476e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}