diff --git "a/checkpoint-18000/trainer_state.json" "b/checkpoint-18000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-18000/trainer_state.json"
@@ -0,0 +1,126090 @@
+{
+  "best_metric": 3.08333420753479,
+  "best_model_checkpoint": "models/GPTNeoX-160M-minipile-final/checkpoint-18000",
+  "epoch": 0.7578947368421053,
+  "eval_steps": 3000,
+  "global_step": 18000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 4.210526315789474e-05,
+      "grad_norm": 3.890625,
+      "learning_rate": 1.0526315789473683e-06,
+      "loss": 7.3413,
+      "step": 1
+    },
+    {
+      "epoch": 8.421052631578948e-05,
+      "grad_norm": 4.5,
+      "learning_rate": 2.1052631578947366e-06,
+      "loss": 7.8082,
+      "step": 2
+    },
+    {
+      "epoch": 0.0001263157894736842,
+      "grad_norm": 3.640625,
+      "learning_rate": 3.157894736842105e-06,
+      "loss": 7.7825,
+      "step": 3
+    },
+    {
+      "epoch": 0.00016842105263157895,
+      "grad_norm": 4.875,
+      "learning_rate": 4.210526315789473e-06,
+      "loss": 8.2535,
+      "step": 4
+    },
+    {
+      "epoch": 0.0002105263157894737,
+      "grad_norm": 3.875,
+      "learning_rate": 5.263157894736842e-06,
+      "loss": 7.7617,
+      "step": 5
+    },
+    {
+      "epoch": 0.0002526315789473684,
+      "grad_norm": 4.0,
+      "learning_rate": 6.31578947368421e-06,
+      "loss": 7.7052,
+      "step": 6
+    },
+    {
+      "epoch": 0.00029473684210526316,
+      "grad_norm": 5.21875,
+      "learning_rate": 7.368421052631579e-06,
+      "loss": 8.0528,
+      "step": 7
+    },
+    {
+      "epoch": 0.0003368421052631579,
+      "grad_norm": 4.4375,
+      "learning_rate": 8.421052631578947e-06,
+      "loss": 8.0914,
+      "step": 8
+    },
+    {
+      "epoch": 0.00037894736842105265,
+      "grad_norm": 4.15625,
+      "learning_rate": 9.473684210526317e-06,
+      "loss": 7.847,
+      "step": 9
+    },
+    {
+      "epoch": 0.0004210526315789474,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.0526315789473684e-05,
+      "loss": 7.3733,
+      "step": 10
+    },
+    {
+      "epoch": 0.0004631578947368421,
+      "grad_norm": 4.46875,
+      "learning_rate": 1.1578947368421053e-05,
+      "loss": 8.2678,
+      "step": 11
+    },
+    {
+      "epoch": 0.0005052631578947368,
+      "grad_norm": 3.890625,
+      "learning_rate": 1.263157894736842e-05,
+      "loss": 7.9476,
+      "step": 12
+    },
+    {
+      "epoch": 0.0005473684210526316,
+      "grad_norm": 4.5,
+      "learning_rate": 1.368421052631579e-05,
+      "loss": 8.2657,
+      "step": 13
+    },
+    {
+      "epoch": 0.0005894736842105263,
+      "grad_norm": 4.625,
+      "learning_rate": 1.4736842105263159e-05,
+      "loss": 7.954,
+      "step": 14
+    },
+    {
+      "epoch": 0.0006315789473684211,
+      "grad_norm": 4.28125,
+      "learning_rate": 1.5789473684210526e-05,
+      "loss": 7.2995,
+      "step": 15
+    },
+    {
+      "epoch": 0.0006736842105263158,
+      "grad_norm": 3.859375,
+      "learning_rate": 1.6842105263157893e-05,
+      "loss": 7.876,
+      "step": 16
+    },
+    {
+      "epoch": 0.0007157894736842105,
+      "grad_norm": 4.3125,
+      "learning_rate": 1.7894736842105264e-05,
+      "loss": 7.925,
+      "step": 17
+    },
+    {
+      "epoch": 0.0007578947368421053,
+      "grad_norm": 3.078125,
+      "learning_rate": 1.8947368421052634e-05,
+      "loss": 7.2259,
+      "step": 18
+    },
+    {
+      "epoch": 0.0008,
+      "grad_norm": 5.34375,
+      "learning_rate": 2e-05,
+      "loss": 8.2554,
+      "step": 19
+    },
+    {
+      "epoch": 0.0008421052631578948,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.105263157894737e-05,
+      "loss": 7.42,
+      "step": 20
+    },
+    {
+      "epoch": 0.0008842105263157895,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.2105263157894736e-05,
+      "loss": 7.3183,
+      "step": 21
+    },
+    {
+      "epoch": 0.0009263157894736842,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.3157894736842107e-05,
+      "loss": 7.8771,
+      "step": 22
+    },
+    {
+      "epoch": 0.0009684210526315789,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.4210526315789474e-05,
+      "loss": 7.4952,
+      "step": 23
+    },
+    {
+      "epoch": 0.0010105263157894737,
+      "grad_norm": 2.71875,
+      "learning_rate": 2.526315789473684e-05,
+      "loss": 7.5569,
+      "step": 24
+    },
+    {
+      "epoch": 0.0010526315789473684,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.631578947368421e-05,
+      "loss": 6.8489,
+      "step": 25
+    },
+    {
+      "epoch": 0.0010947368421052631,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.736842105263158e-05,
+      "loss": 7.8147,
+      "step": 26
+    },
+    {
+      "epoch": 0.0011368421052631579,
+      "grad_norm": 1.9296875,
+      "learning_rate": 2.842105263157895e-05,
+      "loss": 6.961,
+      "step": 27
+    },
+    {
+      "epoch": 0.0011789473684210526,
+      "grad_norm": 2.625,
+      "learning_rate": 2.9473684210526317e-05,
+      "loss": 7.421,
+      "step": 28
+    },
+    {
+      "epoch": 0.0012210526315789474,
+      "grad_norm": 2.03125,
+      "learning_rate": 3.0526315789473684e-05,
+      "loss": 6.9197,
+      "step": 29
+    },
+    {
+      "epoch": 0.0012631578947368421,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.157894736842105e-05,
+      "loss": 7.7633,
+      "step": 30
+    },
+    {
+      "epoch": 0.0013052631578947369,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.263157894736842e-05,
+      "loss": 6.9348,
+      "step": 31
+    },
+    {
+      "epoch": 0.0013473684210526316,
+      "grad_norm": 1.875,
+      "learning_rate": 3.3684210526315786e-05,
+      "loss": 6.9246,
+      "step": 32
+    },
+    {
+      "epoch": 0.0013894736842105264,
+      "grad_norm": 2.1875,
+      "learning_rate": 3.473684210526316e-05,
+      "loss": 7.4443,
+      "step": 33
+    },
+    {
+      "epoch": 0.001431578947368421,
+      "grad_norm": 1.65625,
+      "learning_rate": 3.578947368421053e-05,
+      "loss": 7.2099,
+      "step": 34
+    },
+    {
+      "epoch": 0.0014736842105263158,
+      "grad_norm": 1.859375,
+      "learning_rate": 3.6842105263157895e-05,
+      "loss": 6.8235,
+      "step": 35
+    },
+    {
+      "epoch": 0.0015157894736842106,
+      "grad_norm": 1.953125,
+      "learning_rate": 3.789473684210527e-05,
+      "loss": 6.9099,
+      "step": 36
+    },
+    {
+      "epoch": 0.0015578947368421053,
+      "grad_norm": 1.90625,
+      "learning_rate": 3.8947368421052636e-05,
+      "loss": 6.941,
+      "step": 37
+    },
+    {
+      "epoch": 0.0016,
+      "grad_norm": 1.6328125,
+      "learning_rate": 4e-05,
+      "loss": 6.6937,
+      "step": 38
+    },
+    {
+      "epoch": 0.0016421052631578948,
+      "grad_norm": 1.734375,
+      "learning_rate": 4.105263157894737e-05,
+      "loss": 7.0925,
+      "step": 39
+    },
+    {
+      "epoch": 0.0016842105263157896,
+      "grad_norm": 1.4609375,
+      "learning_rate": 4.210526315789474e-05,
+      "loss": 6.8102,
+      "step": 40
+    },
+    {
+      "epoch": 0.0017263157894736843,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.3157894736842105e-05,
+      "loss": 6.9307,
+      "step": 41
+    },
+    {
+      "epoch": 0.001768421052631579,
+      "grad_norm": 1.6796875,
+      "learning_rate": 4.421052631578947e-05,
+      "loss": 6.3676,
+      "step": 42
+    },
+    {
+      "epoch": 0.0018105263157894736,
+      "grad_norm": 1.78125,
+      "learning_rate": 4.5263157894736846e-05,
+      "loss": 7.1898,
+      "step": 43
+    },
+    {
+      "epoch": 0.0018526315789473683,
+      "grad_norm": 1.4765625,
+      "learning_rate": 4.6315789473684214e-05,
+      "loss": 6.763,
+      "step": 44
+    },
+    {
+      "epoch": 0.001894736842105263,
+      "grad_norm": 1.34375,
+      "learning_rate": 4.736842105263158e-05,
+      "loss": 6.5191,
+      "step": 45
+    },
+    {
+      "epoch": 0.0019368421052631578,
+      "grad_norm": 1.578125,
+      "learning_rate": 4.842105263157895e-05,
+      "loss": 6.8033,
+      "step": 46
+    },
+    {
+      "epoch": 0.0019789473684210528,
+      "grad_norm": 1.765625,
+      "learning_rate": 4.9473684210526315e-05,
+      "loss": 6.9758,
+      "step": 47
+    },
+    {
+      "epoch": 0.0020210526315789473,
+      "grad_norm": 1.8359375,
+      "learning_rate": 5.052631578947368e-05,
+      "loss": 7.1849,
+      "step": 48
+    },
+    {
+      "epoch": 0.0020631578947368423,
+      "grad_norm": 2.296875,
+      "learning_rate": 5.157894736842105e-05,
+      "loss": 5.9802,
+      "step": 49
+    },
+    {
+      "epoch": 0.002105263157894737,
+      "grad_norm": 1.4453125,
+      "learning_rate": 5.263157894736842e-05,
+      "loss": 6.4116,
+      "step": 50
+    },
+    {
+      "epoch": 0.0021473684210526318,
+      "grad_norm": 1.2890625,
+      "learning_rate": 5.368421052631579e-05,
+      "loss": 6.084,
+      "step": 51
+    },
+    {
+      "epoch": 0.0021894736842105263,
+      "grad_norm": 1.3828125,
+      "learning_rate": 5.473684210526316e-05,
+      "loss": 6.541,
+      "step": 52
+    },
+    {
+      "epoch": 0.0022315789473684212,
+      "grad_norm": 1.25,
+      "learning_rate": 5.5789473684210526e-05,
+      "loss": 6.3551,
+      "step": 53
+    },
+    {
+      "epoch": 0.0022736842105263158,
+      "grad_norm": 1.640625,
+      "learning_rate": 5.68421052631579e-05,
+      "loss": 6.6938,
+      "step": 54
+    },
+    {
+      "epoch": 0.0023157894736842107,
+      "grad_norm": 1.265625,
+      "learning_rate": 5.789473684210527e-05,
+      "loss": 6.2223,
+      "step": 55
+    },
+    {
+      "epoch": 0.0023578947368421053,
+      "grad_norm": 1.265625,
+      "learning_rate": 5.8947368421052634e-05,
+      "loss": 6.2336,
+      "step": 56
+    },
+    {
+      "epoch": 0.0024,
+      "grad_norm": 1.0703125,
+      "learning_rate": 6e-05,
+      "loss": 6.0368,
+      "step": 57
+    },
+    {
+      "epoch": 0.0024421052631578948,
+      "grad_norm": 1.1484375,
+      "learning_rate": 6.105263157894737e-05,
+      "loss": 6.2726,
+      "step": 58
+    },
+    {
+      "epoch": 0.0024842105263157893,
+      "grad_norm": 1.578125,
+      "learning_rate": 6.210526315789474e-05,
+      "loss": 6.8905,
+      "step": 59
+    },
+    {
+      "epoch": 0.0025263157894736842,
+      "grad_norm": 1.1328125,
+      "learning_rate": 6.31578947368421e-05,
+      "loss": 6.0762,
+      "step": 60
+    },
+    {
+      "epoch": 0.0025684210526315788,
+      "grad_norm": 1.6640625,
+      "learning_rate": 6.421052631578946e-05,
+      "loss": 6.2807,
+      "step": 61
+    },
+    {
+      "epoch": 0.0026105263157894737,
+      "grad_norm": 1.703125,
+      "learning_rate": 6.526315789473684e-05,
+      "loss": 6.2477,
+      "step": 62
+    },
+    {
+      "epoch": 0.0026526315789473683,
+      "grad_norm": 1.2421875,
+      "learning_rate": 6.631578947368421e-05,
+      "loss": 6.0277,
+      "step": 63
+    },
+    {
+      "epoch": 0.0026947368421052632,
+      "grad_norm": 1.5234375,
+      "learning_rate": 6.736842105263157e-05,
+      "loss": 5.8858,
+      "step": 64
+    },
+    {
+      "epoch": 0.0027368421052631577,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.842105263157896e-05,
+      "loss": 6.3088,
+      "step": 65
+    },
+    {
+      "epoch": 0.0027789473684210527,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.947368421052632e-05,
+      "loss": 6.5203,
+      "step": 66
+    },
+    {
+      "epoch": 0.0028210526315789472,
+      "grad_norm": 1.3671875,
+      "learning_rate": 7.05263157894737e-05,
+      "loss": 6.5611,
+      "step": 67
+    },
+    {
+      "epoch": 0.002863157894736842,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.157894736842105e-05,
+      "loss": 6.0409,
+      "step": 68
+    },
+    {
+      "epoch": 0.0029052631578947367,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.263157894736843e-05,
+      "loss": 5.9745,
+      "step": 69
+    },
+    {
+      "epoch": 0.0029473684210526317,
+      "grad_norm": 1.15625,
+      "learning_rate": 7.368421052631579e-05,
+      "loss": 5.946,
+      "step": 70
+    },
+    {
+      "epoch": 0.002989473684210526,
+      "grad_norm": 1.1953125,
+      "learning_rate": 7.473684210526316e-05,
+      "loss": 6.0341,
+      "step": 71
+    },
+    {
+      "epoch": 0.003031578947368421,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.578947368421054e-05,
+      "loss": 5.6885,
+      "step": 72
+    },
+    {
+      "epoch": 0.0030736842105263157,
+      "grad_norm": 1.2109375,
+      "learning_rate": 7.68421052631579e-05,
+      "loss": 6.4778,
+      "step": 73
+    },
+    {
+      "epoch": 0.0031157894736842107,
+      "grad_norm": 0.9921875,
+      "learning_rate": 7.789473684210527e-05,
+      "loss": 6.1629,
+      "step": 74
+    },
+    {
+      "epoch": 0.003157894736842105,
+      "grad_norm": 1.140625,
+      "learning_rate": 7.894736842105263e-05,
+      "loss": 5.6599,
+      "step": 75
+    },
+    {
+      "epoch": 0.0032,
+      "grad_norm": 2.0,
+      "learning_rate": 8e-05,
+      "loss": 5.901,
+      "step": 76
+    },
+    {
+      "epoch": 0.0032421052631578947,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.105263157894737e-05,
+      "loss": 6.076,
+      "step": 77
+    },
+    {
+      "epoch": 0.0032842105263157896,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.210526315789474e-05,
+      "loss": 5.7718,
+      "step": 78
+    },
+    {
+      "epoch": 0.003326315789473684,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.315789473684212e-05,
+      "loss": 5.7908,
+      "step": 79
+    },
+    {
+      "epoch": 0.003368421052631579,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.421052631578948e-05,
+      "loss": 5.9521,
+      "step": 80
+    },
+    {
+      "epoch": 0.0034105263157894737,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.526315789473685e-05,
+      "loss": 6.0692,
+      "step": 81
+    },
+    {
+      "epoch": 0.0034526315789473686,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.631578947368421e-05,
+      "loss": 5.7741,
+      "step": 82
+    },
+    {
+      "epoch": 0.003494736842105263,
+      "grad_norm": 1.1640625,
+      "learning_rate": 8.736842105263158e-05,
+      "loss": 5.6126,
+      "step": 83
+    },
+    {
+      "epoch": 0.003536842105263158,
+      "grad_norm": 1.2890625,
+      "learning_rate": 8.842105263157894e-05,
+      "loss": 5.8609,
+      "step": 84
+    },
+    {
+      "epoch": 0.0035789473684210526,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.947368421052632e-05,
+      "loss": 5.8477,
+      "step": 85
+    },
+    {
+      "epoch": 0.003621052631578947,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.052631578947369e-05,
+      "loss": 5.8577,
+      "step": 86
+    },
+    {
+      "epoch": 0.003663157894736842,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.157894736842105e-05,
+      "loss": 5.9344,
+      "step": 87
+    },
+    {
+      "epoch": 0.0037052631578947367,
+      "grad_norm": 1.5390625,
+      "learning_rate": 9.263157894736843e-05,
+      "loss": 6.1071,
+      "step": 88
+    },
+    {
+      "epoch": 0.0037473684210526316,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.368421052631579e-05,
+      "loss": 5.7145,
+      "step": 89
+    },
+    {
+      "epoch": 0.003789473684210526,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.473684210526316e-05,
+      "loss": 5.8452,
+      "step": 90
+    },
+    {
+      "epoch": 0.003831578947368421,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.578947368421052e-05,
+      "loss": 5.6488,
+      "step": 91
+    },
+    {
+      "epoch": 0.0038736842105263156,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.68421052631579e-05,
+      "loss": 5.5844,
+      "step": 92
+    },
+    {
+      "epoch": 0.00391578947368421,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.789473684210526e-05,
+      "loss": 5.7262,
+      "step": 93
+    },
+    {
+      "epoch": 0.0039578947368421056,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.894736842105263e-05,
+      "loss": 5.6465,
+      "step": 94
+    },
+    {
+      "epoch": 0.004,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.0001,
+      "loss": 5.7792,
+      "step": 95
+    },
+    {
+      "epoch": 0.004042105263157895,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00010105263157894737,
+      "loss": 5.8167,
+      "step": 96
+    },
+    {
+      "epoch": 0.004084210526315789,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00010210526315789474,
+      "loss": 5.7599,
+      "step": 97
+    },
+    {
+      "epoch": 0.0041263157894736845,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0001031578947368421,
+      "loss": 5.5593,
+      "step": 98
+    },
+    {
+      "epoch": 0.004168421052631579,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.00010421052631578947,
+      "loss": 5.6243,
+      "step": 99
+    },
+    {
+      "epoch": 0.004210526315789474,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00010526315789473683,
+      "loss": 5.756,
+      "step": 100
+    },
+    {
+      "epoch": 0.004252631578947368,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00010631578947368421,
+      "loss": 5.5719,
+      "step": 101
+    },
+    {
+      "epoch": 0.0042947368421052635,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00010736842105263158,
+      "loss": 5.418,
+      "step": 102
+    },
+    {
+      "epoch": 0.004336842105263158,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00010842105263157894,
+      "loss": 5.5703,
+      "step": 103
+    },
+    {
+      "epoch": 0.004378947368421053,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00010947368421052632,
+      "loss": 5.6246,
+      "step": 104
+    },
+    {
+      "epoch": 0.004421052631578947,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00011052631578947368,
+      "loss": 5.8189,
+      "step": 105
+    },
+    {
+      "epoch": 0.0044631578947368425,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00011157894736842105,
+      "loss": 5.4517,
+      "step": 106
+    },
+    {
+      "epoch": 0.004505263157894737,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00011263157894736841,
+      "loss": 5.4263,
+      "step": 107
+    },
+    {
+      "epoch": 0.0045473684210526315,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.0001136842105263158,
+      "loss": 5.5597,
+      "step": 108
+    },
+    {
+      "epoch": 0.004589473684210526,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00011473684210526316,
+      "loss": 5.6423,
+      "step": 109
+    },
+    {
+      "epoch": 0.0046315789473684215,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00011578947368421053,
+      "loss": 5.6542,
+      "step": 110
+    },
+    {
+      "epoch": 0.004673684210526316,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00011684210526315791,
+      "loss": 5.5691,
+      "step": 111
+    },
+    {
+      "epoch": 0.0047157894736842105,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00011789473684210527,
+      "loss": 5.4473,
+      "step": 112
+    },
+    {
+      "epoch": 0.004757894736842105,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00011894736842105264,
+      "loss": 5.4498,
+      "step": 113
+    },
+    {
+      "epoch": 0.0048,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00012,
+      "loss": 5.67,
+      "step": 114
+    },
+    {
+      "epoch": 0.004842105263157895,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00012105263157894738,
+      "loss": 5.3806,
+      "step": 115
+    },
+    {
+      "epoch": 0.0048842105263157895,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00012210526315789474,
+      "loss": 5.4164,
+      "step": 116
+    },
+    {
+      "epoch": 0.004926315789473684,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.0001231578947368421,
+      "loss": 5.2849,
+      "step": 117
+    },
+    {
+      "epoch": 0.0049684210526315786,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00012421052631578949,
+      "loss": 5.4942,
+      "step": 118
+    },
+    {
+      "epoch": 0.005010526315789474,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00012526315789473686,
+      "loss": 5.7359,
+      "step": 119
+    },
+    {
+      "epoch": 0.0050526315789473685,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001263157894736842,
+      "loss": 5.5311,
+      "step": 120
+    },
+    {
+      "epoch": 0.005094736842105263,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00012736842105263158,
+      "loss": 5.3792,
+      "step": 121
+    },
+    {
+      "epoch": 0.0051368421052631575,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00012842105263157893,
+      "loss": 5.4266,
+      "step": 122
+    },
+    {
+      "epoch": 0.005178947368421053,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.00012947368421052633,
+      "loss": 5.401,
+      "step": 123
+    },
+    {
+      "epoch": 0.0052210526315789475,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00013052631578947368,
+      "loss": 5.1973,
+      "step": 124
+    },
+    {
+      "epoch": 0.005263157894736842,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00013157894736842105,
+      "loss": 5.287,
+      "step": 125
+    },
+    {
+      "epoch": 0.0053052631578947365,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00013263157894736842,
+      "loss": 5.3064,
+      "step": 126
+    },
+    {
+      "epoch": 0.005347368421052632,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.0001336842105263158,
+      "loss": 5.2671,
+      "step": 127
+    },
+    {
+      "epoch": 0.0053894736842105264,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00013473684210526314,
+      "loss": 5.3091,
+      "step": 128
+    },
+    {
+      "epoch": 0.005431578947368421,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00013578947368421052,
+      "loss": 5.3557,
+      "step": 129
+    },
+    {
+      "epoch": 0.0054736842105263155,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00013684210526315792,
+      "loss": 5.4329,
+      "step": 130
+    },
+    {
+      "epoch": 0.005515789473684211,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00013789473684210527,
+      "loss": 5.1083,
+      "step": 131
+    },
+    {
+      "epoch": 0.005557894736842105,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00013894736842105264,
+      "loss": 5.3215,
+      "step": 132
+    },
+    {
+      "epoch": 0.0056,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00014000000000000001,
+      "loss": 5.1231,
+      "step": 133
+    },
+    {
+      "epoch": 0.0056421052631578945,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.0001410526315789474,
+      "loss": 5.2299,
+      "step": 134
+    },
+    {
+      "epoch": 0.00568421052631579,
+      "grad_norm": 0.875,
+      "learning_rate": 0.00014210526315789474,
+      "loss": 5.3445,
+      "step": 135
+    },
+    {
+      "epoch": 0.005726315789473684,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0001431578947368421,
+      "loss": 5.0036,
+      "step": 136
+    },
+    {
+      "epoch": 0.005768421052631579,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00014421052631578948,
+      "loss": 4.9686,
+      "step": 137
+    },
+    {
+      "epoch": 0.0058105263157894734,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.00014526315789473686,
+      "loss": 5.2371,
+      "step": 138
+    },
+    {
+      "epoch": 0.005852631578947369,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.0001463157894736842,
+      "loss": 5.2915,
+      "step": 139
+    },
+    {
+      "epoch": 0.005894736842105263,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00014736842105263158,
+      "loss": 5.2411,
+      "step": 140
+    },
+    {
+      "epoch": 0.005936842105263158,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00014842105263157895,
+      "loss": 5.2067,
+      "step": 141
+    },
+    {
+      "epoch": 0.005978947368421052,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00014947368421052633,
+      "loss": 5.2988,
+      "step": 142
+    },
+    {
+      "epoch": 0.006021052631578947,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00015052631578947367,
+      "loss": 5.2165,
+      "step": 143
+    },
+    {
+      "epoch": 0.006063157894736842,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00015157894736842108,
+      "loss": 5.2836,
+      "step": 144
+    },
+    {
+      "epoch": 0.006105263157894737,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00015263157894736842,
+      "loss": 5.1639,
+      "step": 145
+    },
+    {
+      "epoch": 0.006147368421052631,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001536842105263158,
+      "loss": 5.0287,
+      "step": 146
+    },
+    {
+      "epoch": 0.006189473684210526,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00015473684210526314,
+      "loss": 4.9608,
+      "step": 147
+    },
+    {
+      "epoch": 0.006231578947368421,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00015578947368421054,
+      "loss": 5.2198,
+      "step": 148
+    },
+    {
+      "epoch": 0.006273684210526316,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.0001568421052631579,
+      "loss": 5.2419,
+      "step": 149
+    },
+    {
+      "epoch": 0.00631578947368421,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00015789473684210527,
+      "loss": 5.0146,
+      "step": 150
+    },
+    {
+      "epoch": 0.006357894736842105,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00015894736842105264,
+      "loss": 5.2386,
+      "step": 151
+    },
+    {
+      "epoch": 0.0064,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00016,
+      "loss": 5.1775,
+      "step": 152
+    },
+    {
+      "epoch": 0.006442105263157895,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00016105263157894736,
+      "loss": 5.1883,
+      "step": 153
+    },
+    {
+      "epoch": 0.006484210526315789,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.00016210526315789473,
+      "loss": 5.1067,
+      "step": 154
+    },
+    {
+      "epoch": 0.006526315789473684,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.0001631578947368421,
+      "loss": 5.0509,
+      "step": 155
+    },
+    {
+      "epoch": 0.006568421052631579,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00016421052631578948,
+      "loss": 4.9433,
+      "step": 156
+    },
+    {
+      "epoch": 0.006610526315789474,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00016526315789473683,
+      "loss": 5.23,
+      "step": 157
+    },
+    {
+      "epoch": 0.006652631578947368,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00016631578947368423,
+      "loss": 5.023,
+      "step": 158
+    },
+    {
+      "epoch": 0.006694736842105263,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.00016736842105263158,
+      "loss": 5.0362,
+      "step": 159
+    },
+    {
+      "epoch": 0.006736842105263158,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00016842105263157895,
+      "loss": 5.2109,
+      "step": 160
+    },
+    {
+      "epoch": 0.006778947368421053,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.0001694736842105263,
+      "loss": 5.1547,
+      "step": 161
+    },
+    {
+      "epoch": 0.006821052631578947,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.0001705263157894737,
+      "loss": 5.0887,
+      "step": 162
+    },
+    {
+      "epoch": 0.006863157894736842,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00017157894736842105,
+      "loss": 5.193,
+      "step": 163
+    },
+    {
+      "epoch": 0.006905263157894737,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00017263157894736842,
+      "loss": 5.2748,
+      "step": 164
+    },
+    {
+      "epoch": 0.006947368421052632,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001736842105263158,
+      "loss": 5.1514,
+      "step": 165
+    },
+    {
+      "epoch": 0.006989473684210526,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00017473684210526317,
+      "loss": 4.7939,
+      "step": 166
+    },
+    {
+      "epoch": 0.007031578947368421,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00017578947368421052,
+      "loss": 4.9078,
+      "step": 167
+    },
+    {
+      "epoch": 0.007073684210526316,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.0001768421052631579,
+      "loss": 5.2173,
+      "step": 168
+    },
+    {
+      "epoch": 0.007115789473684211,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00017789473684210526,
+      "loss": 4.913,
+      "step": 169
+    },
+    {
+      "epoch": 0.007157894736842105,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00017894736842105264,
+      "loss": 5.1521,
+      "step": 170
+    },
+    {
+      "epoch": 0.0072,
+      "grad_norm": 0.875,
+      "learning_rate": 0.00017999999999999998,
+      "loss": 4.9807,
+      "step": 171
+    },
+    {
+      "epoch": 0.007242105263157894,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018105263157894739,
+      "loss": 4.9046,
+      "step": 172
+    },
+    {
+      "epoch": 0.00728421052631579,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00018210526315789476,
+      "loss": 5.2114,
+      "step": 173
+    },
+    {
+      "epoch": 0.007326315789473684,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0001831578947368421,
+      "loss": 5.0849,
+      "step": 174
+    },
+    {
+      "epoch": 0.007368421052631579,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00018421052631578948,
+      "loss": 4.9578,
+      "step": 175
+    },
+    {
+      "epoch": 0.007410526315789473,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.00018526315789473685,
+      "loss": 5.1629,
+      "step": 176
+    },
+    {
+      "epoch": 0.007452631578947369,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00018631578947368423,
+      "loss": 5.0355,
+      "step": 177
+    },
+    {
+      "epoch": 0.007494736842105263,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00018736842105263158,
+      "loss": 5.1637,
+      "step": 178
+    },
+    {
+      "epoch": 0.007536842105263158,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00018842105263157895,
+      "loss": 4.9553,
+      "step": 179
+    },
+    {
+      "epoch": 0.007578947368421052,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00018947368421052632,
+      "loss": 4.7738,
+      "step": 180
+    },
+    {
+      "epoch": 0.007621052631578948,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001905263157894737,
+      "loss": 5.2183,
+      "step": 181
+    },
+    {
+      "epoch": 0.007663157894736842,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00019157894736842104,
+      "loss": 4.9233,
+      "step": 182
+    },
+    {
+      "epoch": 0.007705263157894737,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00019263157894736845,
+      "loss": 5.0219,
+      "step": 183
+    },
+    {
+      "epoch": 0.007747368421052631,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0001936842105263158,
+      "loss": 4.6389,
+      "step": 184
+    },
+    {
+      "epoch": 0.007789473684210527,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00019473684210526317,
+      "loss": 4.9254,
+      "step": 185
+    },
+    {
+      "epoch": 0.00783157894736842,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0001957894736842105,
+      "loss": 4.8345,
+      "step": 186
+    },
+    {
+      "epoch": 0.007873684210526316,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019684210526315791,
+      "loss": 5.1883,
+      "step": 187
+    },
+    {
+      "epoch": 0.007915789473684211,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00019789473684210526,
+      "loss": 4.916,
+      "step": 188
+    },
+    {
+      "epoch": 0.007957894736842105,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00019894736842105264,
+      "loss": 4.886,
+      "step": 189
+    },
+    {
+      "epoch": 0.008,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0002,
+      "loss": 4.6236,
+      "step": 190
+    },
+    {
+      "epoch": 0.008042105263157896,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00020105263157894738,
+      "loss": 4.8837,
+      "step": 191
+    },
+    {
+      "epoch": 0.00808421052631579,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.00020210526315789473,
+      "loss": 4.6736,
+      "step": 192
+    },
+    {
+      "epoch": 0.008126315789473685,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0002031578947368421,
+      "loss": 4.9368,
+      "step": 193
+    },
+    {
+      "epoch": 0.008168421052631578,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00020421052631578948,
+      "loss": 5.2258,
+      "step": 194
+    },
+    {
+      "epoch": 0.008210526315789474,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00020526315789473685,
+      "loss": 4.7898,
+      "step": 195
+    },
+    {
+      "epoch": 0.008252631578947369,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.0002063157894736842,
+      "loss": 4.9005,
+      "step": 196
+    },
+    {
+      "epoch": 0.008294736842105263,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.0002073684210526316,
+      "loss": 4.882,
+      "step": 197
+    },
+    {
+      "epoch": 0.008336842105263158,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00020842105263157895,
+      "loss": 5.1039,
+      "step": 198
+    },
+    {
+      "epoch": 0.008378947368421052,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00020947368421052632,
+      "loss": 4.9803,
+      "step": 199
+    },
+    {
+      "epoch": 0.008421052631578947,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00021052631578947367,
+      "loss": 4.8663,
+      "step": 200
+    },
+    {
+      "epoch": 0.008463157894736843,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00021157894736842107,
+      "loss": 4.7288,
+      "step": 201
+    },
+    {
+      "epoch": 0.008505263157894736,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00021263157894736842,
+      "loss": 4.7374,
+      "step": 202
+    },
+    {
+      "epoch": 0.008547368421052632,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.0002136842105263158,
+      "loss": 4.6774,
+      "step": 203
+    },
+    {
+      "epoch": 0.008589473684210527,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00021473684210526316,
+      "loss": 4.8405,
+      "step": 204
+    },
+    {
+      "epoch": 0.00863157894736842,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00021578947368421054,
+      "loss": 4.8519,
+      "step": 205
+    },
+    {
+      "epoch": 0.008673684210526316,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00021684210526315789,
+      "loss": 5.0526,
+      "step": 206
+    },
+    {
+      "epoch": 0.00871578947368421,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00021789473684210526,
+      "loss": 5.0063,
+      "step": 207
+    },
+    {
+      "epoch": 0.008757894736842105,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00021894736842105263,
+      "loss": 5.0228,
+      "step": 208
+    },
+    {
+      "epoch": 0.0088,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00022,
+      "loss": 5.1899,
+      "step": 209
+    },
+    {
+      "epoch": 0.008842105263157894,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00022105263157894735,
+      "loss": 4.7837,
+      "step": 210
+    },
+    {
+      "epoch": 0.00888421052631579,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00022210526315789476,
+      "loss": 5.0637,
+      "step": 211
+    },
+    {
+      "epoch": 0.008926315789473685,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.0002231578947368421,
+      "loss": 4.8577,
+      "step": 212
+    },
+    {
+      "epoch": 0.008968421052631579,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00022421052631578948,
+      "loss": 5.0281,
+      "step": 213
+    },
+    {
+      "epoch": 0.009010526315789474,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00022526315789473682,
+      "loss": 4.6837,
+      "step": 214
+    },
+    {
+      "epoch": 0.009052631578947368,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00022631578947368422,
+      "loss": 4.7071,
+      "step": 215
+    },
+    {
+      "epoch": 0.009094736842105263,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0002273684210526316,
+      "loss": 4.5763,
+      "step": 216
+    },
+    {
+      "epoch": 0.009136842105263158,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00022842105263157895,
+      "loss": 4.7718,
+      "step": 217
+    },
+    {
+      "epoch": 0.009178947368421052,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00022947368421052632,
+      "loss": 4.7802,
+      "step": 218
+    },
+    {
+      "epoch": 0.009221052631578948,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.0002305263157894737,
+      "loss": 5.1001,
+      "step": 219
+    },
+    {
+      "epoch": 0.009263157894736843,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00023157894736842107,
+      "loss": 4.8724,
+      "step": 220
+    },
+    {
+      "epoch": 0.009305263157894737,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00023263157894736841,
+      "loss": 4.7159,
+      "step": 221
+    },
+    {
+      "epoch": 0.009347368421052632,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00023368421052631582,
+      "loss": 4.8988,
+      "step": 222
+    },
+    {
+      "epoch": 0.009389473684210526,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00023473684210526316,
+      "loss": 4.8679,
+      "step": 223
+    },
+    {
+      "epoch": 0.009431578947368421,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00023578947368421054,
+      "loss": 4.9539,
+      "step": 224
+    },
+    {
+      "epoch": 0.009473684210526316,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00023684210526315788,
+      "loss": 4.7077,
+      "step": 225
+    },
+    {
+      "epoch": 0.00951578947368421,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00023789473684210529,
+      "loss": 4.9912,
+      "step": 226
+    },
+    {
+      "epoch": 0.009557894736842105,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00023894736842105263,
+      "loss": 4.8173,
+      "step": 227
+    },
+    {
+      "epoch": 0.0096,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00024,
+      "loss": 4.7645,
+      "step": 228
+    },
+    {
+      "epoch": 0.009642105263157895,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00024105263157894738,
+      "loss": 4.9278,
+      "step": 229
+    },
+    {
+      "epoch": 0.00968421052631579,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00024210526315789475,
+      "loss": 4.8724,
+      "step": 230
+    },
+    {
+      "epoch": 0.009726315789473684,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0002431578947368421,
+      "loss": 4.5503,
+      "step": 231
+    },
+    {
+      "epoch": 0.009768421052631579,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0002442105263157895,
+      "loss": 4.741,
+      "step": 232
+    },
+    {
+      "epoch": 0.009810526315789474,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00024526315789473685,
+      "loss": 4.5287,
+      "step": 233
+    },
+    {
+      "epoch": 0.009852631578947368,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0002463157894736842,
+      "loss": 4.9546,
+      "step": 234
+    },
+    {
+      "epoch": 0.009894736842105263,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.0002473684210526316,
+      "loss": 4.6637,
+      "step": 235
+    },
+    {
+      "epoch": 0.009936842105263157,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00024842105263157897,
+      "loss": 4.7514,
+      "step": 236
+    },
+    {
+      "epoch": 0.009978947368421053,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00024947368421052635,
+      "loss": 4.8299,
+      "step": 237
+    },
+    {
+      "epoch": 0.010021052631578948,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.0002505263157894737,
+      "loss": 4.6256,
+      "step": 238
+    },
+    {
+      "epoch": 0.010063157894736842,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.0002515789473684211,
+      "loss": 4.7495,
+      "step": 239
+    },
+    {
+      "epoch": 0.010105263157894737,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.0002526315789473684,
+      "loss": 4.5281,
+      "step": 240
+    },
+    {
+      "epoch": 0.010147368421052632,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0002536842105263158,
+      "loss": 4.8828,
+      "step": 241
+    },
+    {
+      "epoch": 0.010189473684210526,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00025473684210526316,
+      "loss": 4.6277,
+      "step": 242
+    },
+    {
+      "epoch": 0.010231578947368421,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00025578947368421054,
+      "loss": 5.1124,
+      "step": 243
+    },
+    {
+      "epoch": 0.010273684210526315,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00025684210526315786,
+      "loss": 4.9601,
+      "step": 244
+    },
+    {
+      "epoch": 0.01031578947368421,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0002578947368421053,
+      "loss": 4.7439,
+      "step": 245
+    },
+    {
+      "epoch": 0.010357894736842106,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00025894736842105266,
+      "loss": 4.6651,
+      "step": 246
+    },
+    {
+      "epoch": 0.0104,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00026000000000000003,
+      "loss": 4.4402,
+      "step": 247
+    },
+    {
+      "epoch": 0.010442105263157895,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00026105263157894735,
+      "loss": 4.8318,
+      "step": 248
+    },
+    {
+      "epoch": 0.01048421052631579,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.0002621052631578947,
+      "loss": 4.5234,
+      "step": 249
+    },
+    {
+      "epoch": 0.010526315789473684,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0002631578947368421,
+      "loss": 4.8966,
+      "step": 250
+    },
+    {
+      "epoch": 0.01056842105263158,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.0002642105263157895,
+      "loss": 4.5258,
+      "step": 251
+    },
+    {
+      "epoch": 0.010610526315789473,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.00026526315789473685,
+      "loss": 4.5859,
+      "step": 252
+    },
+    {
+      "epoch": 0.010652631578947368,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0002663157894736842,
+      "loss": 4.5474,
+      "step": 253
+    },
+    {
+      "epoch": 0.010694736842105264,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.0002673684210526316,
+      "loss": 5.1724,
+      "step": 254
+    },
+    {
+      "epoch": 0.010736842105263157,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00026842105263157897,
+      "loss": 4.8648,
+      "step": 255
+    },
+    {
+      "epoch": 0.010778947368421053,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0002694736842105263,
+      "loss": 4.8057,
+      "step": 256
+    },
+    {
+      "epoch": 0.010821052631578947,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00027052631578947366,
+      "loss": 4.8077,
+      "step": 257
+    },
+    {
+      "epoch": 0.010863157894736842,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00027157894736842104,
+      "loss": 5.0211,
+      "step": 258
+    },
+    {
+      "epoch": 0.010905263157894737,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00027263157894736847,
+      "loss": 4.9733,
+      "step": 259
+    },
+    {
+      "epoch": 0.010947368421052631,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00027368421052631584,
+      "loss": 4.795,
+      "step": 260
+    },
+    {
+      "epoch": 0.010989473684210526,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00027473684210526316,
+      "loss": 4.6317,
+      "step": 261
+    },
+    {
+      "epoch": 0.011031578947368422,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00027578947368421053,
+      "loss": 4.4151,
+      "step": 262
+    },
+    {
+      "epoch": 0.011073684210526315,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0002768421052631579,
+      "loss": 4.8259,
+      "step": 263
+    },
+    {
+      "epoch": 0.01111578947368421,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.0002778947368421053,
+      "loss": 4.8518,
+      "step": 264
+    },
+    {
+      "epoch": 0.011157894736842104,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.0002789473684210526,
+      "loss": 4.4706,
+      "step": 265
+    },
+    {
+      "epoch": 0.0112,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00028000000000000003,
+      "loss": 4.7413,
+      "step": 266
+    },
+    {
+      "epoch": 0.011242105263157895,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.0002810526315789474,
+      "loss": 4.7639,
+      "step": 267
+    },
+    {
+      "epoch": 0.011284210526315789,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0002821052631578948,
+      "loss": 4.7669,
+      "step": 268
+    },
+    {
+      "epoch": 0.011326315789473684,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.0002831578947368421,
+      "loss": 4.33,
+      "step": 269
+    },
+    {
+      "epoch": 0.01136842105263158,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00028421052631578947,
+      "loss": 4.4285,
+      "step": 270
+    },
+    {
+      "epoch": 0.011410526315789473,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00028526315789473685,
+      "loss": 4.5333,
+      "step": 271
+    },
+    {
+      "epoch": 0.011452631578947369,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.0002863157894736842,
+      "loss": 4.6635,
+      "step": 272
+    },
+    {
+      "epoch": 0.011494736842105262,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0002873684210526316,
+      "loss": 4.5812,
+      "step": 273
+    },
+    {
+      "epoch": 0.011536842105263158,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00028842105263157897,
+      "loss": 4.8065,
+      "step": 274
+    },
+    {
+      "epoch": 0.011578947368421053,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00028947368421052634,
+      "loss": 4.6259,
+      "step": 275
+    },
+    {
+      "epoch": 0.011621052631578947,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0002905263157894737,
+      "loss": 4.648,
+      "step": 276
+    },
+    {
+      "epoch": 0.011663157894736842,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00029157894736842104,
+      "loss": 4.7563,
+      "step": 277
+    },
+    {
+      "epoch": 0.011705263157894738,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.0002926315789473684,
+      "loss": 4.7388,
+      "step": 278
+    },
+    {
+      "epoch": 0.011747368421052631,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0002936842105263158,
+      "loss": 4.5267,
+      "step": 279
+    },
+    {
+      "epoch": 0.011789473684210527,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.00029473684210526316,
+      "loss": 4.7164,
+      "step": 280
+    },
+    {
+      "epoch": 0.01183157894736842,
+      "grad_norm": 0.8515625,
+      "learning_rate": 0.00029578947368421053,
+      "loss": 4.728,
+      "step": 281
+    },
+    {
+      "epoch": 0.011873684210526316,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.0002968421052631579,
+      "loss": 4.6658,
+      "step": 282
+    },
+    {
+      "epoch": 0.011915789473684211,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0002978947368421053,
+      "loss": 4.8512,
+      "step": 283
+    },
+    {
+      "epoch": 0.011957894736842105,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00029894736842105265,
+      "loss": 4.3698,
+      "step": 284
+    },
+    {
+      "epoch": 0.012,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0003,
+      "loss": 4.5798,
+      "step": 285
+    },
+    {
+      "epoch": 0.012042105263157894,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00030105263157894735,
+      "loss": 4.8204,
+      "step": 286
+    },
+    {
+      "epoch": 0.01208421052631579,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0003021052631578947,
+      "loss": 4.8963,
+      "step": 287
+    },
+    {
+      "epoch": 0.012126315789473685,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00030315789473684215,
+      "loss": 4.581,
+      "step": 288
+    },
+    {
+      "epoch": 0.012168421052631578,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00030421052631578947,
+      "loss": 4.5804,
+      "step": 289
+    },
+    {
+      "epoch": 0.012210526315789474,
+      "grad_norm": 0.921875,
+      "learning_rate": 0.00030526315789473684,
+      "loss": 4.5165,
+      "step": 290
+    },
+    {
+      "epoch": 0.01225263157894737,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0003063157894736842,
+      "loss": 4.7213,
+      "step": 291
+    },
+    {
+      "epoch": 0.012294736842105263,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.0003073684210526316,
+      "loss": 4.4559,
+      "step": 292
+    },
+    {
+      "epoch": 0.012336842105263158,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0003084210526315789,
+      "loss": 4.6154,
+      "step": 293
+    },
+    {
+      "epoch": 0.012378947368421052,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.0003094736842105263,
+      "loss": 4.3068,
+      "step": 294
+    },
+    {
+      "epoch": 0.012421052631578947,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0003105263157894737,
+      "loss": 4.6104,
+      "step": 295
+    },
+    {
+      "epoch": 0.012463157894736843,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0003115789473684211,
+      "loss": 4.6329,
+      "step": 296
+    },
+    {
+      "epoch": 0.012505263157894736,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.0003126315789473684,
+      "loss": 4.497,
+      "step": 297
+    },
+    {
+      "epoch": 0.012547368421052632,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0003136842105263158,
+      "loss": 4.5875,
+      "step": 298
+    },
+    {
+      "epoch": 0.012589473684210527,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00031473684210526316,
+      "loss": 4.1527,
+      "step": 299
+    },
+    {
+      "epoch": 0.01263157894736842,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.00031578947368421053,
+      "loss": 4.6874,
+      "step": 300
+    },
+    {
+      "epoch": 0.012673684210526316,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00031684210526315785,
+      "loss": 4.4477,
+      "step": 301
+    },
+    {
+      "epoch": 0.01271578947368421,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0003178947368421053,
+      "loss": 4.5187,
+      "step": 302
+    },
+    {
+      "epoch": 0.012757894736842105,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00031894736842105265,
+      "loss": 4.5969,
+      "step": 303
+    },
+    {
+      "epoch": 0.0128,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00032,
+      "loss": 4.7898,
+      "step": 304
+    },
+    {
+      "epoch": 0.012842105263157894,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0003210526315789474,
+      "loss": 4.7483,
+      "step": 305
+    },
+    {
+      "epoch": 0.01288421052631579,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0003221052631578947,
+      "loss": 4.2825,
+      "step": 306
+    },
+    {
+      "epoch": 0.012926315789473685,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0003231578947368421,
+      "loss": 4.6983,
+      "step": 307
+    },
+    {
+      "epoch": 0.012968421052631579,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00032421052631578947,
+      "loss": 4.1973,
+      "step": 308
+    },
+    {
+      "epoch": 0.013010526315789474,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0003252631578947369,
+      "loss": 4.4216,
+      "step": 309
+    },
+    {
+      "epoch": 0.013052631578947368,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.0003263157894736842,
+      "loss": 4.6639,
+      "step": 310
+    },
+    {
+      "epoch": 0.013094736842105263,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0003273684210526316,
+      "loss": 4.6799,
+      "step": 311
+    },
+    {
+      "epoch": 0.013136842105263159,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00032842105263157896,
+      "loss": 4.2909,
+      "step": 312
+    },
+    {
+      "epoch": 0.013178947368421052,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00032947368421052634,
+      "loss": 4.9743,
+      "step": 313
+    },
+    {
+      "epoch": 0.013221052631578948,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00033052631578947366,
+      "loss": 4.3255,
+      "step": 314
+    },
+    {
+      "epoch": 0.013263157894736841,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00033157894736842103,
+      "loss": 4.5748,
+      "step": 315
+    },
+    {
+      "epoch": 0.013305263157894737,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00033263157894736846,
+      "loss": 4.3673,
+      "step": 316
+    },
+    {
+      "epoch": 0.013347368421052632,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.00033368421052631583,
+      "loss": 4.656,
+      "step": 317
+    },
+    {
+      "epoch": 0.013389473684210526,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00033473684210526315,
+      "loss": 4.5088,
+      "step": 318
+    },
+    {
+      "epoch": 0.013431578947368421,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00033578947368421053,
+      "loss": 4.6969,
+      "step": 319
+    },
+    {
+      "epoch": 0.013473684210526317,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0003368421052631579,
+      "loss": 4.7857,
+      "step": 320
+    },
+    {
+      "epoch": 0.01351578947368421,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.0003378947368421053,
+      "loss": 4.6076,
+      "step": 321
+    },
+    {
+      "epoch": 0.013557894736842106,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.0003389473684210526,
+      "loss": 4.1856,
+      "step": 322
+    },
+    {
+      "epoch": 0.0136,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00034,
+      "loss": 4.5701,
+      "step": 323
+    },
+    {
+      "epoch": 0.013642105263157895,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0003410526315789474,
+      "loss": 4.4323,
+      "step": 324
+    },
+    {
+      "epoch": 0.01368421052631579,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00034210526315789477,
+      "loss": 4.3803,
+      "step": 325
+    },
+    {
+      "epoch": 0.013726315789473684,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.0003431578947368421,
+      "loss": 4.2215,
+      "step": 326
+    },
+    {
+      "epoch": 0.013768421052631579,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00034421052631578947,
+      "loss": 4.7205,
+      "step": 327
+    },
+    {
+      "epoch": 0.013810526315789474,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00034526315789473684,
+      "loss": 4.6807,
+      "step": 328
+    },
+    {
+      "epoch": 0.013852631578947368,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0003463157894736842,
+      "loss": 4.6929,
+      "step": 329
+    },
+    {
+      "epoch": 0.013894736842105264,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.0003473684210526316,
+      "loss": 4.3836,
+      "step": 330
+    },
+    {
+      "epoch": 0.013936842105263157,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00034842105263157896,
+      "loss": 4.5278,
+      "step": 331
+    },
+    {
+      "epoch": 0.013978947368421053,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00034947368421052634,
+      "loss": 4.6396,
+      "step": 332
+    },
+    {
+      "epoch": 0.014021052631578948,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0003505263157894737,
+      "loss": 4.5432,
+      "step": 333
+    },
+    {
+      "epoch": 0.014063157894736842,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00035157894736842103,
+      "loss": 4.6171,
+      "step": 334
+    },
+    {
+      "epoch": 0.014105263157894737,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0003526315789473684,
+      "loss": 4.6272,
+      "step": 335
+    },
+    {
+      "epoch": 0.014147368421052632,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.0003536842105263158,
+      "loss": 4.3247,
+      "step": 336
+    },
+    {
+      "epoch": 0.014189473684210526,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.0003547368421052632,
+      "loss": 4.3696,
+      "step": 337
+    },
+    {
+      "epoch": 0.014231578947368421,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0003557894736842105,
+      "loss": 4.7596,
+      "step": 338
+    },
+    {
+      "epoch": 0.014273684210526315,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0003568421052631579,
+      "loss": 4.3865,
+      "step": 339
+    },
+    {
+      "epoch": 0.01431578947368421,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.0003578947368421053,
+      "loss": 4.4291,
+      "step": 340
+    },
+    {
+      "epoch": 0.014357894736842106,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00035894736842105265,
+      "loss": 4.397,
+      "step": 341
+    },
+    {
+      "epoch": 0.0144,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00035999999999999997,
+      "loss": 4.5021,
+      "step": 342
+    },
+    {
+      "epoch": 0.014442105263157895,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.00036105263157894734,
+      "loss": 4.3352,
+      "step": 343
+    },
+    {
+      "epoch": 0.014484210526315789,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00036210526315789477,
+      "loss": 4.3564,
+      "step": 344
+    },
+    {
+      "epoch": 0.014526315789473684,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00036315789473684214,
+      "loss": 4.92,
+      "step": 345
+    },
+    {
+      "epoch": 0.01456842105263158,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.0003642105263157895,
+      "loss": 4.4194,
+      "step": 346
+    },
+    {
+      "epoch": 0.014610526315789473,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00036526315789473684,
+      "loss": 4.4044,
+      "step": 347
+    },
+    {
+      "epoch": 0.014652631578947369,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0003663157894736842,
+      "loss": 4.3202,
+      "step": 348
+    },
+    {
+      "epoch": 0.014694736842105264,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0003673684210526316,
+      "loss": 4.2268,
+      "step": 349
+    },
+    {
+      "epoch": 0.014736842105263158,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00036842105263157896,
+      "loss": 4.7891,
+      "step": 350
+    },
+    {
+      "epoch": 0.014778947368421053,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00036947368421052633,
+      "loss": 4.4866,
+      "step": 351
+    },
+    {
+      "epoch": 0.014821052631578947,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.0003705263157894737,
+      "loss": 4.3846,
+      "step": 352
+    },
+    {
+      "epoch": 0.014863157894736842,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.0003715789473684211,
+      "loss": 4.5073,
+      "step": 353
+    },
+    {
+      "epoch": 0.014905263157894737,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00037263157894736846,
+      "loss": 4.4347,
+      "step": 354
+    },
+    {
+      "epoch": 0.014947368421052631,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.0003736842105263158,
+      "loss": 4.293,
+      "step": 355
+    },
+    {
+      "epoch": 0.014989473684210526,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00037473684210526315,
+      "loss": 4.4316,
+      "step": 356
+    },
+    {
+      "epoch": 0.015031578947368422,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.0003757894736842105,
+      "loss": 4.4448,
+      "step": 357
+    },
+    {
+      "epoch": 0.015073684210526316,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0003768421052631579,
+      "loss": 4.4419,
+      "step": 358
+    },
+    {
+      "epoch": 0.015115789473684211,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00037789473684210527,
+      "loss": 4.7215,
+      "step": 359
+    },
+    {
+      "epoch": 0.015157894736842105,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00037894736842105265,
+      "loss": 4.4125,
+      "step": 360
+    },
+    {
+      "epoch": 0.0152,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00038,
+      "loss": 4.3999,
+      "step": 361
+    },
+    {
+      "epoch": 0.015242105263157895,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.0003810526315789474,
+      "loss": 4.7075,
+      "step": 362
+    },
+    {
+      "epoch": 0.015284210526315789,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0003821052631578947,
+      "loss": 4.631,
+      "step": 363
+    },
+    {
+      "epoch": 0.015326315789473684,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.0003831578947368421,
+      "loss": 4.5804,
+      "step": 364
+    },
+    {
+      "epoch": 0.015368421052631578,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00038421052631578946,
+      "loss": 4.5522,
+      "step": 365
+    },
+    {
+      "epoch": 0.015410526315789473,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0003852631578947369,
+      "loss": 4.0813,
+      "step": 366
+    },
+    {
+      "epoch": 0.015452631578947369,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.0003863157894736842,
+      "loss": 4.4482,
+      "step": 367
+    },
+    {
+      "epoch": 0.015494736842105263,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0003873684210526316,
+      "loss": 4.4242,
+      "step": 368
+    },
+    {
+      "epoch": 0.015536842105263158,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00038842105263157896,
+      "loss": 4.4984,
+      "step": 369
+    },
+    {
+      "epoch": 0.015578947368421053,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00038947368421052633,
+      "loss": 4.338,
+      "step": 370
+    },
+    {
+      "epoch": 0.015621052631578947,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00039052631578947365,
+      "loss": 4.5405,
+      "step": 371
+    },
+    {
+      "epoch": 0.01566315789473684,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.000391578947368421,
+      "loss": 4.4824,
+      "step": 372
+    },
+    {
+      "epoch": 0.015705263157894736,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00039263157894736846,
+      "loss": 4.3563,
+      "step": 373
+    },
+    {
+      "epoch": 0.01574736842105263,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00039368421052631583,
+      "loss": 4.4242,
+      "step": 374
+    },
+    {
+      "epoch": 0.015789473684210527,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00039473684210526315,
+      "loss": 4.2607,
+      "step": 375
+    },
+    {
+      "epoch": 0.015831578947368422,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0003957894736842105,
+      "loss": 4.2254,
+      "step": 376
+    },
+    {
+      "epoch": 0.015873684210526314,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0003968421052631579,
+      "loss": 4.1182,
+      "step": 377
+    },
+    {
+      "epoch": 0.01591578947368421,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00039789473684210527,
+      "loss": 4.8709,
+      "step": 378
+    },
+    {
+      "epoch": 0.015957894736842105,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0003989473684210526,
+      "loss": 4.3529,
+      "step": 379
+    },
+    {
+      "epoch": 0.016,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0004,
+      "loss": 4.0789,
+      "step": 380
+    },
+    {
+      "epoch": 0.016042105263157896,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0004010526315789474,
+      "loss": 4.2138,
+      "step": 381
+    },
+    {
+      "epoch": 0.01608421052631579,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00040210526315789477,
+      "loss": 4.3717,
+      "step": 382
+    },
+    {
+      "epoch": 0.016126315789473683,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0004031578947368421,
+      "loss": 4.3187,
+      "step": 383
+    },
+    {
+      "epoch": 0.01616842105263158,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00040421052631578946,
+      "loss": 4.4462,
+      "step": 384
+    },
+    {
+      "epoch": 0.016210526315789474,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.00040526315789473684,
+      "loss": 3.9189,
+      "step": 385
+    },
+    {
+      "epoch": 0.01625263157894737,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0004063157894736842,
+      "loss": 4.4457,
+      "step": 386
+    },
+    {
+      "epoch": 0.016294736842105265,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00040736842105263164,
+      "loss": 4.5699,
+      "step": 387
+    },
+    {
+      "epoch": 0.016336842105263157,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00040842105263157896,
+      "loss": 4.2877,
+      "step": 388
+    },
+    {
+      "epoch": 0.016378947368421052,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00040947368421052633,
+      "loss": 4.271,
+      "step": 389
+    },
+    {
+      "epoch": 0.016421052631578947,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0004105263157894737,
+      "loss": 4.2374,
+      "step": 390
+    },
+    {
+      "epoch": 0.016463157894736843,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0004115789473684211,
+      "loss": 4.2013,
+      "step": 391
+    },
+    {
+      "epoch": 0.016505263157894738,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0004126315789473684,
+      "loss": 4.4356,
+      "step": 392
+    },
+    {
+      "epoch": 0.01654736842105263,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0004136842105263158,
+      "loss": 4.4449,
+      "step": 393
+    },
+    {
+      "epoch": 0.016589473684210525,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0004147368421052632,
+      "loss": 4.0937,
+      "step": 394
+    },
+    {
+      "epoch": 0.01663157894736842,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0004157894736842106,
+      "loss": 4.2837,
+      "step": 395
+    },
+    {
+      "epoch": 0.016673684210526316,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0004168421052631579,
+      "loss": 4.4151,
+      "step": 396
+    },
+    {
+      "epoch": 0.01671578947368421,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00041789473684210527,
+      "loss": 4.6205,
+      "step": 397
+    },
+    {
+      "epoch": 0.016757894736842104,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00041894736842105264,
+      "loss": 4.3583,
+      "step": 398
+    },
+    {
+      "epoch": 0.0168,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00042,
+      "loss": 4.3391,
+      "step": 399
+    },
+    {
+      "epoch": 0.016842105263157894,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00042105263157894734,
+      "loss": 4.5594,
+      "step": 400
+    },
+    {
+      "epoch": 0.01688421052631579,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00042210526315789477,
+      "loss": 4.1017,
+      "step": 401
+    },
+    {
+      "epoch": 0.016926315789473685,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00042315789473684214,
+      "loss": 4.1371,
+      "step": 402
+    },
+    {
+      "epoch": 0.01696842105263158,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0004242105263157895,
+      "loss": 4.2816,
+      "step": 403
+    },
+    {
+      "epoch": 0.017010526315789472,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00042526315789473683,
+      "loss": 4.3082,
+      "step": 404
+    },
+    {
+      "epoch": 0.017052631578947368,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0004263157894736842,
+      "loss": 4.4282,
+      "step": 405
+    },
+    {
+      "epoch": 0.017094736842105263,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0004273684210526316,
+      "loss": 4.6419,
+      "step": 406
+    },
+    {
+      "epoch": 0.01713684210526316,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00042842105263157896,
+      "loss": 4.5408,
+      "step": 407
+    },
+    {
+      "epoch": 0.017178947368421054,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00042947368421052633,
+      "loss": 4.3513,
+      "step": 408
+    },
+    {
+      "epoch": 0.017221052631578946,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0004305263157894737,
+      "loss": 4.3757,
+      "step": 409
+    },
+    {
+      "epoch": 0.01726315789473684,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0004315789473684211,
+      "loss": 4.2217,
+      "step": 410
+    },
+    {
+      "epoch": 0.017305263157894737,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00043263157894736845,
+      "loss": 4.2977,
+      "step": 411
+    },
+    {
+      "epoch": 0.017347368421052632,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00043368421052631577,
+      "loss": 4.2425,
+      "step": 412
+    },
+    {
+      "epoch": 0.017389473684210528,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00043473684210526315,
+      "loss": 4.204,
+      "step": 413
+    },
+    {
+      "epoch": 0.01743157894736842,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0004357894736842105,
+      "loss": 4.3425,
+      "step": 414
+    },
+    {
+      "epoch": 0.017473684210526315,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00043684210526315795,
+      "loss": 4.4786,
+      "step": 415
+    },
+    {
+      "epoch": 0.01751578947368421,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00043789473684210527,
+      "loss": 4.1803,
+      "step": 416
+    },
+    {
+      "epoch": 0.017557894736842106,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00043894736842105264,
+      "loss": 4.3962,
+      "step": 417
+    },
+    {
+      "epoch": 0.0176,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00044,
+      "loss": 4.1655,
+      "step": 418
+    },
+    {
+      "epoch": 0.017642105263157896,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0004410526315789474,
+      "loss": 4.4406,
+      "step": 419
+    },
+    {
+      "epoch": 0.01768421052631579,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0004421052631578947,
+      "loss": 3.8846,
+      "step": 420
+    },
+    {
+      "epoch": 0.017726315789473684,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0004431578947368421,
+      "loss": 4.44,
+      "step": 421
+    },
+    {
+      "epoch": 0.01776842105263158,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0004442105263157895,
+      "loss": 4.4924,
+      "step": 422
+    },
+    {
+      "epoch": 0.017810526315789475,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.0004452631578947369,
+      "loss": 4.1203,
+      "step": 423
+    },
+    {
+      "epoch": 0.01785263157894737,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0004463157894736842,
+      "loss": 4.1306,
+      "step": 424
+    },
+    {
+      "epoch": 0.017894736842105262,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0004473684210526316,
+      "loss": 4.5186,
+      "step": 425
+    },
+    {
+      "epoch": 0.017936842105263157,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00044842105263157895,
+      "loss": 4.1231,
+      "step": 426
+    },
+    {
+      "epoch": 0.017978947368421053,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00044947368421052633,
+      "loss": 4.4977,
+      "step": 427
+    },
+    {
+      "epoch": 0.018021052631578948,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00045052631578947365,
+      "loss": 4.1505,
+      "step": 428
+    },
+    {
+      "epoch": 0.018063157894736843,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.0004515789473684211,
+      "loss": 4.3092,
+      "step": 429
+    },
+    {
+      "epoch": 0.018105263157894735,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00045263157894736845,
+      "loss": 4.2571,
+      "step": 430
+    },
+    {
+      "epoch": 0.01814736842105263,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0004536842105263158,
+      "loss": 4.574,
+      "step": 431
+    },
+    {
+      "epoch": 0.018189473684210526,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0004547368421052632,
+      "loss": 4.4953,
+      "step": 432
+    },
+    {
+      "epoch": 0.01823157894736842,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0004557894736842105,
+      "loss": 4.3898,
+      "step": 433
+    },
+    {
+      "epoch": 0.018273684210526317,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0004568421052631579,
+      "loss": 4.3192,
+      "step": 434
+    },
+    {
+      "epoch": 0.01831578947368421,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00045789473684210527,
+      "loss": 4.3572,
+      "step": 435
+    },
+    {
+      "epoch": 0.018357894736842104,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00045894736842105264,
+      "loss": 4.4823,
+      "step": 436
+    },
+    {
+      "epoch": 0.0184,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00046,
+      "loss": 4.3421,
+      "step": 437
+    },
+    {
+      "epoch": 0.018442105263157895,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0004610526315789474,
+      "loss": 4.2476,
+      "step": 438
+    },
+    {
+      "epoch": 0.01848421052631579,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00046210526315789476,
+      "loss": 4.193,
+      "step": 439
+    },
+    {
+      "epoch": 0.018526315789473686,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00046315789473684214,
+      "loss": 4.476,
+      "step": 440
+    },
+    {
+      "epoch": 0.018568421052631578,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.00046421052631578946,
+      "loss": 4.8103,
+      "step": 441
+    },
+    {
+      "epoch": 0.018610526315789473,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00046526315789473683,
+      "loss": 4.257,
+      "step": 442
+    },
+    {
+      "epoch": 0.01865263157894737,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0004663157894736842,
+      "loss": 4.1948,
+      "step": 443
+    },
+    {
+      "epoch": 0.018694736842105264,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00046736842105263163,
+      "loss": 4.1955,
+      "step": 444
+    },
+    {
+      "epoch": 0.01873684210526316,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00046842105263157895,
+      "loss": 4.251,
+      "step": 445
+    },
+    {
+      "epoch": 0.01877894736842105,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0004694736842105263,
+      "loss": 4.3446,
+      "step": 446
+    },
+    {
+      "epoch": 0.018821052631578947,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0004705263157894737,
+      "loss": 4.2658,
+      "step": 447
+    },
+    {
+      "epoch": 0.018863157894736842,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0004715789473684211,
+      "loss": 4.4793,
+      "step": 448
+    },
+    {
+      "epoch": 0.018905263157894737,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0004726315789473684,
+      "loss": 4.1144,
+      "step": 449
+    },
+    {
+      "epoch": 0.018947368421052633,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00047368421052631577,
+      "loss": 4.1679,
+      "step": 450
+    },
+    {
+      "epoch": 0.018989473684210525,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0004747368421052632,
+      "loss": 4.3472,
+      "step": 451
+    },
+    {
+      "epoch": 0.01903157894736842,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00047578947368421057,
+      "loss": 4.237,
+      "step": 452
+    },
+    {
+      "epoch": 0.019073684210526316,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0004768421052631579,
+      "loss": 4.4889,
+      "step": 453
+    },
+    {
+      "epoch": 0.01911578947368421,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00047789473684210526,
+      "loss": 4.0149,
+      "step": 454
+    },
+    {
+      "epoch": 0.019157894736842106,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00047894736842105264,
+      "loss": 4.1984,
+      "step": 455
+    },
+    {
+      "epoch": 0.0192,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00048,
+      "loss": 3.8028,
+      "step": 456
+    },
+    {
+      "epoch": 0.019242105263157894,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00048105263157894733,
+      "loss": 3.8972,
+      "step": 457
+    },
+    {
+      "epoch": 0.01928421052631579,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00048210526315789476,
+      "loss": 4.0164,
+      "step": 458
+    },
+    {
+      "epoch": 0.019326315789473685,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.00048315789473684213,
+      "loss": 4.0093,
+      "step": 459
+    },
+    {
+      "epoch": 0.01936842105263158,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0004842105263157895,
+      "loss": 4.4367,
+      "step": 460
+    },
+    {
+      "epoch": 0.019410526315789475,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00048526315789473683,
+      "loss": 4.4371,
+      "step": 461
+    },
+    {
+      "epoch": 0.019452631578947367,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0004863157894736842,
+      "loss": 4.1765,
+      "step": 462
+    },
+    {
+      "epoch": 0.019494736842105263,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0004873684210526316,
+      "loss": 4.302,
+      "step": 463
+    },
+    {
+      "epoch": 0.019536842105263158,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.000488421052631579,
+      "loss": 4.3553,
+      "step": 464
+    },
+    {
+      "epoch": 0.019578947368421053,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0004894736842105264,
+      "loss": 4.2031,
+      "step": 465
+    },
+    {
+      "epoch": 0.01962105263157895,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0004905263157894737,
+      "loss": 4.2276,
+      "step": 466
+    },
+    {
+      "epoch": 0.01966315789473684,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.000491578947368421,
+      "loss": 4.2627,
+      "step": 467
+    },
+    {
+      "epoch": 0.019705263157894736,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0004926315789473684,
+      "loss": 4.4481,
+      "step": 468
+    },
+    {
+      "epoch": 0.01974736842105263,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0004936842105263158,
+      "loss": 4.4823,
+      "step": 469
+    },
+    {
+      "epoch": 0.019789473684210527,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0004947368421052632,
+      "loss": 4.2895,
+      "step": 470
+    },
+    {
+      "epoch": 0.019831578947368422,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0004957894736842105,
+      "loss": 4.1776,
+      "step": 471
+    },
+    {
+      "epoch": 0.019873684210526314,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0004968421052631579,
+      "loss": 4.1281,
+      "step": 472
+    },
+    {
+      "epoch": 0.01991578947368421,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0004978947368421053,
+      "loss": 4.2832,
+      "step": 473
+    },
+    {
+      "epoch": 0.019957894736842105,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.0004989473684210527,
+      "loss": 4.8963,
+      "step": 474
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0005,
+      "loss": 4.3687,
+      "step": 475
+    },
+    {
+      "epoch": 0.020042105263157896,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.0004999999977226469,
+      "loss": 4.3004,
+      "step": 476
+    },
+    {
+      "epoch": 0.02008421052631579,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0004999999908905876,
+      "loss": 4.3485,
+      "step": 477
+    },
+    {
+      "epoch": 0.020126315789473683,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0004999999795038221,
+      "loss": 4.2256,
+      "step": 478
+    },
+    {
+      "epoch": 0.02016842105263158,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0004999999635623508,
+      "loss": 3.9872,
+      "step": 479
+    },
+    {
+      "epoch": 0.020210526315789474,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0004999999430661738,
+      "loss": 4.0898,
+      "step": 480
+    },
+    {
+      "epoch": 0.02025263157894737,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0004999999180152917,
+      "loss": 4.3916,
+      "step": 481
+    },
+    {
+      "epoch": 0.020294736842105265,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0004999998884097047,
+      "loss": 4.3878,
+      "step": 482
+    },
+    {
+      "epoch": 0.020336842105263157,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0004999998542494136,
+      "loss": 4.3222,
+      "step": 483
+    },
+    {
+      "epoch": 0.020378947368421052,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0004999998155344189,
+      "loss": 4.4022,
+      "step": 484
+    },
+    {
+      "epoch": 0.020421052631578947,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0004999997722647211,
+      "loss": 4.2781,
+      "step": 485
+    },
+    {
+      "epoch": 0.020463157894736843,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0004999997244403214,
+      "loss": 4.3358,
+      "step": 486
+    },
+    {
+      "epoch": 0.020505263157894738,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0004999996720612204,
+      "loss": 3.7881,
+      "step": 487
+    },
+    {
+      "epoch": 0.02054736842105263,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.000499999615127419,
+      "loss": 4.8277,
+      "step": 488
+    },
+    {
+      "epoch": 0.020589473684210526,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0004999995536389185,
+      "loss": 4.2344,
+      "step": 489
+    },
+    {
+      "epoch": 0.02063157894736842,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0004999994875957198,
+      "loss": 4.3953,
+      "step": 490
+    },
+    {
+      "epoch": 0.020673684210526316,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0004999994169978242,
+      "loss": 4.4188,
+      "step": 491
+    },
+    {
+      "epoch": 0.02071578947368421,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0004999993418452329,
+      "loss": 4.385,
+      "step": 492
+    },
+    {
+      "epoch": 0.020757894736842104,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0004999992621379474,
+      "loss": 4.3987,
+      "step": 493
+    },
+    {
+      "epoch": 0.0208,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.000499999177875969,
+      "loss": 4.1516,
+      "step": 494
+    },
+    {
+      "epoch": 0.020842105263157894,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0004999990890592994,
+      "loss": 3.6346,
+      "step": 495
+    },
+    {
+      "epoch": 0.02088421052631579,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0004999989956879402,
+      "loss": 4.209,
+      "step": 496
+    },
+    {
+      "epoch": 0.020926315789473685,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0004999988977618929,
+      "loss": 4.2969,
+      "step": 497
+    },
+    {
+      "epoch": 0.02096842105263158,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0004999987952811595,
+      "loss": 4.2258,
+      "step": 498
+    },
+    {
+      "epoch": 0.021010526315789473,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0004999986882457417,
+      "loss": 3.8085,
+      "step": 499
+    },
+    {
+      "epoch": 0.021052631578947368,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0004999985766556415,
+      "loss": 4.1657,
+      "step": 500
+    },
+    {
+      "epoch": 0.021094736842105263,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0004999984605108611,
+      "loss": 4.2793,
+      "step": 501
+    },
+    {
+      "epoch": 0.02113684210526316,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.0004999983398114024,
+      "loss": 3.8602,
+      "step": 502
+    },
+    {
+      "epoch": 0.021178947368421054,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0004999982145572678,
+      "loss": 4.1351,
+      "step": 503
+    },
+    {
+      "epoch": 0.021221052631578946,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0004999980847484593,
+      "loss": 3.9963,
+      "step": 504
+    },
+    {
+      "epoch": 0.02126315789473684,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.0004999979503849796,
+      "loss": 4.3515,
+      "step": 505
+    },
+    {
+      "epoch": 0.021305263157894737,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0004999978114668309,
+      "loss": 4.1705,
+      "step": 506
+    },
+    {
+      "epoch": 0.021347368421052632,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0004999976679940158,
+      "loss": 4.3977,
+      "step": 507
+    },
+    {
+      "epoch": 0.021389473684210528,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0004999975199665369,
+      "loss": 4.8703,
+      "step": 508
+    },
+    {
+      "epoch": 0.02143157894736842,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.000499997367384397,
+      "loss": 4.594,
+      "step": 509
+    },
+    {
+      "epoch": 0.021473684210526315,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0004999972102475987,
+      "loss": 4.0068,
+      "step": 510
+    },
+    {
+      "epoch": 0.02151578947368421,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0004999970485561451,
+      "loss": 4.3193,
+      "step": 511
+    },
+    {
+      "epoch": 0.021557894736842106,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0004999968823100389,
+      "loss": 4.0248,
+      "step": 512
+    },
+    {
+      "epoch": 0.0216,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0004999967115092833,
+      "loss": 4.5359,
+      "step": 513
+    },
+    {
+      "epoch": 0.021642105263157893,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0004999965361538813,
+      "loss": 4.3914,
+      "step": 514
+    },
+    {
+      "epoch": 0.02168421052631579,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0004999963562438361,
+      "loss": 4.3953,
+      "step": 515
+    },
+    {
+      "epoch": 0.021726315789473684,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0004999961717791512,
+      "loss": 4.6134,
+      "step": 516
+    },
+    {
+      "epoch": 0.02176842105263158,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0004999959827598296,
+      "loss": 4.6595,
+      "step": 517
+    },
+    {
+      "epoch": 0.021810526315789475,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.000499995789185875,
+      "loss": 3.9478,
+      "step": 518
+    },
+    {
+      "epoch": 0.02185263157894737,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0004999955910572908,
+      "loss": 4.493,
+      "step": 519
+    },
+    {
+      "epoch": 0.021894736842105262,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0004999953883740808,
+      "loss": 4.1525,
+      "step": 520
+    },
+    {
+      "epoch": 0.021936842105263157,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0004999951811362484,
+      "loss": 4.106,
+      "step": 521
+    },
+    {
+      "epoch": 0.021978947368421053,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0004999949693437976,
+      "loss": 4.5819,
+      "step": 522
+    },
+    {
+      "epoch": 0.022021052631578948,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0004999947529967323,
+      "loss": 4.3143,
+      "step": 523
+    },
+    {
+      "epoch": 0.022063157894736844,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0004999945320950562,
+      "loss": 4.2218,
+      "step": 524
+    },
+    {
+      "epoch": 0.022105263157894735,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0004999943066387734,
+      "loss": 4.0465,
+      "step": 525
+    },
+    {
+      "epoch": 0.02214736842105263,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0004999940766278882,
+      "loss": 4.2427,
+      "step": 526
+    },
+    {
+      "epoch": 0.022189473684210526,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0004999938420624046,
+      "loss": 3.7561,
+      "step": 527
+    },
+    {
+      "epoch": 0.02223157894736842,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0004999936029423269,
+      "loss": 3.9305,
+      "step": 528
+    },
+    {
+      "epoch": 0.022273684210526317,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0004999933592676596,
+      "loss": 4.5952,
+      "step": 529
+    },
+    {
+      "epoch": 0.02231578947368421,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0004999931110384069,
+      "loss": 3.8931,
+      "step": 530
+    },
+    {
+      "epoch": 0.022357894736842104,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0004999928582545734,
+      "loss": 4.1726,
+      "step": 531
+    },
+    {
+      "epoch": 0.0224,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0004999926009161639,
+      "loss": 3.9406,
+      "step": 532
+    },
+    {
+      "epoch": 0.022442105263157895,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0004999923390231829,
+      "loss": 4.0424,
+      "step": 533
+    },
+    {
+      "epoch": 0.02248421052631579,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0004999920725756352,
+      "loss": 4.2692,
+      "step": 534
+    },
+    {
+      "epoch": 0.022526315789473682,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0004999918015735256,
+      "loss": 3.9708,
+      "step": 535
+    },
+    {
+      "epoch": 0.022568421052631578,
+      "grad_norm": 0.77734375,
+      "learning_rate": 0.0004999915260168592,
+      "loss": 3.944,
+      "step": 536
+    },
+    {
+      "epoch": 0.022610526315789473,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0004999912459056409,
+      "loss": 4.3285,
+      "step": 537
+    },
+    {
+      "epoch": 0.02265263157894737,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0004999909612398759,
+      "loss": 4.3843,
+      "step": 538
+    },
+    {
+      "epoch": 0.022694736842105264,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0004999906720195692,
+      "loss": 4.2188,
+      "step": 539
+    },
+    {
+      "epoch": 0.02273684210526316,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004999903782447262,
+      "loss": 4.3927,
+      "step": 540
+    },
+    {
+      "epoch": 0.02277894736842105,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0004999900799153522,
+      "loss": 4.1103,
+      "step": 541
+    },
+    {
+      "epoch": 0.022821052631578947,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0004999897770314527,
+      "loss": 3.994,
+      "step": 542
+    },
+    {
+      "epoch": 0.022863157894736842,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0004999894695930332,
+      "loss": 4.3853,
+      "step": 543
+    },
+    {
+      "epoch": 0.022905263157894738,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0004999891576000993,
+      "loss": 3.9278,
+      "step": 544
+    },
+    {
+      "epoch": 0.022947368421052633,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0004999888410526566,
+      "loss": 4.1244,
+      "step": 545
+    },
+    {
+      "epoch": 0.022989473684210525,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.000499988519950711,
+      "loss": 4.4078,
+      "step": 546
+    },
+    {
+      "epoch": 0.02303157894736842,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0004999881942942683,
+      "loss": 3.7988,
+      "step": 547
+    },
+    {
+      "epoch": 0.023073684210526316,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0004999878640833345,
+      "loss": 4.393,
+      "step": 548
+    },
+    {
+      "epoch": 0.02311578947368421,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0004999875293179155,
+      "loss": 4.2036,
+      "step": 549
+    },
+    {
+      "epoch": 0.023157894736842106,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0004999871899980173,
+      "loss": 3.8576,
+      "step": 550
+    },
+    {
+      "epoch": 0.0232,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.0004999868461236464,
+      "loss": 3.9127,
+      "step": 551
+    },
+    {
+      "epoch": 0.023242105263157894,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0004999864976948088,
+      "loss": 4.1617,
+      "step": 552
+    },
+    {
+      "epoch": 0.02328421052631579,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0004999861447115109,
+      "loss": 4.3166,
+      "step": 553
+    },
+    {
+      "epoch": 0.023326315789473685,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0004999857871737592,
+      "loss": 4.219,
+      "step": 554
+    },
+    {
+      "epoch": 0.02336842105263158,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0004999854250815602,
+      "loss": 4.4197,
+      "step": 555
+    },
+    {
+      "epoch": 0.023410526315789475,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.0004999850584349205,
+      "loss": 3.9932,
+      "step": 556
+    },
+    {
+      "epoch": 0.023452631578947367,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0004999846872338467,
+      "loss": 4.3658,
+      "step": 557
+    },
+    {
+      "epoch": 0.023494736842105263,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0004999843114783456,
+      "loss": 4.4909,
+      "step": 558
+    },
+    {
+      "epoch": 0.023536842105263158,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0004999839311684241,
+      "loss": 3.5408,
+      "step": 559
+    },
+    {
+      "epoch": 0.023578947368421053,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0004999835463040891,
+      "loss": 4.0307,
+      "step": 560
+    },
+    {
+      "epoch": 0.02362105263157895,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.0004999831568853476,
+      "loss": 4.1097,
+      "step": 561
+    },
+    {
+      "epoch": 0.02366315789473684,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0004999827629122067,
+      "loss": 4.5409,
+      "step": 562
+    },
+    {
+      "epoch": 0.023705263157894736,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0004999823643846736,
+      "loss": 4.5116,
+      "step": 563
+    },
+    {
+      "epoch": 0.02374736842105263,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0004999819613027555,
+      "loss": 4.239,
+      "step": 564
+    },
+    {
+      "epoch": 0.023789473684210527,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0004999815536664598,
+      "loss": 4.1938,
+      "step": 565
+    },
+    {
+      "epoch": 0.023831578947368422,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0004999811414757939,
+      "loss": 3.6577,
+      "step": 566
+    },
+    {
+      "epoch": 0.023873684210526314,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0004999807247307653,
+      "loss": 4.2733,
+      "step": 567
+    },
+    {
+      "epoch": 0.02391578947368421,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0004999803034313817,
+      "loss": 3.7278,
+      "step": 568
+    },
+    {
+      "epoch": 0.023957894736842105,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004999798775776505,
+      "loss": 4.1731,
+      "step": 569
+    },
+    {
+      "epoch": 0.024,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0004999794471695798,
+      "loss": 4.4026,
+      "step": 570
+    },
+    {
+      "epoch": 0.024042105263157896,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0004999790122071773,
+      "loss": 4.2953,
+      "step": 571
+    },
+    {
+      "epoch": 0.024084210526315788,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0004999785726904508,
+      "loss": 4.3452,
+      "step": 572
+    },
+    {
+      "epoch": 0.024126315789473683,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0004999781286194085,
+      "loss": 3.5627,
+      "step": 573
+    },
+    {
+      "epoch": 0.02416842105263158,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0004999776799940584,
+      "loss": 4.3395,
+      "step": 574
+    },
+    {
+      "epoch": 0.024210526315789474,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0004999772268144087,
+      "loss": 4.6307,
+      "step": 575
+    },
+    {
+      "epoch": 0.02425263157894737,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0004999767690804676,
+      "loss": 4.0194,
+      "step": 576
+    },
+    {
+      "epoch": 0.024294736842105265,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.0004999763067922435,
+      "loss": 3.7996,
+      "step": 577
+    },
+    {
+      "epoch": 0.024336842105263157,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0004999758399497447,
+      "loss": 4.2088,
+      "step": 578
+    },
+    {
+      "epoch": 0.024378947368421052,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00049997536855298,
+      "loss": 4.4423,
+      "step": 579
+    },
+    {
+      "epoch": 0.024421052631578948,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0004999748926019577,
+      "loss": 4.206,
+      "step": 580
+    },
+    {
+      "epoch": 0.024463157894736843,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.0004999744120966865,
+      "loss": 4.1363,
+      "step": 581
+    },
+    {
+      "epoch": 0.02450526315789474,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0004999739270371752,
+      "loss": 4.3203,
+      "step": 582
+    },
+    {
+      "epoch": 0.02454736842105263,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0004999734374234328,
+      "loss": 4.4022,
+      "step": 583
+    },
+    {
+      "epoch": 0.024589473684210526,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.000499972943255468,
+      "loss": 4.241,
+      "step": 584
+    },
+    {
+      "epoch": 0.02463157894736842,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0004999724445332898,
+      "loss": 3.9091,
+      "step": 585
+    },
+    {
+      "epoch": 0.024673684210526316,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0004999719412569075,
+      "loss": 4.6019,
+      "step": 586
+    },
+    {
+      "epoch": 0.024715789473684212,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0004999714334263301,
+      "loss": 3.8187,
+      "step": 587
+    },
+    {
+      "epoch": 0.024757894736842104,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0004999709210415669,
+      "loss": 4.2854,
+      "step": 588
+    },
+    {
+      "epoch": 0.0248,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0004999704041026272,
+      "loss": 4.0036,
+      "step": 589
+    },
+    {
+      "epoch": 0.024842105263157895,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0004999698826095204,
+      "loss": 3.6707,
+      "step": 590
+    },
+    {
+      "epoch": 0.02488421052631579,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.000499969356562256,
+      "loss": 4.102,
+      "step": 591
+    },
+    {
+      "epoch": 0.024926315789473685,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0004999688259608438,
+      "loss": 4.2249,
+      "step": 592
+    },
+    {
+      "epoch": 0.024968421052631577,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0004999682908052932,
+      "loss": 4.2307,
+      "step": 593
+    },
+    {
+      "epoch": 0.025010526315789473,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.000499967751095614,
+      "loss": 4.0243,
+      "step": 594
+    },
+    {
+      "epoch": 0.025052631578947368,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0004999672068318161,
+      "loss": 3.7756,
+      "step": 595
+    },
+    {
+      "epoch": 0.025094736842105263,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0004999666580139094,
+      "loss": 4.2975,
+      "step": 596
+    },
+    {
+      "epoch": 0.02513684210526316,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.000499966104641904,
+      "loss": 4.1384,
+      "step": 597
+    },
+    {
+      "epoch": 0.025178947368421054,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0004999655467158097,
+      "loss": 3.7663,
+      "step": 598
+    },
+    {
+      "epoch": 0.025221052631578946,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0004999649842356369,
+      "loss": 4.2195,
+      "step": 599
+    },
+    {
+      "epoch": 0.02526315789473684,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0004999644172013957,
+      "loss": 3.9898,
+      "step": 600
+    },
+    {
+      "epoch": 0.025305263157894737,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0004999638456130967,
+      "loss": 4.5602,
+      "step": 601
+    },
+    {
+      "epoch": 0.025347368421052632,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00049996326947075,
+      "loss": 3.6831,
+      "step": 602
+    },
+    {
+      "epoch": 0.025389473684210528,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0004999626887743663,
+      "loss": 4.4858,
+      "step": 603
+    },
+    {
+      "epoch": 0.02543157894736842,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.000499962103523956,
+      "loss": 4.1604,
+      "step": 604
+    },
+    {
+      "epoch": 0.025473684210526315,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00049996151371953,
+      "loss": 4.045,
+      "step": 605
+    },
+    {
+      "epoch": 0.02551578947368421,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0004999609193610988,
+      "loss": 4.3689,
+      "step": 606
+    },
+    {
+      "epoch": 0.025557894736842106,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0004999603204486734,
+      "loss": 4.0321,
+      "step": 607
+    },
+    {
+      "epoch": 0.0256,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0004999597169822646,
+      "loss": 4.116,
+      "step": 608
+    },
+    {
+      "epoch": 0.025642105263157893,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0004999591089618835,
+      "loss": 4.4186,
+      "step": 609
+    },
+    {
+      "epoch": 0.02568421052631579,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.000499958496387541,
+      "loss": 4.0515,
+      "step": 610
+    },
+    {
+      "epoch": 0.025726315789473684,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0004999578792592485,
+      "loss": 4.1381,
+      "step": 611
+    },
+    {
+      "epoch": 0.02576842105263158,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0004999572575770172,
+      "loss": 3.8973,
+      "step": 612
+    },
+    {
+      "epoch": 0.025810526315789475,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0004999566313408582,
+      "loss": 3.9779,
+      "step": 613
+    },
+    {
+      "epoch": 0.02585263157894737,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0004999560005507831,
+      "loss": 4.0894,
+      "step": 614
+    },
+    {
+      "epoch": 0.025894736842105262,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0004999553652068034,
+      "loss": 4.0895,
+      "step": 615
+    },
+    {
+      "epoch": 0.025936842105263157,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0004999547253089307,
+      "loss": 4.043,
+      "step": 616
+    },
+    {
+      "epoch": 0.025978947368421053,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004999540808571765,
+      "loss": 4.3555,
+      "step": 617
+    },
+    {
+      "epoch": 0.026021052631578948,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0004999534318515526,
+      "loss": 3.954,
+      "step": 618
+    },
+    {
+      "epoch": 0.026063157894736844,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0004999527782920709,
+      "loss": 4.2232,
+      "step": 619
+    },
+    {
+      "epoch": 0.026105263157894736,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0004999521201787433,
+      "loss": 3.8596,
+      "step": 620
+    },
+    {
+      "epoch": 0.02614736842105263,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0004999514575115817,
+      "loss": 4.6642,
+      "step": 621
+    },
+    {
+      "epoch": 0.026189473684210526,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0004999507902905982,
+      "loss": 4.1311,
+      "step": 622
+    },
+    {
+      "epoch": 0.026231578947368422,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.000499950118515805,
+      "loss": 4.2543,
+      "step": 623
+    },
+    {
+      "epoch": 0.026273684210526317,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004999494421872144,
+      "loss": 4.0853,
+      "step": 624
+    },
+    {
+      "epoch": 0.02631578947368421,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0004999487613048386,
+      "loss": 3.9806,
+      "step": 625
+    },
+    {
+      "epoch": 0.026357894736842104,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00049994807586869,
+      "loss": 3.8894,
+      "step": 626
+    },
+    {
+      "epoch": 0.0264,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0004999473858787811,
+      "loss": 3.9208,
+      "step": 627
+    },
+    {
+      "epoch": 0.026442105263157895,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0004999466913351246,
+      "loss": 4.5039,
+      "step": 628
+    },
+    {
+      "epoch": 0.02648421052631579,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0004999459922377331,
+      "loss": 4.1046,
+      "step": 629
+    },
+    {
+      "epoch": 0.026526315789473683,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0004999452885866191,
+      "loss": 3.8371,
+      "step": 630
+    },
+    {
+      "epoch": 0.026568421052631578,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0004999445803817958,
+      "loss": 3.8807,
+      "step": 631
+    },
+    {
+      "epoch": 0.026610526315789473,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004999438676232758,
+      "loss": 4.3093,
+      "step": 632
+    },
+    {
+      "epoch": 0.02665263157894737,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0004999431503110723,
+      "loss": 4.2689,
+      "step": 633
+    },
+    {
+      "epoch": 0.026694736842105264,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0004999424284451982,
+      "loss": 4.0376,
+      "step": 634
+    },
+    {
+      "epoch": 0.02673684210526316,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0004999417020256668,
+      "loss": 3.6316,
+      "step": 635
+    },
+    {
+      "epoch": 0.02677894736842105,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0004999409710524912,
+      "loss": 4.0667,
+      "step": 636
+    },
+    {
+      "epoch": 0.026821052631578947,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0004999402355256848,
+      "loss": 3.9578,
+      "step": 637
+    },
+    {
+      "epoch": 0.026863157894736842,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.000499939495445261,
+      "loss": 4.2133,
+      "step": 638
+    },
+    {
+      "epoch": 0.026905263157894738,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0004999387508112332,
+      "loss": 3.9077,
+      "step": 639
+    },
+    {
+      "epoch": 0.026947368421052633,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.000499938001623615,
+      "loss": 3.5485,
+      "step": 640
+    },
+    {
+      "epoch": 0.026989473684210525,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0004999372478824201,
+      "loss": 4.1327,
+      "step": 641
+    },
+    {
+      "epoch": 0.02703157894736842,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0004999364895876623,
+      "loss": 4.2273,
+      "step": 642
+    },
+    {
+      "epoch": 0.027073684210526316,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0004999357267393552,
+      "loss": 4.1991,
+      "step": 643
+    },
+    {
+      "epoch": 0.02711578947368421,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0004999349593375129,
+      "loss": 3.8135,
+      "step": 644
+    },
+    {
+      "epoch": 0.027157894736842107,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0004999341873821492,
+      "loss": 4.0758,
+      "step": 645
+    },
+    {
+      "epoch": 0.0272,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0004999334108732785,
+      "loss": 3.8673,
+      "step": 646
+    },
+    {
+      "epoch": 0.027242105263157894,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0004999326298109145,
+      "loss": 4.2295,
+      "step": 647
+    },
+    {
+      "epoch": 0.02728421052631579,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004999318441950716,
+      "loss": 4.1584,
+      "step": 648
+    },
+    {
+      "epoch": 0.027326315789473685,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0004999310540257643,
+      "loss": 4.0127,
+      "step": 649
+    },
+    {
+      "epoch": 0.02736842105263158,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0004999302593030068,
+      "loss": 4.3217,
+      "step": 650
+    },
+    {
+      "epoch": 0.027410526315789472,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.0004999294600268137,
+      "loss": 4.0146,
+      "step": 651
+    },
+    {
+      "epoch": 0.027452631578947367,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0004999286561971994,
+      "loss": 4.3334,
+      "step": 652
+    },
+    {
+      "epoch": 0.027494736842105263,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0004999278478141787,
+      "loss": 4.462,
+      "step": 653
+    },
+    {
+      "epoch": 0.027536842105263158,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004999270348777662,
+      "loss": 3.9484,
+      "step": 654
+    },
+    {
+      "epoch": 0.027578947368421054,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0004999262173879769,
+      "loss": 4.4601,
+      "step": 655
+    },
+    {
+      "epoch": 0.02762105263157895,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004999253953448255,
+      "loss": 4.2206,
+      "step": 656
+    },
+    {
+      "epoch": 0.02766315789473684,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0004999245687483271,
+      "loss": 4.341,
+      "step": 657
+    },
+    {
+      "epoch": 0.027705263157894736,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004999237375984965,
+      "loss": 4.1534,
+      "step": 658
+    },
+    {
+      "epoch": 0.02774736842105263,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0004999229018953493,
+      "loss": 4.3511,
+      "step": 659
+    },
+    {
+      "epoch": 0.027789473684210527,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0004999220616389003,
+      "loss": 4.3578,
+      "step": 660
+    },
+    {
+      "epoch": 0.027831578947368422,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.000499921216829165,
+      "loss": 4.2076,
+      "step": 661
+    },
+    {
+      "epoch": 0.027873684210526314,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0004999203674661589,
+      "loss": 4.3109,
+      "step": 662
+    },
+    {
+      "epoch": 0.02791578947368421,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0004999195135498973,
+      "loss": 4.3867,
+      "step": 663
+    },
+    {
+      "epoch": 0.027957894736842105,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004999186550803958,
+      "loss": 4.0917,
+      "step": 664
+    },
+    {
+      "epoch": 0.028,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00049991779205767,
+      "loss": 4.3205,
+      "step": 665
+    },
+    {
+      "epoch": 0.028042105263157896,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0004999169244817356,
+      "loss": 3.965,
+      "step": 666
+    },
+    {
+      "epoch": 0.028084210526315788,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0004999160523526086,
+      "loss": 4.1982,
+      "step": 667
+    },
+    {
+      "epoch": 0.028126315789473683,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0004999151756703048,
+      "loss": 4.1706,
+      "step": 668
+    },
+    {
+      "epoch": 0.02816842105263158,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004999142944348401,
+      "loss": 4.3561,
+      "step": 669
+    },
+    {
+      "epoch": 0.028210526315789474,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0004999134086462305,
+      "loss": 4.0577,
+      "step": 670
+    },
+    {
+      "epoch": 0.02825263157894737,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.0004999125183044923,
+      "loss": 3.8123,
+      "step": 671
+    },
+    {
+      "epoch": 0.028294736842105265,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004999116234096417,
+      "loss": 4.2543,
+      "step": 672
+    },
+    {
+      "epoch": 0.028336842105263157,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0004999107239616948,
+      "loss": 4.1865,
+      "step": 673
+    },
+    {
+      "epoch": 0.028378947368421052,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0004999098199606684,
+      "loss": 4.2058,
+      "step": 674
+    },
+    {
+      "epoch": 0.028421052631578948,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004999089114065784,
+      "loss": 3.7633,
+      "step": 675
+    },
+    {
+      "epoch": 0.028463157894736843,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0004999079982994419,
+      "loss": 4.1803,
+      "step": 676
+    },
+    {
+      "epoch": 0.02850526315789474,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004999070806392753,
+      "loss": 4.0662,
+      "step": 677
+    },
+    {
+      "epoch": 0.02854736842105263,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0004999061584260952,
+      "loss": 3.5857,
+      "step": 678
+    },
+    {
+      "epoch": 0.028589473684210526,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0004999052316599185,
+      "loss": 3.8849,
+      "step": 679
+    },
+    {
+      "epoch": 0.02863157894736842,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0004999043003407623,
+      "loss": 3.686,
+      "step": 680
+    },
+    {
+      "epoch": 0.028673684210526316,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0004999033644686432,
+      "loss": 3.7717,
+      "step": 681
+    },
+    {
+      "epoch": 0.028715789473684212,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0004999024240435784,
+      "loss": 3.8231,
+      "step": 682
+    },
+    {
+      "epoch": 0.028757894736842104,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004999014790655851,
+      "loss": 4.4855,
+      "step": 683
+    },
+    {
+      "epoch": 0.0288,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0004999005295346806,
+      "loss": 4.2384,
+      "step": 684
+    },
+    {
+      "epoch": 0.028842105263157895,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.000499899575450882,
+      "loss": 4.2844,
+      "step": 685
+    },
+    {
+      "epoch": 0.02888421052631579,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0004998986168142066,
+      "loss": 3.9843,
+      "step": 686
+    },
+    {
+      "epoch": 0.028926315789473685,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0004998976536246722,
+      "loss": 4.0072,
+      "step": 687
+    },
+    {
+      "epoch": 0.028968421052631577,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0004998966858822962,
+      "loss": 3.4171,
+      "step": 688
+    },
+    {
+      "epoch": 0.029010526315789473,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0004998957135870961,
+      "loss": 4.1952,
+      "step": 689
+    },
+    {
+      "epoch": 0.029052631578947368,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0004998947367390897,
+      "loss": 4.0596,
+      "step": 690
+    },
+    {
+      "epoch": 0.029094736842105264,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0004998937553382948,
+      "loss": 3.9193,
+      "step": 691
+    },
+    {
+      "epoch": 0.02913684210526316,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0004998927693847294,
+      "loss": 4.1471,
+      "step": 692
+    },
+    {
+      "epoch": 0.029178947368421054,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004998917788784113,
+      "loss": 4.4418,
+      "step": 693
+    },
+    {
+      "epoch": 0.029221052631578946,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0004998907838193586,
+      "loss": 4.2383,
+      "step": 694
+    },
+    {
+      "epoch": 0.02926315789473684,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0004998897842075894,
+      "loss": 4.1382,
+      "step": 695
+    },
+    {
+      "epoch": 0.029305263157894737,
+      "grad_norm": 0.5,
+      "learning_rate": 0.000499888780043122,
+      "loss": 4.1138,
+      "step": 696
+    },
+    {
+      "epoch": 0.029347368421052632,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0004998877713259745,
+      "loss": 4.0818,
+      "step": 697
+    },
+    {
+      "epoch": 0.029389473684210528,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004998867580561655,
+      "loss": 3.9813,
+      "step": 698
+    },
+    {
+      "epoch": 0.02943157894736842,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004998857402337133,
+      "loss": 4.1947,
+      "step": 699
+    },
+    {
+      "epoch": 0.029473684210526315,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004998847178586366,
+      "loss": 4.3387,
+      "step": 700
+    },
+    {
+      "epoch": 0.02951578947368421,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004998836909309541,
+      "loss": 3.9755,
+      "step": 701
+    },
+    {
+      "epoch": 0.029557894736842106,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0004998826594506842,
+      "loss": 4.0868,
+      "step": 702
+    },
+    {
+      "epoch": 0.0296,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0004998816234178458,
+      "loss": 4.3954,
+      "step": 703
+    },
+    {
+      "epoch": 0.029642105263157893,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004998805828324578,
+      "loss": 4.1477,
+      "step": 704
+    },
+    {
+      "epoch": 0.02968421052631579,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0004998795376945392,
+      "loss": 4.2149,
+      "step": 705
+    },
+    {
+      "epoch": 0.029726315789473684,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004998784880041091,
+      "loss": 3.9834,
+      "step": 706
+    },
+    {
+      "epoch": 0.02976842105263158,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0004998774337611864,
+      "loss": 4.2129,
+      "step": 707
+    },
+    {
+      "epoch": 0.029810526315789475,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0004998763749657906,
+      "loss": 4.0314,
+      "step": 708
+    },
+    {
+      "epoch": 0.029852631578947367,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0004998753116179408,
+      "loss": 4.1842,
+      "step": 709
+    },
+    {
+      "epoch": 0.029894736842105262,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004998742437176565,
+      "loss": 3.9284,
+      "step": 710
+    },
+    {
+      "epoch": 0.029936842105263158,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0004998731712649569,
+      "loss": 4.4043,
+      "step": 711
+    },
+    {
+      "epoch": 0.029978947368421053,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0004998720942598618,
+      "loss": 3.5528,
+      "step": 712
+    },
+    {
+      "epoch": 0.03002105263157895,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0004998710127023906,
+      "loss": 3.768,
+      "step": 713
+    },
+    {
+      "epoch": 0.030063157894736844,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0004998699265925633,
+      "loss": 3.937,
+      "step": 714
+    },
+    {
+      "epoch": 0.030105263157894736,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0004998688359303994,
+      "loss": 4.2029,
+      "step": 715
+    },
+    {
+      "epoch": 0.03014736842105263,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.000499867740715919,
+      "loss": 3.8871,
+      "step": 716
+    },
+    {
+      "epoch": 0.030189473684210526,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.0004998666409491419,
+      "loss": 4.0253,
+      "step": 717
+    },
+    {
+      "epoch": 0.030231578947368422,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004998655366300882,
+      "loss": 4.2328,
+      "step": 718
+    },
+    {
+      "epoch": 0.030273684210526317,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0004998644277587781,
+      "loss": 3.9334,
+      "step": 719
+    },
+    {
+      "epoch": 0.03031578947368421,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0004998633143352315,
+      "loss": 3.8762,
+      "step": 720
+    },
+    {
+      "epoch": 0.030357894736842105,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0004998621963594691,
+      "loss": 4.1105,
+      "step": 721
+    },
+    {
+      "epoch": 0.0304,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.0004998610738315108,
+      "loss": 3.9509,
+      "step": 722
+    },
+    {
+      "epoch": 0.030442105263157895,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0004998599467513774,
+      "loss": 4.1413,
+      "step": 723
+    },
+    {
+      "epoch": 0.03048421052631579,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004998588151190895,
+      "loss": 4.4055,
+      "step": 724
+    },
+    {
+      "epoch": 0.030526315789473683,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0004998576789346675,
+      "loss": 3.7939,
+      "step": 725
+    },
+    {
+      "epoch": 0.030568421052631578,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.000499856538198132,
+      "loss": 3.8,
+      "step": 726
+    },
+    {
+      "epoch": 0.030610526315789473,
+      "grad_norm": 0.921875,
+      "learning_rate": 0.0004998553929095041,
+      "loss": 3.8837,
+      "step": 727
+    },
+    {
+      "epoch": 0.03065263157894737,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0004998542430688045,
+      "loss": 3.6112,
+      "step": 728
+    },
+    {
+      "epoch": 0.030694736842105264,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0004998530886760542,
+      "loss": 4.1042,
+      "step": 729
+    },
+    {
+      "epoch": 0.030736842105263156,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.000499851929731274,
+      "loss": 3.7537,
+      "step": 730
+    },
+    {
+      "epoch": 0.03077894736842105,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0004998507662344854,
+      "loss": 3.767,
+      "step": 731
+    },
+    {
+      "epoch": 0.030821052631578947,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0004998495981857093,
+      "loss": 4.0155,
+      "step": 732
+    },
+    {
+      "epoch": 0.030863157894736842,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004998484255849671,
+      "loss": 4.1372,
+      "step": 733
+    },
+    {
+      "epoch": 0.030905263157894738,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0004998472484322802,
+      "loss": 3.9748,
+      "step": 734
+    },
+    {
+      "epoch": 0.030947368421052633,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0004998460667276699,
+      "loss": 3.9526,
+      "step": 735
+    },
+    {
+      "epoch": 0.030989473684210525,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0004998448804711578,
+      "loss": 3.9407,
+      "step": 736
+    },
+    {
+      "epoch": 0.03103157894736842,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004998436896627656,
+      "loss": 4.1717,
+      "step": 737
+    },
+    {
+      "epoch": 0.031073684210526316,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.000499842494302515,
+      "loss": 3.7643,
+      "step": 738
+    },
+    {
+      "epoch": 0.03111578947368421,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004998412943904276,
+      "loss": 3.8482,
+      "step": 739
+    },
+    {
+      "epoch": 0.031157894736842107,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0004998400899265254,
+      "loss": 4.2118,
+      "step": 740
+    },
+    {
+      "epoch": 0.0312,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004998388809108303,
+      "loss": 4.3336,
+      "step": 741
+    },
+    {
+      "epoch": 0.031242105263157894,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0004998376673433644,
+      "loss": 3.9516,
+      "step": 742
+    },
+    {
+      "epoch": 0.03128421052631579,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0004998364492241498,
+      "loss": 3.7602,
+      "step": 743
+    },
+    {
+      "epoch": 0.03132631578947368,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0004998352265532085,
+      "loss": 4.2425,
+      "step": 744
+    },
+    {
+      "epoch": 0.03136842105263158,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0004998339993305629,
+      "loss": 4.1772,
+      "step": 745
+    },
+    {
+      "epoch": 0.03141052631578947,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0004998327675562354,
+      "loss": 3.8244,
+      "step": 746
+    },
+    {
+      "epoch": 0.03145263157894737,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0004998315312302485,
+      "loss": 3.9586,
+      "step": 747
+    },
+    {
+      "epoch": 0.03149473684210526,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0004998302903526246,
+      "loss": 3.7922,
+      "step": 748
+    },
+    {
+      "epoch": 0.031536842105263155,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0004998290449233863,
+      "loss": 3.9835,
+      "step": 749
+    },
+    {
+      "epoch": 0.031578947368421054,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0004998277949425563,
+      "loss": 4.2068,
+      "step": 750
+    },
+    {
+      "epoch": 0.031621052631578946,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0004998265404101576,
+      "loss": 4.128,
+      "step": 751
+    },
+    {
+      "epoch": 0.031663157894736844,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0004998252813262127,
+      "loss": 4.2271,
+      "step": 752
+    },
+    {
+      "epoch": 0.031705263157894736,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0004998240176907448,
+      "loss": 4.006,
+      "step": 753
+    },
+    {
+      "epoch": 0.03174736842105263,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0004998227495037768,
+      "loss": 4.0048,
+      "step": 754
+    },
+    {
+      "epoch": 0.03178947368421053,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0004998214767653319,
+      "loss": 3.9074,
+      "step": 755
+    },
+    {
+      "epoch": 0.03183157894736842,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.000499820199475433,
+      "loss": 3.8607,
+      "step": 756
+    },
+    {
+      "epoch": 0.03187368421052632,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0004998189176341038,
+      "loss": 3.9136,
+      "step": 757
+    },
+    {
+      "epoch": 0.03191578947368421,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0004998176312413674,
+      "loss": 4.0978,
+      "step": 758
+    },
+    {
+      "epoch": 0.0319578947368421,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0004998163402972472,
+      "loss": 4.6721,
+      "step": 759
+    },
+    {
+      "epoch": 0.032,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0004998150448017669,
+      "loss": 4.2847,
+      "step": 760
+    },
+    {
+      "epoch": 0.03204210526315789,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00049981374475495,
+      "loss": 3.9226,
+      "step": 761
+    },
+    {
+      "epoch": 0.03208421052631579,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0004998124401568201,
+      "loss": 3.9979,
+      "step": 762
+    },
+    {
+      "epoch": 0.03212631578947368,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0004998111310074012,
+      "loss": 4.0338,
+      "step": 763
+    },
+    {
+      "epoch": 0.03216842105263158,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0004998098173067169,
+      "loss": 3.9018,
+      "step": 764
+    },
+    {
+      "epoch": 0.032210526315789474,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0004998084990547913,
+      "loss": 4.1493,
+      "step": 765
+    },
+    {
+      "epoch": 0.032252631578947366,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0004998071762516483,
+      "loss": 3.926,
+      "step": 766
+    },
+    {
+      "epoch": 0.032294736842105265,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004998058488973121,
+      "loss": 4.3805,
+      "step": 767
+    },
+    {
+      "epoch": 0.03233684210526316,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004998045169918069,
+      "loss": 3.7523,
+      "step": 768
+    },
+    {
+      "epoch": 0.032378947368421056,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0004998031805351569,
+      "loss": 3.596,
+      "step": 769
+    },
+    {
+      "epoch": 0.03242105263157895,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0004998018395273863,
+      "loss": 4.0849,
+      "step": 770
+    },
+    {
+      "epoch": 0.03246315789473684,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0004998004939685197,
+      "loss": 3.9699,
+      "step": 771
+    },
+    {
+      "epoch": 0.03250526315789474,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0004997991438585817,
+      "loss": 4.0008,
+      "step": 772
+    },
+    {
+      "epoch": 0.03254736842105263,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004997977891975969,
+      "loss": 4.0709,
+      "step": 773
+    },
+    {
+      "epoch": 0.03258947368421053,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0004997964299855897,
+      "loss": 4.0414,
+      "step": 774
+    },
+    {
+      "epoch": 0.03263157894736842,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004997950662225852,
+      "loss": 4.13,
+      "step": 775
+    },
+    {
+      "epoch": 0.03267368421052631,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.000499793697908608,
+      "loss": 3.997,
+      "step": 776
+    },
+    {
+      "epoch": 0.03271578947368421,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0004997923250436831,
+      "loss": 3.9943,
+      "step": 777
+    },
+    {
+      "epoch": 0.032757894736842104,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0004997909476278356,
+      "loss": 3.6848,
+      "step": 778
+    },
+    {
+      "epoch": 0.0328,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0004997895656610905,
+      "loss": 4.4947,
+      "step": 779
+    },
+    {
+      "epoch": 0.032842105263157895,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.000499788179143473,
+      "loss": 3.9596,
+      "step": 780
+    },
+    {
+      "epoch": 0.03288421052631579,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0004997867880750084,
+      "loss": 3.921,
+      "step": 781
+    },
+    {
+      "epoch": 0.032926315789473685,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0004997853924557219,
+      "loss": 4.0804,
+      "step": 782
+    },
+    {
+      "epoch": 0.03296842105263158,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0004997839922856391,
+      "loss": 3.9264,
+      "step": 783
+    },
+    {
+      "epoch": 0.033010526315789476,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0004997825875647855,
+      "loss": 3.608,
+      "step": 784
+    },
+    {
+      "epoch": 0.03305263157894737,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0004997811782931866,
+      "loss": 4.0276,
+      "step": 785
+    },
+    {
+      "epoch": 0.03309473684210526,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.000499779764470868,
+      "loss": 3.7064,
+      "step": 786
+    },
+    {
+      "epoch": 0.03313684210526316,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0004997783460978557,
+      "loss": 4.1538,
+      "step": 787
+    },
+    {
+      "epoch": 0.03317894736842105,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.0004997769231741753,
+      "loss": 4.189,
+      "step": 788
+    },
+    {
+      "epoch": 0.03322105263157895,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.000499775495699853,
+      "loss": 4.2725,
+      "step": 789
+    },
+    {
+      "epoch": 0.03326315789473684,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0004997740636749145,
+      "loss": 3.6191,
+      "step": 790
+    },
+    {
+      "epoch": 0.033305263157894734,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0004997726270993861,
+      "loss": 3.9842,
+      "step": 791
+    },
+    {
+      "epoch": 0.03334736842105263,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004997711859732939,
+      "loss": 4.1349,
+      "step": 792
+    },
+    {
+      "epoch": 0.033389473684210524,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0004997697402966642,
+      "loss": 3.8909,
+      "step": 793
+    },
+    {
+      "epoch": 0.03343157894736842,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0004997682900695232,
+      "loss": 4.1655,
+      "step": 794
+    },
+    {
+      "epoch": 0.033473684210526315,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004997668352918975,
+      "loss": 4.059,
+      "step": 795
+    },
+    {
+      "epoch": 0.03351578947368421,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0004997653759638135,
+      "loss": 3.6017,
+      "step": 796
+    },
+    {
+      "epoch": 0.033557894736842106,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0004997639120852978,
+      "loss": 3.7744,
+      "step": 797
+    },
+    {
+      "epoch": 0.0336,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004997624436563771,
+      "loss": 4.2117,
+      "step": 798
+    },
+    {
+      "epoch": 0.0336421052631579,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004997609706770781,
+      "loss": 4.047,
+      "step": 799
+    },
+    {
+      "epoch": 0.03368421052631579,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004997594931474277,
+      "loss": 4.1946,
+      "step": 800
+    },
+    {
+      "epoch": 0.03372631578947369,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0004997580110674528,
+      "loss": 3.7523,
+      "step": 801
+    },
+    {
+      "epoch": 0.03376842105263158,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0004997565244371805,
+      "loss": 4.0859,
+      "step": 802
+    },
+    {
+      "epoch": 0.03381052631578947,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.0004997550332566375,
+      "loss": 3.7302,
+      "step": 803
+    },
+    {
+      "epoch": 0.03385263157894737,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0004997535375258515,
+      "loss": 3.7161,
+      "step": 804
+    },
+    {
+      "epoch": 0.03389473684210526,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004997520372448494,
+      "loss": 4.2415,
+      "step": 805
+    },
+    {
+      "epoch": 0.03393684210526316,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004997505324136586,
+      "loss": 4.1853,
+      "step": 806
+    },
+    {
+      "epoch": 0.03397894736842105,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0004997490230323066,
+      "loss": 3.8699,
+      "step": 807
+    },
+    {
+      "epoch": 0.034021052631578945,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0004997475091008208,
+      "loss": 4.0161,
+      "step": 808
+    },
+    {
+      "epoch": 0.034063157894736844,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0004997459906192288,
+      "loss": 3.9752,
+      "step": 809
+    },
+    {
+      "epoch": 0.034105263157894736,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0004997444675875583,
+      "loss": 4.1348,
+      "step": 810
+    },
+    {
+      "epoch": 0.034147368421052635,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0004997429400058371,
+      "loss": 4.2644,
+      "step": 811
+    },
+    {
+      "epoch": 0.034189473684210527,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.000499741407874093,
+      "loss": 3.7122,
+      "step": 812
+    },
+    {
+      "epoch": 0.03423157894736842,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0004997398711923537,
+      "loss": 3.6367,
+      "step": 813
+    },
+    {
+      "epoch": 0.03427368421052632,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004997383299606476,
+      "loss": 3.6281,
+      "step": 814
+    },
+    {
+      "epoch": 0.03431578947368421,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0004997367841790024,
+      "loss": 3.7011,
+      "step": 815
+    },
+    {
+      "epoch": 0.03435789473684211,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0004997352338474465,
+      "loss": 3.7414,
+      "step": 816
+    },
+    {
+      "epoch": 0.0344,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004997336789660081,
+      "loss": 4.1041,
+      "step": 817
+    },
+    {
+      "epoch": 0.03444210526315789,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004997321195347154,
+      "loss": 3.8263,
+      "step": 818
+    },
+    {
+      "epoch": 0.03448421052631579,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0004997305555535969,
+      "loss": 4.1188,
+      "step": 819
+    },
+    {
+      "epoch": 0.03452631578947368,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004997289870226812,
+      "loss": 4.0564,
+      "step": 820
+    },
+    {
+      "epoch": 0.03456842105263158,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004997274139419967,
+      "loss": 3.723,
+      "step": 821
+    },
+    {
+      "epoch": 0.034610526315789474,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0004997258363115721,
+      "loss": 4.3223,
+      "step": 822
+    },
+    {
+      "epoch": 0.034652631578947365,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0004997242541314363,
+      "loss": 3.5808,
+      "step": 823
+    },
+    {
+      "epoch": 0.034694736842105264,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0004997226674016179,
+      "loss": 3.8489,
+      "step": 824
+    },
+    {
+      "epoch": 0.034736842105263156,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.000499721076122146,
+      "loss": 3.9126,
+      "step": 825
+    },
+    {
+      "epoch": 0.034778947368421055,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004997194802930494,
+      "loss": 4.1561,
+      "step": 826
+    },
+    {
+      "epoch": 0.03482105263157895,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004997178799143573,
+      "loss": 3.6641,
+      "step": 827
+    },
+    {
+      "epoch": 0.03486315789473684,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004997162749860989,
+      "loss": 4.2048,
+      "step": 828
+    },
+    {
+      "epoch": 0.03490526315789474,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004997146655083034,
+      "loss": 4.0615,
+      "step": 829
+    },
+    {
+      "epoch": 0.03494736842105263,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.000499713051481,
+      "loss": 3.8288,
+      "step": 830
+    },
+    {
+      "epoch": 0.03498947368421053,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0004997114329042181,
+      "loss": 4.1566,
+      "step": 831
+    },
+    {
+      "epoch": 0.03503157894736842,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0004997098097779875,
+      "loss": 4.0533,
+      "step": 832
+    },
+    {
+      "epoch": 0.03507368421052631,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0004997081821023375,
+      "loss": 4.0512,
+      "step": 833
+    },
+    {
+      "epoch": 0.03511578947368421,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004997065498772978,
+      "loss": 3.9289,
+      "step": 834
+    },
+    {
+      "epoch": 0.0351578947368421,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0004997049131028982,
+      "loss": 3.4434,
+      "step": 835
+    },
+    {
+      "epoch": 0.0352,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0004997032717791684,
+      "loss": 3.7508,
+      "step": 836
+    },
+    {
+      "epoch": 0.035242105263157894,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004997016259061384,
+      "loss": 4.1984,
+      "step": 837
+    },
+    {
+      "epoch": 0.03528421052631579,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004996999754838382,
+      "loss": 4.005,
+      "step": 838
+    },
+    {
+      "epoch": 0.035326315789473685,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0004996983205122978,
+      "loss": 4.275,
+      "step": 839
+    },
+    {
+      "epoch": 0.03536842105263158,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004996966609915474,
+      "loss": 3.8773,
+      "step": 840
+    },
+    {
+      "epoch": 0.035410526315789476,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0004996949969216173,
+      "loss": 3.9692,
+      "step": 841
+    },
+    {
+      "epoch": 0.03545263157894737,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0004996933283025376,
+      "loss": 3.4895,
+      "step": 842
+    },
+    {
+      "epoch": 0.035494736842105266,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.000499691655134339,
+      "loss": 4.0123,
+      "step": 843
+    },
+    {
+      "epoch": 0.03553684210526316,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0004996899774170516,
+      "loss": 3.8246,
+      "step": 844
+    },
+    {
+      "epoch": 0.03557894736842105,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004996882951507063,
+      "loss": 4.1701,
+      "step": 845
+    },
+    {
+      "epoch": 0.03562105263157895,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0004996866083353335,
+      "loss": 4.2301,
+      "step": 846
+    },
+    {
+      "epoch": 0.03566315789473684,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0004996849169709644,
+      "loss": 4.0677,
+      "step": 847
+    },
+    {
+      "epoch": 0.03570526315789474,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004996832210576291,
+      "loss": 4.1931,
+      "step": 848
+    },
+    {
+      "epoch": 0.03574736842105263,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0004996815205953591,
+      "loss": 4.0824,
+      "step": 849
+    },
+    {
+      "epoch": 0.035789473684210524,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0004996798155841851,
+      "loss": 4.1235,
+      "step": 850
+    },
+    {
+      "epoch": 0.03583157894736842,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004996781060241382,
+      "loss": 3.8044,
+      "step": 851
+    },
+    {
+      "epoch": 0.035873684210526315,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0004996763919152496,
+      "loss": 3.7707,
+      "step": 852
+    },
+    {
+      "epoch": 0.03591578947368421,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0004996746732575505,
+      "loss": 3.7273,
+      "step": 853
+    },
+    {
+      "epoch": 0.035957894736842105,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0004996729500510722,
+      "loss": 3.8799,
+      "step": 854
+    },
+    {
+      "epoch": 0.036,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0004996712222958462,
+      "loss": 3.8271,
+      "step": 855
+    },
+    {
+      "epoch": 0.036042105263157896,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0004996694899919036,
+      "loss": 3.9493,
+      "step": 856
+    },
+    {
+      "epoch": 0.03608421052631579,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0004996677531392765,
+      "loss": 3.37,
+      "step": 857
+    },
+    {
+      "epoch": 0.03612631578947369,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0004996660117379962,
+      "loss": 3.8475,
+      "step": 858
+    },
+    {
+      "epoch": 0.03616842105263158,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004996642657880946,
+      "loss": 3.9846,
+      "step": 859
+    },
+    {
+      "epoch": 0.03621052631578947,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0004996625152896033,
+      "loss": 4.2411,
+      "step": 860
+    },
+    {
+      "epoch": 0.03625263157894737,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0004996607602425543,
+      "loss": 3.8842,
+      "step": 861
+    },
+    {
+      "epoch": 0.03629473684210526,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0004996590006469797,
+      "loss": 3.9542,
+      "step": 862
+    },
+    {
+      "epoch": 0.03633684210526316,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004996572365029114,
+      "loss": 4.1191,
+      "step": 863
+    },
+    {
+      "epoch": 0.03637894736842105,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0004996554678103816,
+      "loss": 4.015,
+      "step": 864
+    },
+    {
+      "epoch": 0.036421052631578944,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0004996536945694225,
+      "loss": 3.9961,
+      "step": 865
+    },
+    {
+      "epoch": 0.03646315789473684,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0004996519167800664,
+      "loss": 3.6294,
+      "step": 866
+    },
+    {
+      "epoch": 0.036505263157894735,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0004996501344423456,
+      "loss": 3.467,
+      "step": 867
+    },
+    {
+      "epoch": 0.036547368421052634,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0004996483475562929,
+      "loss": 4.0293,
+      "step": 868
+    },
+    {
+      "epoch": 0.036589473684210526,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0004996465561219405,
+      "loss": 3.8073,
+      "step": 869
+    },
+    {
+      "epoch": 0.03663157894736842,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0004996447601393211,
+      "loss": 4.0301,
+      "step": 870
+    },
+    {
+      "epoch": 0.03667368421052632,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0004996429596084676,
+      "loss": 3.9327,
+      "step": 871
+    },
+    {
+      "epoch": 0.03671578947368421,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0004996411545294127,
+      "loss": 3.8282,
+      "step": 872
+    },
+    {
+      "epoch": 0.03675789473684211,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0004996393449021892,
+      "loss": 3.5362,
+      "step": 873
+    },
+    {
+      "epoch": 0.0368,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0004996375307268303,
+      "loss": 3.8343,
+      "step": 874
+    },
+    {
+      "epoch": 0.03684210526315789,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0004996357120033688,
+      "loss": 3.8469,
+      "step": 875
+    },
+    {
+      "epoch": 0.03688421052631579,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0004996338887318379,
+      "loss": 3.9204,
+      "step": 876
+    },
+    {
+      "epoch": 0.03692631578947368,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0004996320609122708,
+      "loss": 4.391,
+      "step": 877
+    },
+    {
+      "epoch": 0.03696842105263158,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0004996302285447011,
+      "loss": 4.0084,
+      "step": 878
+    },
+    {
+      "epoch": 0.03701052631578947,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0004996283916291617,
+      "loss": 4.096,
+      "step": 879
+    },
+    {
+      "epoch": 0.03705263157894737,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0004996265501656864,
+      "loss": 3.7891,
+      "step": 880
+    },
+    {
+      "epoch": 0.037094736842105264,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0004996247041543087,
+      "loss": 4.262,
+      "step": 881
+    },
+    {
+      "epoch": 0.037136842105263156,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0004996228535950621,
+      "loss": 3.758,
+      "step": 882
+    },
+    {
+      "epoch": 0.037178947368421054,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0004996209984879804,
+      "loss": 4.0539,
+      "step": 883
+    },
+    {
+      "epoch": 0.037221052631578946,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0004996191388330976,
+      "loss": 3.8809,
+      "step": 884
+    },
+    {
+      "epoch": 0.037263157894736845,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0004996172746304471,
+      "loss": 4.0158,
+      "step": 885
+    },
+    {
+      "epoch": 0.03730526315789474,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0004996154058800633,
+      "loss": 4.023,
+      "step": 886
+    },
+    {
+      "epoch": 0.03734736842105263,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00049961353258198,
+      "loss": 3.9961,
+      "step": 887
+    },
+    {
+      "epoch": 0.03738947368421053,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004996116547362316,
+      "loss": 4.0106,
+      "step": 888
+    },
+    {
+      "epoch": 0.03743157894736842,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.000499609772342852,
+      "loss": 4.1207,
+      "step": 889
+    },
+    {
+      "epoch": 0.03747368421052632,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0004996078854018756,
+      "loss": 3.699,
+      "step": 890
+    },
+    {
+      "epoch": 0.03751578947368421,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004996059939133368,
+      "loss": 4.0467,
+      "step": 891
+    },
+    {
+      "epoch": 0.0375578947368421,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0004996040978772702,
+      "loss": 3.7281,
+      "step": 892
+    },
+    {
+      "epoch": 0.0376,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00049960219729371,
+      "loss": 4.0397,
+      "step": 893
+    },
+    {
+      "epoch": 0.03764210526315789,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004996002921626912,
+      "loss": 3.7749,
+      "step": 894
+    },
+    {
+      "epoch": 0.03768421052631579,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0004995983824842483,
+      "loss": 4.0509,
+      "step": 895
+    },
+    {
+      "epoch": 0.037726315789473684,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004995964682584161,
+      "loss": 3.9461,
+      "step": 896
+    },
+    {
+      "epoch": 0.037768421052631576,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0004995945494852297,
+      "loss": 3.9287,
+      "step": 897
+    },
+    {
+      "epoch": 0.037810526315789475,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0004995926261647236,
+      "loss": 3.8354,
+      "step": 898
+    },
+    {
+      "epoch": 0.03785263157894737,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004995906982969332,
+      "loss": 4.0678,
+      "step": 899
+    },
+    {
+      "epoch": 0.037894736842105266,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004995887658818936,
+      "loss": 3.6049,
+      "step": 900
+    },
+    {
+      "epoch": 0.03793684210526316,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00049958682891964,
+      "loss": 3.9772,
+      "step": 901
+    },
+    {
+      "epoch": 0.03797894736842105,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0004995848874102075,
+      "loss": 3.5326,
+      "step": 902
+    },
+    {
+      "epoch": 0.03802105263157895,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0004995829413536316,
+      "loss": 3.9914,
+      "step": 903
+    },
+    {
+      "epoch": 0.03806315789473684,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0004995809907499478,
+      "loss": 4.2027,
+      "step": 904
+    },
+    {
+      "epoch": 0.03810526315789474,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004995790355991916,
+      "loss": 4.1437,
+      "step": 905
+    },
+    {
+      "epoch": 0.03814736842105263,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0004995770759013985,
+      "loss": 4.0409,
+      "step": 906
+    },
+    {
+      "epoch": 0.03818947368421052,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004995751116566045,
+      "loss": 4.0225,
+      "step": 907
+    },
+    {
+      "epoch": 0.03823157894736842,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0004995731428648451,
+      "loss": 3.89,
+      "step": 908
+    },
+    {
+      "epoch": 0.038273684210526314,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004995711695261563,
+      "loss": 4.1284,
+      "step": 909
+    },
+    {
+      "epoch": 0.03831578947368421,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.000499569191640574,
+      "loss": 4.3671,
+      "step": 910
+    },
+    {
+      "epoch": 0.038357894736842105,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004995672092081343,
+      "loss": 4.2369,
+      "step": 911
+    },
+    {
+      "epoch": 0.0384,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004995652222288732,
+      "loss": 4.3193,
+      "step": 912
+    },
+    {
+      "epoch": 0.038442105263157896,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0004995632307028271,
+      "loss": 3.8291,
+      "step": 913
+    },
+    {
+      "epoch": 0.03848421052631579,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004995612346300321,
+      "loss": 3.9642,
+      "step": 914
+    },
+    {
+      "epoch": 0.038526315789473686,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0004995592340105246,
+      "loss": 3.4461,
+      "step": 915
+    },
+    {
+      "epoch": 0.03856842105263158,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004995572288443411,
+      "loss": 4.0714,
+      "step": 916
+    },
+    {
+      "epoch": 0.03861052631578948,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0004995552191315181,
+      "loss": 3.7779,
+      "step": 917
+    },
+    {
+      "epoch": 0.03865263157894737,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0004995532048720923,
+      "loss": 3.724,
+      "step": 918
+    },
+    {
+      "epoch": 0.03869473684210526,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004995511860661002,
+      "loss": 4.0334,
+      "step": 919
+    },
+    {
+      "epoch": 0.03873684210526316,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0004995491627135788,
+      "loss": 4.0845,
+      "step": 920
+    },
+    {
+      "epoch": 0.03877894736842105,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0004995471348145648,
+      "loss": 3.9664,
+      "step": 921
+    },
+    {
+      "epoch": 0.03882105263157895,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0004995451023690952,
+      "loss": 3.9028,
+      "step": 922
+    },
+    {
+      "epoch": 0.03886315789473684,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0004995430653772071,
+      "loss": 3.9435,
+      "step": 923
+    },
+    {
+      "epoch": 0.038905263157894734,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004995410238389375,
+      "loss": 4.0502,
+      "step": 924
+    },
+    {
+      "epoch": 0.03894736842105263,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004995389777543237,
+      "loss": 3.8892,
+      "step": 925
+    },
+    {
+      "epoch": 0.038989473684210525,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0004995369271234029,
+      "loss": 4.3862,
+      "step": 926
+    },
+    {
+      "epoch": 0.039031578947368424,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004995348719462124,
+      "loss": 3.6519,
+      "step": 927
+    },
+    {
+      "epoch": 0.039073684210526316,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0004995328122227898,
+      "loss": 3.7276,
+      "step": 928
+    },
+    {
+      "epoch": 0.03911578947368421,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0004995307479531726,
+      "loss": 3.676,
+      "step": 929
+    },
+    {
+      "epoch": 0.03915789473684211,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0004995286791373982,
+      "loss": 3.8373,
+      "step": 930
+    },
+    {
+      "epoch": 0.0392,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0004995266057755045,
+      "loss": 4.1112,
+      "step": 931
+    },
+    {
+      "epoch": 0.0392421052631579,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0004995245278675293,
+      "loss": 3.818,
+      "step": 932
+    },
+    {
+      "epoch": 0.03928421052631579,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0004995224454135104,
+      "loss": 3.8587,
+      "step": 933
+    },
+    {
+      "epoch": 0.03932631578947368,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004995203584134857,
+      "loss": 4.1682,
+      "step": 934
+    },
+    {
+      "epoch": 0.03936842105263158,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0004995182668674932,
+      "loss": 3.7342,
+      "step": 935
+    },
+    {
+      "epoch": 0.03941052631578947,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0004995161707755711,
+      "loss": 3.6353,
+      "step": 936
+    },
+    {
+      "epoch": 0.03945263157894737,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0004995140701377574,
+      "loss": 3.5279,
+      "step": 937
+    },
+    {
+      "epoch": 0.03949473684210526,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0004995119649540907,
+      "loss": 3.6527,
+      "step": 938
+    },
+    {
+      "epoch": 0.039536842105263155,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.000499509855224609,
+      "loss": 3.8956,
+      "step": 939
+    },
+    {
+      "epoch": 0.039578947368421054,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004995077409493511,
+      "loss": 4.0267,
+      "step": 940
+    },
+    {
+      "epoch": 0.039621052631578946,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0004995056221283551,
+      "loss": 3.7905,
+      "step": 941
+    },
+    {
+      "epoch": 0.039663157894736845,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0004995034987616599,
+      "loss": 3.886,
+      "step": 942
+    },
+    {
+      "epoch": 0.03970526315789474,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0004995013708493041,
+      "loss": 4.0504,
+      "step": 943
+    },
+    {
+      "epoch": 0.03974736842105263,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0004994992383913266,
+      "loss": 3.7585,
+      "step": 944
+    },
+    {
+      "epoch": 0.03978947368421053,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.000499497101387766,
+      "loss": 3.9558,
+      "step": 945
+    },
+    {
+      "epoch": 0.03983157894736842,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0004994949598386614,
+      "loss": 3.8778,
+      "step": 946
+    },
+    {
+      "epoch": 0.03987368421052632,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004994928137440518,
+      "loss": 3.9991,
+      "step": 947
+    },
+    {
+      "epoch": 0.03991578947368421,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004994906631039761,
+      "loss": 3.9531,
+      "step": 948
+    },
+    {
+      "epoch": 0.0399578947368421,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0004994885079184739,
+      "loss": 3.9253,
+      "step": 949
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004994863481875841,
+      "loss": 3.8145,
+      "step": 950
+    },
+    {
+      "epoch": 0.04004210526315789,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0004994841839113462,
+      "loss": 3.9797,
+      "step": 951
+    },
+    {
+      "epoch": 0.04008421052631579,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0004994820150897996,
+      "loss": 3.8238,
+      "step": 952
+    },
+    {
+      "epoch": 0.040126315789473684,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0004994798417229838,
+      "loss": 3.6275,
+      "step": 953
+    },
+    {
+      "epoch": 0.04016842105263158,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0004994776638109384,
+      "loss": 3.9721,
+      "step": 954
+    },
+    {
+      "epoch": 0.040210526315789474,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0004994754813537031,
+      "loss": 3.9007,
+      "step": 955
+    },
+    {
+      "epoch": 0.040252631578947366,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0004994732943513177,
+      "loss": 3.5901,
+      "step": 956
+    },
+    {
+      "epoch": 0.040294736842105265,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0004994711028038219,
+      "loss": 4.1228,
+      "step": 957
+    },
+    {
+      "epoch": 0.04033684210526316,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0004994689067112558,
+      "loss": 3.806,
+      "step": 958
+    },
+    {
+      "epoch": 0.040378947368421056,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0004994667060736593,
+      "loss": 3.5774,
+      "step": 959
+    },
+    {
+      "epoch": 0.04042105263157895,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0004994645008910726,
+      "loss": 3.1609,
+      "step": 960
+    },
+    {
+      "epoch": 0.04046315789473684,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004994622911635357,
+      "loss": 4.4032,
+      "step": 961
+    },
+    {
+      "epoch": 0.04050526315789474,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.000499460076891089,
+      "loss": 3.5839,
+      "step": 962
+    },
+    {
+      "epoch": 0.04054736842105263,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0004994578580737728,
+      "loss": 3.6538,
+      "step": 963
+    },
+    {
+      "epoch": 0.04058947368421053,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0004994556347116274,
+      "loss": 4.1319,
+      "step": 964
+    },
+    {
+      "epoch": 0.04063157894736842,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0004994534068046936,
+      "loss": 3.8493,
+      "step": 965
+    },
+    {
+      "epoch": 0.04067368421052631,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0004994511743530119,
+      "loss": 3.775,
+      "step": 966
+    },
+    {
+      "epoch": 0.04071578947368421,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0004994489373566228,
+      "loss": 3.8811,
+      "step": 967
+    },
+    {
+      "epoch": 0.040757894736842104,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.000499446695815567,
+      "loss": 4.1439,
+      "step": 968
+    },
+    {
+      "epoch": 0.0408,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0004994444497298856,
+      "loss": 3.7446,
+      "step": 969
+    },
+    {
+      "epoch": 0.040842105263157895,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004994421990996195,
+      "loss": 3.8694,
+      "step": 970
+    },
+    {
+      "epoch": 0.04088421052631579,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004994399439248096,
+      "loss": 3.8993,
+      "step": 971
+    },
+    {
+      "epoch": 0.040926315789473686,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0004994376842054969,
+      "loss": 3.8486,
+      "step": 972
+    },
+    {
+      "epoch": 0.04096842105263158,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0004994354199417227,
+      "loss": 4.4611,
+      "step": 973
+    },
+    {
+      "epoch": 0.041010526315789476,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004994331511335283,
+      "loss": 4.0182,
+      "step": 974
+    },
+    {
+      "epoch": 0.04105263157894737,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0004994308777809548,
+      "loss": 4.131,
+      "step": 975
+    },
+    {
+      "epoch": 0.04109473684210526,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.000499428599884044,
+      "loss": 3.8359,
+      "step": 976
+    },
+    {
+      "epoch": 0.04113684210526316,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.000499426317442837,
+      "loss": 3.5298,
+      "step": 977
+    },
+    {
+      "epoch": 0.04117894736842105,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004994240304573756,
+      "loss": 4.0759,
+      "step": 978
+    },
+    {
+      "epoch": 0.04122105263157895,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0004994217389277015,
+      "loss": 3.7825,
+      "step": 979
+    },
+    {
+      "epoch": 0.04126315789473684,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0004994194428538563,
+      "loss": 3.7227,
+      "step": 980
+    },
+    {
+      "epoch": 0.041305263157894734,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0004994171422358819,
+      "loss": 4.0787,
+      "step": 981
+    },
+    {
+      "epoch": 0.04134736842105263,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0004994148370738203,
+      "loss": 3.7981,
+      "step": 982
+    },
+    {
+      "epoch": 0.041389473684210525,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004994125273677134,
+      "loss": 3.5939,
+      "step": 983
+    },
+    {
+      "epoch": 0.04143157894736842,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004994102131176033,
+      "loss": 4.1439,
+      "step": 984
+    },
+    {
+      "epoch": 0.041473684210526315,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004994078943235321,
+      "loss": 4.0923,
+      "step": 985
+    },
+    {
+      "epoch": 0.04151578947368421,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.000499405570985542,
+      "loss": 3.883,
+      "step": 986
+    },
+    {
+      "epoch": 0.041557894736842106,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0004994032431036756,
+      "loss": 3.8801,
+      "step": 987
+    },
+    {
+      "epoch": 0.0416,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0004994009106779752,
+      "loss": 3.7071,
+      "step": 988
+    },
+    {
+      "epoch": 0.0416421052631579,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.000499398573708483,
+      "loss": 4.0476,
+      "step": 989
+    },
+    {
+      "epoch": 0.04168421052631579,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0004993962321952421,
+      "loss": 3.3707,
+      "step": 990
+    },
+    {
+      "epoch": 0.04172631578947368,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0004993938861382947,
+      "loss": 3.8226,
+      "step": 991
+    },
+    {
+      "epoch": 0.04176842105263158,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004993915355376838,
+      "loss": 3.9579,
+      "step": 992
+    },
+    {
+      "epoch": 0.04181052631578947,
+      "grad_norm": 0.875,
+      "learning_rate": 0.0004993891803934522,
+      "loss": 3.5042,
+      "step": 993
+    },
+    {
+      "epoch": 0.04185263157894737,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004993868207056427,
+      "loss": 4.0232,
+      "step": 994
+    },
+    {
+      "epoch": 0.04189473684210526,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0004993844564742982,
+      "loss": 4.1442,
+      "step": 995
+    },
+    {
+      "epoch": 0.04193684210526316,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.000499382087699462,
+      "loss": 3.9318,
+      "step": 996
+    },
+    {
+      "epoch": 0.04197894736842105,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0004993797143811773,
+      "loss": 3.8454,
+      "step": 997
+    },
+    {
+      "epoch": 0.042021052631578945,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004993773365194871,
+      "loss": 3.6206,
+      "step": 998
+    },
+    {
+      "epoch": 0.042063157894736844,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004993749541144349,
+      "loss": 3.8465,
+      "step": 999
+    },
+    {
+      "epoch": 0.042105263157894736,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.000499372567166064,
+      "loss": 3.843,
+      "step": 1000
+    },
+    {
+      "epoch": 0.042147368421052635,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0004993701756744179,
+      "loss": 3.5974,
+      "step": 1001
+    },
+    {
+      "epoch": 0.04218947368421053,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004993677796395401,
+      "loss": 3.8475,
+      "step": 1002
+    },
+    {
+      "epoch": 0.04223157894736842,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0004993653790614746,
+      "loss": 3.6086,
+      "step": 1003
+    },
+    {
+      "epoch": 0.04227368421052632,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004993629739402647,
+      "loss": 4.0577,
+      "step": 1004
+    },
+    {
+      "epoch": 0.04231578947368421,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0004993605642759545,
+      "loss": 3.6497,
+      "step": 1005
+    },
+    {
+      "epoch": 0.04235789473684211,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004993581500685877,
+      "loss": 3.6451,
+      "step": 1006
+    },
+    {
+      "epoch": 0.0424,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0004993557313182085,
+      "loss": 3.9034,
+      "step": 1007
+    },
+    {
+      "epoch": 0.04244210526315789,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004993533080248608,
+      "loss": 3.9603,
+      "step": 1008
+    },
+    {
+      "epoch": 0.04248421052631579,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004993508801885889,
+      "loss": 3.7417,
+      "step": 1009
+    },
+    {
+      "epoch": 0.04252631578947368,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0004993484478094368,
+      "loss": 3.6999,
+      "step": 1010
+    },
+    {
+      "epoch": 0.04256842105263158,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.000499346010887449,
+      "loss": 3.9864,
+      "step": 1011
+    },
+    {
+      "epoch": 0.042610526315789474,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004993435694226699,
+      "loss": 3.2921,
+      "step": 1012
+    },
+    {
+      "epoch": 0.042652631578947366,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0004993411234151439,
+      "loss": 3.9383,
+      "step": 1013
+    },
+    {
+      "epoch": 0.042694736842105264,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0004993386728649155,
+      "loss": 3.7808,
+      "step": 1014
+    },
+    {
+      "epoch": 0.042736842105263156,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0004993362177720295,
+      "loss": 3.4354,
+      "step": 1015
+    },
+    {
+      "epoch": 0.042778947368421055,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004993337581365307,
+      "loss": 3.8211,
+      "step": 1016
+    },
+    {
+      "epoch": 0.04282105263157895,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0004993312939584637,
+      "loss": 3.875,
+      "step": 1017
+    },
+    {
+      "epoch": 0.04286315789473684,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0004993288252378734,
+      "loss": 3.6614,
+      "step": 1018
+    },
+    {
+      "epoch": 0.04290526315789474,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0004993263519748049,
+      "loss": 3.3945,
+      "step": 1019
+    },
+    {
+      "epoch": 0.04294736842105263,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004993238741693032,
+      "loss": 3.7737,
+      "step": 1020
+    },
+    {
+      "epoch": 0.04298947368421053,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004993213918214136,
+      "loss": 3.8601,
+      "step": 1021
+    },
+    {
+      "epoch": 0.04303157894736842,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.000499318904931181,
+      "loss": 3.7544,
+      "step": 1022
+    },
+    {
+      "epoch": 0.04307368421052631,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.000499316413498651,
+      "loss": 3.718,
+      "step": 1023
+    },
+    {
+      "epoch": 0.04311578947368421,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.0004993139175238688,
+      "loss": 4.5859,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0431578947368421,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00049931141700688,
+      "loss": 3.5511,
+      "step": 1025
+    },
+    {
+      "epoch": 0.0432,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0004993089119477301,
+      "loss": 4.1038,
+      "step": 1026
+    },
+    {
+      "epoch": 0.043242105263157894,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004993064023464648,
+      "loss": 3.5809,
+      "step": 1027
+    },
+    {
+      "epoch": 0.043284210526315786,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004993038882031296,
+      "loss": 4.0216,
+      "step": 1028
+    },
+    {
+      "epoch": 0.043326315789473685,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0004993013695177706,
+      "loss": 3.6265,
+      "step": 1029
+    },
+    {
+      "epoch": 0.04336842105263158,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0004992988462904337,
+      "loss": 3.9212,
+      "step": 1030
+    },
+    {
+      "epoch": 0.043410526315789476,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004992963185211645,
+      "loss": 3.6245,
+      "step": 1031
+    },
+    {
+      "epoch": 0.04345263157894737,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004992937862100095,
+      "loss": 3.7956,
+      "step": 1032
+    },
+    {
+      "epoch": 0.04349473684210527,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004992912493570144,
+      "loss": 3.4924,
+      "step": 1033
+    },
+    {
+      "epoch": 0.04353684210526316,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004992887079622257,
+      "loss": 4.0111,
+      "step": 1034
+    },
+    {
+      "epoch": 0.04357894736842105,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0004992861620256898,
+      "loss": 3.6975,
+      "step": 1035
+    },
+    {
+      "epoch": 0.04362105263157895,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0004992836115474528,
+      "loss": 3.6044,
+      "step": 1036
+    },
+    {
+      "epoch": 0.04366315789473684,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004992810565275612,
+      "loss": 3.8523,
+      "step": 1037
+    },
+    {
+      "epoch": 0.04370526315789474,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0004992784969660619,
+      "loss": 3.8583,
+      "step": 1038
+    },
+    {
+      "epoch": 0.04374736842105263,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0004992759328630011,
+      "loss": 3.5004,
+      "step": 1039
+    },
+    {
+      "epoch": 0.043789473684210524,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0004992733642184258,
+      "loss": 3.7063,
+      "step": 1040
+    },
+    {
+      "epoch": 0.04383157894736842,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0004992707910323825,
+      "loss": 3.5979,
+      "step": 1041
+    },
+    {
+      "epoch": 0.043873684210526315,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0004992682133049184,
+      "loss": 3.3756,
+      "step": 1042
+    },
+    {
+      "epoch": 0.043915789473684214,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0004992656310360803,
+      "loss": 4.0856,
+      "step": 1043
+    },
+    {
+      "epoch": 0.043957894736842106,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004992630442259153,
+      "loss": 3.8896,
+      "step": 1044
+    },
+    {
+      "epoch": 0.044,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0004992604528744705,
+      "loss": 4.0158,
+      "step": 1045
+    },
+    {
+      "epoch": 0.044042105263157896,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0004992578569817931,
+      "loss": 4.2554,
+      "step": 1046
+    },
+    {
+      "epoch": 0.04408421052631579,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0004992552565479305,
+      "loss": 3.8884,
+      "step": 1047
+    },
+    {
+      "epoch": 0.04412631578947369,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0004992526515729299,
+      "loss": 3.7229,
+      "step": 1048
+    },
+    {
+      "epoch": 0.04416842105263158,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0004992500420568388,
+      "loss": 4.0046,
+      "step": 1049
+    },
+    {
+      "epoch": 0.04421052631578947,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0004992474279997049,
+      "loss": 3.837,
+      "step": 1050
+    },
+    {
+      "epoch": 0.04425263157894737,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0004992448094015756,
+      "loss": 3.8705,
+      "step": 1051
+    },
+    {
+      "epoch": 0.04429473684210526,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0004992421862624988,
+      "loss": 3.7249,
+      "step": 1052
+    },
+    {
+      "epoch": 0.04433684210526316,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004992395585825222,
+      "loss": 3.8203,
+      "step": 1053
+    },
+    {
+      "epoch": 0.04437894736842105,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0004992369263616937,
+      "loss": 3.7441,
+      "step": 1054
+    },
+    {
+      "epoch": 0.044421052631578944,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0004992342896000613,
+      "loss": 3.5447,
+      "step": 1055
+    },
+    {
+      "epoch": 0.04446315789473684,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004992316482976729,
+      "loss": 4.2132,
+      "step": 1056
+    },
+    {
+      "epoch": 0.044505263157894735,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0004992290024545767,
+      "loss": 3.6485,
+      "step": 1057
+    },
+    {
+      "epoch": 0.044547368421052634,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004992263520708209,
+      "loss": 4.137,
+      "step": 1058
+    },
+    {
+      "epoch": 0.044589473684210526,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004992236971464539,
+      "loss": 4.0215,
+      "step": 1059
+    },
+    {
+      "epoch": 0.04463157894736842,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0004992210376815238,
+      "loss": 4.052,
+      "step": 1060
+    },
+    {
+      "epoch": 0.04467368421052632,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0004992183736760793,
+      "loss": 3.8045,
+      "step": 1061
+    },
+    {
+      "epoch": 0.04471578947368421,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004992157051301689,
+      "loss": 3.7097,
+      "step": 1062
+    },
+    {
+      "epoch": 0.04475789473684211,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004992130320438411,
+      "loss": 4.0733,
+      "step": 1063
+    },
+    {
+      "epoch": 0.0448,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0004992103544171447,
+      "loss": 3.9754,
+      "step": 1064
+    },
+    {
+      "epoch": 0.04484210526315789,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0004992076722501284,
+      "loss": 3.7527,
+      "step": 1065
+    },
+    {
+      "epoch": 0.04488421052631579,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0004992049855428412,
+      "loss": 3.8109,
+      "step": 1066
+    },
+    {
+      "epoch": 0.04492631578947368,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004992022942953319,
+      "loss": 3.4605,
+      "step": 1067
+    },
+    {
+      "epoch": 0.04496842105263158,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0004991995985076497,
+      "loss": 3.5055,
+      "step": 1068
+    },
+    {
+      "epoch": 0.04501052631578947,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004991968981798435,
+      "loss": 3.9842,
+      "step": 1069
+    },
+    {
+      "epoch": 0.045052631578947365,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004991941933119626,
+      "loss": 3.8903,
+      "step": 1070
+    },
+    {
+      "epoch": 0.045094736842105264,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004991914839040563,
+      "loss": 3.7382,
+      "step": 1071
+    },
+    {
+      "epoch": 0.045136842105263156,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.000499188769956174,
+      "loss": 3.5897,
+      "step": 1072
+    },
+    {
+      "epoch": 0.045178947368421055,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004991860514683651,
+      "loss": 3.9113,
+      "step": 1073
+    },
+    {
+      "epoch": 0.04522105263157895,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004991833284406791,
+      "loss": 3.7818,
+      "step": 1074
+    },
+    {
+      "epoch": 0.045263157894736845,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004991806008731657,
+      "loss": 4.0869,
+      "step": 1075
+    },
+    {
+      "epoch": 0.04530526315789474,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0004991778687658744,
+      "loss": 4.0519,
+      "step": 1076
+    },
+    {
+      "epoch": 0.04534736842105263,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0004991751321188553,
+      "loss": 3.8419,
+      "step": 1077
+    },
+    {
+      "epoch": 0.04538947368421053,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0004991723909321578,
+      "loss": 3.6765,
+      "step": 1078
+    },
+    {
+      "epoch": 0.04543157894736842,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004991696452058323,
+      "loss": 4.2501,
+      "step": 1079
+    },
+    {
+      "epoch": 0.04547368421052632,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004991668949399286,
+      "loss": 3.5953,
+      "step": 1080
+    },
+    {
+      "epoch": 0.04551578947368421,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004991641401344967,
+      "loss": 3.5416,
+      "step": 1081
+    },
+    {
+      "epoch": 0.0455578947368421,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004991613807895871,
+      "loss": 4.0286,
+      "step": 1082
+    },
+    {
+      "epoch": 0.0456,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004991586169052497,
+      "loss": 3.7822,
+      "step": 1083
+    },
+    {
+      "epoch": 0.045642105263157894,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0004991558484815353,
+      "loss": 3.708,
+      "step": 1084
+    },
+    {
+      "epoch": 0.04568421052631579,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0004991530755184939,
+      "loss": 3.5665,
+      "step": 1085
+    },
+    {
+      "epoch": 0.045726315789473684,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004991502980161763,
+      "loss": 3.9086,
+      "step": 1086
+    },
+    {
+      "epoch": 0.045768421052631576,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0004991475159746329,
+      "loss": 3.6985,
+      "step": 1087
+    },
+    {
+      "epoch": 0.045810526315789475,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004991447293939146,
+      "loss": 3.6663,
+      "step": 1088
+    },
+    {
+      "epoch": 0.04585263157894737,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004991419382740721,
+      "loss": 3.7988,
+      "step": 1089
+    },
+    {
+      "epoch": 0.045894736842105266,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.000499139142615156,
+      "loss": 3.5434,
+      "step": 1090
+    },
+    {
+      "epoch": 0.04593684210526316,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0004991363424172176,
+      "loss": 3.7934,
+      "step": 1091
+    },
+    {
+      "epoch": 0.04597894736842105,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0004991335376803077,
+      "loss": 3.4404,
+      "step": 1092
+    },
+    {
+      "epoch": 0.04602105263157895,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004991307284044775,
+      "loss": 3.4192,
+      "step": 1093
+    },
+    {
+      "epoch": 0.04606315789473684,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004991279145897783,
+      "loss": 3.7616,
+      "step": 1094
+    },
+    {
+      "epoch": 0.04610526315789474,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.000499125096236261,
+      "loss": 4.0635,
+      "step": 1095
+    },
+    {
+      "epoch": 0.04614736842105263,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004991222733439773,
+      "loss": 3.7645,
+      "step": 1096
+    },
+    {
+      "epoch": 0.04618947368421052,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004991194459129784,
+      "loss": 3.8296,
+      "step": 1097
+    },
+    {
+      "epoch": 0.04623157894736842,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.000499116613943316,
+      "loss": 3.7204,
+      "step": 1098
+    },
+    {
+      "epoch": 0.046273684210526314,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0004991137774350414,
+      "loss": 3.5295,
+      "step": 1099
+    },
+    {
+      "epoch": 0.04631578947368421,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004991109363882065,
+      "loss": 3.9595,
+      "step": 1100
+    },
+    {
+      "epoch": 0.046357894736842105,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004991080908028632,
+      "loss": 4.3297,
+      "step": 1101
+    },
+    {
+      "epoch": 0.0464,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004991052406790631,
+      "loss": 3.8261,
+      "step": 1102
+    },
+    {
+      "epoch": 0.046442105263157896,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004991023860168582,
+      "loss": 3.8489,
+      "step": 1103
+    },
+    {
+      "epoch": 0.04648421052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004990995268163005,
+      "loss": 4.0968,
+      "step": 1104
+    },
+    {
+      "epoch": 0.046526315789473686,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0004990966630774421,
+      "loss": 4.3113,
+      "step": 1105
+    },
+    {
+      "epoch": 0.04656842105263158,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004990937948003351,
+      "loss": 4.0536,
+      "step": 1106
+    },
+    {
+      "epoch": 0.04661052631578947,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.0004990909219850319,
+      "loss": 3.5195,
+      "step": 1107
+    },
+    {
+      "epoch": 0.04665263157894737,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0004990880446315848,
+      "loss": 3.1762,
+      "step": 1108
+    },
+    {
+      "epoch": 0.04669473684210526,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0004990851627400462,
+      "loss": 3.4462,
+      "step": 1109
+    },
+    {
+      "epoch": 0.04673684210526316,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004990822763104685,
+      "loss": 3.8925,
+      "step": 1110
+    },
+    {
+      "epoch": 0.04677894736842105,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004990793853429045,
+      "loss": 3.8434,
+      "step": 1111
+    },
+    {
+      "epoch": 0.04682105263157895,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0004990764898374067,
+      "loss": 4.2752,
+      "step": 1112
+    },
+    {
+      "epoch": 0.04686315789473684,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0004990735897940279,
+      "loss": 3.9717,
+      "step": 1113
+    },
+    {
+      "epoch": 0.046905263157894735,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0004990706852128211,
+      "loss": 3.6493,
+      "step": 1114
+    },
+    {
+      "epoch": 0.04694736842105263,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0004990677760938388,
+      "loss": 3.9603,
+      "step": 1115
+    },
+    {
+      "epoch": 0.046989473684210525,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0004990648624371344,
+      "loss": 3.6008,
+      "step": 1116
+    },
+    {
+      "epoch": 0.047031578947368424,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004990619442427609,
+      "loss": 3.6804,
+      "step": 1117
+    },
+    {
+      "epoch": 0.047073684210526316,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004990590215107713,
+      "loss": 3.8254,
+      "step": 1118
+    },
+    {
+      "epoch": 0.04711578947368421,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.000499056094241219,
+      "loss": 3.6268,
+      "step": 1119
+    },
+    {
+      "epoch": 0.04715789473684211,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0004990531624341572,
+      "loss": 3.6993,
+      "step": 1120
+    },
+    {
+      "epoch": 0.0472,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0004990502260896395,
+      "loss": 3.9004,
+      "step": 1121
+    },
+    {
+      "epoch": 0.0472421052631579,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0004990472852077194,
+      "loss": 4.029,
+      "step": 1122
+    },
+    {
+      "epoch": 0.04728421052631579,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004990443397884502,
+      "loss": 4.139,
+      "step": 1123
+    },
+    {
+      "epoch": 0.04732631578947368,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0004990413898318858,
+      "loss": 3.8081,
+      "step": 1124
+    },
+    {
+      "epoch": 0.04736842105263158,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00049903843533808,
+      "loss": 3.8606,
+      "step": 1125
+    },
+    {
+      "epoch": 0.04741052631578947,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0004990354763070864,
+      "loss": 4.0022,
+      "step": 1126
+    },
+    {
+      "epoch": 0.04745263157894737,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004990325127389591,
+      "loss": 3.8362,
+      "step": 1127
+    },
+    {
+      "epoch": 0.04749473684210526,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.000499029544633752,
+      "loss": 3.7651,
+      "step": 1128
+    },
+    {
+      "epoch": 0.047536842105263155,
+      "grad_norm": 0.875,
+      "learning_rate": 0.0004990265719915191,
+      "loss": 3.7897,
+      "step": 1129
+    },
+    {
+      "epoch": 0.047578947368421054,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0004990235948123148,
+      "loss": 3.9444,
+      "step": 1130
+    },
+    {
+      "epoch": 0.047621052631578946,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.000499020613096193,
+      "loss": 3.8025,
+      "step": 1131
+    },
+    {
+      "epoch": 0.047663157894736845,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004990176268432083,
+      "loss": 3.8231,
+      "step": 1132
+    },
+    {
+      "epoch": 0.04770526315789474,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.0004990146360534152,
+      "loss": 4.151,
+      "step": 1133
+    },
+    {
+      "epoch": 0.04774736842105263,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0004990116407268679,
+      "loss": 3.8828,
+      "step": 1134
+    },
+    {
+      "epoch": 0.04778947368421053,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.000499008640863621,
+      "loss": 4.0393,
+      "step": 1135
+    },
+    {
+      "epoch": 0.04783157894736842,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0004990056364637293,
+      "loss": 3.8835,
+      "step": 1136
+    },
+    {
+      "epoch": 0.04787368421052632,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004990026275272474,
+      "loss": 3.6874,
+      "step": 1137
+    },
+    {
+      "epoch": 0.04791578947368421,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004989996140542303,
+      "loss": 4.1264,
+      "step": 1138
+    },
+    {
+      "epoch": 0.0479578947368421,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004989965960447327,
+      "loss": 3.621,
+      "step": 1139
+    },
+    {
+      "epoch": 0.048,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0004989935734988098,
+      "loss": 4.0086,
+      "step": 1140
+    },
+    {
+      "epoch": 0.04804210526315789,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004989905464165165,
+      "loss": 4.1447,
+      "step": 1141
+    },
+    {
+      "epoch": 0.04808421052631579,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.000498987514797908,
+      "loss": 3.8843,
+      "step": 1142
+    },
+    {
+      "epoch": 0.048126315789473684,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0004989844786430394,
+      "loss": 3.6188,
+      "step": 1143
+    },
+    {
+      "epoch": 0.048168421052631576,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0004989814379519663,
+      "loss": 3.9116,
+      "step": 1144
+    },
+    {
+      "epoch": 0.048210526315789475,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004989783927247439,
+      "loss": 4.0312,
+      "step": 1145
+    },
+    {
+      "epoch": 0.048252631578947366,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0004989753429614276,
+      "loss": 3.8522,
+      "step": 1146
+    },
+    {
+      "epoch": 0.048294736842105265,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0004989722886620732,
+      "loss": 3.861,
+      "step": 1147
+    },
+    {
+      "epoch": 0.04833684210526316,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0004989692298267362,
+      "loss": 3.996,
+      "step": 1148
+    },
+    {
+      "epoch": 0.048378947368421056,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0004989661664554724,
+      "loss": 3.5301,
+      "step": 1149
+    },
+    {
+      "epoch": 0.04842105263157895,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004989630985483375,
+      "loss": 3.5198,
+      "step": 1150
+    },
+    {
+      "epoch": 0.04846315789473684,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004989600261053876,
+      "loss": 3.9163,
+      "step": 1151
+    },
+    {
+      "epoch": 0.04850526315789474,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004989569491266785,
+      "loss": 3.6483,
+      "step": 1152
+    },
+    {
+      "epoch": 0.04854736842105263,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0004989538676122662,
+      "loss": 3.655,
+      "step": 1153
+    },
+    {
+      "epoch": 0.04858947368421053,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.000498950781562207,
+      "loss": 3.3422,
+      "step": 1154
+    },
+    {
+      "epoch": 0.04863157894736842,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.000498947690976557,
+      "loss": 4.0828,
+      "step": 1155
+    },
+    {
+      "epoch": 0.04867368421052631,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0004989445958553726,
+      "loss": 4.0799,
+      "step": 1156
+    },
+    {
+      "epoch": 0.04871578947368421,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004989414961987103,
+      "loss": 4.135,
+      "step": 1157
+    },
+    {
+      "epoch": 0.048757894736842104,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0004989383920066264,
+      "loss": 3.7147,
+      "step": 1158
+    },
+    {
+      "epoch": 0.0488,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0004989352832791774,
+      "loss": 3.9812,
+      "step": 1159
+    },
+    {
+      "epoch": 0.048842105263157895,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00049893217001642,
+      "loss": 3.4013,
+      "step": 1160
+    },
+    {
+      "epoch": 0.04888421052631579,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.000498929052218411,
+      "loss": 4.0839,
+      "step": 1161
+    },
+    {
+      "epoch": 0.048926315789473686,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004989259298852071,
+      "loss": 3.9679,
+      "step": 1162
+    },
+    {
+      "epoch": 0.04896842105263158,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0004989228030168653,
+      "loss": 3.6415,
+      "step": 1163
+    },
+    {
+      "epoch": 0.04901052631578948,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0004989196716134424,
+      "loss": 3.9239,
+      "step": 1164
+    },
+    {
+      "epoch": 0.04905263157894737,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004989165356749956,
+      "loss": 3.8078,
+      "step": 1165
+    },
+    {
+      "epoch": 0.04909473684210526,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004989133952015821,
+      "loss": 4.3432,
+      "step": 1166
+    },
+    {
+      "epoch": 0.04913684210526316,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004989102501932589,
+      "loss": 4.0647,
+      "step": 1167
+    },
+    {
+      "epoch": 0.04917894736842105,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004989071006500834,
+      "loss": 3.8704,
+      "step": 1168
+    },
+    {
+      "epoch": 0.04922105263157895,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.000498903946572113,
+      "loss": 3.6771,
+      "step": 1169
+    },
+    {
+      "epoch": 0.04926315789473684,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0004989007879594051,
+      "loss": 3.5376,
+      "step": 1170
+    },
+    {
+      "epoch": 0.049305263157894734,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004988976248120173,
+      "loss": 3.8722,
+      "step": 1171
+    },
+    {
+      "epoch": 0.04934736842105263,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0004988944571300073,
+      "loss": 4.0583,
+      "step": 1172
+    },
+    {
+      "epoch": 0.049389473684210525,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004988912849134326,
+      "loss": 3.6733,
+      "step": 1173
+    },
+    {
+      "epoch": 0.049431578947368424,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0004988881081623512,
+      "loss": 4.1548,
+      "step": 1174
+    },
+    {
+      "epoch": 0.049473684210526316,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.000498884926876821,
+      "loss": 3.4042,
+      "step": 1175
+    },
+    {
+      "epoch": 0.04951578947368421,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004988817410568996,
+      "loss": 3.7907,
+      "step": 1176
+    },
+    {
+      "epoch": 0.049557894736842106,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004988785507026455,
+      "loss": 3.828,
+      "step": 1177
+    },
+    {
+      "epoch": 0.0496,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004988753558141165,
+      "loss": 3.8158,
+      "step": 1178
+    },
+    {
+      "epoch": 0.0496421052631579,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.000498872156391371,
+      "loss": 3.9778,
+      "step": 1179
+    },
+    {
+      "epoch": 0.04968421052631579,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0004988689524344673,
+      "loss": 3.4973,
+      "step": 1180
+    },
+    {
+      "epoch": 0.04972631578947368,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004988657439434637,
+      "loss": 3.7954,
+      "step": 1181
+    },
+    {
+      "epoch": 0.04976842105263158,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0004988625309184186,
+      "loss": 3.8242,
+      "step": 1182
+    },
+    {
+      "epoch": 0.04981052631578947,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0004988593133593905,
+      "loss": 3.7245,
+      "step": 1183
+    },
+    {
+      "epoch": 0.04985263157894737,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0004988560912664382,
+      "loss": 3.3523,
+      "step": 1184
+    },
+    {
+      "epoch": 0.04989473684210526,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004988528646396204,
+      "loss": 3.8366,
+      "step": 1185
+    },
+    {
+      "epoch": 0.049936842105263154,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0004988496334789956,
+      "loss": 3.8055,
+      "step": 1186
+    },
+    {
+      "epoch": 0.04997894736842105,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0004988463977846229,
+      "loss": 3.6484,
+      "step": 1187
+    },
+    {
+      "epoch": 0.050021052631578945,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0004988431575565613,
+      "loss": 3.9612,
+      "step": 1188
+    },
+    {
+      "epoch": 0.050063157894736844,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0004988399127948697,
+      "loss": 3.8424,
+      "step": 1189
+    },
+    {
+      "epoch": 0.050105263157894736,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0004988366634996074,
+      "loss": 4.1103,
+      "step": 1190
+    },
+    {
+      "epoch": 0.050147368421052635,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0004988334096708332,
+      "loss": 3.6221,
+      "step": 1191
+    },
+    {
+      "epoch": 0.05018947368421053,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0004988301513086068,
+      "loss": 4.1425,
+      "step": 1192
+    },
+    {
+      "epoch": 0.05023157894736842,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004988268884129874,
+      "loss": 3.8124,
+      "step": 1193
+    },
+    {
+      "epoch": 0.05027368421052632,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0004988236209840344,
+      "loss": 3.7767,
+      "step": 1194
+    },
+    {
+      "epoch": 0.05031578947368421,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004988203490218075,
+      "loss": 3.7257,
+      "step": 1195
+    },
+    {
+      "epoch": 0.05035789473684211,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004988170725263661,
+      "loss": 3.6231,
+      "step": 1196
+    },
+    {
+      "epoch": 0.0504,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00049881379149777,
+      "loss": 3.7598,
+      "step": 1197
+    },
+    {
+      "epoch": 0.05044210526315789,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.000498810505936079,
+      "loss": 3.8414,
+      "step": 1198
+    },
+    {
+      "epoch": 0.05048421052631579,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0004988072158413528,
+      "loss": 3.775,
+      "step": 1199
+    },
+    {
+      "epoch": 0.05052631578947368,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004988039212136516,
+      "loss": 3.6302,
+      "step": 1200
+    },
+    {
+      "epoch": 0.05056842105263158,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004988006220530353,
+      "loss": 3.7299,
+      "step": 1201
+    },
+    {
+      "epoch": 0.050610526315789474,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0004987973183595639,
+      "loss": 3.7273,
+      "step": 1202
+    },
+    {
+      "epoch": 0.050652631578947366,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0004987940101332977,
+      "loss": 3.6947,
+      "step": 1203
+    },
+    {
+      "epoch": 0.050694736842105265,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.000498790697374297,
+      "loss": 2.9553,
+      "step": 1204
+    },
+    {
+      "epoch": 0.05073684210526316,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.000498787380082622,
+      "loss": 3.6512,
+      "step": 1205
+    },
+    {
+      "epoch": 0.050778947368421055,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004987840582583334,
+      "loss": 3.831,
+      "step": 1206
+    },
+    {
+      "epoch": 0.05082105263157895,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004987807319014916,
+      "loss": 3.875,
+      "step": 1207
+    },
+    {
+      "epoch": 0.05086315789473684,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0004987774010121569,
+      "loss": 3.1948,
+      "step": 1208
+    },
+    {
+      "epoch": 0.05090526315789474,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0004987740655903905,
+      "loss": 3.8225,
+      "step": 1209
+    },
+    {
+      "epoch": 0.05094736842105263,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004987707256362529,
+      "loss": 3.5869,
+      "step": 1210
+    },
+    {
+      "epoch": 0.05098947368421053,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004987673811498049,
+      "loss": 3.9492,
+      "step": 1211
+    },
+    {
+      "epoch": 0.05103157894736842,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0004987640321311075,
+      "loss": 3.985,
+      "step": 1212
+    },
+    {
+      "epoch": 0.05107368421052631,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004987606785802217,
+      "loss": 3.6899,
+      "step": 1213
+    },
+    {
+      "epoch": 0.05111578947368421,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0004987573204972087,
+      "loss": 3.5856,
+      "step": 1214
+    },
+    {
+      "epoch": 0.051157894736842104,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004987539578821296,
+      "loss": 3.689,
+      "step": 1215
+    },
+    {
+      "epoch": 0.0512,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0004987505907350456,
+      "loss": 3.7019,
+      "step": 1216
+    },
+    {
+      "epoch": 0.051242105263157894,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0004987472190560181,
+      "loss": 3.9487,
+      "step": 1217
+    },
+    {
+      "epoch": 0.051284210526315786,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004987438428451087,
+      "loss": 3.7958,
+      "step": 1218
+    },
+    {
+      "epoch": 0.051326315789473685,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004987404621023785,
+      "loss": 4.2904,
+      "step": 1219
+    },
+    {
+      "epoch": 0.05136842105263158,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0004987370768278895,
+      "loss": 3.9304,
+      "step": 1220
+    },
+    {
+      "epoch": 0.051410526315789476,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004987336870217032,
+      "loss": 4.2174,
+      "step": 1221
+    },
+    {
+      "epoch": 0.05145263157894737,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0004987302926838812,
+      "loss": 3.825,
+      "step": 1222
+    },
+    {
+      "epoch": 0.05149473684210526,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0004987268938144858,
+      "loss": 4.1986,
+      "step": 1223
+    },
+    {
+      "epoch": 0.05153684210526316,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004987234904135784,
+      "loss": 3.8196,
+      "step": 1224
+    },
+    {
+      "epoch": 0.05157894736842105,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0004987200824812213,
+      "loss": 3.6349,
+      "step": 1225
+    },
+    {
+      "epoch": 0.05162105263157895,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0004987166700174766,
+      "loss": 3.9788,
+      "step": 1226
+    },
+    {
+      "epoch": 0.05166315789473684,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0004987132530224064,
+      "loss": 3.9148,
+      "step": 1227
+    },
+    {
+      "epoch": 0.05170526315789474,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0004987098314960729,
+      "loss": 3.9758,
+      "step": 1228
+    },
+    {
+      "epoch": 0.05174736842105263,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0004987064054385386,
+      "loss": 3.6117,
+      "step": 1229
+    },
+    {
+      "epoch": 0.051789473684210524,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004987029748498656,
+      "loss": 3.8208,
+      "step": 1230
+    },
+    {
+      "epoch": 0.05183157894736842,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004986995397301167,
+      "loss": 4.0551,
+      "step": 1231
+    },
+    {
+      "epoch": 0.051873684210526315,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0004986961000793545,
+      "loss": 3.6894,
+      "step": 1232
+    },
+    {
+      "epoch": 0.051915789473684214,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004986926558976416,
+      "loss": 3.9324,
+      "step": 1233
+    },
+    {
+      "epoch": 0.051957894736842106,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0004986892071850405,
+      "loss": 3.8361,
+      "step": 1234
+    },
+    {
+      "epoch": 0.052,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0004986857539416143,
+      "loss": 3.7379,
+      "step": 1235
+    },
+    {
+      "epoch": 0.052042105263157896,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004986822961674259,
+      "loss": 4.0313,
+      "step": 1236
+    },
+    {
+      "epoch": 0.05208421052631579,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004986788338625383,
+      "loss": 3.3483,
+      "step": 1237
+    },
+    {
+      "epoch": 0.05212631578947369,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004986753670270145,
+      "loss": 3.6726,
+      "step": 1238
+    },
+    {
+      "epoch": 0.05216842105263158,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004986718956609177,
+      "loss": 3.9699,
+      "step": 1239
+    },
+    {
+      "epoch": 0.05221052631578947,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004986684197643111,
+      "loss": 4.0079,
+      "step": 1240
+    },
+    {
+      "epoch": 0.05225263157894737,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.000498664939337258,
+      "loss": 3.9858,
+      "step": 1241
+    },
+    {
+      "epoch": 0.05229473684210526,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.000498661454379822,
+      "loss": 3.5264,
+      "step": 1242
+    },
+    {
+      "epoch": 0.05233684210526316,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004986579648920665,
+      "loss": 3.901,
+      "step": 1243
+    },
+    {
+      "epoch": 0.05237894736842105,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.000498654470874055,
+      "loss": 4.0074,
+      "step": 1244
+    },
+    {
+      "epoch": 0.052421052631578945,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004986509723258511,
+      "loss": 4.1283,
+      "step": 1245
+    },
+    {
+      "epoch": 0.052463157894736843,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0004986474692475187,
+      "loss": 3.6396,
+      "step": 1246
+    },
+    {
+      "epoch": 0.052505263157894735,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004986439616391216,
+      "loss": 3.4812,
+      "step": 1247
+    },
+    {
+      "epoch": 0.052547368421052634,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0004986404495007236,
+      "loss": 3.998,
+      "step": 1248
+    },
+    {
+      "epoch": 0.052589473684210526,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004986369328323888,
+      "loss": 3.7454,
+      "step": 1249
+    },
+    {
+      "epoch": 0.05263157894736842,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004986334116341812,
+      "loss": 3.8636,
+      "step": 1250
+    },
+    {
+      "epoch": 0.05267368421052632,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004986298859061649,
+      "loss": 3.6326,
+      "step": 1251
+    },
+    {
+      "epoch": 0.05271578947368421,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004986263556484043,
+      "loss": 3.8106,
+      "step": 1252
+    },
+    {
+      "epoch": 0.05275789473684211,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004986228208609636,
+      "loss": 4.4184,
+      "step": 1253
+    },
+    {
+      "epoch": 0.0528,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004986192815439073,
+      "loss": 3.7713,
+      "step": 1254
+    },
+    {
+      "epoch": 0.05284210526315789,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0004986157376972997,
+      "loss": 3.4042,
+      "step": 1255
+    },
+    {
+      "epoch": 0.05288421052631579,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004986121893212056,
+      "loss": 3.9335,
+      "step": 1256
+    },
+    {
+      "epoch": 0.05292631578947368,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0004986086364156893,
+      "loss": 4.0119,
+      "step": 1257
+    },
+    {
+      "epoch": 0.05296842105263158,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004986050789808159,
+      "loss": 3.9728,
+      "step": 1258
+    },
+    {
+      "epoch": 0.05301052631578947,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00049860151701665,
+      "loss": 3.9399,
+      "step": 1259
+    },
+    {
+      "epoch": 0.053052631578947365,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0004985979505232566,
+      "loss": 3.9068,
+      "step": 1260
+    },
+    {
+      "epoch": 0.053094736842105264,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0004985943795007006,
+      "loss": 3.2969,
+      "step": 1261
+    },
+    {
+      "epoch": 0.053136842105263156,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004985908039490471,
+      "loss": 3.9107,
+      "step": 1262
+    },
+    {
+      "epoch": 0.053178947368421055,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0004985872238683613,
+      "loss": 3.9077,
+      "step": 1263
+    },
+    {
+      "epoch": 0.05322105263157895,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004985836392587083,
+      "loss": 4.3125,
+      "step": 1264
+    },
+    {
+      "epoch": 0.05326315789473684,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0004985800501201535,
+      "loss": 3.6438,
+      "step": 1265
+    },
+    {
+      "epoch": 0.05330526315789474,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004985764564527621,
+      "loss": 3.9617,
+      "step": 1266
+    },
+    {
+      "epoch": 0.05334736842105263,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0004985728582565999,
+      "loss": 3.5043,
+      "step": 1267
+    },
+    {
+      "epoch": 0.05338947368421053,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0004985692555317323,
+      "loss": 3.3678,
+      "step": 1268
+    },
+    {
+      "epoch": 0.05343157894736842,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0004985656482782248,
+      "loss": 3.7745,
+      "step": 1269
+    },
+    {
+      "epoch": 0.05347368421052632,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0004985620364961433,
+      "loss": 3.8117,
+      "step": 1270
+    },
+    {
+      "epoch": 0.05351578947368421,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004985584201855535,
+      "loss": 3.7198,
+      "step": 1271
+    },
+    {
+      "epoch": 0.0535578947368421,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004985547993465214,
+      "loss": 3.7349,
+      "step": 1272
+    },
+    {
+      "epoch": 0.0536,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0004985511739791129,
+      "loss": 3.7648,
+      "step": 1273
+    },
+    {
+      "epoch": 0.053642105263157894,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.000498547544083394,
+      "loss": 4.1469,
+      "step": 1274
+    },
+    {
+      "epoch": 0.05368421052631579,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004985439096594308,
+      "loss": 3.9309,
+      "step": 1275
+    },
+    {
+      "epoch": 0.053726315789473685,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004985402707072898,
+      "loss": 3.8224,
+      "step": 1276
+    },
+    {
+      "epoch": 0.053768421052631576,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.000498536627227037,
+      "loss": 3.6667,
+      "step": 1277
+    },
+    {
+      "epoch": 0.053810526315789475,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0004985329792187387,
+      "loss": 3.5203,
+      "step": 1278
+    },
+    {
+      "epoch": 0.05385263157894737,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0004985293266824617,
+      "loss": 4.2161,
+      "step": 1279
+    },
+    {
+      "epoch": 0.053894736842105266,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004985256696182724,
+      "loss": 3.6726,
+      "step": 1280
+    },
+    {
+      "epoch": 0.05393684210526316,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0004985220080262374,
+      "loss": 3.4364,
+      "step": 1281
+    },
+    {
+      "epoch": 0.05397894736842105,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004985183419064233,
+      "loss": 4.0991,
+      "step": 1282
+    },
+    {
+      "epoch": 0.05402105263157895,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004985146712588971,
+      "loss": 3.4965,
+      "step": 1283
+    },
+    {
+      "epoch": 0.05406315789473684,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0004985109960837256,
+      "loss": 3.9651,
+      "step": 1284
+    },
+    {
+      "epoch": 0.05410526315789474,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0004985073163809757,
+      "loss": 3.6569,
+      "step": 1285
+    },
+    {
+      "epoch": 0.05414736842105263,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0004985036321507144,
+      "loss": 3.7453,
+      "step": 1286
+    },
+    {
+      "epoch": 0.05418947368421052,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0004984999433930091,
+      "loss": 3.5282,
+      "step": 1287
+    },
+    {
+      "epoch": 0.05423157894736842,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0004984962501079266,
+      "loss": 3.7703,
+      "step": 1288
+    },
+    {
+      "epoch": 0.054273684210526314,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0004984925522955345,
+      "loss": 3.8629,
+      "step": 1289
+    },
+    {
+      "epoch": 0.05431578947368421,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0004984888499559001,
+      "loss": 3.8146,
+      "step": 1290
+    },
+    {
+      "epoch": 0.054357894736842105,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004984851430890907,
+      "loss": 3.8921,
+      "step": 1291
+    },
+    {
+      "epoch": 0.0544,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.000498481431695174,
+      "loss": 3.4637,
+      "step": 1292
+    },
+    {
+      "epoch": 0.054442105263157896,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004984777157742177,
+      "loss": 3.7084,
+      "step": 1293
+    },
+    {
+      "epoch": 0.05448421052631579,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0004984739953262892,
+      "loss": 3.5203,
+      "step": 1294
+    },
+    {
+      "epoch": 0.05452631578947369,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004984702703514565,
+      "loss": 3.8092,
+      "step": 1295
+    },
+    {
+      "epoch": 0.05456842105263158,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004984665408497874,
+      "loss": 3.7166,
+      "step": 1296
+    },
+    {
+      "epoch": 0.05461052631578947,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004984628068213499,
+      "loss": 3.8697,
+      "step": 1297
+    },
+    {
+      "epoch": 0.05465263157894737,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004984590682662119,
+      "loss": 3.8686,
+      "step": 1298
+    },
+    {
+      "epoch": 0.05469473684210526,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004984553251844417,
+      "loss": 4.2682,
+      "step": 1299
+    },
+    {
+      "epoch": 0.05473684210526316,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004984515775761074,
+      "loss": 3.7709,
+      "step": 1300
+    },
+    {
+      "epoch": 0.05477894736842105,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004984478254412772,
+      "loss": 3.9018,
+      "step": 1301
+    },
+    {
+      "epoch": 0.054821052631578944,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004984440687800197,
+      "loss": 3.4633,
+      "step": 1302
+    },
+    {
+      "epoch": 0.05486315789473684,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.000498440307592403,
+      "loss": 3.7377,
+      "step": 1303
+    },
+    {
+      "epoch": 0.054905263157894735,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004984365418784958,
+      "loss": 4.0835,
+      "step": 1304
+    },
+    {
+      "epoch": 0.054947368421052634,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004984327716383669,
+      "loss": 3.6445,
+      "step": 1305
+    },
+    {
+      "epoch": 0.054989473684210526,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004984289968720846,
+      "loss": 4.3372,
+      "step": 1306
+    },
+    {
+      "epoch": 0.055031578947368424,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.000498425217579718,
+      "loss": 3.402,
+      "step": 1307
+    },
+    {
+      "epoch": 0.055073684210526316,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004984214337613357,
+      "loss": 3.3674,
+      "step": 1308
+    },
+    {
+      "epoch": 0.05511578947368421,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004984176454170068,
+      "loss": 3.8537,
+      "step": 1309
+    },
+    {
+      "epoch": 0.05515789473684211,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0004984138525468003,
+      "loss": 3.4536,
+      "step": 1310
+    },
+    {
+      "epoch": 0.0552,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004984100551507853,
+      "loss": 3.814,
+      "step": 1311
+    },
+    {
+      "epoch": 0.0552421052631579,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004984062532290309,
+      "loss": 3.6992,
+      "step": 1312
+    },
+    {
+      "epoch": 0.05528421052631579,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004984024467816065,
+      "loss": 4.0318,
+      "step": 1313
+    },
+    {
+      "epoch": 0.05532631578947368,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0004983986358085814,
+      "loss": 3.8128,
+      "step": 1314
+    },
+    {
+      "epoch": 0.05536842105263158,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.000498394820310025,
+      "loss": 3.4022,
+      "step": 1315
+    },
+    {
+      "epoch": 0.05541052631578947,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004983910002860067,
+      "loss": 4.1656,
+      "step": 1316
+    },
+    {
+      "epoch": 0.05545263157894737,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0004983871757365963,
+      "loss": 3.4264,
+      "step": 1317
+    },
+    {
+      "epoch": 0.05549473684210526,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0004983833466618634,
+      "loss": 4.0115,
+      "step": 1318
+    },
+    {
+      "epoch": 0.055536842105263155,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004983795130618778,
+      "loss": 3.6987,
+      "step": 1319
+    },
+    {
+      "epoch": 0.055578947368421054,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004983756749367093,
+      "loss": 4.0374,
+      "step": 1320
+    },
+    {
+      "epoch": 0.055621052631578946,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0004983718322864278,
+      "loss": 3.7045,
+      "step": 1321
+    },
+    {
+      "epoch": 0.055663157894736845,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0004983679851111034,
+      "loss": 3.4282,
+      "step": 1322
+    },
+    {
+      "epoch": 0.05570526315789474,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.000498364133410806,
+      "loss": 3.2793,
+      "step": 1323
+    },
+    {
+      "epoch": 0.05574736842105263,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004983602771856061,
+      "loss": 3.9131,
+      "step": 1324
+    },
+    {
+      "epoch": 0.05578947368421053,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004983564164355736,
+      "loss": 3.878,
+      "step": 1325
+    },
+    {
+      "epoch": 0.05583157894736842,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.000498352551160779,
+      "loss": 3.5558,
+      "step": 1326
+    },
+    {
+      "epoch": 0.05587368421052632,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0004983486813612928,
+      "loss": 3.8112,
+      "step": 1327
+    },
+    {
+      "epoch": 0.05591578947368421,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004983448070371854,
+      "loss": 3.7708,
+      "step": 1328
+    },
+    {
+      "epoch": 0.0559578947368421,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0004983409281885273,
+      "loss": 3.6721,
+      "step": 1329
+    },
+    {
+      "epoch": 0.056,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0004983370448153895,
+      "loss": 4.0084,
+      "step": 1330
+    },
+    {
+      "epoch": 0.05604210526315789,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0004983331569178424,
+      "loss": 3.2946,
+      "step": 1331
+    },
+    {
+      "epoch": 0.05608421052631579,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.000498329264495957,
+      "loss": 3.8263,
+      "step": 1332
+    },
+    {
+      "epoch": 0.056126315789473684,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0004983253675498042,
+      "loss": 3.4891,
+      "step": 1333
+    },
+    {
+      "epoch": 0.056168421052631576,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0004983214660794548,
+      "loss": 3.9235,
+      "step": 1334
+    },
+    {
+      "epoch": 0.056210526315789475,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0004983175600849802,
+      "loss": 3.7288,
+      "step": 1335
+    },
+    {
+      "epoch": 0.05625263157894737,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0004983136495664514,
+      "loss": 3.3728,
+      "step": 1336
+    },
+    {
+      "epoch": 0.056294736842105265,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0004983097345239397,
+      "loss": 4.0245,
+      "step": 1337
+    },
+    {
+      "epoch": 0.05633684210526316,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004983058149575163,
+      "loss": 3.7233,
+      "step": 1338
+    },
+    {
+      "epoch": 0.05637894736842105,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0004983018908672527,
+      "loss": 3.6046,
+      "step": 1339
+    },
+    {
+      "epoch": 0.05642105263157895,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0004982979622532204,
+      "loss": 3.8371,
+      "step": 1340
+    },
+    {
+      "epoch": 0.05646315789473684,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0004982940291154911,
+      "loss": 3.9984,
+      "step": 1341
+    },
+    {
+      "epoch": 0.05650526315789474,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0004982900914541361,
+      "loss": 3.7613,
+      "step": 1342
+    },
+    {
+      "epoch": 0.05654736842105263,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0004982861492692275,
+      "loss": 3.3244,
+      "step": 1343
+    },
+    {
+      "epoch": 0.05658947368421053,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.000498282202560837,
+      "loss": 3.3399,
+      "step": 1344
+    },
+    {
+      "epoch": 0.05663157894736842,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004982782513290365,
+      "loss": 3.7781,
+      "step": 1345
+    },
+    {
+      "epoch": 0.056673684210526314,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0004982742955738979,
+      "loss": 4.0354,
+      "step": 1346
+    },
+    {
+      "epoch": 0.05671578947368421,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0004982703352954934,
+      "loss": 3.7529,
+      "step": 1347
+    },
+    {
+      "epoch": 0.056757894736842104,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.000498266370493895,
+      "loss": 3.6758,
+      "step": 1348
+    },
+    {
+      "epoch": 0.0568,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0004982624011691752,
+      "loss": 3.8836,
+      "step": 1349
+    },
+    {
+      "epoch": 0.056842105263157895,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.000498258427321406,
+      "loss": 3.6656,
+      "step": 1350
+    },
+    {
+      "epoch": 0.05688421052631579,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00049825444895066,
+      "loss": 3.819,
+      "step": 1351
+    },
+    {
+      "epoch": 0.056926315789473686,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0004982504660570096,
+      "loss": 3.7427,
+      "step": 1352
+    },
+    {
+      "epoch": 0.05696842105263158,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004982464786405274,
+      "loss": 4.0407,
+      "step": 1353
+    },
+    {
+      "epoch": 0.05701052631578948,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0004982424867012861,
+      "loss": 3.506,
+      "step": 1354
+    },
+    {
+      "epoch": 0.05705263157894737,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004982384902393582,
+      "loss": 3.854,
+      "step": 1355
+    },
+    {
+      "epoch": 0.05709473684210526,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004982344892548168,
+      "loss": 3.7895,
+      "step": 1356
+    },
+    {
+      "epoch": 0.05713684210526316,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0004982304837477348,
+      "loss": 3.8546,
+      "step": 1357
+    },
+    {
+      "epoch": 0.05717894736842105,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0004982264737181848,
+      "loss": 3.6283,
+      "step": 1358
+    },
+    {
+      "epoch": 0.05722105263157895,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004982224591662403,
+      "loss": 3.6073,
+      "step": 1359
+    },
+    {
+      "epoch": 0.05726315789473684,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004982184400919742,
+      "loss": 3.9446,
+      "step": 1360
+    },
+    {
+      "epoch": 0.057305263157894734,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004982144164954596,
+      "loss": 3.8313,
+      "step": 1361
+    },
+    {
+      "epoch": 0.05734736842105263,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004982103883767701,
+      "loss": 3.5819,
+      "step": 1362
+    },
+    {
+      "epoch": 0.057389473684210525,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004982063557359789,
+      "loss": 3.5704,
+      "step": 1363
+    },
+    {
+      "epoch": 0.057431578947368424,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004982023185731597,
+      "loss": 3.7707,
+      "step": 1364
+    },
+    {
+      "epoch": 0.057473684210526316,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0004981982768883857,
+      "loss": 3.6571,
+      "step": 1365
+    },
+    {
+      "epoch": 0.05751578947368421,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0004981942306817308,
+      "loss": 3.4531,
+      "step": 1366
+    },
+    {
+      "epoch": 0.057557894736842106,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004981901799532686,
+      "loss": 4.0188,
+      "step": 1367
+    },
+    {
+      "epoch": 0.0576,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0004981861247030729,
+      "loss": 3.5399,
+      "step": 1368
+    },
+    {
+      "epoch": 0.0576421052631579,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0004981820649312177,
+      "loss": 3.6424,
+      "step": 1369
+    },
+    {
+      "epoch": 0.05768421052631579,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0004981780006377769,
+      "loss": 3.857,
+      "step": 1370
+    },
+    {
+      "epoch": 0.05772631578947368,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004981739318228244,
+      "loss": 3.5255,
+      "step": 1371
+    },
+    {
+      "epoch": 0.05776842105263158,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0004981698584864345,
+      "loss": 3.3467,
+      "step": 1372
+    },
+    {
+      "epoch": 0.05781052631578947,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0004981657806286813,
+      "loss": 4.0037,
+      "step": 1373
+    },
+    {
+      "epoch": 0.05785263157894737,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004981616982496393,
+      "loss": 3.5145,
+      "step": 1374
+    },
+    {
+      "epoch": 0.05789473684210526,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004981576113493827,
+      "loss": 3.9105,
+      "step": 1375
+    },
+    {
+      "epoch": 0.057936842105263155,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.000498153519927986,
+      "loss": 3.6701,
+      "step": 1376
+    },
+    {
+      "epoch": 0.057978947368421054,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004981494239855237,
+      "loss": 3.8234,
+      "step": 1377
+    },
+    {
+      "epoch": 0.058021052631578945,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004981453235220704,
+      "loss": 3.8154,
+      "step": 1378
+    },
+    {
+      "epoch": 0.058063157894736844,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.000498141218537701,
+      "loss": 3.2274,
+      "step": 1379
+    },
+    {
+      "epoch": 0.058105263157894736,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00049813710903249,
+      "loss": 3.6631,
+      "step": 1380
+    },
+    {
+      "epoch": 0.05814736842105263,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0004981329950065125,
+      "loss": 3.4974,
+      "step": 1381
+    },
+    {
+      "epoch": 0.05818947368421053,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004981288764598434,
+      "loss": 4.0409,
+      "step": 1382
+    },
+    {
+      "epoch": 0.05823157894736842,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0004981247533925577,
+      "loss": 3.8766,
+      "step": 1383
+    },
+    {
+      "epoch": 0.05827368421052632,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004981206258047305,
+      "loss": 3.843,
+      "step": 1384
+    },
+    {
+      "epoch": 0.05831578947368421,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004981164936964371,
+      "loss": 4.0245,
+      "step": 1385
+    },
+    {
+      "epoch": 0.05835789473684211,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004981123570677526,
+      "loss": 3.481,
+      "step": 1386
+    },
+    {
+      "epoch": 0.0584,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004981082159187525,
+      "loss": 3.7682,
+      "step": 1387
+    },
+    {
+      "epoch": 0.05844210526315789,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004981040702495123,
+      "loss": 3.8991,
+      "step": 1388
+    },
+    {
+      "epoch": 0.05848421052631579,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0004980999200601074,
+      "loss": 3.7622,
+      "step": 1389
+    },
+    {
+      "epoch": 0.05852631578947368,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004980957653506134,
+      "loss": 3.8332,
+      "step": 1390
+    },
+    {
+      "epoch": 0.05856842105263158,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004980916061211062,
+      "loss": 3.6944,
+      "step": 1391
+    },
+    {
+      "epoch": 0.058610526315789474,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0004980874423716614,
+      "loss": 3.4223,
+      "step": 1392
+    },
+    {
+      "epoch": 0.058652631578947366,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004980832741023547,
+      "loss": 3.6684,
+      "step": 1393
+    },
+    {
+      "epoch": 0.058694736842105265,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004980791013132624,
+      "loss": 4.2956,
+      "step": 1394
+    },
+    {
+      "epoch": 0.05873684210526316,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004980749240044603,
+      "loss": 4.061,
+      "step": 1395
+    },
+    {
+      "epoch": 0.058778947368421056,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0004980707421760246,
+      "loss": 3.5141,
+      "step": 1396
+    },
+    {
+      "epoch": 0.05882105263157895,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004980665558280315,
+      "loss": 3.6213,
+      "step": 1397
+    },
+    {
+      "epoch": 0.05886315789473684,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0004980623649605571,
+      "loss": 3.6913,
+      "step": 1398
+    },
+    {
+      "epoch": 0.05890526315789474,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.000498058169573678,
+      "loss": 3.4471,
+      "step": 1399
+    },
+    {
+      "epoch": 0.05894736842105263,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0004980539696674704,
+      "loss": 3.7601,
+      "step": 1400
+    },
+    {
+      "epoch": 0.05898947368421053,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.000498049765242011,
+      "loss": 3.9941,
+      "step": 1401
+    },
+    {
+      "epoch": 0.05903157894736842,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004980455562973762,
+      "loss": 3.8599,
+      "step": 1402
+    },
+    {
+      "epoch": 0.05907368421052631,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004980413428336428,
+      "loss": 3.4384,
+      "step": 1403
+    },
+    {
+      "epoch": 0.05911578947368421,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0004980371248508878,
+      "loss": 3.9244,
+      "step": 1404
+    },
+    {
+      "epoch": 0.059157894736842104,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004980329023491876,
+      "loss": 3.8382,
+      "step": 1405
+    },
+    {
+      "epoch": 0.0592,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004980286753286195,
+      "loss": 4.0686,
+      "step": 1406
+    },
+    {
+      "epoch": 0.059242105263157895,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004980244437892602,
+      "loss": 3.5042,
+      "step": 1407
+    },
+    {
+      "epoch": 0.059284210526315786,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.000498020207731187,
+      "loss": 3.6917,
+      "step": 1408
+    },
+    {
+      "epoch": 0.059326315789473685,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004980159671544771,
+      "loss": 3.5638,
+      "step": 1409
+    },
+    {
+      "epoch": 0.05936842105263158,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004980117220592076,
+      "loss": 4.0654,
+      "step": 1410
+    },
+    {
+      "epoch": 0.059410526315789476,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.000498007472445456,
+      "loss": 3.9187,
+      "step": 1411
+    },
+    {
+      "epoch": 0.05945263157894737,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004980032183132996,
+      "loss": 3.3682,
+      "step": 1412
+    },
+    {
+      "epoch": 0.05949473684210526,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.000497998959662816,
+      "loss": 3.9349,
+      "step": 1413
+    },
+    {
+      "epoch": 0.05953684210526316,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004979946964940827,
+      "loss": 4.036,
+      "step": 1414
+    },
+    {
+      "epoch": 0.05957894736842105,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0004979904288071774,
+      "loss": 3.2511,
+      "step": 1415
+    },
+    {
+      "epoch": 0.05962105263157895,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004979861566021779,
+      "loss": 3.9683,
+      "step": 1416
+    },
+    {
+      "epoch": 0.05966315789473684,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.000497981879879162,
+      "loss": 3.6874,
+      "step": 1417
+    },
+    {
+      "epoch": 0.059705263157894733,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004979775986382076,
+      "loss": 3.8246,
+      "step": 1418
+    },
+    {
+      "epoch": 0.05974736842105263,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004979733128793927,
+      "loss": 3.8402,
+      "step": 1419
+    },
+    {
+      "epoch": 0.059789473684210524,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004979690226027954,
+      "loss": 3.6947,
+      "step": 1420
+    },
+    {
+      "epoch": 0.05983157894736842,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.0004979647278084939,
+      "loss": 3.9764,
+      "step": 1421
+    },
+    {
+      "epoch": 0.059873684210526315,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004979604284965664,
+      "loss": 4.0506,
+      "step": 1422
+    },
+    {
+      "epoch": 0.059915789473684214,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004979561246670911,
+      "loss": 4.1389,
+      "step": 1423
+    },
+    {
+      "epoch": 0.059957894736842106,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0004979518163201467,
+      "loss": 3.7321,
+      "step": 1424
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0004979475034558115,
+      "loss": 3.6528,
+      "step": 1425
+    },
+    {
+      "epoch": 0.0600421052631579,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0004979431860741641,
+      "loss": 3.5599,
+      "step": 1426
+    },
+    {
+      "epoch": 0.06008421052631579,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0004979388641752833,
+      "loss": 3.9305,
+      "step": 1427
+    },
+    {
+      "epoch": 0.06012631578947369,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0004979345377592475,
+      "loss": 4.0798,
+      "step": 1428
+    },
+    {
+      "epoch": 0.06016842105263158,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0004979302068261358,
+      "loss": 3.3725,
+      "step": 1429
+    },
+    {
+      "epoch": 0.06021052631578947,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0004979258713760271,
+      "loss": 3.6621,
+      "step": 1430
+    },
+    {
+      "epoch": 0.06025263157894737,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.0004979215314090003,
+      "loss": 3.1578,
+      "step": 1431
+    },
+    {
+      "epoch": 0.06029473684210526,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0004979171869251344,
+      "loss": 3.592,
+      "step": 1432
+    },
+    {
+      "epoch": 0.06033684210526316,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004979128379245088,
+      "loss": 3.9118,
+      "step": 1433
+    },
+    {
+      "epoch": 0.06037894736842105,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004979084844072025,
+      "loss": 3.6324,
+      "step": 1434
+    },
+    {
+      "epoch": 0.060421052631578945,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004979041263732948,
+      "loss": 3.9731,
+      "step": 1435
+    },
+    {
+      "epoch": 0.060463157894736844,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0004978997638228654,
+      "loss": 3.941,
+      "step": 1436
+    },
+    {
+      "epoch": 0.060505263157894736,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0004978953967559933,
+      "loss": 3.764,
+      "step": 1437
+    },
+    {
+      "epoch": 0.060547368421052634,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004978910251727586,
+      "loss": 3.9604,
+      "step": 1438
+    },
+    {
+      "epoch": 0.060589473684210526,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0004978866490732406,
+      "loss": 3.8266,
+      "step": 1439
+    },
+    {
+      "epoch": 0.06063157894736842,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0004978822684575191,
+      "loss": 3.8518,
+      "step": 1440
+    },
+    {
+      "epoch": 0.06067368421052632,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004978778833256739,
+      "loss": 3.5031,
+      "step": 1441
+    },
+    {
+      "epoch": 0.06071578947368421,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0004978734936777849,
+      "loss": 3.803,
+      "step": 1442
+    },
+    {
+      "epoch": 0.06075789473684211,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0004978690995139321,
+      "loss": 3.6023,
+      "step": 1443
+    },
+    {
+      "epoch": 0.0608,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0004978647008341956,
+      "loss": 3.9022,
+      "step": 1444
+    },
+    {
+      "epoch": 0.06084210526315789,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0004978602976386554,
+      "loss": 3.6245,
+      "step": 1445
+    },
+    {
+      "epoch": 0.06088421052631579,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0004978558899273918,
+      "loss": 3.909,
+      "step": 1446
+    },
+    {
+      "epoch": 0.06092631578947368,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004978514777004851,
+      "loss": 4.1468,
+      "step": 1447
+    },
+    {
+      "epoch": 0.06096842105263158,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0004978470609580157,
+      "loss": 3.6682,
+      "step": 1448
+    },
+    {
+      "epoch": 0.06101052631578947,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004978426397000642,
+      "loss": 4.0666,
+      "step": 1449
+    },
+    {
+      "epoch": 0.061052631578947365,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0004978382139267108,
+      "loss": 3.8659,
+      "step": 1450
+    },
+    {
+      "epoch": 0.061094736842105264,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0004978337836380364,
+      "loss": 3.6098,
+      "step": 1451
+    },
+    {
+      "epoch": 0.061136842105263156,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004978293488341217,
+      "loss": 3.9778,
+      "step": 1452
+    },
+    {
+      "epoch": 0.061178947368421055,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004978249095150475,
+      "loss": 3.7539,
+      "step": 1453
+    },
+    {
+      "epoch": 0.06122105263157895,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0004978204656808944,
+      "loss": 3.7538,
+      "step": 1454
+    },
+    {
+      "epoch": 0.06126315789473684,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004978160173317438,
+      "loss": 3.9262,
+      "step": 1455
+    },
+    {
+      "epoch": 0.06130526315789474,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0004978115644676765,
+      "loss": 3.6827,
+      "step": 1456
+    },
+    {
+      "epoch": 0.06134736842105263,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004978071070887737,
+      "loss": 3.4441,
+      "step": 1457
+    },
+    {
+      "epoch": 0.06138947368421053,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004978026451951165,
+      "loss": 3.6408,
+      "step": 1458
+    },
+    {
+      "epoch": 0.06143157894736842,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0004977981787867862,
+      "loss": 4.0173,
+      "step": 1459
+    },
+    {
+      "epoch": 0.06147368421052631,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004977937078638644,
+      "loss": 3.8653,
+      "step": 1460
+    },
+    {
+      "epoch": 0.06151578947368421,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004977892324264323,
+      "loss": 4.0035,
+      "step": 1461
+    },
+    {
+      "epoch": 0.0615578947368421,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004977847524745715,
+      "loss": 3.7402,
+      "step": 1462
+    },
+    {
+      "epoch": 0.0616,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0004977802680083637,
+      "loss": 3.412,
+      "step": 1463
+    },
+    {
+      "epoch": 0.061642105263157894,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004977757790278905,
+      "loss": 3.6565,
+      "step": 1464
+    },
+    {
+      "epoch": 0.06168421052631579,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004977712855332337,
+      "loss": 3.5489,
+      "step": 1465
+    },
+    {
+      "epoch": 0.061726315789473685,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004977667875244753,
+      "loss": 3.56,
+      "step": 1466
+    },
+    {
+      "epoch": 0.06176842105263158,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0004977622850016971,
+      "loss": 3.7848,
+      "step": 1467
+    },
+    {
+      "epoch": 0.061810526315789475,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004977577779649811,
+      "loss": 3.6628,
+      "step": 1468
+    },
+    {
+      "epoch": 0.06185263157894737,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0004977532664144096,
+      "loss": 3.4917,
+      "step": 1469
+    },
+    {
+      "epoch": 0.061894736842105266,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004977487503500646,
+      "loss": 3.9277,
+      "step": 1470
+    },
+    {
+      "epoch": 0.06193684210526316,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0004977442297720285,
+      "loss": 4.2681,
+      "step": 1471
+    },
+    {
+      "epoch": 0.06197894736842105,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0004977397046803837,
+      "loss": 3.7431,
+      "step": 1472
+    },
+    {
+      "epoch": 0.06202105263157895,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004977351750752125,
+      "loss": 3.5858,
+      "step": 1473
+    },
+    {
+      "epoch": 0.06206315789473684,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004977306409565974,
+      "loss": 4.2416,
+      "step": 1474
+    },
+    {
+      "epoch": 0.06210526315789474,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004977261023246212,
+      "loss": 3.5031,
+      "step": 1475
+    },
+    {
+      "epoch": 0.06214736842105263,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0004977215591793665,
+      "loss": 3.6567,
+      "step": 1476
+    },
+    {
+      "epoch": 0.062189473684210524,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.000497717011520916,
+      "loss": 4.0074,
+      "step": 1477
+    },
+    {
+      "epoch": 0.06223157894736842,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004977124593493526,
+      "loss": 3.846,
+      "step": 1478
+    },
+    {
+      "epoch": 0.062273684210526314,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0004977079026647591,
+      "loss": 3.6382,
+      "step": 1479
+    },
+    {
+      "epoch": 0.06231578947368421,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004977033414672189,
+      "loss": 3.947,
+      "step": 1480
+    },
+    {
+      "epoch": 0.062357894736842105,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004976987757568148,
+      "loss": 3.5414,
+      "step": 1481
+    },
+    {
+      "epoch": 0.0624,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00049769420553363,
+      "loss": 3.843,
+      "step": 1482
+    },
+    {
+      "epoch": 0.062442105263157896,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004976896307977477,
+      "loss": 3.6364,
+      "step": 1483
+    },
+    {
+      "epoch": 0.06248421052631579,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0004976850515492515,
+      "loss": 3.8939,
+      "step": 1484
+    },
+    {
+      "epoch": 0.06252631578947368,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0004976804677882246,
+      "loss": 3.3034,
+      "step": 1485
+    },
+    {
+      "epoch": 0.06256842105263158,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0004976758795147507,
+      "loss": 3.1922,
+      "step": 1486
+    },
+    {
+      "epoch": 0.06261052631578948,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0004976712867289132,
+      "loss": 3.7994,
+      "step": 1487
+    },
+    {
+      "epoch": 0.06265263157894736,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004976666894307959,
+      "loss": 4.0463,
+      "step": 1488
+    },
+    {
+      "epoch": 0.06269473684210526,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0004976620876204824,
+      "loss": 3.6857,
+      "step": 1489
+    },
+    {
+      "epoch": 0.06273684210526316,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004976574812980567,
+      "loss": 3.6421,
+      "step": 1490
+    },
+    {
+      "epoch": 0.06277894736842106,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004976528704636028,
+      "loss": 3.7976,
+      "step": 1491
+    },
+    {
+      "epoch": 0.06282105263157894,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004976482551172044,
+      "loss": 4.2465,
+      "step": 1492
+    },
+    {
+      "epoch": 0.06286315789473684,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.000497643635258946,
+      "loss": 3.9085,
+      "step": 1493
+    },
+    {
+      "epoch": 0.06290526315789474,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004976390108889113,
+      "loss": 3.5754,
+      "step": 1494
+    },
+    {
+      "epoch": 0.06294736842105263,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0004976343820071849,
+      "loss": 3.7551,
+      "step": 1495
+    },
+    {
+      "epoch": 0.06298947368421053,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.000497629748613851,
+      "loss": 3.6691,
+      "step": 1496
+    },
+    {
+      "epoch": 0.06303157894736842,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004976251107089941,
+      "loss": 3.7454,
+      "step": 1497
+    },
+    {
+      "epoch": 0.06307368421052631,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0004976204682926985,
+      "loss": 3.7499,
+      "step": 1498
+    },
+    {
+      "epoch": 0.06311578947368421,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.000497615821365049,
+      "loss": 4.1201,
+      "step": 1499
+    },
+    {
+      "epoch": 0.06315789473684211,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00049761116992613,
+      "loss": 3.5651,
+      "step": 1500
+    },
+    {
+      "epoch": 0.0632,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0004976065139760267,
+      "loss": 3.3507,
+      "step": 1501
+    },
+    {
+      "epoch": 0.06324210526315789,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004976018535148235,
+      "loss": 3.9816,
+      "step": 1502
+    },
+    {
+      "epoch": 0.06328421052631579,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0004975971885426054,
+      "loss": 3.1666,
+      "step": 1503
+    },
+    {
+      "epoch": 0.06332631578947369,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0004975925190594576,
+      "loss": 3.6817,
+      "step": 1504
+    },
+    {
+      "epoch": 0.06336842105263157,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004975878450654648,
+      "loss": 3.6552,
+      "step": 1505
+    },
+    {
+      "epoch": 0.06341052631578947,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0004975831665607125,
+      "loss": 2.9724,
+      "step": 1506
+    },
+    {
+      "epoch": 0.06345263157894737,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004975784835452858,
+      "loss": 4.1568,
+      "step": 1507
+    },
+    {
+      "epoch": 0.06349473684210526,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00049757379601927,
+      "loss": 3.7084,
+      "step": 1508
+    },
+    {
+      "epoch": 0.06353684210526316,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004975691039827505,
+      "loss": 3.9228,
+      "step": 1509
+    },
+    {
+      "epoch": 0.06357894736842105,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0004975644074358128,
+      "loss": 3.3696,
+      "step": 1510
+    },
+    {
+      "epoch": 0.06362105263157895,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004975597063785424,
+      "loss": 4.021,
+      "step": 1511
+    },
+    {
+      "epoch": 0.06366315789473684,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0004975550008110252,
+      "loss": 3.7964,
+      "step": 1512
+    },
+    {
+      "epoch": 0.06370526315789474,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004975502907333467,
+      "loss": 3.7025,
+      "step": 1513
+    },
+    {
+      "epoch": 0.06374736842105264,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0004975455761455927,
+      "loss": 3.7162,
+      "step": 1514
+    },
+    {
+      "epoch": 0.06378947368421052,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004975408570478491,
+      "loss": 3.7441,
+      "step": 1515
+    },
+    {
+      "epoch": 0.06383157894736842,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004975361334402021,
+      "loss": 3.6904,
+      "step": 1516
+    },
+    {
+      "epoch": 0.06387368421052632,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004975314053227375,
+      "loss": 3.6474,
+      "step": 1517
+    },
+    {
+      "epoch": 0.0639157894736842,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0004975266726955415,
+      "loss": 3.309,
+      "step": 1518
+    },
+    {
+      "epoch": 0.0639578947368421,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0004975219355587004,
+      "loss": 3.7018,
+      "step": 1519
+    },
+    {
+      "epoch": 0.064,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0004975171939123004,
+      "loss": 3.8566,
+      "step": 1520
+    },
+    {
+      "epoch": 0.0640421052631579,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0004975124477564282,
+      "loss": 3.2526,
+      "step": 1521
+    },
+    {
+      "epoch": 0.06408421052631579,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004975076970911697,
+      "loss": 3.9706,
+      "step": 1522
+    },
+    {
+      "epoch": 0.06412631578947368,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004975029419166119,
+      "loss": 4.1502,
+      "step": 1523
+    },
+    {
+      "epoch": 0.06416842105263158,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0004974981822328413,
+      "loss": 3.4427,
+      "step": 1524
+    },
+    {
+      "epoch": 0.06421052631578947,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0004974934180399447,
+      "loss": 3.2783,
+      "step": 1525
+    },
+    {
+      "epoch": 0.06425263157894737,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0004974886493380086,
+      "loss": 3.4603,
+      "step": 1526
+    },
+    {
+      "epoch": 0.06429473684210527,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0004974838761271202,
+      "loss": 3.3505,
+      "step": 1527
+    },
+    {
+      "epoch": 0.06433684210526316,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0004974790984073665,
+      "loss": 3.7598,
+      "step": 1528
+    },
+    {
+      "epoch": 0.06437894736842105,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004974743161788342,
+      "loss": 3.6861,
+      "step": 1529
+    },
+    {
+      "epoch": 0.06442105263157895,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004974695294416107,
+      "loss": 3.8653,
+      "step": 1530
+    },
+    {
+      "epoch": 0.06446315789473685,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004974647381957832,
+      "loss": 3.692,
+      "step": 1531
+    },
+    {
+      "epoch": 0.06450526315789473,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0004974599424414389,
+      "loss": 3.9575,
+      "step": 1532
+    },
+    {
+      "epoch": 0.06454736842105263,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0004974551421786652,
+      "loss": 3.5997,
+      "step": 1533
+    },
+    {
+      "epoch": 0.06458947368421053,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004974503374075496,
+      "loss": 3.7357,
+      "step": 1534
+    },
+    {
+      "epoch": 0.06463157894736841,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0004974455281281794,
+      "loss": 3.5105,
+      "step": 1535
+    },
+    {
+      "epoch": 0.06467368421052631,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004974407143406426,
+      "loss": 4.0164,
+      "step": 1536
+    },
+    {
+      "epoch": 0.06471578947368421,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004974358960450267,
+      "loss": 3.8167,
+      "step": 1537
+    },
+    {
+      "epoch": 0.06475789473684211,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004974310732414194,
+      "loss": 3.5038,
+      "step": 1538
+    },
+    {
+      "epoch": 0.0648,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004974262459299087,
+      "loss": 3.4413,
+      "step": 1539
+    },
+    {
+      "epoch": 0.0648421052631579,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0004974214141105826,
+      "loss": 3.2818,
+      "step": 1540
+    },
+    {
+      "epoch": 0.0648842105263158,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.000497416577783529,
+      "loss": 4.1847,
+      "step": 1541
+    },
+    {
+      "epoch": 0.06492631578947368,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0004974117369488361,
+      "loss": 3.5805,
+      "step": 1542
+    },
+    {
+      "epoch": 0.06496842105263158,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004974068916065919,
+      "loss": 3.6437,
+      "step": 1543
+    },
+    {
+      "epoch": 0.06501052631578948,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004974020417568849,
+      "loss": 3.5432,
+      "step": 1544
+    },
+    {
+      "epoch": 0.06505263157894736,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0004973971873998035,
+      "loss": 3.7226,
+      "step": 1545
+    },
+    {
+      "epoch": 0.06509473684210526,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004973923285354358,
+      "loss": 4.0023,
+      "step": 1546
+    },
+    {
+      "epoch": 0.06513684210526316,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0004973874651638707,
+      "loss": 3.3934,
+      "step": 1547
+    },
+    {
+      "epoch": 0.06517894736842106,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004973825972851967,
+      "loss": 3.8527,
+      "step": 1548
+    },
+    {
+      "epoch": 0.06522105263157894,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0004973777248995023,
+      "loss": 3.6853,
+      "step": 1549
+    },
+    {
+      "epoch": 0.06526315789473684,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004973728480068765,
+      "loss": 3.5738,
+      "step": 1550
+    },
+    {
+      "epoch": 0.06530526315789474,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.000497367966607408,
+      "loss": 3.6356,
+      "step": 1551
+    },
+    {
+      "epoch": 0.06534736842105263,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004973630807011859,
+      "loss": 3.7679,
+      "step": 1552
+    },
+    {
+      "epoch": 0.06538947368421053,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004973581902882989,
+      "loss": 3.5949,
+      "step": 1553
+    },
+    {
+      "epoch": 0.06543157894736842,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004973532953688366,
+      "loss": 3.6411,
+      "step": 1554
+    },
+    {
+      "epoch": 0.06547368421052631,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004973483959428877,
+      "loss": 3.7853,
+      "step": 1555
+    },
+    {
+      "epoch": 0.06551578947368421,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0004973434920105416,
+      "loss": 3.9905,
+      "step": 1556
+    },
+    {
+      "epoch": 0.0655578947368421,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0004973385835718879,
+      "loss": 3.5615,
+      "step": 1557
+    },
+    {
+      "epoch": 0.0656,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0004973336706270156,
+      "loss": 3.3959,
+      "step": 1558
+    },
+    {
+      "epoch": 0.06564210526315789,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004973287531760145,
+      "loss": 3.8301,
+      "step": 1559
+    },
+    {
+      "epoch": 0.06568421052631579,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004973238312189742,
+      "loss": 3.6743,
+      "step": 1560
+    },
+    {
+      "epoch": 0.06572631578947369,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004973189047559842,
+      "loss": 3.6623,
+      "step": 1561
+    },
+    {
+      "epoch": 0.06576842105263157,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004973139737871344,
+      "loss": 3.6671,
+      "step": 1562
+    },
+    {
+      "epoch": 0.06581052631578947,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0004973090383125145,
+      "loss": 3.7153,
+      "step": 1563
+    },
+    {
+      "epoch": 0.06585263157894737,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0004973040983322146,
+      "loss": 4.0499,
+      "step": 1564
+    },
+    {
+      "epoch": 0.06589473684210527,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004972991538463246,
+      "loss": 3.4217,
+      "step": 1565
+    },
+    {
+      "epoch": 0.06593684210526315,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0004972942048549345,
+      "loss": 3.7176,
+      "step": 1566
+    },
+    {
+      "epoch": 0.06597894736842105,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0004972892513581345,
+      "loss": 3.6522,
+      "step": 1567
+    },
+    {
+      "epoch": 0.06602105263157895,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.000497284293356015,
+      "loss": 3.7333,
+      "step": 1568
+    },
+    {
+      "epoch": 0.06606315789473684,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0004972793308486662,
+      "loss": 3.7564,
+      "step": 1569
+    },
+    {
+      "epoch": 0.06610526315789474,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0004972743638361785,
+      "loss": 3.4878,
+      "step": 1570
+    },
+    {
+      "epoch": 0.06614736842105264,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004972693923186425,
+      "loss": 3.5301,
+      "step": 1571
+    },
+    {
+      "epoch": 0.06618947368421052,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0004972644162961486,
+      "loss": 4.0258,
+      "step": 1572
+    },
+    {
+      "epoch": 0.06623157894736842,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0004972594357687876,
+      "loss": 3.8434,
+      "step": 1573
+    },
+    {
+      "epoch": 0.06627368421052632,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0004972544507366502,
+      "loss": 3.5917,
+      "step": 1574
+    },
+    {
+      "epoch": 0.06631578947368422,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0004972494611998272,
+      "loss": 3.9154,
+      "step": 1575
+    },
+    {
+      "epoch": 0.0663578947368421,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0004972444671584095,
+      "loss": 3.4825,
+      "step": 1576
+    },
+    {
+      "epoch": 0.0664,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0004972394686124882,
+      "loss": 3.868,
+      "step": 1577
+    },
+    {
+      "epoch": 0.0664421052631579,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0004972344655621542,
+      "loss": 3.8297,
+      "step": 1578
+    },
+    {
+      "epoch": 0.06648421052631578,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004972294580074989,
+      "loss": 3.8074,
+      "step": 1579
+    },
+    {
+      "epoch": 0.06652631578947368,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0004972244459486132,
+      "loss": 3.4433,
+      "step": 1580
+    },
+    {
+      "epoch": 0.06656842105263158,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004972194293855885,
+      "loss": 3.8557,
+      "step": 1581
+    },
+    {
+      "epoch": 0.06661052631578947,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004972144083185164,
+      "loss": 4.0276,
+      "step": 1582
+    },
+    {
+      "epoch": 0.06665263157894737,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004972093827474882,
+      "loss": 3.5499,
+      "step": 1583
+    },
+    {
+      "epoch": 0.06669473684210526,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004972043526725955,
+      "loss": 3.9461,
+      "step": 1584
+    },
+    {
+      "epoch": 0.06673684210526316,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004971993180939299,
+      "loss": 3.9613,
+      "step": 1585
+    },
+    {
+      "epoch": 0.06677894736842105,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0004971942790115832,
+      "loss": 3.4843,
+      "step": 1586
+    },
+    {
+      "epoch": 0.06682105263157895,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004971892354256473,
+      "loss": 3.4749,
+      "step": 1587
+    },
+    {
+      "epoch": 0.06686315789473685,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004971841873362138,
+      "loss": 3.5181,
+      "step": 1588
+    },
+    {
+      "epoch": 0.06690526315789473,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0004971791347433749,
+      "loss": 3.1767,
+      "step": 1589
+    },
+    {
+      "epoch": 0.06694736842105263,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004971740776472227,
+      "loss": 3.8569,
+      "step": 1590
+    },
+    {
+      "epoch": 0.06698947368421053,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.000497169016047849,
+      "loss": 3.6002,
+      "step": 1591
+    },
+    {
+      "epoch": 0.06703157894736841,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004971639499453465,
+      "loss": 3.8235,
+      "step": 1592
+    },
+    {
+      "epoch": 0.06707368421052631,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.000497158879339807,
+      "loss": 3.5475,
+      "step": 1593
+    },
+    {
+      "epoch": 0.06711578947368421,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004971538042313232,
+      "loss": 3.3942,
+      "step": 1594
+    },
+    {
+      "epoch": 0.06715789473684211,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004971487246199875,
+      "loss": 3.2053,
+      "step": 1595
+    },
+    {
+      "epoch": 0.0672,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004971436405058925,
+      "loss": 4.1573,
+      "step": 1596
+    },
+    {
+      "epoch": 0.0672421052631579,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004971385518891306,
+      "loss": 3.417,
+      "step": 1597
+    },
+    {
+      "epoch": 0.0672842105263158,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004971334587697947,
+      "loss": 3.1885,
+      "step": 1598
+    },
+    {
+      "epoch": 0.06732631578947368,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0004971283611479776,
+      "loss": 3.7261,
+      "step": 1599
+    },
+    {
+      "epoch": 0.06736842105263158,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0004971232590237721,
+      "loss": 3.7957,
+      "step": 1600
+    },
+    {
+      "epoch": 0.06741052631578948,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004971181523972712,
+      "loss": 3.4128,
+      "step": 1601
+    },
+    {
+      "epoch": 0.06745263157894738,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004971130412685679,
+      "loss": 3.5927,
+      "step": 1602
+    },
+    {
+      "epoch": 0.06749473684210526,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004971079256377554,
+      "loss": 3.6683,
+      "step": 1603
+    },
+    {
+      "epoch": 0.06753684210526316,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004971028055049267,
+      "loss": 3.7017,
+      "step": 1604
+    },
+    {
+      "epoch": 0.06757894736842106,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004970976808701753,
+      "loss": 3.9397,
+      "step": 1605
+    },
+    {
+      "epoch": 0.06762105263157894,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004970925517335944,
+      "loss": 3.6519,
+      "step": 1606
+    },
+    {
+      "epoch": 0.06766315789473684,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004970874180952776,
+      "loss": 3.7639,
+      "step": 1607
+    },
+    {
+      "epoch": 0.06770526315789474,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004970822799553184,
+      "loss": 3.2546,
+      "step": 1608
+    },
+    {
+      "epoch": 0.06774736842105263,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0004970771373138103,
+      "loss": 3.998,
+      "step": 1609
+    },
+    {
+      "epoch": 0.06778947368421052,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0004970719901708471,
+      "loss": 3.7759,
+      "step": 1610
+    },
+    {
+      "epoch": 0.06783157894736842,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0004970668385265225,
+      "loss": 3.8989,
+      "step": 1611
+    },
+    {
+      "epoch": 0.06787368421052632,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0004970616823809304,
+      "loss": 3.6788,
+      "step": 1612
+    },
+    {
+      "epoch": 0.06791578947368421,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0004970565217341647,
+      "loss": 3.7188,
+      "step": 1613
+    },
+    {
+      "epoch": 0.0679578947368421,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004970513565863196,
+      "loss": 3.9735,
+      "step": 1614
+    },
+    {
+      "epoch": 0.068,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004970461869374889,
+      "loss": 3.5269,
+      "step": 1615
+    },
+    {
+      "epoch": 0.06804210526315789,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.000497041012787767,
+      "loss": 3.755,
+      "step": 1616
+    },
+    {
+      "epoch": 0.06808421052631579,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004970358341372482,
+      "loss": 4.1299,
+      "step": 1617
+    },
+    {
+      "epoch": 0.06812631578947369,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004970306509860267,
+      "loss": 3.6043,
+      "step": 1618
+    },
+    {
+      "epoch": 0.06816842105263157,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004970254633341969,
+      "loss": 3.9046,
+      "step": 1619
+    },
+    {
+      "epoch": 0.06821052631578947,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004970202711818535,
+      "loss": 3.6095,
+      "step": 1620
+    },
+    {
+      "epoch": 0.06825263157894737,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004970150745290909,
+      "loss": 3.499,
+      "step": 1621
+    },
+    {
+      "epoch": 0.06829473684210527,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0004970098733760039,
+      "loss": 3.5376,
+      "step": 1622
+    },
+    {
+      "epoch": 0.06833684210526315,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0004970046677226873,
+      "loss": 4.0292,
+      "step": 1623
+    },
+    {
+      "epoch": 0.06837894736842105,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0004969994575692358,
+      "loss": 3.5207,
+      "step": 1624
+    },
+    {
+      "epoch": 0.06842105263157895,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0004969942429157443,
+      "loss": 3.3669,
+      "step": 1625
+    },
+    {
+      "epoch": 0.06846315789473684,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0004969890237623082,
+      "loss": 3.4828,
+      "step": 1626
+    },
+    {
+      "epoch": 0.06850526315789474,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.000496983800109022,
+      "loss": 3.7579,
+      "step": 1627
+    },
+    {
+      "epoch": 0.06854736842105263,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004969785719559813,
+      "loss": 3.8547,
+      "step": 1628
+    },
+    {
+      "epoch": 0.06858947368421052,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004969733393032812,
+      "loss": 3.5979,
+      "step": 1629
+    },
+    {
+      "epoch": 0.06863157894736842,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.000496968102151017,
+      "loss": 3.35,
+      "step": 1630
+    },
+    {
+      "epoch": 0.06867368421052632,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004969628604992841,
+      "loss": 3.2912,
+      "step": 1631
+    },
+    {
+      "epoch": 0.06871578947368422,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0004969576143481781,
+      "loss": 3.8525,
+      "step": 1632
+    },
+    {
+      "epoch": 0.0687578947368421,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0004969523636977946,
+      "loss": 3.9714,
+      "step": 1633
+    },
+    {
+      "epoch": 0.0688,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0004969471085482291,
+      "loss": 3.6215,
+      "step": 1634
+    },
+    {
+      "epoch": 0.0688421052631579,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004969418488995775,
+      "loss": 3.7858,
+      "step": 1635
+    },
+    {
+      "epoch": 0.06888421052631578,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004969365847519356,
+      "loss": 3.814,
+      "step": 1636
+    },
+    {
+      "epoch": 0.06892631578947368,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0004969313161053991,
+      "loss": 3.4875,
+      "step": 1637
+    },
+    {
+      "epoch": 0.06896842105263158,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0004969260429600643,
+      "loss": 3.6464,
+      "step": 1638
+    },
+    {
+      "epoch": 0.06901052631578948,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.000496920765316027,
+      "loss": 3.5257,
+      "step": 1639
+    },
+    {
+      "epoch": 0.06905263157894737,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004969154831733836,
+      "loss": 3.6062,
+      "step": 1640
+    },
+    {
+      "epoch": 0.06909473684210526,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004969101965322301,
+      "loss": 3.8344,
+      "step": 1641
+    },
+    {
+      "epoch": 0.06913684210526316,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.000496904905392663,
+      "loss": 3.4854,
+      "step": 1642
+    },
+    {
+      "epoch": 0.06917894736842105,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0004968996097547787,
+      "loss": 3.5177,
+      "step": 1643
+    },
+    {
+      "epoch": 0.06922105263157895,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004968943096186735,
+      "loss": 3.7809,
+      "step": 1644
+    },
+    {
+      "epoch": 0.06926315789473685,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.000496889004984444,
+      "loss": 3.8504,
+      "step": 1645
+    },
+    {
+      "epoch": 0.06930526315789473,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.000496883695852187,
+      "loss": 3.9813,
+      "step": 1646
+    },
+    {
+      "epoch": 0.06934736842105263,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004968783822219991,
+      "loss": 3.2324,
+      "step": 1647
+    },
+    {
+      "epoch": 0.06938947368421053,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0004968730640939773,
+      "loss": 4.0656,
+      "step": 1648
+    },
+    {
+      "epoch": 0.06943157894736843,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0004968677414682181,
+      "loss": 3.6067,
+      "step": 1649
+    },
+    {
+      "epoch": 0.06947368421052631,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0004968624143448188,
+      "loss": 3.816,
+      "step": 1650
+    },
+    {
+      "epoch": 0.06951578947368421,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004968570827238764,
+      "loss": 3.3084,
+      "step": 1651
+    },
+    {
+      "epoch": 0.06955789473684211,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.000496851746605488,
+      "loss": 3.9131,
+      "step": 1652
+    },
+    {
+      "epoch": 0.0696,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004968464059897508,
+      "loss": 3.8056,
+      "step": 1653
+    },
+    {
+      "epoch": 0.0696421052631579,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0004968410608767621,
+      "loss": 3.2827,
+      "step": 1654
+    },
+    {
+      "epoch": 0.06968421052631579,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0004968357112666193,
+      "loss": 3.5338,
+      "step": 1655
+    },
+    {
+      "epoch": 0.06972631578947368,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00049683035715942,
+      "loss": 3.9412,
+      "step": 1656
+    },
+    {
+      "epoch": 0.06976842105263158,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004968249985552615,
+      "loss": 3.8525,
+      "step": 1657
+    },
+    {
+      "epoch": 0.06981052631578948,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004968196354542413,
+      "loss": 3.5625,
+      "step": 1658
+    },
+    {
+      "epoch": 0.06985263157894737,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0004968142678564577,
+      "loss": 4.066,
+      "step": 1659
+    },
+    {
+      "epoch": 0.06989473684210526,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0004968088957620079,
+      "loss": 4.0047,
+      "step": 1660
+    },
+    {
+      "epoch": 0.06993684210526316,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0004968035191709901,
+      "loss": 3.4038,
+      "step": 1661
+    },
+    {
+      "epoch": 0.06997894736842106,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004967981380835021,
+      "loss": 3.6796,
+      "step": 1662
+    },
+    {
+      "epoch": 0.07002105263157894,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004967927524996422,
+      "loss": 3.7461,
+      "step": 1663
+    },
+    {
+      "epoch": 0.07006315789473684,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004967873624195081,
+      "loss": 3.4546,
+      "step": 1664
+    },
+    {
+      "epoch": 0.07010526315789474,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004967819678431983,
+      "loss": 3.6146,
+      "step": 1665
+    },
+    {
+      "epoch": 0.07014736842105262,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.000496776568770811,
+      "loss": 3.7792,
+      "step": 1666
+    },
+    {
+      "epoch": 0.07018947368421052,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004967711652024446,
+      "loss": 3.7992,
+      "step": 1667
+    },
+    {
+      "epoch": 0.07023157894736842,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004967657571381975,
+      "loss": 3.6278,
+      "step": 1668
+    },
+    {
+      "epoch": 0.07027368421052632,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004967603445781683,
+      "loss": 3.7263,
+      "step": 1669
+    },
+    {
+      "epoch": 0.0703157894736842,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004967549275224553,
+      "loss": 3.5737,
+      "step": 1670
+    },
+    {
+      "epoch": 0.0703578947368421,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004967495059711577,
+      "loss": 3.9559,
+      "step": 1671
+    },
+    {
+      "epoch": 0.0704,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004967440799243739,
+      "loss": 3.9898,
+      "step": 1672
+    },
+    {
+      "epoch": 0.07044210526315789,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004967386493822029,
+      "loss": 3.8508,
+      "step": 1673
+    },
+    {
+      "epoch": 0.07048421052631579,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004967332143447436,
+      "loss": 3.3655,
+      "step": 1674
+    },
+    {
+      "epoch": 0.07052631578947369,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0004967277748120951,
+      "loss": 3.6784,
+      "step": 1675
+    },
+    {
+      "epoch": 0.07056842105263159,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004967223307843565,
+      "loss": 3.8198,
+      "step": 1676
+    },
+    {
+      "epoch": 0.07061052631578947,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004967168822616268,
+      "loss": 3.51,
+      "step": 1677
+    },
+    {
+      "epoch": 0.07065263157894737,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0004967114292440054,
+      "loss": 3.3994,
+      "step": 1678
+    },
+    {
+      "epoch": 0.07069473684210527,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004967059717315916,
+      "loss": 3.5857,
+      "step": 1679
+    },
+    {
+      "epoch": 0.07073684210526315,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004967005097244849,
+      "loss": 3.3842,
+      "step": 1680
+    },
+    {
+      "epoch": 0.07077894736842105,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0004966950432227848,
+      "loss": 3.4506,
+      "step": 1681
+    },
+    {
+      "epoch": 0.07082105263157895,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004966895722265908,
+      "loss": 3.8047,
+      "step": 1682
+    },
+    {
+      "epoch": 0.07086315789473684,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0004966840967360027,
+      "loss": 3.6717,
+      "step": 1683
+    },
+    {
+      "epoch": 0.07090526315789474,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004966786167511203,
+      "loss": 3.8388,
+      "step": 1684
+    },
+    {
+      "epoch": 0.07094736842105263,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004966731322720431,
+      "loss": 3.5484,
+      "step": 1685
+    },
+    {
+      "epoch": 0.07098947368421053,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004966676432988714,
+      "loss": 3.2095,
+      "step": 1686
+    },
+    {
+      "epoch": 0.07103157894736842,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004966621498317051,
+      "loss": 3.9041,
+      "step": 1687
+    },
+    {
+      "epoch": 0.07107368421052632,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004966566518706442,
+      "loss": 3.6749,
+      "step": 1688
+    },
+    {
+      "epoch": 0.07111578947368422,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004966511494157888,
+      "loss": 4.2981,
+      "step": 1689
+    },
+    {
+      "epoch": 0.0711578947368421,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004966456424672394,
+      "loss": 3.5101,
+      "step": 1690
+    },
+    {
+      "epoch": 0.0712,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004966401310250962,
+      "loss": 3.6502,
+      "step": 1691
+    },
+    {
+      "epoch": 0.0712421052631579,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004966346150894595,
+      "loss": 3.2637,
+      "step": 1692
+    },
+    {
+      "epoch": 0.07128421052631578,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00049662909466043,
+      "loss": 3.3594,
+      "step": 1693
+    },
+    {
+      "epoch": 0.07132631578947368,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004966235697381082,
+      "loss": 3.831,
+      "step": 1694
+    },
+    {
+      "epoch": 0.07136842105263158,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004966180403225946,
+      "loss": 3.3443,
+      "step": 1695
+    },
+    {
+      "epoch": 0.07141052631578948,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004966125064139902,
+      "loss": 3.6084,
+      "step": 1696
+    },
+    {
+      "epoch": 0.07145263157894736,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0004966069680123955,
+      "loss": 3.3885,
+      "step": 1697
+    },
+    {
+      "epoch": 0.07149473684210526,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004966014251179117,
+      "loss": 3.6821,
+      "step": 1698
+    },
+    {
+      "epoch": 0.07153684210526316,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004965958777306398,
+      "loss": 3.2782,
+      "step": 1699
+    },
+    {
+      "epoch": 0.07157894736842105,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004965903258506806,
+      "loss": 4.0316,
+      "step": 1700
+    },
+    {
+      "epoch": 0.07162105263157895,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0004965847694781355,
+      "loss": 3.7619,
+      "step": 1701
+    },
+    {
+      "epoch": 0.07166315789473685,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0004965792086131054,
+      "loss": 3.5016,
+      "step": 1702
+    },
+    {
+      "epoch": 0.07170526315789473,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0004965736432556921,
+      "loss": 3.6455,
+      "step": 1703
+    },
+    {
+      "epoch": 0.07174736842105263,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004965680734059966,
+      "loss": 4.0863,
+      "step": 1704
+    },
+    {
+      "epoch": 0.07178947368421053,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004965624990641205,
+      "loss": 3.7722,
+      "step": 1705
+    },
+    {
+      "epoch": 0.07183157894736843,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004965569202301655,
+      "loss": 4.0669,
+      "step": 1706
+    },
+    {
+      "epoch": 0.07187368421052631,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004965513369042329,
+      "loss": 3.6657,
+      "step": 1707
+    },
+    {
+      "epoch": 0.07191578947368421,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004965457490864248,
+      "loss": 3.5376,
+      "step": 1708
+    },
+    {
+      "epoch": 0.07195789473684211,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0004965401567768428,
+      "loss": 3.5939,
+      "step": 1709
+    },
+    {
+      "epoch": 0.072,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004965345599755888,
+      "loss": 3.7482,
+      "step": 1710
+    },
+    {
+      "epoch": 0.0720421052631579,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0004965289586827647,
+      "loss": 3.5953,
+      "step": 1711
+    },
+    {
+      "epoch": 0.07208421052631579,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004965233528984727,
+      "loss": 3.8089,
+      "step": 1712
+    },
+    {
+      "epoch": 0.07212631578947368,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0004965177426228149,
+      "loss": 3.478,
+      "step": 1713
+    },
+    {
+      "epoch": 0.07216842105263158,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004965121278558934,
+      "loss": 3.593,
+      "step": 1714
+    },
+    {
+      "epoch": 0.07221052631578948,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0004965065085978106,
+      "loss": 3.3264,
+      "step": 1715
+    },
+    {
+      "epoch": 0.07225263157894737,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004965008848486689,
+      "loss": 4.226,
+      "step": 1716
+    },
+    {
+      "epoch": 0.07229473684210526,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004964952566085707,
+      "loss": 3.3619,
+      "step": 1717
+    },
+    {
+      "epoch": 0.07233684210526316,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0004964896238776184,
+      "loss": 3.2105,
+      "step": 1718
+    },
+    {
+      "epoch": 0.07237894736842106,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004964839866559149,
+      "loss": 3.546,
+      "step": 1719
+    },
+    {
+      "epoch": 0.07242105263157894,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004964783449435627,
+      "loss": 3.5829,
+      "step": 1720
+    },
+    {
+      "epoch": 0.07246315789473684,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004964726987406648,
+      "loss": 3.9014,
+      "step": 1721
+    },
+    {
+      "epoch": 0.07250526315789474,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0004964670480473238,
+      "loss": 2.9922,
+      "step": 1722
+    },
+    {
+      "epoch": 0.07254736842105264,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004964613928636428,
+      "loss": 3.7412,
+      "step": 1723
+    },
+    {
+      "epoch": 0.07258947368421052,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004964557331897248,
+      "loss": 4.0009,
+      "step": 1724
+    },
+    {
+      "epoch": 0.07263157894736842,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004964500690256728,
+      "loss": 3.964,
+      "step": 1725
+    },
+    {
+      "epoch": 0.07267368421052632,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004964444003715902,
+      "loss": 3.4966,
+      "step": 1726
+    },
+    {
+      "epoch": 0.0727157894736842,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0004964387272275802,
+      "loss": 3.8888,
+      "step": 1727
+    },
+    {
+      "epoch": 0.0727578947368421,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0004964330495937462,
+      "loss": 3.4576,
+      "step": 1728
+    },
+    {
+      "epoch": 0.0728,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004964273674701916,
+      "loss": 3.4533,
+      "step": 1729
+    },
+    {
+      "epoch": 0.07284210526315789,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0004964216808570198,
+      "loss": 3.2211,
+      "step": 1730
+    },
+    {
+      "epoch": 0.07288421052631579,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0004964159897543345,
+      "loss": 4.1171,
+      "step": 1731
+    },
+    {
+      "epoch": 0.07292631578947369,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004964102941622395,
+      "loss": 3.5748,
+      "step": 1732
+    },
+    {
+      "epoch": 0.07296842105263159,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004964045940808383,
+      "loss": 3.3029,
+      "step": 1733
+    },
+    {
+      "epoch": 0.07301052631578947,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.000496398889510235,
+      "loss": 3.5237,
+      "step": 1734
+    },
+    {
+      "epoch": 0.07305263157894737,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004963931804505335,
+      "loss": 3.8931,
+      "step": 1735
+    },
+    {
+      "epoch": 0.07309473684210527,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0004963874669018377,
+      "loss": 3.3809,
+      "step": 1736
+    },
+    {
+      "epoch": 0.07313684210526315,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004963817488642517,
+      "loss": 4.1551,
+      "step": 1737
+    },
+    {
+      "epoch": 0.07317894736842105,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0004963760263378798,
+      "loss": 3.2452,
+      "step": 1738
+    },
+    {
+      "epoch": 0.07322105263157895,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004963702993228262,
+      "loss": 3.7427,
+      "step": 1739
+    },
+    {
+      "epoch": 0.07326315789473684,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004963645678191951,
+      "loss": 3.4735,
+      "step": 1740
+    },
+    {
+      "epoch": 0.07330526315789473,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004963588318270912,
+      "loss": 3.3544,
+      "step": 1741
+    },
+    {
+      "epoch": 0.07334736842105263,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004963530913466187,
+      "loss": 3.595,
+      "step": 1742
+    },
+    {
+      "epoch": 0.07338947368421053,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004963473463778824,
+      "loss": 3.8544,
+      "step": 1743
+    },
+    {
+      "epoch": 0.07343157894736842,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004963415969209868,
+      "loss": 3.6384,
+      "step": 1744
+    },
+    {
+      "epoch": 0.07347368421052632,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0004963358429760368,
+      "loss": 3.4096,
+      "step": 1745
+    },
+    {
+      "epoch": 0.07351578947368421,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0004963300845431373,
+      "loss": 3.9575,
+      "step": 1746
+    },
+    {
+      "epoch": 0.0735578947368421,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004963243216223929,
+      "loss": 3.944,
+      "step": 1747
+    },
+    {
+      "epoch": 0.0736,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0004963185542139089,
+      "loss": 3.8351,
+      "step": 1748
+    },
+    {
+      "epoch": 0.0736421052631579,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0004963127823177902,
+      "loss": 3.7207,
+      "step": 1749
+    },
+    {
+      "epoch": 0.07368421052631578,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.000496307005934142,
+      "loss": 3.6938,
+      "step": 1750
+    },
+    {
+      "epoch": 0.07372631578947368,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0004963012250630696,
+      "loss": 3.3901,
+      "step": 1751
+    },
+    {
+      "epoch": 0.07376842105263158,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0004962954397046783,
+      "loss": 3.7485,
+      "step": 1752
+    },
+    {
+      "epoch": 0.07381052631578948,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004962896498590734,
+      "loss": 4.225,
+      "step": 1753
+    },
+    {
+      "epoch": 0.07385263157894736,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0004962838555263604,
+      "loss": 3.3185,
+      "step": 1754
+    },
+    {
+      "epoch": 0.07389473684210526,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0004962780567066451,
+      "loss": 3.4933,
+      "step": 1755
+    },
+    {
+      "epoch": 0.07393684210526316,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004962722534000328,
+      "loss": 3.4005,
+      "step": 1756
+    },
+    {
+      "epoch": 0.07397894736842105,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004962664456066295,
+      "loss": 4.0651,
+      "step": 1757
+    },
+    {
+      "epoch": 0.07402105263157895,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004962606333265408,
+      "loss": 4.0011,
+      "step": 1758
+    },
+    {
+      "epoch": 0.07406315789473684,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0004962548165598728,
+      "loss": 3.5673,
+      "step": 1759
+    },
+    {
+      "epoch": 0.07410526315789474,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004962489953067314,
+      "loss": 3.8407,
+      "step": 1760
+    },
+    {
+      "epoch": 0.07414736842105263,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004962431695672226,
+      "loss": 3.7317,
+      "step": 1761
+    },
+    {
+      "epoch": 0.07418947368421053,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004962373393414526,
+      "loss": 3.8278,
+      "step": 1762
+    },
+    {
+      "epoch": 0.07423157894736843,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0004962315046295276,
+      "loss": 3.7913,
+      "step": 1763
+    },
+    {
+      "epoch": 0.07427368421052631,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004962256654315538,
+      "loss": 3.8082,
+      "step": 1764
+    },
+    {
+      "epoch": 0.07431578947368421,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004962198217476378,
+      "loss": 3.9171,
+      "step": 1765
+    },
+    {
+      "epoch": 0.07435789473684211,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.0004962139735778859,
+      "loss": 3.9598,
+      "step": 1766
+    },
+    {
+      "epoch": 0.0744,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0004962081209224046,
+      "loss": 3.6397,
+      "step": 1767
+    },
+    {
+      "epoch": 0.07444210526315789,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0004962022637813007,
+      "loss": 3.1378,
+      "step": 1768
+    },
+    {
+      "epoch": 0.07448421052631579,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004961964021546808,
+      "loss": 4.1881,
+      "step": 1769
+    },
+    {
+      "epoch": 0.07452631578947369,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0004961905360426519,
+      "loss": 3.4501,
+      "step": 1770
+    },
+    {
+      "epoch": 0.07456842105263158,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004961846654453205,
+      "loss": 3.9726,
+      "step": 1771
+    },
+    {
+      "epoch": 0.07461052631578947,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0004961787903627938,
+      "loss": 3.8051,
+      "step": 1772
+    },
+    {
+      "epoch": 0.07465263157894737,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0004961729107951788,
+      "loss": 3.8196,
+      "step": 1773
+    },
+    {
+      "epoch": 0.07469473684210526,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0004961670267425826,
+      "loss": 3.5223,
+      "step": 1774
+    },
+    {
+      "epoch": 0.07473684210526316,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004961611382051123,
+      "loss": 3.9418,
+      "step": 1775
+    },
+    {
+      "epoch": 0.07477894736842106,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004961552451828755,
+      "loss": 3.6962,
+      "step": 1776
+    },
+    {
+      "epoch": 0.07482105263157894,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004961493476759792,
+      "loss": 3.713,
+      "step": 1777
+    },
+    {
+      "epoch": 0.07486315789473684,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.000496143445684531,
+      "loss": 3.0897,
+      "step": 1778
+    },
+    {
+      "epoch": 0.07490526315789474,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004961375392086386,
+      "loss": 3.9444,
+      "step": 1779
+    },
+    {
+      "epoch": 0.07494736842105264,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004961316282484093,
+      "loss": 3.5383,
+      "step": 1780
+    },
+    {
+      "epoch": 0.07498947368421052,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0004961257128039509,
+      "loss": 3.8087,
+      "step": 1781
+    },
+    {
+      "epoch": 0.07503157894736842,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0004961197928753712,
+      "loss": 3.4868,
+      "step": 1782
+    },
+    {
+      "epoch": 0.07507368421052632,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004961138684627781,
+      "loss": 3.7377,
+      "step": 1783
+    },
+    {
+      "epoch": 0.0751157894736842,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004961079395662795,
+      "loss": 3.5204,
+      "step": 1784
+    },
+    {
+      "epoch": 0.0751578947368421,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004961020061859834,
+      "loss": 3.6972,
+      "step": 1785
+    },
+    {
+      "epoch": 0.0752,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004960960683219978,
+      "loss": 3.7269,
+      "step": 1786
+    },
+    {
+      "epoch": 0.07524210526315789,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004960901259744311,
+      "loss": 3.6722,
+      "step": 1787
+    },
+    {
+      "epoch": 0.07528421052631579,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004960841791433915,
+      "loss": 3.8057,
+      "step": 1788
+    },
+    {
+      "epoch": 0.07532631578947369,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004960782278289872,
+      "loss": 3.5651,
+      "step": 1789
+    },
+    {
+      "epoch": 0.07536842105263158,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004960722720313267,
+      "loss": 4.2334,
+      "step": 1790
+    },
+    {
+      "epoch": 0.07541052631578947,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0004960663117505185,
+      "loss": 3.6041,
+      "step": 1791
+    },
+    {
+      "epoch": 0.07545263157894737,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004960603469866713,
+      "loss": 3.3041,
+      "step": 1792
+    },
+    {
+      "epoch": 0.07549473684210527,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004960543777398936,
+      "loss": 3.0062,
+      "step": 1793
+    },
+    {
+      "epoch": 0.07553684210526315,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004960484040102943,
+      "loss": 3.8303,
+      "step": 1794
+    },
+    {
+      "epoch": 0.07557894736842105,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004960424257979822,
+      "loss": 3.6535,
+      "step": 1795
+    },
+    {
+      "epoch": 0.07562105263157895,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004960364431030662,
+      "loss": 3.8331,
+      "step": 1796
+    },
+    {
+      "epoch": 0.07566315789473685,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004960304559256553,
+      "loss": 3.7033,
+      "step": 1797
+    },
+    {
+      "epoch": 0.07570526315789473,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004960244642658585,
+      "loss": 3.6106,
+      "step": 1798
+    },
+    {
+      "epoch": 0.07574736842105263,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004960184681237849,
+      "loss": 3.596,
+      "step": 1799
+    },
+    {
+      "epoch": 0.07578947368421053,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0004960124674995441,
+      "loss": 4.1051,
+      "step": 1800
+    },
+    {
+      "epoch": 0.07583157894736842,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004960064623932451,
+      "loss": 3.5837,
+      "step": 1801
+    },
+    {
+      "epoch": 0.07587368421052632,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004960004528049974,
+      "loss": 3.7843,
+      "step": 1802
+    },
+    {
+      "epoch": 0.07591578947368421,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0004959944387349105,
+      "loss": 3.5628,
+      "step": 1803
+    },
+    {
+      "epoch": 0.0759578947368421,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.000495988420183094,
+      "loss": 4.0603,
+      "step": 1804
+    },
+    {
+      "epoch": 0.076,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0004959823971496574,
+      "loss": 3.3416,
+      "step": 1805
+    },
+    {
+      "epoch": 0.0760421052631579,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0004959763696347106,
+      "loss": 3.6872,
+      "step": 1806
+    },
+    {
+      "epoch": 0.0760842105263158,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004959703376383634,
+      "loss": 3.8652,
+      "step": 1807
+    },
+    {
+      "epoch": 0.07612631578947368,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004959643011607256,
+      "loss": 3.5389,
+      "step": 1808
+    },
+    {
+      "epoch": 0.07616842105263158,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0004959582602019072,
+      "loss": 3.517,
+      "step": 1809
+    },
+    {
+      "epoch": 0.07621052631578948,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004959522147620183,
+      "loss": 3.8383,
+      "step": 1810
+    },
+    {
+      "epoch": 0.07625263157894736,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004959461648411692,
+      "loss": 3.6336,
+      "step": 1811
+    },
+    {
+      "epoch": 0.07629473684210526,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0004959401104394697,
+      "loss": 3.5757,
+      "step": 1812
+    },
+    {
+      "epoch": 0.07633684210526316,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004959340515570305,
+      "loss": 4.0782,
+      "step": 1813
+    },
+    {
+      "epoch": 0.07637894736842105,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004959279881939618,
+      "loss": 3.6384,
+      "step": 1814
+    },
+    {
+      "epoch": 0.07642105263157895,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004959219203503741,
+      "loss": 3.8301,
+      "step": 1815
+    },
+    {
+      "epoch": 0.07646315789473684,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.000495915848026378,
+      "loss": 3.281,
+      "step": 1816
+    },
+    {
+      "epoch": 0.07650526315789474,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004959097712220841,
+      "loss": 3.7455,
+      "step": 1817
+    },
+    {
+      "epoch": 0.07654736842105263,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.000495903689937603,
+      "loss": 4.0063,
+      "step": 1818
+    },
+    {
+      "epoch": 0.07658947368421053,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004958976041730457,
+      "loss": 3.8636,
+      "step": 1819
+    },
+    {
+      "epoch": 0.07663157894736843,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004958915139285229,
+      "loss": 3.7563,
+      "step": 1820
+    },
+    {
+      "epoch": 0.07667368421052631,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004958854192041456,
+      "loss": 4.0655,
+      "step": 1821
+    },
+    {
+      "epoch": 0.07671578947368421,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.000495879320000025,
+      "loss": 4.0406,
+      "step": 1822
+    },
+    {
+      "epoch": 0.07675789473684211,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.000495873216316272,
+      "loss": 3.626,
+      "step": 1823
+    },
+    {
+      "epoch": 0.0768,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004958671081529979,
+      "loss": 3.9336,
+      "step": 1824
+    },
+    {
+      "epoch": 0.07684210526315789,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004958609955103139,
+      "loss": 3.8481,
+      "step": 1825
+    },
+    {
+      "epoch": 0.07688421052631579,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004958548783883316,
+      "loss": 3.5711,
+      "step": 1826
+    },
+    {
+      "epoch": 0.07692631578947369,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004958487567871621,
+      "loss": 3.6206,
+      "step": 1827
+    },
+    {
+      "epoch": 0.07696842105263157,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004958426307069173,
+      "loss": 3.7727,
+      "step": 1828
+    },
+    {
+      "epoch": 0.07701052631578947,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0004958365001477084,
+      "loss": 3.5596,
+      "step": 1829
+    },
+    {
+      "epoch": 0.07705263157894737,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004958303651096475,
+      "loss": 3.4394,
+      "step": 1830
+    },
+    {
+      "epoch": 0.07709473684210526,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004958242255928461,
+      "loss": 3.5369,
+      "step": 1831
+    },
+    {
+      "epoch": 0.07713684210526316,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0004958180815974162,
+      "loss": 3.4367,
+      "step": 1832
+    },
+    {
+      "epoch": 0.07717894736842106,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004958119331234696,
+      "loss": 2.9811,
+      "step": 1833
+    },
+    {
+      "epoch": 0.07722105263157895,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0004958057801711185,
+      "loss": 3.5743,
+      "step": 1834
+    },
+    {
+      "epoch": 0.07726315789473684,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004957996227404747,
+      "loss": 3.6136,
+      "step": 1835
+    },
+    {
+      "epoch": 0.07730526315789474,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004957934608316507,
+      "loss": 3.6534,
+      "step": 1836
+    },
+    {
+      "epoch": 0.07734736842105264,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0004957872944447588,
+      "loss": 3.388,
+      "step": 1837
+    },
+    {
+      "epoch": 0.07738947368421052,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.000495781123579911,
+      "loss": 3.944,
+      "step": 1838
+    },
+    {
+      "epoch": 0.07743157894736842,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004957749482372199,
+      "loss": 3.9884,
+      "step": 1839
+    },
+    {
+      "epoch": 0.07747368421052632,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0004957687684167981,
+      "loss": 3.8765,
+      "step": 1840
+    },
+    {
+      "epoch": 0.0775157894736842,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.000495762584118758,
+      "loss": 3.8636,
+      "step": 1841
+    },
+    {
+      "epoch": 0.0775578947368421,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004957563953432125,
+      "loss": 3.6129,
+      "step": 1842
+    },
+    {
+      "epoch": 0.0776,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004957502020902741,
+      "loss": 3.6204,
+      "step": 1843
+    },
+    {
+      "epoch": 0.0776421052631579,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004957440043600558,
+      "loss": 4.1154,
+      "step": 1844
+    },
+    {
+      "epoch": 0.07768421052631579,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004957378021526705,
+      "loss": 4.3149,
+      "step": 1845
+    },
+    {
+      "epoch": 0.07772631578947369,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0004957315954682312,
+      "loss": 3.4905,
+      "step": 1846
+    },
+    {
+      "epoch": 0.07776842105263158,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0004957253843068509,
+      "loss": 3.4097,
+      "step": 1847
+    },
+    {
+      "epoch": 0.07781052631578947,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004957191686686429,
+      "loss": 3.6746,
+      "step": 1848
+    },
+    {
+      "epoch": 0.07785263157894737,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004957129485537203,
+      "loss": 3.4614,
+      "step": 1849
+    },
+    {
+      "epoch": 0.07789473684210527,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004957067239621965,
+      "loss": 3.552,
+      "step": 1850
+    },
+    {
+      "epoch": 0.07793684210526315,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0004957004948941848,
+      "loss": 3.7481,
+      "step": 1851
+    },
+    {
+      "epoch": 0.07797894736842105,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004956942613497989,
+      "loss": 3.4045,
+      "step": 1852
+    },
+    {
+      "epoch": 0.07802105263157895,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0004956880233291521,
+      "loss": 3.5658,
+      "step": 1853
+    },
+    {
+      "epoch": 0.07806315789473685,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0004956817808323583,
+      "loss": 3.382,
+      "step": 1854
+    },
+    {
+      "epoch": 0.07810526315789473,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.000495675533859531,
+      "loss": 3.6634,
+      "step": 1855
+    },
+    {
+      "epoch": 0.07814736842105263,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004956692824107842,
+      "loss": 3.3574,
+      "step": 1856
+    },
+    {
+      "epoch": 0.07818947368421053,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004956630264862317,
+      "loss": 3.3688,
+      "step": 1857
+    },
+    {
+      "epoch": 0.07823157894736842,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0004956567660859876,
+      "loss": 3.4702,
+      "step": 1858
+    },
+    {
+      "epoch": 0.07827368421052631,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004956505012101658,
+      "loss": 3.3042,
+      "step": 1859
+    },
+    {
+      "epoch": 0.07831578947368421,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004956442318588805,
+      "loss": 3.6902,
+      "step": 1860
+    },
+    {
+      "epoch": 0.0783578947368421,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004956379580322459,
+      "loss": 3.6241,
+      "step": 1861
+    },
+    {
+      "epoch": 0.0784,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004956316797303762,
+      "loss": 3.8502,
+      "step": 1862
+    },
+    {
+      "epoch": 0.0784421052631579,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.000495625396953386,
+      "loss": 3.7562,
+      "step": 1863
+    },
+    {
+      "epoch": 0.0784842105263158,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0004956191097013898,
+      "loss": 3.8918,
+      "step": 1864
+    },
+    {
+      "epoch": 0.07852631578947368,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004956128179745018,
+      "loss": 3.6625,
+      "step": 1865
+    },
+    {
+      "epoch": 0.07856842105263158,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.000495606521772837,
+      "loss": 3.9097,
+      "step": 1866
+    },
+    {
+      "epoch": 0.07861052631578948,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004956002210965098,
+      "loss": 3.82,
+      "step": 1867
+    },
+    {
+      "epoch": 0.07865263157894736,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004955939159456353,
+      "loss": 3.3097,
+      "step": 1868
+    },
+    {
+      "epoch": 0.07869473684210526,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.0004955876063203281,
+      "loss": 2.8061,
+      "step": 1869
+    },
+    {
+      "epoch": 0.07873684210526316,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004955812922207033,
+      "loss": 3.5491,
+      "step": 1870
+    },
+    {
+      "epoch": 0.07877894736842106,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004955749736468758,
+      "loss": 3.8769,
+      "step": 1871
+    },
+    {
+      "epoch": 0.07882105263157894,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.000495568650598961,
+      "loss": 3.2703,
+      "step": 1872
+    },
+    {
+      "epoch": 0.07886315789473684,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0004955623230770738,
+      "loss": 3.4897,
+      "step": 1873
+    },
+    {
+      "epoch": 0.07890526315789474,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0004955559910813297,
+      "loss": 3.4976,
+      "step": 1874
+    },
+    {
+      "epoch": 0.07894736842105263,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0004955496546118439,
+      "loss": 3.6242,
+      "step": 1875
+    },
+    {
+      "epoch": 0.07898947368421053,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0004955433136687318,
+      "loss": 3.8319,
+      "step": 1876
+    },
+    {
+      "epoch": 0.07903157894736842,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004955369682521091,
+      "loss": 3.6944,
+      "step": 1877
+    },
+    {
+      "epoch": 0.07907368421052631,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004955306183620915,
+      "loss": 3.8731,
+      "step": 1878
+    },
+    {
+      "epoch": 0.07911578947368421,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004955242639987943,
+      "loss": 3.7115,
+      "step": 1879
+    },
+    {
+      "epoch": 0.07915789473684211,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0004955179051623335,
+      "loss": 3.8529,
+      "step": 1880
+    },
+    {
+      "epoch": 0.0792,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.000495511541852825,
+      "loss": 3.3012,
+      "step": 1881
+    },
+    {
+      "epoch": 0.07924210526315789,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0004955051740703848,
+      "loss": 3.3343,
+      "step": 1882
+    },
+    {
+      "epoch": 0.07928421052631579,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004954988018151285,
+      "loss": 3.323,
+      "step": 1883
+    },
+    {
+      "epoch": 0.07932631578947369,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0004954924250871728,
+      "loss": 4.0457,
+      "step": 1884
+    },
+    {
+      "epoch": 0.07936842105263157,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0004954860438866333,
+      "loss": 3.6411,
+      "step": 1885
+    },
+    {
+      "epoch": 0.07941052631578947,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004954796582136266,
+      "loss": 3.7893,
+      "step": 1886
+    },
+    {
+      "epoch": 0.07945263157894737,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004954732680682689,
+      "loss": 3.5343,
+      "step": 1887
+    },
+    {
+      "epoch": 0.07949473684210526,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004954668734506768,
+      "loss": 3.7463,
+      "step": 1888
+    },
+    {
+      "epoch": 0.07953684210526316,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004954604743609666,
+      "loss": 3.5578,
+      "step": 1889
+    },
+    {
+      "epoch": 0.07957894736842105,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.000495454070799255,
+      "loss": 3.3886,
+      "step": 1890
+    },
+    {
+      "epoch": 0.07962105263157895,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0004954476627656586,
+      "loss": 3.4701,
+      "step": 1891
+    },
+    {
+      "epoch": 0.07966315789473684,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0004954412502602942,
+      "loss": 3.3631,
+      "step": 1892
+    },
+    {
+      "epoch": 0.07970526315789474,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004954348332832785,
+      "loss": 3.7645,
+      "step": 1893
+    },
+    {
+      "epoch": 0.07974736842105264,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0004954284118347287,
+      "loss": 3.7881,
+      "step": 1894
+    },
+    {
+      "epoch": 0.07978947368421052,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0004954219859147614,
+      "loss": 3.7712,
+      "step": 1895
+    },
+    {
+      "epoch": 0.07983157894736842,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004954155555234941,
+      "loss": 3.6624,
+      "step": 1896
+    },
+    {
+      "epoch": 0.07987368421052632,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0004954091206610435,
+      "loss": 3.4891,
+      "step": 1897
+    },
+    {
+      "epoch": 0.0799157894736842,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004954026813275272,
+      "loss": 3.681,
+      "step": 1898
+    },
+    {
+      "epoch": 0.0799578947368421,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004953962375230623,
+      "loss": 3.2556,
+      "step": 1899
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004953897892477664,
+      "loss": 3.9049,
+      "step": 1900
+    },
+    {
+      "epoch": 0.0800421052631579,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004953833365017568,
+      "loss": 3.9772,
+      "step": 1901
+    },
+    {
+      "epoch": 0.08008421052631579,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.000495376879285151,
+      "loss": 3.7265,
+      "step": 1902
+    },
+    {
+      "epoch": 0.08012631578947368,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0004953704175980669,
+      "loss": 3.6198,
+      "step": 1903
+    },
+    {
+      "epoch": 0.08016842105263158,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004953639514406221,
+      "loss": 3.9809,
+      "step": 1904
+    },
+    {
+      "epoch": 0.08021052631578947,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.0004953574808129344,
+      "loss": 3.5986,
+      "step": 1905
+    },
+    {
+      "epoch": 0.08025263157894737,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004953510057151216,
+      "loss": 3.7484,
+      "step": 1906
+    },
+    {
+      "epoch": 0.08029473684210527,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004953445261473018,
+      "loss": 3.5561,
+      "step": 1907
+    },
+    {
+      "epoch": 0.08033684210526316,
+      "grad_norm": 0.5,
+      "learning_rate": 0.000495338042109593,
+      "loss": 3.354,
+      "step": 1908
+    },
+    {
+      "epoch": 0.08037894736842105,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004953315536021134,
+      "loss": 3.554,
+      "step": 1909
+    },
+    {
+      "epoch": 0.08042105263157895,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.000495325060624981,
+      "loss": 3.6499,
+      "step": 1910
+    },
+    {
+      "epoch": 0.08046315789473685,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0004953185631783143,
+      "loss": 3.4099,
+      "step": 1911
+    },
+    {
+      "epoch": 0.08050526315789473,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004953120612622316,
+      "loss": 3.3944,
+      "step": 1912
+    },
+    {
+      "epoch": 0.08054736842105263,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004953055548768514,
+      "loss": 3.4921,
+      "step": 1913
+    },
+    {
+      "epoch": 0.08058947368421053,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0004952990440222923,
+      "loss": 3.6982,
+      "step": 1914
+    },
+    {
+      "epoch": 0.08063157894736842,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004952925286986727,
+      "loss": 3.4956,
+      "step": 1915
+    },
+    {
+      "epoch": 0.08067368421052631,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004952860089061116,
+      "loss": 3.9061,
+      "step": 1916
+    },
+    {
+      "epoch": 0.08071578947368421,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004952794846447274,
+      "loss": 3.4516,
+      "step": 1917
+    },
+    {
+      "epoch": 0.08075789473684211,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004952729559146393,
+      "loss": 3.528,
+      "step": 1918
+    },
+    {
+      "epoch": 0.0808,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004952664227159661,
+      "loss": 3.8554,
+      "step": 1919
+    },
+    {
+      "epoch": 0.0808421052631579,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0004952598850488268,
+      "loss": 3.1512,
+      "step": 1920
+    },
+    {
+      "epoch": 0.0808842105263158,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004952533429133407,
+      "loss": 3.45,
+      "step": 1921
+    },
+    {
+      "epoch": 0.08092631578947368,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0004952467963096268,
+      "loss": 3.3787,
+      "step": 1922
+    },
+    {
+      "epoch": 0.08096842105263158,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0004952402452378043,
+      "loss": 3.9287,
+      "step": 1923
+    },
+    {
+      "epoch": 0.08101052631578948,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004952336896979929,
+      "loss": 3.8087,
+      "step": 1924
+    },
+    {
+      "epoch": 0.08105263157894736,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004952271296903117,
+      "loss": 3.6194,
+      "step": 1925
+    },
+    {
+      "epoch": 0.08109473684210526,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004952205652148802,
+      "loss": 3.6588,
+      "step": 1926
+    },
+    {
+      "epoch": 0.08113684210526316,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0004952139962718183,
+      "loss": 3.461,
+      "step": 1927
+    },
+    {
+      "epoch": 0.08117894736842106,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0004952074228612455,
+      "loss": 3.3798,
+      "step": 1928
+    },
+    {
+      "epoch": 0.08122105263157894,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004952008449832815,
+      "loss": 4.0183,
+      "step": 1929
+    },
+    {
+      "epoch": 0.08126315789473684,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004951942626380462,
+      "loss": 3.591,
+      "step": 1930
+    },
+    {
+      "epoch": 0.08130526315789474,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004951876758256594,
+      "loss": 3.7295,
+      "step": 1931
+    },
+    {
+      "epoch": 0.08134736842105263,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004951810845462415,
+      "loss": 4.018,
+      "step": 1932
+    },
+    {
+      "epoch": 0.08138947368421053,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004951744887999121,
+      "loss": 3.831,
+      "step": 1933
+    },
+    {
+      "epoch": 0.08143157894736842,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004951678885867917,
+      "loss": 3.7,
+      "step": 1934
+    },
+    {
+      "epoch": 0.08147368421052631,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004951612839070005,
+      "loss": 3.4071,
+      "step": 1935
+    },
+    {
+      "epoch": 0.08151578947368421,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004951546747606585,
+      "loss": 3.9388,
+      "step": 1936
+    },
+    {
+      "epoch": 0.08155789473684211,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004951480611478866,
+      "loss": 3.5639,
+      "step": 1937
+    },
+    {
+      "epoch": 0.0816,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.000495141443068805,
+      "loss": 3.745,
+      "step": 1938
+    },
+    {
+      "epoch": 0.08164210526315789,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004951348205235344,
+      "loss": 3.8362,
+      "step": 1939
+    },
+    {
+      "epoch": 0.08168421052631579,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0004951281935121953,
+      "loss": 3.6108,
+      "step": 1940
+    },
+    {
+      "epoch": 0.08172631578947369,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0004951215620349086,
+      "loss": 3.4748,
+      "step": 1941
+    },
+    {
+      "epoch": 0.08176842105263157,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004951149260917949,
+      "loss": 3.7909,
+      "step": 1942
+    },
+    {
+      "epoch": 0.08181052631578947,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0004951082856829753,
+      "loss": 3.6138,
+      "step": 1943
+    },
+    {
+      "epoch": 0.08185263157894737,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0004951016408085708,
+      "loss": 3.6875,
+      "step": 1944
+    },
+    {
+      "epoch": 0.08189473684210526,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004950949914687023,
+      "loss": 2.9941,
+      "step": 1945
+    },
+    {
+      "epoch": 0.08193684210526316,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0004950883376634911,
+      "loss": 3.4746,
+      "step": 1946
+    },
+    {
+      "epoch": 0.08197894736842105,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004950816793930584,
+      "loss": 3.4375,
+      "step": 1947
+    },
+    {
+      "epoch": 0.08202105263157895,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004950750166575254,
+      "loss": 3.7499,
+      "step": 1948
+    },
+    {
+      "epoch": 0.08206315789473684,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004950683494570135,
+      "loss": 3.7055,
+      "step": 1949
+    },
+    {
+      "epoch": 0.08210526315789474,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004950616777916442,
+      "loss": 3.4872,
+      "step": 1950
+    },
+    {
+      "epoch": 0.08214736842105264,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0004950550016615392,
+      "loss": 3.6569,
+      "step": 1951
+    },
+    {
+      "epoch": 0.08218947368421052,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004950483210668199,
+      "loss": 3.7398,
+      "step": 1952
+    },
+    {
+      "epoch": 0.08223157894736842,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0004950416360076081,
+      "loss": 3.2988,
+      "step": 1953
+    },
+    {
+      "epoch": 0.08227368421052632,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0004950349464840257,
+      "loss": 3.714,
+      "step": 1954
+    },
+    {
+      "epoch": 0.08231578947368422,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004950282524961944,
+      "loss": 3.4952,
+      "step": 1955
+    },
+    {
+      "epoch": 0.0823578947368421,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004950215540442362,
+      "loss": 3.7531,
+      "step": 1956
+    },
+    {
+      "epoch": 0.0824,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004950148511282732,
+      "loss": 3.6568,
+      "step": 1957
+    },
+    {
+      "epoch": 0.0824421052631579,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0004950081437484277,
+      "loss": 3.3777,
+      "step": 1958
+    },
+    {
+      "epoch": 0.08248421052631578,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004950014319048214,
+      "loss": 3.5235,
+      "step": 1959
+    },
+    {
+      "epoch": 0.08252631578947368,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.000494994715597577,
+      "loss": 3.3017,
+      "step": 1960
+    },
+    {
+      "epoch": 0.08256842105263158,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004949879948268167,
+      "loss": 3.6792,
+      "step": 1961
+    },
+    {
+      "epoch": 0.08261052631578947,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.000494981269592663,
+      "loss": 3.7127,
+      "step": 1962
+    },
+    {
+      "epoch": 0.08265263157894737,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0004949745398952385,
+      "loss": 3.4819,
+      "step": 1963
+    },
+    {
+      "epoch": 0.08269473684210527,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0004949678057346656,
+      "loss": 3.3384,
+      "step": 1964
+    },
+    {
+      "epoch": 0.08273684210526316,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004949610671110672,
+      "loss": 3.8225,
+      "step": 1965
+    },
+    {
+      "epoch": 0.08277894736842105,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0004949543240245659,
+      "loss": 3.2495,
+      "step": 1966
+    },
+    {
+      "epoch": 0.08282105263157895,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004949475764752845,
+      "loss": 3.3999,
+      "step": 1967
+    },
+    {
+      "epoch": 0.08286315789473685,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004949408244633464,
+      "loss": 3.5412,
+      "step": 1968
+    },
+    {
+      "epoch": 0.08290526315789473,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.000494934067988874,
+      "loss": 3.2931,
+      "step": 1969
+    },
+    {
+      "epoch": 0.08294736842105263,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004949273070519906,
+      "loss": 3.4803,
+      "step": 1970
+    },
+    {
+      "epoch": 0.08298947368421053,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004949205416528196,
+      "loss": 3.8478,
+      "step": 1971
+    },
+    {
+      "epoch": 0.08303157894736841,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0004949137717914842,
+      "loss": 3.6972,
+      "step": 1972
+    },
+    {
+      "epoch": 0.08307368421052631,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004949069974681075,
+      "loss": 3.6583,
+      "step": 1973
+    },
+    {
+      "epoch": 0.08311578947368421,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004949002186828131,
+      "loss": 3.9406,
+      "step": 1974
+    },
+    {
+      "epoch": 0.08315789473684211,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004948934354357244,
+      "loss": 3.8938,
+      "step": 1975
+    },
+    {
+      "epoch": 0.0832,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004948866477269651,
+      "loss": 3.6138,
+      "step": 1976
+    },
+    {
+      "epoch": 0.0832421052631579,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004948798555566589,
+      "loss": 3.387,
+      "step": 1977
+    },
+    {
+      "epoch": 0.0832842105263158,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004948730589249292,
+      "loss": 3.7547,
+      "step": 1978
+    },
+    {
+      "epoch": 0.08332631578947368,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004948662578319003,
+      "loss": 3.7165,
+      "step": 1979
+    },
+    {
+      "epoch": 0.08336842105263158,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0004948594522776958,
+      "loss": 3.4931,
+      "step": 1980
+    },
+    {
+      "epoch": 0.08341052631578948,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004948526422624398,
+      "loss": 3.7863,
+      "step": 1981
+    },
+    {
+      "epoch": 0.08345263157894736,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004948458277862563,
+      "loss": 3.8176,
+      "step": 1982
+    },
+    {
+      "epoch": 0.08349473684210526,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004948390088492696,
+      "loss": 3.5962,
+      "step": 1983
+    },
+    {
+      "epoch": 0.08353684210526316,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004948321854516038,
+      "loss": 3.9091,
+      "step": 1984
+    },
+    {
+      "epoch": 0.08357894736842106,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0004948253575933832,
+      "loss": 3.9116,
+      "step": 1985
+    },
+    {
+      "epoch": 0.08362105263157894,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004948185252747323,
+      "loss": 4.0613,
+      "step": 1986
+    },
+    {
+      "epoch": 0.08366315789473684,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004948116884957754,
+      "loss": 3.6634,
+      "step": 1987
+    },
+    {
+      "epoch": 0.08370526315789474,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0004948048472566373,
+      "loss": 3.7137,
+      "step": 1988
+    },
+    {
+      "epoch": 0.08374736842105263,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0004947980015574425,
+      "loss": 3.6376,
+      "step": 1989
+    },
+    {
+      "epoch": 0.08378947368421052,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004947911513983156,
+      "loss": 3.6237,
+      "step": 1990
+    },
+    {
+      "epoch": 0.08383157894736842,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004947842967793817,
+      "loss": 3.8604,
+      "step": 1991
+    },
+    {
+      "epoch": 0.08387368421052632,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004947774377007656,
+      "loss": 3.8249,
+      "step": 1992
+    },
+    {
+      "epoch": 0.08391578947368421,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.000494770574162592,
+      "loss": 3.6279,
+      "step": 1993
+    },
+    {
+      "epoch": 0.0839578947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004947637061649861,
+      "loss": 3.6857,
+      "step": 1994
+    },
+    {
+      "epoch": 0.084,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0004947568337080732,
+      "loss": 3.928,
+      "step": 1995
+    },
+    {
+      "epoch": 0.08404210526315789,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004947499567919784,
+      "loss": 3.6396,
+      "step": 1996
+    },
+    {
+      "epoch": 0.08408421052631579,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004947430754168268,
+      "loss": 3.6607,
+      "step": 1997
+    },
+    {
+      "epoch": 0.08412631578947369,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.000494736189582744,
+      "loss": 3.5467,
+      "step": 1998
+    },
+    {
+      "epoch": 0.08416842105263157,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004947292992898554,
+      "loss": 3.7738,
+      "step": 1999
+    },
+    {
+      "epoch": 0.08421052631578947,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004947224045382865,
+      "loss": 4.0989,
+      "step": 2000
+    },
+    {
+      "epoch": 0.08425263157894737,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.000494715505328163,
+      "loss": 3.777,
+      "step": 2001
+    },
+    {
+      "epoch": 0.08429473684210527,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004947086016596104,
+      "loss": 3.3926,
+      "step": 2002
+    },
+    {
+      "epoch": 0.08433684210526315,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004947016935327546,
+      "loss": 3.0078,
+      "step": 2003
+    },
+    {
+      "epoch": 0.08437894736842105,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0004946947809477216,
+      "loss": 3.3413,
+      "step": 2004
+    },
+    {
+      "epoch": 0.08442105263157895,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.000494687863904637,
+      "loss": 3.5306,
+      "step": 2005
+    },
+    {
+      "epoch": 0.08446315789473684,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004946809424036272,
+      "loss": 3.9017,
+      "step": 2006
+    },
+    {
+      "epoch": 0.08450526315789474,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.000494674016444818,
+      "loss": 3.4367,
+      "step": 2007
+    },
+    {
+      "epoch": 0.08454736842105263,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0004946670860283358,
+      "loss": 4.0367,
+      "step": 2008
+    },
+    {
+      "epoch": 0.08458947368421052,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0004946601511543066,
+      "loss": 3.4446,
+      "step": 2009
+    },
+    {
+      "epoch": 0.08463157894736842,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004946532118228571,
+      "loss": 3.7409,
+      "step": 2010
+    },
+    {
+      "epoch": 0.08467368421052632,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004946462680341134,
+      "loss": 4.0729,
+      "step": 2011
+    },
+    {
+      "epoch": 0.08471578947368422,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004946393197882022,
+      "loss": 3.8701,
+      "step": 2012
+    },
+    {
+      "epoch": 0.0847578947368421,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00049463236708525,
+      "loss": 3.8745,
+      "step": 2013
+    },
+    {
+      "epoch": 0.0848,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004946254099253835,
+      "loss": 3.9885,
+      "step": 2014
+    },
+    {
+      "epoch": 0.0848421052631579,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004946184483087295,
+      "loss": 3.3153,
+      "step": 2015
+    },
+    {
+      "epoch": 0.08488421052631578,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004946114822354147,
+      "loss": 3.5487,
+      "step": 2016
+    },
+    {
+      "epoch": 0.08492631578947368,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004946045117055662,
+      "loss": 3.6788,
+      "step": 2017
+    },
+    {
+      "epoch": 0.08496842105263158,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004945975367193109,
+      "loss": 3.8843,
+      "step": 2018
+    },
+    {
+      "epoch": 0.08501052631578947,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004945905572767758,
+      "loss": 2.9441,
+      "step": 2019
+    },
+    {
+      "epoch": 0.08505263157894737,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004945835733780881,
+      "loss": 3.5928,
+      "step": 2020
+    },
+    {
+      "epoch": 0.08509473684210526,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0004945765850233752,
+      "loss": 3.7109,
+      "step": 2021
+    },
+    {
+      "epoch": 0.08513684210526316,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0004945695922127642,
+      "loss": 3.631,
+      "step": 2022
+    },
+    {
+      "epoch": 0.08517894736842105,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0004945625949463826,
+      "loss": 3.331,
+      "step": 2023
+    },
+    {
+      "epoch": 0.08522105263157895,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004945555932243578,
+      "loss": 3.6429,
+      "step": 2024
+    },
+    {
+      "epoch": 0.08526315789473685,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004945485870468176,
+      "loss": 3.4923,
+      "step": 2025
+    },
+    {
+      "epoch": 0.08530526315789473,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0004945415764138893,
+      "loss": 3.7182,
+      "step": 2026
+    },
+    {
+      "epoch": 0.08534736842105263,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.000494534561325701,
+      "loss": 3.9691,
+      "step": 2027
+    },
+    {
+      "epoch": 0.08538947368421053,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004945275417823802,
+      "loss": 3.2748,
+      "step": 2028
+    },
+    {
+      "epoch": 0.08543157894736843,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0004945205177840548,
+      "loss": 3.4739,
+      "step": 2029
+    },
+    {
+      "epoch": 0.08547368421052631,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.000494513489330853,
+      "loss": 3.5763,
+      "step": 2030
+    },
+    {
+      "epoch": 0.08551578947368421,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004945064564229026,
+      "loss": 3.5381,
+      "step": 2031
+    },
+    {
+      "epoch": 0.08555789473684211,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004944994190603319,
+      "loss": 3.1597,
+      "step": 2032
+    },
+    {
+      "epoch": 0.0856,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004944923772432691,
+      "loss": 3.36,
+      "step": 2033
+    },
+    {
+      "epoch": 0.0856421052631579,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004944853309718424,
+      "loss": 3.6844,
+      "step": 2034
+    },
+    {
+      "epoch": 0.0856842105263158,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004944782802461803,
+      "loss": 3.7089,
+      "step": 2035
+    },
+    {
+      "epoch": 0.08572631578947368,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004944712250664112,
+      "loss": 3.9458,
+      "step": 2036
+    },
+    {
+      "epoch": 0.08576842105263158,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004944641654326635,
+      "loss": 4.1766,
+      "step": 2037
+    },
+    {
+      "epoch": 0.08581052631578948,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.000494457101345066,
+      "loss": 3.8978,
+      "step": 2038
+    },
+    {
+      "epoch": 0.08585263157894737,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004944500328037474,
+      "loss": 3.4751,
+      "step": 2039
+    },
+    {
+      "epoch": 0.08589473684210526,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0004944429598088364,
+      "loss": 3.9218,
+      "step": 2040
+    },
+    {
+      "epoch": 0.08593684210526316,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0004944358823604617,
+      "loss": 3.6082,
+      "step": 2041
+    },
+    {
+      "epoch": 0.08597894736842106,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004944288004587527,
+      "loss": 3.7253,
+      "step": 2042
+    },
+    {
+      "epoch": 0.08602105263157894,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0004944217141038379,
+      "loss": 3.6628,
+      "step": 2043
+    },
+    {
+      "epoch": 0.08606315789473684,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004944146232958468,
+      "loss": 3.5219,
+      "step": 2044
+    },
+    {
+      "epoch": 0.08610526315789474,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004944075280349084,
+      "loss": 3.3887,
+      "step": 2045
+    },
+    {
+      "epoch": 0.08614736842105263,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.000494400428321152,
+      "loss": 3.8472,
+      "step": 2046
+    },
+    {
+      "epoch": 0.08618947368421052,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.000494393324154707,
+      "loss": 3.5136,
+      "step": 2047
+    },
+    {
+      "epoch": 0.08623157894736842,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004943862155357027,
+      "loss": 3.5147,
+      "step": 2048
+    },
+    {
+      "epoch": 0.08627368421052632,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004943791024642687,
+      "loss": 3.7999,
+      "step": 2049
+    },
+    {
+      "epoch": 0.0863157894736842,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004943719849405347,
+      "loss": 3.4143,
+      "step": 2050
+    },
+    {
+      "epoch": 0.0863578947368421,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004943648629646302,
+      "loss": 3.9442,
+      "step": 2051
+    },
+    {
+      "epoch": 0.0864,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004943577365366849,
+      "loss": 3.8247,
+      "step": 2052
+    },
+    {
+      "epoch": 0.08644210526315789,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004943506056568289,
+      "loss": 3.3835,
+      "step": 2053
+    },
+    {
+      "epoch": 0.08648421052631579,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004943434703251919,
+      "loss": 3.6517,
+      "step": 2054
+    },
+    {
+      "epoch": 0.08652631578947369,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.000494336330541904,
+      "loss": 3.9773,
+      "step": 2055
+    },
+    {
+      "epoch": 0.08656842105263157,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0004943291863070952,
+      "loss": 3.5868,
+      "step": 2056
+    },
+    {
+      "epoch": 0.08661052631578947,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0004943220376208957,
+      "loss": 3.5879,
+      "step": 2057
+    },
+    {
+      "epoch": 0.08665263157894737,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004943148844834359,
+      "loss": 3.6377,
+      "step": 2058
+    },
+    {
+      "epoch": 0.08669473684210527,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004943077268948457,
+      "loss": 3.6065,
+      "step": 2059
+    },
+    {
+      "epoch": 0.08673684210526315,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0004943005648552559,
+      "loss": 3.7017,
+      "step": 2060
+    },
+    {
+      "epoch": 0.08677894736842105,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004942933983647969,
+      "loss": 3.7522,
+      "step": 2061
+    },
+    {
+      "epoch": 0.08682105263157895,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0004942862274235991,
+      "loss": 3.1223,
+      "step": 2062
+    },
+    {
+      "epoch": 0.08686315789473684,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0004942790520317934,
+      "loss": 3.437,
+      "step": 2063
+    },
+    {
+      "epoch": 0.08690526315789474,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004942718721895102,
+      "loss": 3.9889,
+      "step": 2064
+    },
+    {
+      "epoch": 0.08694736842105263,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004942646878968806,
+      "loss": 3.6701,
+      "step": 2065
+    },
+    {
+      "epoch": 0.08698947368421053,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004942574991540354,
+      "loss": 3.6973,
+      "step": 2066
+    },
+    {
+      "epoch": 0.08703157894736842,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004942503059611055,
+      "loss": 3.7488,
+      "step": 2067
+    },
+    {
+      "epoch": 0.08707368421052632,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004942431083182221,
+      "loss": 3.9706,
+      "step": 2068
+    },
+    {
+      "epoch": 0.08711578947368422,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004942359062255162,
+      "loss": 3.8694,
+      "step": 2069
+    },
+    {
+      "epoch": 0.0871578947368421,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.000494228699683119,
+      "loss": 3.487,
+      "step": 2070
+    },
+    {
+      "epoch": 0.0872,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.000494221488691162,
+      "loss": 4.178,
+      "step": 2071
+    },
+    {
+      "epoch": 0.0872421052631579,
+      "grad_norm": 11.4375,
+      "learning_rate": 0.0004942142732497762,
+      "loss": 3.3172,
+      "step": 2072
+    },
+    {
+      "epoch": 0.08728421052631578,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0004942070533590935,
+      "loss": 3.5073,
+      "step": 2073
+    },
+    {
+      "epoch": 0.08732631578947368,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004941998290192451,
+      "loss": 3.7137,
+      "step": 2074
+    },
+    {
+      "epoch": 0.08736842105263158,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0004941926002303628,
+      "loss": 3.9667,
+      "step": 2075
+    },
+    {
+      "epoch": 0.08741052631578948,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004941853669925781,
+      "loss": 3.8093,
+      "step": 2076
+    },
+    {
+      "epoch": 0.08745263157894737,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.000494178129306023,
+      "loss": 3.6918,
+      "step": 2077
+    },
+    {
+      "epoch": 0.08749473684210526,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004941708871708294,
+      "loss": 3.5568,
+      "step": 2078
+    },
+    {
+      "epoch": 0.08753684210526316,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.000494163640587129,
+      "loss": 3.285,
+      "step": 2079
+    },
+    {
+      "epoch": 0.08757894736842105,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.000494156389555054,
+      "loss": 3.6124,
+      "step": 2080
+    },
+    {
+      "epoch": 0.08762105263157895,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0004941491340747364,
+      "loss": 3.7445,
+      "step": 2081
+    },
+    {
+      "epoch": 0.08766315789473685,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004941418741463084,
+      "loss": 3.584,
+      "step": 2082
+    },
+    {
+      "epoch": 0.08770526315789473,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0004941346097699023,
+      "loss": 3.3727,
+      "step": 2083
+    },
+    {
+      "epoch": 0.08774736842105263,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004941273409456506,
+      "loss": 3.5898,
+      "step": 2084
+    },
+    {
+      "epoch": 0.08778947368421053,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004941200676736855,
+      "loss": 3.6588,
+      "step": 2085
+    },
+    {
+      "epoch": 0.08783157894736843,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004941127899541396,
+      "loss": 3.5977,
+      "step": 2086
+    },
+    {
+      "epoch": 0.08787368421052631,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004941055077871454,
+      "loss": 3.3557,
+      "step": 2087
+    },
+    {
+      "epoch": 0.08791578947368421,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004940982211728359,
+      "loss": 3.4079,
+      "step": 2088
+    },
+    {
+      "epoch": 0.08795789473684211,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.0004940909301113434,
+      "loss": 3.249,
+      "step": 2089
+    },
+    {
+      "epoch": 0.088,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004940836346028011,
+      "loss": 2.947,
+      "step": 2090
+    },
+    {
+      "epoch": 0.0880421052631579,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0004940763346473416,
+      "loss": 3.5199,
+      "step": 2091
+    },
+    {
+      "epoch": 0.08808421052631579,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0004940690302450982,
+      "loss": 4.0699,
+      "step": 2092
+    },
+    {
+      "epoch": 0.08812631578947368,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0004940617213962038,
+      "loss": 3.3557,
+      "step": 2093
+    },
+    {
+      "epoch": 0.08816842105263158,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004940544081007915,
+      "loss": 3.7499,
+      "step": 2094
+    },
+    {
+      "epoch": 0.08821052631578948,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0004940470903589948,
+      "loss": 3.6581,
+      "step": 2095
+    },
+    {
+      "epoch": 0.08825263157894737,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0004940397681709467,
+      "loss": 3.5073,
+      "step": 2096
+    },
+    {
+      "epoch": 0.08829473684210526,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004940324415367808,
+      "loss": 3.5221,
+      "step": 2097
+    },
+    {
+      "epoch": 0.08833684210526316,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004940251104566306,
+      "loss": 3.4831,
+      "step": 2098
+    },
+    {
+      "epoch": 0.08837894736842106,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0004940177749306296,
+      "loss": 3.818,
+      "step": 2099
+    },
+    {
+      "epoch": 0.08842105263157894,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004940104349589114,
+      "loss": 3.8387,
+      "step": 2100
+    },
+    {
+      "epoch": 0.08846315789473684,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0004940030905416098,
+      "loss": 3.7556,
+      "step": 2101
+    },
+    {
+      "epoch": 0.08850526315789474,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004939957416788586,
+      "loss": 3.4967,
+      "step": 2102
+    },
+    {
+      "epoch": 0.08854736842105264,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004939883883707917,
+      "loss": 3.8585,
+      "step": 2103
+    },
+    {
+      "epoch": 0.08858947368421052,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.000493981030617543,
+      "loss": 3.7497,
+      "step": 2104
+    },
+    {
+      "epoch": 0.08863157894736842,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004939736684192466,
+      "loss": 3.755,
+      "step": 2105
+    },
+    {
+      "epoch": 0.08867368421052632,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004939663017760365,
+      "loss": 3.5746,
+      "step": 2106
+    },
+    {
+      "epoch": 0.0887157894736842,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0004939589306880472,
+      "loss": 3.689,
+      "step": 2107
+    },
+    {
+      "epoch": 0.0887578947368421,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004939515551554127,
+      "loss": 3.5345,
+      "step": 2108
+    },
+    {
+      "epoch": 0.0888,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004939441751782676,
+      "loss": 3.5887,
+      "step": 2109
+    },
+    {
+      "epoch": 0.08884210526315789,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0004939367907567462,
+      "loss": 3.1686,
+      "step": 2110
+    },
+    {
+      "epoch": 0.08888421052631579,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004939294018909831,
+      "loss": 3.5004,
+      "step": 2111
+    },
+    {
+      "epoch": 0.08892631578947369,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004939220085811129,
+      "loss": 3.7309,
+      "step": 2112
+    },
+    {
+      "epoch": 0.08896842105263159,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004939146108272703,
+      "loss": 3.5626,
+      "step": 2113
+    },
+    {
+      "epoch": 0.08901052631578947,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0004939072086295901,
+      "loss": 3.3308,
+      "step": 2114
+    },
+    {
+      "epoch": 0.08905263157894737,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004938998019882071,
+      "loss": 3.8362,
+      "step": 2115
+    },
+    {
+      "epoch": 0.08909473684210527,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0004938923909032562,
+      "loss": 3.6301,
+      "step": 2116
+    },
+    {
+      "epoch": 0.08913684210526315,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004938849753748727,
+      "loss": 3.7757,
+      "step": 2117
+    },
+    {
+      "epoch": 0.08917894736842105,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004938775554031913,
+      "loss": 3.7551,
+      "step": 2118
+    },
+    {
+      "epoch": 0.08922105263157895,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004938701309883474,
+      "loss": 3.7182,
+      "step": 2119
+    },
+    {
+      "epoch": 0.08926315789473684,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004938627021304764,
+      "loss": 4.178,
+      "step": 2120
+    },
+    {
+      "epoch": 0.08930526315789473,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004938552688297134,
+      "loss": 3.5459,
+      "step": 2121
+    },
+    {
+      "epoch": 0.08934736842105263,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.000493847831086194,
+      "loss": 3.6382,
+      "step": 2122
+    },
+    {
+      "epoch": 0.08938947368421053,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004938403889000535,
+      "loss": 3.8423,
+      "step": 2123
+    },
+    {
+      "epoch": 0.08943157894736842,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004938329422714277,
+      "loss": 3.4575,
+      "step": 2124
+    },
+    {
+      "epoch": 0.08947368421052632,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004938254912004522,
+      "loss": 3.7989,
+      "step": 2125
+    },
+    {
+      "epoch": 0.08951578947368422,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004938180356872626,
+      "loss": 3.4769,
+      "step": 2126
+    },
+    {
+      "epoch": 0.0895578947368421,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.000493810575731995,
+      "loss": 3.3955,
+      "step": 2127
+    },
+    {
+      "epoch": 0.0896,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004938031113347852,
+      "loss": 3.6091,
+      "step": 2128
+    },
+    {
+      "epoch": 0.0896421052631579,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.000493795642495769,
+      "loss": 3.6599,
+      "step": 2129
+    },
+    {
+      "epoch": 0.08968421052631578,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004937881692150827,
+      "loss": 3.7955,
+      "step": 2130
+    },
+    {
+      "epoch": 0.08972631578947368,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004937806914928625,
+      "loss": 3.1032,
+      "step": 2131
+    },
+    {
+      "epoch": 0.08976842105263158,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004937732093292446,
+      "loss": 3.592,
+      "step": 2132
+    },
+    {
+      "epoch": 0.08981052631578948,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.000493765722724365,
+      "loss": 4.0776,
+      "step": 2133
+    },
+    {
+      "epoch": 0.08985263157894736,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004937582316783604,
+      "loss": 3.6436,
+      "step": 2134
+    },
+    {
+      "epoch": 0.08989473684210526,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004937507361913674,
+      "loss": 3.7856,
+      "step": 2135
+    },
+    {
+      "epoch": 0.08993684210526316,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004937432362635222,
+      "loss": 3.6941,
+      "step": 2136
+    },
+    {
+      "epoch": 0.08997894736842105,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004937357318949617,
+      "loss": 3.5192,
+      "step": 2137
+    },
+    {
+      "epoch": 0.09002105263157895,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004937282230858225,
+      "loss": 3.4524,
+      "step": 2138
+    },
+    {
+      "epoch": 0.09006315789473684,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004937207098362416,
+      "loss": 4.1238,
+      "step": 2139
+    },
+    {
+      "epoch": 0.09010526315789473,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004937131921463555,
+      "loss": 3.8706,
+      "step": 2140
+    },
+    {
+      "epoch": 0.09014736842105263,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004937056700163015,
+      "loss": 3.6403,
+      "step": 2141
+    },
+    {
+      "epoch": 0.09018947368421053,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0004936981434462166,
+      "loss": 3.5454,
+      "step": 2142
+    },
+    {
+      "epoch": 0.09023157894736843,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0004936906124362378,
+      "loss": 2.9797,
+      "step": 2143
+    },
+    {
+      "epoch": 0.09027368421052631,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0004936830769865023,
+      "loss": 3.6078,
+      "step": 2144
+    },
+    {
+      "epoch": 0.09031578947368421,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004936755370971475,
+      "loss": 3.3638,
+      "step": 2145
+    },
+    {
+      "epoch": 0.09035789473684211,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0004936679927683107,
+      "loss": 3.2906,
+      "step": 2146
+    },
+    {
+      "epoch": 0.0904,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004936604440001295,
+      "loss": 3.4049,
+      "step": 2147
+    },
+    {
+      "epoch": 0.0904421052631579,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004936528907927411,
+      "loss": 3.6378,
+      "step": 2148
+    },
+    {
+      "epoch": 0.09048421052631579,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0004936453331462835,
+      "loss": 3.2355,
+      "step": 2149
+    },
+    {
+      "epoch": 0.09052631578947369,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004936377710608941,
+      "loss": 3.1031,
+      "step": 2150
+    },
+    {
+      "epoch": 0.09056842105263158,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004936302045367107,
+      "loss": 3.7966,
+      "step": 2151
+    },
+    {
+      "epoch": 0.09061052631578947,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004936226335738714,
+      "loss": 3.812,
+      "step": 2152
+    },
+    {
+      "epoch": 0.09065263157894737,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004936150581725139,
+      "loss": 3.615,
+      "step": 2153
+    },
+    {
+      "epoch": 0.09069473684210526,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004936074783327762,
+      "loss": 3.5712,
+      "step": 2154
+    },
+    {
+      "epoch": 0.09073684210526316,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004935998940547965,
+      "loss": 3.1613,
+      "step": 2155
+    },
+    {
+      "epoch": 0.09077894736842106,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004935923053387129,
+      "loss": 3.7646,
+      "step": 2156
+    },
+    {
+      "epoch": 0.09082105263157894,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004935847121846638,
+      "loss": 3.3472,
+      "step": 2157
+    },
+    {
+      "epoch": 0.09086315789473684,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004935771145927874,
+      "loss": 3.17,
+      "step": 2158
+    },
+    {
+      "epoch": 0.09090526315789474,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004935695125632222,
+      "loss": 3.614,
+      "step": 2159
+    },
+    {
+      "epoch": 0.09094736842105264,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004935619060961067,
+      "loss": 4.2274,
+      "step": 2160
+    },
+    {
+      "epoch": 0.09098947368421052,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004935542951915794,
+      "loss": 3.4428,
+      "step": 2161
+    },
+    {
+      "epoch": 0.09103157894736842,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.000493546679849779,
+      "loss": 3.5008,
+      "step": 2162
+    },
+    {
+      "epoch": 0.09107368421052632,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004935390600708442,
+      "loss": 3.7241,
+      "step": 2163
+    },
+    {
+      "epoch": 0.0911157894736842,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004935314358549139,
+      "loss": 3.148,
+      "step": 2164
+    },
+    {
+      "epoch": 0.0911578947368421,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.000493523807202127,
+      "loss": 3.4836,
+      "step": 2165
+    },
+    {
+      "epoch": 0.0912,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004935161741126225,
+      "loss": 3.4649,
+      "step": 2166
+    },
+    {
+      "epoch": 0.09124210526315789,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004935085365865393,
+      "loss": 3.8163,
+      "step": 2167
+    },
+    {
+      "epoch": 0.09128421052631579,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004935008946240168,
+      "loss": 3.69,
+      "step": 2168
+    },
+    {
+      "epoch": 0.09132631578947369,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.000493493248225194,
+      "loss": 3.6096,
+      "step": 2169
+    },
+    {
+      "epoch": 0.09136842105263158,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0004934855973902105,
+      "loss": 3.4264,
+      "step": 2170
+    },
+    {
+      "epoch": 0.09141052631578947,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004934779421192053,
+      "loss": 3.6218,
+      "step": 2171
+    },
+    {
+      "epoch": 0.09145263157894737,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0004934702824123182,
+      "loss": 3.8792,
+      "step": 2172
+    },
+    {
+      "epoch": 0.09149473684210527,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004934626182696886,
+      "loss": 3.2383,
+      "step": 2173
+    },
+    {
+      "epoch": 0.09153684210526315,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.000493454949691456,
+      "loss": 3.7411,
+      "step": 2174
+    },
+    {
+      "epoch": 0.09157894736842105,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0004934472766777604,
+      "loss": 4.013,
+      "step": 2175
+    },
+    {
+      "epoch": 0.09162105263157895,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004934395992287414,
+      "loss": 3.7162,
+      "step": 2176
+    },
+    {
+      "epoch": 0.09166315789473684,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.000493431917344539,
+      "loss": 3.3383,
+      "step": 2177
+    },
+    {
+      "epoch": 0.09170526315789473,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004934242310252931,
+      "loss": 3.5431,
+      "step": 2178
+    },
+    {
+      "epoch": 0.09174736842105263,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004934165402711436,
+      "loss": 3.6037,
+      "step": 2179
+    },
+    {
+      "epoch": 0.09178947368421053,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004934088450822308,
+      "loss": 3.4008,
+      "step": 2180
+    },
+    {
+      "epoch": 0.09183157894736842,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004934011454586949,
+      "loss": 3.8943,
+      "step": 2181
+    },
+    {
+      "epoch": 0.09187368421052632,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.000493393441400676,
+      "loss": 4.0653,
+      "step": 2182
+    },
+    {
+      "epoch": 0.09191578947368421,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0004933857329083147,
+      "loss": 3.7465,
+      "step": 2183
+    },
+    {
+      "epoch": 0.0919578947368421,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004933780199817512,
+      "loss": 3.7839,
+      "step": 2184
+    },
+    {
+      "epoch": 0.092,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0004933703026211262,
+      "loss": 3.198,
+      "step": 2185
+    },
+    {
+      "epoch": 0.0920421052631579,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0004933625808265802,
+      "loss": 3.4851,
+      "step": 2186
+    },
+    {
+      "epoch": 0.0920842105263158,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004933548545982539,
+      "loss": 3.4378,
+      "step": 2187
+    },
+    {
+      "epoch": 0.09212631578947368,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004933471239362881,
+      "loss": 3.6771,
+      "step": 2188
+    },
+    {
+      "epoch": 0.09216842105263158,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004933393888408237,
+      "loss": 3.4539,
+      "step": 2189
+    },
+    {
+      "epoch": 0.09221052631578948,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004933316493120015,
+      "loss": 3.6512,
+      "step": 2190
+    },
+    {
+      "epoch": 0.09225263157894736,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004933239053499625,
+      "loss": 3.5342,
+      "step": 2191
+    },
+    {
+      "epoch": 0.09229473684210526,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004933161569548479,
+      "loss": 3.9198,
+      "step": 2192
+    },
+    {
+      "epoch": 0.09233684210526316,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004933084041267988,
+      "loss": 3.8718,
+      "step": 2193
+    },
+    {
+      "epoch": 0.09237894736842105,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004933006468659564,
+      "loss": 3.8724,
+      "step": 2194
+    },
+    {
+      "epoch": 0.09242105263157895,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004932928851724621,
+      "loss": 3.4539,
+      "step": 2195
+    },
+    {
+      "epoch": 0.09246315789473684,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004932851190464573,
+      "loss": 3.4672,
+      "step": 2196
+    },
+    {
+      "epoch": 0.09250526315789474,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004932773484880835,
+      "loss": 3.6806,
+      "step": 2197
+    },
+    {
+      "epoch": 0.09254736842105263,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004932695734974822,
+      "loss": 3.746,
+      "step": 2198
+    },
+    {
+      "epoch": 0.09258947368421053,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.0004932617940747951,
+      "loss": 3.318,
+      "step": 2199
+    },
+    {
+      "epoch": 0.09263157894736843,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004932540102201641,
+      "loss": 3.4834,
+      "step": 2200
+    },
+    {
+      "epoch": 0.09267368421052631,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0004932462219337305,
+      "loss": 3.6298,
+      "step": 2201
+    },
+    {
+      "epoch": 0.09271578947368421,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0004932384292156369,
+      "loss": 3.8754,
+      "step": 2202
+    },
+    {
+      "epoch": 0.09275789473684211,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0004932306320660247,
+      "loss": 3.4735,
+      "step": 2203
+    },
+    {
+      "epoch": 0.0928,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004932228304850363,
+      "loss": 3.9172,
+      "step": 2204
+    },
+    {
+      "epoch": 0.09284210526315789,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004932150244728136,
+      "loss": 3.7408,
+      "step": 2205
+    },
+    {
+      "epoch": 0.09288421052631579,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0004932072140294989,
+      "loss": 3.5298,
+      "step": 2206
+    },
+    {
+      "epoch": 0.09292631578947369,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004931993991552346,
+      "loss": 3.4128,
+      "step": 2207
+    },
+    {
+      "epoch": 0.09296842105263158,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0004931915798501629,
+      "loss": 3.6295,
+      "step": 2208
+    },
+    {
+      "epoch": 0.09301052631578947,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004931837561144264,
+      "loss": 4.0245,
+      "step": 2209
+    },
+    {
+      "epoch": 0.09305263157894737,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004931759279481676,
+      "loss": 3.4963,
+      "step": 2210
+    },
+    {
+      "epoch": 0.09309473684210526,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004931680953515292,
+      "loss": 3.6967,
+      "step": 2211
+    },
+    {
+      "epoch": 0.09313684210526316,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004931602583246537,
+      "loss": 3.6086,
+      "step": 2212
+    },
+    {
+      "epoch": 0.09317894736842106,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0004931524168676841,
+      "loss": 3.4076,
+      "step": 2213
+    },
+    {
+      "epoch": 0.09322105263157894,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.000493144570980763,
+      "loss": 3.6261,
+      "step": 2214
+    },
+    {
+      "epoch": 0.09326315789473684,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004931367206640337,
+      "loss": 3.3957,
+      "step": 2215
+    },
+    {
+      "epoch": 0.09330526315789474,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0004931288659176388,
+      "loss": 3.3958,
+      "step": 2216
+    },
+    {
+      "epoch": 0.09334736842105264,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004931210067417218,
+      "loss": 3.6099,
+      "step": 2217
+    },
+    {
+      "epoch": 0.09338947368421052,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004931131431364257,
+      "loss": 3.6716,
+      "step": 2218
+    },
+    {
+      "epoch": 0.09343157894736842,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004931052751018937,
+      "loss": 3.8295,
+      "step": 2219
+    },
+    {
+      "epoch": 0.09347368421052632,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0004930974026382693,
+      "loss": 3.9262,
+      "step": 2220
+    },
+    {
+      "epoch": 0.0935157894736842,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004930895257456957,
+      "loss": 3.3495,
+      "step": 2221
+    },
+    {
+      "epoch": 0.0935578947368421,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0004930816444243167,
+      "loss": 3.7346,
+      "step": 2222
+    },
+    {
+      "epoch": 0.0936,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004930737586742758,
+      "loss": 3.5685,
+      "step": 2223
+    },
+    {
+      "epoch": 0.0936421052631579,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0004930658684957164,
+      "loss": 3.3972,
+      "step": 2224
+    },
+    {
+      "epoch": 0.09368421052631579,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0004930579738887826,
+      "loss": 3.2524,
+      "step": 2225
+    },
+    {
+      "epoch": 0.09372631578947369,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004930500748536182,
+      "loss": 3.6159,
+      "step": 2226
+    },
+    {
+      "epoch": 0.09376842105263158,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004930421713903668,
+      "loss": 3.5875,
+      "step": 2227
+    },
+    {
+      "epoch": 0.09381052631578947,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0004930342634991727,
+      "loss": 3.1611,
+      "step": 2228
+    },
+    {
+      "epoch": 0.09385263157894737,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0004930263511801799,
+      "loss": 3.639,
+      "step": 2229
+    },
+    {
+      "epoch": 0.09389473684210527,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0004930184344335326,
+      "loss": 3.1428,
+      "step": 2230
+    },
+    {
+      "epoch": 0.09393684210526315,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004930105132593748,
+      "loss": 3.7563,
+      "step": 2231
+    },
+    {
+      "epoch": 0.09397894736842105,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.000493002587657851,
+      "loss": 3.37,
+      "step": 2232
+    },
+    {
+      "epoch": 0.09402105263157895,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004929946576291056,
+      "loss": 3.644,
+      "step": 2233
+    },
+    {
+      "epoch": 0.09406315789473685,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.000492986723173283,
+      "loss": 3.8953,
+      "step": 2234
+    },
+    {
+      "epoch": 0.09410526315789473,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0004929787842905278,
+      "loss": 3.7189,
+      "step": 2235
+    },
+    {
+      "epoch": 0.09414736842105263,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004929708409809847,
+      "loss": 3.628,
+      "step": 2236
+    },
+    {
+      "epoch": 0.09418947368421053,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0004929628932447984,
+      "loss": 3.9377,
+      "step": 2237
+    },
+    {
+      "epoch": 0.09423157894736842,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004929549410821136,
+      "loss": 3.6712,
+      "step": 2238
+    },
+    {
+      "epoch": 0.09427368421052632,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0004929469844930752,
+      "loss": 3.3246,
+      "step": 2239
+    },
+    {
+      "epoch": 0.09431578947368421,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004929390234778283,
+      "loss": 3.7811,
+      "step": 2240
+    },
+    {
+      "epoch": 0.0943578947368421,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004929310580365178,
+      "loss": 3.3557,
+      "step": 2241
+    },
+    {
+      "epoch": 0.0944,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004929230881692887,
+      "loss": 3.7594,
+      "step": 2242
+    },
+    {
+      "epoch": 0.0944421052631579,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0004929151138762866,
+      "loss": 3.636,
+      "step": 2243
+    },
+    {
+      "epoch": 0.0944842105263158,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004929071351576564,
+      "loss": 3.7785,
+      "step": 2244
+    },
+    {
+      "epoch": 0.09452631578947368,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004928991520135436,
+      "loss": 3.8262,
+      "step": 2245
+    },
+    {
+      "epoch": 0.09456842105263158,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0004928911644440936,
+      "loss": 3.4451,
+      "step": 2246
+    },
+    {
+      "epoch": 0.09461052631578948,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.000492883172449452,
+      "loss": 3.7754,
+      "step": 2247
+    },
+    {
+      "epoch": 0.09465263157894736,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004928751760297645,
+      "loss": 3.327,
+      "step": 2248
+    },
+    {
+      "epoch": 0.09469473684210526,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004928671751851765,
+      "loss": 3.4821,
+      "step": 2249
+    },
+    {
+      "epoch": 0.09473684210526316,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.000492859169915834,
+      "loss": 3.619,
+      "step": 2250
+    },
+    {
+      "epoch": 0.09477894736842105,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0004928511602218827,
+      "loss": 3.7875,
+      "step": 2251
+    },
+    {
+      "epoch": 0.09482105263157894,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004928431461034687,
+      "loss": 3.6277,
+      "step": 2252
+    },
+    {
+      "epoch": 0.09486315789473684,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004928351275607379,
+      "loss": 3.2966,
+      "step": 2253
+    },
+    {
+      "epoch": 0.09490526315789474,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004928271045938363,
+      "loss": 3.3082,
+      "step": 2254
+    },
+    {
+      "epoch": 0.09494736842105263,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004928190772029102,
+      "loss": 3.9054,
+      "step": 2255
+    },
+    {
+      "epoch": 0.09498947368421053,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004928110453881059,
+      "loss": 3.4825,
+      "step": 2256
+    },
+    {
+      "epoch": 0.09503157894736843,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004928030091495695,
+      "loss": 3.6731,
+      "step": 2257
+    },
+    {
+      "epoch": 0.09507368421052631,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004927949684874477,
+      "loss": 3.6112,
+      "step": 2258
+    },
+    {
+      "epoch": 0.09511578947368421,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0004927869234018868,
+      "loss": 3.0327,
+      "step": 2259
+    },
+    {
+      "epoch": 0.09515789473684211,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004927788738930334,
+      "loss": 3.6778,
+      "step": 2260
+    },
+    {
+      "epoch": 0.0952,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004927708199610341,
+      "loss": 3.5951,
+      "step": 2261
+    },
+    {
+      "epoch": 0.09524210526315789,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004927627616060358,
+      "loss": 3.2553,
+      "step": 2262
+    },
+    {
+      "epoch": 0.09528421052631579,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0004927546988281853,
+      "loss": 3.3389,
+      "step": 2263
+    },
+    {
+      "epoch": 0.09532631578947369,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004927466316276292,
+      "loss": 3.8305,
+      "step": 2264
+    },
+    {
+      "epoch": 0.09536842105263157,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004927385600045148,
+      "loss": 3.824,
+      "step": 2265
+    },
+    {
+      "epoch": 0.09541052631578947,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.000492730483958989,
+      "loss": 3.4389,
+      "step": 2266
+    },
+    {
+      "epoch": 0.09545263157894737,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.000492722403491199,
+      "loss": 3.6056,
+      "step": 2267
+    },
+    {
+      "epoch": 0.09549473684210526,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004927143186012922,
+      "loss": 3.8661,
+      "step": 2268
+    },
+    {
+      "epoch": 0.09553684210526316,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004927062292894153,
+      "loss": 3.6917,
+      "step": 2269
+    },
+    {
+      "epoch": 0.09557894736842105,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004926981355557164,
+      "loss": 3.6152,
+      "step": 2270
+    },
+    {
+      "epoch": 0.09562105263157895,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004926900374003424,
+      "loss": 3.4671,
+      "step": 2271
+    },
+    {
+      "epoch": 0.09566315789473684,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004926819348234412,
+      "loss": 3.482,
+      "step": 2272
+    },
+    {
+      "epoch": 0.09570526315789474,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004926738278251603,
+      "loss": 3.4053,
+      "step": 2273
+    },
+    {
+      "epoch": 0.09574736842105264,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004926657164056474,
+      "loss": 3.6836,
+      "step": 2274
+    },
+    {
+      "epoch": 0.09578947368421052,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004926576005650501,
+      "loss": 3.472,
+      "step": 2275
+    },
+    {
+      "epoch": 0.09583157894736842,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004926494803035166,
+      "loss": 3.4592,
+      "step": 2276
+    },
+    {
+      "epoch": 0.09587368421052632,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004926413556211946,
+      "loss": 3.8507,
+      "step": 2277
+    },
+    {
+      "epoch": 0.0959157894736842,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004926332265182321,
+      "loss": 3.8539,
+      "step": 2278
+    },
+    {
+      "epoch": 0.0959578947368421,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004926250929947776,
+      "loss": 3.2598,
+      "step": 2279
+    },
+    {
+      "epoch": 0.096,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004926169550509787,
+      "loss": 3.3632,
+      "step": 2280
+    },
+    {
+      "epoch": 0.0960421052631579,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.000492608812686984,
+      "loss": 3.3726,
+      "step": 2281
+    },
+    {
+      "epoch": 0.09608421052631579,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004926006659029419,
+      "loss": 3.678,
+      "step": 2282
+    },
+    {
+      "epoch": 0.09612631578947368,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004925925146990005,
+      "loss": 3.8254,
+      "step": 2283
+    },
+    {
+      "epoch": 0.09616842105263158,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004925843590753087,
+      "loss": 4.0319,
+      "step": 2284
+    },
+    {
+      "epoch": 0.09621052631578947,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004925761990320148,
+      "loss": 3.8538,
+      "step": 2285
+    },
+    {
+      "epoch": 0.09625263157894737,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004925680345692676,
+      "loss": 3.7418,
+      "step": 2286
+    },
+    {
+      "epoch": 0.09629473684210527,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004925598656872158,
+      "loss": 3.5188,
+      "step": 2287
+    },
+    {
+      "epoch": 0.09633684210526315,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0004925516923860083,
+      "loss": 3.0186,
+      "step": 2288
+    },
+    {
+      "epoch": 0.09637894736842105,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0004925435146657939,
+      "loss": 3.8819,
+      "step": 2289
+    },
+    {
+      "epoch": 0.09642105263157895,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0004925353325267217,
+      "loss": 3.1003,
+      "step": 2290
+    },
+    {
+      "epoch": 0.09646315789473685,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0004925271459689406,
+      "loss": 3.5093,
+      "step": 2291
+    },
+    {
+      "epoch": 0.09650526315789473,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004925189549925999,
+      "loss": 3.5151,
+      "step": 2292
+    },
+    {
+      "epoch": 0.09654736842105263,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0004925107595978489,
+      "loss": 3.4106,
+      "step": 2293
+    },
+    {
+      "epoch": 0.09658947368421053,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004925025597848367,
+      "loss": 3.8088,
+      "step": 2294
+    },
+    {
+      "epoch": 0.09663157894736842,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004924943555537128,
+      "loss": 4.0872,
+      "step": 2295
+    },
+    {
+      "epoch": 0.09667368421052631,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004924861469046266,
+      "loss": 3.5939,
+      "step": 2296
+    },
+    {
+      "epoch": 0.09671578947368421,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004924779338377278,
+      "loss": 3.5667,
+      "step": 2297
+    },
+    {
+      "epoch": 0.09675789473684211,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0004924697163531658,
+      "loss": 3.5407,
+      "step": 2298
+    },
+    {
+      "epoch": 0.0968,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004924614944510907,
+      "loss": 3.7918,
+      "step": 2299
+    },
+    {
+      "epoch": 0.0968421052631579,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004924532681316519,
+      "loss": 3.1883,
+      "step": 2300
+    },
+    {
+      "epoch": 0.0968842105263158,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004924450373949994,
+      "loss": 3.9175,
+      "step": 2301
+    },
+    {
+      "epoch": 0.09692631578947368,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004924368022412832,
+      "loss": 3.8284,
+      "step": 2302
+    },
+    {
+      "epoch": 0.09696842105263158,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004924285626706534,
+      "loss": 3.4134,
+      "step": 2303
+    },
+    {
+      "epoch": 0.09701052631578948,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004924203186832599,
+      "loss": 3.6822,
+      "step": 2304
+    },
+    {
+      "epoch": 0.09705263157894736,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004924120702792531,
+      "loss": 3.8275,
+      "step": 2305
+    },
+    {
+      "epoch": 0.09709473684210526,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004924038174587833,
+      "loss": 3.7106,
+      "step": 2306
+    },
+    {
+      "epoch": 0.09713684210526316,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004923955602220007,
+      "loss": 3.8975,
+      "step": 2307
+    },
+    {
+      "epoch": 0.09717894736842106,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004923872985690557,
+      "loss": 3.6689,
+      "step": 2308
+    },
+    {
+      "epoch": 0.09722105263157894,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.000492379032500099,
+      "loss": 3.1689,
+      "step": 2309
+    },
+    {
+      "epoch": 0.09726315789473684,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0004923707620152811,
+      "loss": 3.555,
+      "step": 2310
+    },
+    {
+      "epoch": 0.09730526315789474,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004923624871147526,
+      "loss": 3.2337,
+      "step": 2311
+    },
+    {
+      "epoch": 0.09734736842105263,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004923542077986644,
+      "loss": 3.4281,
+      "step": 2312
+    },
+    {
+      "epoch": 0.09738947368421053,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004923459240671673,
+      "loss": 3.3522,
+      "step": 2313
+    },
+    {
+      "epoch": 0.09743157894736842,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004923376359204123,
+      "loss": 3.7305,
+      "step": 2314
+    },
+    {
+      "epoch": 0.09747368421052631,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0004923293433585503,
+      "loss": 3.1243,
+      "step": 2315
+    },
+    {
+      "epoch": 0.09751578947368421,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004923210463817322,
+      "loss": 3.749,
+      "step": 2316
+    },
+    {
+      "epoch": 0.09755789473684211,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004923127449901095,
+      "loss": 3.7855,
+      "step": 2317
+    },
+    {
+      "epoch": 0.0976,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004923044391838332,
+      "loss": 3.8114,
+      "step": 2318
+    },
+    {
+      "epoch": 0.09764210526315789,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004922961289630549,
+      "loss": 3.3494,
+      "step": 2319
+    },
+    {
+      "epoch": 0.09768421052631579,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004922878143279256,
+      "loss": 3.6752,
+      "step": 2320
+    },
+    {
+      "epoch": 0.09772631578947369,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004922794952785972,
+      "loss": 3.9347,
+      "step": 2321
+    },
+    {
+      "epoch": 0.09776842105263157,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004922711718152209,
+      "loss": 3.7123,
+      "step": 2322
+    },
+    {
+      "epoch": 0.09781052631578947,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0004922628439379485,
+      "loss": 3.6127,
+      "step": 2323
+    },
+    {
+      "epoch": 0.09785263157894737,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004922545116469318,
+      "loss": 3.7444,
+      "step": 2324
+    },
+    {
+      "epoch": 0.09789473684210526,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004922461749423226,
+      "loss": 3.5492,
+      "step": 2325
+    },
+    {
+      "epoch": 0.09793684210526316,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004922378338242725,
+      "loss": 3.0203,
+      "step": 2326
+    },
+    {
+      "epoch": 0.09797894736842105,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004922294882929339,
+      "loss": 3.1889,
+      "step": 2327
+    },
+    {
+      "epoch": 0.09802105263157895,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0004922211383484586,
+      "loss": 3.7574,
+      "step": 2328
+    },
+    {
+      "epoch": 0.09806315789473684,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004922127839909986,
+      "loss": 3.5968,
+      "step": 2329
+    },
+    {
+      "epoch": 0.09810526315789474,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004922044252207065,
+      "loss": 3.6444,
+      "step": 2330
+    },
+    {
+      "epoch": 0.09814736842105264,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004921960620377343,
+      "loss": 3.5392,
+      "step": 2331
+    },
+    {
+      "epoch": 0.09818947368421052,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004921876944422342,
+      "loss": 2.9158,
+      "step": 2332
+    },
+    {
+      "epoch": 0.09823157894736842,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0004921793224343591,
+      "loss": 3.4225,
+      "step": 2333
+    },
+    {
+      "epoch": 0.09827368421052632,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004921709460142612,
+      "loss": 3.5111,
+      "step": 2334
+    },
+    {
+      "epoch": 0.0983157894736842,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0004921625651820933,
+      "loss": 3.7589,
+      "step": 2335
+    },
+    {
+      "epoch": 0.0983578947368421,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004921541799380078,
+      "loss": 3.2662,
+      "step": 2336
+    },
+    {
+      "epoch": 0.0984,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004921457902821578,
+      "loss": 3.5344,
+      "step": 2337
+    },
+    {
+      "epoch": 0.0984421052631579,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.000492137396214696,
+      "loss": 3.3708,
+      "step": 2338
+    },
+    {
+      "epoch": 0.09848421052631579,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004921289977357753,
+      "loss": 3.5139,
+      "step": 2339
+    },
+    {
+      "epoch": 0.09852631578947368,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004921205948455488,
+      "loss": 3.6257,
+      "step": 2340
+    },
+    {
+      "epoch": 0.09856842105263158,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0004921121875441695,
+      "loss": 3.9827,
+      "step": 2341
+    },
+    {
+      "epoch": 0.09861052631578947,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004921037758317905,
+      "loss": 3.6639,
+      "step": 2342
+    },
+    {
+      "epoch": 0.09865263157894737,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004920953597085654,
+      "loss": 3.5262,
+      "step": 2343
+    },
+    {
+      "epoch": 0.09869473684210527,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0004920869391746472,
+      "loss": 4.0008,
+      "step": 2344
+    },
+    {
+      "epoch": 0.09873684210526316,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004920785142301893,
+      "loss": 3.1332,
+      "step": 2345
+    },
+    {
+      "epoch": 0.09877894736842105,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004920700848753454,
+      "loss": 3.6987,
+      "step": 2346
+    },
+    {
+      "epoch": 0.09882105263157895,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.000492061651110269,
+      "loss": 3.0578,
+      "step": 2347
+    },
+    {
+      "epoch": 0.09886315789473685,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004920532129351138,
+      "loss": 3.6252,
+      "step": 2348
+    },
+    {
+      "epoch": 0.09890526315789473,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0004920447703500333,
+      "loss": 3.3714,
+      "step": 2349
+    },
+    {
+      "epoch": 0.09894736842105263,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004920363233551816,
+      "loss": 3.0817,
+      "step": 2350
+    },
+    {
+      "epoch": 0.09898947368421053,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004920278719507123,
+      "loss": 3.938,
+      "step": 2351
+    },
+    {
+      "epoch": 0.09903157894736841,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0004920194161367797,
+      "loss": 3.577,
+      "step": 2352
+    },
+    {
+      "epoch": 0.09907368421052631,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004920109559135376,
+      "loss": 3.6999,
+      "step": 2353
+    },
+    {
+      "epoch": 0.09911578947368421,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004920024912811403,
+      "loss": 4.1408,
+      "step": 2354
+    },
+    {
+      "epoch": 0.09915789473684211,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0004919940222397419,
+      "loss": 3.2699,
+      "step": 2355
+    },
+    {
+      "epoch": 0.0992,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0004919855487894968,
+      "loss": 3.665,
+      "step": 2356
+    },
+    {
+      "epoch": 0.0992421052631579,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004919770709305593,
+      "loss": 3.7943,
+      "step": 2357
+    },
+    {
+      "epoch": 0.0992842105263158,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.000491968588663084,
+      "loss": 3.3793,
+      "step": 2358
+    },
+    {
+      "epoch": 0.09932631578947368,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.000491960101987225,
+      "loss": 3.5438,
+      "step": 2359
+    },
+    {
+      "epoch": 0.09936842105263158,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004919516109031375,
+      "loss": 4.0149,
+      "step": 2360
+    },
+    {
+      "epoch": 0.09941052631578948,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0004919431154109758,
+      "loss": 3.5727,
+      "step": 2361
+    },
+    {
+      "epoch": 0.09945263157894736,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004919346155108949,
+      "loss": 3.3026,
+      "step": 2362
+    },
+    {
+      "epoch": 0.09949473684210526,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004919261112030495,
+      "loss": 3.7856,
+      "step": 2363
+    },
+    {
+      "epoch": 0.09953684210526316,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0004919176024875947,
+      "loss": 3.9578,
+      "step": 2364
+    },
+    {
+      "epoch": 0.09957894736842106,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004919090893646854,
+      "loss": 3.8835,
+      "step": 2365
+    },
+    {
+      "epoch": 0.09962105263157894,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004919005718344767,
+      "loss": 3.2881,
+      "step": 2366
+    },
+    {
+      "epoch": 0.09966315789473684,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0004918920498971238,
+      "loss": 3.4665,
+      "step": 2367
+    },
+    {
+      "epoch": 0.09970526315789474,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004918835235527819,
+      "loss": 3.7954,
+      "step": 2368
+    },
+    {
+      "epoch": 0.09974736842105263,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004918749928016064,
+      "loss": 3.0782,
+      "step": 2369
+    },
+    {
+      "epoch": 0.09978947368421053,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0004918664576437528,
+      "loss": 3.4417,
+      "step": 2370
+    },
+    {
+      "epoch": 0.09983157894736842,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004918579180793765,
+      "loss": 3.1913,
+      "step": 2371
+    },
+    {
+      "epoch": 0.09987368421052631,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004918493741086331,
+      "loss": 3.7737,
+      "step": 2372
+    },
+    {
+      "epoch": 0.09991578947368421,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004918408257316784,
+      "loss": 3.4303,
+      "step": 2373
+    },
+    {
+      "epoch": 0.0999578947368421,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004918322729486677,
+      "loss": 3.4441,
+      "step": 2374
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004918237157597574,
+      "loss": 3.574,
+      "step": 2375
+    },
+    {
+      "epoch": 0.10004210526315789,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.000491815154165103,
+      "loss": 3.3758,
+      "step": 2376
+    },
+    {
+      "epoch": 0.10008421052631579,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004918065881648607,
+      "loss": 3.5957,
+      "step": 2377
+    },
+    {
+      "epoch": 0.10012631578947369,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004917980177591864,
+      "loss": 3.654,
+      "step": 2378
+    },
+    {
+      "epoch": 0.10016842105263157,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004917894429482363,
+      "loss": 3.5593,
+      "step": 2379
+    },
+    {
+      "epoch": 0.10021052631578947,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004917808637321667,
+      "loss": 3.145,
+      "step": 2380
+    },
+    {
+      "epoch": 0.10025263157894737,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004917722801111337,
+      "loss": 3.5548,
+      "step": 2381
+    },
+    {
+      "epoch": 0.10029473684210527,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.000491763692085294,
+      "loss": 2.8117,
+      "step": 2382
+    },
+    {
+      "epoch": 0.10033684210526315,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004917550996548039,
+      "loss": 4.078,
+      "step": 2383
+    },
+    {
+      "epoch": 0.10037894736842105,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004917465028198199,
+      "loss": 3.5865,
+      "step": 2384
+    },
+    {
+      "epoch": 0.10042105263157895,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004917379015804985,
+      "loss": 3.3883,
+      "step": 2385
+    },
+    {
+      "epoch": 0.10046315789473684,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004917292959369968,
+      "loss": 3.7073,
+      "step": 2386
+    },
+    {
+      "epoch": 0.10050526315789474,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004917206858894713,
+      "loss": 3.8791,
+      "step": 2387
+    },
+    {
+      "epoch": 0.10054736842105264,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004917120714380788,
+      "loss": 3.6739,
+      "step": 2388
+    },
+    {
+      "epoch": 0.10058947368421052,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004917034525829764,
+      "loss": 3.7219,
+      "step": 2389
+    },
+    {
+      "epoch": 0.10063157894736842,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004916948293243211,
+      "loss": 3.509,
+      "step": 2390
+    },
+    {
+      "epoch": 0.10067368421052632,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00049168620166227,
+      "loss": 3.5365,
+      "step": 2391
+    },
+    {
+      "epoch": 0.10071578947368422,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004916775695969803,
+      "loss": 3.7747,
+      "step": 2392
+    },
+    {
+      "epoch": 0.1007578947368421,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004916689331286091,
+      "loss": 4.096,
+      "step": 2393
+    },
+    {
+      "epoch": 0.1008,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.000491660292257314,
+      "loss": 3.8376,
+      "step": 2394
+    },
+    {
+      "epoch": 0.1008421052631579,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004916516469832524,
+      "loss": 3.5274,
+      "step": 2395
+    },
+    {
+      "epoch": 0.10088421052631578,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004916429973065815,
+      "loss": 3.6971,
+      "step": 2396
+    },
+    {
+      "epoch": 0.10092631578947368,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004916343432274594,
+      "loss": 3.5289,
+      "step": 2397
+    },
+    {
+      "epoch": 0.10096842105263158,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004916256847460433,
+      "loss": 3.685,
+      "step": 2398
+    },
+    {
+      "epoch": 0.10101052631578947,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004916170218624911,
+      "loss": 3.7434,
+      "step": 2399
+    },
+    {
+      "epoch": 0.10105263157894737,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004916083545769607,
+      "loss": 3.5422,
+      "step": 2400
+    },
+    {
+      "epoch": 0.10109473684210527,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004915996828896099,
+      "loss": 3.3303,
+      "step": 2401
+    },
+    {
+      "epoch": 0.10113684210526316,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0004915910068005968,
+      "loss": 3.482,
+      "step": 2402
+    },
+    {
+      "epoch": 0.10117894736842105,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004915823263100794,
+      "loss": 3.6612,
+      "step": 2403
+    },
+    {
+      "epoch": 0.10122105263157895,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004915736414182159,
+      "loss": 3.6481,
+      "step": 2404
+    },
+    {
+      "epoch": 0.10126315789473685,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004915649521251644,
+      "loss": 3.3203,
+      "step": 2405
+    },
+    {
+      "epoch": 0.10130526315789473,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0004915562584310834,
+      "loss": 3.3243,
+      "step": 2406
+    },
+    {
+      "epoch": 0.10134736842105263,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004915475603361312,
+      "loss": 3.9871,
+      "step": 2407
+    },
+    {
+      "epoch": 0.10138947368421053,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0004915388578404661,
+      "loss": 3.5335,
+      "step": 2408
+    },
+    {
+      "epoch": 0.10143157894736841,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004915301509442469,
+      "loss": 3.9484,
+      "step": 2409
+    },
+    {
+      "epoch": 0.10147368421052631,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0004915214396476322,
+      "loss": 3.4999,
+      "step": 2410
+    },
+    {
+      "epoch": 0.10151578947368421,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004915127239507806,
+      "loss": 3.4293,
+      "step": 2411
+    },
+    {
+      "epoch": 0.10155789473684211,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004915040038538509,
+      "loss": 3.7979,
+      "step": 2412
+    },
+    {
+      "epoch": 0.1016,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.000491495279357002,
+      "loss": 3.7688,
+      "step": 2413
+    },
+    {
+      "epoch": 0.1016421052631579,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0004914865504603929,
+      "loss": 4.1445,
+      "step": 2414
+    },
+    {
+      "epoch": 0.1016842105263158,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0004914778171641826,
+      "loss": 3.4013,
+      "step": 2415
+    },
+    {
+      "epoch": 0.10172631578947368,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004914690794685302,
+      "loss": 3.2828,
+      "step": 2416
+    },
+    {
+      "epoch": 0.10176842105263158,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004914603373735948,
+      "loss": 3.3925,
+      "step": 2417
+    },
+    {
+      "epoch": 0.10181052631578948,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004914515908795358,
+      "loss": 3.3367,
+      "step": 2418
+    },
+    {
+      "epoch": 0.10185263157894738,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004914428399865126,
+      "loss": 3.6096,
+      "step": 2419
+    },
+    {
+      "epoch": 0.10189473684210526,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004914340846946843,
+      "loss": 3.6033,
+      "step": 2420
+    },
+    {
+      "epoch": 0.10193684210526316,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004914253250042109,
+      "loss": 3.0198,
+      "step": 2421
+    },
+    {
+      "epoch": 0.10197894736842106,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0004914165609152516,
+      "loss": 3.4526,
+      "step": 2422
+    },
+    {
+      "epoch": 0.10202105263157894,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0004914077924279663,
+      "loss": 3.216,
+      "step": 2423
+    },
+    {
+      "epoch": 0.10206315789473684,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004913990195425146,
+      "loss": 3.4327,
+      "step": 2424
+    },
+    {
+      "epoch": 0.10210526315789474,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004913902422590564,
+      "loss": 3.4962,
+      "step": 2425
+    },
+    {
+      "epoch": 0.10214736842105263,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0004913814605777517,
+      "loss": 3.1794,
+      "step": 2426
+    },
+    {
+      "epoch": 0.10218947368421052,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004913726744987602,
+      "loss": 3.7942,
+      "step": 2427
+    },
+    {
+      "epoch": 0.10223157894736842,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004913638840222424,
+      "loss": 3.6939,
+      "step": 2428
+    },
+    {
+      "epoch": 0.10227368421052632,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004913550891483581,
+      "loss": 3.6534,
+      "step": 2429
+    },
+    {
+      "epoch": 0.10231578947368421,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004913462898772677,
+      "loss": 3.2149,
+      "step": 2430
+    },
+    {
+      "epoch": 0.1023578947368421,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0004913374862091314,
+      "loss": 3.4913,
+      "step": 2431
+    },
+    {
+      "epoch": 0.1024,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004913286781441097,
+      "loss": 3.7831,
+      "step": 2432
+    },
+    {
+      "epoch": 0.10244210526315789,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004913198656823631,
+      "loss": 3.4988,
+      "step": 2433
+    },
+    {
+      "epoch": 0.10248421052631579,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.000491311048824052,
+      "loss": 3.6084,
+      "step": 2434
+    },
+    {
+      "epoch": 0.10252631578947369,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004913022275693372,
+      "loss": 3.3943,
+      "step": 2435
+    },
+    {
+      "epoch": 0.10256842105263157,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004912934019183793,
+      "loss": 3.4985,
+      "step": 2436
+    },
+    {
+      "epoch": 0.10261052631578947,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004912845718713391,
+      "loss": 3.4263,
+      "step": 2437
+    },
+    {
+      "epoch": 0.10265263157894737,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0004912757374283775,
+      "loss": 3.5648,
+      "step": 2438
+    },
+    {
+      "epoch": 0.10269473684210527,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004912668985896555,
+      "loss": 3.6311,
+      "step": 2439
+    },
+    {
+      "epoch": 0.10273684210526315,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.000491258055355334,
+      "loss": 3.0916,
+      "step": 2440
+    },
+    {
+      "epoch": 0.10277894736842105,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004912492077255743,
+      "loss": 3.679,
+      "step": 2441
+    },
+    {
+      "epoch": 0.10282105263157895,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004912403557005374,
+      "loss": 3.87,
+      "step": 2442
+    },
+    {
+      "epoch": 0.10286315789473684,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0004912314992803847,
+      "loss": 3.3916,
+      "step": 2443
+    },
+    {
+      "epoch": 0.10290526315789474,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004912226384652775,
+      "loss": 3.4416,
+      "step": 2444
+    },
+    {
+      "epoch": 0.10294736842105263,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004912137732553772,
+      "loss": 3.6113,
+      "step": 2445
+    },
+    {
+      "epoch": 0.10298947368421052,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0004912049036508454,
+      "loss": 3.6274,
+      "step": 2446
+    },
+    {
+      "epoch": 0.10303157894736842,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004911960296518438,
+      "loss": 3.4642,
+      "step": 2447
+    },
+    {
+      "epoch": 0.10307368421052632,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004911871512585337,
+      "loss": 3.9321,
+      "step": 2448
+    },
+    {
+      "epoch": 0.10311578947368422,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0004911782684710773,
+      "loss": 3.8722,
+      "step": 2449
+    },
+    {
+      "epoch": 0.1031578947368421,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004911693812896361,
+      "loss": 3.4834,
+      "step": 2450
+    },
+    {
+      "epoch": 0.1032,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004911604897143722,
+      "loss": 3.4988,
+      "step": 2451
+    },
+    {
+      "epoch": 0.1032421052631579,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004911515937454474,
+      "loss": 3.6018,
+      "step": 2452
+    },
+    {
+      "epoch": 0.10328421052631578,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.000491142693383024,
+      "loss": 3.6775,
+      "step": 2453
+    },
+    {
+      "epoch": 0.10332631578947368,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004911337886272641,
+      "loss": 3.3132,
+      "step": 2454
+    },
+    {
+      "epoch": 0.10336842105263158,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004911248794783299,
+      "loss": 3.5588,
+      "step": 2455
+    },
+    {
+      "epoch": 0.10341052631578948,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004911159659363836,
+      "loss": 3.8202,
+      "step": 2456
+    },
+    {
+      "epoch": 0.10345263157894737,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004911070480015878,
+      "loss": 3.5973,
+      "step": 2457
+    },
+    {
+      "epoch": 0.10349473684210526,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004910981256741048,
+      "loss": 3.3257,
+      "step": 2458
+    },
+    {
+      "epoch": 0.10353684210526316,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004910891989540973,
+      "loss": 3.6352,
+      "step": 2459
+    },
+    {
+      "epoch": 0.10357894736842105,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0004910802678417279,
+      "loss": 3.6312,
+      "step": 2460
+    },
+    {
+      "epoch": 0.10362105263157895,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004910713323371591,
+      "loss": 3.8383,
+      "step": 2461
+    },
+    {
+      "epoch": 0.10366315789473685,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.000491062392440554,
+      "loss": 4.0315,
+      "step": 2462
+    },
+    {
+      "epoch": 0.10370526315789473,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004910534481520753,
+      "loss": 3.4518,
+      "step": 2463
+    },
+    {
+      "epoch": 0.10374736842105263,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004910444994718861,
+      "loss": 3.489,
+      "step": 2464
+    },
+    {
+      "epoch": 0.10378947368421053,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004910355464001492,
+      "loss": 3.5052,
+      "step": 2465
+    },
+    {
+      "epoch": 0.10383157894736843,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004910265889370279,
+      "loss": 3.3843,
+      "step": 2466
+    },
+    {
+      "epoch": 0.10387368421052631,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0004910176270826854,
+      "loss": 3.4297,
+      "step": 2467
+    },
+    {
+      "epoch": 0.10391578947368421,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004910086608372849,
+      "loss": 3.7052,
+      "step": 2468
+    },
+    {
+      "epoch": 0.10395789473684211,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004909996902009897,
+      "loss": 3.5446,
+      "step": 2469
+    },
+    {
+      "epoch": 0.104,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004909907151739633,
+      "loss": 3.178,
+      "step": 2470
+    },
+    {
+      "epoch": 0.1040421052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004909817357563693,
+      "loss": 3.6128,
+      "step": 2471
+    },
+    {
+      "epoch": 0.10408421052631579,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0004909727519483712,
+      "loss": 3.6213,
+      "step": 2472
+    },
+    {
+      "epoch": 0.10412631578947368,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004909637637501327,
+      "loss": 3.6342,
+      "step": 2473
+    },
+    {
+      "epoch": 0.10416842105263158,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004909547711618175,
+      "loss": 3.5634,
+      "step": 2474
+    },
+    {
+      "epoch": 0.10421052631578948,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0004909457741835895,
+      "loss": 3.7878,
+      "step": 2475
+    },
+    {
+      "epoch": 0.10425263157894737,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004909367728156126,
+      "loss": 3.2269,
+      "step": 2476
+    },
+    {
+      "epoch": 0.10429473684210526,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004909277670580508,
+      "loss": 3.3652,
+      "step": 2477
+    },
+    {
+      "epoch": 0.10433684210526316,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004909187569110681,
+      "loss": 3.4177,
+      "step": 2478
+    },
+    {
+      "epoch": 0.10437894736842106,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0004909097423748288,
+      "loss": 3.4736,
+      "step": 2479
+    },
+    {
+      "epoch": 0.10442105263157894,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.000490900723449497,
+      "loss": 3.6626,
+      "step": 2480
+    },
+    {
+      "epoch": 0.10446315789473684,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004908917001352371,
+      "loss": 3.4584,
+      "step": 2481
+    },
+    {
+      "epoch": 0.10450526315789474,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0004908826724322134,
+      "loss": 4.0237,
+      "step": 2482
+    },
+    {
+      "epoch": 0.10454736842105262,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0004908736403405905,
+      "loss": 3.6316,
+      "step": 2483
+    },
+    {
+      "epoch": 0.10458947368421052,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004908646038605329,
+      "loss": 3.5141,
+      "step": 2484
+    },
+    {
+      "epoch": 0.10463157894736842,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004908555629922052,
+      "loss": 3.7593,
+      "step": 2485
+    },
+    {
+      "epoch": 0.10467368421052632,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004908465177357721,
+      "loss": 3.6262,
+      "step": 2486
+    },
+    {
+      "epoch": 0.1047157894736842,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0004908374680913984,
+      "loss": 3.2696,
+      "step": 2487
+    },
+    {
+      "epoch": 0.1047578947368421,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.000490828414059249,
+      "loss": 3.5004,
+      "step": 2488
+    },
+    {
+      "epoch": 0.1048,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.000490819355639489,
+      "loss": 3.8349,
+      "step": 2489
+    },
+    {
+      "epoch": 0.10484210526315789,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0004908102928322832,
+      "loss": 3.6042,
+      "step": 2490
+    },
+    {
+      "epoch": 0.10488421052631579,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0004908012256377969,
+      "loss": 3.3381,
+      "step": 2491
+    },
+    {
+      "epoch": 0.10492631578947369,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004907921540561951,
+      "loss": 3.089,
+      "step": 2492
+    },
+    {
+      "epoch": 0.10496842105263159,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004907830780876431,
+      "loss": 3.8093,
+      "step": 2493
+    },
+    {
+      "epoch": 0.10501052631578947,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0004907739977323064,
+      "loss": 3.778,
+      "step": 2494
+    },
+    {
+      "epoch": 0.10505263157894737,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004907649129903504,
+      "loss": 4.023,
+      "step": 2495
+    },
+    {
+      "epoch": 0.10509473684210527,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0004907558238619405,
+      "loss": 3.0695,
+      "step": 2496
+    },
+    {
+      "epoch": 0.10513684210526315,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0004907467303472423,
+      "loss": 3.5901,
+      "step": 2497
+    },
+    {
+      "epoch": 0.10517894736842105,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0004907376324464218,
+      "loss": 3.4782,
+      "step": 2498
+    },
+    {
+      "epoch": 0.10522105263157895,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004907285301596442,
+      "loss": 3.2867,
+      "step": 2499
+    },
+    {
+      "epoch": 0.10526315789473684,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004907194234870758,
+      "loss": 3.462,
+      "step": 2500
+    },
+    {
+      "epoch": 0.10530526315789474,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0004907103124288823,
+      "loss": 3.423,
+      "step": 2501
+    },
+    {
+      "epoch": 0.10534736842105263,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0004907011969852297,
+      "loss": 3.3231,
+      "step": 2502
+    },
+    {
+      "epoch": 0.10538947368421053,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004906920771562842,
+      "loss": 3.5672,
+      "step": 2503
+    },
+    {
+      "epoch": 0.10543157894736842,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004906829529422118,
+      "loss": 3.4666,
+      "step": 2504
+    },
+    {
+      "epoch": 0.10547368421052632,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004906738243431787,
+      "loss": 3.8139,
+      "step": 2505
+    },
+    {
+      "epoch": 0.10551578947368422,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004906646913593514,
+      "loss": 3.0866,
+      "step": 2506
+    },
+    {
+      "epoch": 0.1055578947368421,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004906555539908962,
+      "loss": 3.0625,
+      "step": 2507
+    },
+    {
+      "epoch": 0.1056,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004906464122379796,
+      "loss": 3.3318,
+      "step": 2508
+    },
+    {
+      "epoch": 0.1056421052631579,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.000490637266100768,
+      "loss": 3.6122,
+      "step": 2509
+    },
+    {
+      "epoch": 0.10568421052631578,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004906281155794282,
+      "loss": 2.7703,
+      "step": 2510
+    },
+    {
+      "epoch": 0.10572631578947368,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0004906189606741268,
+      "loss": 3.3437,
+      "step": 2511
+    },
+    {
+      "epoch": 0.10576842105263158,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004906098013850307,
+      "loss": 3.3754,
+      "step": 2512
+    },
+    {
+      "epoch": 0.10581052631578948,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004906006377123068,
+      "loss": 3.4719,
+      "step": 2513
+    },
+    {
+      "epoch": 0.10585263157894736,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004905914696561219,
+      "loss": 3.7294,
+      "step": 2514
+    },
+    {
+      "epoch": 0.10589473684210526,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004905822972166432,
+      "loss": 3.3914,
+      "step": 2515
+    },
+    {
+      "epoch": 0.10593684210526316,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0004905731203940375,
+      "loss": 3.5486,
+      "step": 2516
+    },
+    {
+      "epoch": 0.10597894736842105,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004905639391884724,
+      "loss": 3.5795,
+      "step": 2517
+    },
+    {
+      "epoch": 0.10602105263157895,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.000490554753600115,
+      "loss": 3.7017,
+      "step": 2518
+    },
+    {
+      "epoch": 0.10606315789473685,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004905455636291324,
+      "loss": 4.0968,
+      "step": 2519
+    },
+    {
+      "epoch": 0.10610526315789473,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004905363692756925,
+      "loss": 3.7371,
+      "step": 2520
+    },
+    {
+      "epoch": 0.10614736842105263,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004905271705399624,
+      "loss": 2.97,
+      "step": 2521
+    },
+    {
+      "epoch": 0.10618947368421053,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00049051796742211,
+      "loss": 3.3626,
+      "step": 2522
+    },
+    {
+      "epoch": 0.10623157894736843,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0004905087599223027,
+      "loss": 3.4704,
+      "step": 2523
+    },
+    {
+      "epoch": 0.10627368421052631,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004904995480407086,
+      "loss": 3.9547,
+      "step": 2524
+    },
+    {
+      "epoch": 0.10631578947368421,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.000490490331777495,
+      "loss": 3.9125,
+      "step": 2525
+    },
+    {
+      "epoch": 0.10635789473684211,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0004904811111328303,
+      "loss": 3.0285,
+      "step": 2526
+    },
+    {
+      "epoch": 0.1064,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004904718861068823,
+      "loss": 3.566,
+      "step": 2527
+    },
+    {
+      "epoch": 0.1064421052631579,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004904626566998191,
+      "loss": 3.4938,
+      "step": 2528
+    },
+    {
+      "epoch": 0.10648421052631579,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004904534229118087,
+      "loss": 3.5643,
+      "step": 2529
+    },
+    {
+      "epoch": 0.10652631578947368,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004904441847430196,
+      "loss": 3.4715,
+      "step": 2530
+    },
+    {
+      "epoch": 0.10656842105263158,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00049043494219362,
+      "loss": 3.9355,
+      "step": 2531
+    },
+    {
+      "epoch": 0.10661052631578948,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004904256952637781,
+      "loss": 3.3853,
+      "step": 2532
+    },
+    {
+      "epoch": 0.10665263157894737,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004904164439536626,
+      "loss": 3.1788,
+      "step": 2533
+    },
+    {
+      "epoch": 0.10669473684210526,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004904071882634419,
+      "loss": 3.379,
+      "step": 2534
+    },
+    {
+      "epoch": 0.10673684210526316,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004903979281932849,
+      "loss": 3.947,
+      "step": 2535
+    },
+    {
+      "epoch": 0.10677894736842106,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004903886637433599,
+      "loss": 3.6698,
+      "step": 2536
+    },
+    {
+      "epoch": 0.10682105263157894,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004903793949138359,
+      "loss": 3.8876,
+      "step": 2537
+    },
+    {
+      "epoch": 0.10686315789473684,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004903701217048818,
+      "loss": 3.754,
+      "step": 2538
+    },
+    {
+      "epoch": 0.10690526315789474,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004903608441166666,
+      "loss": 3.6785,
+      "step": 2539
+    },
+    {
+      "epoch": 0.10694736842105264,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.000490351562149359,
+      "loss": 4.0393,
+      "step": 2540
+    },
+    {
+      "epoch": 0.10698947368421052,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004903422758031285,
+      "loss": 3.7675,
+      "step": 2541
+    },
+    {
+      "epoch": 0.10703157894736842,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0004903329850781442,
+      "loss": 3.4964,
+      "step": 2542
+    },
+    {
+      "epoch": 0.10707368421052632,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004903236899745752,
+      "loss": 3.729,
+      "step": 2543
+    },
+    {
+      "epoch": 0.1071157894736842,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.000490314390492591,
+      "loss": 3.4319,
+      "step": 2544
+    },
+    {
+      "epoch": 0.1071578947368421,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0004903050866323608,
+      "loss": 3.559,
+      "step": 2545
+    },
+    {
+      "epoch": 0.1072,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004902957783940544,
+      "loss": 3.3122,
+      "step": 2546
+    },
+    {
+      "epoch": 0.10724210526315789,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004902864657778411,
+      "loss": 3.2039,
+      "step": 2547
+    },
+    {
+      "epoch": 0.10728421052631579,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004902771487838909,
+      "loss": 3.7317,
+      "step": 2548
+    },
+    {
+      "epoch": 0.10732631578947369,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0004902678274123734,
+      "loss": 3.4333,
+      "step": 2549
+    },
+    {
+      "epoch": 0.10736842105263159,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004902585016634583,
+      "loss": 3.4448,
+      "step": 2550
+    },
+    {
+      "epoch": 0.10741052631578947,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004902491715373155,
+      "loss": 3.1972,
+      "step": 2551
+    },
+    {
+      "epoch": 0.10745263157894737,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004902398370341153,
+      "loss": 3.6141,
+      "step": 2552
+    },
+    {
+      "epoch": 0.10749473684210527,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004902304981540273,
+      "loss": 3.4586,
+      "step": 2553
+    },
+    {
+      "epoch": 0.10753684210526315,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.000490221154897222,
+      "loss": 3.4054,
+      "step": 2554
+    },
+    {
+      "epoch": 0.10757894736842105,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004902118072638695,
+      "loss": 3.5277,
+      "step": 2555
+    },
+    {
+      "epoch": 0.10762105263157895,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.0004902024552541401,
+      "loss": 3.6185,
+      "step": 2556
+    },
+    {
+      "epoch": 0.10766315789473684,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004901930988682042,
+      "loss": 3.6268,
+      "step": 2557
+    },
+    {
+      "epoch": 0.10770526315789473,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0004901837381062322,
+      "loss": 3.7803,
+      "step": 2558
+    },
+    {
+      "epoch": 0.10774736842105263,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004901743729683947,
+      "loss": 3.164,
+      "step": 2559
+    },
+    {
+      "epoch": 0.10778947368421053,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004901650034548625,
+      "loss": 3.5852,
+      "step": 2560
+    },
+    {
+      "epoch": 0.10783157894736842,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004901556295658059,
+      "loss": 3.9368,
+      "step": 2561
+    },
+    {
+      "epoch": 0.10787368421052632,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.000490146251301396,
+      "loss": 3.9976,
+      "step": 2562
+    },
+    {
+      "epoch": 0.10791578947368421,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0004901368686618035,
+      "loss": 3.8391,
+      "step": 2563
+    },
+    {
+      "epoch": 0.1079578947368421,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004901274816471995,
+      "loss": 3.6153,
+      "step": 2564
+    },
+    {
+      "epoch": 0.108,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004901180902577548,
+      "loss": 3.579,
+      "step": 2565
+    },
+    {
+      "epoch": 0.1080421052631579,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004901086944936407,
+      "loss": 3.8054,
+      "step": 2566
+    },
+    {
+      "epoch": 0.10808421052631578,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0004900992943550283,
+      "loss": 3.3685,
+      "step": 2567
+    },
+    {
+      "epoch": 0.10812631578947368,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004900898898420889,
+      "loss": 3.5342,
+      "step": 2568
+    },
+    {
+      "epoch": 0.10816842105263158,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004900804809549937,
+      "loss": 3.3373,
+      "step": 2569
+    },
+    {
+      "epoch": 0.10821052631578948,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004900710676939143,
+      "loss": 3.4235,
+      "step": 2570
+    },
+    {
+      "epoch": 0.10825263157894736,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.000490061650059022,
+      "loss": 3.7912,
+      "step": 2571
+    },
+    {
+      "epoch": 0.10829473684210526,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004900522280504887,
+      "loss": 3.8327,
+      "step": 2572
+    },
+    {
+      "epoch": 0.10833684210526316,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0004900428016684857,
+      "loss": 3.5385,
+      "step": 2573
+    },
+    {
+      "epoch": 0.10837894736842105,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004900333709131849,
+      "loss": 3.7324,
+      "step": 2574
+    },
+    {
+      "epoch": 0.10842105263157895,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004900239357847582,
+      "loss": 3.1748,
+      "step": 2575
+    },
+    {
+      "epoch": 0.10846315789473684,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004900144962833773,
+      "loss": 3.3892,
+      "step": 2576
+    },
+    {
+      "epoch": 0.10850526315789474,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0004900050524092143,
+      "loss": 3.7106,
+      "step": 2577
+    },
+    {
+      "epoch": 0.10854736842105263,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004899956041624413,
+      "loss": 3.1601,
+      "step": 2578
+    },
+    {
+      "epoch": 0.10858947368421053,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004899861515432303,
+      "loss": 3.5534,
+      "step": 2579
+    },
+    {
+      "epoch": 0.10863157894736843,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004899766945517537,
+      "loss": 3.7678,
+      "step": 2580
+    },
+    {
+      "epoch": 0.10867368421052631,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004899672331881836,
+      "loss": 3.7374,
+      "step": 2581
+    },
+    {
+      "epoch": 0.10871578947368421,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004899577674526925,
+      "loss": 3.9345,
+      "step": 2582
+    },
+    {
+      "epoch": 0.10875789473684211,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004899482973454529,
+      "loss": 3.3318,
+      "step": 2583
+    },
+    {
+      "epoch": 0.1088,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.000489938822866637,
+      "loss": 3.7552,
+      "step": 2584
+    },
+    {
+      "epoch": 0.10884210526315789,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004899293440164178,
+      "loss": 3.2896,
+      "step": 2585
+    },
+    {
+      "epoch": 0.10888421052631579,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.000489919860794968,
+      "loss": 3.5565,
+      "step": 2586
+    },
+    {
+      "epoch": 0.10892631578947369,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0004899103732024601,
+      "loss": 3.2471,
+      "step": 2587
+    },
+    {
+      "epoch": 0.10896842105263158,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004899008812390671,
+      "loss": 3.795,
+      "step": 2588
+    },
+    {
+      "epoch": 0.10901052631578947,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004898913849049619,
+      "loss": 3.6525,
+      "step": 2589
+    },
+    {
+      "epoch": 0.10905263157894737,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004898818842003175,
+      "loss": 3.3793,
+      "step": 2590
+    },
+    {
+      "epoch": 0.10909473684210526,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.000489872379125307,
+      "loss": 3.6333,
+      "step": 2591
+    },
+    {
+      "epoch": 0.10913684210526316,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004898628696801037,
+      "loss": 3.4143,
+      "step": 2592
+    },
+    {
+      "epoch": 0.10917894736842106,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0004898533558648807,
+      "loss": 3.3189,
+      "step": 2593
+    },
+    {
+      "epoch": 0.10922105263157894,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004898438376798114,
+      "loss": 3.565,
+      "step": 2594
+    },
+    {
+      "epoch": 0.10926315789473684,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.000489834315125069,
+      "loss": 3.4209,
+      "step": 2595
+    },
+    {
+      "epoch": 0.10930526315789474,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004898247882008273,
+      "loss": 3.5324,
+      "step": 2596
+    },
+    {
+      "epoch": 0.10934736842105264,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0004898152569072598,
+      "loss": 3.5261,
+      "step": 2597
+    },
+    {
+      "epoch": 0.10938947368421052,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004898057212445401,
+      "loss": 3.5859,
+      "step": 2598
+    },
+    {
+      "epoch": 0.10943157894736842,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004897961812128418,
+      "loss": 3.4298,
+      "step": 2599
+    },
+    {
+      "epoch": 0.10947368421052632,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.000489786636812339,
+      "loss": 3.317,
+      "step": 2600
+    },
+    {
+      "epoch": 0.1095157894736842,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004897770880432052,
+      "loss": 3.5587,
+      "step": 2601
+    },
+    {
+      "epoch": 0.1095578947368421,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0004897675349056147,
+      "loss": 3.6695,
+      "step": 2602
+    },
+    {
+      "epoch": 0.1096,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0004897579773997414,
+      "loss": 3.3769,
+      "step": 2603
+    },
+    {
+      "epoch": 0.10964210526315789,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004897484155257594,
+      "loss": 3.6357,
+      "step": 2604
+    },
+    {
+      "epoch": 0.10968421052631579,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004897388492838432,
+      "loss": 3.4751,
+      "step": 2605
+    },
+    {
+      "epoch": 0.10972631578947369,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004897292786741666,
+      "loss": 3.2206,
+      "step": 2606
+    },
+    {
+      "epoch": 0.10976842105263158,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004897197036969044,
+      "loss": 3.3609,
+      "step": 2607
+    },
+    {
+      "epoch": 0.10981052631578947,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004897101243522306,
+      "loss": 3.4097,
+      "step": 2608
+    },
+    {
+      "epoch": 0.10985263157894737,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004897005406403202,
+      "loss": 3.6122,
+      "step": 2609
+    },
+    {
+      "epoch": 0.10989473684210527,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004896909525613475,
+      "loss": 3.6775,
+      "step": 2610
+    },
+    {
+      "epoch": 0.10993684210526315,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004896813601154873,
+      "loss": 3.4825,
+      "step": 2611
+    },
+    {
+      "epoch": 0.10997894736842105,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004896717633029143,
+      "loss": 3.9252,
+      "step": 2612
+    },
+    {
+      "epoch": 0.11002105263157895,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0004896621621238034,
+      "loss": 3.3762,
+      "step": 2613
+    },
+    {
+      "epoch": 0.11006315789473685,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004896525565783296,
+      "loss": 3.6748,
+      "step": 2614
+    },
+    {
+      "epoch": 0.11010526315789473,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004896429466666676,
+      "loss": 3.1753,
+      "step": 2615
+    },
+    {
+      "epoch": 0.11014736842105263,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004896333323889928,
+      "loss": 3.45,
+      "step": 2616
+    },
+    {
+      "epoch": 0.11018947368421053,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004896237137454802,
+      "loss": 3.7695,
+      "step": 2617
+    },
+    {
+      "epoch": 0.11023157894736842,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004896140907363051,
+      "loss": 3.924,
+      "step": 2618
+    },
+    {
+      "epoch": 0.11027368421052632,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004896044633616428,
+      "loss": 3.8208,
+      "step": 2619
+    },
+    {
+      "epoch": 0.11031578947368421,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004895948316216687,
+      "loss": 3.3617,
+      "step": 2620
+    },
+    {
+      "epoch": 0.1103578947368421,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004895851955165582,
+      "loss": 3.7592,
+      "step": 2621
+    },
+    {
+      "epoch": 0.1104,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0004895755550464871,
+      "loss": 3.3189,
+      "step": 2622
+    },
+    {
+      "epoch": 0.1104421052631579,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004895659102116308,
+      "loss": 3.3376,
+      "step": 2623
+    },
+    {
+      "epoch": 0.1104842105263158,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0004895562610121652,
+      "loss": 3.3798,
+      "step": 2624
+    },
+    {
+      "epoch": 0.11052631578947368,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004895466074482657,
+      "loss": 3.7928,
+      "step": 2625
+    },
+    {
+      "epoch": 0.11056842105263158,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004895369495201086,
+      "loss": 3.0828,
+      "step": 2626
+    },
+    {
+      "epoch": 0.11061052631578948,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004895272872278699,
+      "loss": 3.6787,
+      "step": 2627
+    },
+    {
+      "epoch": 0.11065263157894736,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004895176205717254,
+      "loss": 3.791,
+      "step": 2628
+    },
+    {
+      "epoch": 0.11069473684210526,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.000489507949551851,
+      "loss": 3.423,
+      "step": 2629
+    },
+    {
+      "epoch": 0.11073684210526316,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004894982741684235,
+      "loss": 3.4624,
+      "step": 2630
+    },
+    {
+      "epoch": 0.11077894736842105,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004894885944216186,
+      "loss": 3.4961,
+      "step": 2631
+    },
+    {
+      "epoch": 0.11082105263157895,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004894789103116131,
+      "loss": 3.6301,
+      "step": 2632
+    },
+    {
+      "epoch": 0.11086315789473684,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004894692218385832,
+      "loss": 3.8849,
+      "step": 2633
+    },
+    {
+      "epoch": 0.11090526315789474,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0004894595290027054,
+      "loss": 3.2536,
+      "step": 2634
+    },
+    {
+      "epoch": 0.11094736842105263,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004894498318041562,
+      "loss": 3.9343,
+      "step": 2635
+    },
+    {
+      "epoch": 0.11098947368421053,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004894401302431126,
+      "loss": 3.5315,
+      "step": 2636
+    },
+    {
+      "epoch": 0.11103157894736843,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004894304243197511,
+      "loss": 3.3575,
+      "step": 2637
+    },
+    {
+      "epoch": 0.11107368421052631,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0004894207140342486,
+      "loss": 3.2135,
+      "step": 2638
+    },
+    {
+      "epoch": 0.11111578947368421,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.000489410999386782,
+      "loss": 3.5283,
+      "step": 2639
+    },
+    {
+      "epoch": 0.11115789473684211,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004894012803775283,
+      "loss": 3.7674,
+      "step": 2640
+    },
+    {
+      "epoch": 0.1112,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004893915570066645,
+      "loss": 3.4459,
+      "step": 2641
+    },
+    {
+      "epoch": 0.11124210526315789,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004893818292743678,
+      "loss": 3.5808,
+      "step": 2642
+    },
+    {
+      "epoch": 0.11128421052631579,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004893720971808154,
+      "loss": 3.3435,
+      "step": 2643
+    },
+    {
+      "epoch": 0.11132631578947369,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004893623607261847,
+      "loss": 3.5229,
+      "step": 2644
+    },
+    {
+      "epoch": 0.11136842105263157,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0004893526199106531,
+      "loss": 3.2473,
+      "step": 2645
+    },
+    {
+      "epoch": 0.11141052631578947,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0004893428747343979,
+      "loss": 3.7486,
+      "step": 2646
+    },
+    {
+      "epoch": 0.11145263157894737,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004893331251975968,
+      "loss": 3.56,
+      "step": 2647
+    },
+    {
+      "epoch": 0.11149473684210526,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0004893233713004272,
+      "loss": 3.3711,
+      "step": 2648
+    },
+    {
+      "epoch": 0.11153684210526316,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004893136130430671,
+      "loss": 3.1145,
+      "step": 2649
+    },
+    {
+      "epoch": 0.11157894736842106,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004893038504256941,
+      "loss": 3.5733,
+      "step": 2650
+    },
+    {
+      "epoch": 0.11162105263157895,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0004892940834484862,
+      "loss": 2.9337,
+      "step": 2651
+    },
+    {
+      "epoch": 0.11166315789473684,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004892843121116213,
+      "loss": 3.6927,
+      "step": 2652
+    },
+    {
+      "epoch": 0.11170526315789474,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004892745364152773,
+      "loss": 3.8813,
+      "step": 2653
+    },
+    {
+      "epoch": 0.11174736842105264,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004892647563596323,
+      "loss": 3.6019,
+      "step": 2654
+    },
+    {
+      "epoch": 0.11178947368421052,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004892549719448647,
+      "loss": 3.5056,
+      "step": 2655
+    },
+    {
+      "epoch": 0.11183157894736842,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004892451831711525,
+      "loss": 3.6645,
+      "step": 2656
+    },
+    {
+      "epoch": 0.11187368421052632,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0004892353900386742,
+      "loss": 3.4746,
+      "step": 2657
+    },
+    {
+      "epoch": 0.1119157894736842,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004892255925476082,
+      "loss": 3.7596,
+      "step": 2658
+    },
+    {
+      "epoch": 0.1119578947368421,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.000489215790698133,
+      "loss": 3.4502,
+      "step": 2659
+    },
+    {
+      "epoch": 0.112,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004892059844904272,
+      "loss": 3.1957,
+      "step": 2660
+    },
+    {
+      "epoch": 0.1120421052631579,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004891961739246694,
+      "loss": 3.5748,
+      "step": 2661
+    },
+    {
+      "epoch": 0.11208421052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004891863590010383,
+      "loss": 3.6574,
+      "step": 2662
+    },
+    {
+      "epoch": 0.11212631578947369,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004891765397197128,
+      "loss": 3.4892,
+      "step": 2663
+    },
+    {
+      "epoch": 0.11216842105263158,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004891667160808719,
+      "loss": 3.8335,
+      "step": 2664
+    },
+    {
+      "epoch": 0.11221052631578947,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0004891568880846941,
+      "loss": 3.1086,
+      "step": 2665
+    },
+    {
+      "epoch": 0.11225263157894737,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004891470557313591,
+      "loss": 3.1166,
+      "step": 2666
+    },
+    {
+      "epoch": 0.11229473684210527,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0004891372190210456,
+      "loss": 3.6589,
+      "step": 2667
+    },
+    {
+      "epoch": 0.11233684210526315,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.000489127377953933,
+      "loss": 3.8796,
+      "step": 2668
+    },
+    {
+      "epoch": 0.11237894736842105,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004891175325302006,
+      "loss": 3.6773,
+      "step": 2669
+    },
+    {
+      "epoch": 0.11242105263157895,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0004891076827500276,
+      "loss": 3.8707,
+      "step": 2670
+    },
+    {
+      "epoch": 0.11246315789473685,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004890978286135935,
+      "loss": 3.5764,
+      "step": 2671
+    },
+    {
+      "epoch": 0.11250526315789473,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.000489087970121078,
+      "loss": 3.4786,
+      "step": 2672
+    },
+    {
+      "epoch": 0.11254736842105263,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0004890781072726606,
+      "loss": 3.2337,
+      "step": 2673
+    },
+    {
+      "epoch": 0.11258947368421053,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.000489068240068521,
+      "loss": 3.5644,
+      "step": 2674
+    },
+    {
+      "epoch": 0.11263157894736842,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0004890583685088389,
+      "loss": 3.9321,
+      "step": 2675
+    },
+    {
+      "epoch": 0.11267368421052631,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004890484925937943,
+      "loss": 3.7626,
+      "step": 2676
+    },
+    {
+      "epoch": 0.11271578947368421,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.000489038612323567,
+      "loss": 3.5162,
+      "step": 2677
+    },
+    {
+      "epoch": 0.1127578947368421,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.000489028727698337,
+      "loss": 3.3456,
+      "step": 2678
+    },
+    {
+      "epoch": 0.1128,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004890188387182843,
+      "loss": 3.7814,
+      "step": 2679
+    },
+    {
+      "epoch": 0.1128421052631579,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004890089453835894,
+      "loss": 3.2508,
+      "step": 2680
+    },
+    {
+      "epoch": 0.1128842105263158,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0004889990476944322,
+      "loss": 2.7378,
+      "step": 2681
+    },
+    {
+      "epoch": 0.11292631578947368,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004889891456509933,
+      "loss": 3.6601,
+      "step": 2682
+    },
+    {
+      "epoch": 0.11296842105263158,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004889792392534529,
+      "loss": 3.8327,
+      "step": 2683
+    },
+    {
+      "epoch": 0.11301052631578948,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004889693285019916,
+      "loss": 3.8476,
+      "step": 2684
+    },
+    {
+      "epoch": 0.11305263157894736,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004889594133967899,
+      "loss": 3.2264,
+      "step": 2685
+    },
+    {
+      "epoch": 0.11309473684210526,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004889494939380284,
+      "loss": 3.5233,
+      "step": 2686
+    },
+    {
+      "epoch": 0.11313684210526316,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004889395701258879,
+      "loss": 3.7289,
+      "step": 2687
+    },
+    {
+      "epoch": 0.11317894736842106,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004889296419605492,
+      "loss": 3.5828,
+      "step": 2688
+    },
+    {
+      "epoch": 0.11322105263157894,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004889197094421931,
+      "loss": 3.5343,
+      "step": 2689
+    },
+    {
+      "epoch": 0.11326315789473684,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004889097725710007,
+      "loss": 3.3791,
+      "step": 2690
+    },
+    {
+      "epoch": 0.11330526315789474,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004888998313471528,
+      "loss": 3.8507,
+      "step": 2691
+    },
+    {
+      "epoch": 0.11334736842105263,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004888898857708309,
+      "loss": 3.5045,
+      "step": 2692
+    },
+    {
+      "epoch": 0.11338947368421053,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004888799358422157,
+      "loss": 3.6004,
+      "step": 2693
+    },
+    {
+      "epoch": 0.11343157894736842,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004888699815614888,
+      "loss": 3.4196,
+      "step": 2694
+    },
+    {
+      "epoch": 0.11347368421052631,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004888600229288316,
+      "loss": 3.849,
+      "step": 2695
+    },
+    {
+      "epoch": 0.11351578947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004888500599444255,
+      "loss": 3.2949,
+      "step": 2696
+    },
+    {
+      "epoch": 0.11355789473684211,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004888400926084518,
+      "loss": 3.658,
+      "step": 2697
+    },
+    {
+      "epoch": 0.1136,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004888301209210923,
+      "loss": 3.9084,
+      "step": 2698
+    },
+    {
+      "epoch": 0.11364210526315789,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004888201448825286,
+      "loss": 3.1516,
+      "step": 2699
+    },
+    {
+      "epoch": 0.11368421052631579,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004888101644929424,
+      "loss": 3.7905,
+      "step": 2700
+    },
+    {
+      "epoch": 0.11372631578947369,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004888001797525157,
+      "loss": 3.3805,
+      "step": 2701
+    },
+    {
+      "epoch": 0.11376842105263157,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004887901906614301,
+      "loss": 3.376,
+      "step": 2702
+    },
+    {
+      "epoch": 0.11381052631578947,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.000488780197219868,
+      "loss": 3.6209,
+      "step": 2703
+    },
+    {
+      "epoch": 0.11385263157894737,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004887701994280112,
+      "loss": 3.6253,
+      "step": 2704
+    },
+    {
+      "epoch": 0.11389473684210526,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.000488760197286042,
+      "loss": 3.7679,
+      "step": 2705
+    },
+    {
+      "epoch": 0.11393684210526316,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004887501907941425,
+      "loss": 3.6703,
+      "step": 2706
+    },
+    {
+      "epoch": 0.11397894736842105,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004887401799524948,
+      "loss": 3.5127,
+      "step": 2707
+    },
+    {
+      "epoch": 0.11402105263157895,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004887301647612818,
+      "loss": 3.535,
+      "step": 2708
+    },
+    {
+      "epoch": 0.11406315789473684,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004887201452206856,
+      "loss": 3.7991,
+      "step": 2709
+    },
+    {
+      "epoch": 0.11410526315789474,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004887101213308889,
+      "loss": 3.4245,
+      "step": 2710
+    },
+    {
+      "epoch": 0.11414736842105264,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004887000930920742,
+      "loss": 3.8205,
+      "step": 2711
+    },
+    {
+      "epoch": 0.11418947368421052,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004886900605044242,
+      "loss": 3.3188,
+      "step": 2712
+    },
+    {
+      "epoch": 0.11423157894736842,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004886800235681219,
+      "loss": 3.961,
+      "step": 2713
+    },
+    {
+      "epoch": 0.11427368421052632,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004886699822833499,
+      "loss": 3.601,
+      "step": 2714
+    },
+    {
+      "epoch": 0.1143157894736842,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004886599366502913,
+      "loss": 3.4259,
+      "step": 2715
+    },
+    {
+      "epoch": 0.1143578947368421,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.000488649886669129,
+      "loss": 3.6399,
+      "step": 2716
+    },
+    {
+      "epoch": 0.1144,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004886398323400463,
+      "loss": 3.694,
+      "step": 2717
+    },
+    {
+      "epoch": 0.1144421052631579,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004886297736632261,
+      "loss": 3.5827,
+      "step": 2718
+    },
+    {
+      "epoch": 0.11448421052631579,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0004886197106388519,
+      "loss": 3.6914,
+      "step": 2719
+    },
+    {
+      "epoch": 0.11452631578947368,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004886096432671069,
+      "loss": 4.0794,
+      "step": 2720
+    },
+    {
+      "epoch": 0.11456842105263158,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0004885995715481746,
+      "loss": 3.3042,
+      "step": 2721
+    },
+    {
+      "epoch": 0.11461052631578947,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004885894954822385,
+      "loss": 3.8311,
+      "step": 2722
+    },
+    {
+      "epoch": 0.11465263157894737,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004885794150694819,
+      "loss": 3.9166,
+      "step": 2723
+    },
+    {
+      "epoch": 0.11469473684210527,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004885693303100889,
+      "loss": 3.1083,
+      "step": 2724
+    },
+    {
+      "epoch": 0.11473684210526315,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004885592412042428,
+      "loss": 3.5434,
+      "step": 2725
+    },
+    {
+      "epoch": 0.11477894736842105,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004885491477521278,
+      "loss": 3.2582,
+      "step": 2726
+    },
+    {
+      "epoch": 0.11482105263157895,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0004885390499539276,
+      "loss": 3.36,
+      "step": 2727
+    },
+    {
+      "epoch": 0.11486315789473685,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004885289478098261,
+      "loss": 3.3487,
+      "step": 2728
+    },
+    {
+      "epoch": 0.11490526315789473,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004885188413200074,
+      "loss": 3.6309,
+      "step": 2729
+    },
+    {
+      "epoch": 0.11494736842105263,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004885087304846557,
+      "loss": 3.7837,
+      "step": 2730
+    },
+    {
+      "epoch": 0.11498947368421053,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004884986153039553,
+      "loss": 3.6787,
+      "step": 2731
+    },
+    {
+      "epoch": 0.11503157894736842,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004884884957780902,
+      "loss": 3.9876,
+      "step": 2732
+    },
+    {
+      "epoch": 0.11507368421052631,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.000488478371907245,
+      "loss": 4.0725,
+      "step": 2733
+    },
+    {
+      "epoch": 0.11511578947368421,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.000488468243691604,
+      "loss": 3.281,
+      "step": 2734
+    },
+    {
+      "epoch": 0.11515789473684211,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004884581111313518,
+      "loss": 3.6996,
+      "step": 2735
+    },
+    {
+      "epoch": 0.1152,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.000488447974226673,
+      "loss": 3.57,
+      "step": 2736
+    },
+    {
+      "epoch": 0.1152421052631579,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004884378329777523,
+      "loss": 3.2319,
+      "step": 2737
+    },
+    {
+      "epoch": 0.1152842105263158,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0004884276873847745,
+      "loss": 3.3286,
+      "step": 2738
+    },
+    {
+      "epoch": 0.11532631578947368,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004884175374479243,
+      "loss": 3.768,
+      "step": 2739
+    },
+    {
+      "epoch": 0.11536842105263158,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004884073831673867,
+      "loss": 3.3615,
+      "step": 2740
+    },
+    {
+      "epoch": 0.11541052631578948,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004883972245433466,
+      "loss": 3.321,
+      "step": 2741
+    },
+    {
+      "epoch": 0.11545263157894736,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004883870615759893,
+      "loss": 3.9683,
+      "step": 2742
+    },
+    {
+      "epoch": 0.11549473684210526,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004883768942654998,
+      "loss": 3.2698,
+      "step": 2743
+    },
+    {
+      "epoch": 0.11553684210526316,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0004883667226120633,
+      "loss": 3.3373,
+      "step": 2744
+    },
+    {
+      "epoch": 0.11557894736842106,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004883565466158652,
+      "loss": 3.784,
+      "step": 2745
+    },
+    {
+      "epoch": 0.11562105263157894,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.000488346366277091,
+      "loss": 3.6288,
+      "step": 2746
+    },
+    {
+      "epoch": 0.11566315789473684,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0004883361815959259,
+      "loss": 3.5148,
+      "step": 2747
+    },
+    {
+      "epoch": 0.11570526315789474,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004883259925725557,
+      "loss": 3.5405,
+      "step": 2748
+    },
+    {
+      "epoch": 0.11574736842105263,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.000488315799207166,
+      "loss": 3.5724,
+      "step": 2749
+    },
+    {
+      "epoch": 0.11578947368421053,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004883056014999423,
+      "loss": 3.1807,
+      "step": 2750
+    },
+    {
+      "epoch": 0.11583157894736842,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004882953994510706,
+      "loss": 3.8557,
+      "step": 2751
+    },
+    {
+      "epoch": 0.11587368421052631,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004882851930607368,
+      "loss": 3.8998,
+      "step": 2752
+    },
+    {
+      "epoch": 0.11591578947368421,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004882749823291266,
+      "loss": 3.2544,
+      "step": 2753
+    },
+    {
+      "epoch": 0.11595789473684211,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004882647672564263,
+      "loss": 3.7189,
+      "step": 2754
+    },
+    {
+      "epoch": 0.116,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004882545478428218,
+      "loss": 3.4328,
+      "step": 2755
+    },
+    {
+      "epoch": 0.11604210526315789,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004882443240884994,
+      "loss": 3.3119,
+      "step": 2756
+    },
+    {
+      "epoch": 0.11608421052631579,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00048823409599364537,
+      "loss": 3.2452,
+      "step": 2757
+    },
+    {
+      "epoch": 0.11612631578947369,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0004882238635584461,
+      "loss": 3.264,
+      "step": 2758
+    },
+    {
+      "epoch": 0.11616842105263157,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00048821362678308776,
+      "loss": 3.5905,
+      "step": 2759
+    },
+    {
+      "epoch": 0.11621052631578947,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004882033856677571,
+      "loss": 3.1328,
+      "step": 2760
+    },
+    {
+      "epoch": 0.11625263157894737,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00048819314021264064,
+      "loss": 3.5047,
+      "step": 2761
+    },
+    {
+      "epoch": 0.11629473684210526,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004881828904179251,
+      "loss": 3.4413,
+      "step": 2762
+    },
+    {
+      "epoch": 0.11633684210526316,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00048817263628379707,
+      "loss": 3.2281,
+      "step": 2763
+    },
+    {
+      "epoch": 0.11637894736842105,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00048816237781044346,
+      "loss": 3.7667,
+      "step": 2764
+    },
+    {
+      "epoch": 0.11642105263157895,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004881521149980511,
+      "loss": 3.384,
+      "step": 2765
+    },
+    {
+      "epoch": 0.11646315789473684,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004881418478468071,
+      "loss": 3.4954,
+      "step": 2766
+    },
+    {
+      "epoch": 0.11650526315789474,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00048813157635689845,
+      "loss": 3.7482,
+      "step": 2767
+    },
+    {
+      "epoch": 0.11654736842105264,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004881213005285123,
+      "loss": 3.6393,
+      "step": 2768
+    },
+    {
+      "epoch": 0.11658947368421052,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00048811102036183576,
+      "loss": 3.4446,
+      "step": 2769
+    },
+    {
+      "epoch": 0.11663157894736842,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00048810073585705616,
+      "loss": 3.6504,
+      "step": 2770
+    },
+    {
+      "epoch": 0.11667368421052632,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.000488090447014361,
+      "loss": 3.8029,
+      "step": 2771
+    },
+    {
+      "epoch": 0.11671578947368422,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0004880801538339376,
+      "loss": 3.0719,
+      "step": 2772
+    },
+    {
+      "epoch": 0.1167578947368421,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004880698563159736,
+      "loss": 3.4642,
+      "step": 2773
+    },
+    {
+      "epoch": 0.1168,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004880595544606564,
+      "loss": 3.6255,
+      "step": 2774
+    },
+    {
+      "epoch": 0.1168421052631579,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004880492482681739,
+      "loss": 3.2833,
+      "step": 2775
+    },
+    {
+      "epoch": 0.11688421052631578,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004880389377387138,
+      "loss": 3.3541,
+      "step": 2776
+    },
+    {
+      "epoch": 0.11692631578947368,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00048802862287246394,
+      "loss": 3.6164,
+      "step": 2777
+    },
+    {
+      "epoch": 0.11696842105263158,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00048801830366961224,
+      "loss": 3.9764,
+      "step": 2778
+    },
+    {
+      "epoch": 0.11701052631578947,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004880079801303468,
+      "loss": 3.7477,
+      "step": 2779
+    },
+    {
+      "epoch": 0.11705263157894737,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00048799765225485543,
+      "loss": 4.0114,
+      "step": 2780
+    },
+    {
+      "epoch": 0.11709473684210527,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004879873200433266,
+      "loss": 3.0378,
+      "step": 2781
+    },
+    {
+      "epoch": 0.11713684210526316,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0004879769834959483,
+      "loss": 3.3359,
+      "step": 2782
+    },
+    {
+      "epoch": 0.11717894736842105,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004879666426129091,
+      "loss": 2.9993,
+      "step": 2783
+    },
+    {
+      "epoch": 0.11722105263157895,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004879562973943972,
+      "loss": 4.038,
+      "step": 2784
+    },
+    {
+      "epoch": 0.11726315789473685,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0004879459478406012,
+      "loss": 3.5074,
+      "step": 2785
+    },
+    {
+      "epoch": 0.11730526315789473,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00048793559395170953,
+      "loss": 3.7285,
+      "step": 2786
+    },
+    {
+      "epoch": 0.11734736842105263,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.000487925235727911,
+      "loss": 3.5123,
+      "step": 2787
+    },
+    {
+      "epoch": 0.11738947368421053,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00048791487316939415,
+      "loss": 3.5816,
+      "step": 2788
+    },
+    {
+      "epoch": 0.11743157894736841,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004879045062763479,
+      "loss": 3.5543,
+      "step": 2789
+    },
+    {
+      "epoch": 0.11747368421052631,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.000487894135048961,
+      "loss": 3.526,
+      "step": 2790
+    },
+    {
+      "epoch": 0.11751578947368421,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00048788375948742254,
+      "loss": 3.49,
+      "step": 2791
+    },
+    {
+      "epoch": 0.11755789473684211,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004878733795919215,
+      "loss": 3.5965,
+      "step": 2792
+    },
+    {
+      "epoch": 0.1176,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0004878629953626469,
+      "loss": 3.8373,
+      "step": 2793
+    },
+    {
+      "epoch": 0.1176421052631579,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00048785260679978803,
+      "loss": 3.3511,
+      "step": 2794
+    },
+    {
+      "epoch": 0.1176842105263158,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004878422139035341,
+      "loss": 3.3516,
+      "step": 2795
+    },
+    {
+      "epoch": 0.11772631578947368,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004878318166740745,
+      "loss": 3.2327,
+      "step": 2796
+    },
+    {
+      "epoch": 0.11776842105263158,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00048782141511159873,
+      "loss": 3.0295,
+      "step": 2797
+    },
+    {
+      "epoch": 0.11781052631578948,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004878110092162961,
+      "loss": 3.4337,
+      "step": 2798
+    },
+    {
+      "epoch": 0.11785263157894736,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0004878005989883564,
+      "loss": 3.44,
+      "step": 2799
+    },
+    {
+      "epoch": 0.11789473684210526,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004877901844279691,
+      "loss": 3.4228,
+      "step": 2800
+    },
+    {
+      "epoch": 0.11793684210526316,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00048777976553532397,
+      "loss": 3.546,
+      "step": 2801
+    },
+    {
+      "epoch": 0.11797894736842106,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00048776934231061105,
+      "loss": 3.1686,
+      "step": 2802
+    },
+    {
+      "epoch": 0.11802105263157894,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00048775891475402,
+      "loss": 3.8679,
+      "step": 2803
+    },
+    {
+      "epoch": 0.11806315789473684,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004877484828657409,
+      "loss": 3.788,
+      "step": 2804
+    },
+    {
+      "epoch": 0.11810526315789474,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00048773804664596377,
+      "loss": 3.1713,
+      "step": 2805
+    },
+    {
+      "epoch": 0.11814736842105263,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00048772760609487874,
+      "loss": 3.7858,
+      "step": 2806
+    },
+    {
+      "epoch": 0.11818947368421052,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00048771716121267605,
+      "loss": 3.552,
+      "step": 2807
+    },
+    {
+      "epoch": 0.11823157894736842,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004877067119995461,
+      "loss": 4.2121,
+      "step": 2808
+    },
+    {
+      "epoch": 0.11827368421052632,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.000487696258455679,
+      "loss": 3.6594,
+      "step": 2809
+    },
+    {
+      "epoch": 0.11831578947368421,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004876858005812654,
+      "loss": 2.9684,
+      "step": 2810
+    },
+    {
+      "epoch": 0.1183578947368421,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004876753383764958,
+      "loss": 3.3871,
+      "step": 2811
+    },
+    {
+      "epoch": 0.1184,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004876648718415608,
+      "loss": 3.8453,
+      "step": 2812
+    },
+    {
+      "epoch": 0.11844210526315789,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00048765440097665104,
+      "loss": 3.3664,
+      "step": 2813
+    },
+    {
+      "epoch": 0.11848421052631579,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004876439257819574,
+      "loss": 3.5623,
+      "step": 2814
+    },
+    {
+      "epoch": 0.11852631578947369,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004876334462576706,
+      "loss": 3.5241,
+      "step": 2815
+    },
+    {
+      "epoch": 0.11856842105263157,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00048762296240398165,
+      "loss": 3.3805,
+      "step": 2816
+    },
+    {
+      "epoch": 0.11861052631578947,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00048761247422108146,
+      "loss": 3.6599,
+      "step": 2817
+    },
+    {
+      "epoch": 0.11865263157894737,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004876019817091612,
+      "loss": 3.5569,
+      "step": 2818
+    },
+    {
+      "epoch": 0.11869473684210527,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00048759148486841197,
+      "loss": 3.6372,
+      "step": 2819
+    },
+    {
+      "epoch": 0.11873684210526315,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00048758098369902517,
+      "loss": 3.1221,
+      "step": 2820
+    },
+    {
+      "epoch": 0.11877894736842105,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004875704782011918,
+      "loss": 3.2028,
+      "step": 2821
+    },
+    {
+      "epoch": 0.11882105263157895,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004875599683751036,
+      "loss": 3.3473,
+      "step": 2822
+    },
+    {
+      "epoch": 0.11886315789473684,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00048754945422095186,
+      "loss": 3.5971,
+      "step": 2823
+    },
+    {
+      "epoch": 0.11890526315789474,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00048753893573892816,
+      "loss": 3.4471,
+      "step": 2824
+    },
+    {
+      "epoch": 0.11894736842105263,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004875284129292242,
+      "loss": 3.1798,
+      "step": 2825
+    },
+    {
+      "epoch": 0.11898947368421052,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004875178857920316,
+      "loss": 3.0938,
+      "step": 2826
+    },
+    {
+      "epoch": 0.11903157894736842,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00048750735432754217,
+      "loss": 3.5277,
+      "step": 2827
+    },
+    {
+      "epoch": 0.11907368421052632,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00048749681853594787,
+      "loss": 3.5053,
+      "step": 2828
+    },
+    {
+      "epoch": 0.11911578947368422,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00048748627841744055,
+      "loss": 3.189,
+      "step": 2829
+    },
+    {
+      "epoch": 0.1191578947368421,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00048747573397221224,
+      "loss": 3.4325,
+      "step": 2830
+    },
+    {
+      "epoch": 0.1192,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0004874651852004551,
+      "loss": 3.7074,
+      "step": 2831
+    },
+    {
+      "epoch": 0.1192421052631579,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004874546321023613,
+      "loss": 3.5767,
+      "step": 2832
+    },
+    {
+      "epoch": 0.11928421052631578,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004874440746781231,
+      "loss": 3.9523,
+      "step": 2833
+    },
+    {
+      "epoch": 0.11932631578947368,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00048743351292793283,
+      "loss": 3.741,
+      "step": 2834
+    },
+    {
+      "epoch": 0.11936842105263158,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004874229468519829,
+      "loss": 3.6629,
+      "step": 2835
+    },
+    {
+      "epoch": 0.11941052631578947,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00048741237645046585,
+      "loss": 3.4801,
+      "step": 2836
+    },
+    {
+      "epoch": 0.11945263157894737,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004874018017235743,
+      "loss": 3.4843,
+      "step": 2837
+    },
+    {
+      "epoch": 0.11949473684210526,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004873912226715008,
+      "loss": 3.7619,
+      "step": 2838
+    },
+    {
+      "epoch": 0.11953684210526316,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00048738063929443823,
+      "loss": 3.7195,
+      "step": 2839
+    },
+    {
+      "epoch": 0.11957894736842105,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004873700515925793,
+      "loss": 3.4025,
+      "step": 2840
+    },
+    {
+      "epoch": 0.11962105263157895,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00048735945956611686,
+      "loss": 3.4672,
+      "step": 2841
+    },
+    {
+      "epoch": 0.11966315789473685,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00048734886321524405,
+      "loss": 3.3723,
+      "step": 2842
+    },
+    {
+      "epoch": 0.11970526315789473,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004873382625401537,
+      "loss": 3.6559,
+      "step": 2843
+    },
+    {
+      "epoch": 0.11974736842105263,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00048732765754103916,
+      "loss": 3.6285,
+      "step": 2844
+    },
+    {
+      "epoch": 0.11978947368421053,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00048731704821809355,
+      "loss": 3.5868,
+      "step": 2845
+    },
+    {
+      "epoch": 0.11983157894736843,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00048730643457151014,
+      "loss": 3.3669,
+      "step": 2846
+    },
+    {
+      "epoch": 0.11987368421052631,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00048729581660148224,
+      "loss": 3.4807,
+      "step": 2847
+    },
+    {
+      "epoch": 0.11991578947368421,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004872851943082035,
+      "loss": 3.4516,
+      "step": 2848
+    },
+    {
+      "epoch": 0.11995789473684211,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0004872745676918673,
+      "loss": 3.4762,
+      "step": 2849
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0004872639367526672,
+      "loss": 3.5862,
+      "step": 2850
+    },
+    {
+      "epoch": 0.1200421052631579,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00048725330149079695,
+      "loss": 3.5795,
+      "step": 2851
+    },
+    {
+      "epoch": 0.1200842105263158,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0004872426619064503,
+      "loss": 3.6206,
+      "step": 2852
+    },
+    {
+      "epoch": 0.12012631578947368,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00048723201799982113,
+      "loss": 3.2186,
+      "step": 2853
+    },
+    {
+      "epoch": 0.12016842105263158,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004872213697711033,
+      "loss": 3.8518,
+      "step": 2854
+    },
+    {
+      "epoch": 0.12021052631578948,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0004872107172204909,
+      "loss": 3.7301,
+      "step": 2855
+    },
+    {
+      "epoch": 0.12025263157894737,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00048720006034817787,
+      "loss": 3.5111,
+      "step": 2856
+    },
+    {
+      "epoch": 0.12029473684210526,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0004871893991543585,
+      "loss": 3.4307,
+      "step": 2857
+    },
+    {
+      "epoch": 0.12033684210526316,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00048717873363922695,
+      "loss": 2.8608,
+      "step": 2858
+    },
+    {
+      "epoch": 0.12037894736842106,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004871680638029775,
+      "loss": 3.2442,
+      "step": 2859
+    },
+    {
+      "epoch": 0.12042105263157894,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00048715738964580466,
+      "loss": 3.6237,
+      "step": 2860
+    },
+    {
+      "epoch": 0.12046315789473684,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00048714671116790277,
+      "loss": 3.688,
+      "step": 2861
+    },
+    {
+      "epoch": 0.12050526315789474,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004871360283694665,
+      "loss": 3.6212,
+      "step": 2862
+    },
+    {
+      "epoch": 0.12054736842105263,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004871253412506903,
+      "loss": 3.434,
+      "step": 2863
+    },
+    {
+      "epoch": 0.12058947368421052,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00048711464981176914,
+      "loss": 3.51,
+      "step": 2864
+    },
+    {
+      "epoch": 0.12063157894736842,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004871039540528975,
+      "loss": 3.8943,
+      "step": 2865
+    },
+    {
+      "epoch": 0.12067368421052632,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0004870932539742705,
+      "loss": 3.6839,
+      "step": 2866
+    },
+    {
+      "epoch": 0.1207157894736842,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.000487082549576083,
+      "loss": 3.286,
+      "step": 2867
+    },
+    {
+      "epoch": 0.1207578947368421,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00048707184085853,
+      "loss": 3.9478,
+      "step": 2868
+    },
+    {
+      "epoch": 0.1208,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004870611278218066,
+      "loss": 3.3878,
+      "step": 2869
+    },
+    {
+      "epoch": 0.12084210526315789,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.000487050410466108,
+      "loss": 3.9745,
+      "step": 2870
+    },
+    {
+      "epoch": 0.12088421052631579,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0004870396887916294,
+      "loss": 3.292,
+      "step": 2871
+    },
+    {
+      "epoch": 0.12092631578947369,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004870289627985662,
+      "loss": 3.6294,
+      "step": 2872
+    },
+    {
+      "epoch": 0.12096842105263157,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0004870182324871138,
+      "loss": 3.6152,
+      "step": 2873
+    },
+    {
+      "epoch": 0.12101052631578947,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0004870074978574677,
+      "loss": 3.1454,
+      "step": 2874
+    },
+    {
+      "epoch": 0.12105263157894737,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00048699675890982345,
+      "loss": 3.2372,
+      "step": 2875
+    },
+    {
+      "epoch": 0.12109473684210527,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004869860156443767,
+      "loss": 3.4052,
+      "step": 2876
+    },
+    {
+      "epoch": 0.12113684210526315,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004869752680613232,
+      "loss": 3.4905,
+      "step": 2877
+    },
+    {
+      "epoch": 0.12117894736842105,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00048696451616085873,
+      "loss": 3.5956,
+      "step": 2878
+    },
+    {
+      "epoch": 0.12122105263157895,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004869537599431793,
+      "loss": 3.4402,
+      "step": 2879
+    },
+    {
+      "epoch": 0.12126315789473684,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00048694299940848064,
+      "loss": 3.4183,
+      "step": 2880
+    },
+    {
+      "epoch": 0.12130526315789474,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00048693223455695896,
+      "loss": 3.8453,
+      "step": 2881
+    },
+    {
+      "epoch": 0.12134736842105263,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00048692146538881036,
+      "loss": 3.2207,
+      "step": 2882
+    },
+    {
+      "epoch": 0.12138947368421053,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.000486910691904231,
+      "loss": 3.7992,
+      "step": 2883
+    },
+    {
+      "epoch": 0.12143157894736842,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00048689991410341726,
+      "loss": 3.466,
+      "step": 2884
+    },
+    {
+      "epoch": 0.12147368421052632,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004868891319865654,
+      "loss": 3.4106,
+      "step": 2885
+    },
+    {
+      "epoch": 0.12151578947368422,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004868783455538719,
+      "loss": 3.4247,
+      "step": 2886
+    },
+    {
+      "epoch": 0.1215578947368421,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00048686755480553326,
+      "loss": 3.6108,
+      "step": 2887
+    },
+    {
+      "epoch": 0.1216,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004868567597417461,
+      "loss": 3.4444,
+      "step": 2888
+    },
+    {
+      "epoch": 0.1216421052631579,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.000486845960362707,
+      "loss": 3.5251,
+      "step": 2889
+    },
+    {
+      "epoch": 0.12168421052631578,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004868351566686129,
+      "loss": 3.5891,
+      "step": 2890
+    },
+    {
+      "epoch": 0.12172631578947368,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004868243486596604,
+      "loss": 3.5605,
+      "step": 2891
+    },
+    {
+      "epoch": 0.12176842105263158,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00048681353633604665,
+      "loss": 3.6923,
+      "step": 2892
+    },
+    {
+      "epoch": 0.12181052631578948,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004868027196979684,
+      "loss": 3.2688,
+      "step": 2893
+    },
+    {
+      "epoch": 0.12185263157894737,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00048679189874562283,
+      "loss": 3.2256,
+      "step": 2894
+    },
+    {
+      "epoch": 0.12189473684210526,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004867810734792072,
+      "loss": 3.4213,
+      "step": 2895
+    },
+    {
+      "epoch": 0.12193684210526316,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004867702438989185,
+      "loss": 3.5332,
+      "step": 2896
+    },
+    {
+      "epoch": 0.12197894736842105,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00048675941000495416,
+      "loss": 3.1365,
+      "step": 2897
+    },
+    {
+      "epoch": 0.12202105263157895,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004867485717975116,
+      "loss": 3.3783,
+      "step": 2898
+    },
+    {
+      "epoch": 0.12206315789473685,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00048673772927678815,
+      "loss": 3.6153,
+      "step": 2899
+    },
+    {
+      "epoch": 0.12210526315789473,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004867268824429815,
+      "loss": 3.749,
+      "step": 2900
+    },
+    {
+      "epoch": 0.12214736842105263,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00048671603129628923,
+      "loss": 3.7498,
+      "step": 2901
+    },
+    {
+      "epoch": 0.12218947368421053,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004867051758369089,
+      "loss": 3.1852,
+      "step": 2902
+    },
+    {
+      "epoch": 0.12223157894736843,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0004866943160650384,
+      "loss": 3.7043,
+      "step": 2903
+    },
+    {
+      "epoch": 0.12227368421052631,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00048668345198087565,
+      "loss": 3.7886,
+      "step": 2904
+    },
+    {
+      "epoch": 0.12231578947368421,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00048667258358461846,
+      "loss": 3.3222,
+      "step": 2905
+    },
+    {
+      "epoch": 0.12235789473684211,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00048666171087646484,
+      "loss": 3.5955,
+      "step": 2906
+    },
+    {
+      "epoch": 0.1224,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.000486650833856613,
+      "loss": 3.2265,
+      "step": 2907
+    },
+    {
+      "epoch": 0.1224421052631579,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00048663995252526095,
+      "loss": 3.2892,
+      "step": 2908
+    },
+    {
+      "epoch": 0.12248421052631579,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.000486629066882607,
+      "loss": 3.7501,
+      "step": 2909
+    },
+    {
+      "epoch": 0.12252631578947368,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00048661817692884947,
+      "loss": 3.7994,
+      "step": 2910
+    },
+    {
+      "epoch": 0.12256842105263158,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00048660728266418683,
+      "loss": 4.1502,
+      "step": 2911
+    },
+    {
+      "epoch": 0.12261052631578948,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00048659638408881745,
+      "loss": 2.9858,
+      "step": 2912
+    },
+    {
+      "epoch": 0.12265263157894737,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00048658548120293995,
+      "loss": 3.5554,
+      "step": 2913
+    },
+    {
+      "epoch": 0.12269473684210526,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00048657457400675297,
+      "loss": 3.1423,
+      "step": 2914
+    },
+    {
+      "epoch": 0.12273684210526316,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004865636625004551,
+      "loss": 2.8558,
+      "step": 2915
+    },
+    {
+      "epoch": 0.12277894736842106,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004865527466842454,
+      "loss": 3.8267,
+      "step": 2916
+    },
+    {
+      "epoch": 0.12282105263157894,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00048654182655832255,
+      "loss": 3.6986,
+      "step": 2917
+    },
+    {
+      "epoch": 0.12286315789473684,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004865309021228856,
+      "loss": 3.452,
+      "step": 2918
+    },
+    {
+      "epoch": 0.12290526315789474,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00048651997337813345,
+      "loss": 3.5154,
+      "step": 2919
+    },
+    {
+      "epoch": 0.12294736842105262,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00048650904032426534,
+      "loss": 3.613,
+      "step": 2920
+    },
+    {
+      "epoch": 0.12298947368421052,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00048649810296148035,
+      "loss": 3.2096,
+      "step": 2921
+    },
+    {
+      "epoch": 0.12303157894736842,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004864871612899778,
+      "loss": 3.2185,
+      "step": 2922
+    },
+    {
+      "epoch": 0.12307368421052632,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00048647621530995705,
+      "loss": 3.4172,
+      "step": 2923
+    },
+    {
+      "epoch": 0.1231157894736842,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00048646526502161755,
+      "loss": 3.9241,
+      "step": 2924
+    },
+    {
+      "epoch": 0.1231578947368421,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00048645431042515866,
+      "loss": 3.54,
+      "step": 2925
+    },
+    {
+      "epoch": 0.1232,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004864433515207801,
+      "loss": 3.7357,
+      "step": 2926
+    },
+    {
+      "epoch": 0.12324210526315789,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0004864323883086815,
+      "loss": 3.7652,
+      "step": 2927
+    },
+    {
+      "epoch": 0.12328421052631579,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004864214207890626,
+      "loss": 3.8698,
+      "step": 2928
+    },
+    {
+      "epoch": 0.12332631578947369,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0004864104489621232,
+      "loss": 3.6557,
+      "step": 2929
+    },
+    {
+      "epoch": 0.12336842105263159,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00048639947282806315,
+      "loss": 3.3412,
+      "step": 2930
+    },
+    {
+      "epoch": 0.12341052631578947,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004863884923870825,
+      "loss": 3.2352,
+      "step": 2931
+    },
+    {
+      "epoch": 0.12345263157894737,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0004863775076393812,
+      "loss": 3.5197,
+      "step": 2932
+    },
+    {
+      "epoch": 0.12349473684210527,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00048636651858515947,
+      "loss": 3.3437,
+      "step": 2933
+    },
+    {
+      "epoch": 0.12353684210526315,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004863555252246175,
+      "loss": 3.5777,
+      "step": 2934
+    },
+    {
+      "epoch": 0.12357894736842105,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00048634452755795554,
+      "loss": 3.2911,
+      "step": 2935
+    },
+    {
+      "epoch": 0.12362105263157895,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00048633352558537403,
+      "loss": 3.523,
+      "step": 2936
+    },
+    {
+      "epoch": 0.12366315789473684,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004863225193070734,
+      "loss": 3.7026,
+      "step": 2937
+    },
+    {
+      "epoch": 0.12370526315789473,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.000486311508723254,
+      "loss": 3.6308,
+      "step": 2938
+    },
+    {
+      "epoch": 0.12374736842105263,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00048630049383411667,
+      "loss": 2.9412,
+      "step": 2939
+    },
+    {
+      "epoch": 0.12378947368421053,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00048628947463986197,
+      "loss": 3.5675,
+      "step": 2940
+    },
+    {
+      "epoch": 0.12383157894736842,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004862784511406907,
+      "loss": 3.6594,
+      "step": 2941
+    },
+    {
+      "epoch": 0.12387368421052632,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0004862674233368036,
+      "loss": 3.6711,
+      "step": 2942
+    },
+    {
+      "epoch": 0.12391578947368422,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004862563912284017,
+      "loss": 3.1911,
+      "step": 2943
+    },
+    {
+      "epoch": 0.1239578947368421,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004862453548156859,
+      "loss": 4.2802,
+      "step": 2944
+    },
+    {
+      "epoch": 0.124,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0004862343140988573,
+      "loss": 3.4156,
+      "step": 2945
+    },
+    {
+      "epoch": 0.1240421052631579,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004862232690781171,
+      "loss": 3.6335,
+      "step": 2946
+    },
+    {
+      "epoch": 0.12408421052631578,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00048621221975366646,
+      "loss": 3.6205,
+      "step": 2947
+    },
+    {
+      "epoch": 0.12412631578947368,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00048620116612570674,
+      "loss": 3.3457,
+      "step": 2948
+    },
+    {
+      "epoch": 0.12416842105263158,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004861901081944393,
+      "loss": 3.4073,
+      "step": 2949
+    },
+    {
+      "epoch": 0.12421052631578948,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004861790459600656,
+      "loss": 3.4518,
+      "step": 2950
+    },
+    {
+      "epoch": 0.12425263157894736,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004861679794227872,
+      "loss": 3.3027,
+      "step": 2951
+    },
+    {
+      "epoch": 0.12429473684210526,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004861569085828057,
+      "loss": 3.0642,
+      "step": 2952
+    },
+    {
+      "epoch": 0.12433684210526316,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004861458334403227,
+      "loss": 3.5175,
+      "step": 2953
+    },
+    {
+      "epoch": 0.12437894736842105,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004861347539955402,
+      "loss": 3.9531,
+      "step": 2954
+    },
+    {
+      "epoch": 0.12442105263157895,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004861236702486599,
+      "loss": 3.1812,
+      "step": 2955
+    },
+    {
+      "epoch": 0.12446315789473684,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0004861125821998837,
+      "loss": 3.2253,
+      "step": 2956
+    },
+    {
+      "epoch": 0.12450526315789473,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.0004861014898494137,
+      "loss": 3.2094,
+      "step": 2957
+    },
+    {
+      "epoch": 0.12454736842105263,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.000486090393197452,
+      "loss": 3.2554,
+      "step": 2958
+    },
+    {
+      "epoch": 0.12458947368421053,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004860792922442007,
+      "loss": 3.3683,
+      "step": 2959
+    },
+    {
+      "epoch": 0.12463157894736843,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0004860681869898621,
+      "loss": 2.9148,
+      "step": 2960
+    },
+    {
+      "epoch": 0.12467368421052631,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004860570774346384,
+      "loss": 3.3476,
+      "step": 2961
+    },
+    {
+      "epoch": 0.12471578947368421,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00048604596357873227,
+      "loss": 3.367,
+      "step": 2962
+    },
+    {
+      "epoch": 0.12475789473684211,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004860348454223459,
+      "loss": 3.3253,
+      "step": 2963
+    },
+    {
+      "epoch": 0.1248,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00048602372296568204,
+      "loss": 3.1474,
+      "step": 2964
+    },
+    {
+      "epoch": 0.1248421052631579,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00048601259620894324,
+      "loss": 3.5001,
+      "step": 2965
+    },
+    {
+      "epoch": 0.12488421052631579,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00048600146515233224,
+      "loss": 3.3656,
+      "step": 2966
+    },
+    {
+      "epoch": 0.12492631578947369,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00048599032979605185,
+      "loss": 3.1679,
+      "step": 2967
+    },
+    {
+      "epoch": 0.12496842105263158,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004859791901403049,
+      "loss": 3.3416,
+      "step": 2968
+    },
+    {
+      "epoch": 0.12501052631578946,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00048596804618529445,
+      "loss": 3.5502,
+      "step": 2969
+    },
+    {
+      "epoch": 0.12505263157894736,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0004859568979312233,
+      "loss": 3.7684,
+      "step": 2970
+    },
+    {
+      "epoch": 0.12509473684210526,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00048594574537829484,
+      "loss": 3.3721,
+      "step": 2971
+    },
+    {
+      "epoch": 0.12513684210526316,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0004859345885267121,
+      "loss": 3.3118,
+      "step": 2972
+    },
+    {
+      "epoch": 0.12517894736842106,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00048592342737667835,
+      "loss": 3.5161,
+      "step": 2973
+    },
+    {
+      "epoch": 0.12522105263157896,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00048591226192839696,
+      "loss": 3.6436,
+      "step": 2974
+    },
+    {
+      "epoch": 0.12526315789473685,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00048590109218207134,
+      "loss": 3.4589,
+      "step": 2975
+    },
+    {
+      "epoch": 0.12530526315789473,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00048588991813790493,
+      "loss": 3.3623,
+      "step": 2976
+    },
+    {
+      "epoch": 0.12534736842105262,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004858787397961014,
+      "loss": 3.6315,
+      "step": 2977
+    },
+    {
+      "epoch": 0.12538947368421052,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004858675571568645,
+      "loss": 3.6184,
+      "step": 2978
+    },
+    {
+      "epoch": 0.12543157894736842,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00048585637022039775,
+      "loss": 3.2735,
+      "step": 2979
+    },
+    {
+      "epoch": 0.12547368421052632,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.000485845178986905,
+      "loss": 3.8127,
+      "step": 2980
+    },
+    {
+      "epoch": 0.12551578947368422,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0004858339834565902,
+      "loss": 3.6341,
+      "step": 2981
+    },
+    {
+      "epoch": 0.12555789473684212,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00048582278362965743,
+      "loss": 3.5892,
+      "step": 2982
+    },
+    {
+      "epoch": 0.1256,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00048581157950631056,
+      "loss": 3.3656,
+      "step": 2983
+    },
+    {
+      "epoch": 0.1256421052631579,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00048580037108675376,
+      "loss": 3.484,
+      "step": 2984
+    },
+    {
+      "epoch": 0.1256842105263158,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00048578915837119125,
+      "loss": 3.3778,
+      "step": 2985
+    },
+    {
+      "epoch": 0.12572631578947369,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00048577794135982734,
+      "loss": 3.1639,
+      "step": 2986
+    },
+    {
+      "epoch": 0.12576842105263158,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004857667200528664,
+      "loss": 3.4191,
+      "step": 2987
+    },
+    {
+      "epoch": 0.12581052631578948,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00048575549445051275,
+      "loss": 3.4421,
+      "step": 2988
+    },
+    {
+      "epoch": 0.12585263157894735,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00048574426455297103,
+      "loss": 3.8453,
+      "step": 2989
+    },
+    {
+      "epoch": 0.12589473684210525,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004857330303604458,
+      "loss": 3.3646,
+      "step": 2990
+    },
+    {
+      "epoch": 0.12593684210526315,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00048572179187314176,
+      "loss": 2.9505,
+      "step": 2991
+    },
+    {
+      "epoch": 0.12597894736842105,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00048571054909126354,
+      "loss": 3.3229,
+      "step": 2992
+    },
+    {
+      "epoch": 0.12602105263157895,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004856993020150162,
+      "loss": 3.2146,
+      "step": 2993
+    },
+    {
+      "epoch": 0.12606315789473685,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004856880506446044,
+      "loss": 3.8323,
+      "step": 2994
+    },
+    {
+      "epoch": 0.12610526315789475,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004856767949802333,
+      "loss": 3.2727,
+      "step": 2995
+    },
+    {
+      "epoch": 0.12614736842105262,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00048566553502210787,
+      "loss": 3.3491,
+      "step": 2996
+    },
+    {
+      "epoch": 0.12618947368421052,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00048565427077043333,
+      "loss": 3.8218,
+      "step": 2997
+    },
+    {
+      "epoch": 0.12623157894736842,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004856430022254148,
+      "loss": 3.9196,
+      "step": 2998
+    },
+    {
+      "epoch": 0.12627368421052632,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00048563172938725777,
+      "loss": 3.5469,
+      "step": 2999
+    },
+    {
+      "epoch": 0.12631578947368421,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00048562045225616733,
+      "loss": 3.2952,
+      "step": 3000
+    },
+    {
+      "epoch": 0.12631578947368421,
+      "eval_loss": 3.4825093746185303,
+      "eval_runtime": 345.0058,
+      "eval_samples_per_second": 43.478,
+      "eval_steps_per_second": 5.435,
+      "step": 3000
+    },
+    {
+      "epoch": 0.1263578947368421,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0004856091708323491,
+      "loss": 3.5591,
+      "step": 3001
+    },
+    {
+      "epoch": 0.1264,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00048559788511600875,
+      "loss": 3.7928,
+      "step": 3002
+    },
+    {
+      "epoch": 0.12644210526315788,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004855865951073516,
+      "loss": 3.5254,
+      "step": 3003
+    },
+    {
+      "epoch": 0.12648421052631578,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00048557530080658355,
+      "loss": 3.3097,
+      "step": 3004
+    },
+    {
+      "epoch": 0.12652631578947368,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004855640022139103,
+      "loss": 3.6043,
+      "step": 3005
+    },
+    {
+      "epoch": 0.12656842105263158,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00048555269932953774,
+      "loss": 3.6455,
+      "step": 3006
+    },
+    {
+      "epoch": 0.12661052631578948,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00048554139215367176,
+      "loss": 3.27,
+      "step": 3007
+    },
+    {
+      "epoch": 0.12665263157894738,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00048553008068651825,
+      "loss": 3.0601,
+      "step": 3008
+    },
+    {
+      "epoch": 0.12669473684210528,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004855187649282836,
+      "loss": 3.8757,
+      "step": 3009
+    },
+    {
+      "epoch": 0.12673684210526315,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00048550744487917363,
+      "loss": 3.4292,
+      "step": 3010
+    },
+    {
+      "epoch": 0.12677894736842105,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00048549612053939474,
+      "loss": 3.6556,
+      "step": 3011
+    },
+    {
+      "epoch": 0.12682105263157895,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00048548479190915327,
+      "loss": 3.515,
+      "step": 3012
+    },
+    {
+      "epoch": 0.12686315789473684,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00048547345898865556,
+      "loss": 3.6684,
+      "step": 3013
+    },
+    {
+      "epoch": 0.12690526315789474,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004854621217781081,
+      "loss": 3.6245,
+      "step": 3014
+    },
+    {
+      "epoch": 0.12694736842105264,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00048545078027771745,
+      "loss": 3.3477,
+      "step": 3015
+    },
+    {
+      "epoch": 0.1269894736842105,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00048543943448769024,
+      "loss": 3.7547,
+      "step": 3016
+    },
+    {
+      "epoch": 0.1270315789473684,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004854280844082331,
+      "loss": 3.2349,
+      "step": 3017
+    },
+    {
+      "epoch": 0.1270736842105263,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0004854167300395529,
+      "loss": 3.1902,
+      "step": 3018
+    },
+    {
+      "epoch": 0.1271157894736842,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004854053713818565,
+      "loss": 3.125,
+      "step": 3019
+    },
+    {
+      "epoch": 0.1271578947368421,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0004853940084353509,
+      "loss": 3.5249,
+      "step": 3020
+    },
+    {
+      "epoch": 0.1272,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004853826412002429,
+      "loss": 3.5504,
+      "step": 3021
+    },
+    {
+      "epoch": 0.1272421052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00048537126967673977,
+      "loss": 3.4727,
+      "step": 3022
+    },
+    {
+      "epoch": 0.12728421052631578,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004853598938650487,
+      "loss": 3.4245,
+      "step": 3023
+    },
+    {
+      "epoch": 0.12732631578947368,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004853485137653769,
+      "loss": 3.3179,
+      "step": 3024
+    },
+    {
+      "epoch": 0.12736842105263158,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004853371293779317,
+      "loss": 3.5749,
+      "step": 3025
+    },
+    {
+      "epoch": 0.12741052631578947,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004853257407029205,
+      "loss": 3.7831,
+      "step": 3026
+    },
+    {
+      "epoch": 0.12745263157894737,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00048531434774055085,
+      "loss": 3.2463,
+      "step": 3027
+    },
+    {
+      "epoch": 0.12749473684210527,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004853029504910302,
+      "loss": 3.6175,
+      "step": 3028
+    },
+    {
+      "epoch": 0.12753684210526317,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00048529154895456627,
+      "loss": 3.7884,
+      "step": 3029
+    },
+    {
+      "epoch": 0.12757894736842104,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004852801431313668,
+      "loss": 3.0851,
+      "step": 3030
+    },
+    {
+      "epoch": 0.12762105263157894,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00048526873302163955,
+      "loss": 3.7011,
+      "step": 3031
+    },
+    {
+      "epoch": 0.12766315789473684,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004852573186255924,
+      "loss": 3.18,
+      "step": 3032
+    },
+    {
+      "epoch": 0.12770526315789474,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00048524589994343334,
+      "loss": 3.4369,
+      "step": 3033
+    },
+    {
+      "epoch": 0.12774736842105264,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00048523447697537036,
+      "loss": 3.3593,
+      "step": 3034
+    },
+    {
+      "epoch": 0.12778947368421054,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004852230497216116,
+      "loss": 3.5531,
+      "step": 3035
+    },
+    {
+      "epoch": 0.1278315789473684,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004852116181823653,
+      "loss": 3.7948,
+      "step": 3036
+    },
+    {
+      "epoch": 0.1278736842105263,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004852001823578397,
+      "loss": 3.771,
+      "step": 3037
+    },
+    {
+      "epoch": 0.1279157894736842,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00048518874224824307,
+      "loss": 3.5598,
+      "step": 3038
+    },
+    {
+      "epoch": 0.1279578947368421,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00048517729785378384,
+      "loss": 3.2631,
+      "step": 3039
+    },
+    {
+      "epoch": 0.128,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00048516584917467065,
+      "loss": 3.7522,
+      "step": 3040
+    },
+    {
+      "epoch": 0.1280421052631579,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00048515439621111194,
+      "loss": 3.7963,
+      "step": 3041
+    },
+    {
+      "epoch": 0.1280842105263158,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00048514293896331653,
+      "loss": 3.4834,
+      "step": 3042
+    },
+    {
+      "epoch": 0.12812631578947367,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00048513147743149286,
+      "loss": 3.4996,
+      "step": 3043
+    },
+    {
+      "epoch": 0.12816842105263157,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004851200116158501,
+      "loss": 3.6002,
+      "step": 3044
+    },
+    {
+      "epoch": 0.12821052631578947,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.000485108541516597,
+      "loss": 3.4594,
+      "step": 3045
+    },
+    {
+      "epoch": 0.12825263157894737,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00048509706713394243,
+      "loss": 3.5628,
+      "step": 3046
+    },
+    {
+      "epoch": 0.12829473684210527,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004850855884680956,
+      "loss": 3.5081,
+      "step": 3047
+    },
+    {
+      "epoch": 0.12833684210526317,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00048507410551926547,
+      "loss": 3.8763,
+      "step": 3048
+    },
+    {
+      "epoch": 0.12837894736842106,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004850626182876614,
+      "loss": 3.5405,
+      "step": 3049
+    },
+    {
+      "epoch": 0.12842105263157894,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00048505112677349254,
+      "loss": 3.4144,
+      "step": 3050
+    },
+    {
+      "epoch": 0.12846315789473683,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004850396309769684,
+      "loss": 3.5738,
+      "step": 3051
+    },
+    {
+      "epoch": 0.12850526315789473,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004850281308982983,
+      "loss": 3.5143,
+      "step": 3052
+    },
+    {
+      "epoch": 0.12854736842105263,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00048501662653769185,
+      "loss": 3.697,
+      "step": 3053
+    },
+    {
+      "epoch": 0.12858947368421053,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00048500511789535856,
+      "loss": 3.5978,
+      "step": 3054
+    },
+    {
+      "epoch": 0.12863157894736843,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00048499360497150813,
+      "loss": 3.4588,
+      "step": 3055
+    },
+    {
+      "epoch": 0.12867368421052633,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004849820877663504,
+      "loss": 3.4352,
+      "step": 3056
+    },
+    {
+      "epoch": 0.1287157894736842,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00048497056628009507,
+      "loss": 3.4422,
+      "step": 3057
+    },
+    {
+      "epoch": 0.1287578947368421,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00048495904051295207,
+      "loss": 3.2563,
+      "step": 3058
+    },
+    {
+      "epoch": 0.1288,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00048494751046513143,
+      "loss": 3.4515,
+      "step": 3059
+    },
+    {
+      "epoch": 0.1288421052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00048493597613684326,
+      "loss": 3.4291,
+      "step": 3060
+    },
+    {
+      "epoch": 0.1288842105263158,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004849244375282976,
+      "loss": 3.293,
+      "step": 3061
+    },
+    {
+      "epoch": 0.1289263157894737,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00048491289463970465,
+      "loss": 3.4308,
+      "step": 3062
+    },
+    {
+      "epoch": 0.12896842105263157,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004849013474712749,
+      "loss": 3.5792,
+      "step": 3063
+    },
+    {
+      "epoch": 0.12901052631578946,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004848897960232185,
+      "loss": 3.3259,
+      "step": 3064
+    },
+    {
+      "epoch": 0.12905263157894736,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004848782402957461,
+      "loss": 3.366,
+      "step": 3065
+    },
+    {
+      "epoch": 0.12909473684210526,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.000484866680289068,
+      "loss": 3.274,
+      "step": 3066
+    },
+    {
+      "epoch": 0.12913684210526316,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00048485511600339506,
+      "loss": 3.1575,
+      "step": 3067
+    },
+    {
+      "epoch": 0.12917894736842106,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00048484354743893776,
+      "loss": 3.4722,
+      "step": 3068
+    },
+    {
+      "epoch": 0.12922105263157896,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.000484831974595907,
+      "loss": 3.0451,
+      "step": 3069
+    },
+    {
+      "epoch": 0.12926315789473683,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00048482039747451354,
+      "loss": 3.2021,
+      "step": 3070
+    },
+    {
+      "epoch": 0.12930526315789473,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004848088160749684,
+      "loss": 3.6545,
+      "step": 3071
+    },
+    {
+      "epoch": 0.12934736842105263,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0004847972303974825,
+      "loss": 3.2949,
+      "step": 3072
+    },
+    {
+      "epoch": 0.12938947368421053,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00048478564044226693,
+      "loss": 3.1198,
+      "step": 3073
+    },
+    {
+      "epoch": 0.12943157894736843,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004847740462095329,
+      "loss": 3.5714,
+      "step": 3074
+    },
+    {
+      "epoch": 0.12947368421052632,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00048476244769949154,
+      "loss": 3.0656,
+      "step": 3075
+    },
+    {
+      "epoch": 0.12951578947368422,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004847508449123542,
+      "loss": 3.8121,
+      "step": 3076
+    },
+    {
+      "epoch": 0.1295578947368421,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004847392378483323,
+      "loss": 3.7964,
+      "step": 3077
+    },
+    {
+      "epoch": 0.1296,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004847276265076373,
+      "loss": 3.32,
+      "step": 3078
+    },
+    {
+      "epoch": 0.1296421052631579,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004847160108904808,
+      "loss": 3.8884,
+      "step": 3079
+    },
+    {
+      "epoch": 0.1296842105263158,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00048470439099707426,
+      "loss": 3.9961,
+      "step": 3080
+    },
+    {
+      "epoch": 0.1297263157894737,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00048469276682762955,
+      "loss": 3.2673,
+      "step": 3081
+    },
+    {
+      "epoch": 0.1297684210526316,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004846811383823584,
+      "loss": 3.7614,
+      "step": 3082
+    },
+    {
+      "epoch": 0.1298105263157895,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00048466950566147256,
+      "loss": 3.625,
+      "step": 3083
+    },
+    {
+      "epoch": 0.12985263157894736,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00048465786866518415,
+      "loss": 3.2057,
+      "step": 3084
+    },
+    {
+      "epoch": 0.12989473684210526,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00048464622739370503,
+      "loss": 3.5994,
+      "step": 3085
+    },
+    {
+      "epoch": 0.12993684210526316,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004846345818472473,
+      "loss": 3.3678,
+      "step": 3086
+    },
+    {
+      "epoch": 0.12997894736842106,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00048462293202602324,
+      "loss": 3.6239,
+      "step": 3087
+    },
+    {
+      "epoch": 0.13002105263157895,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00048461127793024503,
+      "loss": 3.7816,
+      "step": 3088
+    },
+    {
+      "epoch": 0.13006315789473685,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00048459961956012494,
+      "loss": 3.4049,
+      "step": 3089
+    },
+    {
+      "epoch": 0.13010526315789472,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00048458795691587545,
+      "loss": 3.3318,
+      "step": 3090
+    },
+    {
+      "epoch": 0.13014736842105262,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.000484576289997709,
+      "loss": 3.1791,
+      "step": 3091
+    },
+    {
+      "epoch": 0.13018947368421052,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00048456461880583827,
+      "loss": 3.7217,
+      "step": 3092
+    },
+    {
+      "epoch": 0.13023157894736842,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004845529433404757,
+      "loss": 3.5749,
+      "step": 3093
+    },
+    {
+      "epoch": 0.13027368421052632,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00048454126360183405,
+      "loss": 3.7437,
+      "step": 3094
+    },
+    {
+      "epoch": 0.13031578947368422,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004845295795901262,
+      "loss": 3.2497,
+      "step": 3095
+    },
+    {
+      "epoch": 0.13035789473684212,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.000484517891305565,
+      "loss": 3.5373,
+      "step": 3096
+    },
+    {
+      "epoch": 0.1304,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004845061987483633,
+      "loss": 3.5362,
+      "step": 3097
+    },
+    {
+      "epoch": 0.1304421052631579,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00048449450191873425,
+      "loss": 3.3323,
+      "step": 3098
+    },
+    {
+      "epoch": 0.13048421052631579,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0004844828008168909,
+      "loss": 3.4124,
+      "step": 3099
+    },
+    {
+      "epoch": 0.13052631578947368,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00048447109544304636,
+      "loss": 3.486,
+      "step": 3100
+    },
+    {
+      "epoch": 0.13056842105263158,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.000484459385797414,
+      "loss": 3.8117,
+      "step": 3101
+    },
+    {
+      "epoch": 0.13061052631578948,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004844476718802071,
+      "loss": 3.6356,
+      "step": 3102
+    },
+    {
+      "epoch": 0.13065263157894738,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004844359536916391,
+      "loss": 3.1973,
+      "step": 3103
+    },
+    {
+      "epoch": 0.13069473684210525,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00048442423123192343,
+      "loss": 3.3644,
+      "step": 3104
+    },
+    {
+      "epoch": 0.13073684210526315,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00048441250450127375,
+      "loss": 3.1139,
+      "step": 3105
+    },
+    {
+      "epoch": 0.13077894736842105,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004844007734999036,
+      "loss": 3.6364,
+      "step": 3106
+    },
+    {
+      "epoch": 0.13082105263157895,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0004843890382280268,
+      "loss": 3.7265,
+      "step": 3107
+    },
+    {
+      "epoch": 0.13086315789473685,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004843772986858571,
+      "loss": 3.3619,
+      "step": 3108
+    },
+    {
+      "epoch": 0.13090526315789475,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00048436555487360843,
+      "loss": 3.5016,
+      "step": 3109
+    },
+    {
+      "epoch": 0.13094736842105262,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00048435380679149475,
+      "loss": 3.5004,
+      "step": 3110
+    },
+    {
+      "epoch": 0.13098947368421052,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004843420544397301,
+      "loss": 3.8717,
+      "step": 3111
+    },
+    {
+      "epoch": 0.13103157894736842,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00048433029781852846,
+      "loss": 3.8099,
+      "step": 3112
+    },
+    {
+      "epoch": 0.13107368421052631,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004843185369281041,
+      "loss": 3.6289,
+      "step": 3113
+    },
+    {
+      "epoch": 0.1311157894736842,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00048430677176867137,
+      "loss": 3.7775,
+      "step": 3114
+    },
+    {
+      "epoch": 0.1311578947368421,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00048429500234044455,
+      "loss": 3.4422,
+      "step": 3115
+    },
+    {
+      "epoch": 0.1312,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00048428322864363805,
+      "loss": 3.0621,
+      "step": 3116
+    },
+    {
+      "epoch": 0.13124210526315788,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00048427145067846643,
+      "loss": 3.2286,
+      "step": 3117
+    },
+    {
+      "epoch": 0.13128421052631578,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00048425966844514425,
+      "loss": 3.6919,
+      "step": 3118
+    },
+    {
+      "epoch": 0.13132631578947368,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0004842478819438861,
+      "loss": 3.1122,
+      "step": 3119
+    },
+    {
+      "epoch": 0.13136842105263158,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00048423609117490685,
+      "loss": 3.1712,
+      "step": 3120
+    },
+    {
+      "epoch": 0.13141052631578948,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00048422429613842114,
+      "loss": 3.373,
+      "step": 3121
+    },
+    {
+      "epoch": 0.13145263157894738,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00048421249683464404,
+      "loss": 3.466,
+      "step": 3122
+    },
+    {
+      "epoch": 0.13149473684210528,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00048420069326379035,
+      "loss": 3.3927,
+      "step": 3123
+    },
+    {
+      "epoch": 0.13153684210526315,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004841888854260753,
+      "loss": 3.65,
+      "step": 3124
+    },
+    {
+      "epoch": 0.13157894736842105,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00048417707332171385,
+      "loss": 3.5691,
+      "step": 3125
+    },
+    {
+      "epoch": 0.13162105263157894,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00048416525695092126,
+      "loss": 3.3102,
+      "step": 3126
+    },
+    {
+      "epoch": 0.13166315789473684,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00048415343631391287,
+      "loss": 3.4244,
+      "step": 3127
+    },
+    {
+      "epoch": 0.13170526315789474,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00048414161141090395,
+      "loss": 3.105,
+      "step": 3128
+    },
+    {
+      "epoch": 0.13174736842105264,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00048412978224211003,
+      "loss": 3.538,
+      "step": 3129
+    },
+    {
+      "epoch": 0.13178947368421054,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0004841179488077465,
+      "loss": 3.6223,
+      "step": 3130
+    },
+    {
+      "epoch": 0.1318315789473684,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00048410611110802907,
+      "loss": 3.293,
+      "step": 3131
+    },
+    {
+      "epoch": 0.1318736842105263,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004840942691431733,
+      "loss": 3.8934,
+      "step": 3132
+    },
+    {
+      "epoch": 0.1319157894736842,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0004840824229133951,
+      "loss": 3.3952,
+      "step": 3133
+    },
+    {
+      "epoch": 0.1319578947368421,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00048407057241891003,
+      "loss": 3.6741,
+      "step": 3134
+    },
+    {
+      "epoch": 0.132,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004840587176599343,
+      "loss": 3.6919,
+      "step": 3135
+    },
+    {
+      "epoch": 0.1320421052631579,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00048404685863668375,
+      "loss": 3.3624,
+      "step": 3136
+    },
+    {
+      "epoch": 0.13208421052631578,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004840349953493743,
+      "loss": 3.5521,
+      "step": 3137
+    },
+    {
+      "epoch": 0.13212631578947368,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00048402312779822235,
+      "loss": 3.434,
+      "step": 3138
+    },
+    {
+      "epoch": 0.13216842105263157,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00048401125598344387,
+      "loss": 3.7669,
+      "step": 3139
+    },
+    {
+      "epoch": 0.13221052631578947,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00048399937990525535,
+      "loss": 3.3957,
+      "step": 3140
+    },
+    {
+      "epoch": 0.13225263157894737,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.000483987499563873,
+      "loss": 3.6902,
+      "step": 3141
+    },
+    {
+      "epoch": 0.13229473684210527,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004839756149595133,
+      "loss": 3.4689,
+      "step": 3142
+    },
+    {
+      "epoch": 0.13233684210526317,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004839637260923929,
+      "loss": 3.8819,
+      "step": 3143
+    },
+    {
+      "epoch": 0.13237894736842104,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00048395183296272824,
+      "loss": 3.9212,
+      "step": 3144
+    },
+    {
+      "epoch": 0.13242105263157894,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004839399355707361,
+      "loss": 3.294,
+      "step": 3145
+    },
+    {
+      "epoch": 0.13246315789473684,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004839280339166331,
+      "loss": 3.4012,
+      "step": 3146
+    },
+    {
+      "epoch": 0.13250526315789474,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00048391612800063633,
+      "loss": 3.4675,
+      "step": 3147
+    },
+    {
+      "epoch": 0.13254736842105264,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004839042178229625,
+      "loss": 3.4778,
+      "step": 3148
+    },
+    {
+      "epoch": 0.13258947368421053,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0004838923033838286,
+      "loss": 3.4909,
+      "step": 3149
+    },
+    {
+      "epoch": 0.13263157894736843,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00048388038468345177,
+      "loss": 3.248,
+      "step": 3150
+    },
+    {
+      "epoch": 0.1326736842105263,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00048386846172204914,
+      "loss": 3.7677,
+      "step": 3151
+    },
+    {
+      "epoch": 0.1327157894736842,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004838565344998379,
+      "loss": 3.4439,
+      "step": 3152
+    },
+    {
+      "epoch": 0.1327578947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004838446030170354,
+      "loss": 3.6969,
+      "step": 3153
+    },
+    {
+      "epoch": 0.1328,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.000483832667273859,
+      "loss": 2.9953,
+      "step": 3154
+    },
+    {
+      "epoch": 0.1328421052631579,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00048382072727052607,
+      "loss": 3.5185,
+      "step": 3155
+    },
+    {
+      "epoch": 0.1328842105263158,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00048380878300725426,
+      "loss": 2.7888,
+      "step": 3156
+    },
+    {
+      "epoch": 0.1329263157894737,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00048379683448426114,
+      "loss": 3.1143,
+      "step": 3157
+    },
+    {
+      "epoch": 0.13296842105263157,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00048378488170176446,
+      "loss": 3.5351,
+      "step": 3158
+    },
+    {
+      "epoch": 0.13301052631578947,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00048377292465998187,
+      "loss": 3.7393,
+      "step": 3159
+    },
+    {
+      "epoch": 0.13305263157894737,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004837609633591312,
+      "loss": 3.1264,
+      "step": 3160
+    },
+    {
+      "epoch": 0.13309473684210527,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00048374899779943047,
+      "loss": 3.4538,
+      "step": 3161
+    },
+    {
+      "epoch": 0.13313684210526316,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004837370279810977,
+      "loss": 3.3446,
+      "step": 3162
+    },
+    {
+      "epoch": 0.13317894736842106,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0004837250539043509,
+      "loss": 3.0745,
+      "step": 3163
+    },
+    {
+      "epoch": 0.13322105263157893,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004837130755694082,
+      "loss": 3.2596,
+      "step": 3164
+    },
+    {
+      "epoch": 0.13326315789473683,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004837010929764879,
+      "loss": 3.3567,
+      "step": 3165
+    },
+    {
+      "epoch": 0.13330526315789473,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004836891061258082,
+      "loss": 3.7764,
+      "step": 3166
+    },
+    {
+      "epoch": 0.13334736842105263,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004836771150175876,
+      "loss": 3.1345,
+      "step": 3167
+    },
+    {
+      "epoch": 0.13338947368421053,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0004836651196520445,
+      "loss": 3.6898,
+      "step": 3168
+    },
+    {
+      "epoch": 0.13343157894736843,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00048365312002939757,
+      "loss": 3.3996,
+      "step": 3169
+    },
+    {
+      "epoch": 0.13347368421052633,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004836411161498652,
+      "loss": 3.7285,
+      "step": 3170
+    },
+    {
+      "epoch": 0.1335157894736842,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004836291080136663,
+      "loss": 3.2869,
+      "step": 3171
+    },
+    {
+      "epoch": 0.1335578947368421,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004836170956210195,
+      "loss": 3.6923,
+      "step": 3172
+    },
+    {
+      "epoch": 0.1336,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00048360507897214373,
+      "loss": 3.4182,
+      "step": 3173
+    },
+    {
+      "epoch": 0.1336421052631579,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004835930580672579,
+      "loss": 3.4935,
+      "step": 3174
+    },
+    {
+      "epoch": 0.1336842105263158,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00048358103290658096,
+      "loss": 3.4496,
+      "step": 3175
+    },
+    {
+      "epoch": 0.1337263157894737,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004835690034903321,
+      "loss": 3.7415,
+      "step": 3176
+    },
+    {
+      "epoch": 0.1337684210526316,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004835569698187304,
+      "loss": 3.9538,
+      "step": 3177
+    },
+    {
+      "epoch": 0.13381052631578946,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004835449318919952,
+      "loss": 3.4964,
+      "step": 3178
+    },
+    {
+      "epoch": 0.13385263157894736,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00048353288971034566,
+      "loss": 3.6151,
+      "step": 3179
+    },
+    {
+      "epoch": 0.13389473684210526,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004835208432740013,
+      "loss": 4.0144,
+      "step": 3180
+    },
+    {
+      "epoch": 0.13393684210526316,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004835087925831815,
+      "loss": 3.6099,
+      "step": 3181
+    },
+    {
+      "epoch": 0.13397894736842106,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004834967376381059,
+      "loss": 3.0776,
+      "step": 3182
+    },
+    {
+      "epoch": 0.13402105263157896,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004834846784389941,
+      "loss": 3.4687,
+      "step": 3183
+    },
+    {
+      "epoch": 0.13406315789473683,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00048347261498606574,
+      "loss": 3.3375,
+      "step": 3184
+    },
+    {
+      "epoch": 0.13410526315789473,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0004834605472795407,
+      "loss": 3.7815,
+      "step": 3185
+    },
+    {
+      "epoch": 0.13414736842105263,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00048344847531963883,
+      "loss": 3.3755,
+      "step": 3186
+    },
+    {
+      "epoch": 0.13418947368421053,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00048343639910658,
+      "loss": 3.6461,
+      "step": 3187
+    },
+    {
+      "epoch": 0.13423157894736842,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004834243186405842,
+      "loss": 3.3575,
+      "step": 3188
+    },
+    {
+      "epoch": 0.13427368421052632,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00048341223392187163,
+      "loss": 3.2938,
+      "step": 3189
+    },
+    {
+      "epoch": 0.13431578947368422,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004834001449506624,
+      "loss": 3.2255,
+      "step": 3190
+    },
+    {
+      "epoch": 0.1343578947368421,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00048338805172717676,
+      "loss": 3.5113,
+      "step": 3191
+    },
+    {
+      "epoch": 0.1344,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00048337595425163504,
+      "loss": 3.8051,
+      "step": 3192
+    },
+    {
+      "epoch": 0.1344421052631579,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004833638525242576,
+      "loss": 3.8248,
+      "step": 3193
+    },
+    {
+      "epoch": 0.1344842105263158,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00048335174654526503,
+      "loss": 3.2356,
+      "step": 3194
+    },
+    {
+      "epoch": 0.1345263157894737,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00048333963631487774,
+      "loss": 3.6372,
+      "step": 3195
+    },
+    {
+      "epoch": 0.1345684210526316,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0004833275218333165,
+      "loss": 3.3769,
+      "step": 3196
+    },
+    {
+      "epoch": 0.13461052631578949,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00048331540310080193,
+      "loss": 3.3041,
+      "step": 3197
+    },
+    {
+      "epoch": 0.13465263157894736,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004833032801175549,
+      "loss": 3.7184,
+      "step": 3198
+    },
+    {
+      "epoch": 0.13469473684210526,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004832911528837962,
+      "loss": 3.6978,
+      "step": 3199
+    },
+    {
+      "epoch": 0.13473684210526315,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004832790213997467,
+      "loss": 3.1034,
+      "step": 3200
+    },
+    {
+      "epoch": 0.13477894736842105,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00048326688566562763,
+      "loss": 3.2976,
+      "step": 3201
+    },
+    {
+      "epoch": 0.13482105263157895,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00048325474568165994,
+      "loss": 3.6054,
+      "step": 3202
+    },
+    {
+      "epoch": 0.13486315789473685,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00048324260144806485,
+      "loss": 3.5848,
+      "step": 3203
+    },
+    {
+      "epoch": 0.13490526315789475,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00048323045296506364,
+      "loss": 3.325,
+      "step": 3204
+    },
+    {
+      "epoch": 0.13494736842105262,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0004832183002328776,
+      "loss": 3.447,
+      "step": 3205
+    },
+    {
+      "epoch": 0.13498947368421052,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00048320614325172816,
+      "loss": 3.6158,
+      "step": 3206
+    },
+    {
+      "epoch": 0.13503157894736842,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00048319398202183677,
+      "loss": 3.7218,
+      "step": 3207
+    },
+    {
+      "epoch": 0.13507368421052632,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.000483181816543425,
+      "loss": 3.6022,
+      "step": 3208
+    },
+    {
+      "epoch": 0.13511578947368422,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004831696468167146,
+      "loss": 3.3156,
+      "step": 3209
+    },
+    {
+      "epoch": 0.13515789473684212,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00048315747284192713,
+      "loss": 3.5495,
+      "step": 3210
+    },
+    {
+      "epoch": 0.1352,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00048314529461928446,
+      "loss": 3.1416,
+      "step": 3211
+    },
+    {
+      "epoch": 0.13524210526315789,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00048313311214900846,
+      "loss": 3.738,
+      "step": 3212
+    },
+    {
+      "epoch": 0.13528421052631578,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.000483120925431321,
+      "loss": 3.6982,
+      "step": 3213
+    },
+    {
+      "epoch": 0.13532631578947368,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004831087344664443,
+      "loss": 2.9038,
+      "step": 3214
+    },
+    {
+      "epoch": 0.13536842105263158,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004830965392546003,
+      "loss": 3.2407,
+      "step": 3215
+    },
+    {
+      "epoch": 0.13541052631578948,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004830843397960113,
+      "loss": 3.7073,
+      "step": 3216
+    },
+    {
+      "epoch": 0.13545263157894738,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00048307213609089937,
+      "loss": 3.3639,
+      "step": 3217
+    },
+    {
+      "epoch": 0.13549473684210525,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00048305992813948707,
+      "loss": 3.1163,
+      "step": 3218
+    },
+    {
+      "epoch": 0.13553684210526315,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00048304771594199667,
+      "loss": 3.4708,
+      "step": 3219
+    },
+    {
+      "epoch": 0.13557894736842105,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00048303549949865065,
+      "loss": 3.644,
+      "step": 3220
+    },
+    {
+      "epoch": 0.13562105263157895,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004830232788096717,
+      "loss": 3.6264,
+      "step": 3221
+    },
+    {
+      "epoch": 0.13566315789473685,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004830110538752824,
+      "loss": 3.1388,
+      "step": 3222
+    },
+    {
+      "epoch": 0.13570526315789475,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004829988246957055,
+      "loss": 3.5765,
+      "step": 3223
+    },
+    {
+      "epoch": 0.13574736842105264,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00048298659127116373,
+      "loss": 2.7384,
+      "step": 3224
+    },
+    {
+      "epoch": 0.13578947368421052,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00048297435360188004,
+      "loss": 3.4258,
+      "step": 3225
+    },
+    {
+      "epoch": 0.13583157894736841,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0004829621116880774,
+      "loss": 3.2441,
+      "step": 3226
+    },
+    {
+      "epoch": 0.1358736842105263,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00048294986552997873,
+      "loss": 3.0285,
+      "step": 3227
+    },
+    {
+      "epoch": 0.1359157894736842,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004829376151278072,
+      "loss": 3.6065,
+      "step": 3228
+    },
+    {
+      "epoch": 0.1359578947368421,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.000482925360481786,
+      "loss": 3.0228,
+      "step": 3229
+    },
+    {
+      "epoch": 0.136,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004829131015921385,
+      "loss": 3.9504,
+      "step": 3230
+    },
+    {
+      "epoch": 0.13604210526315788,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00048290083845908794,
+      "loss": 3.0421,
+      "step": 3231
+    },
+    {
+      "epoch": 0.13608421052631578,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0004828885710828577,
+      "loss": 2.6915,
+      "step": 3232
+    },
+    {
+      "epoch": 0.13612631578947368,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004828762994636714,
+      "loss": 3.3249,
+      "step": 3233
+    },
+    {
+      "epoch": 0.13616842105263158,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00048286402360175245,
+      "loss": 3.8109,
+      "step": 3234
+    },
+    {
+      "epoch": 0.13621052631578948,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00048285174349732465,
+      "loss": 3.4175,
+      "step": 3235
+    },
+    {
+      "epoch": 0.13625263157894738,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00048283945915061167,
+      "loss": 3.5594,
+      "step": 3236
+    },
+    {
+      "epoch": 0.13629473684210527,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00048282717056183735,
+      "loss": 3.5825,
+      "step": 3237
+    },
+    {
+      "epoch": 0.13633684210526315,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00048281487773122546,
+      "loss": 3.6342,
+      "step": 3238
+    },
+    {
+      "epoch": 0.13637894736842104,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00048280258065900005,
+      "loss": 3.2347,
+      "step": 3239
+    },
+    {
+      "epoch": 0.13642105263157894,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00048279027934538523,
+      "loss": 3.7541,
+      "step": 3240
+    },
+    {
+      "epoch": 0.13646315789473684,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.000482777973790605,
+      "loss": 3.2116,
+      "step": 3241
+    },
+    {
+      "epoch": 0.13650526315789474,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00048276566399488356,
+      "loss": 3.4717,
+      "step": 3242
+    },
+    {
+      "epoch": 0.13654736842105264,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00048275334995844524,
+      "loss": 3.5744,
+      "step": 3243
+    },
+    {
+      "epoch": 0.13658947368421054,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004827410316815144,
+      "loss": 3.4858,
+      "step": 3244
+    },
+    {
+      "epoch": 0.1366315789473684,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0004827287091643153,
+      "loss": 3.1273,
+      "step": 3245
+    },
+    {
+      "epoch": 0.1366736842105263,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004827163824070727,
+      "loss": 3.9362,
+      "step": 3246
+    },
+    {
+      "epoch": 0.1367157894736842,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00048270405141001094,
+      "loss": 3.6746,
+      "step": 3247
+    },
+    {
+      "epoch": 0.1367578947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004826917161733548,
+      "loss": 3.5528,
+      "step": 3248
+    },
+    {
+      "epoch": 0.1368,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0004826793766973291,
+      "loss": 3.4151,
+      "step": 3249
+    },
+    {
+      "epoch": 0.1368421052631579,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0004826670329821584,
+      "loss": 3.0746,
+      "step": 3250
+    },
+    {
+      "epoch": 0.1368842105263158,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004826546850280678,
+      "loss": 3.5138,
+      "step": 3251
+    },
+    {
+      "epoch": 0.13692631578947367,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00048264233283528226,
+      "loss": 3.0448,
+      "step": 3252
+    },
+    {
+      "epoch": 0.13696842105263157,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00048262997640402663,
+      "loss": 3.4632,
+      "step": 3253
+    },
+    {
+      "epoch": 0.13701052631578947,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004826176157345262,
+      "loss": 3.5807,
+      "step": 3254
+    },
+    {
+      "epoch": 0.13705263157894737,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00048260525082700613,
+      "loss": 3.4383,
+      "step": 3255
+    },
+    {
+      "epoch": 0.13709473684210527,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00048259288168169174,
+      "loss": 3.4302,
+      "step": 3256
+    },
+    {
+      "epoch": 0.13713684210526317,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004825805082988083,
+      "loss": 3.1451,
+      "step": 3257
+    },
+    {
+      "epoch": 0.13717894736842104,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00048256813067858124,
+      "loss": 3.2636,
+      "step": 3258
+    },
+    {
+      "epoch": 0.13722105263157894,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004825557488212361,
+      "loss": 3.311,
+      "step": 3259
+    },
+    {
+      "epoch": 0.13726315789473684,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0004825433627269985,
+      "loss": 3.2212,
+      "step": 3260
+    },
+    {
+      "epoch": 0.13730526315789474,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.000482530972396094,
+      "loss": 3.6675,
+      "step": 3261
+    },
+    {
+      "epoch": 0.13734736842105263,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004825185778287485,
+      "loss": 3.5987,
+      "step": 3262
+    },
+    {
+      "epoch": 0.13738947368421053,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004825061790251876,
+      "loss": 3.5565,
+      "step": 3263
+    },
+    {
+      "epoch": 0.13743157894736843,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00048249377598563736,
+      "loss": 3.4649,
+      "step": 3264
+    },
+    {
+      "epoch": 0.1374736842105263,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00048248136871032366,
+      "loss": 3.4036,
+      "step": 3265
+    },
+    {
+      "epoch": 0.1375157894736842,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004824689571994726,
+      "loss": 3.6233,
+      "step": 3266
+    },
+    {
+      "epoch": 0.1375578947368421,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004824565414533102,
+      "loss": 4.049,
+      "step": 3267
+    },
+    {
+      "epoch": 0.1376,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00048244412147206283,
+      "loss": 3.5397,
+      "step": 3268
+    },
+    {
+      "epoch": 0.1376421052631579,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0004824316972559567,
+      "loss": 3.2536,
+      "step": 3269
+    },
+    {
+      "epoch": 0.1376842105263158,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004824192688052181,
+      "loss": 3.5472,
+      "step": 3270
+    },
+    {
+      "epoch": 0.1377263157894737,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004824068361200735,
+      "loss": 3.6646,
+      "step": 3271
+    },
+    {
+      "epoch": 0.13776842105263157,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004823943992007494,
+      "loss": 3.7678,
+      "step": 3272
+    },
+    {
+      "epoch": 0.13781052631578947,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004823819580474724,
+      "loss": 3.3231,
+      "step": 3273
+    },
+    {
+      "epoch": 0.13785263157894737,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004823695126604692,
+      "loss": 2.7148,
+      "step": 3274
+    },
+    {
+      "epoch": 0.13789473684210526,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00048235706303996647,
+      "loss": 3.9458,
+      "step": 3275
+    },
+    {
+      "epoch": 0.13793684210526316,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0004823446091861911,
+      "loss": 3.4602,
+      "step": 3276
+    },
+    {
+      "epoch": 0.13797894736842106,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0004823321510993699,
+      "loss": 3.2659,
+      "step": 3277
+    },
+    {
+      "epoch": 0.13802105263157896,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00048231968877972985,
+      "loss": 3.678,
+      "step": 3278
+    },
+    {
+      "epoch": 0.13806315789473683,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00048230722222749814,
+      "loss": 3.2327,
+      "step": 3279
+    },
+    {
+      "epoch": 0.13810526315789473,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004822947514429017,
+      "loss": 3.7413,
+      "step": 3280
+    },
+    {
+      "epoch": 0.13814736842105263,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0004822822764261678,
+      "loss": 2.7387,
+      "step": 3281
+    },
+    {
+      "epoch": 0.13818947368421053,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004822697971775238,
+      "loss": 4.0238,
+      "step": 3282
+    },
+    {
+      "epoch": 0.13823157894736843,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00048225731369719706,
+      "loss": 3.5222,
+      "step": 3283
+    },
+    {
+      "epoch": 0.13827368421052633,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004822448259854149,
+      "loss": 3.8242,
+      "step": 3284
+    },
+    {
+      "epoch": 0.1383157894736842,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004822323340424049,
+      "loss": 3.5235,
+      "step": 3285
+    },
+    {
+      "epoch": 0.1383578947368421,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00048221983786839455,
+      "loss": 3.6023,
+      "step": 3286
+    },
+    {
+      "epoch": 0.1384,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00048220733746361165,
+      "loss": 3.0167,
+      "step": 3287
+    },
+    {
+      "epoch": 0.1384421052631579,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00048219483282828396,
+      "loss": 3.7563,
+      "step": 3288
+    },
+    {
+      "epoch": 0.1384842105263158,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00048218232396263916,
+      "loss": 3.2722,
+      "step": 3289
+    },
+    {
+      "epoch": 0.1385263157894737,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00048216981086690526,
+      "loss": 3.6911,
+      "step": 3290
+    },
+    {
+      "epoch": 0.1385684210526316,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0004821572935413102,
+      "loss": 3.5001,
+      "step": 3291
+    },
+    {
+      "epoch": 0.13861052631578946,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.000482144771986082,
+      "loss": 3.7663,
+      "step": 3292
+    },
+    {
+      "epoch": 0.13865263157894736,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004821322462014488,
+      "loss": 3.5049,
+      "step": 3293
+    },
+    {
+      "epoch": 0.13869473684210526,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004821197161876389,
+      "loss": 3.1934,
+      "step": 3294
+    },
+    {
+      "epoch": 0.13873684210526316,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00048210718194488045,
+      "loss": 3.5232,
+      "step": 3295
+    },
+    {
+      "epoch": 0.13877894736842106,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004820946434734018,
+      "loss": 3.6145,
+      "step": 3296
+    },
+    {
+      "epoch": 0.13882105263157896,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00048208210077343153,
+      "loss": 3.9118,
+      "step": 3297
+    },
+    {
+      "epoch": 0.13886315789473686,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.000482069553845198,
+      "loss": 3.6828,
+      "step": 3298
+    },
+    {
+      "epoch": 0.13890526315789473,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00048205700268892995,
+      "loss": 3.4494,
+      "step": 3299
+    },
+    {
+      "epoch": 0.13894736842105262,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00048204444730485586,
+      "loss": 3.5507,
+      "step": 3300
+    },
+    {
+      "epoch": 0.13898947368421052,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00048203188769320465,
+      "loss": 3.5164,
+      "step": 3301
+    },
+    {
+      "epoch": 0.13903157894736842,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004820193238542051,
+      "loss": 3.5431,
+      "step": 3302
+    },
+    {
+      "epoch": 0.13907368421052632,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004820067557880861,
+      "loss": 3.6186,
+      "step": 3303
+    },
+    {
+      "epoch": 0.13911578947368422,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004819941834950765,
+      "loss": 3.3359,
+      "step": 3304
+    },
+    {
+      "epoch": 0.1391578947368421,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004819816069754055,
+      "loss": 3.5231,
+      "step": 3305
+    },
+    {
+      "epoch": 0.1392,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00048196902622930216,
+      "loss": 2.6678,
+      "step": 3306
+    },
+    {
+      "epoch": 0.1392421052631579,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004819564412569958,
+      "loss": 3.8929,
+      "step": 3307
+    },
+    {
+      "epoch": 0.1392842105263158,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00048194385205871555,
+      "loss": 3.952,
+      "step": 3308
+    },
+    {
+      "epoch": 0.1393263157894737,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00048193125863469087,
+      "loss": 3.4279,
+      "step": 3309
+    },
+    {
+      "epoch": 0.13936842105263159,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004819186609851511,
+      "loss": 3.6198,
+      "step": 3310
+    },
+    {
+      "epoch": 0.13941052631578948,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00048190605911032594,
+      "loss": 3.4354,
+      "step": 3311
+    },
+    {
+      "epoch": 0.13945263157894736,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00048189345301044475,
+      "loss": 3.2515,
+      "step": 3312
+    },
+    {
+      "epoch": 0.13949473684210525,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004818808426857374,
+      "loss": 3.341,
+      "step": 3313
+    },
+    {
+      "epoch": 0.13953684210526315,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004818682281364335,
+      "loss": 3.3847,
+      "step": 3314
+    },
+    {
+      "epoch": 0.13957894736842105,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004818556093627629,
+      "loss": 3.8515,
+      "step": 3315
+    },
+    {
+      "epoch": 0.13962105263157895,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004818429863649555,
+      "loss": 3.6405,
+      "step": 3316
+    },
+    {
+      "epoch": 0.13966315789473685,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00048183035914324134,
+      "loss": 3.5672,
+      "step": 3317
+    },
+    {
+      "epoch": 0.13970526315789475,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00048181772769785044,
+      "loss": 3.0749,
+      "step": 3318
+    },
+    {
+      "epoch": 0.13974736842105262,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0004818050920290129,
+      "loss": 3.253,
+      "step": 3319
+    },
+    {
+      "epoch": 0.13978947368421052,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004817924521369589,
+      "loss": 3.264,
+      "step": 3320
+    },
+    {
+      "epoch": 0.13983157894736842,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004817798080219188,
+      "loss": 3.7743,
+      "step": 3321
+    },
+    {
+      "epoch": 0.13987368421052632,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00048176715968412294,
+      "loss": 3.5263,
+      "step": 3322
+    },
+    {
+      "epoch": 0.13991578947368422,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00048175450712380176,
+      "loss": 3.5551,
+      "step": 3323
+    },
+    {
+      "epoch": 0.13995789473684211,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004817418503411857,
+      "loss": 3.5418,
+      "step": 3324
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004817291893365054,
+      "loss": 3.8286,
+      "step": 3325
+    },
+    {
+      "epoch": 0.14004210526315788,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00048171652410999166,
+      "loss": 3.2367,
+      "step": 3326
+    },
+    {
+      "epoch": 0.14008421052631578,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00048170385466187506,
+      "loss": 3.7586,
+      "step": 3327
+    },
+    {
+      "epoch": 0.14012631578947368,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004816911809923864,
+      "loss": 3.8166,
+      "step": 3328
+    },
+    {
+      "epoch": 0.14016842105263158,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00048167850310175663,
+      "loss": 3.5673,
+      "step": 3329
+    },
+    {
+      "epoch": 0.14021052631578948,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0004816658209902168,
+      "loss": 3.3051,
+      "step": 3330
+    },
+    {
+      "epoch": 0.14025263157894738,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.000481653134657998,
+      "loss": 3.8526,
+      "step": 3331
+    },
+    {
+      "epoch": 0.14029473684210525,
+      "grad_norm": 0.875,
+      "learning_rate": 0.00048164044410533115,
+      "loss": 3.2402,
+      "step": 3332
+    },
+    {
+      "epoch": 0.14033684210526315,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0004816277493324476,
+      "loss": 3.9009,
+      "step": 3333
+    },
+    {
+      "epoch": 0.14037894736842105,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00048161505033957866,
+      "loss": 3.3513,
+      "step": 3334
+    },
+    {
+      "epoch": 0.14042105263157895,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00048160234712695557,
+      "loss": 3.0041,
+      "step": 3335
+    },
+    {
+      "epoch": 0.14046315789473685,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00048158963969480986,
+      "loss": 3.3732,
+      "step": 3336
+    },
+    {
+      "epoch": 0.14050526315789474,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.000481576928043373,
+      "loss": 3.0949,
+      "step": 3337
+    },
+    {
+      "epoch": 0.14054736842105264,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004815642121728766,
+      "loss": 3.7713,
+      "step": 3338
+    },
+    {
+      "epoch": 0.14058947368421051,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004815514920835524,
+      "loss": 3.4124,
+      "step": 3339
+    },
+    {
+      "epoch": 0.1406315789473684,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00048153876777563197,
+      "loss": 3.7876,
+      "step": 3340
+    },
+    {
+      "epoch": 0.1406736842105263,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00048152603924934734,
+      "loss": 3.3635,
+      "step": 3341
+    },
+    {
+      "epoch": 0.1407157894736842,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0004815133065049303,
+      "loss": 3.5514,
+      "step": 3342
+    },
+    {
+      "epoch": 0.1407578947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004815005695426128,
+      "loss": 3.1351,
+      "step": 3343
+    },
+    {
+      "epoch": 0.1408,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00048148782836262687,
+      "loss": 3.4575,
+      "step": 3344
+    },
+    {
+      "epoch": 0.1408421052631579,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004814750829652048,
+      "loss": 3.3083,
+      "step": 3345
+    },
+    {
+      "epoch": 0.14088421052631578,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004814623333505786,
+      "loss": 3.274,
+      "step": 3346
+    },
+    {
+      "epoch": 0.14092631578947368,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00048144957951898066,
+      "loss": 3.2509,
+      "step": 3347
+    },
+    {
+      "epoch": 0.14096842105263158,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00048143682147064337,
+      "loss": 3.0702,
+      "step": 3348
+    },
+    {
+      "epoch": 0.14101052631578948,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00048142405920579905,
+      "loss": 3.2826,
+      "step": 3349
+    },
+    {
+      "epoch": 0.14105263157894737,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004814112927246804,
+      "loss": 3.1565,
+      "step": 3350
+    },
+    {
+      "epoch": 0.14109473684210527,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00048139852202751975,
+      "loss": 3.6337,
+      "step": 3351
+    },
+    {
+      "epoch": 0.14113684210526317,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00048138574711455,
+      "loss": 3.4784,
+      "step": 3352
+    },
+    {
+      "epoch": 0.14117894736842104,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00048137296798600384,
+      "loss": 3.7547,
+      "step": 3353
+    },
+    {
+      "epoch": 0.14122105263157894,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00048136018464211395,
+      "loss": 3.458,
+      "step": 3354
+    },
+    {
+      "epoch": 0.14126315789473684,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004813473970831134,
+      "loss": 3.8346,
+      "step": 3355
+    },
+    {
+      "epoch": 0.14130526315789474,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00048133460530923506,
+      "loss": 3.2314,
+      "step": 3356
+    },
+    {
+      "epoch": 0.14134736842105264,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.000481321809320712,
+      "loss": 3.7825,
+      "step": 3357
+    },
+    {
+      "epoch": 0.14138947368421054,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004813090091177774,
+      "loss": 3.6057,
+      "step": 3358
+    },
+    {
+      "epoch": 0.1414315789473684,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004812962047006645,
+      "loss": 3.528,
+      "step": 3359
+    },
+    {
+      "epoch": 0.1414736842105263,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00048128339606960647,
+      "loss": 3.3825,
+      "step": 3360
+    },
+    {
+      "epoch": 0.1415157894736842,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00048127058322483665,
+      "loss": 3.257,
+      "step": 3361
+    },
+    {
+      "epoch": 0.1415578947368421,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004812577661665886,
+      "loss": 3.5798,
+      "step": 3362
+    },
+    {
+      "epoch": 0.1416,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00048124494489509574,
+      "loss": 3.4369,
+      "step": 3363
+    },
+    {
+      "epoch": 0.1416421052631579,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00048123211941059166,
+      "loss": 3.6946,
+      "step": 3364
+    },
+    {
+      "epoch": 0.1416842105263158,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004812192897133101,
+      "loss": 3.1618,
+      "step": 3365
+    },
+    {
+      "epoch": 0.14172631578947367,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004812064558034847,
+      "loss": 3.3881,
+      "step": 3366
+    },
+    {
+      "epoch": 0.14176842105263157,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004811936176813494,
+      "loss": 3.5791,
+      "step": 3367
+    },
+    {
+      "epoch": 0.14181052631578947,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.000481180775347138,
+      "loss": 3.5384,
+      "step": 3368
+    },
+    {
+      "epoch": 0.14185263157894737,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00048116792880108445,
+      "loss": 3.3532,
+      "step": 3369
+    },
+    {
+      "epoch": 0.14189473684210527,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004811550780434229,
+      "loss": 3.7962,
+      "step": 3370
+    },
+    {
+      "epoch": 0.14193684210526317,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00048114222307438735,
+      "loss": 4.0457,
+      "step": 3371
+    },
+    {
+      "epoch": 0.14197894736842107,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004811293638942122,
+      "loss": 3.8407,
+      "step": 3372
+    },
+    {
+      "epoch": 0.14202105263157894,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004811165005031315,
+      "loss": 3.4283,
+      "step": 3373
+    },
+    {
+      "epoch": 0.14206315789473684,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00048110363290137983,
+      "loss": 3.4009,
+      "step": 3374
+    },
+    {
+      "epoch": 0.14210526315789473,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004810907610891914,
+      "loss": 3.3084,
+      "step": 3375
+    },
+    {
+      "epoch": 0.14214736842105263,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00048107788506680083,
+      "loss": 3.3244,
+      "step": 3376
+    },
+    {
+      "epoch": 0.14218947368421053,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004810650048344427,
+      "loss": 3.7103,
+      "step": 3377
+    },
+    {
+      "epoch": 0.14223157894736843,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00048105212039235173,
+      "loss": 3.5199,
+      "step": 3378
+    },
+    {
+      "epoch": 0.1422736842105263,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004810392317407626,
+      "loss": 3.365,
+      "step": 3379
+    },
+    {
+      "epoch": 0.1423157894736842,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0004810263388799101,
+      "loss": 3.1618,
+      "step": 3380
+    },
+    {
+      "epoch": 0.1423578947368421,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004810134418100291,
+      "loss": 3.895,
+      "step": 3381
+    },
+    {
+      "epoch": 0.1424,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004810005405313547,
+      "loss": 3.4511,
+      "step": 3382
+    },
+    {
+      "epoch": 0.1424421052631579,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0004809876350441218,
+      "loss": 2.8591,
+      "step": 3383
+    },
+    {
+      "epoch": 0.1424842105263158,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004809747253485656,
+      "loss": 3.61,
+      "step": 3384
+    },
+    {
+      "epoch": 0.1425263157894737,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00048096181144492137,
+      "loss": 3.213,
+      "step": 3385
+    },
+    {
+      "epoch": 0.14256842105263157,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004809488933334242,
+      "loss": 3.751,
+      "step": 3386
+    },
+    {
+      "epoch": 0.14261052631578947,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0004809359710143096,
+      "loss": 3.4314,
+      "step": 3387
+    },
+    {
+      "epoch": 0.14265263157894736,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004809230444878129,
+      "loss": 3.7577,
+      "step": 3388
+    },
+    {
+      "epoch": 0.14269473684210526,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004809101137541697,
+      "loss": 3.6778,
+      "step": 3389
+    },
+    {
+      "epoch": 0.14273684210526316,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004808971788136155,
+      "loss": 2.7928,
+      "step": 3390
+    },
+    {
+      "epoch": 0.14277894736842106,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004808842396663861,
+      "loss": 3.4381,
+      "step": 3391
+    },
+    {
+      "epoch": 0.14282105263157896,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.000480871296312717,
+      "loss": 3.5115,
+      "step": 3392
+    },
+    {
+      "epoch": 0.14286315789473683,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00048085834875284417,
+      "loss": 3.7029,
+      "step": 3393
+    },
+    {
+      "epoch": 0.14290526315789473,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004808453969870035,
+      "loss": 3.1848,
+      "step": 3394
+    },
+    {
+      "epoch": 0.14294736842105263,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00048083244101543093,
+      "loss": 3.6566,
+      "step": 3395
+    },
+    {
+      "epoch": 0.14298947368421053,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004808194808383625,
+      "loss": 3.3442,
+      "step": 3396
+    },
+    {
+      "epoch": 0.14303157894736843,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00048080651645603437,
+      "loss": 3.4952,
+      "step": 3397
+    },
+    {
+      "epoch": 0.14307368421052633,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00048079354786868263,
+      "loss": 3.5902,
+      "step": 3398
+    },
+    {
+      "epoch": 0.14311578947368422,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004807805750765436,
+      "loss": 3.2667,
+      "step": 3399
+    },
+    {
+      "epoch": 0.1431578947368421,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004807675980798537,
+      "loss": 3.2706,
+      "step": 3400
+    },
+    {
+      "epoch": 0.1432,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00048075461687884934,
+      "loss": 3.3424,
+      "step": 3401
+    },
+    {
+      "epoch": 0.1432421052631579,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004807416314737669,
+      "loss": 3.6431,
+      "step": 3402
+    },
+    {
+      "epoch": 0.1432842105263158,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004807286418648431,
+      "loss": 3.3001,
+      "step": 3403
+    },
+    {
+      "epoch": 0.1433263157894737,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00048071564805231455,
+      "loss": 3.4262,
+      "step": 3404
+    },
+    {
+      "epoch": 0.1433684210526316,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00048070265003641793,
+      "loss": 3.517,
+      "step": 3405
+    },
+    {
+      "epoch": 0.14341052631578946,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004806896478173901,
+      "loss": 3.4288,
+      "step": 3406
+    },
+    {
+      "epoch": 0.14345263157894736,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00048067664139546796,
+      "loss": 3.4006,
+      "step": 3407
+    },
+    {
+      "epoch": 0.14349473684210526,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004806636307708884,
+      "loss": 3.5285,
+      "step": 3408
+    },
+    {
+      "epoch": 0.14353684210526316,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004806506159438886,
+      "loss": 3.0659,
+      "step": 3409
+    },
+    {
+      "epoch": 0.14357894736842106,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004806375969147055,
+      "loss": 3.4186,
+      "step": 3410
+    },
+    {
+      "epoch": 0.14362105263157895,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00048062457368357636,
+      "loss": 3.5544,
+      "step": 3411
+    },
+    {
+      "epoch": 0.14366315789473685,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00048061154625073853,
+      "loss": 3.2955,
+      "step": 3412
+    },
+    {
+      "epoch": 0.14370526315789472,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004805985146164292,
+      "loss": 3.7367,
+      "step": 3413
+    },
+    {
+      "epoch": 0.14374736842105262,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.000480585478780886,
+      "loss": 3.4163,
+      "step": 3414
+    },
+    {
+      "epoch": 0.14378947368421052,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004805724387443462,
+      "loss": 3.2796,
+      "step": 3415
+    },
+    {
+      "epoch": 0.14383157894736842,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004805593945070475,
+      "loss": 3.2515,
+      "step": 3416
+    },
+    {
+      "epoch": 0.14387368421052632,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004805463460692275,
+      "loss": 3.6293,
+      "step": 3417
+    },
+    {
+      "epoch": 0.14391578947368422,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.000480533293431124,
+      "loss": 3.5968,
+      "step": 3418
+    },
+    {
+      "epoch": 0.14395789473684212,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004805202365929747,
+      "loss": 3.0503,
+      "step": 3419
+    },
+    {
+      "epoch": 0.144,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00048050717555501765,
+      "loss": 3.338,
+      "step": 3420
+    },
+    {
+      "epoch": 0.1440421052631579,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00048049411031749064,
+      "loss": 3.5539,
+      "step": 3421
+    },
+    {
+      "epoch": 0.1440842105263158,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004804810408806317,
+      "loss": 3.2507,
+      "step": 3422
+    },
+    {
+      "epoch": 0.14412631578947369,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0004804679672446791,
+      "loss": 3.4739,
+      "step": 3423
+    },
+    {
+      "epoch": 0.14416842105263158,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00048045488940987084,
+      "loss": 3.807,
+      "step": 3424
+    },
+    {
+      "epoch": 0.14421052631578948,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00048044180737644534,
+      "loss": 3.4709,
+      "step": 3425
+    },
+    {
+      "epoch": 0.14425263157894735,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00048042872114464083,
+      "loss": 3.5264,
+      "step": 3426
+    },
+    {
+      "epoch": 0.14429473684210525,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00048041563071469574,
+      "loss": 3.4984,
+      "step": 3427
+    },
+    {
+      "epoch": 0.14433684210526315,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00048040253608684863,
+      "loss": 3.1885,
+      "step": 3428
+    },
+    {
+      "epoch": 0.14437894736842105,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.000480389437261338,
+      "loss": 3.532,
+      "step": 3429
+    },
+    {
+      "epoch": 0.14442105263157895,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00048037633423840255,
+      "loss": 3.5646,
+      "step": 3430
+    },
+    {
+      "epoch": 0.14446315789473685,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00048036322701828095,
+      "loss": 3.235,
+      "step": 3431
+    },
+    {
+      "epoch": 0.14450526315789475,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00048035011560121203,
+      "loss": 3.3639,
+      "step": 3432
+    },
+    {
+      "epoch": 0.14454736842105262,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00048033699998743464,
+      "loss": 3.6266,
+      "step": 3433
+    },
+    {
+      "epoch": 0.14458947368421052,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0004803238801771877,
+      "loss": 3.1089,
+      "step": 3434
+    },
+    {
+      "epoch": 0.14463157894736842,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004803107561707103,
+      "loss": 3.3758,
+      "step": 3435
+    },
+    {
+      "epoch": 0.14467368421052632,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004802976279682416,
+      "loss": 3.4435,
+      "step": 3436
+    },
+    {
+      "epoch": 0.14471578947368421,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004802844955700206,
+      "loss": 3.4905,
+      "step": 3437
+    },
+    {
+      "epoch": 0.1447578947368421,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004802713589762868,
+      "loss": 3.4209,
+      "step": 3438
+    },
+    {
+      "epoch": 0.1448,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00048025821818727934,
+      "loss": 3.3824,
+      "step": 3439
+    },
+    {
+      "epoch": 0.14484210526315788,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004802450732032376,
+      "loss": 4.0705,
+      "step": 3440
+    },
+    {
+      "epoch": 0.14488421052631578,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00048023192402440123,
+      "loss": 3.7318,
+      "step": 3441
+    },
+    {
+      "epoch": 0.14492631578947368,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004802187706510097,
+      "loss": 3.4073,
+      "step": 3442
+    },
+    {
+      "epoch": 0.14496842105263158,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004802056130833027,
+      "loss": 3.1488,
+      "step": 3443
+    },
+    {
+      "epoch": 0.14501052631578948,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00048019245132151987,
+      "loss": 3.2328,
+      "step": 3444
+    },
+    {
+      "epoch": 0.14505263157894738,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00048017928536590106,
+      "loss": 3.7851,
+      "step": 3445
+    },
+    {
+      "epoch": 0.14509473684210528,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004801661152166861,
+      "loss": 3.3842,
+      "step": 3446
+    },
+    {
+      "epoch": 0.14513684210526315,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00048015294087411497,
+      "loss": 3.3484,
+      "step": 3447
+    },
+    {
+      "epoch": 0.14517894736842105,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0004801397623384277,
+      "loss": 3.6755,
+      "step": 3448
+    },
+    {
+      "epoch": 0.14522105263157895,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004801265796098643,
+      "loss": 3.3168,
+      "step": 3449
+    },
+    {
+      "epoch": 0.14526315789473684,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00048011339268866506,
+      "loss": 3.2477,
+      "step": 3450
+    },
+    {
+      "epoch": 0.14530526315789474,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004801002015750702,
+      "loss": 2.9345,
+      "step": 3451
+    },
+    {
+      "epoch": 0.14534736842105264,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00048008700626931994,
+      "loss": 3.5804,
+      "step": 3452
+    },
+    {
+      "epoch": 0.1453894736842105,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004800738067716548,
+      "loss": 3.8735,
+      "step": 3453
+    },
+    {
+      "epoch": 0.1454315789473684,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00048006060308231524,
+      "loss": 3.658,
+      "step": 3454
+    },
+    {
+      "epoch": 0.1454736842105263,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00048004739520154184,
+      "loss": 3.5298,
+      "step": 3455
+    },
+    {
+      "epoch": 0.1455157894736842,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00048003418312957515,
+      "loss": 2.9864,
+      "step": 3456
+    },
+    {
+      "epoch": 0.1455578947368421,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00048002096686665597,
+      "loss": 3.5737,
+      "step": 3457
+    },
+    {
+      "epoch": 0.1456,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00048000774641302493,
+      "loss": 3.3253,
+      "step": 3458
+    },
+    {
+      "epoch": 0.1456421052631579,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0004799945217689231,
+      "loss": 3.3111,
+      "step": 3459
+    },
+    {
+      "epoch": 0.14568421052631578,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00047998129293459123,
+      "loss": 3.4269,
+      "step": 3460
+    },
+    {
+      "epoch": 0.14572631578947368,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004799680599102705,
+      "loss": 3.7297,
+      "step": 3461
+    },
+    {
+      "epoch": 0.14576842105263157,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004799548226962019,
+      "loss": 3.2455,
+      "step": 3462
+    },
+    {
+      "epoch": 0.14581052631578947,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004799415812926266,
+      "loss": 3.1751,
+      "step": 3463
+    },
+    {
+      "epoch": 0.14585263157894737,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0004799283356997859,
+      "loss": 3.5139,
+      "step": 3464
+    },
+    {
+      "epoch": 0.14589473684210527,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.000479915085917921,
+      "loss": 3.6596,
+      "step": 3465
+    },
+    {
+      "epoch": 0.14593684210526317,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004799018319472734,
+      "loss": 3.5211,
+      "step": 3466
+    },
+    {
+      "epoch": 0.14597894736842104,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004798885737880846,
+      "loss": 4.0496,
+      "step": 3467
+    },
+    {
+      "epoch": 0.14602105263157894,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.000479875311440596,
+      "loss": 3.496,
+      "step": 3468
+    },
+    {
+      "epoch": 0.14606315789473684,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00047986204490504936,
+      "loss": 3.4902,
+      "step": 3469
+    },
+    {
+      "epoch": 0.14610526315789474,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0004798487741816864,
+      "loss": 3.5114,
+      "step": 3470
+    },
+    {
+      "epoch": 0.14614736842105264,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004798354992707488,
+      "loss": 3.246,
+      "step": 3471
+    },
+    {
+      "epoch": 0.14618947368421054,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004798222201724785,
+      "loss": 3.6993,
+      "step": 3472
+    },
+    {
+      "epoch": 0.14623157894736843,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00047980893688711723,
+      "loss": 3.5203,
+      "step": 3473
+    },
+    {
+      "epoch": 0.1462736842105263,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00047979564941490725,
+      "loss": 3.2323,
+      "step": 3474
+    },
+    {
+      "epoch": 0.1463157894736842,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004797823577560905,
+      "loss": 3.3565,
+      "step": 3475
+    },
+    {
+      "epoch": 0.1463578947368421,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0004797690619109092,
+      "loss": 3.4393,
+      "step": 3476
+    },
+    {
+      "epoch": 0.1464,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004797557618796056,
+      "loss": 3.1628,
+      "step": 3477
+    },
+    {
+      "epoch": 0.1464421052631579,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00047974245766242186,
+      "loss": 3.8626,
+      "step": 3478
+    },
+    {
+      "epoch": 0.1464842105263158,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00047972914925960053,
+      "loss": 3.2123,
+      "step": 3479
+    },
+    {
+      "epoch": 0.14652631578947367,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00047971583667138405,
+      "loss": 3.5529,
+      "step": 3480
+    },
+    {
+      "epoch": 0.14656842105263157,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00047970251989801487,
+      "loss": 3.9338,
+      "step": 3481
+    },
+    {
+      "epoch": 0.14661052631578947,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004796891989397357,
+      "loss": 3.3719,
+      "step": 3482
+    },
+    {
+      "epoch": 0.14665263157894737,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00047967587379678926,
+      "loss": 3.3891,
+      "step": 3483
+    },
+    {
+      "epoch": 0.14669473684210527,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004796625444694182,
+      "loss": 3.6581,
+      "step": 3484
+    },
+    {
+      "epoch": 0.14673684210526317,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004796492109578654,
+      "loss": 3.1821,
+      "step": 3485
+    },
+    {
+      "epoch": 0.14677894736842106,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004796358732623738,
+      "loss": 3.3633,
+      "step": 3486
+    },
+    {
+      "epoch": 0.14682105263157894,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00047962253138318645,
+      "loss": 3.3571,
+      "step": 3487
+    },
+    {
+      "epoch": 0.14686315789473683,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0004796091853205463,
+      "loss": 3.294,
+      "step": 3488
+    },
+    {
+      "epoch": 0.14690526315789473,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00047959583507469665,
+      "loss": 3.5613,
+      "step": 3489
+    },
+    {
+      "epoch": 0.14694736842105263,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00047958248064588067,
+      "loss": 3.4197,
+      "step": 3490
+    },
+    {
+      "epoch": 0.14698947368421053,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00047956912203434156,
+      "loss": 3.3872,
+      "step": 3491
+    },
+    {
+      "epoch": 0.14703157894736843,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004795557592403228,
+      "loss": 3.6493,
+      "step": 3492
+    },
+    {
+      "epoch": 0.14707368421052633,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00047954239226406786,
+      "loss": 3.1529,
+      "step": 3493
+    },
+    {
+      "epoch": 0.1471157894736842,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004795290211058202,
+      "loss": 3.3958,
+      "step": 3494
+    },
+    {
+      "epoch": 0.1471578947368421,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00047951564576582354,
+      "loss": 3.5002,
+      "step": 3495
+    },
+    {
+      "epoch": 0.1472,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004795022662443214,
+      "loss": 3.2861,
+      "step": 3496
+    },
+    {
+      "epoch": 0.1472421052631579,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00047948888254155766,
+      "loss": 3.1008,
+      "step": 3497
+    },
+    {
+      "epoch": 0.1472842105263158,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00047947549465777604,
+      "loss": 3.5641,
+      "step": 3498
+    },
+    {
+      "epoch": 0.1473263157894737,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004794621025932206,
+      "loss": 3.5222,
+      "step": 3499
+    },
+    {
+      "epoch": 0.14736842105263157,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00047944870634813525,
+      "loss": 3.2415,
+      "step": 3500
+    },
+    {
+      "epoch": 0.14741052631578946,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00047943530592276407,
+      "loss": 3.731,
+      "step": 3501
+    },
+    {
+      "epoch": 0.14745263157894736,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004794219013173512,
+      "loss": 3.7334,
+      "step": 3502
+    },
+    {
+      "epoch": 0.14749473684210526,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004794084925321408,
+      "loss": 3.4148,
+      "step": 3503
+    },
+    {
+      "epoch": 0.14753684210526316,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00047939507956737724,
+      "loss": 3.5669,
+      "step": 3504
+    },
+    {
+      "epoch": 0.14757894736842106,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004793816624233048,
+      "loss": 3.0733,
+      "step": 3505
+    },
+    {
+      "epoch": 0.14762105263157896,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004793682411001681,
+      "loss": 3.6561,
+      "step": 3506
+    },
+    {
+      "epoch": 0.14766315789473683,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004793548155982115,
+      "loss": 3.5954,
+      "step": 3507
+    },
+    {
+      "epoch": 0.14770526315789473,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00047934138591767964,
+      "loss": 3.4994,
+      "step": 3508
+    },
+    {
+      "epoch": 0.14774736842105263,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004793279520588171,
+      "loss": 3.0289,
+      "step": 3509
+    },
+    {
+      "epoch": 0.14778947368421053,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004793145140218688,
+      "loss": 3.6115,
+      "step": 3510
+    },
+    {
+      "epoch": 0.14783157894736842,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004793010718070795,
+      "loss": 3.8799,
+      "step": 3511
+    },
+    {
+      "epoch": 0.14787368421052632,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00047928762541469405,
+      "loss": 3.542,
+      "step": 3512
+    },
+    {
+      "epoch": 0.14791578947368422,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004792741748449575,
+      "loss": 3.0782,
+      "step": 3513
+    },
+    {
+      "epoch": 0.1479578947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004792607200981149,
+      "loss": 3.4163,
+      "step": 3514
+    },
+    {
+      "epoch": 0.148,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004792472611744113,
+      "loss": 3.6587,
+      "step": 3515
+    },
+    {
+      "epoch": 0.1480421052631579,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00047923379807409197,
+      "loss": 3.3142,
+      "step": 3516
+    },
+    {
+      "epoch": 0.1480842105263158,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00047922033079740215,
+      "loss": 3.4841,
+      "step": 3517
+    },
+    {
+      "epoch": 0.1481263157894737,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004792068593445873,
+      "loss": 3.3279,
+      "step": 3518
+    },
+    {
+      "epoch": 0.1481684210526316,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004791933837158927,
+      "loss": 3.4343,
+      "step": 3519
+    },
+    {
+      "epoch": 0.1482105263157895,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00047917990391156397,
+      "loss": 3.4753,
+      "step": 3520
+    },
+    {
+      "epoch": 0.14825263157894736,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0004791664199318467,
+      "loss": 3.5413,
+      "step": 3521
+    },
+    {
+      "epoch": 0.14829473684210526,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00047915293177698646,
+      "loss": 3.775,
+      "step": 3522
+    },
+    {
+      "epoch": 0.14833684210526316,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00047913943944722905,
+      "loss": 3.553,
+      "step": 3523
+    },
+    {
+      "epoch": 0.14837894736842105,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0004791259429428203,
+      "loss": 3.7543,
+      "step": 3524
+    },
+    {
+      "epoch": 0.14842105263157895,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004791124422640061,
+      "loss": 4.0371,
+      "step": 3525
+    },
+    {
+      "epoch": 0.14846315789473685,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004790989374110324,
+      "loss": 2.9852,
+      "step": 3526
+    },
+    {
+      "epoch": 0.14850526315789472,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0004790854283841452,
+      "loss": 2.8469,
+      "step": 3527
+    },
+    {
+      "epoch": 0.14854736842105262,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004790719151835907,
+      "loss": 3.8494,
+      "step": 3528
+    },
+    {
+      "epoch": 0.14858947368421052,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00047905839780961504,
+      "loss": 3.2385,
+      "step": 3529
+    },
+    {
+      "epoch": 0.14863157894736842,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00047904487626246453,
+      "loss": 3.5745,
+      "step": 3530
+    },
+    {
+      "epoch": 0.14867368421052632,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00047903135054238547,
+      "loss": 3.3895,
+      "step": 3531
+    },
+    {
+      "epoch": 0.14871578947368422,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00047901782064962425,
+      "loss": 3.3836,
+      "step": 3532
+    },
+    {
+      "epoch": 0.14875789473684212,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00047900428658442746,
+      "loss": 3.6977,
+      "step": 3533
+    },
+    {
+      "epoch": 0.1488,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00047899074834704166,
+      "loss": 3.0294,
+      "step": 3534
+    },
+    {
+      "epoch": 0.1488421052631579,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00047897720593771345,
+      "loss": 3.279,
+      "step": 3535
+    },
+    {
+      "epoch": 0.14888421052631579,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00047896365935668964,
+      "loss": 3.8636,
+      "step": 3536
+    },
+    {
+      "epoch": 0.14892631578947368,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004789501086042169,
+      "loss": 3.4007,
+      "step": 3537
+    },
+    {
+      "epoch": 0.14896842105263158,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00047893655368054223,
+      "loss": 3.2157,
+      "step": 3538
+    },
+    {
+      "epoch": 0.14901052631578948,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004789229945859125,
+      "loss": 3.2495,
+      "step": 3539
+    },
+    {
+      "epoch": 0.14905263157894738,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00047890943132057487,
+      "loss": 3.777,
+      "step": 3540
+    },
+    {
+      "epoch": 0.14909473684210525,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00047889586388477623,
+      "loss": 3.8531,
+      "step": 3541
+    },
+    {
+      "epoch": 0.14913684210526315,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.000478882292278764,
+      "loss": 3.6161,
+      "step": 3542
+    },
+    {
+      "epoch": 0.14917894736842105,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004788687165027853,
+      "loss": 3.3703,
+      "step": 3543
+    },
+    {
+      "epoch": 0.14922105263157895,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0004788551365570875,
+      "loss": 3.7496,
+      "step": 3544
+    },
+    {
+      "epoch": 0.14926315789473685,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00047884155244191804,
+      "loss": 3.7035,
+      "step": 3545
+    },
+    {
+      "epoch": 0.14930526315789475,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004788279641575243,
+      "loss": 3.6786,
+      "step": 3546
+    },
+    {
+      "epoch": 0.14934736842105265,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00047881437170415393,
+      "loss": 3.7735,
+      "step": 3547
+    },
+    {
+      "epoch": 0.14938947368421052,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00047880077508205457,
+      "loss": 3.521,
+      "step": 3548
+    },
+    {
+      "epoch": 0.14943157894736842,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004787871742914739,
+      "loss": 3.4225,
+      "step": 3549
+    },
+    {
+      "epoch": 0.14947368421052631,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00047877356933265973,
+      "loss": 3.6627,
+      "step": 3550
+    },
+    {
+      "epoch": 0.1495157894736842,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00047875996020586,
+      "loss": 3.2554,
+      "step": 3551
+    },
+    {
+      "epoch": 0.1495578947368421,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004787463469113225,
+      "loss": 3.639,
+      "step": 3552
+    },
+    {
+      "epoch": 0.1496,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00047873272944929534,
+      "loss": 3.5686,
+      "step": 3553
+    },
+    {
+      "epoch": 0.14964210526315788,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00047871910782002657,
+      "loss": 3.3285,
+      "step": 3554
+    },
+    {
+      "epoch": 0.14968421052631578,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004787054820237644,
+      "loss": 3.3723,
+      "step": 3555
+    },
+    {
+      "epoch": 0.14972631578947368,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004786918520607572,
+      "loss": 3.0776,
+      "step": 3556
+    },
+    {
+      "epoch": 0.14976842105263158,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00047867821793125297,
+      "loss": 3.2772,
+      "step": 3557
+    },
+    {
+      "epoch": 0.14981052631578948,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00047866457963550036,
+      "loss": 3.5471,
+      "step": 3558
+    },
+    {
+      "epoch": 0.14985263157894738,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00047865093717374774,
+      "loss": 3.4563,
+      "step": 3559
+    },
+    {
+      "epoch": 0.14989473684210528,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004786372905462438,
+      "loss": 3.5108,
+      "step": 3560
+    },
+    {
+      "epoch": 0.14993684210526315,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00047862363975323695,
+      "loss": 3.7157,
+      "step": 3561
+    },
+    {
+      "epoch": 0.14997894736842105,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004786099847949761,
+      "loss": 3.3045,
+      "step": 3562
+    },
+    {
+      "epoch": 0.15002105263157894,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0004785963256717099,
+      "loss": 3.3637,
+      "step": 3563
+    },
+    {
+      "epoch": 0.15006315789473684,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00047858266238368725,
+      "loss": 2.8851,
+      "step": 3564
+    },
+    {
+      "epoch": 0.15010526315789474,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.000478568994931157,
+      "loss": 3.5491,
+      "step": 3565
+    },
+    {
+      "epoch": 0.15014736842105264,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00047855532331436836,
+      "loss": 3.09,
+      "step": 3566
+    },
+    {
+      "epoch": 0.15018947368421054,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004785416475335701,
+      "loss": 3.1478,
+      "step": 3567
+    },
+    {
+      "epoch": 0.1502315789473684,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00047852796758901165,
+      "loss": 3.9951,
+      "step": 3568
+    },
+    {
+      "epoch": 0.1502736842105263,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004785142834809421,
+      "loss": 2.8634,
+      "step": 3569
+    },
+    {
+      "epoch": 0.1503157894736842,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00047850059520961085,
+      "loss": 3.5533,
+      "step": 3570
+    },
+    {
+      "epoch": 0.1503578947368421,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00047848690277526717,
+      "loss": 3.6855,
+      "step": 3571
+    },
+    {
+      "epoch": 0.1504,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004784732061781606,
+      "loss": 3.2511,
+      "step": 3572
+    },
+    {
+      "epoch": 0.1504421052631579,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00047845950541854066,
+      "loss": 3.2379,
+      "step": 3573
+    },
+    {
+      "epoch": 0.15048421052631578,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.000478445800496657,
+      "loss": 3.2198,
+      "step": 3574
+    },
+    {
+      "epoch": 0.15052631578947367,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00047843209141275924,
+      "loss": 3.1124,
+      "step": 3575
+    },
+    {
+      "epoch": 0.15056842105263157,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004784183781670972,
+      "loss": 3.7152,
+      "step": 3576
+    },
+    {
+      "epoch": 0.15061052631578947,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0004784046607599206,
+      "loss": 3.2085,
+      "step": 3577
+    },
+    {
+      "epoch": 0.15065263157894737,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004783909391914795,
+      "loss": 3.5067,
+      "step": 3578
+    },
+    {
+      "epoch": 0.15069473684210527,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004783772134620238,
+      "loss": 3.93,
+      "step": 3579
+    },
+    {
+      "epoch": 0.15073684210526317,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0004783634835718037,
+      "loss": 3.4735,
+      "step": 3580
+    },
+    {
+      "epoch": 0.15077894736842104,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00047834974952106916,
+      "loss": 3.7857,
+      "step": 3581
+    },
+    {
+      "epoch": 0.15082105263157894,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00047833601131007055,
+      "loss": 3.4935,
+      "step": 3582
+    },
+    {
+      "epoch": 0.15086315789473684,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0004783222689390581,
+      "loss": 3.5528,
+      "step": 3583
+    },
+    {
+      "epoch": 0.15090526315789474,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00047830852240828204,
+      "loss": 3.3636,
+      "step": 3584
+    },
+    {
+      "epoch": 0.15094736842105264,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004782947717179931,
+      "loss": 3.7314,
+      "step": 3585
+    },
+    {
+      "epoch": 0.15098947368421053,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004782810168684416,
+      "loss": 3.4479,
+      "step": 3586
+    },
+    {
+      "epoch": 0.15103157894736843,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004782672578598782,
+      "loss": 3.0838,
+      "step": 3587
+    },
+    {
+      "epoch": 0.1510736842105263,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004782534946925535,
+      "loss": 2.8466,
+      "step": 3588
+    },
+    {
+      "epoch": 0.1511157894736842,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004782397273667184,
+      "loss": 3.5797,
+      "step": 3589
+    },
+    {
+      "epoch": 0.1511578947368421,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00047822595588262363,
+      "loss": 3.3296,
+      "step": 3590
+    },
+    {
+      "epoch": 0.1512,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00047821218024052004,
+      "loss": 3.4251,
+      "step": 3591
+    },
+    {
+      "epoch": 0.1512421052631579,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00047819840044065876,
+      "loss": 3.3906,
+      "step": 3592
+    },
+    {
+      "epoch": 0.1512842105263158,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00047818461648329073,
+      "loss": 3.8013,
+      "step": 3593
+    },
+    {
+      "epoch": 0.1513263157894737,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.000478170828368667,
+      "loss": 3.228,
+      "step": 3594
+    },
+    {
+      "epoch": 0.15136842105263157,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00047815703609703893,
+      "loss": 3.4233,
+      "step": 3595
+    },
+    {
+      "epoch": 0.15141052631578947,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004781432396686577,
+      "loss": 3.0338,
+      "step": 3596
+    },
+    {
+      "epoch": 0.15145263157894737,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004781294390837747,
+      "loss": 3.4687,
+      "step": 3597
+    },
+    {
+      "epoch": 0.15149473684210527,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00047811563434264144,
+      "loss": 3.3073,
+      "step": 3598
+    },
+    {
+      "epoch": 0.15153684210526316,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004781018254455093,
+      "loss": 3.5941,
+      "step": 3599
+    },
+    {
+      "epoch": 0.15157894736842106,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004780880123926299,
+      "loss": 3.4236,
+      "step": 3600
+    },
+    {
+      "epoch": 0.15162105263157893,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00047807419518425487,
+      "loss": 3.2942,
+      "step": 3601
+    },
+    {
+      "epoch": 0.15166315789473683,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00047806037382063605,
+      "loss": 3.4013,
+      "step": 3602
+    },
+    {
+      "epoch": 0.15170526315789473,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00047804654830202513,
+      "loss": 3.45,
+      "step": 3603
+    },
+    {
+      "epoch": 0.15174736842105263,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.000478032718628674,
+      "loss": 3.314,
+      "step": 3604
+    },
+    {
+      "epoch": 0.15178947368421053,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004780188848008348,
+      "loss": 3.3036,
+      "step": 3605
+    },
+    {
+      "epoch": 0.15183157894736843,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00047800504681875936,
+      "loss": 3.2694,
+      "step": 3606
+    },
+    {
+      "epoch": 0.15187368421052633,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00047799120468269986,
+      "loss": 3.479,
+      "step": 3607
+    },
+    {
+      "epoch": 0.1519157894736842,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00047797735839290843,
+      "loss": 3.7281,
+      "step": 3608
+    },
+    {
+      "epoch": 0.1519578947368421,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004779635079496375,
+      "loss": 2.9445,
+      "step": 3609
+    },
+    {
+      "epoch": 0.152,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00047794965335313925,
+      "loss": 3.4902,
+      "step": 3610
+    },
+    {
+      "epoch": 0.1520421052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00047793579460366615,
+      "loss": 3.0122,
+      "step": 3611
+    },
+    {
+      "epoch": 0.1520842105263158,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004779219317014707,
+      "loss": 3.3085,
+      "step": 3612
+    },
+    {
+      "epoch": 0.1521263157894737,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004779080646468054,
+      "loss": 3.3256,
+      "step": 3613
+    },
+    {
+      "epoch": 0.1521684210526316,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.000477894193439923,
+      "loss": 3.2456,
+      "step": 3614
+    },
+    {
+      "epoch": 0.15221052631578946,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00047788031808107615,
+      "loss": 3.4174,
+      "step": 3615
+    },
+    {
+      "epoch": 0.15225263157894736,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0004778664385705176,
+      "loss": 3.0583,
+      "step": 3616
+    },
+    {
+      "epoch": 0.15229473684210526,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004778525549085003,
+      "loss": 3.5168,
+      "step": 3617
+    },
+    {
+      "epoch": 0.15233684210526316,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00047783866709527725,
+      "loss": 3.3866,
+      "step": 3618
+    },
+    {
+      "epoch": 0.15237894736842106,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004778247751311013,
+      "loss": 2.9958,
+      "step": 3619
+    },
+    {
+      "epoch": 0.15242105263157896,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0004778108790162256,
+      "loss": 3.2339,
+      "step": 3620
+    },
+    {
+      "epoch": 0.15246315789473686,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00047779697875090346,
+      "loss": 3.7221,
+      "step": 3621
+    },
+    {
+      "epoch": 0.15250526315789473,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0004777830743353879,
+      "loss": 3.3945,
+      "step": 3622
+    },
+    {
+      "epoch": 0.15254736842105263,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004777691657699324,
+      "loss": 3.793,
+      "step": 3623
+    },
+    {
+      "epoch": 0.15258947368421052,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00047775525305479035,
+      "loss": 3.2065,
+      "step": 3624
+    },
+    {
+      "epoch": 0.15263157894736842,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00047774133619021516,
+      "loss": 3.7421,
+      "step": 3625
+    },
+    {
+      "epoch": 0.15267368421052632,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004777274151764604,
+      "loss": 3.5376,
+      "step": 3626
+    },
+    {
+      "epoch": 0.15271578947368422,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004777134900137797,
+      "loss": 3.6357,
+      "step": 3627
+    },
+    {
+      "epoch": 0.1527578947368421,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004776995607024268,
+      "loss": 3.7376,
+      "step": 3628
+    },
+    {
+      "epoch": 0.1528,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0004776856272426555,
+      "loss": 2.9966,
+      "step": 3629
+    },
+    {
+      "epoch": 0.1528421052631579,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0004776716896347195,
+      "loss": 3.1757,
+      "step": 3630
+    },
+    {
+      "epoch": 0.1528842105263158,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00047765774787887286,
+      "loss": 3.6047,
+      "step": 3631
+    },
+    {
+      "epoch": 0.1529263157894737,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004776438019753695,
+      "loss": 3.292,
+      "step": 3632
+    },
+    {
+      "epoch": 0.1529684210526316,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004776298519244635,
+      "loss": 3.5381,
+      "step": 3633
+    },
+    {
+      "epoch": 0.15301052631578949,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00047761589772640913,
+      "loss": 3.1568,
+      "step": 3634
+    },
+    {
+      "epoch": 0.15305263157894736,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004776019393814605,
+      "loss": 3.3782,
+      "step": 3635
+    },
+    {
+      "epoch": 0.15309473684210526,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00047758797688987197,
+      "loss": 3.7427,
+      "step": 3636
+    },
+    {
+      "epoch": 0.15313684210526315,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00047757401025189794,
+      "loss": 3.1935,
+      "step": 3637
+    },
+    {
+      "epoch": 0.15317894736842105,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00047756003946779283,
+      "loss": 3.6259,
+      "step": 3638
+    },
+    {
+      "epoch": 0.15322105263157895,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00047754606453781116,
+      "loss": 3.1283,
+      "step": 3639
+    },
+    {
+      "epoch": 0.15326315789473685,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004775320854622075,
+      "loss": 3.7776,
+      "step": 3640
+    },
+    {
+      "epoch": 0.15330526315789475,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0004775181022412366,
+      "loss": 3.5322,
+      "step": 3641
+    },
+    {
+      "epoch": 0.15334736842105262,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00047750411487515324,
+      "loss": 3.6808,
+      "step": 3642
+    },
+    {
+      "epoch": 0.15338947368421052,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00047749012336421214,
+      "loss": 3.767,
+      "step": 3643
+    },
+    {
+      "epoch": 0.15343157894736842,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00047747612770866835,
+      "loss": 3.5188,
+      "step": 3644
+    },
+    {
+      "epoch": 0.15347368421052632,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00047746212790877673,
+      "loss": 3.615,
+      "step": 3645
+    },
+    {
+      "epoch": 0.15351578947368422,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004774481239647925,
+      "loss": 3.8009,
+      "step": 3646
+    },
+    {
+      "epoch": 0.15355789473684212,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004774341158769706,
+      "loss": 3.3402,
+      "step": 3647
+    },
+    {
+      "epoch": 0.1536,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00047742010364556635,
+      "loss": 3.5347,
+      "step": 3648
+    },
+    {
+      "epoch": 0.15364210526315789,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.000477406087270835,
+      "loss": 3.5282,
+      "step": 3649
+    },
+    {
+      "epoch": 0.15368421052631578,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.000477392066753032,
+      "loss": 3.3169,
+      "step": 3650
+    },
+    {
+      "epoch": 0.15372631578947368,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00047737804209241265,
+      "loss": 3.406,
+      "step": 3651
+    },
+    {
+      "epoch": 0.15376842105263158,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00047736401328923265,
+      "loss": 3.3344,
+      "step": 3652
+    },
+    {
+      "epoch": 0.15381052631578948,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00047734998034374733,
+      "loss": 2.9495,
+      "step": 3653
+    },
+    {
+      "epoch": 0.15385263157894738,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00047733594325621257,
+      "loss": 3.7441,
+      "step": 3654
+    },
+    {
+      "epoch": 0.15389473684210525,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00047732190202688397,
+      "loss": 3.4942,
+      "step": 3655
+    },
+    {
+      "epoch": 0.15393684210526315,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00047730785665601745,
+      "loss": 3.6626,
+      "step": 3656
+    },
+    {
+      "epoch": 0.15397894736842105,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00047729380714386883,
+      "loss": 3.3828,
+      "step": 3657
+    },
+    {
+      "epoch": 0.15402105263157895,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00047727975349069415,
+      "loss": 3.2138,
+      "step": 3658
+    },
+    {
+      "epoch": 0.15406315789473685,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004772656956967494,
+      "loss": 3.0591,
+      "step": 3659
+    },
+    {
+      "epoch": 0.15410526315789475,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00047725163376229063,
+      "loss": 3.4689,
+      "step": 3660
+    },
+    {
+      "epoch": 0.15414736842105264,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004772375676875741,
+      "loss": 3.5733,
+      "step": 3661
+    },
+    {
+      "epoch": 0.15418947368421052,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004772234974728562,
+      "loss": 3.3081,
+      "step": 3662
+    },
+    {
+      "epoch": 0.15423157894736841,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004772094231183931,
+      "loss": 3.3391,
+      "step": 3663
+    },
+    {
+      "epoch": 0.1542736842105263,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00047719534462444126,
+      "loss": 3.4436,
+      "step": 3664
+    },
+    {
+      "epoch": 0.1543157894736842,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004771812619912571,
+      "loss": 3.4334,
+      "step": 3665
+    },
+    {
+      "epoch": 0.1543578947368421,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00047716717521909734,
+      "loss": 3.4928,
+      "step": 3666
+    },
+    {
+      "epoch": 0.1544,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004771530843082186,
+      "loss": 3.8065,
+      "step": 3667
+    },
+    {
+      "epoch": 0.1544421052631579,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00047713898925887756,
+      "loss": 3.7351,
+      "step": 3668
+    },
+    {
+      "epoch": 0.15448421052631578,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.000477124890071331,
+      "loss": 3.0596,
+      "step": 3669
+    },
+    {
+      "epoch": 0.15452631578947368,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00047711078674583573,
+      "loss": 3.7457,
+      "step": 3670
+    },
+    {
+      "epoch": 0.15456842105263158,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0004770966792826489,
+      "loss": 3.5734,
+      "step": 3671
+    },
+    {
+      "epoch": 0.15461052631578948,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0004770825676820273,
+      "loss": 3.1431,
+      "step": 3672
+    },
+    {
+      "epoch": 0.15465263157894737,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004770684519442281,
+      "loss": 3.5353,
+      "step": 3673
+    },
+    {
+      "epoch": 0.15469473684210527,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00047705433206950853,
+      "loss": 3.8456,
+      "step": 3674
+    },
+    {
+      "epoch": 0.15473684210526314,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004770402080581259,
+      "loss": 3.717,
+      "step": 3675
+    },
+    {
+      "epoch": 0.15477894736842104,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004770260799103374,
+      "loss": 3.7989,
+      "step": 3676
+    },
+    {
+      "epoch": 0.15482105263157894,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00047701194762640037,
+      "loss": 3.2728,
+      "step": 3677
+    },
+    {
+      "epoch": 0.15486315789473684,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004769978112065725,
+      "loss": 3.4157,
+      "step": 3678
+    },
+    {
+      "epoch": 0.15490526315789474,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00047698367065111114,
+      "loss": 3.5842,
+      "step": 3679
+    },
+    {
+      "epoch": 0.15494736842105264,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00047696952596027397,
+      "loss": 3.2744,
+      "step": 3680
+    },
+    {
+      "epoch": 0.15498947368421054,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004769553771343188,
+      "loss": 3.4852,
+      "step": 3681
+    },
+    {
+      "epoch": 0.1550315789473684,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004769412241735033,
+      "loss": 3.3286,
+      "step": 3682
+    },
+    {
+      "epoch": 0.1550736842105263,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00047692706707808536,
+      "loss": 3.5173,
+      "step": 3683
+    },
+    {
+      "epoch": 0.1551157894736842,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00047691290584832286,
+      "loss": 3.1805,
+      "step": 3684
+    },
+    {
+      "epoch": 0.1551578947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004768987404844738,
+      "loss": 3.1129,
+      "step": 3685
+    },
+    {
+      "epoch": 0.1552,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004768845709867964,
+      "loss": 3.4441,
+      "step": 3686
+    },
+    {
+      "epoch": 0.1552421052631579,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00047687039735554854,
+      "loss": 3.3472,
+      "step": 3687
+    },
+    {
+      "epoch": 0.1552842105263158,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004768562195909887,
+      "loss": 3.4257,
+      "step": 3688
+    },
+    {
+      "epoch": 0.15532631578947367,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.000476842037693375,
+      "loss": 3.6577,
+      "step": 3689
+    },
+    {
+      "epoch": 0.15536842105263157,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00047682785166296593,
+      "loss": 3.8201,
+      "step": 3690
+    },
+    {
+      "epoch": 0.15541052631578947,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00047681366150002,
+      "loss": 3.284,
+      "step": 3691
+    },
+    {
+      "epoch": 0.15545263157894737,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0004767994672047956,
+      "loss": 3.3328,
+      "step": 3692
+    },
+    {
+      "epoch": 0.15549473684210527,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00047678526877755135,
+      "loss": 3.1374,
+      "step": 3693
+    },
+    {
+      "epoch": 0.15553684210526317,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00047677106621854597,
+      "loss": 3.5267,
+      "step": 3694
+    },
+    {
+      "epoch": 0.15557894736842104,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00047675685952803826,
+      "loss": 3.0731,
+      "step": 3695
+    },
+    {
+      "epoch": 0.15562105263157894,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.000476742648706287,
+      "loss": 3.0489,
+      "step": 3696
+    },
+    {
+      "epoch": 0.15566315789473684,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00047672843375355103,
+      "loss": 3.9336,
+      "step": 3697
+    },
+    {
+      "epoch": 0.15570526315789474,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0004767142146700894,
+      "loss": 3.2461,
+      "step": 3698
+    },
+    {
+      "epoch": 0.15574736842105263,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00047669999145616117,
+      "loss": 3.6933,
+      "step": 3699
+    },
+    {
+      "epoch": 0.15578947368421053,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00047668576411202546,
+      "loss": 3.2744,
+      "step": 3700
+    },
+    {
+      "epoch": 0.15583157894736843,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004766715326379415,
+      "loss": 3.7561,
+      "step": 3701
+    },
+    {
+      "epoch": 0.1558736842105263,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004766572970341685,
+      "loss": 3.5275,
+      "step": 3702
+    },
+    {
+      "epoch": 0.1559157894736842,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004766430573009659,
+      "loss": 3.4197,
+      "step": 3703
+    },
+    {
+      "epoch": 0.1559578947368421,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004766288134385931,
+      "loss": 3.5674,
+      "step": 3704
+    },
+    {
+      "epoch": 0.156,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004766145654473095,
+      "loss": 3.2447,
+      "step": 3705
+    },
+    {
+      "epoch": 0.1560421052631579,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00047660031332737493,
+      "loss": 3.7222,
+      "step": 3706
+    },
+    {
+      "epoch": 0.1560842105263158,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00047658605707904877,
+      "loss": 3.4923,
+      "step": 3707
+    },
+    {
+      "epoch": 0.1561263157894737,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00047657179670259096,
+      "loss": 3.2968,
+      "step": 3708
+    },
+    {
+      "epoch": 0.15616842105263157,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00047655753219826117,
+      "loss": 3.9744,
+      "step": 3709
+    },
+    {
+      "epoch": 0.15621052631578947,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00047654326356631933,
+      "loss": 3.9199,
+      "step": 3710
+    },
+    {
+      "epoch": 0.15625263157894737,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00047652899080702544,
+      "loss": 3.4477,
+      "step": 3711
+    },
+    {
+      "epoch": 0.15629473684210526,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004765147139206395,
+      "loss": 3.721,
+      "step": 3712
+    },
+    {
+      "epoch": 0.15633684210526316,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00047650043290742164,
+      "loss": 3.2584,
+      "step": 3713
+    },
+    {
+      "epoch": 0.15637894736842106,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00047648614776763193,
+      "loss": 3.4564,
+      "step": 3714
+    },
+    {
+      "epoch": 0.15642105263157896,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00047647185850153073,
+      "loss": 3.3904,
+      "step": 3715
+    },
+    {
+      "epoch": 0.15646315789473683,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004764575651093784,
+      "loss": 3.2006,
+      "step": 3716
+    },
+    {
+      "epoch": 0.15650526315789473,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00047644326759143534,
+      "loss": 3.6703,
+      "step": 3717
+    },
+    {
+      "epoch": 0.15654736842105263,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.000476428965947962,
+      "loss": 3.3534,
+      "step": 3718
+    },
+    {
+      "epoch": 0.15658947368421053,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004764146601792189,
+      "loss": 3.8089,
+      "step": 3719
+    },
+    {
+      "epoch": 0.15663157894736843,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00047640035028546674,
+      "loss": 3.2193,
+      "step": 3720
+    },
+    {
+      "epoch": 0.15667368421052633,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004763860362669662,
+      "loss": 3.7856,
+      "step": 3721
+    },
+    {
+      "epoch": 0.1567157894736842,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00047637171812397804,
+      "loss": 3.4054,
+      "step": 3722
+    },
+    {
+      "epoch": 0.1567578947368421,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00047635739585676323,
+      "loss": 3.1605,
+      "step": 3723
+    },
+    {
+      "epoch": 0.1568,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004763430694655826,
+      "loss": 3.3979,
+      "step": 3724
+    },
+    {
+      "epoch": 0.1568421052631579,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00047632873895069717,
+      "loss": 3.9123,
+      "step": 3725
+    },
+    {
+      "epoch": 0.1568842105263158,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00047631440431236807,
+      "loss": 3.3996,
+      "step": 3726
+    },
+    {
+      "epoch": 0.1569263157894737,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004763000655508564,
+      "loss": 3.3632,
+      "step": 3727
+    },
+    {
+      "epoch": 0.1569684210526316,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00047628572266642346,
+      "loss": 3.5738,
+      "step": 3728
+    },
+    {
+      "epoch": 0.15701052631578946,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004762713756593304,
+      "loss": 3.3733,
+      "step": 3729
+    },
+    {
+      "epoch": 0.15705263157894736,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004762570245298389,
+      "loss": 3.9693,
+      "step": 3730
+    },
+    {
+      "epoch": 0.15709473684210526,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00047624266927821015,
+      "loss": 3.5044,
+      "step": 3731
+    },
+    {
+      "epoch": 0.15713684210526316,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004762283099047059,
+      "loss": 3.719,
+      "step": 3732
+    },
+    {
+      "epoch": 0.15717894736842106,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00047621394640958754,
+      "loss": 3.3341,
+      "step": 3733
+    },
+    {
+      "epoch": 0.15722105263157896,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004761995787931169,
+      "loss": 3.5223,
+      "step": 3734
+    },
+    {
+      "epoch": 0.15726315789473685,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004761852070555557,
+      "loss": 3.2013,
+      "step": 3735
+    },
+    {
+      "epoch": 0.15730526315789473,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00047617083119716583,
+      "loss": 3.6286,
+      "step": 3736
+    },
+    {
+      "epoch": 0.15734736842105262,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004761564512182091,
+      "loss": 3.1929,
+      "step": 3737
+    },
+    {
+      "epoch": 0.15738947368421052,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004761420671189476,
+      "loss": 3.4403,
+      "step": 3738
+    },
+    {
+      "epoch": 0.15743157894736842,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004761276788996433,
+      "loss": 3.5543,
+      "step": 3739
+    },
+    {
+      "epoch": 0.15747368421052632,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004761132865605584,
+      "loss": 3.4269,
+      "step": 3740
+    },
+    {
+      "epoch": 0.15751578947368422,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004760988901019551,
+      "loss": 3.7293,
+      "step": 3741
+    },
+    {
+      "epoch": 0.15755789473684212,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0004760844895240957,
+      "loss": 3.4919,
+      "step": 3742
+    },
+    {
+      "epoch": 0.1576,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00047607008482724244,
+      "loss": 3.3613,
+      "step": 3743
+    },
+    {
+      "epoch": 0.1576421052631579,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004760556760116579,
+      "loss": 3.3115,
+      "step": 3744
+    },
+    {
+      "epoch": 0.1576842105263158,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004760412630776046,
+      "loss": 3.8654,
+      "step": 3745
+    },
+    {
+      "epoch": 0.1577263157894737,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00047602684602534505,
+      "loss": 3.4589,
+      "step": 3746
+    },
+    {
+      "epoch": 0.15776842105263159,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004760124248551419,
+      "loss": 3.1835,
+      "step": 3747
+    },
+    {
+      "epoch": 0.15781052631578948,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.000475997999567258,
+      "loss": 3.7373,
+      "step": 3748
+    },
+    {
+      "epoch": 0.15785263157894736,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00047598357016195603,
+      "loss": 3.8226,
+      "step": 3749
+    },
+    {
+      "epoch": 0.15789473684210525,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00047596913663949895,
+      "loss": 3.135,
+      "step": 3750
+    },
+    {
+      "epoch": 0.15793684210526315,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00047595469900014965,
+      "loss": 3.3915,
+      "step": 3751
+    },
+    {
+      "epoch": 0.15797894736842105,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004759402572441712,
+      "loss": 3.1679,
+      "step": 3752
+    },
+    {
+      "epoch": 0.15802105263157895,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004759258113718269,
+      "loss": 3.5198,
+      "step": 3753
+    },
+    {
+      "epoch": 0.15806315789473685,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00047591136138337963,
+      "loss": 3.5471,
+      "step": 3754
+    },
+    {
+      "epoch": 0.15810526315789475,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00047589690727909285,
+      "loss": 3.4873,
+      "step": 3755
+    },
+    {
+      "epoch": 0.15814736842105262,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00047588244905922975,
+      "loss": 3.2597,
+      "step": 3756
+    },
+    {
+      "epoch": 0.15818947368421052,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004758679867240539,
+      "loss": 3.579,
+      "step": 3757
+    },
+    {
+      "epoch": 0.15823157894736842,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004758535202738287,
+      "loss": 3.6976,
+      "step": 3758
+    },
+    {
+      "epoch": 0.15827368421052632,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004758390497088178,
+      "loss": 2.9278,
+      "step": 3759
+    },
+    {
+      "epoch": 0.15831578947368422,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004758245750292847,
+      "loss": 3.2774,
+      "step": 3760
+    },
+    {
+      "epoch": 0.15835789473684211,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0004758100962354931,
+      "loss": 3.2356,
+      "step": 3761
+    },
+    {
+      "epoch": 0.1584,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0004757956133277069,
+      "loss": 3.2109,
+      "step": 3762
+    },
+    {
+      "epoch": 0.15844210526315788,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00047578112630618994,
+      "loss": 3.089,
+      "step": 3763
+    },
+    {
+      "epoch": 0.15848421052631578,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00047576663517120624,
+      "loss": 3.2914,
+      "step": 3764
+    },
+    {
+      "epoch": 0.15852631578947368,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004757521399230196,
+      "loss": 3.529,
+      "step": 3765
+    },
+    {
+      "epoch": 0.15856842105263158,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004757376405618943,
+      "loss": 3.5075,
+      "step": 3766
+    },
+    {
+      "epoch": 0.15861052631578948,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00047572313708809435,
+      "loss": 3.6595,
+      "step": 3767
+    },
+    {
+      "epoch": 0.15865263157894738,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004757086295018841,
+      "loss": 3.4359,
+      "step": 3768
+    },
+    {
+      "epoch": 0.15869473684210525,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00047569411780352776,
+      "loss": 3.3874,
+      "step": 3769
+    },
+    {
+      "epoch": 0.15873684210526315,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004756796019932898,
+      "loss": 3.4118,
+      "step": 3770
+    },
+    {
+      "epoch": 0.15877894736842105,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00047566508207143467,
+      "loss": 3.8719,
+      "step": 3771
+    },
+    {
+      "epoch": 0.15882105263157895,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004756505580382269,
+      "loss": 3.8064,
+      "step": 3772
+    },
+    {
+      "epoch": 0.15886315789473685,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00047563602989393106,
+      "loss": 3.4762,
+      "step": 3773
+    },
+    {
+      "epoch": 0.15890526315789474,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00047562149763881185,
+      "loss": 3.2884,
+      "step": 3774
+    },
+    {
+      "epoch": 0.15894736842105264,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00047560696127313406,
+      "loss": 3.764,
+      "step": 3775
+    },
+    {
+      "epoch": 0.1589894736842105,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004755924207971626,
+      "loss": 3.3026,
+      "step": 3776
+    },
+    {
+      "epoch": 0.1590315789473684,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00047557787621116214,
+      "loss": 3.6832,
+      "step": 3777
+    },
+    {
+      "epoch": 0.1590736842105263,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0004755633275153979,
+      "loss": 3.3365,
+      "step": 3778
+    },
+    {
+      "epoch": 0.1591157894736842,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004755487747101349,
+      "loss": 3.6425,
+      "step": 3779
+    },
+    {
+      "epoch": 0.1591578947368421,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004755342177956382,
+      "loss": 3.5729,
+      "step": 3780
+    },
+    {
+      "epoch": 0.1592,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00047551965677217304,
+      "loss": 3.4463,
+      "step": 3781
+    },
+    {
+      "epoch": 0.1592421052631579,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00047550509164000465,
+      "loss": 3.0716,
+      "step": 3782
+    },
+    {
+      "epoch": 0.15928421052631578,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00047549052239939855,
+      "loss": 3.2612,
+      "step": 3783
+    },
+    {
+      "epoch": 0.15932631578947368,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00047547594905062,
+      "loss": 3.5127,
+      "step": 3784
+    },
+    {
+      "epoch": 0.15936842105263158,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00047546137159393465,
+      "loss": 3.3692,
+      "step": 3785
+    },
+    {
+      "epoch": 0.15941052631578947,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00047544679002960795,
+      "loss": 3.2505,
+      "step": 3786
+    },
+    {
+      "epoch": 0.15945263157894737,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004754322043579056,
+      "loss": 3.6312,
+      "step": 3787
+    },
+    {
+      "epoch": 0.15949473684210527,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00047541761457909346,
+      "loss": 3.9617,
+      "step": 3788
+    },
+    {
+      "epoch": 0.15953684210526317,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00047540302069343715,
+      "loss": 3.4276,
+      "step": 3789
+    },
+    {
+      "epoch": 0.15957894736842104,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004753884227012027,
+      "loss": 3.1666,
+      "step": 3790
+    },
+    {
+      "epoch": 0.15962105263157894,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.000475373820602656,
+      "loss": 3.4347,
+      "step": 3791
+    },
+    {
+      "epoch": 0.15966315789473684,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00047535921439806306,
+      "loss": 3.2315,
+      "step": 3792
+    },
+    {
+      "epoch": 0.15970526315789474,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00047534460408768997,
+      "loss": 3.122,
+      "step": 3793
+    },
+    {
+      "epoch": 0.15974736842105264,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004753299896718031,
+      "loss": 3.6026,
+      "step": 3794
+    },
+    {
+      "epoch": 0.15978947368421054,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004753153711506685,
+      "loss": 2.9973,
+      "step": 3795
+    },
+    {
+      "epoch": 0.1598315789473684,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004753007485245525,
+      "loss": 3.1226,
+      "step": 3796
+    },
+    {
+      "epoch": 0.1598736842105263,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00047528612179372165,
+      "loss": 3.3758,
+      "step": 3797
+    },
+    {
+      "epoch": 0.1599157894736842,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00047527149095844234,
+      "loss": 3.4076,
+      "step": 3798
+    },
+    {
+      "epoch": 0.1599578947368421,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004752568560189812,
+      "loss": 3.3045,
+      "step": 3799
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00047524221697560476,
+      "loss": 3.2325,
+      "step": 3800
+    },
+    {
+      "epoch": 0.1600421052631579,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00047522757382857986,
+      "loss": 3.5248,
+      "step": 3801
+    },
+    {
+      "epoch": 0.1600842105263158,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00047521292657817305,
+      "loss": 2.9584,
+      "step": 3802
+    },
+    {
+      "epoch": 0.16012631578947367,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004751982752246514,
+      "loss": 3.4668,
+      "step": 3803
+    },
+    {
+      "epoch": 0.16016842105263157,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00047518361976828184,
+      "loss": 3.263,
+      "step": 3804
+    },
+    {
+      "epoch": 0.16021052631578947,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0004751689602093312,
+      "loss": 3.3976,
+      "step": 3805
+    },
+    {
+      "epoch": 0.16025263157894737,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00047515429654806674,
+      "loss": 3.3254,
+      "step": 3806
+    },
+    {
+      "epoch": 0.16029473684210527,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00047513962878475556,
+      "loss": 3.0458,
+      "step": 3807
+    },
+    {
+      "epoch": 0.16033684210526317,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004751249569196648,
+      "loss": 3.6979,
+      "step": 3808
+    },
+    {
+      "epoch": 0.16037894736842107,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00047511028095306186,
+      "loss": 3.3904,
+      "step": 3809
+    },
+    {
+      "epoch": 0.16042105263157894,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004750956008852141,
+      "loss": 4.0054,
+      "step": 3810
+    },
+    {
+      "epoch": 0.16046315789473684,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.00047508091671638895,
+      "loss": 3.0281,
+      "step": 3811
+    },
+    {
+      "epoch": 0.16050526315789473,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.000475066228446854,
+      "loss": 3.5083,
+      "step": 3812
+    },
+    {
+      "epoch": 0.16054736842105263,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004750515360768768,
+      "loss": 2.986,
+      "step": 3813
+    },
+    {
+      "epoch": 0.16058947368421053,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00047503683960672504,
+      "loss": 3.5373,
+      "step": 3814
+    },
+    {
+      "epoch": 0.16063157894736843,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004750221390366665,
+      "loss": 3.4918,
+      "step": 3815
+    },
+    {
+      "epoch": 0.16067368421052633,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00047500743436696893,
+      "loss": 3.5649,
+      "step": 3816
+    },
+    {
+      "epoch": 0.1607157894736842,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004749927255979002,
+      "loss": 3.7729,
+      "step": 3817
+    },
+    {
+      "epoch": 0.1607578947368421,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00047497801272972844,
+      "loss": 3.7871,
+      "step": 3818
+    },
+    {
+      "epoch": 0.1608,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00047496329576272167,
+      "loss": 3.1167,
+      "step": 3819
+    },
+    {
+      "epoch": 0.1608421052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00047494857469714794,
+      "loss": 3.0299,
+      "step": 3820
+    },
+    {
+      "epoch": 0.1608842105263158,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00047493384953327546,
+      "loss": 3.4207,
+      "step": 3821
+    },
+    {
+      "epoch": 0.1609263157894737,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00047491912027137253,
+      "loss": 3.7361,
+      "step": 3822
+    },
+    {
+      "epoch": 0.16096842105263157,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00047490438691170754,
+      "loss": 3.4605,
+      "step": 3823
+    },
+    {
+      "epoch": 0.16101052631578947,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00047488964945454883,
+      "loss": 3.5144,
+      "step": 3824
+    },
+    {
+      "epoch": 0.16105263157894736,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00047487490790016495,
+      "loss": 3.9343,
+      "step": 3825
+    },
+    {
+      "epoch": 0.16109473684210526,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004748601622488244,
+      "loss": 3.5133,
+      "step": 3826
+    },
+    {
+      "epoch": 0.16113684210526316,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00047484541250079605,
+      "loss": 3.2526,
+      "step": 3827
+    },
+    {
+      "epoch": 0.16117894736842106,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00047483065865634835,
+      "loss": 3.6039,
+      "step": 3828
+    },
+    {
+      "epoch": 0.16122105263157896,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00047481590071575026,
+      "loss": 3.2024,
+      "step": 3829
+    },
+    {
+      "epoch": 0.16126315789473683,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004748011386792706,
+      "loss": 3.0853,
+      "step": 3830
+    },
+    {
+      "epoch": 0.16130526315789473,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0004747863725471783,
+      "loss": 3.3361,
+      "step": 3831
+    },
+    {
+      "epoch": 0.16134736842105263,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00047477160231974245,
+      "loss": 3.5184,
+      "step": 3832
+    },
+    {
+      "epoch": 0.16138947368421053,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004747568279972321,
+      "loss": 3.8712,
+      "step": 3833
+    },
+    {
+      "epoch": 0.16143157894736843,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00047474204957991635,
+      "loss": 3.1647,
+      "step": 3834
+    },
+    {
+      "epoch": 0.16147368421052632,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0004747272670680646,
+      "loss": 3.6008,
+      "step": 3835
+    },
+    {
+      "epoch": 0.16151578947368422,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.000474712480461946,
+      "loss": 3.2742,
+      "step": 3836
+    },
+    {
+      "epoch": 0.1615578947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00047469768976183006,
+      "loss": 3.6445,
+      "step": 3837
+    },
+    {
+      "epoch": 0.1616,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00047468289496798626,
+      "loss": 3.7055,
+      "step": 3838
+    },
+    {
+      "epoch": 0.1616421052631579,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004746680960806841,
+      "loss": 3.2288,
+      "step": 3839
+    },
+    {
+      "epoch": 0.1616842105263158,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004746532931001932,
+      "loss": 3.508,
+      "step": 3840
+    },
+    {
+      "epoch": 0.1617263157894737,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00047463848602678326,
+      "loss": 3.2557,
+      "step": 3841
+    },
+    {
+      "epoch": 0.1617684210526316,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00047462367486072397,
+      "loss": 3.515,
+      "step": 3842
+    },
+    {
+      "epoch": 0.16181052631578946,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0004746088596022854,
+      "loss": 3.4366,
+      "step": 3843
+    },
+    {
+      "epoch": 0.16185263157894736,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00047459404025173717,
+      "loss": 3.4384,
+      "step": 3844
+    },
+    {
+      "epoch": 0.16189473684210526,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004745792168093495,
+      "loss": 3.5839,
+      "step": 3845
+    },
+    {
+      "epoch": 0.16193684210526316,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004745643892753923,
+      "loss": 3.6662,
+      "step": 3846
+    },
+    {
+      "epoch": 0.16197894736842106,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00047454955765013587,
+      "loss": 3.2906,
+      "step": 3847
+    },
+    {
+      "epoch": 0.16202105263157895,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00047453472193385025,
+      "loss": 3.3933,
+      "step": 3848
+    },
+    {
+      "epoch": 0.16206315789473685,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00047451988212680586,
+      "loss": 3.6889,
+      "step": 3849
+    },
+    {
+      "epoch": 0.16210526315789472,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00047450503822927295,
+      "loss": 3.2869,
+      "step": 3850
+    },
+    {
+      "epoch": 0.16214736842105262,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00047449019024152207,
+      "loss": 3.1754,
+      "step": 3851
+    },
+    {
+      "epoch": 0.16218947368421052,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004744753381638237,
+      "loss": 3.2455,
+      "step": 3852
+    },
+    {
+      "epoch": 0.16223157894736842,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004744604819964484,
+      "loss": 3.0194,
+      "step": 3853
+    },
+    {
+      "epoch": 0.16227368421052632,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004744456217396669,
+      "loss": 3.1745,
+      "step": 3854
+    },
+    {
+      "epoch": 0.16231578947368422,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00047443075739374985,
+      "loss": 3.6739,
+      "step": 3855
+    },
+    {
+      "epoch": 0.16235789473684212,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.000474415888958968,
+      "loss": 3.1853,
+      "step": 3856
+    },
+    {
+      "epoch": 0.1624,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0004744010164355925,
+      "loss": 3.4155,
+      "step": 3857
+    },
+    {
+      "epoch": 0.1624421052631579,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00047438613982389403,
+      "loss": 3.6003,
+      "step": 3858
+    },
+    {
+      "epoch": 0.1624842105263158,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004743712591241438,
+      "loss": 3.3342,
+      "step": 3859
+    },
+    {
+      "epoch": 0.16252631578947369,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004743563743366128,
+      "loss": 3.5112,
+      "step": 3860
+    },
+    {
+      "epoch": 0.16256842105263158,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004743414854615723,
+      "loss": 3.3818,
+      "step": 3861
+    },
+    {
+      "epoch": 0.16261052631578948,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00047432659249929356,
+      "loss": 3.4501,
+      "step": 3862
+    },
+    {
+      "epoch": 0.16265263157894738,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004743116954500478,
+      "loss": 4.0091,
+      "step": 3863
+    },
+    {
+      "epoch": 0.16269473684210525,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00047429679431410654,
+      "loss": 3.5009,
+      "step": 3864
+    },
+    {
+      "epoch": 0.16273684210526315,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004742818890917412,
+      "loss": 3.1644,
+      "step": 3865
+    },
+    {
+      "epoch": 0.16277894736842105,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0004742669797832234,
+      "loss": 3.2342,
+      "step": 3866
+    },
+    {
+      "epoch": 0.16282105263157895,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00047425206638882465,
+      "loss": 3.1711,
+      "step": 3867
+    },
+    {
+      "epoch": 0.16286315789473685,
+      "grad_norm": 0.81640625,
+      "learning_rate": 0.0004742371489088169,
+      "loss": 3.1086,
+      "step": 3868
+    },
+    {
+      "epoch": 0.16290526315789475,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00047422222734347163,
+      "loss": 3.2091,
+      "step": 3869
+    },
+    {
+      "epoch": 0.16294736842105262,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004742073016930608,
+      "loss": 3.4166,
+      "step": 3870
+    },
+    {
+      "epoch": 0.16298947368421052,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00047419237195785646,
+      "loss": 3.3303,
+      "step": 3871
+    },
+    {
+      "epoch": 0.16303157894736842,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00047417743813813054,
+      "loss": 3.4595,
+      "step": 3872
+    },
+    {
+      "epoch": 0.16307368421052632,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00047416250023415497,
+      "loss": 3.305,
+      "step": 3873
+    },
+    {
+      "epoch": 0.16311578947368421,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00047414755824620215,
+      "loss": 3.7613,
+      "step": 3874
+    },
+    {
+      "epoch": 0.1631578947368421,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00047413261217454407,
+      "loss": 3.4918,
+      "step": 3875
+    },
+    {
+      "epoch": 0.1632,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00047411766201945323,
+      "loss": 3.0577,
+      "step": 3876
+    },
+    {
+      "epoch": 0.16324210526315788,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00047410270778120193,
+      "loss": 3.3608,
+      "step": 3877
+    },
+    {
+      "epoch": 0.16328421052631578,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004740877494600625,
+      "loss": 3.3662,
+      "step": 3878
+    },
+    {
+      "epoch": 0.16332631578947368,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00047407278705630763,
+      "loss": 3.5284,
+      "step": 3879
+    },
+    {
+      "epoch": 0.16336842105263158,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0004740578205702099,
+      "loss": 3.7682,
+      "step": 3880
+    },
+    {
+      "epoch": 0.16341052631578948,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00047404285000204183,
+      "loss": 3.6055,
+      "step": 3881
+    },
+    {
+      "epoch": 0.16345263157894738,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004740278753520764,
+      "loss": 3.4189,
+      "step": 3882
+    },
+    {
+      "epoch": 0.16349473684210528,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0004740128966205862,
+      "loss": 3.3415,
+      "step": 3883
+    },
+    {
+      "epoch": 0.16353684210526315,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00047399791380784426,
+      "loss": 3.6714,
+      "step": 3884
+    },
+    {
+      "epoch": 0.16357894736842105,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004739829269141235,
+      "loss": 3.5951,
+      "step": 3885
+    },
+    {
+      "epoch": 0.16362105263157894,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.000473967935939697,
+      "loss": 3.6779,
+      "step": 3886
+    },
+    {
+      "epoch": 0.16366315789473684,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00047395294088483785,
+      "loss": 3.3035,
+      "step": 3887
+    },
+    {
+      "epoch": 0.16370526315789474,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004739379417498193,
+      "loss": 3.0263,
+      "step": 3888
+    },
+    {
+      "epoch": 0.16374736842105264,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0004739229385349145,
+      "loss": 3.5854,
+      "step": 3889
+    },
+    {
+      "epoch": 0.1637894736842105,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0004739079312403969,
+      "loss": 3.5468,
+      "step": 3890
+    },
+    {
+      "epoch": 0.1638315789473684,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00047389291986653983,
+      "loss": 3.2046,
+      "step": 3891
+    },
+    {
+      "epoch": 0.1638736842105263,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00047387790441361685,
+      "loss": 2.9892,
+      "step": 3892
+    },
+    {
+      "epoch": 0.1639157894736842,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004738628848819015,
+      "loss": 3.5071,
+      "step": 3893
+    },
+    {
+      "epoch": 0.1639578947368421,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004738478612716674,
+      "loss": 3.605,
+      "step": 3894
+    },
+    {
+      "epoch": 0.164,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004738328335831883,
+      "loss": 3.847,
+      "step": 3895
+    },
+    {
+      "epoch": 0.1640421052631579,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00047381780181673797,
+      "loss": 3.5521,
+      "step": 3896
+    },
+    {
+      "epoch": 0.16408421052631578,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004738027659725902,
+      "loss": 3.8377,
+      "step": 3897
+    },
+    {
+      "epoch": 0.16412631578947368,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00047378772605101904,
+      "loss": 3.6925,
+      "step": 3898
+    },
+    {
+      "epoch": 0.16416842105263157,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0004737726820522985,
+      "loss": 3.4188,
+      "step": 3899
+    },
+    {
+      "epoch": 0.16421052631578947,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00047375763397670257,
+      "loss": 3.0865,
+      "step": 3900
+    },
+    {
+      "epoch": 0.16425263157894737,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004737425818245055,
+      "loss": 3.2189,
+      "step": 3901
+    },
+    {
+      "epoch": 0.16429473684210527,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00047372752559598136,
+      "loss": 4.0025,
+      "step": 3902
+    },
+    {
+      "epoch": 0.16433684210526317,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004737124652914047,
+      "loss": 3.2244,
+      "step": 3903
+    },
+    {
+      "epoch": 0.16437894736842104,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00047369740091104967,
+      "loss": 3.7539,
+      "step": 3904
+    },
+    {
+      "epoch": 0.16442105263157894,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004736823324551909,
+      "loss": 3.6803,
+      "step": 3905
+    },
+    {
+      "epoch": 0.16446315789473684,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004736672599241028,
+      "loss": 2.9834,
+      "step": 3906
+    },
+    {
+      "epoch": 0.16450526315789474,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00047365218331806003,
+      "loss": 3.1287,
+      "step": 3907
+    },
+    {
+      "epoch": 0.16454736842105264,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004736371026373373,
+      "loss": 2.9637,
+      "step": 3908
+    },
+    {
+      "epoch": 0.16458947368421054,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00047362201788220926,
+      "loss": 3.3041,
+      "step": 3909
+    },
+    {
+      "epoch": 0.16463157894736843,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004736069290529509,
+      "loss": 3.5508,
+      "step": 3910
+    },
+    {
+      "epoch": 0.1646736842105263,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004735918361498369,
+      "loss": 3.7763,
+      "step": 3911
+    },
+    {
+      "epoch": 0.1647157894736842,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004735767391731424,
+      "loss": 3.4247,
+      "step": 3912
+    },
+    {
+      "epoch": 0.1647578947368421,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00047356163812314244,
+      "loss": 2.9776,
+      "step": 3913
+    },
+    {
+      "epoch": 0.1648,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004735465330001121,
+      "loss": 3.7389,
+      "step": 3914
+    },
+    {
+      "epoch": 0.1648421052631579,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00047353142380432656,
+      "loss": 3.4266,
+      "step": 3915
+    },
+    {
+      "epoch": 0.1648842105263158,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00047351631053606116,
+      "loss": 3.3735,
+      "step": 3916
+    },
+    {
+      "epoch": 0.16492631578947367,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004735011931955912,
+      "loss": 3.7591,
+      "step": 3917
+    },
+    {
+      "epoch": 0.16496842105263157,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00047348607178319203,
+      "loss": 2.6511,
+      "step": 3918
+    },
+    {
+      "epoch": 0.16501052631578947,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00047347094629913923,
+      "loss": 3.7593,
+      "step": 3919
+    },
+    {
+      "epoch": 0.16505263157894737,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00047345581674370843,
+      "loss": 3.5246,
+      "step": 3920
+    },
+    {
+      "epoch": 0.16509473684210527,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004734406831171751,
+      "loss": 3.4634,
+      "step": 3921
+    },
+    {
+      "epoch": 0.16513684210526317,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00047342554541981514,
+      "loss": 3.5734,
+      "step": 3922
+    },
+    {
+      "epoch": 0.16517894736842106,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00047341040365190423,
+      "loss": 3.3072,
+      "step": 3923
+    },
+    {
+      "epoch": 0.16522105263157894,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004733952578137183,
+      "loss": 3.6486,
+      "step": 3924
+    },
+    {
+      "epoch": 0.16526315789473683,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0004733801079055332,
+      "loss": 3.0872,
+      "step": 3925
+    },
+    {
+      "epoch": 0.16530526315789473,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.000473364953927625,
+      "loss": 3.6241,
+      "step": 3926
+    },
+    {
+      "epoch": 0.16534736842105263,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004733497958802697,
+      "loss": 3.3633,
+      "step": 3927
+    },
+    {
+      "epoch": 0.16538947368421053,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0004733346337637437,
+      "loss": 3.5949,
+      "step": 3928
+    },
+    {
+      "epoch": 0.16543157894736843,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.000473319467578323,
+      "loss": 3.2528,
+      "step": 3929
+    },
+    {
+      "epoch": 0.16547368421052633,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00047330429732428403,
+      "loss": 3.6965,
+      "step": 3930
+    },
+    {
+      "epoch": 0.1655157894736842,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004732891230019031,
+      "loss": 3.3838,
+      "step": 3931
+    },
+    {
+      "epoch": 0.1655578947368421,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004732739446114567,
+      "loss": 3.3009,
+      "step": 3932
+    },
+    {
+      "epoch": 0.1656,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004732587621532214,
+      "loss": 3.4809,
+      "step": 3933
+    },
+    {
+      "epoch": 0.1656421052631579,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00047324357562747373,
+      "loss": 3.422,
+      "step": 3934
+    },
+    {
+      "epoch": 0.1656842105263158,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004732283850344904,
+      "loss": 3.6986,
+      "step": 3935
+    },
+    {
+      "epoch": 0.1657263157894737,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00047321319037454826,
+      "loss": 3.7558,
+      "step": 3936
+    },
+    {
+      "epoch": 0.1657684210526316,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.000473197991647924,
+      "loss": 3.4695,
+      "step": 3937
+    },
+    {
+      "epoch": 0.16581052631578946,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00047318278885489454,
+      "loss": 3.7542,
+      "step": 3938
+    },
+    {
+      "epoch": 0.16585263157894736,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00047316758199573695,
+      "loss": 3.1024,
+      "step": 3939
+    },
+    {
+      "epoch": 0.16589473684210526,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004731523710707282,
+      "loss": 3.3538,
+      "step": 3940
+    },
+    {
+      "epoch": 0.16593684210526316,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00047313715608014544,
+      "loss": 3.2272,
+      "step": 3941
+    },
+    {
+      "epoch": 0.16597894736842106,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00047312193702426586,
+      "loss": 3.2253,
+      "step": 3942
+    },
+    {
+      "epoch": 0.16602105263157896,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004731067139033668,
+      "loss": 3.8651,
+      "step": 3943
+    },
+    {
+      "epoch": 0.16606315789473683,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004730914867177255,
+      "loss": 3.5505,
+      "step": 3944
+    },
+    {
+      "epoch": 0.16610526315789473,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004730762554676195,
+      "loss": 3.5339,
+      "step": 3945
+    },
+    {
+      "epoch": 0.16614736842105263,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0004730610201533262,
+      "loss": 3.5887,
+      "step": 3946
+    },
+    {
+      "epoch": 0.16618947368421053,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00047304578077512317,
+      "loss": 3.4064,
+      "step": 3947
+    },
+    {
+      "epoch": 0.16623157894736842,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004730305373332881,
+      "loss": 3.2213,
+      "step": 3948
+    },
+    {
+      "epoch": 0.16627368421052632,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004730152898280987,
+      "loss": 3.8282,
+      "step": 3949
+    },
+    {
+      "epoch": 0.16631578947368422,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00047300003825983275,
+      "loss": 3.599,
+      "step": 3950
+    },
+    {
+      "epoch": 0.1663578947368421,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004729847826287681,
+      "loss": 3.0654,
+      "step": 3951
+    },
+    {
+      "epoch": 0.1664,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004729695229351827,
+      "loss": 3.5216,
+      "step": 3952
+    },
+    {
+      "epoch": 0.1664421052631579,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00047295425917935456,
+      "loss": 3.3799,
+      "step": 3953
+    },
+    {
+      "epoch": 0.1664842105263158,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0004729389913615618,
+      "loss": 3.5818,
+      "step": 3954
+    },
+    {
+      "epoch": 0.1665263157894737,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00047292371948208257,
+      "loss": 2.9763,
+      "step": 3955
+    },
+    {
+      "epoch": 0.1665684210526316,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004729084435411951,
+      "loss": 3.4696,
+      "step": 3956
+    },
+    {
+      "epoch": 0.1666105263157895,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00047289316353917764,
+      "loss": 3.1098,
+      "step": 3957
+    },
+    {
+      "epoch": 0.16665263157894736,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00047287787947630875,
+      "loss": 3.5092,
+      "step": 3958
+    },
+    {
+      "epoch": 0.16669473684210526,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004728625913528667,
+      "loss": 3.5431,
+      "step": 3959
+    },
+    {
+      "epoch": 0.16673684210526316,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00047284729916913004,
+      "loss": 3.5243,
+      "step": 3960
+    },
+    {
+      "epoch": 0.16677894736842105,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00047283200292537746,
+      "loss": 3.5311,
+      "step": 3961
+    },
+    {
+      "epoch": 0.16682105263157895,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00047281670262188757,
+      "loss": 3.3587,
+      "step": 3962
+    },
+    {
+      "epoch": 0.16686315789473685,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004728013982589392,
+      "loss": 3.3555,
+      "step": 3963
+    },
+    {
+      "epoch": 0.16690526315789472,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00047278608983681116,
+      "loss": 3.6365,
+      "step": 3964
+    },
+    {
+      "epoch": 0.16694736842105262,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00047277077735578235,
+      "loss": 3.847,
+      "step": 3965
+    },
+    {
+      "epoch": 0.16698947368421052,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0004727554608161316,
+      "loss": 3.6163,
+      "step": 3966
+    },
+    {
+      "epoch": 0.16703157894736842,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004727401402181382,
+      "loss": 3.5799,
+      "step": 3967
+    },
+    {
+      "epoch": 0.16707368421052632,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.000472724815562081,
+      "loss": 2.9099,
+      "step": 3968
+    },
+    {
+      "epoch": 0.16711578947368422,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004727094868482395,
+      "loss": 3.2216,
+      "step": 3969
+    },
+    {
+      "epoch": 0.16715789473684212,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00047269415407689275,
+      "loss": 3.39,
+      "step": 3970
+    },
+    {
+      "epoch": 0.1672,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00047267881724832015,
+      "loss": 3.1619,
+      "step": 3971
+    },
+    {
+      "epoch": 0.1672421052631579,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00047266347636280127,
+      "loss": 2.9004,
+      "step": 3972
+    },
+    {
+      "epoch": 0.16728421052631579,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004726481314206154,
+      "loss": 3.7035,
+      "step": 3973
+    },
+    {
+      "epoch": 0.16732631578947368,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004726327824220421,
+      "loss": 3.3023,
+      "step": 3974
+    },
+    {
+      "epoch": 0.16736842105263158,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004726174293673612,
+      "loss": 3.5242,
+      "step": 3975
+    },
+    {
+      "epoch": 0.16741052631578948,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00047260207225685217,
+      "loss": 3.2183,
+      "step": 3976
+    },
+    {
+      "epoch": 0.16745263157894738,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00047258671109079496,
+      "loss": 3.1849,
+      "step": 3977
+    },
+    {
+      "epoch": 0.16749473684210525,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004725713458694695,
+      "loss": 3.3391,
+      "step": 3978
+    },
+    {
+      "epoch": 0.16753684210526315,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00047255597659315556,
+      "loss": 3.3541,
+      "step": 3979
+    },
+    {
+      "epoch": 0.16757894736842105,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004725406032621332,
+      "loss": 2.9318,
+      "step": 3980
+    },
+    {
+      "epoch": 0.16762105263157895,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0004725252258766826,
+      "loss": 3.4536,
+      "step": 3981
+    },
+    {
+      "epoch": 0.16766315789473685,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00047250984443708376,
+      "loss": 3.83,
+      "step": 3982
+    },
+    {
+      "epoch": 0.16770526315789475,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004724944589436171,
+      "loss": 3.4686,
+      "step": 3983
+    },
+    {
+      "epoch": 0.16774736842105265,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004724790693965627,
+      "loss": 3.4259,
+      "step": 3984
+    },
+    {
+      "epoch": 0.16778947368421052,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00047246367579620107,
+      "loss": 3.8927,
+      "step": 3985
+    },
+    {
+      "epoch": 0.16783157894736841,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00047244827814281275,
+      "loss": 3.5099,
+      "step": 3986
+    },
+    {
+      "epoch": 0.1678736842105263,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0004724328764366781,
+      "loss": 3.8294,
+      "step": 3987
+    },
+    {
+      "epoch": 0.1679157894736842,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00047241747067807777,
+      "loss": 3.027,
+      "step": 3988
+    },
+    {
+      "epoch": 0.1679578947368421,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00047240206086729256,
+      "loss": 3.297,
+      "step": 3989
+    },
+    {
+      "epoch": 0.168,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.000472386647004603,
+      "loss": 3.7308,
+      "step": 3990
+    },
+    {
+      "epoch": 0.16804210526315788,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00047237122909029,
+      "loss": 3.5469,
+      "step": 3991
+    },
+    {
+      "epoch": 0.16808421052631578,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00047235580712463453,
+      "loss": 3.0115,
+      "step": 3992
+    },
+    {
+      "epoch": 0.16812631578947368,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00047234038110791754,
+      "loss": 3.5591,
+      "step": 3993
+    },
+    {
+      "epoch": 0.16816842105263158,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00047232495104042007,
+      "loss": 3.2605,
+      "step": 3994
+    },
+    {
+      "epoch": 0.16821052631578948,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004723095169224231,
+      "loss": 3.4928,
+      "step": 3995
+    },
+    {
+      "epoch": 0.16825263157894738,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00047229407875420807,
+      "loss": 3.5546,
+      "step": 3996
+    },
+    {
+      "epoch": 0.16829473684210527,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.000472278636536056,
+      "loss": 3.3614,
+      "step": 3997
+    },
+    {
+      "epoch": 0.16833684210526315,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00047226319026824837,
+      "loss": 3.9368,
+      "step": 3998
+    },
+    {
+      "epoch": 0.16837894736842104,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00047224773995106664,
+      "loss": 2.9254,
+      "step": 3999
+    },
+    {
+      "epoch": 0.16842105263157894,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004722322855847921,
+      "loss": 3.2352,
+      "step": 4000
+    },
+    {
+      "epoch": 0.16846315789473684,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.0004722168271697065,
+      "loss": 3.6208,
+      "step": 4001
+    },
+    {
+      "epoch": 0.16850526315789474,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00047220136470609134,
+      "loss": 3.4051,
+      "step": 4002
+    },
+    {
+      "epoch": 0.16854736842105264,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00047218589819422843,
+      "loss": 3.5408,
+      "step": 4003
+    },
+    {
+      "epoch": 0.16858947368421054,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00047217042763439957,
+      "loss": 3.2939,
+      "step": 4004
+    },
+    {
+      "epoch": 0.1686315789473684,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004721549530268865,
+      "loss": 3.4704,
+      "step": 4005
+    },
+    {
+      "epoch": 0.1686736842105263,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0004721394743719711,
+      "loss": 3.4414,
+      "step": 4006
+    },
+    {
+      "epoch": 0.1687157894736842,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004721239916699357,
+      "loss": 3.266,
+      "step": 4007
+    },
+    {
+      "epoch": 0.1687578947368421,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.000472108504921062,
+      "loss": 3.3826,
+      "step": 4008
+    },
+    {
+      "epoch": 0.1688,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004720930141256324,
+      "loss": 3.8842,
+      "step": 4009
+    },
+    {
+      "epoch": 0.1688421052631579,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00047207751928392895,
+      "loss": 3.4885,
+      "step": 4010
+    },
+    {
+      "epoch": 0.1688842105263158,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00047206202039623405,
+      "loss": 3.577,
+      "step": 4011
+    },
+    {
+      "epoch": 0.16892631578947367,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004720465174628301,
+      "loss": 3.3415,
+      "step": 4012
+    },
+    {
+      "epoch": 0.16896842105263157,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00047203101048399945,
+      "loss": 3.3593,
+      "step": 4013
+    },
+    {
+      "epoch": 0.16901052631578947,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00047201549946002466,
+      "loss": 3.2469,
+      "step": 4014
+    },
+    {
+      "epoch": 0.16905263157894737,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0004719999843911884,
+      "loss": 3.6397,
+      "step": 4015
+    },
+    {
+      "epoch": 0.16909473684210527,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004719844652777732,
+      "loss": 2.9888,
+      "step": 4016
+    },
+    {
+      "epoch": 0.16913684210526317,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00047196894212006183,
+      "loss": 3.0445,
+      "step": 4017
+    },
+    {
+      "epoch": 0.16917894736842104,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004719534149183372,
+      "loss": 3.3861,
+      "step": 4018
+    },
+    {
+      "epoch": 0.16922105263157894,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004719378836728821,
+      "loss": 3.5823,
+      "step": 4019
+    },
+    {
+      "epoch": 0.16926315789473684,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00047192234838397954,
+      "loss": 3.2576,
+      "step": 4020
+    },
+    {
+      "epoch": 0.16930526315789474,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004719068090519125,
+      "loss": 3.6116,
+      "step": 4021
+    },
+    {
+      "epoch": 0.16934736842105264,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00047189126567696415,
+      "loss": 3.3864,
+      "step": 4022
+    },
+    {
+      "epoch": 0.16938947368421053,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004718757182594177,
+      "loss": 3.7278,
+      "step": 4023
+    },
+    {
+      "epoch": 0.16943157894736843,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00047186016679955637,
+      "loss": 3.1557,
+      "step": 4024
+    },
+    {
+      "epoch": 0.1694736842105263,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00047184461129766344,
+      "loss": 3.4391,
+      "step": 4025
+    },
+    {
+      "epoch": 0.1695157894736842,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004718290517540223,
+      "loss": 3.3408,
+      "step": 4026
+    },
+    {
+      "epoch": 0.1695578947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004718134881689165,
+      "loss": 3.5917,
+      "step": 4027
+    },
+    {
+      "epoch": 0.1696,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004717979205426296,
+      "loss": 3.1851,
+      "step": 4028
+    },
+    {
+      "epoch": 0.1696421052631579,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00047178234887544515,
+      "loss": 3.6029,
+      "step": 4029
+    },
+    {
+      "epoch": 0.1696842105263158,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004717667731676469,
+      "loss": 3.7685,
+      "step": 4030
+    },
+    {
+      "epoch": 0.1697263157894737,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00047175119341951865,
+      "loss": 3.5113,
+      "step": 4031
+    },
+    {
+      "epoch": 0.16976842105263157,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00047173560963134414,
+      "loss": 3.6431,
+      "step": 4032
+    },
+    {
+      "epoch": 0.16981052631578947,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004717200218034074,
+      "loss": 3.5522,
+      "step": 4033
+    },
+    {
+      "epoch": 0.16985263157894737,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00047170442993599233,
+      "loss": 3.8096,
+      "step": 4034
+    },
+    {
+      "epoch": 0.16989473684210527,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0004716888340293831,
+      "loss": 3.5516,
+      "step": 4035
+    },
+    {
+      "epoch": 0.16993684210526316,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00047167323408386377,
+      "loss": 3.4304,
+      "step": 4036
+    },
+    {
+      "epoch": 0.16997894736842106,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004716576300997185,
+      "loss": 2.9175,
+      "step": 4037
+    },
+    {
+      "epoch": 0.17002105263157893,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00047164202207723173,
+      "loss": 3.6158,
+      "step": 4038
+    },
+    {
+      "epoch": 0.17006315789473683,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004716264100166877,
+      "loss": 3.4811,
+      "step": 4039
+    },
+    {
+      "epoch": 0.17010526315789473,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004716107939183708,
+      "loss": 3.5036,
+      "step": 4040
+    },
+    {
+      "epoch": 0.17014736842105263,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00047159517378256576,
+      "loss": 3.512,
+      "step": 4041
+    },
+    {
+      "epoch": 0.17018947368421053,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00047157954960955694,
+      "loss": 2.9685,
+      "step": 4042
+    },
+    {
+      "epoch": 0.17023157894736843,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004715639213996291,
+      "loss": 3.396,
+      "step": 4043
+    },
+    {
+      "epoch": 0.17027368421052633,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004715482891530669,
+      "loss": 3.2873,
+      "step": 4044
+    },
+    {
+      "epoch": 0.1703157894736842,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0004715326528701552,
+      "loss": 3.4103,
+      "step": 4045
+    },
+    {
+      "epoch": 0.1703578947368421,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00047151701255117887,
+      "loss": 3.4841,
+      "step": 4046
+    },
+    {
+      "epoch": 0.1704,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0004715013681964229,
+      "loss": 3.4816,
+      "step": 4047
+    },
+    {
+      "epoch": 0.1704421052631579,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004714857198061722,
+      "loss": 3.6729,
+      "step": 4048
+    },
+    {
+      "epoch": 0.1704842105263158,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00047147006738071183,
+      "loss": 3.3521,
+      "step": 4049
+    },
+    {
+      "epoch": 0.1705263157894737,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004714544109203272,
+      "loss": 3.5727,
+      "step": 4050
+    },
+    {
+      "epoch": 0.1705684210526316,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00047143875042530325,
+      "loss": 3.8321,
+      "step": 4051
+    },
+    {
+      "epoch": 0.17061052631578946,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004714230858959255,
+      "loss": 3.649,
+      "step": 4052
+    },
+    {
+      "epoch": 0.17065263157894736,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00047140741733247936,
+      "loss": 3.1874,
+      "step": 4053
+    },
+    {
+      "epoch": 0.17069473684210526,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004713917447352501,
+      "loss": 3.4428,
+      "step": 4054
+    },
+    {
+      "epoch": 0.17073684210526316,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004713760681045235,
+      "loss": 3.3829,
+      "step": 4055
+    },
+    {
+      "epoch": 0.17077894736842106,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.000471360387440585,
+      "loss": 3.4407,
+      "step": 4056
+    },
+    {
+      "epoch": 0.17082105263157896,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004713447027437203,
+      "loss": 3.1996,
+      "step": 4057
+    },
+    {
+      "epoch": 0.17086315789473686,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00047132901401421523,
+      "loss": 2.883,
+      "step": 4058
+    },
+    {
+      "epoch": 0.17090526315789473,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004713133212523556,
+      "loss": 3.6038,
+      "step": 4059
+    },
+    {
+      "epoch": 0.17094736842105263,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00047129762445842716,
+      "loss": 3.7883,
+      "step": 4060
+    },
+    {
+      "epoch": 0.17098947368421052,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00047128192363271614,
+      "loss": 3.4081,
+      "step": 4061
+    },
+    {
+      "epoch": 0.17103157894736842,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004712662187755085,
+      "loss": 3.2596,
+      "step": 4062
+    },
+    {
+      "epoch": 0.17107368421052632,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00047125050988709017,
+      "loss": 3.6669,
+      "step": 4063
+    },
+    {
+      "epoch": 0.17111578947368422,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.0004712347969677476,
+      "loss": 3.5728,
+      "step": 4064
+    },
+    {
+      "epoch": 0.1711578947368421,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.000471219080017767,
+      "loss": 3.3157,
+      "step": 4065
+    },
+    {
+      "epoch": 0.1712,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004712033590374346,
+      "loss": 3.0154,
+      "step": 4066
+    },
+    {
+      "epoch": 0.1712421052631579,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00047118763402703695,
+      "loss": 3.3101,
+      "step": 4067
+    },
+    {
+      "epoch": 0.1712842105263158,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004711719049868605,
+      "loss": 3.4353,
+      "step": 4068
+    },
+    {
+      "epoch": 0.1713263157894737,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004711561719171918,
+      "loss": 3.1108,
+      "step": 4069
+    },
+    {
+      "epoch": 0.1713684210526316,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00047114043481831747,
+      "loss": 3.4348,
+      "step": 4070
+    },
+    {
+      "epoch": 0.17141052631578949,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004711246936905243,
+      "loss": 3.1932,
+      "step": 4071
+    },
+    {
+      "epoch": 0.17145263157894736,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.000471108948534099,
+      "loss": 3.4375,
+      "step": 4072
+    },
+    {
+      "epoch": 0.17149473684210526,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.0004710931993493285,
+      "loss": 3.4676,
+      "step": 4073
+    },
+    {
+      "epoch": 0.17153684210526315,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00047107744613649957,
+      "loss": 3.8069,
+      "step": 4074
+    },
+    {
+      "epoch": 0.17157894736842105,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004710616888958994,
+      "loss": 3.628,
+      "step": 4075
+    },
+    {
+      "epoch": 0.17162105263157895,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00047104592762781496,
+      "loss": 3.5614,
+      "step": 4076
+    },
+    {
+      "epoch": 0.17166315789473685,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0004710301623325335,
+      "loss": 3.4875,
+      "step": 4077
+    },
+    {
+      "epoch": 0.17170526315789475,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004710143930103421,
+      "loss": 2.96,
+      "step": 4078
+    },
+    {
+      "epoch": 0.17174736842105262,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.0004709986196615281,
+      "loss": 3.5211,
+      "step": 4079
+    },
+    {
+      "epoch": 0.17178947368421052,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00047098284228637904,
+      "loss": 3.5459,
+      "step": 4080
+    },
+    {
+      "epoch": 0.17183157894736842,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004709670608851822,
+      "loss": 3.6898,
+      "step": 4081
+    },
+    {
+      "epoch": 0.17187368421052632,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00047095127545822513,
+      "loss": 3.6728,
+      "step": 4082
+    },
+    {
+      "epoch": 0.17191578947368422,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004709354860057954,
+      "loss": 3.3401,
+      "step": 4083
+    },
+    {
+      "epoch": 0.17195789473684212,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0004709196925281807,
+      "loss": 3.3327,
+      "step": 4084
+    },
+    {
+      "epoch": 0.172,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004709038950256688,
+      "loss": 2.6864,
+      "step": 4085
+    },
+    {
+      "epoch": 0.17204210526315789,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00047088809349854754,
+      "loss": 3.6756,
+      "step": 4086
+    },
+    {
+      "epoch": 0.17208421052631578,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004708722879471047,
+      "loss": 3.7004,
+      "step": 4087
+    },
+    {
+      "epoch": 0.17212631578947368,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00047085647837162825,
+      "loss": 3.4447,
+      "step": 4088
+    },
+    {
+      "epoch": 0.17216842105263158,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00047084066477240627,
+      "loss": 3.1243,
+      "step": 4089
+    },
+    {
+      "epoch": 0.17221052631578948,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00047082484714972684,
+      "loss": 3.8072,
+      "step": 4090
+    },
+    {
+      "epoch": 0.17225263157894738,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00047080902550387824,
+      "loss": 3.5478,
+      "step": 4091
+    },
+    {
+      "epoch": 0.17229473684210525,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00047079319983514854,
+      "loss": 3.5313,
+      "step": 4092
+    },
+    {
+      "epoch": 0.17233684210526315,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00047077737014382626,
+      "loss": 3.3409,
+      "step": 4093
+    },
+    {
+      "epoch": 0.17237894736842105,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00047076153643019956,
+      "loss": 3.6879,
+      "step": 4094
+    },
+    {
+      "epoch": 0.17242105263157895,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004707456986945572,
+      "loss": 3.5188,
+      "step": 4095
+    },
+    {
+      "epoch": 0.17246315789473685,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004707298569371875,
+      "loss": 3.3724,
+      "step": 4096
+    },
+    {
+      "epoch": 0.17250526315789474,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00047071401115837913,
+      "loss": 2.9252,
+      "step": 4097
+    },
+    {
+      "epoch": 0.17254736842105264,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00047069816135842085,
+      "loss": 3.7512,
+      "step": 4098
+    },
+    {
+      "epoch": 0.17258947368421051,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00047068230753760133,
+      "loss": 3.769,
+      "step": 4099
+    },
+    {
+      "epoch": 0.1726315789473684,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00047066644969620953,
+      "loss": 3.0994,
+      "step": 4100
+    },
+    {
+      "epoch": 0.1726736842105263,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004706505878345342,
+      "loss": 3.5695,
+      "step": 4101
+    },
+    {
+      "epoch": 0.1727157894736842,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004706347219528645,
+      "loss": 3.497,
+      "step": 4102
+    },
+    {
+      "epoch": 0.1727578947368421,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004706188520514894,
+      "loss": 3.8816,
+      "step": 4103
+    },
+    {
+      "epoch": 0.1728,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00047060297813069794,
+      "loss": 3.4126,
+      "step": 4104
+    },
+    {
+      "epoch": 0.1728421052631579,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0004705871001907795,
+      "loss": 2.8575,
+      "step": 4105
+    },
+    {
+      "epoch": 0.17288421052631578,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00047057121823202325,
+      "loss": 3.4735,
+      "step": 4106
+    },
+    {
+      "epoch": 0.17292631578947368,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0004705553322547186,
+      "loss": 3.662,
+      "step": 4107
+    },
+    {
+      "epoch": 0.17296842105263158,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0004705394422591548,
+      "loss": 3.2555,
+      "step": 4108
+    },
+    {
+      "epoch": 0.17301052631578948,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004705235482456217,
+      "loss": 2.9976,
+      "step": 4109
+    },
+    {
+      "epoch": 0.17305263157894737,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004705076502144085,
+      "loss": 3.5981,
+      "step": 4110
+    },
+    {
+      "epoch": 0.17309473684210527,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00047049174816580504,
+      "loss": 3.4769,
+      "step": 4111
+    },
+    {
+      "epoch": 0.17313684210526314,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00047047584210010095,
+      "loss": 3.7224,
+      "step": 4112
+    },
+    {
+      "epoch": 0.17317894736842104,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00047045993201758616,
+      "loss": 3.6483,
+      "step": 4113
+    },
+    {
+      "epoch": 0.17322105263157894,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004704440179185504,
+      "loss": 3.2328,
+      "step": 4114
+    },
+    {
+      "epoch": 0.17326315789473684,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004704280998032836,
+      "loss": 3.5565,
+      "step": 4115
+    },
+    {
+      "epoch": 0.17330526315789474,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004704121776720759,
+      "loss": 3.5066,
+      "step": 4116
+    },
+    {
+      "epoch": 0.17334736842105264,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004703962515252172,
+      "loss": 3.5862,
+      "step": 4117
+    },
+    {
+      "epoch": 0.17338947368421054,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004703803213629978,
+      "loss": 3.3561,
+      "step": 4118
+    },
+    {
+      "epoch": 0.1734315789473684,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00047036438718570784,
+      "loss": 3.28,
+      "step": 4119
+    },
+    {
+      "epoch": 0.1734736842105263,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004703484489936377,
+      "loss": 3.61,
+      "step": 4120
+    },
+    {
+      "epoch": 0.1735157894736842,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00047033250678707775,
+      "loss": 3.604,
+      "step": 4121
+    },
+    {
+      "epoch": 0.1735578947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00047031656056631833,
+      "loss": 3.5327,
+      "step": 4122
+    },
+    {
+      "epoch": 0.1736,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004703006103316501,
+      "loss": 3.7875,
+      "step": 4123
+    },
+    {
+      "epoch": 0.1736421052631579,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0004702846560833636,
+      "loss": 2.7759,
+      "step": 4124
+    },
+    {
+      "epoch": 0.1736842105263158,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00047026869782174947,
+      "loss": 3.1796,
+      "step": 4125
+    },
+    {
+      "epoch": 0.17372631578947367,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004702527355470985,
+      "loss": 3.5315,
+      "step": 4126
+    },
+    {
+      "epoch": 0.17376842105263157,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00047023676925970136,
+      "loss": 3.8077,
+      "step": 4127
+    },
+    {
+      "epoch": 0.17381052631578947,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00047022079895984916,
+      "loss": 3.5666,
+      "step": 4128
+    },
+    {
+      "epoch": 0.17385263157894737,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00047020482464783277,
+      "loss": 3.3682,
+      "step": 4129
+    },
+    {
+      "epoch": 0.17389473684210527,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004701888463239431,
+      "loss": 3.7405,
+      "step": 4130
+    },
+    {
+      "epoch": 0.17393684210526317,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004701728639884714,
+      "loss": 3.5147,
+      "step": 4131
+    },
+    {
+      "epoch": 0.17397894736842107,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004701568776417089,
+      "loss": 3.1774,
+      "step": 4132
+    },
+    {
+      "epoch": 0.17402105263157894,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00047014088728394664,
+      "loss": 3.3915,
+      "step": 4133
+    },
+    {
+      "epoch": 0.17406315789473684,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00047012489291547614,
+      "loss": 3.7362,
+      "step": 4134
+    },
+    {
+      "epoch": 0.17410526315789474,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004701088945365887,
+      "loss": 3.1461,
+      "step": 4135
+    },
+    {
+      "epoch": 0.17414736842105263,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00047009289214757584,
+      "loss": 3.4335,
+      "step": 4136
+    },
+    {
+      "epoch": 0.17418947368421053,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00047007688574872907,
+      "loss": 3.2442,
+      "step": 4137
+    },
+    {
+      "epoch": 0.17423157894736843,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00047006087534034004,
+      "loss": 3.362,
+      "step": 4138
+    },
+    {
+      "epoch": 0.1742736842105263,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004700448609227004,
+      "loss": 3.0886,
+      "step": 4139
+    },
+    {
+      "epoch": 0.1743157894736842,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004700288424961019,
+      "loss": 3.3822,
+      "step": 4140
+    },
+    {
+      "epoch": 0.1743578947368421,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00047001282006083647,
+      "loss": 2.7435,
+      "step": 4141
+    },
+    {
+      "epoch": 0.1744,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0004699967936171959,
+      "loss": 3.337,
+      "step": 4142
+    },
+    {
+      "epoch": 0.1744421052631579,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00046998076316547233,
+      "loss": 3.075,
+      "step": 4143
+    },
+    {
+      "epoch": 0.1744842105263158,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004699647287059577,
+      "loss": 3.0559,
+      "step": 4144
+    },
+    {
+      "epoch": 0.1745263157894737,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004699486902389441,
+      "loss": 3.8189,
+      "step": 4145
+    },
+    {
+      "epoch": 0.17456842105263157,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004699326477647239,
+      "loss": 3.336,
+      "step": 4146
+    },
+    {
+      "epoch": 0.17461052631578947,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004699166012835891,
+      "loss": 3.6742,
+      "step": 4147
+    },
+    {
+      "epoch": 0.17465263157894736,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004699005507958324,
+      "loss": 3.2918,
+      "step": 4148
+    },
+    {
+      "epoch": 0.17469473684210526,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004698844963017459,
+      "loss": 3.2911,
+      "step": 4149
+    },
+    {
+      "epoch": 0.17473684210526316,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00046986843780162223,
+      "loss": 2.9933,
+      "step": 4150
+    },
+    {
+      "epoch": 0.17477894736842106,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004698523752957541,
+      "loss": 3.4779,
+      "step": 4151
+    },
+    {
+      "epoch": 0.17482105263157896,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00046983630878443385,
+      "loss": 2.9505,
+      "step": 4152
+    },
+    {
+      "epoch": 0.17486315789473683,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00046982023826795444,
+      "loss": 3.1004,
+      "step": 4153
+    },
+    {
+      "epoch": 0.17490526315789473,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004698041637466085,
+      "loss": 3.6035,
+      "step": 4154
+    },
+    {
+      "epoch": 0.17494736842105263,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004697880852206891,
+      "loss": 3.3295,
+      "step": 4155
+    },
+    {
+      "epoch": 0.17498947368421053,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00046977200269048886,
+      "loss": 3.3053,
+      "step": 4156
+    },
+    {
+      "epoch": 0.17503157894736843,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00046975591615630103,
+      "loss": 3.6766,
+      "step": 4157
+    },
+    {
+      "epoch": 0.17507368421052633,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004697398256184186,
+      "loss": 3.2469,
+      "step": 4158
+    },
+    {
+      "epoch": 0.1751157894736842,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0004697237310771347,
+      "loss": 3.5089,
+      "step": 4159
+    },
+    {
+      "epoch": 0.1751578947368421,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00046970763253274264,
+      "loss": 3.1863,
+      "step": 4160
+    },
+    {
+      "epoch": 0.1752,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00046969152998553565,
+      "loss": 3.7891,
+      "step": 4161
+    },
+    {
+      "epoch": 0.1752421052631579,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00046967542343580705,
+      "loss": 3.0643,
+      "step": 4162
+    },
+    {
+      "epoch": 0.1752842105263158,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00046965931288385043,
+      "loss": 3.2654,
+      "step": 4163
+    },
+    {
+      "epoch": 0.1753263157894737,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004696431983299592,
+      "loss": 3.5448,
+      "step": 4164
+    },
+    {
+      "epoch": 0.1753684210526316,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00046962707977442694,
+      "loss": 3.5483,
+      "step": 4165
+    },
+    {
+      "epoch": 0.17541052631578946,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00046961095721754733,
+      "loss": 2.9364,
+      "step": 4166
+    },
+    {
+      "epoch": 0.17545263157894736,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004695948306596141,
+      "loss": 2.904,
+      "step": 4167
+    },
+    {
+      "epoch": 0.17549473684210526,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004695787001009211,
+      "loss": 3.6715,
+      "step": 4168
+    },
+    {
+      "epoch": 0.17553684210526316,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004695625655417622,
+      "loss": 3.2459,
+      "step": 4169
+    },
+    {
+      "epoch": 0.17557894736842106,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00046954642698243134,
+      "loss": 3.3887,
+      "step": 4170
+    },
+    {
+      "epoch": 0.17562105263157896,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00046953028442322245,
+      "loss": 3.2032,
+      "step": 4171
+    },
+    {
+      "epoch": 0.17566315789473685,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00046951413786442976,
+      "loss": 3.6078,
+      "step": 4172
+    },
+    {
+      "epoch": 0.17570526315789473,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004694979873063474,
+      "loss": 3.439,
+      "step": 4173
+    },
+    {
+      "epoch": 0.17574736842105262,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004694818327492696,
+      "loss": 3.3076,
+      "step": 4174
+    },
+    {
+      "epoch": 0.17578947368421052,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004694656741934907,
+      "loss": 3.4701,
+      "step": 4175
+    },
+    {
+      "epoch": 0.17583157894736842,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00046944951163930507,
+      "loss": 3.715,
+      "step": 4176
+    },
+    {
+      "epoch": 0.17587368421052632,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004694333450870072,
+      "loss": 3.3777,
+      "step": 4177
+    },
+    {
+      "epoch": 0.17591578947368422,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004694171745368916,
+      "loss": 3.5243,
+      "step": 4178
+    },
+    {
+      "epoch": 0.17595789473684212,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004694009999892529,
+      "loss": 3.5761,
+      "step": 4179
+    },
+    {
+      "epoch": 0.176,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00046938482144438576,
+      "loss": 3.5781,
+      "step": 4180
+    },
+    {
+      "epoch": 0.1760421052631579,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00046936863890258487,
+      "loss": 3.4956,
+      "step": 4181
+    },
+    {
+      "epoch": 0.1760842105263158,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004693524523641452,
+      "loss": 3.5015,
+      "step": 4182
+    },
+    {
+      "epoch": 0.1761263157894737,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00046933626182936157,
+      "loss": 3.8282,
+      "step": 4183
+    },
+    {
+      "epoch": 0.17616842105263159,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00046932006729852896,
+      "loss": 3.7916,
+      "step": 4184
+    },
+    {
+      "epoch": 0.17621052631578948,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004693038687719424,
+      "loss": 3.513,
+      "step": 4185
+    },
+    {
+      "epoch": 0.17625263157894736,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.000469287666249897,
+      "loss": 3.2565,
+      "step": 4186
+    },
+    {
+      "epoch": 0.17629473684210525,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.000469271459732688,
+      "loss": 3.0477,
+      "step": 4187
+    },
+    {
+      "epoch": 0.17633684210526315,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00046925524922061064,
+      "loss": 3.2048,
+      "step": 4188
+    },
+    {
+      "epoch": 0.17637894736842105,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00046923903471396026,
+      "loss": 3.5178,
+      "step": 4189
+    },
+    {
+      "epoch": 0.17642105263157895,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004692228162130322,
+      "loss": 3.6415,
+      "step": 4190
+    },
+    {
+      "epoch": 0.17646315789473685,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004692065937181221,
+      "loss": 3.3499,
+      "step": 4191
+    },
+    {
+      "epoch": 0.17650526315789475,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00046919036722952535,
+      "loss": 3.6412,
+      "step": 4192
+    },
+    {
+      "epoch": 0.17654736842105262,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00046917413674753765,
+      "loss": 3.2624,
+      "step": 4193
+    },
+    {
+      "epoch": 0.17658947368421052,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00046915790227245477,
+      "loss": 3.6382,
+      "step": 4194
+    },
+    {
+      "epoch": 0.17663157894736842,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00046914166380457236,
+      "loss": 3.1782,
+      "step": 4195
+    },
+    {
+      "epoch": 0.17667368421052632,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004691254213441863,
+      "loss": 3.3404,
+      "step": 4196
+    },
+    {
+      "epoch": 0.17671578947368421,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0004691091748915925,
+      "loss": 3.2054,
+      "step": 4197
+    },
+    {
+      "epoch": 0.1767578947368421,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004690929244470871,
+      "loss": 3.6556,
+      "step": 4198
+    },
+    {
+      "epoch": 0.1768,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004690766700109659,
+      "loss": 3.4804,
+      "step": 4199
+    },
+    {
+      "epoch": 0.17684210526315788,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004690604115835252,
+      "loss": 3.5481,
+      "step": 4200
+    },
+    {
+      "epoch": 0.17688421052631578,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0004690441491650613,
+      "loss": 3.4507,
+      "step": 4201
+    },
+    {
+      "epoch": 0.17692631578947368,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00046902788275587025,
+      "loss": 3.4143,
+      "step": 4202
+    },
+    {
+      "epoch": 0.17696842105263158,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004690116123562486,
+      "loss": 3.4218,
+      "step": 4203
+    },
+    {
+      "epoch": 0.17701052631578948,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0004689953379664926,
+      "loss": 3.5022,
+      "step": 4204
+    },
+    {
+      "epoch": 0.17705263157894738,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004689790595868989,
+      "loss": 3.6886,
+      "step": 4205
+    },
+    {
+      "epoch": 0.17709473684210528,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00046896277721776407,
+      "loss": 3.2538,
+      "step": 4206
+    },
+    {
+      "epoch": 0.17713684210526315,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004689464908593847,
+      "loss": 3.1598,
+      "step": 4207
+    },
+    {
+      "epoch": 0.17717894736842105,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004689302005120575,
+      "loss": 3.6936,
+      "step": 4208
+    },
+    {
+      "epoch": 0.17722105263157895,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00046891390617607927,
+      "loss": 3.2474,
+      "step": 4209
+    },
+    {
+      "epoch": 0.17726315789473684,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004688976078517468,
+      "loss": 3.3635,
+      "step": 4210
+    },
+    {
+      "epoch": 0.17730526315789474,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00046888130553935726,
+      "loss": 3.2507,
+      "step": 4211
+    },
+    {
+      "epoch": 0.17734736842105264,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0004688649992392074,
+      "loss": 3.348,
+      "step": 4212
+    },
+    {
+      "epoch": 0.1773894736842105,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004688486889515945,
+      "loss": 3.3024,
+      "step": 4213
+    },
+    {
+      "epoch": 0.1774315789473684,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004688323746768156,
+      "loss": 3.0525,
+      "step": 4214
+    },
+    {
+      "epoch": 0.1774736842105263,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004688160564151679,
+      "loss": 3.294,
+      "step": 4215
+    },
+    {
+      "epoch": 0.1775157894736842,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00046879973416694875,
+      "loss": 3.0432,
+      "step": 4216
+    },
+    {
+      "epoch": 0.1775578947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004687834079324556,
+      "loss": 3.237,
+      "step": 4217
+    },
+    {
+      "epoch": 0.1776,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004687670777119858,
+      "loss": 3.3216,
+      "step": 4218
+    },
+    {
+      "epoch": 0.1776421052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004687507435058368,
+      "loss": 3.2273,
+      "step": 4219
+    },
+    {
+      "epoch": 0.17768421052631578,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004687344053143063,
+      "loss": 3.4829,
+      "step": 4220
+    },
+    {
+      "epoch": 0.17772631578947368,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00046871806313769204,
+      "loss": 3.2,
+      "step": 4221
+    },
+    {
+      "epoch": 0.17776842105263158,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00046870171697629157,
+      "loss": 3.3576,
+      "step": 4222
+    },
+    {
+      "epoch": 0.17781052631578947,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00046868536683040274,
+      "loss": 3.3935,
+      "step": 4223
+    },
+    {
+      "epoch": 0.17785263157894737,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004686690127003236,
+      "loss": 3.3322,
+      "step": 4224
+    },
+    {
+      "epoch": 0.17789473684210527,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004686526545863519,
+      "loss": 3.6801,
+      "step": 4225
+    },
+    {
+      "epoch": 0.17793684210526317,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004686362924887857,
+      "loss": 3.4058,
+      "step": 4226
+    },
+    {
+      "epoch": 0.17797894736842104,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004686199264079232,
+      "loss": 3.0714,
+      "step": 4227
+    },
+    {
+      "epoch": 0.17802105263157894,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0004686035563440625,
+      "loss": 3.3951,
+      "step": 4228
+    },
+    {
+      "epoch": 0.17806315789473684,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00046858718229750177,
+      "loss": 3.2651,
+      "step": 4229
+    },
+    {
+      "epoch": 0.17810526315789474,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0004685708042685395,
+      "loss": 3.783,
+      "step": 4230
+    },
+    {
+      "epoch": 0.17814736842105264,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00046855442225747393,
+      "loss": 3.4605,
+      "step": 4231
+    },
+    {
+      "epoch": 0.17818947368421054,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00046853803626460363,
+      "loss": 3.2673,
+      "step": 4232
+    },
+    {
+      "epoch": 0.1782315789473684,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.000468521646290227,
+      "loss": 3.2684,
+      "step": 4233
+    },
+    {
+      "epoch": 0.1782736842105263,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004685052523346428,
+      "loss": 3.5541,
+      "step": 4234
+    },
+    {
+      "epoch": 0.1783157894736842,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00046848885439814965,
+      "loss": 3.7024,
+      "step": 4235
+    },
+    {
+      "epoch": 0.1783578947368421,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004684724524810462,
+      "loss": 3.5113,
+      "step": 4236
+    },
+    {
+      "epoch": 0.1784,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0004684560465836315,
+      "loss": 3.4779,
+      "step": 4237
+    },
+    {
+      "epoch": 0.1784421052631579,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004684396367062042,
+      "loss": 3.1146,
+      "step": 4238
+    },
+    {
+      "epoch": 0.1784842105263158,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00046842322284906335,
+      "loss": 3.4847,
+      "step": 4239
+    },
+    {
+      "epoch": 0.17852631578947367,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004684068050125081,
+      "loss": 2.9235,
+      "step": 4240
+    },
+    {
+      "epoch": 0.17856842105263157,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00046839038319683744,
+      "loss": 3.4952,
+      "step": 4241
+    },
+    {
+      "epoch": 0.17861052631578947,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00046837395740235055,
+      "loss": 3.2787,
+      "step": 4242
+    },
+    {
+      "epoch": 0.17865263157894737,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004683575276293468,
+      "loss": 3.611,
+      "step": 4243
+    },
+    {
+      "epoch": 0.17869473684210527,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00046834109387812554,
+      "loss": 3.792,
+      "step": 4244
+    },
+    {
+      "epoch": 0.17873684210526317,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004683246561489859,
+      "loss": 3.3878,
+      "step": 4245
+    },
+    {
+      "epoch": 0.17877894736842107,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00046830821444222766,
+      "loss": 3.4518,
+      "step": 4246
+    },
+    {
+      "epoch": 0.17882105263157894,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00046829176875815026,
+      "loss": 3.124,
+      "step": 4247
+    },
+    {
+      "epoch": 0.17886315789473684,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004682753190970533,
+      "loss": 3.5221,
+      "step": 4248
+    },
+    {
+      "epoch": 0.17890526315789473,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00046825886545923644,
+      "loss": 3.2836,
+      "step": 4249
+    },
+    {
+      "epoch": 0.17894736842105263,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004682424078449995,
+      "loss": 3.4374,
+      "step": 4250
+    },
+    {
+      "epoch": 0.17898947368421053,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00046822594625464236,
+      "loss": 3.6944,
+      "step": 4251
+    },
+    {
+      "epoch": 0.17903157894736843,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004682094806884649,
+      "loss": 3.4054,
+      "step": 4252
+    },
+    {
+      "epoch": 0.17907368421052633,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00046819301114676704,
+      "loss": 3.46,
+      "step": 4253
+    },
+    {
+      "epoch": 0.1791157894736842,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004681765376298489,
+      "loss": 3.4442,
+      "step": 4254
+    },
+    {
+      "epoch": 0.1791578947368421,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0004681600601380106,
+      "loss": 3.7417,
+      "step": 4255
+    },
+    {
+      "epoch": 0.1792,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004681435786715523,
+      "loss": 3.4556,
+      "step": 4256
+    },
+    {
+      "epoch": 0.1792421052631579,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004681270932307743,
+      "loss": 3.3891,
+      "step": 4257
+    },
+    {
+      "epoch": 0.1792842105263158,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.000468110603815977,
+      "loss": 3.3337,
+      "step": 4258
+    },
+    {
+      "epoch": 0.1793263157894737,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00046809411042746075,
+      "loss": 3.515,
+      "step": 4259
+    },
+    {
+      "epoch": 0.17936842105263157,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004680776130655261,
+      "loss": 3.4017,
+      "step": 4260
+    },
+    {
+      "epoch": 0.17941052631578946,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0004680611117304735,
+      "loss": 3.9449,
+      "step": 4261
+    },
+    {
+      "epoch": 0.17945263157894736,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004680446064226037,
+      "loss": 3.157,
+      "step": 4262
+    },
+    {
+      "epoch": 0.17949473684210526,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004680280971422173,
+      "loss": 3.3798,
+      "step": 4263
+    },
+    {
+      "epoch": 0.17953684210526316,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00046801158388961515,
+      "loss": 3.5937,
+      "step": 4264
+    },
+    {
+      "epoch": 0.17957894736842106,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00046799506666509816,
+      "loss": 3.2309,
+      "step": 4265
+    },
+    {
+      "epoch": 0.17962105263157896,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00046797854546896717,
+      "loss": 3.2346,
+      "step": 4266
+    },
+    {
+      "epoch": 0.17966315789473683,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004679620203015232,
+      "loss": 3.336,
+      "step": 4267
+    },
+    {
+      "epoch": 0.17970526315789473,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004679454911630673,
+      "loss": 3.4914,
+      "step": 4268
+    },
+    {
+      "epoch": 0.17974736842105263,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00046792895805390064,
+      "loss": 3.5245,
+      "step": 4269
+    },
+    {
+      "epoch": 0.17978947368421053,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004679124209743244,
+      "loss": 3.7937,
+      "step": 4270
+    },
+    {
+      "epoch": 0.17983157894736843,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0004678958799246399,
+      "loss": 3.2519,
+      "step": 4271
+    },
+    {
+      "epoch": 0.17987368421052632,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00046787933490514846,
+      "loss": 3.4968,
+      "step": 4272
+    },
+    {
+      "epoch": 0.17991578947368422,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004678627859161515,
+      "loss": 3.4933,
+      "step": 4273
+    },
+    {
+      "epoch": 0.1799578947368421,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004678462329579506,
+      "loss": 3.4709,
+      "step": 4274
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00046782967603084736,
+      "loss": 3.194,
+      "step": 4275
+    },
+    {
+      "epoch": 0.1800421052631579,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004678131151351433,
+      "loss": 3.3363,
+      "step": 4276
+    },
+    {
+      "epoch": 0.1800842105263158,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004677965502711402,
+      "loss": 3.0924,
+      "step": 4277
+    },
+    {
+      "epoch": 0.1801263157894737,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00046777998143913985,
+      "loss": 3.1704,
+      "step": 4278
+    },
+    {
+      "epoch": 0.1801684210526316,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0004677634086394441,
+      "loss": 3.2711,
+      "step": 4279
+    },
+    {
+      "epoch": 0.18021052631578946,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.000467746831872355,
+      "loss": 3.5172,
+      "step": 4280
+    },
+    {
+      "epoch": 0.18025263157894736,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004677302511381744,
+      "loss": 3.2466,
+      "step": 4281
+    },
+    {
+      "epoch": 0.18029473684210526,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004677136664372045,
+      "loss": 2.9772,
+      "step": 4282
+    },
+    {
+      "epoch": 0.18033684210526316,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0004676970777697473,
+      "loss": 3.4742,
+      "step": 4283
+    },
+    {
+      "epoch": 0.18037894736842106,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00046768048513610517,
+      "loss": 3.5518,
+      "step": 4284
+    },
+    {
+      "epoch": 0.18042105263157895,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00046766388853658036,
+      "loss": 3.5894,
+      "step": 4285
+    },
+    {
+      "epoch": 0.18046315789473685,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00046764728797147524,
+      "loss": 3.1369,
+      "step": 4286
+    },
+    {
+      "epoch": 0.18050526315789472,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00046763068344109226,
+      "loss": 3.4405,
+      "step": 4287
+    },
+    {
+      "epoch": 0.18054736842105262,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00046761407494573407,
+      "loss": 3.1766,
+      "step": 4288
+    },
+    {
+      "epoch": 0.18058947368421052,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004675974624857029,
+      "loss": 3.6131,
+      "step": 4289
+    },
+    {
+      "epoch": 0.18063157894736842,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004675808460613018,
+      "loss": 3.4915,
+      "step": 4290
+    },
+    {
+      "epoch": 0.18067368421052632,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004675642256728332,
+      "loss": 3.9375,
+      "step": 4291
+    },
+    {
+      "epoch": 0.18071578947368422,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00046754760132060015,
+      "loss": 3.8096,
+      "step": 4292
+    },
+    {
+      "epoch": 0.18075789473684212,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00046753097300490534,
+      "loss": 3.6347,
+      "step": 4293
+    },
+    {
+      "epoch": 0.1808,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0004675143407260518,
+      "loss": 3.318,
+      "step": 4294
+    },
+    {
+      "epoch": 0.1808421052631579,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004674977044843426,
+      "loss": 2.9944,
+      "step": 4295
+    },
+    {
+      "epoch": 0.1808842105263158,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004674810642800806,
+      "loss": 3.7377,
+      "step": 4296
+    },
+    {
+      "epoch": 0.18092631578947369,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0004674644201135694,
+      "loss": 3.5213,
+      "step": 4297
+    },
+    {
+      "epoch": 0.18096842105263158,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0004674477719851118,
+      "loss": 3.2998,
+      "step": 4298
+    },
+    {
+      "epoch": 0.18101052631578948,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004674311198950113,
+      "loss": 3.3592,
+      "step": 4299
+    },
+    {
+      "epoch": 0.18105263157894738,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00046741446384357133,
+      "loss": 3.5434,
+      "step": 4300
+    },
+    {
+      "epoch": 0.18109473684210525,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004673978038310952,
+      "loss": 3.2923,
+      "step": 4301
+    },
+    {
+      "epoch": 0.18113684210526315,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00046738113985788656,
+      "loss": 3.0153,
+      "step": 4302
+    },
+    {
+      "epoch": 0.18117894736842105,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.000467364471924249,
+      "loss": 3.2006,
+      "step": 4303
+    },
+    {
+      "epoch": 0.18122105263157895,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00046734780003048604,
+      "loss": 3.7719,
+      "step": 4304
+    },
+    {
+      "epoch": 0.18126315789473685,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00046733112417690165,
+      "loss": 3.3493,
+      "step": 4305
+    },
+    {
+      "epoch": 0.18130526315789475,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004673144443637995,
+      "loss": 2.6939,
+      "step": 4306
+    },
+    {
+      "epoch": 0.18134736842105262,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004672977605914835,
+      "loss": 3.4434,
+      "step": 4307
+    },
+    {
+      "epoch": 0.18138947368421052,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004672810728602576,
+      "loss": 3.5016,
+      "step": 4308
+    },
+    {
+      "epoch": 0.18143157894736842,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00046726438117042585,
+      "loss": 3.7426,
+      "step": 4309
+    },
+    {
+      "epoch": 0.18147368421052631,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00046724768552229237,
+      "loss": 3.2753,
+      "step": 4310
+    },
+    {
+      "epoch": 0.1815157894736842,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00046723098591616133,
+      "loss": 3.2095,
+      "step": 4311
+    },
+    {
+      "epoch": 0.1815578947368421,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00046721428235233696,
+      "loss": 2.8532,
+      "step": 4312
+    },
+    {
+      "epoch": 0.1816,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004671975748311236,
+      "loss": 3.577,
+      "step": 4313
+    },
+    {
+      "epoch": 0.18164210526315788,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004671808633528255,
+      "loss": 3.7791,
+      "step": 4314
+    },
+    {
+      "epoch": 0.18168421052631578,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004671641479177474,
+      "loss": 3.4846,
+      "step": 4315
+    },
+    {
+      "epoch": 0.18172631578947368,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004671474285261936,
+      "loss": 3.9367,
+      "step": 4316
+    },
+    {
+      "epoch": 0.18176842105263158,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004671307051784688,
+      "loss": 3.732,
+      "step": 4317
+    },
+    {
+      "epoch": 0.18181052631578948,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00046711397787487766,
+      "loss": 3.7979,
+      "step": 4318
+    },
+    {
+      "epoch": 0.18185263157894738,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00046709724661572494,
+      "loss": 3.3694,
+      "step": 4319
+    },
+    {
+      "epoch": 0.18189473684210528,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004670805114013155,
+      "loss": 3.1885,
+      "step": 4320
+    },
+    {
+      "epoch": 0.18193684210526315,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004670637722319542,
+      "loss": 3.1586,
+      "step": 4321
+    },
+    {
+      "epoch": 0.18197894736842105,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.000467047029107946,
+      "loss": 3.4971,
+      "step": 4322
+    },
+    {
+      "epoch": 0.18202105263157894,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004670302820295959,
+      "loss": 3.3221,
+      "step": 4323
+    },
+    {
+      "epoch": 0.18206315789473684,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0004670135309972091,
+      "loss": 3.2375,
+      "step": 4324
+    },
+    {
+      "epoch": 0.18210526315789474,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004669967760110908,
+      "loss": 3.1671,
+      "step": 4325
+    },
+    {
+      "epoch": 0.18214736842105264,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00046698001707154614,
+      "loss": 3.7695,
+      "step": 4326
+    },
+    {
+      "epoch": 0.18218947368421054,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004669632541788805,
+      "loss": 3.3264,
+      "step": 4327
+    },
+    {
+      "epoch": 0.1822315789473684,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004669464873333993,
+      "loss": 3.2035,
+      "step": 4328
+    },
+    {
+      "epoch": 0.1822736842105263,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00046692971653540804,
+      "loss": 3.2817,
+      "step": 4329
+    },
+    {
+      "epoch": 0.1823157894736842,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00046691294178521214,
+      "loss": 3.4109,
+      "step": 4330
+    },
+    {
+      "epoch": 0.1823578947368421,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00046689616308311744,
+      "loss": 3.2351,
+      "step": 4331
+    },
+    {
+      "epoch": 0.1824,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004668793804294294,
+      "loss": 3.4091,
+      "step": 4332
+    },
+    {
+      "epoch": 0.1824421052631579,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004668625938244538,
+      "loss": 3.3767,
+      "step": 4333
+    },
+    {
+      "epoch": 0.18248421052631578,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00046684580326849666,
+      "loss": 3.4928,
+      "step": 4334
+    },
+    {
+      "epoch": 0.18252631578947368,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0004668290087618638,
+      "loss": 3.0895,
+      "step": 4335
+    },
+    {
+      "epoch": 0.18256842105263157,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004668122103048611,
+      "loss": 3.3407,
+      "step": 4336
+    },
+    {
+      "epoch": 0.18261052631578947,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00046679540789779463,
+      "loss": 3.6747,
+      "step": 4337
+    },
+    {
+      "epoch": 0.18265263157894737,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004667786015409706,
+      "loss": 3.9243,
+      "step": 4338
+    },
+    {
+      "epoch": 0.18269473684210527,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004667617912346952,
+      "loss": 3.0803,
+      "step": 4339
+    },
+    {
+      "epoch": 0.18273684210526317,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00046674497697927456,
+      "loss": 3.0517,
+      "step": 4340
+    },
+    {
+      "epoch": 0.18277894736842104,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004667281587750152,
+      "loss": 3.5114,
+      "step": 4341
+    },
+    {
+      "epoch": 0.18282105263157894,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00046671133662222334,
+      "loss": 3.7038,
+      "step": 4342
+    },
+    {
+      "epoch": 0.18286315789473684,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0004666945105212056,
+      "loss": 3.0405,
+      "step": 4343
+    },
+    {
+      "epoch": 0.18290526315789474,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00046667768047226846,
+      "loss": 3.3121,
+      "step": 4344
+    },
+    {
+      "epoch": 0.18294736842105264,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004666608464757186,
+      "loss": 3.5014,
+      "step": 4345
+    },
+    {
+      "epoch": 0.18298947368421054,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004666440085318626,
+      "loss": 3.4629,
+      "step": 4346
+    },
+    {
+      "epoch": 0.18303157894736843,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0004666271666410074,
+      "loss": 2.9162,
+      "step": 4347
+    },
+    {
+      "epoch": 0.1830736842105263,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004666103208034597,
+      "loss": 3.4681,
+      "step": 4348
+    },
+    {
+      "epoch": 0.1831157894736842,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00046659347101952656,
+      "loss": 3.255,
+      "step": 4349
+    },
+    {
+      "epoch": 0.1831578947368421,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004665766172895147,
+      "loss": 3.046,
+      "step": 4350
+    },
+    {
+      "epoch": 0.1832,
+      "grad_norm": 5.46875,
+      "learning_rate": 0.00046655975961373147,
+      "loss": 3.4244,
+      "step": 4351
+    },
+    {
+      "epoch": 0.1832421052631579,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00046654289799248383,
+      "loss": 2.9638,
+      "step": 4352
+    },
+    {
+      "epoch": 0.1832842105263158,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00046652603242607906,
+      "loss": 3.5785,
+      "step": 4353
+    },
+    {
+      "epoch": 0.18332631578947367,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00046650916291482425,
+      "loss": 3.2352,
+      "step": 4354
+    },
+    {
+      "epoch": 0.18336842105263157,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.000466492289459027,
+      "loss": 3.4128,
+      "step": 4355
+    },
+    {
+      "epoch": 0.18341052631578947,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00046647541205899456,
+      "loss": 3.1858,
+      "step": 4356
+    },
+    {
+      "epoch": 0.18345263157894737,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004664585307150345,
+      "loss": 3.5665,
+      "step": 4357
+    },
+    {
+      "epoch": 0.18349473684210527,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00046644164542745427,
+      "loss": 3.7898,
+      "step": 4358
+    },
+    {
+      "epoch": 0.18353684210526316,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00046642475619656165,
+      "loss": 3.0775,
+      "step": 4359
+    },
+    {
+      "epoch": 0.18357894736842106,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00046640786302266425,
+      "loss": 3.3362,
+      "step": 4360
+    },
+    {
+      "epoch": 0.18362105263157893,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00046639096590606977,
+      "loss": 3.3037,
+      "step": 4361
+    },
+    {
+      "epoch": 0.18366315789473683,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00046637406484708623,
+      "loss": 3.9068,
+      "step": 4362
+    },
+    {
+      "epoch": 0.18370526315789473,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004663571598460214,
+      "loss": 3.4112,
+      "step": 4363
+    },
+    {
+      "epoch": 0.18374736842105263,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004663402509031833,
+      "loss": 3.1825,
+      "step": 4364
+    },
+    {
+      "epoch": 0.18378947368421053,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004663233380188802,
+      "loss": 3.2201,
+      "step": 4365
+    },
+    {
+      "epoch": 0.18383157894736843,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004663064211934198,
+      "loss": 3.3133,
+      "step": 4366
+    },
+    {
+      "epoch": 0.18387368421052633,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004662895004271107,
+      "loss": 3.0813,
+      "step": 4367
+    },
+    {
+      "epoch": 0.1839157894736842,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00046627257572026107,
+      "loss": 3.5289,
+      "step": 4368
+    },
+    {
+      "epoch": 0.1839578947368421,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0004662556470731791,
+      "loss": 3.5994,
+      "step": 4369
+    },
+    {
+      "epoch": 0.184,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004662387144861734,
+      "loss": 3.9798,
+      "step": 4370
+    },
+    {
+      "epoch": 0.1840421052631579,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00046622177795955236,
+      "loss": 3.0208,
+      "step": 4371
+    },
+    {
+      "epoch": 0.1840842105263158,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00046620483749362465,
+      "loss": 3.6957,
+      "step": 4372
+    },
+    {
+      "epoch": 0.1841263157894737,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00046618789308869877,
+      "loss": 3.1677,
+      "step": 4373
+    },
+    {
+      "epoch": 0.1841684210526316,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004661709447450835,
+      "loss": 3.7236,
+      "step": 4374
+    },
+    {
+      "epoch": 0.18421052631578946,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00046615399246308754,
+      "loss": 3.4274,
+      "step": 4375
+    },
+    {
+      "epoch": 0.18425263157894736,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004661370362430199,
+      "loss": 3.6138,
+      "step": 4376
+    },
+    {
+      "epoch": 0.18429473684210526,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00046612007608518935,
+      "loss": 2.9665,
+      "step": 4377
+    },
+    {
+      "epoch": 0.18433684210526316,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00046610311198990497,
+      "loss": 3.4138,
+      "step": 4378
+    },
+    {
+      "epoch": 0.18437894736842106,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004660861439574758,
+      "loss": 2.9325,
+      "step": 4379
+    },
+    {
+      "epoch": 0.18442105263157896,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00046606917198821094,
+      "loss": 3.3308,
+      "step": 4380
+    },
+    {
+      "epoch": 0.18446315789473683,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004660521960824198,
+      "loss": 3.5993,
+      "step": 4381
+    },
+    {
+      "epoch": 0.18450526315789473,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004660352162404113,
+      "loss": 3.3553,
+      "step": 4382
+    },
+    {
+      "epoch": 0.18454736842105263,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00046601823246249506,
+      "loss": 3.3686,
+      "step": 4383
+    },
+    {
+      "epoch": 0.18458947368421053,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004660012447489805,
+      "loss": 3.2605,
+      "step": 4384
+    },
+    {
+      "epoch": 0.18463157894736842,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.000465984253100177,
+      "loss": 3.3257,
+      "step": 4385
+    },
+    {
+      "epoch": 0.18467368421052632,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00046596725751639413,
+      "loss": 3.4743,
+      "step": 4386
+    },
+    {
+      "epoch": 0.18471578947368422,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00046595025799794166,
+      "loss": 3.061,
+      "step": 4387
+    },
+    {
+      "epoch": 0.1847578947368421,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00046593325454512914,
+      "loss": 3.3951,
+      "step": 4388
+    },
+    {
+      "epoch": 0.1848,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00046591624715826654,
+      "loss": 3.3823,
+      "step": 4389
+    },
+    {
+      "epoch": 0.1848421052631579,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00046589923583766345,
+      "loss": 3.1218,
+      "step": 4390
+    },
+    {
+      "epoch": 0.1848842105263158,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004658822205836301,
+      "loss": 3.3523,
+      "step": 4391
+    },
+    {
+      "epoch": 0.1849263157894737,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00046586520139647625,
+      "loss": 3.2584,
+      "step": 4392
+    },
+    {
+      "epoch": 0.1849684210526316,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00046584817827651214,
+      "loss": 3.2319,
+      "step": 4393
+    },
+    {
+      "epoch": 0.1850105263157895,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00046583115122404787,
+      "loss": 3.1867,
+      "step": 4394
+    },
+    {
+      "epoch": 0.18505263157894736,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004658141202393935,
+      "loss": 3.7308,
+      "step": 4395
+    },
+    {
+      "epoch": 0.18509473684210526,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00046579708532285945,
+      "loss": 3.4538,
+      "step": 4396
+    },
+    {
+      "epoch": 0.18513684210526316,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00046578004647475607,
+      "loss": 2.8032,
+      "step": 4397
+    },
+    {
+      "epoch": 0.18517894736842105,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004657630036953938,
+      "loss": 3.5403,
+      "step": 4398
+    },
+    {
+      "epoch": 0.18522105263157895,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004657459569850831,
+      "loss": 3.6156,
+      "step": 4399
+    },
+    {
+      "epoch": 0.18526315789473685,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00046572890634413456,
+      "loss": 3.9091,
+      "step": 4400
+    },
+    {
+      "epoch": 0.18530526315789475,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004657118517728588,
+      "loss": 3.2509,
+      "step": 4401
+    },
+    {
+      "epoch": 0.18534736842105262,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00046569479327156655,
+      "loss": 3.4803,
+      "step": 4402
+    },
+    {
+      "epoch": 0.18538947368421052,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004656777308405686,
+      "loss": 3.1033,
+      "step": 4403
+    },
+    {
+      "epoch": 0.18543157894736842,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00046566066448017586,
+      "loss": 3.2882,
+      "step": 4404
+    },
+    {
+      "epoch": 0.18547368421052632,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00046564359419069915,
+      "loss": 3.4529,
+      "step": 4405
+    },
+    {
+      "epoch": 0.18551578947368422,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00046562651997244955,
+      "loss": 3.442,
+      "step": 4406
+    },
+    {
+      "epoch": 0.18555789473684212,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00046560944182573807,
+      "loss": 3.1917,
+      "step": 4407
+    },
+    {
+      "epoch": 0.1856,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00046559235975087587,
+      "loss": 3.0058,
+      "step": 4408
+    },
+    {
+      "epoch": 0.18564210526315789,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004655752737481742,
+      "loss": 3.4125,
+      "step": 4409
+    },
+    {
+      "epoch": 0.18568421052631578,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00046555818381794435,
+      "loss": 3.3958,
+      "step": 4410
+    },
+    {
+      "epoch": 0.18572631578947368,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0004655410899604977,
+      "loss": 3.4025,
+      "step": 4411
+    },
+    {
+      "epoch": 0.18576842105263158,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004655239921761456,
+      "loss": 3.0672,
+      "step": 4412
+    },
+    {
+      "epoch": 0.18581052631578948,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004655068904651996,
+      "loss": 3.8892,
+      "step": 4413
+    },
+    {
+      "epoch": 0.18585263157894738,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004654897848279713,
+      "loss": 3.4316,
+      "step": 4414
+    },
+    {
+      "epoch": 0.18589473684210525,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00046547267526477224,
+      "loss": 3.1254,
+      "step": 4415
+    },
+    {
+      "epoch": 0.18593684210526315,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004654555617759143,
+      "loss": 3.299,
+      "step": 4416
+    },
+    {
+      "epoch": 0.18597894736842105,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00046543844436170914,
+      "loss": 3.5446,
+      "step": 4417
+    },
+    {
+      "epoch": 0.18602105263157895,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00046542132302246857,
+      "loss": 3.2613,
+      "step": 4418
+    },
+    {
+      "epoch": 0.18606315789473685,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004654041977585047,
+      "loss": 3.5915,
+      "step": 4419
+    },
+    {
+      "epoch": 0.18610526315789475,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004653870685701294,
+      "loss": 3.2797,
+      "step": 4420
+    },
+    {
+      "epoch": 0.18614736842105264,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004653699354576548,
+      "loss": 3.2533,
+      "step": 4421
+    },
+    {
+      "epoch": 0.18618947368421052,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.000465352798421393,
+      "loss": 3.034,
+      "step": 4422
+    },
+    {
+      "epoch": 0.18623157894736841,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004653356574616563,
+      "loss": 3.3064,
+      "step": 4423
+    },
+    {
+      "epoch": 0.1862736842105263,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00046531851257875683,
+      "loss": 3.3461,
+      "step": 4424
+    },
+    {
+      "epoch": 0.1863157894736842,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00046530136377300716,
+      "loss": 2.9325,
+      "step": 4425
+    },
+    {
+      "epoch": 0.1863578947368421,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004652842110447195,
+      "loss": 3.5022,
+      "step": 4426
+    },
+    {
+      "epoch": 0.1864,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004652670543942065,
+      "loss": 3.4207,
+      "step": 4427
+    },
+    {
+      "epoch": 0.18644210526315788,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00046524989382178074,
+      "loss": 3.3801,
+      "step": 4428
+    },
+    {
+      "epoch": 0.18648421052631578,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004652327293277548,
+      "loss": 3.8296,
+      "step": 4429
+    },
+    {
+      "epoch": 0.18652631578947368,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004652155609124414,
+      "loss": 3.2296,
+      "step": 4430
+    },
+    {
+      "epoch": 0.18656842105263158,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00046519838857615335,
+      "loss": 3.0834,
+      "step": 4431
+    },
+    {
+      "epoch": 0.18661052631578948,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004651812123192035,
+      "loss": 3.469,
+      "step": 4432
+    },
+    {
+      "epoch": 0.18665263157894738,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004651640321419048,
+      "loss": 3.1003,
+      "step": 4433
+    },
+    {
+      "epoch": 0.18669473684210527,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004651468480445703,
+      "loss": 3.6867,
+      "step": 4434
+    },
+    {
+      "epoch": 0.18673684210526315,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.000465129660027513,
+      "loss": 3.2798,
+      "step": 4435
+    },
+    {
+      "epoch": 0.18677894736842104,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.000465112468091046,
+      "loss": 3.0654,
+      "step": 4436
+    },
+    {
+      "epoch": 0.18682105263157894,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004650952722354826,
+      "loss": 3.282,
+      "step": 4437
+    },
+    {
+      "epoch": 0.18686315789473684,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004650780724611361,
+      "loss": 3.398,
+      "step": 4438
+    },
+    {
+      "epoch": 0.18690526315789474,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00046506086876831976,
+      "loss": 3.4586,
+      "step": 4439
+    },
+    {
+      "epoch": 0.18694736842105264,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004650436611573471,
+      "loss": 3.7511,
+      "step": 4440
+    },
+    {
+      "epoch": 0.18698947368421054,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004650264496285317,
+      "loss": 3.7843,
+      "step": 4441
+    },
+    {
+      "epoch": 0.1870315789473684,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0004650092341821869,
+      "loss": 3.3451,
+      "step": 4442
+    },
+    {
+      "epoch": 0.1870736842105263,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00046499201481862653,
+      "loss": 3.4381,
+      "step": 4443
+    },
+    {
+      "epoch": 0.1871157894736842,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004649747915381642,
+      "loss": 3.3073,
+      "step": 4444
+    },
+    {
+      "epoch": 0.1871578947368421,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004649575643411139,
+      "loss": 3.0032,
+      "step": 4445
+    },
+    {
+      "epoch": 0.1872,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00046494033322778917,
+      "loss": 3.5015,
+      "step": 4446
+    },
+    {
+      "epoch": 0.1872421052631579,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0004649230981985042,
+      "loss": 2.8596,
+      "step": 4447
+    },
+    {
+      "epoch": 0.1872842105263158,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004649058592535729,
+      "loss": 3.6203,
+      "step": 4448
+    },
+    {
+      "epoch": 0.18732631578947367,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004648886163933094,
+      "loss": 3.1498,
+      "step": 4449
+    },
+    {
+      "epoch": 0.18736842105263157,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00046487136961802765,
+      "loss": 3.2854,
+      "step": 4450
+    },
+    {
+      "epoch": 0.18741052631578947,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00046485411892804217,
+      "loss": 3.1656,
+      "step": 4451
+    },
+    {
+      "epoch": 0.18745263157894737,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.000464836864323667,
+      "loss": 3.347,
+      "step": 4452
+    },
+    {
+      "epoch": 0.18749473684210527,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00046481960580521664,
+      "loss": 3.4064,
+      "step": 4453
+    },
+    {
+      "epoch": 0.18753684210526317,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00046480234337300543,
+      "loss": 3.0811,
+      "step": 4454
+    },
+    {
+      "epoch": 0.18757894736842104,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00046478507702734783,
+      "loss": 3.2662,
+      "step": 4455
+    },
+    {
+      "epoch": 0.18762105263157894,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004647678067685586,
+      "loss": 3.2447,
+      "step": 4456
+    },
+    {
+      "epoch": 0.18766315789473684,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004647505325969522,
+      "loss": 3.4242,
+      "step": 4457
+    },
+    {
+      "epoch": 0.18770526315789474,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00046473325451284347,
+      "loss": 3.6144,
+      "step": 4458
+    },
+    {
+      "epoch": 0.18774736842105264,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00046471597251654715,
+      "loss": 3.3922,
+      "step": 4459
+    },
+    {
+      "epoch": 0.18778947368421053,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00046469868660837805,
+      "loss": 3.15,
+      "step": 4460
+    },
+    {
+      "epoch": 0.18783157894736843,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004646813967886512,
+      "loss": 3.5424,
+      "step": 4461
+    },
+    {
+      "epoch": 0.1878736842105263,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00046466410305768147,
+      "loss": 3.2203,
+      "step": 4462
+    },
+    {
+      "epoch": 0.1879157894736842,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00046464680541578396,
+      "loss": 3.47,
+      "step": 4463
+    },
+    {
+      "epoch": 0.1879578947368421,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00046462950386327395,
+      "loss": 3.0017,
+      "step": 4464
+    },
+    {
+      "epoch": 0.188,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0004646121984004665,
+      "loss": 3.4457,
+      "step": 4465
+    },
+    {
+      "epoch": 0.1880421052631579,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00046459488902767703,
+      "loss": 3.3676,
+      "step": 4466
+    },
+    {
+      "epoch": 0.1880842105263158,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00046457757574522074,
+      "loss": 2.9927,
+      "step": 4467
+    },
+    {
+      "epoch": 0.1881263157894737,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004645602585534132,
+      "loss": 3.2295,
+      "step": 4468
+    },
+    {
+      "epoch": 0.18816842105263157,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004645429374525698,
+      "loss": 3.4414,
+      "step": 4469
+    },
+    {
+      "epoch": 0.18821052631578947,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004645256124430062,
+      "loss": 3.1934,
+      "step": 4470
+    },
+    {
+      "epoch": 0.18825263157894737,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.000464508283525038,
+      "loss": 3.0759,
+      "step": 4471
+    },
+    {
+      "epoch": 0.18829473684210526,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004644909506989808,
+      "loss": 3.4296,
+      "step": 4472
+    },
+    {
+      "epoch": 0.18833684210526316,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00046447361396515066,
+      "loss": 3.2265,
+      "step": 4473
+    },
+    {
+      "epoch": 0.18837894736842106,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004644562733238632,
+      "loss": 3.3423,
+      "step": 4474
+    },
+    {
+      "epoch": 0.18842105263157893,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00046443892877543434,
+      "loss": 2.8664,
+      "step": 4475
+    },
+    {
+      "epoch": 0.18846315789473683,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004644215803201803,
+      "loss": 3.5501,
+      "step": 4476
+    },
+    {
+      "epoch": 0.18850526315789473,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00046440422795841687,
+      "loss": 3.3441,
+      "step": 4477
+    },
+    {
+      "epoch": 0.18854736842105263,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004643868716904604,
+      "loss": 3.5883,
+      "step": 4478
+    },
+    {
+      "epoch": 0.18858947368421053,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.000464369511516627,
+      "loss": 3.4756,
+      "step": 4479
+    },
+    {
+      "epoch": 0.18863157894736843,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00046435214743723296,
+      "loss": 3.1863,
+      "step": 4480
+    },
+    {
+      "epoch": 0.18867368421052633,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00046433477945259465,
+      "loss": 3.5316,
+      "step": 4481
+    },
+    {
+      "epoch": 0.1887157894736842,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004643174075630285,
+      "loss": 3.5398,
+      "step": 4482
+    },
+    {
+      "epoch": 0.1887578947368421,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00046430003176885107,
+      "loss": 3.3589,
+      "step": 4483
+    },
+    {
+      "epoch": 0.1888,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00046428265207037876,
+      "loss": 3.7751,
+      "step": 4484
+    },
+    {
+      "epoch": 0.1888421052631579,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004642652684679283,
+      "loss": 3.1204,
+      "step": 4485
+    },
+    {
+      "epoch": 0.1888842105263158,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004642478809618164,
+      "loss": 3.6897,
+      "step": 4486
+    },
+    {
+      "epoch": 0.1889263157894737,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0004642304895523599,
+      "loss": 3.4737,
+      "step": 4487
+    },
+    {
+      "epoch": 0.1889684210526316,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00046421309423987556,
+      "loss": 3.6536,
+      "step": 4488
+    },
+    {
+      "epoch": 0.18901052631578946,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00046419569502468026,
+      "loss": 3.6034,
+      "step": 4489
+    },
+    {
+      "epoch": 0.18905263157894736,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00046417829190709115,
+      "loss": 3.3792,
+      "step": 4490
+    },
+    {
+      "epoch": 0.18909473684210526,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004641608848874252,
+      "loss": 3.3922,
+      "step": 4491
+    },
+    {
+      "epoch": 0.18913684210526316,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00046414347396599956,
+      "loss": 3.6969,
+      "step": 4492
+    },
+    {
+      "epoch": 0.18917894736842106,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00046412605914313145,
+      "loss": 3.4315,
+      "step": 4493
+    },
+    {
+      "epoch": 0.18922105263157896,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00046410864041913804,
+      "loss": 3.6181,
+      "step": 4494
+    },
+    {
+      "epoch": 0.18926315789473686,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00046409121779433684,
+      "loss": 3.3158,
+      "step": 4495
+    },
+    {
+      "epoch": 0.18930526315789473,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004640737912690452,
+      "loss": 3.4135,
+      "step": 4496
+    },
+    {
+      "epoch": 0.18934736842105263,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004640563608435806,
+      "loss": 3.5823,
+      "step": 4497
+    },
+    {
+      "epoch": 0.18938947368421052,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00046403892651826056,
+      "loss": 3.4215,
+      "step": 4498
+    },
+    {
+      "epoch": 0.18943157894736842,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00046402148829340284,
+      "loss": 3.55,
+      "step": 4499
+    },
+    {
+      "epoch": 0.18947368421052632,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.000464004046169325,
+      "loss": 3.1173,
+      "step": 4500
+    },
+    {
+      "epoch": 0.18951578947368422,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00046398660014634495,
+      "loss": 3.1488,
+      "step": 4501
+    },
+    {
+      "epoch": 0.1895578947368421,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004639691502247805,
+      "loss": 3.5974,
+      "step": 4502
+    },
+    {
+      "epoch": 0.1896,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00046395169640494943,
+      "loss": 3.386,
+      "step": 4503
+    },
+    {
+      "epoch": 0.1896421052631579,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004639342386871699,
+      "loss": 3.2139,
+      "step": 4504
+    },
+    {
+      "epoch": 0.1896842105263158,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00046391677707175984,
+      "loss": 3.4286,
+      "step": 4505
+    },
+    {
+      "epoch": 0.1897263157894737,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004638993115590375,
+      "loss": 3.4419,
+      "step": 4506
+    },
+    {
+      "epoch": 0.1897684210526316,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00046388184214932106,
+      "loss": 3.2082,
+      "step": 4507
+    },
+    {
+      "epoch": 0.18981052631578949,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004638643688429287,
+      "loss": 3.6518,
+      "step": 4508
+    },
+    {
+      "epoch": 0.18985263157894736,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004638468916401788,
+      "loss": 3.9826,
+      "step": 4509
+    },
+    {
+      "epoch": 0.18989473684210526,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004638294105413898,
+      "loss": 3.1157,
+      "step": 4510
+    },
+    {
+      "epoch": 0.18993684210526315,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00046381192554688025,
+      "loss": 3.6597,
+      "step": 4511
+    },
+    {
+      "epoch": 0.18997894736842105,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0004637944366569686,
+      "loss": 3.486,
+      "step": 4512
+    },
+    {
+      "epoch": 0.19002105263157895,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004637769438719735,
+      "loss": 3.5676,
+      "step": 4513
+    },
+    {
+      "epoch": 0.19006315789473685,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00046375944719221366,
+      "loss": 3.2566,
+      "step": 4514
+    },
+    {
+      "epoch": 0.19010526315789475,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0004637419466180078,
+      "loss": 3.4759,
+      "step": 4515
+    },
+    {
+      "epoch": 0.19014736842105262,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004637244421496749,
+      "loss": 3.5676,
+      "step": 4516
+    },
+    {
+      "epoch": 0.19018947368421052,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00046370693378753377,
+      "loss": 3.2276,
+      "step": 4517
+    },
+    {
+      "epoch": 0.19023157894736842,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004636894215319033,
+      "loss": 3.2458,
+      "step": 4518
+    },
+    {
+      "epoch": 0.19027368421052632,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00046367190538310277,
+      "loss": 3.2596,
+      "step": 4519
+    },
+    {
+      "epoch": 0.19031578947368422,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0004636543853414511,
+      "loss": 3.7653,
+      "step": 4520
+    },
+    {
+      "epoch": 0.19035789473684211,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004636368614072676,
+      "loss": 3.7527,
+      "step": 4521
+    },
+    {
+      "epoch": 0.1904,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004636193335808715,
+      "loss": 3.5574,
+      "step": 4522
+    },
+    {
+      "epoch": 0.19044210526315788,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004636018018625821,
+      "loss": 3.4611,
+      "step": 4523
+    },
+    {
+      "epoch": 0.19048421052631578,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004635842662527189,
+      "loss": 2.6897,
+      "step": 4524
+    },
+    {
+      "epoch": 0.19052631578947368,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0004635667267516013,
+      "loss": 3.4963,
+      "step": 4525
+    },
+    {
+      "epoch": 0.19056842105263158,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00046354918335954874,
+      "loss": 3.6976,
+      "step": 4526
+    },
+    {
+      "epoch": 0.19061052631578948,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00046353163607688115,
+      "loss": 3.523,
+      "step": 4527
+    },
+    {
+      "epoch": 0.19065263157894738,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.000463514084903918,
+      "loss": 3.7666,
+      "step": 4528
+    },
+    {
+      "epoch": 0.19069473684210525,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.000463496529840979,
+      "loss": 3.3872,
+      "step": 4529
+    },
+    {
+      "epoch": 0.19073684210526315,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004634789708883842,
+      "loss": 3.4158,
+      "step": 4530
+    },
+    {
+      "epoch": 0.19077894736842105,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00046346140804645326,
+      "loss": 3.4758,
+      "step": 4531
+    },
+    {
+      "epoch": 0.19082105263157895,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004634438413155063,
+      "loss": 3.3974,
+      "step": 4532
+    },
+    {
+      "epoch": 0.19086315789473685,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004634262706958634,
+      "loss": 2.8901,
+      "step": 4533
+    },
+    {
+      "epoch": 0.19090526315789474,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004634086961878446,
+      "loss": 3.608,
+      "step": 4534
+    },
+    {
+      "epoch": 0.19094736842105264,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00046339111779177005,
+      "loss": 3.8581,
+      "step": 4535
+    },
+    {
+      "epoch": 0.19098947368421051,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00046337353550796013,
+      "loss": 3.2613,
+      "step": 4536
+    },
+    {
+      "epoch": 0.1910315789473684,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.000463355949336735,
+      "loss": 2.9671,
+      "step": 4537
+    },
+    {
+      "epoch": 0.1910736842105263,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0004633383592784152,
+      "loss": 3.9004,
+      "step": 4538
+    },
+    {
+      "epoch": 0.1911157894736842,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00046332076533332116,
+      "loss": 3.4956,
+      "step": 4539
+    },
+    {
+      "epoch": 0.1911578947368421,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004633031675017735,
+      "loss": 3.2314,
+      "step": 4540
+    },
+    {
+      "epoch": 0.1912,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00046328556578409267,
+      "loss": 3.8581,
+      "step": 4541
+    },
+    {
+      "epoch": 0.1912421052631579,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00046326796018059937,
+      "loss": 3.1897,
+      "step": 4542
+    },
+    {
+      "epoch": 0.19128421052631578,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004632503506916145,
+      "loss": 3.6953,
+      "step": 4543
+    },
+    {
+      "epoch": 0.19132631578947368,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004632327373174587,
+      "loss": 3.5021,
+      "step": 4544
+    },
+    {
+      "epoch": 0.19136842105263158,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00046321512005845314,
+      "loss": 3.2975,
+      "step": 4545
+    },
+    {
+      "epoch": 0.19141052631578948,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004631974989149185,
+      "loss": 3.2312,
+      "step": 4546
+    },
+    {
+      "epoch": 0.19145263157894737,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004631798738871759,
+      "loss": 3.6851,
+      "step": 4547
+    },
+    {
+      "epoch": 0.19149473684210527,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00046316224497554656,
+      "loss": 3.1115,
+      "step": 4548
+    },
+    {
+      "epoch": 0.19153684210526314,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004631446121803515,
+      "loss": 3.1139,
+      "step": 4549
+    },
+    {
+      "epoch": 0.19157894736842104,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00046312697550191206,
+      "loss": 3.1891,
+      "step": 4550
+    },
+    {
+      "epoch": 0.19162105263157894,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00046310933494054954,
+      "loss": 3.4793,
+      "step": 4551
+    },
+    {
+      "epoch": 0.19166315789473684,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.0004630916904965853,
+      "loss": 3.0458,
+      "step": 4552
+    },
+    {
+      "epoch": 0.19170526315789474,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004630740421703409,
+      "loss": 3.2249,
+      "step": 4553
+    },
+    {
+      "epoch": 0.19174736842105264,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00046305638996213777,
+      "loss": 3.0532,
+      "step": 4554
+    },
+    {
+      "epoch": 0.19178947368421054,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00046303873387229755,
+      "loss": 3.3277,
+      "step": 4555
+    },
+    {
+      "epoch": 0.1918315789473684,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0004630210739011419,
+      "loss": 3.5759,
+      "step": 4556
+    },
+    {
+      "epoch": 0.1918736842105263,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0004630034100489926,
+      "loss": 3.2083,
+      "step": 4557
+    },
+    {
+      "epoch": 0.1919157894736842,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00046298574231617144,
+      "loss": 3.17,
+      "step": 4558
+    },
+    {
+      "epoch": 0.1919578947368421,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004629680707030002,
+      "loss": 3.1794,
+      "step": 4559
+    },
+    {
+      "epoch": 0.192,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0004629503952098011,
+      "loss": 3.2657,
+      "step": 4560
+    },
+    {
+      "epoch": 0.1920421052631579,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0004629327158368959,
+      "loss": 3.4405,
+      "step": 4561
+    },
+    {
+      "epoch": 0.1920842105263158,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004629150325846068,
+      "loss": 3.9288,
+      "step": 4562
+    },
+    {
+      "epoch": 0.19212631578947367,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.000462897345453256,
+      "loss": 3.5464,
+      "step": 4563
+    },
+    {
+      "epoch": 0.19216842105263157,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004628796544431657,
+      "loss": 3.5072,
+      "step": 4564
+    },
+    {
+      "epoch": 0.19221052631578947,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00046286195955465824,
+      "loss": 3.4338,
+      "step": 4565
+    },
+    {
+      "epoch": 0.19225263157894737,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.000462844260788056,
+      "loss": 3.0108,
+      "step": 4566
+    },
+    {
+      "epoch": 0.19229473684210527,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00046282655814368134,
+      "loss": 3.4061,
+      "step": 4567
+    },
+    {
+      "epoch": 0.19233684210526317,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0004628088516218569,
+      "loss": 3.4992,
+      "step": 4568
+    },
+    {
+      "epoch": 0.19237894736842107,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0004627911412229052,
+      "loss": 2.7391,
+      "step": 4569
+    },
+    {
+      "epoch": 0.19242105263157894,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00046277342694714895,
+      "loss": 3.5581,
+      "step": 4570
+    },
+    {
+      "epoch": 0.19246315789473684,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004627557087949108,
+      "loss": 3.6555,
+      "step": 4571
+    },
+    {
+      "epoch": 0.19250526315789473,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00046273798676651367,
+      "loss": 3.6472,
+      "step": 4572
+    },
+    {
+      "epoch": 0.19254736842105263,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00046272026086228036,
+      "loss": 3.7668,
+      "step": 4573
+    },
+    {
+      "epoch": 0.19258947368421053,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004627025310825339,
+      "loss": 3.4273,
+      "step": 4574
+    },
+    {
+      "epoch": 0.19263157894736843,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00046268479742759714,
+      "loss": 3.5233,
+      "step": 4575
+    },
+    {
+      "epoch": 0.1926736842105263,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0004626670598977933,
+      "loss": 3.2058,
+      "step": 4576
+    },
+    {
+      "epoch": 0.1927157894736842,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00046264931849344547,
+      "loss": 3.0236,
+      "step": 4577
+    },
+    {
+      "epoch": 0.1927578947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00046263157321487695,
+      "loss": 3.2673,
+      "step": 4578
+    },
+    {
+      "epoch": 0.1928,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00046261382406241104,
+      "loss": 3.5448,
+      "step": 4579
+    },
+    {
+      "epoch": 0.1928421052631579,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.000462596071036371,
+      "loss": 3.0543,
+      "step": 4580
+    },
+    {
+      "epoch": 0.1928842105263158,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004625783141370804,
+      "loss": 3.5139,
+      "step": 4581
+    },
+    {
+      "epoch": 0.1929263157894737,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004625605533648626,
+      "loss": 2.7327,
+      "step": 4582
+    },
+    {
+      "epoch": 0.19296842105263157,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004625427887200413,
+      "loss": 3.6466,
+      "step": 4583
+    },
+    {
+      "epoch": 0.19301052631578947,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00046252502020294016,
+      "loss": 3.2897,
+      "step": 4584
+    },
+    {
+      "epoch": 0.19305263157894736,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0004625072478138828,
+      "loss": 3.5112,
+      "step": 4585
+    },
+    {
+      "epoch": 0.19309473684210526,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004624894715531931,
+      "loss": 3.4037,
+      "step": 4586
+    },
+    {
+      "epoch": 0.19313684210526316,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004624716914211949,
+      "loss": 3.0236,
+      "step": 4587
+    },
+    {
+      "epoch": 0.19317894736842106,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004624539074182121,
+      "loss": 3.5033,
+      "step": 4588
+    },
+    {
+      "epoch": 0.19322105263157896,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0004624361195445688,
+      "loss": 3.2123,
+      "step": 4589
+    },
+    {
+      "epoch": 0.19326315789473683,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0004624183278005889,
+      "loss": 2.9114,
+      "step": 4590
+    },
+    {
+      "epoch": 0.19330526315789473,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00046240053218659673,
+      "loss": 3.2894,
+      "step": 4591
+    },
+    {
+      "epoch": 0.19334736842105263,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00046238273270291636,
+      "loss": 3.2422,
+      "step": 4592
+    },
+    {
+      "epoch": 0.19338947368421053,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00046236492934987217,
+      "loss": 3.2904,
+      "step": 4593
+    },
+    {
+      "epoch": 0.19343157894736843,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004623471221277885,
+      "loss": 3.5449,
+      "step": 4594
+    },
+    {
+      "epoch": 0.19347368421052633,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0004623293110369897,
+      "loss": 3.7585,
+      "step": 4595
+    },
+    {
+      "epoch": 0.19351578947368422,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00046231149607780045,
+      "loss": 3.6394,
+      "step": 4596
+    },
+    {
+      "epoch": 0.1935578947368421,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00046229367725054505,
+      "loss": 3.8085,
+      "step": 4597
+    },
+    {
+      "epoch": 0.1936,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00046227585455554845,
+      "loss": 3.2591,
+      "step": 4598
+    },
+    {
+      "epoch": 0.1936421052631579,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00046225802799313505,
+      "loss": 3.3793,
+      "step": 4599
+    },
+    {
+      "epoch": 0.1936842105263158,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00046224019756362984,
+      "loss": 3.9237,
+      "step": 4600
+    },
+    {
+      "epoch": 0.1937263157894737,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004622223632673576,
+      "loss": 3.448,
+      "step": 4601
+    },
+    {
+      "epoch": 0.1937684210526316,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004622045251046432,
+      "loss": 3.3553,
+      "step": 4602
+    },
+    {
+      "epoch": 0.19381052631578946,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00046218668307581174,
+      "loss": 3.3716,
+      "step": 4603
+    },
+    {
+      "epoch": 0.19385263157894736,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00046216883718118825,
+      "loss": 3.4757,
+      "step": 4604
+    },
+    {
+      "epoch": 0.19389473684210526,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004621509874210977,
+      "loss": 3.5335,
+      "step": 4605
+    },
+    {
+      "epoch": 0.19393684210526316,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004621331337958656,
+      "loss": 3.3825,
+      "step": 4606
+    },
+    {
+      "epoch": 0.19397894736842106,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00046211527630581696,
+      "loss": 2.9603,
+      "step": 4607
+    },
+    {
+      "epoch": 0.19402105263157896,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00046209741495127724,
+      "loss": 3.5945,
+      "step": 4608
+    },
+    {
+      "epoch": 0.19406315789473685,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00046207954973257175,
+      "loss": 3.1631,
+      "step": 4609
+    },
+    {
+      "epoch": 0.19410526315789473,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00046206168065002613,
+      "loss": 2.9896,
+      "step": 4610
+    },
+    {
+      "epoch": 0.19414736842105262,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0004620438077039658,
+      "loss": 3.2623,
+      "step": 4611
+    },
+    {
+      "epoch": 0.19418947368421052,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004620259308947165,
+      "loss": 3.1442,
+      "step": 4612
+    },
+    {
+      "epoch": 0.19423157894736842,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00046200805022260377,
+      "loss": 3.0004,
+      "step": 4613
+    },
+    {
+      "epoch": 0.19427368421052632,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004619901656879535,
+      "loss": 3.3269,
+      "step": 4614
+    },
+    {
+      "epoch": 0.19431578947368422,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00046197227729109146,
+      "loss": 3.2997,
+      "step": 4615
+    },
+    {
+      "epoch": 0.19435789473684212,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00046195438503234367,
+      "loss": 3.3216,
+      "step": 4616
+    },
+    {
+      "epoch": 0.1944,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.000461936488912036,
+      "loss": 3.2208,
+      "step": 4617
+    },
+    {
+      "epoch": 0.1944421052631579,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0004619185889304945,
+      "loss": 3.4379,
+      "step": 4618
+    },
+    {
+      "epoch": 0.1944842105263158,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004619006850880453,
+      "loss": 3.0507,
+      "step": 4619
+    },
+    {
+      "epoch": 0.19452631578947369,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00046188277738501454,
+      "loss": 3.3616,
+      "step": 4620
+    },
+    {
+      "epoch": 0.19456842105263158,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0004618648658217286,
+      "loss": 3.1964,
+      "step": 4621
+    },
+    {
+      "epoch": 0.19461052631578948,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004618469503985137,
+      "loss": 3.2916,
+      "step": 4622
+    },
+    {
+      "epoch": 0.19465263157894735,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004618290311156963,
+      "loss": 3.4239,
+      "step": 4623
+    },
+    {
+      "epoch": 0.19469473684210525,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00046181110797360284,
+      "loss": 3.3782,
+      "step": 4624
+    },
+    {
+      "epoch": 0.19473684210526315,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00046179318097255986,
+      "loss": 3.4291,
+      "step": 4625
+    },
+    {
+      "epoch": 0.19477894736842105,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00046177525011289387,
+      "loss": 3.6038,
+      "step": 4626
+    },
+    {
+      "epoch": 0.19482105263157895,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0004617573153949318,
+      "loss": 3.7854,
+      "step": 4627
+    },
+    {
+      "epoch": 0.19486315789473685,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004617393768190001,
+      "loss": 3.4795,
+      "step": 4628
+    },
+    {
+      "epoch": 0.19490526315789475,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00046172143438542586,
+      "loss": 3.0325,
+      "step": 4629
+    },
+    {
+      "epoch": 0.19494736842105262,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004617034880945358,
+      "loss": 2.9884,
+      "step": 4630
+    },
+    {
+      "epoch": 0.19498947368421052,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00046168553794665693,
+      "loss": 3.435,
+      "step": 4631
+    },
+    {
+      "epoch": 0.19503157894736842,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00046166758394211626,
+      "loss": 3.5957,
+      "step": 4632
+    },
+    {
+      "epoch": 0.19507368421052632,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00046164962608124095,
+      "loss": 3.0424,
+      "step": 4633
+    },
+    {
+      "epoch": 0.19511578947368421,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004616316643643582,
+      "loss": 3.3512,
+      "step": 4634
+    },
+    {
+      "epoch": 0.1951578947368421,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00046161369879179503,
+      "loss": 3.4092,
+      "step": 4635
+    },
+    {
+      "epoch": 0.1952,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00046159572936387895,
+      "loss": 3.3861,
+      "step": 4636
+    },
+    {
+      "epoch": 0.19524210526315788,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00046157775608093735,
+      "loss": 3.3005,
+      "step": 4637
+    },
+    {
+      "epoch": 0.19528421052631578,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0004615597789432976,
+      "loss": 3.4246,
+      "step": 4638
+    },
+    {
+      "epoch": 0.19532631578947368,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00046154179795128723,
+      "loss": 3.3376,
+      "step": 4639
+    },
+    {
+      "epoch": 0.19536842105263158,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00046152381310523384,
+      "loss": 3.1749,
+      "step": 4640
+    },
+    {
+      "epoch": 0.19541052631578948,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004615058244054652,
+      "loss": 3.6176,
+      "step": 4641
+    },
+    {
+      "epoch": 0.19545263157894738,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0004614878318523088,
+      "loss": 3.3055,
+      "step": 4642
+    },
+    {
+      "epoch": 0.19549473684210528,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00046146983544609273,
+      "loss": 3.6996,
+      "step": 4643
+    },
+    {
+      "epoch": 0.19553684210526315,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00046145183518714463,
+      "loss": 2.9007,
+      "step": 4644
+    },
+    {
+      "epoch": 0.19557894736842105,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00046143383107579263,
+      "loss": 3.1735,
+      "step": 4645
+    },
+    {
+      "epoch": 0.19562105263157895,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0004614158231123646,
+      "loss": 3.333,
+      "step": 4646
+    },
+    {
+      "epoch": 0.19566315789473684,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00046139781129718866,
+      "loss": 3.3323,
+      "step": 4647
+    },
+    {
+      "epoch": 0.19570526315789474,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.000461379795630593,
+      "loss": 3.4821,
+      "step": 4648
+    },
+    {
+      "epoch": 0.19574736842105264,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004613617761129058,
+      "loss": 3.6953,
+      "step": 4649
+    },
+    {
+      "epoch": 0.1957894736842105,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004613437527444554,
+      "loss": 3.6383,
+      "step": 4650
+    },
+    {
+      "epoch": 0.1958315789473684,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0004613257255255702,
+      "loss": 3.2791,
+      "step": 4651
+    },
+    {
+      "epoch": 0.1958736842105263,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0004613076944565785,
+      "loss": 3.3195,
+      "step": 4652
+    },
+    {
+      "epoch": 0.1959157894736842,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004612896595378089,
+      "loss": 3.3579,
+      "step": 4653
+    },
+    {
+      "epoch": 0.1959578947368421,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00046127162076958994,
+      "loss": 3.3907,
+      "step": 4654
+    },
+    {
+      "epoch": 0.196,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004612535781522503,
+      "loss": 3.6124,
+      "step": 4655
+    },
+    {
+      "epoch": 0.1960421052631579,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004612355316861187,
+      "loss": 3.6001,
+      "step": 4656
+    },
+    {
+      "epoch": 0.19608421052631578,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00046121748137152384,
+      "loss": 3.1411,
+      "step": 4657
+    },
+    {
+      "epoch": 0.19612631578947368,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00046119942720879473,
+      "loss": 3.6213,
+      "step": 4658
+    },
+    {
+      "epoch": 0.19616842105263158,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00046118136919826014,
+      "loss": 3.0817,
+      "step": 4659
+    },
+    {
+      "epoch": 0.19621052631578947,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004611633073402491,
+      "loss": 3.4891,
+      "step": 4660
+    },
+    {
+      "epoch": 0.19625263157894737,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0004611452416350909,
+      "loss": 3.4957,
+      "step": 4661
+    },
+    {
+      "epoch": 0.19629473684210527,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004611271720831143,
+      "loss": 3.6496,
+      "step": 4662
+    },
+    {
+      "epoch": 0.19633684210526317,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0004611090986846487,
+      "loss": 3.3819,
+      "step": 4663
+    },
+    {
+      "epoch": 0.19637894736842104,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004610910214400235,
+      "loss": 3.478,
+      "step": 4664
+    },
+    {
+      "epoch": 0.19642105263157894,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0004610729403495677,
+      "loss": 3.3809,
+      "step": 4665
+    },
+    {
+      "epoch": 0.19646315789473684,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.000461054855413611,
+      "loss": 3.405,
+      "step": 4666
+    },
+    {
+      "epoch": 0.19650526315789474,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00046103676663248293,
+      "loss": 3.1792,
+      "step": 4667
+    },
+    {
+      "epoch": 0.19654736842105264,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00046101867400651287,
+      "loss": 3.8562,
+      "step": 4668
+    },
+    {
+      "epoch": 0.19658947368421054,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00046100057753603043,
+      "loss": 3.3059,
+      "step": 4669
+    },
+    {
+      "epoch": 0.1966315789473684,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.0004609824772213654,
+      "loss": 3.5399,
+      "step": 4670
+    },
+    {
+      "epoch": 0.1966736842105263,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00046096437306284766,
+      "loss": 3.4827,
+      "step": 4671
+    },
+    {
+      "epoch": 0.1967157894736842,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00046094626506080685,
+      "loss": 3.186,
+      "step": 4672
+    },
+    {
+      "epoch": 0.1967578947368421,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00046092815321557287,
+      "loss": 3.6666,
+      "step": 4673
+    },
+    {
+      "epoch": 0.1968,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00046091003752747584,
+      "loss": 3.3509,
+      "step": 4674
+    },
+    {
+      "epoch": 0.1968421052631579,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004608919179968457,
+      "loss": 3.2043,
+      "step": 4675
+    },
+    {
+      "epoch": 0.1968842105263158,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004608737946240126,
+      "loss": 3.1292,
+      "step": 4676
+    },
+    {
+      "epoch": 0.19692631578947367,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0004608556674093067,
+      "loss": 2.6994,
+      "step": 4677
+    },
+    {
+      "epoch": 0.19696842105263157,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00046083753635305834,
+      "loss": 3.4107,
+      "step": 4678
+    },
+    {
+      "epoch": 0.19701052631578947,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00046081940145559783,
+      "loss": 3.6181,
+      "step": 4679
+    },
+    {
+      "epoch": 0.19705263157894737,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00046080126271725544,
+      "loss": 3.1048,
+      "step": 4680
+    },
+    {
+      "epoch": 0.19709473684210527,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00046078312013836175,
+      "loss": 3.0991,
+      "step": 4681
+    },
+    {
+      "epoch": 0.19713684210526317,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00046076497371924733,
+      "loss": 3.2433,
+      "step": 4682
+    },
+    {
+      "epoch": 0.19717894736842106,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00046074682346024264,
+      "loss": 2.6989,
+      "step": 4683
+    },
+    {
+      "epoch": 0.19722105263157894,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00046072866936167854,
+      "loss": 3.5098,
+      "step": 4684
+    },
+    {
+      "epoch": 0.19726315789473683,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004607105114238855,
+      "loss": 3.2954,
+      "step": 4685
+    },
+    {
+      "epoch": 0.19730526315789473,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004606923496471947,
+      "loss": 3.4342,
+      "step": 4686
+    },
+    {
+      "epoch": 0.19734736842105263,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004606741840319368,
+      "loss": 3.8941,
+      "step": 4687
+    },
+    {
+      "epoch": 0.19738947368421053,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00046065601457844277,
+      "loss": 3.826,
+      "step": 4688
+    },
+    {
+      "epoch": 0.19743157894736843,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004606378412870437,
+      "loss": 3.5838,
+      "step": 4689
+    },
+    {
+      "epoch": 0.19747368421052633,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004606196641580706,
+      "loss": 3.4745,
+      "step": 4690
+    },
+    {
+      "epoch": 0.1975157894736842,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00046060148319185474,
+      "loss": 2.8536,
+      "step": 4691
+    },
+    {
+      "epoch": 0.1975578947368421,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004605832983887273,
+      "loss": 3.1668,
+      "step": 4692
+    },
+    {
+      "epoch": 0.1976,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00046056510974901957,
+      "loss": 3.4817,
+      "step": 4693
+    },
+    {
+      "epoch": 0.1976421052631579,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004605469172730629,
+      "loss": 3.6233,
+      "step": 4694
+    },
+    {
+      "epoch": 0.1976842105263158,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004605287209611888,
+      "loss": 3.3508,
+      "step": 4695
+    },
+    {
+      "epoch": 0.1977263157894737,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0004605105208137288,
+      "loss": 3.308,
+      "step": 4696
+    },
+    {
+      "epoch": 0.19776842105263157,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00046049231683101445,
+      "loss": 2.9944,
+      "step": 4697
+    },
+    {
+      "epoch": 0.19781052631578946,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00046047410901337737,
+      "loss": 3.5698,
+      "step": 4698
+    },
+    {
+      "epoch": 0.19785263157894736,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00046045589736114933,
+      "loss": 3.5788,
+      "step": 4699
+    },
+    {
+      "epoch": 0.19789473684210526,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00046043768187466216,
+      "loss": 3.4954,
+      "step": 4700
+    },
+    {
+      "epoch": 0.19793684210526316,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00046041946255424767,
+      "loss": 3.1578,
+      "step": 4701
+    },
+    {
+      "epoch": 0.19797894736842106,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00046040123940023774,
+      "loss": 3.0436,
+      "step": 4702
+    },
+    {
+      "epoch": 0.19802105263157896,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004603830124129645,
+      "loss": 3.1099,
+      "step": 4703
+    },
+    {
+      "epoch": 0.19806315789473683,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00046036478159275997,
+      "loss": 3.659,
+      "step": 4704
+    },
+    {
+      "epoch": 0.19810526315789473,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00046034654693995626,
+      "loss": 3.3704,
+      "step": 4705
+    },
+    {
+      "epoch": 0.19814736842105263,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004603283084548856,
+      "loss": 3.6293,
+      "step": 4706
+    },
+    {
+      "epoch": 0.19818947368421053,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004603100661378803,
+      "loss": 3.2193,
+      "step": 4707
+    },
+    {
+      "epoch": 0.19823157894736843,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004602918199892727,
+      "loss": 3.2627,
+      "step": 4708
+    },
+    {
+      "epoch": 0.19827368421052632,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00046027357000939524,
+      "loss": 3.2634,
+      "step": 4709
+    },
+    {
+      "epoch": 0.19831578947368422,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00046025531619858036,
+      "loss": 3.4003,
+      "step": 4710
+    },
+    {
+      "epoch": 0.1983578947368421,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004602370585571607,
+      "loss": 3.6085,
+      "step": 4711
+    },
+    {
+      "epoch": 0.1984,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00046021879708546884,
+      "loss": 2.9857,
+      "step": 4712
+    },
+    {
+      "epoch": 0.1984421052631579,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00046020053178383747,
+      "loss": 3.7464,
+      "step": 4713
+    },
+    {
+      "epoch": 0.1984842105263158,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004601822626525994,
+      "loss": 3.5108,
+      "step": 4714
+    },
+    {
+      "epoch": 0.1985263157894737,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004601639896920874,
+      "loss": 3.222,
+      "step": 4715
+    },
+    {
+      "epoch": 0.1985684210526316,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004601457129026345,
+      "loss": 2.9223,
+      "step": 4716
+    },
+    {
+      "epoch": 0.1986105263157895,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00046012743228457365,
+      "loss": 3.4981,
+      "step": 4717
+    },
+    {
+      "epoch": 0.19865263157894736,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00046010914783823785,
+      "loss": 3.2794,
+      "step": 4718
+    },
+    {
+      "epoch": 0.19869473684210526,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00046009085956396023,
+      "loss": 3.3854,
+      "step": 4719
+    },
+    {
+      "epoch": 0.19873684210526316,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.000460072567462074,
+      "loss": 3.7012,
+      "step": 4720
+    },
+    {
+      "epoch": 0.19877894736842106,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00046005427153291247,
+      "loss": 3.6206,
+      "step": 4721
+    },
+    {
+      "epoch": 0.19882105263157895,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00046003597177680884,
+      "loss": 3.8439,
+      "step": 4722
+    },
+    {
+      "epoch": 0.19886315789473685,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00046001766819409664,
+      "loss": 3.293,
+      "step": 4723
+    },
+    {
+      "epoch": 0.19890526315789472,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004599993607851092,
+      "loss": 3.616,
+      "step": 4724
+    },
+    {
+      "epoch": 0.19894736842105262,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00045998104955018024,
+      "loss": 3.369,
+      "step": 4725
+    },
+    {
+      "epoch": 0.19898947368421052,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.0004599627344896432,
+      "loss": 3.7414,
+      "step": 4726
+    },
+    {
+      "epoch": 0.19903157894736842,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004599444156038319,
+      "loss": 3.4799,
+      "step": 4727
+    },
+    {
+      "epoch": 0.19907368421052632,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00045992609289308,
+      "loss": 2.8465,
+      "step": 4728
+    },
+    {
+      "epoch": 0.19911578947368422,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00045990776635772133,
+      "loss": 3.36,
+      "step": 4729
+    },
+    {
+      "epoch": 0.19915789473684212,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00045988943599808974,
+      "loss": 3.2382,
+      "step": 4730
+    },
+    {
+      "epoch": 0.1992,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004598711018145193,
+      "loss": 3.2816,
+      "step": 4731
+    },
+    {
+      "epoch": 0.1992421052631579,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.000459852763807344,
+      "loss": 3.1602,
+      "step": 4732
+    },
+    {
+      "epoch": 0.19928421052631579,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004598344219768978,
+      "loss": 3.5339,
+      "step": 4733
+    },
+    {
+      "epoch": 0.19932631578947368,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004598160763235151,
+      "loss": 3.5453,
+      "step": 4734
+    },
+    {
+      "epoch": 0.19936842105263158,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00045979772684753,
+      "loss": 3.3341,
+      "step": 4735
+    },
+    {
+      "epoch": 0.19941052631578948,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00045977937354927675,
+      "loss": 3.141,
+      "step": 4736
+    },
+    {
+      "epoch": 0.19945263157894738,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00045976101642908985,
+      "loss": 3.5318,
+      "step": 4737
+    },
+    {
+      "epoch": 0.19949473684210525,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00045974265548730367,
+      "loss": 3.2621,
+      "step": 4738
+    },
+    {
+      "epoch": 0.19953684210526315,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00045972429072425273,
+      "loss": 3.3693,
+      "step": 4739
+    },
+    {
+      "epoch": 0.19957894736842105,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004597059221402716,
+      "loss": 3.5329,
+      "step": 4740
+    },
+    {
+      "epoch": 0.19962105263157895,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.000459687549735695,
+      "loss": 3.6713,
+      "step": 4741
+    },
+    {
+      "epoch": 0.19966315789473685,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00045966917351085765,
+      "loss": 3.366,
+      "step": 4742
+    },
+    {
+      "epoch": 0.19970526315789475,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004596507934660943,
+      "loss": 3.6754,
+      "step": 4743
+    },
+    {
+      "epoch": 0.19974736842105262,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0004596324096017398,
+      "loss": 3.3526,
+      "step": 4744
+    },
+    {
+      "epoch": 0.19978947368421052,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00045961402191812905,
+      "loss": 3.4811,
+      "step": 4745
+    },
+    {
+      "epoch": 0.19983157894736842,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00045959563041559715,
+      "loss": 3.3433,
+      "step": 4746
+    },
+    {
+      "epoch": 0.19987368421052631,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00045957723509447917,
+      "loss": 3.3926,
+      "step": 4747
+    },
+    {
+      "epoch": 0.1999157894736842,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00045955883595511014,
+      "loss": 3.8522,
+      "step": 4748
+    },
+    {
+      "epoch": 0.1999578947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00045954043299782546,
+      "loss": 3.4289,
+      "step": 4749
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00045952202622296013,
+      "loss": 3.3673,
+      "step": 4750
+    },
+    {
+      "epoch": 0.20004210526315788,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004595036156308498,
+      "loss": 3.0865,
+      "step": 4751
+    },
+    {
+      "epoch": 0.20008421052631578,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004594852012218297,
+      "loss": 3.1619,
+      "step": 4752
+    },
+    {
+      "epoch": 0.20012631578947368,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00045946678299623535,
+      "loss": 3.3714,
+      "step": 4753
+    },
+    {
+      "epoch": 0.20016842105263158,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004594483609544023,
+      "loss": 3.5813,
+      "step": 4754
+    },
+    {
+      "epoch": 0.20021052631578948,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0004594299350966663,
+      "loss": 3.5254,
+      "step": 4755
+    },
+    {
+      "epoch": 0.20025263157894738,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00045941150542336284,
+      "loss": 3.4176,
+      "step": 4756
+    },
+    {
+      "epoch": 0.20029473684210528,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004593930719348279,
+      "loss": 3.0222,
+      "step": 4757
+    },
+    {
+      "epoch": 0.20033684210526315,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00045937463463139715,
+      "loss": 3.5983,
+      "step": 4758
+    },
+    {
+      "epoch": 0.20037894736842105,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00045935619351340664,
+      "loss": 3.2003,
+      "step": 4759
+    },
+    {
+      "epoch": 0.20042105263157894,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00045933774858119213,
+      "loss": 3.3637,
+      "step": 4760
+    },
+    {
+      "epoch": 0.20046315789473684,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0004593192998350899,
+      "loss": 3.5309,
+      "step": 4761
+    },
+    {
+      "epoch": 0.20050526315789474,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00045930084727543595,
+      "loss": 3.4571,
+      "step": 4762
+    },
+    {
+      "epoch": 0.20054736842105264,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00045928239090256654,
+      "loss": 3.2884,
+      "step": 4763
+    },
+    {
+      "epoch": 0.20058947368421054,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00045926393071681775,
+      "loss": 3.0831,
+      "step": 4764
+    },
+    {
+      "epoch": 0.2006315789473684,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004592454667185261,
+      "loss": 3.57,
+      "step": 4765
+    },
+    {
+      "epoch": 0.2006736842105263,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004592269989080279,
+      "loss": 3.3712,
+      "step": 4766
+    },
+    {
+      "epoch": 0.2007157894736842,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00045920852728565954,
+      "loss": 3.4081,
+      "step": 4767
+    },
+    {
+      "epoch": 0.2007578947368421,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00045919005185175764,
+      "loss": 3.5499,
+      "step": 4768
+    },
+    {
+      "epoch": 0.2008,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0004591715726066589,
+      "loss": 2.6562,
+      "step": 4769
+    },
+    {
+      "epoch": 0.2008421052631579,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00045915308955069966,
+      "loss": 3.3569,
+      "step": 4770
+    },
+    {
+      "epoch": 0.20088421052631578,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.000459134602684217,
+      "loss": 3.0852,
+      "step": 4771
+    },
+    {
+      "epoch": 0.20092631578947368,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00045911611200754763,
+      "loss": 3.257,
+      "step": 4772
+    },
+    {
+      "epoch": 0.20096842105263157,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004590976175210284,
+      "loss": 2.9791,
+      "step": 4773
+    },
+    {
+      "epoch": 0.20101052631578947,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00045907911922499624,
+      "loss": 3.4335,
+      "step": 4774
+    },
+    {
+      "epoch": 0.20105263157894737,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00045906061711978817,
+      "loss": 3.0195,
+      "step": 4775
+    },
+    {
+      "epoch": 0.20109473684210527,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004590421112057413,
+      "loss": 3.5424,
+      "step": 4776
+    },
+    {
+      "epoch": 0.20113684210526317,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00045902360148319276,
+      "loss": 3.0324,
+      "step": 4777
+    },
+    {
+      "epoch": 0.20117894736842104,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004590050879524798,
+      "loss": 3.0185,
+      "step": 4778
+    },
+    {
+      "epoch": 0.20122105263157894,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004589865706139397,
+      "loss": 3.5402,
+      "step": 4779
+    },
+    {
+      "epoch": 0.20126315789473684,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004589680494679099,
+      "loss": 3.4977,
+      "step": 4780
+    },
+    {
+      "epoch": 0.20130526315789474,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0004589495245147277,
+      "loss": 3.7023,
+      "step": 4781
+    },
+    {
+      "epoch": 0.20134736842105264,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004589309957547307,
+      "loss": 3.3714,
+      "step": 4782
+    },
+    {
+      "epoch": 0.20138947368421053,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0004589124631882564,
+      "loss": 3.2131,
+      "step": 4783
+    },
+    {
+      "epoch": 0.20143157894736843,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0004588939268156425,
+      "loss": 3.7069,
+      "step": 4784
+    },
+    {
+      "epoch": 0.2014736842105263,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004588753866372267,
+      "loss": 3.8827,
+      "step": 4785
+    },
+    {
+      "epoch": 0.2015157894736842,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00045885684265334684,
+      "loss": 3.6997,
+      "step": 4786
+    },
+    {
+      "epoch": 0.2015578947368421,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004588382948643406,
+      "loss": 3.2795,
+      "step": 4787
+    },
+    {
+      "epoch": 0.2016,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00045881974327054604,
+      "loss": 3.3301,
+      "step": 4788
+    },
+    {
+      "epoch": 0.2016421052631579,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00045880118787230105,
+      "loss": 3.3275,
+      "step": 4789
+    },
+    {
+      "epoch": 0.2016842105263158,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0004587826286699439,
+      "loss": 2.9759,
+      "step": 4790
+    },
+    {
+      "epoch": 0.2017263157894737,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00045876406566381246,
+      "loss": 3.4448,
+      "step": 4791
+    },
+    {
+      "epoch": 0.20176842105263157,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.000458745498854245,
+      "loss": 3.567,
+      "step": 4792
+    },
+    {
+      "epoch": 0.20181052631578947,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0004587269282415799,
+      "loss": 3.2422,
+      "step": 4793
+    },
+    {
+      "epoch": 0.20185263157894737,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004587083538261554,
+      "loss": 3.1943,
+      "step": 4794
+    },
+    {
+      "epoch": 0.20189473684210527,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004586897756083098,
+      "loss": 3.2912,
+      "step": 4795
+    },
+    {
+      "epoch": 0.20193684210526316,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00045867119358838183,
+      "loss": 3.5231,
+      "step": 4796
+    },
+    {
+      "epoch": 0.20197894736842106,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00045865260776670983,
+      "loss": 3.1257,
+      "step": 4797
+    },
+    {
+      "epoch": 0.20202105263157893,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00045863401814363246,
+      "loss": 3.237,
+      "step": 4798
+    },
+    {
+      "epoch": 0.20206315789473683,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004586154247194885,
+      "loss": 3.3493,
+      "step": 4799
+    },
+    {
+      "epoch": 0.20210526315789473,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00045859682749461653,
+      "loss": 3.3089,
+      "step": 4800
+    },
+    {
+      "epoch": 0.20214736842105263,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004585782264693555,
+      "loss": 3.1251,
+      "step": 4801
+    },
+    {
+      "epoch": 0.20218947368421053,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00045855962164404417,
+      "loss": 3.708,
+      "step": 4802
+    },
+    {
+      "epoch": 0.20223157894736843,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004585410130190217,
+      "loss": 3.7962,
+      "step": 4803
+    },
+    {
+      "epoch": 0.20227368421052633,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004585224005946269,
+      "loss": 3.7885,
+      "step": 4804
+    },
+    {
+      "epoch": 0.2023157894736842,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.000458503784371199,
+      "loss": 3.2529,
+      "step": 4805
+    },
+    {
+      "epoch": 0.2023578947368421,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00045848516434907717,
+      "loss": 3.2713,
+      "step": 4806
+    },
+    {
+      "epoch": 0.2024,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0004584665405286006,
+      "loss": 3.2695,
+      "step": 4807
+    },
+    {
+      "epoch": 0.2024421052631579,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004584479129101085,
+      "loss": 3.056,
+      "step": 4808
+    },
+    {
+      "epoch": 0.2024842105263158,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00045842928149394046,
+      "loss": 3.1304,
+      "step": 4809
+    },
+    {
+      "epoch": 0.2025263157894737,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004584106462804357,
+      "loss": 3.2922,
+      "step": 4810
+    },
+    {
+      "epoch": 0.2025684210526316,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004583920072699339,
+      "loss": 3.6885,
+      "step": 4811
+    },
+    {
+      "epoch": 0.20261052631578946,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004583733644627746,
+      "loss": 3.5408,
+      "step": 4812
+    },
+    {
+      "epoch": 0.20265263157894736,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004583547178592974,
+      "loss": 2.9682,
+      "step": 4813
+    },
+    {
+      "epoch": 0.20269473684210526,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.000458336067459842,
+      "loss": 3.1798,
+      "step": 4814
+    },
+    {
+      "epoch": 0.20273684210526316,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00045831741326474835,
+      "loss": 3.387,
+      "step": 4815
+    },
+    {
+      "epoch": 0.20277894736842106,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004582987552743562,
+      "loss": 3.6772,
+      "step": 4816
+    },
+    {
+      "epoch": 0.20282105263157896,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00045828009348900537,
+      "loss": 3.3466,
+      "step": 4817
+    },
+    {
+      "epoch": 0.20286315789473683,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004582614279090359,
+      "loss": 3.4067,
+      "step": 4818
+    },
+    {
+      "epoch": 0.20290526315789473,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.000458242758534788,
+      "loss": 3.0989,
+      "step": 4819
+    },
+    {
+      "epoch": 0.20294736842105263,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004582240853666018,
+      "loss": 3.3641,
+      "step": 4820
+    },
+    {
+      "epoch": 0.20298947368421053,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004582054084048173,
+      "loss": 2.9067,
+      "step": 4821
+    },
+    {
+      "epoch": 0.20303157894736842,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004581867276497749,
+      "loss": 3.5387,
+      "step": 4822
+    },
+    {
+      "epoch": 0.20307368421052632,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00045816804310181493,
+      "loss": 3.6092,
+      "step": 4823
+    },
+    {
+      "epoch": 0.20311578947368422,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00045814935476127784,
+      "loss": 3.6426,
+      "step": 4824
+    },
+    {
+      "epoch": 0.2031578947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.000458130662628504,
+      "loss": 3.531,
+      "step": 4825
+    },
+    {
+      "epoch": 0.2032,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00045811196670383415,
+      "loss": 3.583,
+      "step": 4826
+    },
+    {
+      "epoch": 0.2032421052631579,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004580932669876087,
+      "loss": 3.159,
+      "step": 4827
+    },
+    {
+      "epoch": 0.2032842105263158,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0004580745634801684,
+      "loss": 3.3291,
+      "step": 4828
+    },
+    {
+      "epoch": 0.2033263157894737,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00045805585618185406,
+      "loss": 3.4634,
+      "step": 4829
+    },
+    {
+      "epoch": 0.2033684210526316,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0004580371450930065,
+      "loss": 3.1244,
+      "step": 4830
+    },
+    {
+      "epoch": 0.20341052631578949,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004580184302139666,
+      "loss": 3.0725,
+      "step": 4831
+    },
+    {
+      "epoch": 0.20345263157894736,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00045799971154507536,
+      "loss": 3.7631,
+      "step": 4832
+    },
+    {
+      "epoch": 0.20349473684210526,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00045798098908667365,
+      "loss": 3.0916,
+      "step": 4833
+    },
+    {
+      "epoch": 0.20353684210526315,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00045796226283910277,
+      "loss": 3.3006,
+      "step": 4834
+    },
+    {
+      "epoch": 0.20357894736842105,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00045794353280270375,
+      "loss": 3.1864,
+      "step": 4835
+    },
+    {
+      "epoch": 0.20362105263157895,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004579247989778179,
+      "loss": 3.3763,
+      "step": 4836
+    },
+    {
+      "epoch": 0.20366315789473685,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004579060613647865,
+      "loss": 3.3983,
+      "step": 4837
+    },
+    {
+      "epoch": 0.20370526315789475,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.000457887319963951,
+      "loss": 3.445,
+      "step": 4838
+    },
+    {
+      "epoch": 0.20374736842105262,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004578685747756528,
+      "loss": 3.4286,
+      "step": 4839
+    },
+    {
+      "epoch": 0.20378947368421052,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00045784982580023334,
+      "loss": 3.1014,
+      "step": 4840
+    },
+    {
+      "epoch": 0.20383157894736842,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004578310730380343,
+      "loss": 3.7364,
+      "step": 4841
+    },
+    {
+      "epoch": 0.20387368421052632,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00045781231648939723,
+      "loss": 3.2253,
+      "step": 4842
+    },
+    {
+      "epoch": 0.20391578947368422,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.000457793556154664,
+      "loss": 3.611,
+      "step": 4843
+    },
+    {
+      "epoch": 0.20395789473684212,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004577747920341763,
+      "loss": 3.1724,
+      "step": 4844
+    },
+    {
+      "epoch": 0.204,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.000457756024128276,
+      "loss": 3.1954,
+      "step": 4845
+    },
+    {
+      "epoch": 0.20404210526315789,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00045773725243730505,
+      "loss": 3.2098,
+      "step": 4846
+    },
+    {
+      "epoch": 0.20408421052631578,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0004577184769616055,
+      "loss": 3.0881,
+      "step": 4847
+    },
+    {
+      "epoch": 0.20412631578947368,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00045769969770151933,
+      "loss": 3.1638,
+      "step": 4848
+    },
+    {
+      "epoch": 0.20416842105263158,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00045768091465738864,
+      "loss": 3.2756,
+      "step": 4849
+    },
+    {
+      "epoch": 0.20421052631578948,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00045766212782955576,
+      "loss": 3.456,
+      "step": 4850
+    },
+    {
+      "epoch": 0.20425263157894738,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00045764333721836286,
+      "loss": 3.23,
+      "step": 4851
+    },
+    {
+      "epoch": 0.20429473684210525,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004576245428241524,
+      "loss": 3.3233,
+      "step": 4852
+    },
+    {
+      "epoch": 0.20433684210526315,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0004576057446472667,
+      "loss": 2.8044,
+      "step": 4853
+    },
+    {
+      "epoch": 0.20437894736842105,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004575869426880482,
+      "loss": 3.1849,
+      "step": 4854
+    },
+    {
+      "epoch": 0.20442105263157895,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004575681369468395,
+      "loss": 3.3302,
+      "step": 4855
+    },
+    {
+      "epoch": 0.20446315789473685,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004575493274239833,
+      "loss": 3.3471,
+      "step": 4856
+    },
+    {
+      "epoch": 0.20450526315789475,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0004575305141198222,
+      "loss": 3.2048,
+      "step": 4857
+    },
+    {
+      "epoch": 0.20454736842105264,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00045751169703469897,
+      "loss": 3.5034,
+      "step": 4858
+    },
+    {
+      "epoch": 0.20458947368421052,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004574928761689564,
+      "loss": 3.8968,
+      "step": 4859
+    },
+    {
+      "epoch": 0.20463157894736841,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004574740515229374,
+      "loss": 3.3529,
+      "step": 4860
+    },
+    {
+      "epoch": 0.2046736842105263,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00045745522309698506,
+      "loss": 3.7565,
+      "step": 4861
+    },
+    {
+      "epoch": 0.2047157894736842,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004574363908914422,
+      "loss": 3.8542,
+      "step": 4862
+    },
+    {
+      "epoch": 0.2047578947368421,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00045741755490665204,
+      "loss": 3.021,
+      "step": 4863
+    },
+    {
+      "epoch": 0.2048,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00045739871514295785,
+      "loss": 2.9405,
+      "step": 4864
+    },
+    {
+      "epoch": 0.20484210526315788,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004573798716007026,
+      "loss": 3.1506,
+      "step": 4865
+    },
+    {
+      "epoch": 0.20488421052631578,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00045736102428022983,
+      "loss": 3.4479,
+      "step": 4866
+    },
+    {
+      "epoch": 0.20492631578947368,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00045734217318188286,
+      "loss": 3.5769,
+      "step": 4867
+    },
+    {
+      "epoch": 0.20496842105263158,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.000457323318306005,
+      "loss": 3.5091,
+      "step": 4868
+    },
+    {
+      "epoch": 0.20501052631578948,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00045730445965293996,
+      "loss": 3.3422,
+      "step": 4869
+    },
+    {
+      "epoch": 0.20505263157894738,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00045728559722303127,
+      "loss": 2.8421,
+      "step": 4870
+    },
+    {
+      "epoch": 0.20509473684210527,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004572667310166225,
+      "loss": 3.1501,
+      "step": 4871
+    },
+    {
+      "epoch": 0.20513684210526315,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004572478610340574,
+      "loss": 3.6378,
+      "step": 4872
+    },
+    {
+      "epoch": 0.20517894736842104,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00045722898727567984,
+      "loss": 3.2267,
+      "step": 4873
+    },
+    {
+      "epoch": 0.20522105263157894,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00045721010974183356,
+      "loss": 3.6669,
+      "step": 4874
+    },
+    {
+      "epoch": 0.20526315789473684,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00045719122843286254,
+      "loss": 3.2277,
+      "step": 4875
+    },
+    {
+      "epoch": 0.20530526315789474,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00045717234334911085,
+      "loss": 3.5051,
+      "step": 4876
+    },
+    {
+      "epoch": 0.20534736842105264,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004571534544909224,
+      "loss": 3.6637,
+      "step": 4877
+    },
+    {
+      "epoch": 0.20538947368421054,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004571345618586414,
+      "loss": 3.2662,
+      "step": 4878
+    },
+    {
+      "epoch": 0.2054315789473684,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00045711566545261214,
+      "loss": 3.4582,
+      "step": 4879
+    },
+    {
+      "epoch": 0.2054736842105263,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004570967652731788,
+      "loss": 3.3523,
+      "step": 4880
+    },
+    {
+      "epoch": 0.2055157894736842,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00045707786132068563,
+      "loss": 3.1303,
+      "step": 4881
+    },
+    {
+      "epoch": 0.2055578947368421,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004570589535954772,
+      "loss": 3.6112,
+      "step": 4882
+    },
+    {
+      "epoch": 0.2056,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004570400420978979,
+      "loss": 3.0822,
+      "step": 4883
+    },
+    {
+      "epoch": 0.2056421052631579,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00045702112682829234,
+      "loss": 3.4396,
+      "step": 4884
+    },
+    {
+      "epoch": 0.2056842105263158,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.000457002207787005,
+      "loss": 3.0132,
+      "step": 4885
+    },
+    {
+      "epoch": 0.20572631578947367,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00045698328497438073,
+      "loss": 3.7644,
+      "step": 4886
+    },
+    {
+      "epoch": 0.20576842105263157,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004569643583907642,
+      "loss": 3.7482,
+      "step": 4887
+    },
+    {
+      "epoch": 0.20581052631578947,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004569454280365002,
+      "loss": 3.4677,
+      "step": 4888
+    },
+    {
+      "epoch": 0.20585263157894737,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004569264939119336,
+      "loss": 3.2267,
+      "step": 4889
+    },
+    {
+      "epoch": 0.20589473684210527,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004569075560174095,
+      "loss": 3.4683,
+      "step": 4890
+    },
+    {
+      "epoch": 0.20593684210526317,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00045688861435327276,
+      "loss": 2.7343,
+      "step": 4891
+    },
+    {
+      "epoch": 0.20597894736842104,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00045686966891986865,
+      "loss": 3.3173,
+      "step": 4892
+    },
+    {
+      "epoch": 0.20602105263157894,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004568507197175422,
+      "loss": 3.542,
+      "step": 4893
+    },
+    {
+      "epoch": 0.20606315789473684,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0004568317667466386,
+      "loss": 3.409,
+      "step": 4894
+    },
+    {
+      "epoch": 0.20610526315789474,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004568128100075033,
+      "loss": 3.2829,
+      "step": 4895
+    },
+    {
+      "epoch": 0.20614736842105263,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00045679384950048167,
+      "loss": 3.3108,
+      "step": 4896
+    },
+    {
+      "epoch": 0.20618947368421053,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00045677488522591895,
+      "loss": 3.0382,
+      "step": 4897
+    },
+    {
+      "epoch": 0.20623157894736843,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00045675591718416075,
+      "loss": 3.2969,
+      "step": 4898
+    },
+    {
+      "epoch": 0.2062736842105263,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00045673694537555276,
+      "loss": 3.7491,
+      "step": 4899
+    },
+    {
+      "epoch": 0.2063157894736842,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004567179698004405,
+      "loss": 3.5097,
+      "step": 4900
+    },
+    {
+      "epoch": 0.2063578947368421,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004566989904591697,
+      "loss": 3.8193,
+      "step": 4901
+    },
+    {
+      "epoch": 0.2064,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00045668000735208617,
+      "loss": 3.0735,
+      "step": 4902
+    },
+    {
+      "epoch": 0.2064421052631579,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00045666102047953573,
+      "loss": 3.3607,
+      "step": 4903
+    },
+    {
+      "epoch": 0.2064842105263158,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00045664202984186433,
+      "loss": 3.3636,
+      "step": 4904
+    },
+    {
+      "epoch": 0.2065263157894737,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00045662303543941797,
+      "loss": 3.6543,
+      "step": 4905
+    },
+    {
+      "epoch": 0.20656842105263157,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00045660403727254263,
+      "loss": 3.4569,
+      "step": 4906
+    },
+    {
+      "epoch": 0.20661052631578947,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00045658503534158444,
+      "loss": 3.1884,
+      "step": 4907
+    },
+    {
+      "epoch": 0.20665263157894737,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00045656602964688975,
+      "loss": 3.5567,
+      "step": 4908
+    },
+    {
+      "epoch": 0.20669473684210526,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0004565470201888046,
+      "loss": 3.2059,
+      "step": 4909
+    },
+    {
+      "epoch": 0.20673684210526316,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004565280069676755,
+      "loss": 2.9784,
+      "step": 4910
+    },
+    {
+      "epoch": 0.20677894736842106,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004565089899838487,
+      "loss": 3.223,
+      "step": 4911
+    },
+    {
+      "epoch": 0.20682105263157896,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00045648996923767084,
+      "loss": 3.146,
+      "step": 4912
+    },
+    {
+      "epoch": 0.20686315789473683,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00045647094472948827,
+      "loss": 3.3057,
+      "step": 4913
+    },
+    {
+      "epoch": 0.20690526315789473,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00045645191645964767,
+      "loss": 3.2634,
+      "step": 4914
+    },
+    {
+      "epoch": 0.20694736842105263,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004564328844284957,
+      "loss": 3.5713,
+      "step": 4915
+    },
+    {
+      "epoch": 0.20698947368421053,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00045641384863637924,
+      "loss": 3.5453,
+      "step": 4916
+    },
+    {
+      "epoch": 0.20703157894736843,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00045639480908364497,
+      "loss": 3.3009,
+      "step": 4917
+    },
+    {
+      "epoch": 0.20707368421052633,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004563757657706397,
+      "loss": 3.7999,
+      "step": 4918
+    },
+    {
+      "epoch": 0.2071157894736842,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004563567186977105,
+      "loss": 3.432,
+      "step": 4919
+    },
+    {
+      "epoch": 0.2071578947368421,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00045633766786520435,
+      "loss": 3.0901,
+      "step": 4920
+    },
+    {
+      "epoch": 0.2072,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00045631861327346835,
+      "loss": 3.4383,
+      "step": 4921
+    },
+    {
+      "epoch": 0.2072421052631579,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004562995549228496,
+      "loss": 2.9675,
+      "step": 4922
+    },
+    {
+      "epoch": 0.2072842105263158,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00045628049281369533,
+      "loss": 3.1739,
+      "step": 4923
+    },
+    {
+      "epoch": 0.2073263157894737,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00045626142694635285,
+      "loss": 3.3017,
+      "step": 4924
+    },
+    {
+      "epoch": 0.2073684210526316,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004562423573211696,
+      "loss": 3.4198,
+      "step": 4925
+    },
+    {
+      "epoch": 0.20741052631578946,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00045622328393849285,
+      "loss": 3.1396,
+      "step": 4926
+    },
+    {
+      "epoch": 0.20745263157894736,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00045620420679867013,
+      "loss": 3.1957,
+      "step": 4927
+    },
+    {
+      "epoch": 0.20749473684210526,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00045618512590204916,
+      "loss": 3.1129,
+      "step": 4928
+    },
+    {
+      "epoch": 0.20753684210526316,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004561660412489773,
+      "loss": 3.1561,
+      "step": 4929
+    },
+    {
+      "epoch": 0.20757894736842106,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00045614695283980255,
+      "loss": 3.5342,
+      "step": 4930
+    },
+    {
+      "epoch": 0.20762105263157896,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0004561278606748725,
+      "loss": 3.1264,
+      "step": 4931
+    },
+    {
+      "epoch": 0.20766315789473686,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.000456108764754535,
+      "loss": 3.1724,
+      "step": 4932
+    },
+    {
+      "epoch": 0.20770526315789473,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00045608966507913794,
+      "loss": 3.3363,
+      "step": 4933
+    },
+    {
+      "epoch": 0.20774736842105263,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00045607056164902936,
+      "loss": 3.0936,
+      "step": 4934
+    },
+    {
+      "epoch": 0.20778947368421052,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00045605145446455726,
+      "loss": 3.6235,
+      "step": 4935
+    },
+    {
+      "epoch": 0.20783157894736842,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00045603234352606983,
+      "loss": 3.071,
+      "step": 4936
+    },
+    {
+      "epoch": 0.20787368421052632,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00045601322883391505,
+      "loss": 3.4229,
+      "step": 4937
+    },
+    {
+      "epoch": 0.20791578947368422,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004559941103884414,
+      "loss": 3.5181,
+      "step": 4938
+    },
+    {
+      "epoch": 0.2079578947368421,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00045597498818999706,
+      "loss": 3.0271,
+      "step": 4939
+    },
+    {
+      "epoch": 0.208,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0004559558622389304,
+      "loss": 3.0135,
+      "step": 4940
+    },
+    {
+      "epoch": 0.2080421052631579,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00045593673253558996,
+      "loss": 3.7471,
+      "step": 4941
+    },
+    {
+      "epoch": 0.2080842105263158,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004559175990803242,
+      "loss": 3.1358,
+      "step": 4942
+    },
+    {
+      "epoch": 0.2081263157894737,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004558984618734818,
+      "loss": 3.1351,
+      "step": 4943
+    },
+    {
+      "epoch": 0.20816842105263159,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0004558793209154113,
+      "loss": 3.2653,
+      "step": 4944
+    },
+    {
+      "epoch": 0.20821052631578948,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004558601762064614,
+      "loss": 3.4543,
+      "step": 4945
+    },
+    {
+      "epoch": 0.20825263157894736,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00045584102774698105,
+      "loss": 3.485,
+      "step": 4946
+    },
+    {
+      "epoch": 0.20829473684210525,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.000455821875537319,
+      "loss": 3.4329,
+      "step": 4947
+    },
+    {
+      "epoch": 0.20833684210526315,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004558027195778242,
+      "loss": 3.7951,
+      "step": 4948
+    },
+    {
+      "epoch": 0.20837894736842105,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004557835598688457,
+      "loss": 3.5596,
+      "step": 4949
+    },
+    {
+      "epoch": 0.20842105263157895,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00045576439641073247,
+      "loss": 3.4217,
+      "step": 4950
+    },
+    {
+      "epoch": 0.20846315789473685,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004557452292038338,
+      "loss": 3.7025,
+      "step": 4951
+    },
+    {
+      "epoch": 0.20850526315789475,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004557260582484987,
+      "loss": 3.4397,
+      "step": 4952
+    },
+    {
+      "epoch": 0.20854736842105262,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004557068835450765,
+      "loss": 3.3109,
+      "step": 4953
+    },
+    {
+      "epoch": 0.20858947368421052,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00045568770509391666,
+      "loss": 3.5784,
+      "step": 4954
+    },
+    {
+      "epoch": 0.20863157894736842,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00045566852289536845,
+      "loss": 3.5591,
+      "step": 4955
+    },
+    {
+      "epoch": 0.20867368421052632,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00045564933694978147,
+      "loss": 3.0718,
+      "step": 4956
+    },
+    {
+      "epoch": 0.20871578947368422,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004556301472575052,
+      "loss": 3.4293,
+      "step": 4957
+    },
+    {
+      "epoch": 0.20875789473684211,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00045561095381888917,
+      "loss": 3.6398,
+      "step": 4958
+    },
+    {
+      "epoch": 0.2088,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004555917566342832,
+      "loss": 3.1091,
+      "step": 4959
+    },
+    {
+      "epoch": 0.20884210526315788,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.000455572555704037,
+      "loss": 3.0131,
+      "step": 4960
+    },
+    {
+      "epoch": 0.20888421052631578,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004555533510285003,
+      "loss": 3.2339,
+      "step": 4961
+    },
+    {
+      "epoch": 0.20892631578947368,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004555341426080231,
+      "loss": 3.3304,
+      "step": 4962
+    },
+    {
+      "epoch": 0.20896842105263158,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00045551493044295534,
+      "loss": 3.1964,
+      "step": 4963
+    },
+    {
+      "epoch": 0.20901052631578948,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00045549571453364703,
+      "loss": 3.4942,
+      "step": 4964
+    },
+    {
+      "epoch": 0.20905263157894738,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00045547649488044816,
+      "loss": 3.5193,
+      "step": 4965
+    },
+    {
+      "epoch": 0.20909473684210525,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004554572714837091,
+      "loss": 3.3332,
+      "step": 4966
+    },
+    {
+      "epoch": 0.20913684210526315,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00045543804434377985,
+      "loss": 3.3037,
+      "step": 4967
+    },
+    {
+      "epoch": 0.20917894736842105,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00045541881346101087,
+      "loss": 3.0467,
+      "step": 4968
+    },
+    {
+      "epoch": 0.20922105263157895,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00045539957883575244,
+      "loss": 3.2079,
+      "step": 4969
+    },
+    {
+      "epoch": 0.20926315789473685,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00045538034046835503,
+      "loss": 3.2942,
+      "step": 4970
+    },
+    {
+      "epoch": 0.20930526315789474,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004553610983591691,
+      "loss": 3.3587,
+      "step": 4971
+    },
+    {
+      "epoch": 0.20934736842105264,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004553418525085453,
+      "loss": 3.3851,
+      "step": 4972
+    },
+    {
+      "epoch": 0.20938947368421051,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00045532260291683416,
+      "loss": 3.4295,
+      "step": 4973
+    },
+    {
+      "epoch": 0.2094315789473684,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00045530334958438645,
+      "loss": 3.2109,
+      "step": 4974
+    },
+    {
+      "epoch": 0.2094736842105263,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.000455284092511553,
+      "loss": 3.3413,
+      "step": 4975
+    },
+    {
+      "epoch": 0.2095157894736842,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004552648316986845,
+      "loss": 3.4819,
+      "step": 4976
+    },
+    {
+      "epoch": 0.2095578947368421,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.000455245567146132,
+      "loss": 3.3589,
+      "step": 4977
+    },
+    {
+      "epoch": 0.2096,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0004552262988542464,
+      "loss": 3.427,
+      "step": 4978
+    },
+    {
+      "epoch": 0.2096421052631579,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00045520702682337876,
+      "loss": 3.5033,
+      "step": 4979
+    },
+    {
+      "epoch": 0.20968421052631578,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0004551877510538802,
+      "loss": 3.6604,
+      "step": 4980
+    },
+    {
+      "epoch": 0.20972631578947368,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00045516847154610185,
+      "loss": 3.6099,
+      "step": 4981
+    },
+    {
+      "epoch": 0.20976842105263158,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0004551491883003951,
+      "loss": 3.2591,
+      "step": 4982
+    },
+    {
+      "epoch": 0.20981052631578948,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00045512990131711114,
+      "loss": 3.6699,
+      "step": 4983
+    },
+    {
+      "epoch": 0.20985263157894737,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00045511061059660144,
+      "loss": 3.2984,
+      "step": 4984
+    },
+    {
+      "epoch": 0.20989473684210527,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00045509131613921726,
+      "loss": 3.3488,
+      "step": 4985
+    },
+    {
+      "epoch": 0.20993684210526317,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00045507201794531044,
+      "loss": 3.7559,
+      "step": 4986
+    },
+    {
+      "epoch": 0.20997894736842104,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00045505271601523235,
+      "loss": 3.5581,
+      "step": 4987
+    },
+    {
+      "epoch": 0.21002105263157894,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004550334103493347,
+      "loss": 3.7323,
+      "step": 4988
+    },
+    {
+      "epoch": 0.21006315789473684,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00045501410094796917,
+      "loss": 3.2852,
+      "step": 4989
+    },
+    {
+      "epoch": 0.21010526315789474,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00045499478781148766,
+      "loss": 3.7183,
+      "step": 4990
+    },
+    {
+      "epoch": 0.21014736842105264,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.000454975470940242,
+      "loss": 3.9877,
+      "step": 4991
+    },
+    {
+      "epoch": 0.21018947368421054,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00045495615033458404,
+      "loss": 3.6588,
+      "step": 4992
+    },
+    {
+      "epoch": 0.2102315789473684,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00045493682599486584,
+      "loss": 2.7438,
+      "step": 4993
+    },
+    {
+      "epoch": 0.2102736842105263,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004549174979214394,
+      "loss": 3.5225,
+      "step": 4994
+    },
+    {
+      "epoch": 0.2103157894736842,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00045489816611465704,
+      "loss": 3.0684,
+      "step": 4995
+    },
+    {
+      "epoch": 0.2103578947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004548788305748708,
+      "loss": 3.4274,
+      "step": 4996
+    },
+    {
+      "epoch": 0.2104,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.000454859491302433,
+      "loss": 3.198,
+      "step": 4997
+    },
+    {
+      "epoch": 0.2104421052631579,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00045484014829769593,
+      "loss": 3.8203,
+      "step": 4998
+    },
+    {
+      "epoch": 0.2104842105263158,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.000454820801561012,
+      "loss": 3.8871,
+      "step": 4999
+    },
+    {
+      "epoch": 0.21052631578947367,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00045480145109273387,
+      "loss": 3.4842,
+      "step": 5000
+    },
+    {
+      "epoch": 0.21056842105263157,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004547820968932138,
+      "loss": 3.4447,
+      "step": 5001
+    },
+    {
+      "epoch": 0.21061052631578947,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004547627389628045,
+      "loss": 3.35,
+      "step": 5002
+    },
+    {
+      "epoch": 0.21065263157894737,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0004547433773018588,
+      "loss": 3.1837,
+      "step": 5003
+    },
+    {
+      "epoch": 0.21069473684210527,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00045472401191072934,
+      "loss": 3.2302,
+      "step": 5004
+    },
+    {
+      "epoch": 0.21073684210526317,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00045470464278976873,
+      "loss": 3.1655,
+      "step": 5005
+    },
+    {
+      "epoch": 0.21077894736842107,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00045468526993933025,
+      "loss": 3.4476,
+      "step": 5006
+    },
+    {
+      "epoch": 0.21082105263157894,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004546658933597666,
+      "loss": 3.1952,
+      "step": 5007
+    },
+    {
+      "epoch": 0.21086315789473684,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004546465130514308,
+      "loss": 2.9211,
+      "step": 5008
+    },
+    {
+      "epoch": 0.21090526315789473,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00045462712901467593,
+      "loss": 3.5233,
+      "step": 5009
+    },
+    {
+      "epoch": 0.21094736842105263,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004546077412498553,
+      "loss": 3.2,
+      "step": 5010
+    },
+    {
+      "epoch": 0.21098947368421053,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.000454588349757322,
+      "loss": 3.2371,
+      "step": 5011
+    },
+    {
+      "epoch": 0.21103157894736843,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004545689545374294,
+      "loss": 3.3118,
+      "step": 5012
+    },
+    {
+      "epoch": 0.2110736842105263,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00045454955559053067,
+      "loss": 2.7364,
+      "step": 5013
+    },
+    {
+      "epoch": 0.2111157894736842,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0004545301529169795,
+      "loss": 3.2338,
+      "step": 5014
+    },
+    {
+      "epoch": 0.2111578947368421,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004545107465171292,
+      "loss": 3.6475,
+      "step": 5015
+    },
+    {
+      "epoch": 0.2112,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004544913363913334,
+      "loss": 3.2459,
+      "step": 5016
+    },
+    {
+      "epoch": 0.2112421052631579,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004544719225399457,
+      "loss": 3.4978,
+      "step": 5017
+    },
+    {
+      "epoch": 0.2112842105263158,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004544525049633199,
+      "loss": 3.5852,
+      "step": 5018
+    },
+    {
+      "epoch": 0.2113263157894737,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004544330836618096,
+      "loss": 3.4177,
+      "step": 5019
+    },
+    {
+      "epoch": 0.21136842105263157,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00045441365863576873,
+      "loss": 3.4454,
+      "step": 5020
+    },
+    {
+      "epoch": 0.21141052631578947,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004543942298855512,
+      "loss": 3.3222,
+      "step": 5021
+    },
+    {
+      "epoch": 0.21145263157894736,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004543747974115109,
+      "loss": 3.0539,
+      "step": 5022
+    },
+    {
+      "epoch": 0.21149473684210526,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00045435536121400194,
+      "loss": 3.5983,
+      "step": 5023
+    },
+    {
+      "epoch": 0.21153684210526316,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00045433592129337844,
+      "loss": 3.5145,
+      "step": 5024
+    },
+    {
+      "epoch": 0.21157894736842106,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004543164776499945,
+      "loss": 3.4578,
+      "step": 5025
+    },
+    {
+      "epoch": 0.21162105263157896,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00045429703028420444,
+      "loss": 3.4846,
+      "step": 5026
+    },
+    {
+      "epoch": 0.21166315789473683,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004542775791963625,
+      "loss": 3.6145,
+      "step": 5027
+    },
+    {
+      "epoch": 0.21170526315789473,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00045425812438682313,
+      "loss": 3.1894,
+      "step": 5028
+    },
+    {
+      "epoch": 0.21174736842105263,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004542386658559406,
+      "loss": 3.3676,
+      "step": 5029
+    },
+    {
+      "epoch": 0.21178947368421053,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00045421920360406965,
+      "loss": 3.5897,
+      "step": 5030
+    },
+    {
+      "epoch": 0.21183157894736843,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00045419973763156474,
+      "loss": 3.6415,
+      "step": 5031
+    },
+    {
+      "epoch": 0.21187368421052633,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004541802679387806,
+      "loss": 3.4499,
+      "step": 5032
+    },
+    {
+      "epoch": 0.21191578947368422,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004541607945260718,
+      "loss": 3.0771,
+      "step": 5033
+    },
+    {
+      "epoch": 0.2119578947368421,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0004541413173937932,
+      "loss": 3.5545,
+      "step": 5034
+    },
+    {
+      "epoch": 0.212,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00045412183654229966,
+      "loss": 3.219,
+      "step": 5035
+    },
+    {
+      "epoch": 0.2120421052631579,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004541023519719462,
+      "loss": 3.4325,
+      "step": 5036
+    },
+    {
+      "epoch": 0.2120842105263158,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004540828636830875,
+      "loss": 3.3257,
+      "step": 5037
+    },
+    {
+      "epoch": 0.2121263157894737,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004540633716760789,
+      "loss": 3.0871,
+      "step": 5038
+    },
+    {
+      "epoch": 0.2121684210526316,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004540438759512755,
+      "loss": 3.248,
+      "step": 5039
+    },
+    {
+      "epoch": 0.21221052631578946,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00045402437650903237,
+      "loss": 3.2909,
+      "step": 5040
+    },
+    {
+      "epoch": 0.21225263157894736,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004540048733497048,
+      "loss": 3.6888,
+      "step": 5041
+    },
+    {
+      "epoch": 0.21229473684210526,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0004539853664736482,
+      "loss": 3.2015,
+      "step": 5042
+    },
+    {
+      "epoch": 0.21233684210526316,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004539658558812178,
+      "loss": 3.1744,
+      "step": 5043
+    },
+    {
+      "epoch": 0.21237894736842106,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004539463415727692,
+      "loss": 3.5649,
+      "step": 5044
+    },
+    {
+      "epoch": 0.21242105263157895,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004539268235486579,
+      "loss": 3.6747,
+      "step": 5045
+    },
+    {
+      "epoch": 0.21246315789473685,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00045390730180923947,
+      "loss": 3.2573,
+      "step": 5046
+    },
+    {
+      "epoch": 0.21250526315789472,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00045388777635486954,
+      "loss": 2.9376,
+      "step": 5047
+    },
+    {
+      "epoch": 0.21254736842105262,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004538682471859039,
+      "loss": 3.3826,
+      "step": 5048
+    },
+    {
+      "epoch": 0.21258947368421052,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00045384871430269834,
+      "loss": 3.7058,
+      "step": 5049
+    },
+    {
+      "epoch": 0.21263157894736842,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00045382917770560885,
+      "loss": 3.041,
+      "step": 5050
+    },
+    {
+      "epoch": 0.21267368421052632,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.000453809637394991,
+      "loss": 2.7356,
+      "step": 5051
+    },
+    {
+      "epoch": 0.21271578947368422,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004537900933712012,
+      "loss": 2.9815,
+      "step": 5052
+    },
+    {
+      "epoch": 0.21275789473684212,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0004537705456345953,
+      "loss": 3.2517,
+      "step": 5053
+    },
+    {
+      "epoch": 0.2128,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00045375099418552947,
+      "loss": 3.5416,
+      "step": 5054
+    },
+    {
+      "epoch": 0.2128421052631579,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00045373143902435986,
+      "loss": 3.5455,
+      "step": 5055
+    },
+    {
+      "epoch": 0.2128842105263158,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004537118801514428,
+      "loss": 3.2139,
+      "step": 5056
+    },
+    {
+      "epoch": 0.21292631578947369,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00045369231756713476,
+      "loss": 2.9617,
+      "step": 5057
+    },
+    {
+      "epoch": 0.21296842105263158,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00045367275127179197,
+      "loss": 3.3845,
+      "step": 5058
+    },
+    {
+      "epoch": 0.21301052631578948,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00045365318126577094,
+      "loss": 3.5087,
+      "step": 5059
+    },
+    {
+      "epoch": 0.21305263157894735,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0004536336075494282,
+      "loss": 3.1787,
+      "step": 5060
+    },
+    {
+      "epoch": 0.21309473684210525,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00045361403012312044,
+      "loss": 3.0566,
+      "step": 5061
+    },
+    {
+      "epoch": 0.21313684210526315,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004535944489872043,
+      "loss": 3.137,
+      "step": 5062
+    },
+    {
+      "epoch": 0.21317894736842105,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004535748641420364,
+      "loss": 3.7776,
+      "step": 5063
+    },
+    {
+      "epoch": 0.21322105263157895,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004535552755879738,
+      "loss": 3.4562,
+      "step": 5064
+    },
+    {
+      "epoch": 0.21326315789473685,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004535356833253732,
+      "loss": 3.5042,
+      "step": 5065
+    },
+    {
+      "epoch": 0.21330526315789475,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00045351608735459155,
+      "loss": 3.3419,
+      "step": 5066
+    },
+    {
+      "epoch": 0.21334736842105262,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0004534964876759859,
+      "loss": 3.6556,
+      "step": 5067
+    },
+    {
+      "epoch": 0.21338947368421052,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004534768842899134,
+      "loss": 3.3249,
+      "step": 5068
+    },
+    {
+      "epoch": 0.21343157894736842,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00045345727719673116,
+      "loss": 3.4426,
+      "step": 5069
+    },
+    {
+      "epoch": 0.21347368421052632,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004534376663967964,
+      "loss": 3.2209,
+      "step": 5070
+    },
+    {
+      "epoch": 0.21351578947368421,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00045341805189046634,
+      "loss": 3.4591,
+      "step": 5071
+    },
+    {
+      "epoch": 0.2135578947368421,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004533984336780984,
+      "loss": 3.2081,
+      "step": 5072
+    },
+    {
+      "epoch": 0.2136,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004533788117600499,
+      "loss": 3.2632,
+      "step": 5073
+    },
+    {
+      "epoch": 0.21364210526315788,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00045335918613667847,
+      "loss": 3.4523,
+      "step": 5074
+    },
+    {
+      "epoch": 0.21368421052631578,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00045333955680834165,
+      "loss": 3.5172,
+      "step": 5075
+    },
+    {
+      "epoch": 0.21372631578947368,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.000453319923775397,
+      "loss": 3.3143,
+      "step": 5076
+    },
+    {
+      "epoch": 0.21376842105263158,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004533002870382022,
+      "loss": 3.6705,
+      "step": 5077
+    },
+    {
+      "epoch": 0.21381052631578948,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00045328064659711507,
+      "loss": 3.6537,
+      "step": 5078
+    },
+    {
+      "epoch": 0.21385263157894738,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00045326100245249334,
+      "loss": 2.9073,
+      "step": 5079
+    },
+    {
+      "epoch": 0.21389473684210528,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00045324135460469506,
+      "loss": 3.3215,
+      "step": 5080
+    },
+    {
+      "epoch": 0.21393684210526315,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004532217030540781,
+      "loss": 3.5495,
+      "step": 5081
+    },
+    {
+      "epoch": 0.21397894736842105,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004532020478010004,
+      "loss": 3.2773,
+      "step": 5082
+    },
+    {
+      "epoch": 0.21402105263157895,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004531823888458201,
+      "loss": 3.427,
+      "step": 5083
+    },
+    {
+      "epoch": 0.21406315789473684,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00045316272618889554,
+      "loss": 3.2485,
+      "step": 5084
+    },
+    {
+      "epoch": 0.21410526315789474,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00045314305983058475,
+      "loss": 3.1793,
+      "step": 5085
+    },
+    {
+      "epoch": 0.21414736842105264,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0004531233897712461,
+      "loss": 3.2069,
+      "step": 5086
+    },
+    {
+      "epoch": 0.2141894736842105,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004531037160112379,
+      "loss": 3.405,
+      "step": 5087
+    },
+    {
+      "epoch": 0.2142315789473684,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004530840385509187,
+      "loss": 3.4204,
+      "step": 5088
+    },
+    {
+      "epoch": 0.2142736842105263,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004530643573906469,
+      "loss": 3.1079,
+      "step": 5089
+    },
+    {
+      "epoch": 0.2143157894736842,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004530446725307811,
+      "loss": 3.5732,
+      "step": 5090
+    },
+    {
+      "epoch": 0.2143578947368421,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00045302498397168,
+      "loss": 2.8091,
+      "step": 5091
+    },
+    {
+      "epoch": 0.2144,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004530052917137022,
+      "loss": 3.1624,
+      "step": 5092
+    },
+    {
+      "epoch": 0.2144421052631579,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004529855957572064,
+      "loss": 3.3619,
+      "step": 5093
+    },
+    {
+      "epoch": 0.21448421052631578,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004529658961025517,
+      "loss": 2.9494,
+      "step": 5094
+    },
+    {
+      "epoch": 0.21452631578947368,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004529461927500967,
+      "loss": 3.6869,
+      "step": 5095
+    },
+    {
+      "epoch": 0.21456842105263157,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004529264857002007,
+      "loss": 3.4809,
+      "step": 5096
+    },
+    {
+      "epoch": 0.21461052631578947,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004529067749532224,
+      "loss": 2.87,
+      "step": 5097
+    },
+    {
+      "epoch": 0.21465263157894737,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00045288706050952113,
+      "loss": 3.2988,
+      "step": 5098
+    },
+    {
+      "epoch": 0.21469473684210527,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00045286734236945605,
+      "loss": 3.6759,
+      "step": 5099
+    },
+    {
+      "epoch": 0.21473684210526317,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004528476205333862,
+      "loss": 3.4897,
+      "step": 5100
+    },
+    {
+      "epoch": 0.21477894736842104,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004528278950016712,
+      "loss": 3.3378,
+      "step": 5101
+    },
+    {
+      "epoch": 0.21482105263157894,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004528081657746702,
+      "loss": 3.4187,
+      "step": 5102
+    },
+    {
+      "epoch": 0.21486315789473684,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00045278843285274266,
+      "loss": 3.4823,
+      "step": 5103
+    },
+    {
+      "epoch": 0.21490526315789474,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004527686962362482,
+      "loss": 3.1691,
+      "step": 5104
+    },
+    {
+      "epoch": 0.21494736842105264,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00045274895592554634,
+      "loss": 3.2961,
+      "step": 5105
+    },
+    {
+      "epoch": 0.21498947368421054,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004527292119209967,
+      "loss": 2.6684,
+      "step": 5106
+    },
+    {
+      "epoch": 0.21503157894736843,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00045270946422295904,
+      "loss": 3.134,
+      "step": 5107
+    },
+    {
+      "epoch": 0.2150736842105263,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004526897128317931,
+      "loss": 3.2913,
+      "step": 5108
+    },
+    {
+      "epoch": 0.2151157894736842,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.0004526699577478588,
+      "loss": 3.2627,
+      "step": 5109
+    },
+    {
+      "epoch": 0.2151578947368421,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0004526501989715159,
+      "loss": 3.5957,
+      "step": 5110
+    },
+    {
+      "epoch": 0.2152,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00045263043650312453,
+      "loss": 3.3444,
+      "step": 5111
+    },
+    {
+      "epoch": 0.2152421052631579,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004526106703430447,
+      "loss": 3.3949,
+      "step": 5112
+    },
+    {
+      "epoch": 0.2152842105263158,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004525909004916365,
+      "loss": 3.7753,
+      "step": 5113
+    },
+    {
+      "epoch": 0.21532631578947367,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00045257112694926005,
+      "loss": 3.4975,
+      "step": 5114
+    },
+    {
+      "epoch": 0.21536842105263157,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004525513497162758,
+      "loss": 3.5303,
+      "step": 5115
+    },
+    {
+      "epoch": 0.21541052631578947,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0004525315687930439,
+      "loss": 3.1494,
+      "step": 5116
+    },
+    {
+      "epoch": 0.21545263157894737,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004525117841799248,
+      "loss": 3.5098,
+      "step": 5117
+    },
+    {
+      "epoch": 0.21549473684210527,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004524919958772788,
+      "loss": 3.4156,
+      "step": 5118
+    },
+    {
+      "epoch": 0.21553684210526317,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004524722038854667,
+      "loss": 3.4661,
+      "step": 5119
+    },
+    {
+      "epoch": 0.21557894736842106,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00045245240820484894,
+      "loss": 3.7576,
+      "step": 5120
+    },
+    {
+      "epoch": 0.21562105263157894,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004524326088357862,
+      "loss": 3.0054,
+      "step": 5121
+    },
+    {
+      "epoch": 0.21566315789473683,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004524128057786391,
+      "loss": 3.3508,
+      "step": 5122
+    },
+    {
+      "epoch": 0.21570526315789473,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004523929990337685,
+      "loss": 3.0819,
+      "step": 5123
+    },
+    {
+      "epoch": 0.21574736842105263,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00045237318860153527,
+      "loss": 3.3521,
+      "step": 5124
+    },
+    {
+      "epoch": 0.21578947368421053,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00045235337448230027,
+      "loss": 3.0247,
+      "step": 5125
+    },
+    {
+      "epoch": 0.21583157894736843,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004523335566764246,
+      "loss": 3.7046,
+      "step": 5126
+    },
+    {
+      "epoch": 0.21587368421052633,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00045231373518426934,
+      "loss": 3.3787,
+      "step": 5127
+    },
+    {
+      "epoch": 0.2159157894736842,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00045229391000619544,
+      "loss": 3.2496,
+      "step": 5128
+    },
+    {
+      "epoch": 0.2159578947368421,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004522740811425642,
+      "loss": 3.1712,
+      "step": 5129
+    },
+    {
+      "epoch": 0.216,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004522542485937369,
+      "loss": 3.0353,
+      "step": 5130
+    },
+    {
+      "epoch": 0.2160421052631579,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004522344123600748,
+      "loss": 3.213,
+      "step": 5131
+    },
+    {
+      "epoch": 0.2160842105263158,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00045221457244193933,
+      "loss": 3.7437,
+      "step": 5132
+    },
+    {
+      "epoch": 0.2161263157894737,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00045219472883969194,
+      "loss": 3.1857,
+      "step": 5133
+    },
+    {
+      "epoch": 0.21616842105263157,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00045217488155369415,
+      "loss": 3.3486,
+      "step": 5134
+    },
+    {
+      "epoch": 0.21621052631578946,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00045215503058430753,
+      "loss": 3.2492,
+      "step": 5135
+    },
+    {
+      "epoch": 0.21625263157894736,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004521351759318938,
+      "loss": 3.6465,
+      "step": 5136
+    },
+    {
+      "epoch": 0.21629473684210526,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0004521153175968147,
+      "loss": 3.0847,
+      "step": 5137
+    },
+    {
+      "epoch": 0.21633684210526316,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00045209545557943194,
+      "loss": 3.0244,
+      "step": 5138
+    },
+    {
+      "epoch": 0.21637894736842106,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00045207558988010746,
+      "loss": 3.5579,
+      "step": 5139
+    },
+    {
+      "epoch": 0.21642105263157896,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00045205572049920314,
+      "loss": 3.1917,
+      "step": 5140
+    },
+    {
+      "epoch": 0.21646315789473683,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00045203584743708106,
+      "loss": 3.401,
+      "step": 5141
+    },
+    {
+      "epoch": 0.21650526315789473,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00045201597069410317,
+      "loss": 3.3406,
+      "step": 5142
+    },
+    {
+      "epoch": 0.21654736842105263,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004519960902706316,
+      "loss": 3.3438,
+      "step": 5143
+    },
+    {
+      "epoch": 0.21658947368421053,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00045197620616702864,
+      "loss": 3.2168,
+      "step": 5144
+    },
+    {
+      "epoch": 0.21663157894736843,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00045195631838365656,
+      "loss": 3.2245,
+      "step": 5145
+    },
+    {
+      "epoch": 0.21667368421052632,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00045193642692087767,
+      "loss": 3.2725,
+      "step": 5146
+    },
+    {
+      "epoch": 0.21671578947368422,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00045191653177905424,
+      "loss": 3.8455,
+      "step": 5147
+    },
+    {
+      "epoch": 0.2167578947368421,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004518966329585489,
+      "loss": 3.3189,
+      "step": 5148
+    },
+    {
+      "epoch": 0.2168,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004518767304597241,
+      "loss": 3.0322,
+      "step": 5149
+    },
+    {
+      "epoch": 0.2168421052631579,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004518568242829425,
+      "loss": 3.2413,
+      "step": 5150
+    },
+    {
+      "epoch": 0.2168842105263158,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004518369144285668,
+      "loss": 3.7623,
+      "step": 5151
+    },
+    {
+      "epoch": 0.2169263157894737,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00045181700089695956,
+      "loss": 3.2499,
+      "step": 5152
+    },
+    {
+      "epoch": 0.2169684210526316,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004517970836884837,
+      "loss": 3.6117,
+      "step": 5153
+    },
+    {
+      "epoch": 0.2170105263157895,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004517771628035021,
+      "loss": 3.2788,
+      "step": 5154
+    },
+    {
+      "epoch": 0.21705263157894736,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004517572382423777,
+      "loss": 3.7973,
+      "step": 5155
+    },
+    {
+      "epoch": 0.21709473684210526,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00045173731000547346,
+      "loss": 3.312,
+      "step": 5156
+    },
+    {
+      "epoch": 0.21713684210526316,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00045171737809315246,
+      "loss": 3.3567,
+      "step": 5157
+    },
+    {
+      "epoch": 0.21717894736842105,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004516974425057778,
+      "loss": 3.1575,
+      "step": 5158
+    },
+    {
+      "epoch": 0.21722105263157895,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004516775032437128,
+      "loss": 3.251,
+      "step": 5159
+    },
+    {
+      "epoch": 0.21726315789473685,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00045165756030732056,
+      "loss": 3.3178,
+      "step": 5160
+    },
+    {
+      "epoch": 0.21730526315789472,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00045163761369696457,
+      "loss": 3.0763,
+      "step": 5161
+    },
+    {
+      "epoch": 0.21734736842105262,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00045161766341300815,
+      "loss": 3.2707,
+      "step": 5162
+    },
+    {
+      "epoch": 0.21738947368421052,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0004515977094558148,
+      "loss": 3.491,
+      "step": 5163
+    },
+    {
+      "epoch": 0.21743157894736842,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004515777518257481,
+      "loss": 3.0893,
+      "step": 5164
+    },
+    {
+      "epoch": 0.21747368421052632,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00045155779052317155,
+      "loss": 3.0381,
+      "step": 5165
+    },
+    {
+      "epoch": 0.21751578947368422,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00045153782554844884,
+      "loss": 3.355,
+      "step": 5166
+    },
+    {
+      "epoch": 0.21755789473684212,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004515178569019438,
+      "loss": 2.8812,
+      "step": 5167
+    },
+    {
+      "epoch": 0.2176,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004514978845840202,
+      "loss": 2.6799,
+      "step": 5168
+    },
+    {
+      "epoch": 0.2176421052631579,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00045147790859504186,
+      "loss": 3.5175,
+      "step": 5169
+    },
+    {
+      "epoch": 0.21768421052631579,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00045145792893537274,
+      "loss": 3.2856,
+      "step": 5170
+    },
+    {
+      "epoch": 0.21772631578947368,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0004514379456053769,
+      "loss": 3.003,
+      "step": 5171
+    },
+    {
+      "epoch": 0.21776842105263158,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00045141795860541837,
+      "loss": 3.3975,
+      "step": 5172
+    },
+    {
+      "epoch": 0.21781052631578948,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00045139796793586127,
+      "loss": 3.1275,
+      "step": 5173
+    },
+    {
+      "epoch": 0.21785263157894738,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00045137797359706976,
+      "loss": 3.5616,
+      "step": 5174
+    },
+    {
+      "epoch": 0.21789473684210525,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004513579755894083,
+      "loss": 3.2332,
+      "step": 5175
+    },
+    {
+      "epoch": 0.21793684210526315,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004513379739132411,
+      "loss": 3.2272,
+      "step": 5176
+    },
+    {
+      "epoch": 0.21797894736842105,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004513179685689325,
+      "loss": 3.0212,
+      "step": 5177
+    },
+    {
+      "epoch": 0.21802105263157895,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00045129795955684704,
+      "loss": 3.3967,
+      "step": 5178
+    },
+    {
+      "epoch": 0.21806315789473685,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00045127794687734935,
+      "loss": 3.3554,
+      "step": 5179
+    },
+    {
+      "epoch": 0.21810526315789475,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.0004512579305308039,
+      "loss": 3.2028,
+      "step": 5180
+    },
+    {
+      "epoch": 0.21814736842105265,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00045123791051757546,
+      "loss": 3.3141,
+      "step": 5181
+    },
+    {
+      "epoch": 0.21818947368421052,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004512178868380287,
+      "loss": 3.7758,
+      "step": 5182
+    },
+    {
+      "epoch": 0.21823157894736842,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004511978594925285,
+      "loss": 3.1238,
+      "step": 5183
+    },
+    {
+      "epoch": 0.21827368421052631,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004511778284814397,
+      "loss": 2.844,
+      "step": 5184
+    },
+    {
+      "epoch": 0.2183157894736842,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00045115779380512725,
+      "loss": 3.4275,
+      "step": 5185
+    },
+    {
+      "epoch": 0.2183578947368421,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004511377554639561,
+      "loss": 3.464,
+      "step": 5186
+    },
+    {
+      "epoch": 0.2184,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00045111771345829137,
+      "loss": 3.1737,
+      "step": 5187
+    },
+    {
+      "epoch": 0.21844210526315788,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00045109766778849824,
+      "loss": 3.2259,
+      "step": 5188
+    },
+    {
+      "epoch": 0.21848421052631578,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00045107761845494186,
+      "loss": 3.5325,
+      "step": 5189
+    },
+    {
+      "epoch": 0.21852631578947368,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004510575654579876,
+      "loss": 3.1308,
+      "step": 5190
+    },
+    {
+      "epoch": 0.21856842105263158,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00045103750879800064,
+      "loss": 3.0776,
+      "step": 5191
+    },
+    {
+      "epoch": 0.21861052631578948,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00045101744847534655,
+      "loss": 3.226,
+      "step": 5192
+    },
+    {
+      "epoch": 0.21865263157894738,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004509973844903906,
+      "loss": 2.9164,
+      "step": 5193
+    },
+    {
+      "epoch": 0.21869473684210528,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004509773168434986,
+      "loss": 3.288,
+      "step": 5194
+    },
+    {
+      "epoch": 0.21873684210526315,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00045095724553503603,
+      "loss": 2.9882,
+      "step": 5195
+    },
+    {
+      "epoch": 0.21877894736842105,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004509371705653685,
+      "loss": 3.3799,
+      "step": 5196
+    },
+    {
+      "epoch": 0.21882105263157894,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00045091709193486186,
+      "loss": 3.2181,
+      "step": 5197
+    },
+    {
+      "epoch": 0.21886315789473684,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004508970096438818,
+      "loss": 3.1634,
+      "step": 5198
+    },
+    {
+      "epoch": 0.21890526315789474,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004508769236927943,
+      "loss": 3.5145,
+      "step": 5199
+    },
+    {
+      "epoch": 0.21894736842105264,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004508568340819653,
+      "loss": 3.3697,
+      "step": 5200
+    },
+    {
+      "epoch": 0.21898947368421054,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.0004508367408117608,
+      "loss": 3.5037,
+      "step": 5201
+    },
+    {
+      "epoch": 0.2190315789473684,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00045081664388254677,
+      "loss": 2.9871,
+      "step": 5202
+    },
+    {
+      "epoch": 0.2190736842105263,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00045079654329468956,
+      "loss": 3.28,
+      "step": 5203
+    },
+    {
+      "epoch": 0.2191157894736842,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004507764390485551,
+      "loss": 3.4845,
+      "step": 5204
+    },
+    {
+      "epoch": 0.2191578947368421,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00045075633114451,
+      "loss": 3.5159,
+      "step": 5205
+    },
+    {
+      "epoch": 0.2192,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00045073621958292034,
+      "loss": 3.5112,
+      "step": 5206
+    },
+    {
+      "epoch": 0.2192421052631579,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00045071610436415256,
+      "loss": 3.3459,
+      "step": 5207
+    },
+    {
+      "epoch": 0.21928421052631578,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00045069598548857327,
+      "loss": 3.6525,
+      "step": 5208
+    },
+    {
+      "epoch": 0.21932631578947367,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004506758629565489,
+      "loss": 3.294,
+      "step": 5209
+    },
+    {
+      "epoch": 0.21936842105263157,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00045065573676844606,
+      "loss": 3.412,
+      "step": 5210
+    },
+    {
+      "epoch": 0.21941052631578947,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004506356069246315,
+      "loss": 3.3039,
+      "step": 5211
+    },
+    {
+      "epoch": 0.21945263157894737,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004506154734254719,
+      "loss": 3.7136,
+      "step": 5212
+    },
+    {
+      "epoch": 0.21949473684210527,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004505953362713341,
+      "loss": 3.3052,
+      "step": 5213
+    },
+    {
+      "epoch": 0.21953684210526317,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00045057519546258496,
+      "loss": 3.655,
+      "step": 5214
+    },
+    {
+      "epoch": 0.21957894736842104,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00045055505099959143,
+      "loss": 3.343,
+      "step": 5215
+    },
+    {
+      "epoch": 0.21962105263157894,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004505349028827205,
+      "loss": 3.1474,
+      "step": 5216
+    },
+    {
+      "epoch": 0.21966315789473684,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004505147511123393,
+      "loss": 3.4399,
+      "step": 5217
+    },
+    {
+      "epoch": 0.21970526315789474,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004504945956888149,
+      "loss": 3.2831,
+      "step": 5218
+    },
+    {
+      "epoch": 0.21974736842105264,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004504744366125145,
+      "loss": 3.4026,
+      "step": 5219
+    },
+    {
+      "epoch": 0.21978947368421053,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0004504542738838054,
+      "loss": 3.3068,
+      "step": 5220
+    },
+    {
+      "epoch": 0.21983157894736843,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00045043410750305506,
+      "loss": 3.3471,
+      "step": 5221
+    },
+    {
+      "epoch": 0.2198736842105263,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00045041393747063073,
+      "loss": 3.2265,
+      "step": 5222
+    },
+    {
+      "epoch": 0.2199157894736842,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00045039376378689993,
+      "loss": 3.2977,
+      "step": 5223
+    },
+    {
+      "epoch": 0.2199578947368421,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004503735864522302,
+      "loss": 3.1652,
+      "step": 5224
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00045035340546698916,
+      "loss": 3.0997,
+      "step": 5225
+    },
+    {
+      "epoch": 0.2200421052631579,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0004503332208315445,
+      "loss": 2.7862,
+      "step": 5226
+    },
+    {
+      "epoch": 0.2200842105263158,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004503130325462639,
+      "loss": 3.5518,
+      "step": 5227
+    },
+    {
+      "epoch": 0.2201263157894737,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004502928406115152,
+      "loss": 3.3894,
+      "step": 5228
+    },
+    {
+      "epoch": 0.22016842105263157,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004502726450276663,
+      "loss": 3.65,
+      "step": 5229
+    },
+    {
+      "epoch": 0.22021052631578947,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004502524457950852,
+      "loss": 3.6042,
+      "step": 5230
+    },
+    {
+      "epoch": 0.22025263157894737,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00045023224291413966,
+      "loss": 3.3507,
+      "step": 5231
+    },
+    {
+      "epoch": 0.22029473684210527,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00045021203638519804,
+      "loss": 2.7982,
+      "step": 5232
+    },
+    {
+      "epoch": 0.22033684210526316,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00045019182620862833,
+      "loss": 3.0429,
+      "step": 5233
+    },
+    {
+      "epoch": 0.22037894736842106,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00045017161238479876,
+      "loss": 3.4239,
+      "step": 5234
+    },
+    {
+      "epoch": 0.22042105263157893,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004501513949140776,
+      "loss": 3.3614,
+      "step": 5235
+    },
+    {
+      "epoch": 0.22046315789473683,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00045013117379683314,
+      "loss": 3.3063,
+      "step": 5236
+    },
+    {
+      "epoch": 0.22050526315789473,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0004501109490334339,
+      "loss": 3.3593,
+      "step": 5237
+    },
+    {
+      "epoch": 0.22054736842105263,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00045009072062424826,
+      "loss": 3.1511,
+      "step": 5238
+    },
+    {
+      "epoch": 0.22058947368421053,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004500704885696448,
+      "loss": 3.3692,
+      "step": 5239
+    },
+    {
+      "epoch": 0.22063157894736843,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004500502528699921,
+      "loss": 3.6416,
+      "step": 5240
+    },
+    {
+      "epoch": 0.22067368421052633,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00045003001352565886,
+      "loss": 3.2563,
+      "step": 5241
+    },
+    {
+      "epoch": 0.2207157894736842,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004500097705370138,
+      "loss": 3.4161,
+      "step": 5242
+    },
+    {
+      "epoch": 0.2207578947368421,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00044998952390442573,
+      "loss": 3.4393,
+      "step": 5243
+    },
+    {
+      "epoch": 0.2208,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00044996927362826354,
+      "loss": 3.4941,
+      "step": 5244
+    },
+    {
+      "epoch": 0.2208421052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.000449949019708896,
+      "loss": 2.8758,
+      "step": 5245
+    },
+    {
+      "epoch": 0.2208842105263158,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00044992876214669243,
+      "loss": 3.814,
+      "step": 5246
+    },
+    {
+      "epoch": 0.2209263157894737,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00044990850094202164,
+      "loss": 3.0105,
+      "step": 5247
+    },
+    {
+      "epoch": 0.2209684210526316,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004498882360952528,
+      "loss": 3.795,
+      "step": 5248
+    },
+    {
+      "epoch": 0.22101052631578946,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0004498679676067552,
+      "loss": 3.4465,
+      "step": 5249
+    },
+    {
+      "epoch": 0.22105263157894736,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004498476954768981,
+      "loss": 3.0624,
+      "step": 5250
+    },
+    {
+      "epoch": 0.22109473684210526,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00044982741970605076,
+      "loss": 3.5291,
+      "step": 5251
+    },
+    {
+      "epoch": 0.22113684210526316,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004498071402945826,
+      "loss": 3.3404,
+      "step": 5252
+    },
+    {
+      "epoch": 0.22117894736842106,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00044978685724286315,
+      "loss": 3.6174,
+      "step": 5253
+    },
+    {
+      "epoch": 0.22122105263157896,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00044976657055126193,
+      "loss": 3.5667,
+      "step": 5254
+    },
+    {
+      "epoch": 0.22126315789473683,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0004497462802201484,
+      "loss": 3.2798,
+      "step": 5255
+    },
+    {
+      "epoch": 0.22130526315789473,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00044972598624989246,
+      "loss": 3.3554,
+      "step": 5256
+    },
+    {
+      "epoch": 0.22134736842105263,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00044970568864086363,
+      "loss": 3.1223,
+      "step": 5257
+    },
+    {
+      "epoch": 0.22138947368421052,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004496853873934318,
+      "loss": 3.5604,
+      "step": 5258
+    },
+    {
+      "epoch": 0.22143157894736842,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004496650825079669,
+      "loss": 3.7436,
+      "step": 5259
+    },
+    {
+      "epoch": 0.22147368421052632,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0004496447739848387,
+      "loss": 3.0985,
+      "step": 5260
+    },
+    {
+      "epoch": 0.22151578947368422,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0004496244618244174,
+      "loss": 3.3216,
+      "step": 5261
+    },
+    {
+      "epoch": 0.2215578947368421,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004496041460270729,
+      "loss": 3.6747,
+      "step": 5262
+    },
+    {
+      "epoch": 0.2216,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00044958382659317534,
+      "loss": 3.4734,
+      "step": 5263
+    },
+    {
+      "epoch": 0.2216421052631579,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00044956350352309506,
+      "loss": 3.6832,
+      "step": 5264
+    },
+    {
+      "epoch": 0.2216842105263158,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00044954317681720216,
+      "loss": 3.6663,
+      "step": 5265
+    },
+    {
+      "epoch": 0.2217263157894737,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00044952284647586704,
+      "loss": 3.6828,
+      "step": 5266
+    },
+    {
+      "epoch": 0.2217684210526316,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00044950251249946003,
+      "loss": 3.691,
+      "step": 5267
+    },
+    {
+      "epoch": 0.22181052631578949,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00044948217488835176,
+      "loss": 3.2944,
+      "step": 5268
+    },
+    {
+      "epoch": 0.22185263157894736,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004494618336429126,
+      "loss": 2.6757,
+      "step": 5269
+    },
+    {
+      "epoch": 0.22189473684210526,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004494414887635132,
+      "loss": 3.1438,
+      "step": 5270
+    },
+    {
+      "epoch": 0.22193684210526315,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00044942114025052417,
+      "loss": 3.0991,
+      "step": 5271
+    },
+    {
+      "epoch": 0.22197894736842105,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00044940078810431626,
+      "loss": 3.4627,
+      "step": 5272
+    },
+    {
+      "epoch": 0.22202105263157895,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00044938043232526037,
+      "loss": 3.7105,
+      "step": 5273
+    },
+    {
+      "epoch": 0.22206315789473685,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00044936007291372713,
+      "loss": 3.3262,
+      "step": 5274
+    },
+    {
+      "epoch": 0.22210526315789475,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004493397098700877,
+      "loss": 3.8209,
+      "step": 5275
+    },
+    {
+      "epoch": 0.22214736842105262,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.000449319343194713,
+      "loss": 3.5708,
+      "step": 5276
+    },
+    {
+      "epoch": 0.22218947368421052,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.000449298972887974,
+      "loss": 3.3666,
+      "step": 5277
+    },
+    {
+      "epoch": 0.22223157894736842,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00044927859895024184,
+      "loss": 3.5422,
+      "step": 5278
+    },
+    {
+      "epoch": 0.22227368421052632,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004492582213818878,
+      "loss": 3.7804,
+      "step": 5279
+    },
+    {
+      "epoch": 0.22231578947368422,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00044923784018328314,
+      "loss": 3.3119,
+      "step": 5280
+    },
+    {
+      "epoch": 0.22235789473684212,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004492174553547991,
+      "loss": 3.2462,
+      "step": 5281
+    },
+    {
+      "epoch": 0.2224,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00044919706689680697,
+      "loss": 3.061,
+      "step": 5282
+    },
+    {
+      "epoch": 0.22244210526315789,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00044917667480967843,
+      "loss": 3.3981,
+      "step": 5283
+    },
+    {
+      "epoch": 0.22248421052631578,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00044915627909378485,
+      "loss": 3.7243,
+      "step": 5284
+    },
+    {
+      "epoch": 0.22252631578947368,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.000449135879749498,
+      "loss": 3.2684,
+      "step": 5285
+    },
+    {
+      "epoch": 0.22256842105263158,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004491154767771892,
+      "loss": 3.8089,
+      "step": 5286
+    },
+    {
+      "epoch": 0.22261052631578948,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004490950701772305,
+      "loss": 3.0875,
+      "step": 5287
+    },
+    {
+      "epoch": 0.22265263157894738,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004490746599499935,
+      "loss": 3.5963,
+      "step": 5288
+    },
+    {
+      "epoch": 0.22269473684210525,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004490542460958501,
+      "loss": 3.5365,
+      "step": 5289
+    },
+    {
+      "epoch": 0.22273684210526315,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004490338286151723,
+      "loss": 3.1544,
+      "step": 5290
+    },
+    {
+      "epoch": 0.22277894736842105,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004490134075083319,
+      "loss": 3.5545,
+      "step": 5291
+    },
+    {
+      "epoch": 0.22282105263157895,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00044899298277570104,
+      "loss": 3.4912,
+      "step": 5292
+    },
+    {
+      "epoch": 0.22286315789473685,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004489725544176519,
+      "loss": 3.7183,
+      "step": 5293
+    },
+    {
+      "epoch": 0.22290526315789475,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00044895212243455655,
+      "loss": 2.8311,
+      "step": 5294
+    },
+    {
+      "epoch": 0.22294736842105264,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004489316868267873,
+      "loss": 3.3294,
+      "step": 5295
+    },
+    {
+      "epoch": 0.22298947368421052,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00044891124759471646,
+      "loss": 3.2148,
+      "step": 5296
+    },
+    {
+      "epoch": 0.22303157894736841,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004488908047387164,
+      "loss": 3.6311,
+      "step": 5297
+    },
+    {
+      "epoch": 0.2230736842105263,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00044887035825915957,
+      "loss": 3.3109,
+      "step": 5298
+    },
+    {
+      "epoch": 0.2231157894736842,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004488499081564185,
+      "loss": 3.8074,
+      "step": 5299
+    },
+    {
+      "epoch": 0.2231578947368421,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00044882945443086564,
+      "loss": 3.6378,
+      "step": 5300
+    },
+    {
+      "epoch": 0.2232,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00044880899708287387,
+      "loss": 3.2694,
+      "step": 5301
+    },
+    {
+      "epoch": 0.2232421052631579,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004487885361128157,
+      "loss": 3.2198,
+      "step": 5302
+    },
+    {
+      "epoch": 0.22328421052631578,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004487680715210639,
+      "loss": 3.299,
+      "step": 5303
+    },
+    {
+      "epoch": 0.22332631578947368,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004487476033079915,
+      "loss": 3.1346,
+      "step": 5304
+    },
+    {
+      "epoch": 0.22336842105263158,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00044872713147397125,
+      "loss": 3.0447,
+      "step": 5305
+    },
+    {
+      "epoch": 0.22341052631578948,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00044870665601937613,
+      "loss": 3.1736,
+      "step": 5306
+    },
+    {
+      "epoch": 0.22345263157894737,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004486861769445793,
+      "loss": 3.3146,
+      "step": 5307
+    },
+    {
+      "epoch": 0.22349473684210527,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00044866569424995364,
+      "loss": 3.0451,
+      "step": 5308
+    },
+    {
+      "epoch": 0.22353684210526314,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00044864520793587247,
+      "loss": 3.5234,
+      "step": 5309
+    },
+    {
+      "epoch": 0.22357894736842104,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004486247180027091,
+      "loss": 3.0574,
+      "step": 5310
+    },
+    {
+      "epoch": 0.22362105263157894,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00044860422445083673,
+      "loss": 3.5278,
+      "step": 5311
+    },
+    {
+      "epoch": 0.22366315789473684,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0004485837272806287,
+      "loss": 3.6568,
+      "step": 5312
+    },
+    {
+      "epoch": 0.22370526315789474,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00044856322649245844,
+      "loss": 3.4264,
+      "step": 5313
+    },
+    {
+      "epoch": 0.22374736842105264,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00044854272208669953,
+      "loss": 3.7323,
+      "step": 5314
+    },
+    {
+      "epoch": 0.22378947368421054,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0004485222140637255,
+      "loss": 3.2967,
+      "step": 5315
+    },
+    {
+      "epoch": 0.2238315789473684,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00044850170242391,
+      "loss": 3.366,
+      "step": 5316
+    },
+    {
+      "epoch": 0.2238736842105263,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00044848118716762675,
+      "loss": 3.1688,
+      "step": 5317
+    },
+    {
+      "epoch": 0.2239157894736842,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00044846066829524946,
+      "loss": 3.3749,
+      "step": 5318
+    },
+    {
+      "epoch": 0.2239578947368421,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004484401458071519,
+      "loss": 3.4601,
+      "step": 5319
+    },
+    {
+      "epoch": 0.224,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00044841961970370814,
+      "loss": 3.0214,
+      "step": 5320
+    },
+    {
+      "epoch": 0.2240421052631579,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.000448399089985292,
+      "loss": 3.2473,
+      "step": 5321
+    },
+    {
+      "epoch": 0.2240842105263158,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004483785566522776,
+      "loss": 2.9274,
+      "step": 5322
+    },
+    {
+      "epoch": 0.22412631578947367,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00044835801970503886,
+      "loss": 3.3684,
+      "step": 5323
+    },
+    {
+      "epoch": 0.22416842105263157,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00044833747914395014,
+      "loss": 3.6371,
+      "step": 5324
+    },
+    {
+      "epoch": 0.22421052631578947,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00044831693496938553,
+      "loss": 3.7025,
+      "step": 5325
+    },
+    {
+      "epoch": 0.22425263157894737,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004482963871817195,
+      "loss": 3.3776,
+      "step": 5326
+    },
+    {
+      "epoch": 0.22429473684210527,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004482758357813262,
+      "loss": 3.3394,
+      "step": 5327
+    },
+    {
+      "epoch": 0.22433684210526317,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00044825528076858015,
+      "loss": 3.4329,
+      "step": 5328
+    },
+    {
+      "epoch": 0.22437894736842104,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0004482347221438558,
+      "loss": 2.8252,
+      "step": 5329
+    },
+    {
+      "epoch": 0.22442105263157894,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00044821415990752776,
+      "loss": 3.1779,
+      "step": 5330
+    },
+    {
+      "epoch": 0.22446315789473684,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00044819359405997065,
+      "loss": 2.9593,
+      "step": 5331
+    },
+    {
+      "epoch": 0.22450526315789474,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004481730246015591,
+      "loss": 3.3424,
+      "step": 5332
+    },
+    {
+      "epoch": 0.22454736842105263,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004481524515326678,
+      "loss": 3.5201,
+      "step": 5333
+    },
+    {
+      "epoch": 0.22458947368421053,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004481318748536717,
+      "loss": 3.0077,
+      "step": 5334
+    },
+    {
+      "epoch": 0.22463157894736843,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004481112945649457,
+      "loss": 3.2576,
+      "step": 5335
+    },
+    {
+      "epoch": 0.2246736842105263,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00044809071066686455,
+      "loss": 3.2089,
+      "step": 5336
+    },
+    {
+      "epoch": 0.2247157894736842,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00044807012315980355,
+      "loss": 2.9588,
+      "step": 5337
+    },
+    {
+      "epoch": 0.2247578947368421,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00044804953204413754,
+      "loss": 3.2408,
+      "step": 5338
+    },
+    {
+      "epoch": 0.2248,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004480289373202418,
+      "loss": 3.6674,
+      "step": 5339
+    },
+    {
+      "epoch": 0.2248421052631579,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004480083389884915,
+      "loss": 3.6298,
+      "step": 5340
+    },
+    {
+      "epoch": 0.2248842105263158,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00044798773704926185,
+      "loss": 3.1455,
+      "step": 5341
+    },
+    {
+      "epoch": 0.2249263157894737,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00044796713150292835,
+      "loss": 2.7858,
+      "step": 5342
+    },
+    {
+      "epoch": 0.22496842105263157,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004479465223498662,
+      "loss": 3.0402,
+      "step": 5343
+    },
+    {
+      "epoch": 0.22501052631578947,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004479259095904511,
+      "loss": 3.6379,
+      "step": 5344
+    },
+    {
+      "epoch": 0.22505263157894737,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00044790529322505845,
+      "loss": 3.3131,
+      "step": 5345
+    },
+    {
+      "epoch": 0.22509473684210526,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004478846732540639,
+      "loss": 2.9419,
+      "step": 5346
+    },
+    {
+      "epoch": 0.22513684210526316,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004478640496778431,
+      "loss": 3.4673,
+      "step": 5347
+    },
+    {
+      "epoch": 0.22517894736842106,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00044784342249677174,
+      "loss": 3.4492,
+      "step": 5348
+    },
+    {
+      "epoch": 0.22522105263157896,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0004478227917112257,
+      "loss": 3.5149,
+      "step": 5349
+    },
+    {
+      "epoch": 0.22526315789473683,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004478021573215808,
+      "loss": 3.6039,
+      "step": 5350
+    },
+    {
+      "epoch": 0.22530526315789473,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0004477815193282131,
+      "loss": 2.7898,
+      "step": 5351
+    },
+    {
+      "epoch": 0.22534736842105263,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00044776087773149844,
+      "loss": 3.0764,
+      "step": 5352
+    },
+    {
+      "epoch": 0.22538947368421053,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00044774023253181294,
+      "loss": 3.5454,
+      "step": 5353
+    },
+    {
+      "epoch": 0.22543157894736843,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00044771958372953277,
+      "loss": 3.3953,
+      "step": 5354
+    },
+    {
+      "epoch": 0.22547368421052633,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0004476989313250341,
+      "loss": 3.5499,
+      "step": 5355
+    },
+    {
+      "epoch": 0.2255157894736842,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00044767827531869316,
+      "loss": 3.4708,
+      "step": 5356
+    },
+    {
+      "epoch": 0.2255578947368421,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004476576157108863,
+      "loss": 3.5151,
+      "step": 5357
+    },
+    {
+      "epoch": 0.2256,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00044763695250198994,
+      "loss": 3.656,
+      "step": 5358
+    },
+    {
+      "epoch": 0.2256421052631579,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004476162856923805,
+      "loss": 3.3099,
+      "step": 5359
+    },
+    {
+      "epoch": 0.2256842105263158,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00044759561528243455,
+      "loss": 3.3123,
+      "step": 5360
+    },
+    {
+      "epoch": 0.2257263157894737,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004475749412725286,
+      "loss": 2.8758,
+      "step": 5361
+    },
+    {
+      "epoch": 0.2257684210526316,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00044755426366303944,
+      "loss": 3.0864,
+      "step": 5362
+    },
+    {
+      "epoch": 0.22581052631578946,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0004475335824543436,
+      "loss": 3.1212,
+      "step": 5363
+    },
+    {
+      "epoch": 0.22585263157894736,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004475128976468181,
+      "loss": 3.3591,
+      "step": 5364
+    },
+    {
+      "epoch": 0.22589473684210526,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004474922092408397,
+      "loss": 3.6221,
+      "step": 5365
+    },
+    {
+      "epoch": 0.22593684210526316,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00044747151723678515,
+      "loss": 3.3725,
+      "step": 5366
+    },
+    {
+      "epoch": 0.22597894736842106,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00044745082163503167,
+      "loss": 3.2719,
+      "step": 5367
+    },
+    {
+      "epoch": 0.22602105263157896,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00044743012243595614,
+      "loss": 3.4701,
+      "step": 5368
+    },
+    {
+      "epoch": 0.22606315789473685,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004474094196399358,
+      "loss": 3.4782,
+      "step": 5369
+    },
+    {
+      "epoch": 0.22610526315789473,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00044738871324734775,
+      "loss": 3.5036,
+      "step": 5370
+    },
+    {
+      "epoch": 0.22614736842105262,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0004473680032585694,
+      "loss": 3.4156,
+      "step": 5371
+    },
+    {
+      "epoch": 0.22618947368421052,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00044734728967397774,
+      "loss": 3.387,
+      "step": 5372
+    },
+    {
+      "epoch": 0.22623157894736842,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00044732657249395047,
+      "loss": 3.2278,
+      "step": 5373
+    },
+    {
+      "epoch": 0.22627368421052632,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0004473058517188648,
+      "loss": 3.0927,
+      "step": 5374
+    },
+    {
+      "epoch": 0.22631578947368422,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00044728512734909845,
+      "loss": 3.4747,
+      "step": 5375
+    },
+    {
+      "epoch": 0.22635789473684212,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0004472643993850288,
+      "loss": 3.2073,
+      "step": 5376
+    },
+    {
+      "epoch": 0.2264,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004472436678270336,
+      "loss": 3.5488,
+      "step": 5377
+    },
+    {
+      "epoch": 0.2264421052631579,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004472229326754905,
+      "loss": 3.3754,
+      "step": 5378
+    },
+    {
+      "epoch": 0.2264842105263158,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00044720219393077726,
+      "loss": 3.1922,
+      "step": 5379
+    },
+    {
+      "epoch": 0.2265263157894737,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00044718145159327184,
+      "loss": 3.1251,
+      "step": 5380
+    },
+    {
+      "epoch": 0.22656842105263159,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00044716070566335197,
+      "loss": 3.0596,
+      "step": 5381
+    },
+    {
+      "epoch": 0.22661052631578948,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004471399561413957,
+      "loss": 3.2554,
+      "step": 5382
+    },
+    {
+      "epoch": 0.22665263157894736,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004471192030277811,
+      "loss": 2.9799,
+      "step": 5383
+    },
+    {
+      "epoch": 0.22669473684210525,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004470984463228862,
+      "loss": 3.1611,
+      "step": 5384
+    },
+    {
+      "epoch": 0.22673684210526315,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00044707768602708924,
+      "loss": 3.3698,
+      "step": 5385
+    },
+    {
+      "epoch": 0.22677894736842105,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.0004470569221407683,
+      "loss": 3.4199,
+      "step": 5386
+    },
+    {
+      "epoch": 0.22682105263157895,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004470361546643019,
+      "loss": 3.0724,
+      "step": 5387
+    },
+    {
+      "epoch": 0.22686315789473685,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00044701538359806823,
+      "loss": 2.797,
+      "step": 5388
+    },
+    {
+      "epoch": 0.22690526315789475,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004469946089424457,
+      "loss": 3.6683,
+      "step": 5389
+    },
+    {
+      "epoch": 0.22694736842105262,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004469738306978128,
+      "loss": 3.3828,
+      "step": 5390
+    },
+    {
+      "epoch": 0.22698947368421052,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004469530488645483,
+      "loss": 3.5654,
+      "step": 5391
+    },
+    {
+      "epoch": 0.22703157894736842,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004469322634430306,
+      "loss": 3.688,
+      "step": 5392
+    },
+    {
+      "epoch": 0.22707368421052632,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004469114744336384,
+      "loss": 3.4679,
+      "step": 5393
+    },
+    {
+      "epoch": 0.22711578947368422,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004468906818367505,
+      "loss": 3.5697,
+      "step": 5394
+    },
+    {
+      "epoch": 0.22715789473684211,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00044686988565274577,
+      "loss": 3.4219,
+      "step": 5395
+    },
+    {
+      "epoch": 0.2272,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00044684908588200304,
+      "loss": 2.9927,
+      "step": 5396
+    },
+    {
+      "epoch": 0.22724210526315788,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0004468282825249012,
+      "loss": 3.5558,
+      "step": 5397
+    },
+    {
+      "epoch": 0.22728421052631578,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00044680747558181933,
+      "loss": 2.798,
+      "step": 5398
+    },
+    {
+      "epoch": 0.22732631578947368,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0004467866650531365,
+      "loss": 3.3905,
+      "step": 5399
+    },
+    {
+      "epoch": 0.22736842105263158,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00044676585093923185,
+      "loss": 3.158,
+      "step": 5400
+    },
+    {
+      "epoch": 0.22741052631578948,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00044674503324048455,
+      "loss": 3.6752,
+      "step": 5401
+    },
+    {
+      "epoch": 0.22745263157894738,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00044672421195727396,
+      "loss": 3.1355,
+      "step": 5402
+    },
+    {
+      "epoch": 0.22749473684210525,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00044670338708997933,
+      "loss": 3.0773,
+      "step": 5403
+    },
+    {
+      "epoch": 0.22753684210526315,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00044668255863898013,
+      "loss": 3.7283,
+      "step": 5404
+    },
+    {
+      "epoch": 0.22757894736842105,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004466617266046558,
+      "loss": 3.6276,
+      "step": 5405
+    },
+    {
+      "epoch": 0.22762105263157895,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00044664089098738586,
+      "loss": 3.3026,
+      "step": 5406
+    },
+    {
+      "epoch": 0.22766315789473685,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00044662005178754995,
+      "loss": 3.5225,
+      "step": 5407
+    },
+    {
+      "epoch": 0.22770526315789474,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004465992090055277,
+      "loss": 3.5985,
+      "step": 5408
+    },
+    {
+      "epoch": 0.22774736842105264,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004465783626416989,
+      "loss": 3.7124,
+      "step": 5409
+    },
+    {
+      "epoch": 0.22778947368421051,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004465575126964433,
+      "loss": 3.3157,
+      "step": 5410
+    },
+    {
+      "epoch": 0.2278315789473684,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00044653665917014076,
+      "loss": 3.2714,
+      "step": 5411
+    },
+    {
+      "epoch": 0.2278736842105263,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004465158020631711,
+      "loss": 3.5093,
+      "step": 5412
+    },
+    {
+      "epoch": 0.2279157894736842,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00044649494137591453,
+      "loss": 3.3758,
+      "step": 5413
+    },
+    {
+      "epoch": 0.2279578947368421,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.000446474077108751,
+      "loss": 3.3725,
+      "step": 5414
+    },
+    {
+      "epoch": 0.228,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004464532092620606,
+      "loss": 3.4444,
+      "step": 5415
+    },
+    {
+      "epoch": 0.2280421052631579,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004464323378362236,
+      "loss": 2.7441,
+      "step": 5416
+    },
+    {
+      "epoch": 0.22808421052631578,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00044641146283162016,
+      "loss": 4.0745,
+      "step": 5417
+    },
+    {
+      "epoch": 0.22812631578947368,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00044639058424863066,
+      "loss": 3.1495,
+      "step": 5418
+    },
+    {
+      "epoch": 0.22816842105263158,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00044636970208763543,
+      "loss": 2.872,
+      "step": 5419
+    },
+    {
+      "epoch": 0.22821052631578947,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.000446348816349015,
+      "loss": 3.4887,
+      "step": 5420
+    },
+    {
+      "epoch": 0.22825263157894737,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004463279270331498,
+      "loss": 3.6268,
+      "step": 5421
+    },
+    {
+      "epoch": 0.22829473684210527,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004463070341404205,
+      "loss": 3.7353,
+      "step": 5422
+    },
+    {
+      "epoch": 0.22833684210526317,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00044628613767120755,
+      "loss": 3.7054,
+      "step": 5423
+    },
+    {
+      "epoch": 0.22837894736842104,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00044626523762589194,
+      "loss": 3.2892,
+      "step": 5424
+    },
+    {
+      "epoch": 0.22842105263157894,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00044624433400485426,
+      "loss": 2.9682,
+      "step": 5425
+    },
+    {
+      "epoch": 0.22846315789473684,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.0004462234268084754,
+      "loss": 3.6513,
+      "step": 5426
+    },
+    {
+      "epoch": 0.22850526315789474,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00044620251603713623,
+      "loss": 3.6864,
+      "step": 5427
+    },
+    {
+      "epoch": 0.22854736842105264,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004461816016912178,
+      "loss": 2.9612,
+      "step": 5428
+    },
+    {
+      "epoch": 0.22858947368421054,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00044616068377110096,
+      "loss": 3.3855,
+      "step": 5429
+    },
+    {
+      "epoch": 0.2286315789473684,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004461397622771671,
+      "loss": 3.2489,
+      "step": 5430
+    },
+    {
+      "epoch": 0.2286736842105263,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00044611883720979716,
+      "loss": 3.078,
+      "step": 5431
+    },
+    {
+      "epoch": 0.2287157894736842,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004460979085693724,
+      "loss": 3.7057,
+      "step": 5432
+    },
+    {
+      "epoch": 0.2287578947368421,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004460769763562742,
+      "loss": 3.3223,
+      "step": 5433
+    },
+    {
+      "epoch": 0.2288,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004460560405708839,
+      "loss": 3.4355,
+      "step": 5434
+    },
+    {
+      "epoch": 0.2288421052631579,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004460351012135828,
+      "loss": 3.4338,
+      "step": 5435
+    },
+    {
+      "epoch": 0.2288842105263158,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00044601415828475255,
+      "loss": 3.2501,
+      "step": 5436
+    },
+    {
+      "epoch": 0.22892631578947367,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004459932117847747,
+      "loss": 3.2722,
+      "step": 5437
+    },
+    {
+      "epoch": 0.22896842105263157,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004459722617140307,
+      "loss": 3.1988,
+      "step": 5438
+    },
+    {
+      "epoch": 0.22901052631578947,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004459513080729024,
+      "loss": 3.2104,
+      "step": 5439
+    },
+    {
+      "epoch": 0.22905263157894737,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004459303508617715,
+      "loss": 3.0857,
+      "step": 5440
+    },
+    {
+      "epoch": 0.22909473684210527,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004459093900810198,
+      "loss": 3.3463,
+      "step": 5441
+    },
+    {
+      "epoch": 0.22913684210526317,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004458884257310293,
+      "loss": 2.6772,
+      "step": 5442
+    },
+    {
+      "epoch": 0.22917894736842107,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004458674578121817,
+      "loss": 3.5425,
+      "step": 5443
+    },
+    {
+      "epoch": 0.22922105263157894,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00044584648632485925,
+      "loss": 3.4621,
+      "step": 5444
+    },
+    {
+      "epoch": 0.22926315789473684,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0004458255112694439,
+      "loss": 3.3796,
+      "step": 5445
+    },
+    {
+      "epoch": 0.22930526315789473,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0004458045326463177,
+      "loss": 2.9716,
+      "step": 5446
+    },
+    {
+      "epoch": 0.22934736842105263,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004457835504558632,
+      "loss": 3.1893,
+      "step": 5447
+    },
+    {
+      "epoch": 0.22938947368421053,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004457625646984623,
+      "loss": 3.3768,
+      "step": 5448
+    },
+    {
+      "epoch": 0.22943157894736843,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004457415753744975,
+      "loss": 3.7506,
+      "step": 5449
+    },
+    {
+      "epoch": 0.2294736842105263,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004457205824843512,
+      "loss": 3.4001,
+      "step": 5450
+    },
+    {
+      "epoch": 0.2295157894736842,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004456995860284059,
+      "loss": 3.6028,
+      "step": 5451
+    },
+    {
+      "epoch": 0.2295578947368421,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00044567858600704397,
+      "loss": 2.8487,
+      "step": 5452
+    },
+    {
+      "epoch": 0.2296,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00044565758242064813,
+      "loss": 3.4237,
+      "step": 5453
+    },
+    {
+      "epoch": 0.2296421052631579,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0004456365752696011,
+      "loss": 3.3297,
+      "step": 5454
+    },
+    {
+      "epoch": 0.2296842105263158,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00044561556455428554,
+      "loss": 3.5052,
+      "step": 5455
+    },
+    {
+      "epoch": 0.2297263157894737,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00044559455027508424,
+      "loss": 3.2671,
+      "step": 5456
+    },
+    {
+      "epoch": 0.22976842105263157,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00044557353243238,
+      "loss": 3.3195,
+      "step": 5457
+    },
+    {
+      "epoch": 0.22981052631578947,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004455525110265557,
+      "loss": 3.3865,
+      "step": 5458
+    },
+    {
+      "epoch": 0.22985263157894736,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004455314860579945,
+      "loss": 3.7336,
+      "step": 5459
+    },
+    {
+      "epoch": 0.22989473684210526,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004455104575270794,
+      "loss": 3.104,
+      "step": 5460
+    },
+    {
+      "epoch": 0.22993684210526316,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00044548942543419344,
+      "loss": 3.3694,
+      "step": 5461
+    },
+    {
+      "epoch": 0.22997894736842106,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00044546838977971986,
+      "loss": 3.5706,
+      "step": 5462
+    },
+    {
+      "epoch": 0.23002105263157896,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00044544735056404185,
+      "loss": 3.9186,
+      "step": 5463
+    },
+    {
+      "epoch": 0.23006315789473683,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004454263077875428,
+      "loss": 3.4084,
+      "step": 5464
+    },
+    {
+      "epoch": 0.23010526315789473,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.000445405261450606,
+      "loss": 3.7675,
+      "step": 5465
+    },
+    {
+      "epoch": 0.23014736842105263,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00044538421155361496,
+      "loss": 3.5539,
+      "step": 5466
+    },
+    {
+      "epoch": 0.23018947368421053,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00044536315809695317,
+      "loss": 3.8473,
+      "step": 5467
+    },
+    {
+      "epoch": 0.23023157894736843,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00044534210108100413,
+      "loss": 3.4921,
+      "step": 5468
+    },
+    {
+      "epoch": 0.23027368421052632,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00044532104050615153,
+      "loss": 3.2213,
+      "step": 5469
+    },
+    {
+      "epoch": 0.23031578947368422,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004452999763727791,
+      "loss": 3.0958,
+      "step": 5470
+    },
+    {
+      "epoch": 0.2303578947368421,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00044527890868127054,
+      "loss": 2.8111,
+      "step": 5471
+    },
+    {
+      "epoch": 0.2304,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0004452578374320097,
+      "loss": 3.3192,
+      "step": 5472
+    },
+    {
+      "epoch": 0.2304421052631579,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004452367626253805,
+      "loss": 3.177,
+      "step": 5473
+    },
+    {
+      "epoch": 0.2304842105263158,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004452156842617668,
+      "loss": 2.9855,
+      "step": 5474
+    },
+    {
+      "epoch": 0.2305263157894737,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0004451946023415528,
+      "loss": 2.9031,
+      "step": 5475
+    },
+    {
+      "epoch": 0.2305684210526316,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004451735168651225,
+      "loss": 2.9309,
+      "step": 5476
+    },
+    {
+      "epoch": 0.23061052631578946,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00044515242783286005,
+      "loss": 3.6076,
+      "step": 5477
+    },
+    {
+      "epoch": 0.23065263157894736,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0004451313352451496,
+      "loss": 3.3416,
+      "step": 5478
+    },
+    {
+      "epoch": 0.23069473684210526,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00044511023910237546,
+      "loss": 3.4338,
+      "step": 5479
+    },
+    {
+      "epoch": 0.23073684210526316,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004450891394049221,
+      "loss": 3.3928,
+      "step": 5480
+    },
+    {
+      "epoch": 0.23077894736842106,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004450680361531738,
+      "loss": 3.6654,
+      "step": 5481
+    },
+    {
+      "epoch": 0.23082105263157895,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00044504692934751506,
+      "loss": 2.9869,
+      "step": 5482
+    },
+    {
+      "epoch": 0.23086315789473685,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00044502581898833043,
+      "loss": 3.0357,
+      "step": 5483
+    },
+    {
+      "epoch": 0.23090526315789472,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00044500470507600456,
+      "loss": 3.2127,
+      "step": 5484
+    },
+    {
+      "epoch": 0.23094736842105262,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004449835876109221,
+      "loss": 3.1452,
+      "step": 5485
+    },
+    {
+      "epoch": 0.23098947368421052,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004449624665934677,
+      "loss": 2.8172,
+      "step": 5486
+    },
+    {
+      "epoch": 0.23103157894736842,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00044494134202402626,
+      "loss": 3.4768,
+      "step": 5487
+    },
+    {
+      "epoch": 0.23107368421052632,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00044492021390298264,
+      "loss": 3.1722,
+      "step": 5488
+    },
+    {
+      "epoch": 0.23111578947368422,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00044489908223072175,
+      "loss": 3.2747,
+      "step": 5489
+    },
+    {
+      "epoch": 0.23115789473684212,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004448779470076285,
+      "loss": 3.2161,
+      "step": 5490
+    },
+    {
+      "epoch": 0.2312,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00044485680823408815,
+      "loss": 3.298,
+      "step": 5491
+    },
+    {
+      "epoch": 0.2312421052631579,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00044483566591048564,
+      "loss": 3.785,
+      "step": 5492
+    },
+    {
+      "epoch": 0.2312842105263158,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004448145200372062,
+      "loss": 3.5281,
+      "step": 5493
+    },
+    {
+      "epoch": 0.23132631578947369,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004447933706146351,
+      "loss": 3.3212,
+      "step": 5494
+    },
+    {
+      "epoch": 0.23136842105263158,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004447722176431577,
+      "loss": 3.3681,
+      "step": 5495
+    },
+    {
+      "epoch": 0.23141052631578948,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00044475106112315933,
+      "loss": 3.539,
+      "step": 5496
+    },
+    {
+      "epoch": 0.23145263157894738,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004447299010550254,
+      "loss": 3.5777,
+      "step": 5497
+    },
+    {
+      "epoch": 0.23149473684210525,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00044470873743914153,
+      "loss": 3.551,
+      "step": 5498
+    },
+    {
+      "epoch": 0.23153684210526315,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004446875702758932,
+      "loss": 3.1899,
+      "step": 5499
+    },
+    {
+      "epoch": 0.23157894736842105,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004446663995656661,
+      "loss": 3.4669,
+      "step": 5500
+    },
+    {
+      "epoch": 0.23162105263157895,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00044464522530884593,
+      "loss": 3.4734,
+      "step": 5501
+    },
+    {
+      "epoch": 0.23166315789473685,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00044462404750581847,
+      "loss": 3.2174,
+      "step": 5502
+    },
+    {
+      "epoch": 0.23170526315789475,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00044460286615696955,
+      "loss": 3.2266,
+      "step": 5503
+    },
+    {
+      "epoch": 0.23174736842105262,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00044458168126268504,
+      "loss": 3.0194,
+      "step": 5504
+    },
+    {
+      "epoch": 0.23178947368421052,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00044456049282335085,
+      "loss": 3.5788,
+      "step": 5505
+    },
+    {
+      "epoch": 0.23183157894736842,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004445393008393532,
+      "loss": 3.3057,
+      "step": 5506
+    },
+    {
+      "epoch": 0.23187368421052632,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004445181053110779,
+      "loss": 3.0511,
+      "step": 5507
+    },
+    {
+      "epoch": 0.23191578947368421,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004444969062389115,
+      "loss": 3.1072,
+      "step": 5508
+    },
+    {
+      "epoch": 0.2319578947368421,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00044447570362323986,
+      "loss": 3.2669,
+      "step": 5509
+    },
+    {
+      "epoch": 0.232,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00044445449746444935,
+      "loss": 3.2213,
+      "step": 5510
+    },
+    {
+      "epoch": 0.23204210526315788,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004444332877629265,
+      "loss": 3.2559,
+      "step": 5511
+    },
+    {
+      "epoch": 0.23208421052631578,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0004444120745190575,
+      "loss": 3.5853,
+      "step": 5512
+    },
+    {
+      "epoch": 0.23212631578947368,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00044439085773322897,
+      "loss": 3.0958,
+      "step": 5513
+    },
+    {
+      "epoch": 0.23216842105263158,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00044436963740582735,
+      "loss": 3.284,
+      "step": 5514
+    },
+    {
+      "epoch": 0.23221052631578948,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00044434841353723935,
+      "loss": 3.6203,
+      "step": 5515
+    },
+    {
+      "epoch": 0.23225263157894738,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00044432718612785163,
+      "loss": 3.1295,
+      "step": 5516
+    },
+    {
+      "epoch": 0.23229473684210528,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004443059551780509,
+      "loss": 3.1918,
+      "step": 5517
+    },
+    {
+      "epoch": 0.23233684210526315,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0004442847206882239,
+      "loss": 3.3971,
+      "step": 5518
+    },
+    {
+      "epoch": 0.23237894736842105,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00044426348265875766,
+      "loss": 3.1022,
+      "step": 5519
+    },
+    {
+      "epoch": 0.23242105263157894,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00044424224109003896,
+      "loss": 3.1693,
+      "step": 5520
+    },
+    {
+      "epoch": 0.23246315789473684,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0004442209959824549,
+      "loss": 3.4511,
+      "step": 5521
+    },
+    {
+      "epoch": 0.23250526315789474,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004441997473363924,
+      "loss": 3.1361,
+      "step": 5522
+    },
+    {
+      "epoch": 0.23254736842105264,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00044417849515223875,
+      "loss": 3.0397,
+      "step": 5523
+    },
+    {
+      "epoch": 0.2325894736842105,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00044415723943038105,
+      "loss": 3.5254,
+      "step": 5524
+    },
+    {
+      "epoch": 0.2326315789473684,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004441359801712066,
+      "loss": 3.3133,
+      "step": 5525
+    },
+    {
+      "epoch": 0.2326736842105263,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004441147173751027,
+      "loss": 3.3942,
+      "step": 5526
+    },
+    {
+      "epoch": 0.2327157894736842,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00044409345104245664,
+      "loss": 3.2951,
+      "step": 5527
+    },
+    {
+      "epoch": 0.2327578947368421,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.000444072181173656,
+      "loss": 3.2359,
+      "step": 5528
+    },
+    {
+      "epoch": 0.2328,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004440509077690883,
+      "loss": 3.3952,
+      "step": 5529
+    },
+    {
+      "epoch": 0.2328421052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.000444029630829141,
+      "loss": 3.2425,
+      "step": 5530
+    },
+    {
+      "epoch": 0.23288421052631578,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004440083503542018,
+      "loss": 2.7607,
+      "step": 5531
+    },
+    {
+      "epoch": 0.23292631578947368,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00044398706634465846,
+      "loss": 3.7614,
+      "step": 5532
+    },
+    {
+      "epoch": 0.23296842105263157,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004439657788008986,
+      "loss": 3.2997,
+      "step": 5533
+    },
+    {
+      "epoch": 0.23301052631578947,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004439444877233102,
+      "loss": 3.2707,
+      "step": 5534
+    },
+    {
+      "epoch": 0.23305263157894737,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00044392319311228114,
+      "loss": 3.5786,
+      "step": 5535
+    },
+    {
+      "epoch": 0.23309473684210527,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00044390189496819934,
+      "loss": 3.3597,
+      "step": 5536
+    },
+    {
+      "epoch": 0.23313684210526317,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0004438805932914528,
+      "loss": 3.3336,
+      "step": 5537
+    },
+    {
+      "epoch": 0.23317894736842104,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004438592880824297,
+      "loss": 3.1167,
+      "step": 5538
+    },
+    {
+      "epoch": 0.23322105263157894,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00044383797934151813,
+      "loss": 3.4276,
+      "step": 5539
+    },
+    {
+      "epoch": 0.23326315789473684,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0004438166670691063,
+      "loss": 3.4127,
+      "step": 5540
+    },
+    {
+      "epoch": 0.23330526315789474,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00044379535126558257,
+      "loss": 3.1334,
+      "step": 5541
+    },
+    {
+      "epoch": 0.23334736842105264,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004437740319313352,
+      "loss": 3.3359,
+      "step": 5542
+    },
+    {
+      "epoch": 0.23338947368421054,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004437527090667527,
+      "loss": 3.2754,
+      "step": 5543
+    },
+    {
+      "epoch": 0.23343157894736843,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004437313826722234,
+      "loss": 3.2022,
+      "step": 5544
+    },
+    {
+      "epoch": 0.2334736842105263,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00044371005274813604,
+      "loss": 3.3855,
+      "step": 5545
+    },
+    {
+      "epoch": 0.2335157894736842,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00044368871929487895,
+      "loss": 3.1619,
+      "step": 5546
+    },
+    {
+      "epoch": 0.2335578947368421,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00044366738231284107,
+      "loss": 3.4004,
+      "step": 5547
+    },
+    {
+      "epoch": 0.2336,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004436460418024111,
+      "loss": 3.3367,
+      "step": 5548
+    },
+    {
+      "epoch": 0.2336421052631579,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004436246977639777,
+      "loss": 3.4574,
+      "step": 5549
+    },
+    {
+      "epoch": 0.2336842105263158,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00044360335019792983,
+      "loss": 3.1847,
+      "step": 5550
+    },
+    {
+      "epoch": 0.23372631578947367,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00044358199910465634,
+      "loss": 3.7824,
+      "step": 5551
+    },
+    {
+      "epoch": 0.23376842105263157,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00044356064448454637,
+      "loss": 3.3178,
+      "step": 5552
+    },
+    {
+      "epoch": 0.23381052631578947,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00044353928633798887,
+      "loss": 2.8391,
+      "step": 5553
+    },
+    {
+      "epoch": 0.23385263157894737,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004435179246653729,
+      "loss": 3.5665,
+      "step": 5554
+    },
+    {
+      "epoch": 0.23389473684210527,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00044349655946708777,
+      "loss": 3.6181,
+      "step": 5555
+    },
+    {
+      "epoch": 0.23393684210526317,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004434751907435227,
+      "loss": 3.6585,
+      "step": 5556
+    },
+    {
+      "epoch": 0.23397894736842106,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004434538184950669,
+      "loss": 3.349,
+      "step": 5557
+    },
+    {
+      "epoch": 0.23402105263157894,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00044343244272210985,
+      "loss": 3.1035,
+      "step": 5558
+    },
+    {
+      "epoch": 0.23406315789473683,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.000443411063425041,
+      "loss": 3.4502,
+      "step": 5559
+    },
+    {
+      "epoch": 0.23410526315789473,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004433896806042498,
+      "loss": 3.5399,
+      "step": 5560
+    },
+    {
+      "epoch": 0.23414736842105263,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0004433682942601258,
+      "loss": 3.7573,
+      "step": 5561
+    },
+    {
+      "epoch": 0.23418947368421053,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00044334690439305877,
+      "loss": 3.1991,
+      "step": 5562
+    },
+    {
+      "epoch": 0.23423157894736843,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0004433255110034382,
+      "loss": 3.3179,
+      "step": 5563
+    },
+    {
+      "epoch": 0.23427368421052633,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00044330411409165404,
+      "loss": 3.0712,
+      "step": 5564
+    },
+    {
+      "epoch": 0.2343157894736842,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004432827136580961,
+      "loss": 3.1028,
+      "step": 5565
+    },
+    {
+      "epoch": 0.2343578947368421,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00044326130970315415,
+      "loss": 3.5401,
+      "step": 5566
+    },
+    {
+      "epoch": 0.2344,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00044323990222721823,
+      "loss": 3.0381,
+      "step": 5567
+    },
+    {
+      "epoch": 0.2344421052631579,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00044321849123067826,
+      "loss": 3.0968,
+      "step": 5568
+    },
+    {
+      "epoch": 0.2344842105263158,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004431970767139245,
+      "loss": 3.4412,
+      "step": 5569
+    },
+    {
+      "epoch": 0.2345263157894737,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.000443175658677347,
+      "loss": 3.1902,
+      "step": 5570
+    },
+    {
+      "epoch": 0.2345684210526316,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0004431542371213359,
+      "loss": 3.2213,
+      "step": 5571
+    },
+    {
+      "epoch": 0.23461052631578946,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00044313281204628156,
+      "loss": 3.3623,
+      "step": 5572
+    },
+    {
+      "epoch": 0.23465263157894736,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00044311138345257443,
+      "loss": 3.2425,
+      "step": 5573
+    },
+    {
+      "epoch": 0.23469473684210526,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00044308995134060457,
+      "loss": 3.1465,
+      "step": 5574
+    },
+    {
+      "epoch": 0.23473684210526316,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00044306851571076285,
+      "loss": 3.6587,
+      "step": 5575
+    },
+    {
+      "epoch": 0.23477894736842106,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0004430470765634396,
+      "loss": 3.4262,
+      "step": 5576
+    },
+    {
+      "epoch": 0.23482105263157896,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00044302563389902533,
+      "loss": 2.9537,
+      "step": 5577
+    },
+    {
+      "epoch": 0.23486315789473683,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00044300418771791087,
+      "loss": 3.2086,
+      "step": 5578
+    },
+    {
+      "epoch": 0.23490526315789473,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00044298273802048687,
+      "loss": 2.7803,
+      "step": 5579
+    },
+    {
+      "epoch": 0.23494736842105263,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00044296128480714414,
+      "loss": 3.0605,
+      "step": 5580
+    },
+    {
+      "epoch": 0.23498947368421053,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004429398280782735,
+      "loss": 2.8034,
+      "step": 5581
+    },
+    {
+      "epoch": 0.23503157894736842,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00044291836783426597,
+      "loss": 3.5961,
+      "step": 5582
+    },
+    {
+      "epoch": 0.23507368421052632,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00044289690407551234,
+      "loss": 3.3363,
+      "step": 5583
+    },
+    {
+      "epoch": 0.23511578947368422,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00044287543680240384,
+      "loss": 3.5325,
+      "step": 5584
+    },
+    {
+      "epoch": 0.2351578947368421,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00044285396601533147,
+      "loss": 3.241,
+      "step": 5585
+    },
+    {
+      "epoch": 0.2352,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00044283249171468643,
+      "loss": 3.6339,
+      "step": 5586
+    },
+    {
+      "epoch": 0.2352421052631579,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00044281101390086,
+      "loss": 3.0311,
+      "step": 5587
+    },
+    {
+      "epoch": 0.2352842105263158,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00044278953257424335,
+      "loss": 3.4906,
+      "step": 5588
+    },
+    {
+      "epoch": 0.2353263157894737,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00044276804773522795,
+      "loss": 3.3244,
+      "step": 5589
+    },
+    {
+      "epoch": 0.2353684210526316,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0004427465593842053,
+      "loss": 2.8849,
+      "step": 5590
+    },
+    {
+      "epoch": 0.2354105263157895,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004427250675215667,
+      "loss": 2.844,
+      "step": 5591
+    },
+    {
+      "epoch": 0.23545263157894736,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004427035721477039,
+      "loss": 3.3261,
+      "step": 5592
+    },
+    {
+      "epoch": 0.23549473684210526,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004426820732630084,
+      "loss": 3.3929,
+      "step": 5593
+    },
+    {
+      "epoch": 0.23553684210526316,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004426605708678719,
+      "loss": 3.1813,
+      "step": 5594
+    },
+    {
+      "epoch": 0.23557894736842105,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004426390649626862,
+      "loss": 3.0942,
+      "step": 5595
+    },
+    {
+      "epoch": 0.23562105263157895,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004426175555478431,
+      "loss": 3.5262,
+      "step": 5596
+    },
+    {
+      "epoch": 0.23566315789473685,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00044259604262373436,
+      "loss": 2.9177,
+      "step": 5597
+    },
+    {
+      "epoch": 0.23570526315789472,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004425745261907521,
+      "loss": 3.4885,
+      "step": 5598
+    },
+    {
+      "epoch": 0.23574736842105262,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004425530062492882,
+      "loss": 3.5132,
+      "step": 5599
+    },
+    {
+      "epoch": 0.23578947368421052,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004425314827997348,
+      "loss": 3.1141,
+      "step": 5600
+    },
+    {
+      "epoch": 0.23583157894736842,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00044250995584248394,
+      "loss": 3.1155,
+      "step": 5601
+    },
+    {
+      "epoch": 0.23587368421052632,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00044248842537792786,
+      "loss": 3.081,
+      "step": 5602
+    },
+    {
+      "epoch": 0.23591578947368422,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00044246689140645893,
+      "loss": 3.4895,
+      "step": 5603
+    },
+    {
+      "epoch": 0.23595789473684212,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00044244535392846935,
+      "loss": 3.3859,
+      "step": 5604
+    },
+    {
+      "epoch": 0.236,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00044242381294435154,
+      "loss": 3.5936,
+      "step": 5605
+    },
+    {
+      "epoch": 0.2360421052631579,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004424022684544979,
+      "loss": 3.3618,
+      "step": 5606
+    },
+    {
+      "epoch": 0.23608421052631579,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.000442380720459301,
+      "loss": 3.2641,
+      "step": 5607
+    },
+    {
+      "epoch": 0.23612631578947368,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004423591689591534,
+      "loss": 3.5162,
+      "step": 5608
+    },
+    {
+      "epoch": 0.23616842105263158,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00044233761395444783,
+      "loss": 3.2502,
+      "step": 5609
+    },
+    {
+      "epoch": 0.23621052631578948,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00044231605544557684,
+      "loss": 3.3254,
+      "step": 5610
+    },
+    {
+      "epoch": 0.23625263157894738,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004422944934329333,
+      "loss": 3.3557,
+      "step": 5611
+    },
+    {
+      "epoch": 0.23629473684210525,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004422729279169101,
+      "loss": 2.804,
+      "step": 5612
+    },
+    {
+      "epoch": 0.23633684210526315,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00044225135889789996,
+      "loss": 3.3707,
+      "step": 5613
+    },
+    {
+      "epoch": 0.23637894736842105,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.000442229786376296,
+      "loss": 3.6086,
+      "step": 5614
+    },
+    {
+      "epoch": 0.23642105263157895,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00044220821035249115,
+      "loss": 3.3041,
+      "step": 5615
+    },
+    {
+      "epoch": 0.23646315789473685,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004421866308268786,
+      "loss": 3.4646,
+      "step": 5616
+    },
+    {
+      "epoch": 0.23650526315789475,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00044216504779985143,
+      "loss": 3.0456,
+      "step": 5617
+    },
+    {
+      "epoch": 0.23654736842105265,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004421434612718028,
+      "loss": 2.9969,
+      "step": 5618
+    },
+    {
+      "epoch": 0.23658947368421052,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004421218712431262,
+      "loss": 2.7777,
+      "step": 5619
+    },
+    {
+      "epoch": 0.23663157894736842,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004421002777142148,
+      "loss": 3.0334,
+      "step": 5620
+    },
+    {
+      "epoch": 0.23667368421052631,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.000442078680685462,
+      "loss": 3.2926,
+      "step": 5621
+    },
+    {
+      "epoch": 0.2367157894736842,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004420570801572613,
+      "loss": 3.4331,
+      "step": 5622
+    },
+    {
+      "epoch": 0.2367578947368421,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004420354761300063,
+      "loss": 3.0736,
+      "step": 5623
+    },
+    {
+      "epoch": 0.2368,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004420138686040906,
+      "loss": 3.0868,
+      "step": 5624
+    },
+    {
+      "epoch": 0.23684210526315788,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004419922575799078,
+      "loss": 3.392,
+      "step": 5625
+    },
+    {
+      "epoch": 0.23688421052631578,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00044197064305785157,
+      "loss": 3.2685,
+      "step": 5626
+    },
+    {
+      "epoch": 0.23692631578947368,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00044194902503831584,
+      "loss": 3.1218,
+      "step": 5627
+    },
+    {
+      "epoch": 0.23696842105263158,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00044192740352169436,
+      "loss": 3.7138,
+      "step": 5628
+    },
+    {
+      "epoch": 0.23701052631578948,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004419057785083812,
+      "loss": 3.2949,
+      "step": 5629
+    },
+    {
+      "epoch": 0.23705263157894738,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00044188414999877016,
+      "loss": 3.2364,
+      "step": 5630
+    },
+    {
+      "epoch": 0.23709473684210527,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00044186251799325534,
+      "loss": 3.7482,
+      "step": 5631
+    },
+    {
+      "epoch": 0.23713684210526315,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004418408824922309,
+      "loss": 3.507,
+      "step": 5632
+    },
+    {
+      "epoch": 0.23717894736842104,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00044181924349609105,
+      "loss": 2.9173,
+      "step": 5633
+    },
+    {
+      "epoch": 0.23722105263157894,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00044179760100522986,
+      "loss": 3.4549,
+      "step": 5634
+    },
+    {
+      "epoch": 0.23726315789473684,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00044177595502004177,
+      "loss": 3.0569,
+      "step": 5635
+    },
+    {
+      "epoch": 0.23730526315789474,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0004417543055409211,
+      "loss": 3.2633,
+      "step": 5636
+    },
+    {
+      "epoch": 0.23734736842105264,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00044173265256826236,
+      "loss": 3.1601,
+      "step": 5637
+    },
+    {
+      "epoch": 0.23738947368421054,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0004417109961024598,
+      "loss": 2.8502,
+      "step": 5638
+    },
+    {
+      "epoch": 0.2374315789473684,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004416893361439083,
+      "loss": 3.1021,
+      "step": 5639
+    },
+    {
+      "epoch": 0.2374736842105263,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00044166767269300224,
+      "loss": 3.3733,
+      "step": 5640
+    },
+    {
+      "epoch": 0.2375157894736842,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004416460057501364,
+      "loss": 3.1021,
+      "step": 5641
+    },
+    {
+      "epoch": 0.2375578947368421,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00044162433531570554,
+      "loss": 3.3575,
+      "step": 5642
+    },
+    {
+      "epoch": 0.2376,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004416026613901044,
+      "loss": 3.6462,
+      "step": 5643
+    },
+    {
+      "epoch": 0.2376421052631579,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004415809839737279,
+      "loss": 3.2634,
+      "step": 5644
+    },
+    {
+      "epoch": 0.23768421052631578,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.000441559303066971,
+      "loss": 3.3765,
+      "step": 5645
+    },
+    {
+      "epoch": 0.23772631578947367,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0004415376186702287,
+      "loss": 3.5372,
+      "step": 5646
+    },
+    {
+      "epoch": 0.23776842105263157,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00044151593078389594,
+      "loss": 3.2292,
+      "step": 5647
+    },
+    {
+      "epoch": 0.23781052631578947,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.000441494239408368,
+      "loss": 3.3993,
+      "step": 5648
+    },
+    {
+      "epoch": 0.23785263157894737,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00044147254454403995,
+      "loss": 3.7321,
+      "step": 5649
+    },
+    {
+      "epoch": 0.23789473684210527,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00044145084619130715,
+      "loss": 3.308,
+      "step": 5650
+    },
+    {
+      "epoch": 0.23793684210526317,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00044142914435056494,
+      "loss": 3.3918,
+      "step": 5651
+    },
+    {
+      "epoch": 0.23797894736842104,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004414074390222086,
+      "loss": 3.5144,
+      "step": 5652
+    },
+    {
+      "epoch": 0.23802105263157894,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004413857302066336,
+      "loss": 3.1241,
+      "step": 5653
+    },
+    {
+      "epoch": 0.23806315789473684,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00044136401790423547,
+      "loss": 3.5303,
+      "step": 5654
+    },
+    {
+      "epoch": 0.23810526315789474,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004413423021154098,
+      "loss": 3.3143,
+      "step": 5655
+    },
+    {
+      "epoch": 0.23814736842105264,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00044132058284055217,
+      "loss": 3.1831,
+      "step": 5656
+    },
+    {
+      "epoch": 0.23818947368421053,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0004412988600800583,
+      "loss": 3.2469,
+      "step": 5657
+    },
+    {
+      "epoch": 0.23823157894736843,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.000441277133834324,
+      "loss": 3.5906,
+      "step": 5658
+    },
+    {
+      "epoch": 0.2382736842105263,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00044125540410374503,
+      "loss": 3.3973,
+      "step": 5659
+    },
+    {
+      "epoch": 0.2383157894736842,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004412336708887174,
+      "loss": 3.5825,
+      "step": 5660
+    },
+    {
+      "epoch": 0.2383578947368421,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004412119341896369,
+      "loss": 3.2473,
+      "step": 5661
+    },
+    {
+      "epoch": 0.2384,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0004411901940068996,
+      "loss": 3.0988,
+      "step": 5662
+    },
+    {
+      "epoch": 0.2384421052631579,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004411684503409017,
+      "loss": 2.9111,
+      "step": 5663
+    },
+    {
+      "epoch": 0.2384842105263158,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00044114670319203916,
+      "loss": 3.5432,
+      "step": 5664
+    },
+    {
+      "epoch": 0.2385263157894737,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0004411249525607083,
+      "loss": 3.7182,
+      "step": 5665
+    },
+    {
+      "epoch": 0.23856842105263157,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0004411031984473054,
+      "loss": 3.716,
+      "step": 5666
+    },
+    {
+      "epoch": 0.23861052631578947,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004410814408522267,
+      "loss": 3.7485,
+      "step": 5667
+    },
+    {
+      "epoch": 0.23865263157894737,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00044105967977586867,
+      "loss": 3.5136,
+      "step": 5668
+    },
+    {
+      "epoch": 0.23869473684210527,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0004410379152186278,
+      "loss": 2.983,
+      "step": 5669
+    },
+    {
+      "epoch": 0.23873684210526316,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00044101614718090057,
+      "loss": 3.216,
+      "step": 5670
+    },
+    {
+      "epoch": 0.23877894736842106,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00044099437566308354,
+      "loss": 2.8832,
+      "step": 5671
+    },
+    {
+      "epoch": 0.23882105263157893,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004409726006655734,
+      "loss": 3.5001,
+      "step": 5672
+    },
+    {
+      "epoch": 0.23886315789473683,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004409508221887669,
+      "loss": 3.0544,
+      "step": 5673
+    },
+    {
+      "epoch": 0.23890526315789473,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0004409290402330608,
+      "loss": 3.0664,
+      "step": 5674
+    },
+    {
+      "epoch": 0.23894736842105263,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00044090725479885187,
+      "loss": 3.3974,
+      "step": 5675
+    },
+    {
+      "epoch": 0.23898947368421053,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.000440885465886537,
+      "loss": 3.3597,
+      "step": 5676
+    },
+    {
+      "epoch": 0.23903157894736843,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00044086367349651333,
+      "loss": 3.4833,
+      "step": 5677
+    },
+    {
+      "epoch": 0.23907368421052633,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0004408418776291777,
+      "loss": 3.2379,
+      "step": 5678
+    },
+    {
+      "epoch": 0.2391157894736842,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0004408200782849274,
+      "loss": 3.5677,
+      "step": 5679
+    },
+    {
+      "epoch": 0.2391578947368421,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004407982754641594,
+      "loss": 3.1961,
+      "step": 5680
+    },
+    {
+      "epoch": 0.2392,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.000440776469167271,
+      "loss": 3.1835,
+      "step": 5681
+    },
+    {
+      "epoch": 0.2392421052631579,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004407546593946595,
+      "loss": 2.9983,
+      "step": 5682
+    },
+    {
+      "epoch": 0.2392842105263158,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00044073284614672224,
+      "loss": 3.4474,
+      "step": 5683
+    },
+    {
+      "epoch": 0.2393263157894737,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004407110294238566,
+      "loss": 3.2739,
+      "step": 5684
+    },
+    {
+      "epoch": 0.2393684210526316,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004406892092264601,
+      "loss": 3.0298,
+      "step": 5685
+    },
+    {
+      "epoch": 0.23941052631578946,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00044066738555493027,
+      "loss": 3.0882,
+      "step": 5686
+    },
+    {
+      "epoch": 0.23945263157894736,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004406455584096647,
+      "loss": 2.8994,
+      "step": 5687
+    },
+    {
+      "epoch": 0.23949473684210526,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00044062372779106105,
+      "loss": 3.1491,
+      "step": 5688
+    },
+    {
+      "epoch": 0.23953684210526316,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.000440601893699517,
+      "loss": 3.1292,
+      "step": 5689
+    },
+    {
+      "epoch": 0.23957894736842106,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0004405800561354304,
+      "loss": 3.425,
+      "step": 5690
+    },
+    {
+      "epoch": 0.23962105263157896,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00044055821509919916,
+      "loss": 3.1172,
+      "step": 5691
+    },
+    {
+      "epoch": 0.23966315789473686,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00044053637059122107,
+      "loss": 3.8108,
+      "step": 5692
+    },
+    {
+      "epoch": 0.23970526315789473,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004405145226118943,
+      "loss": 3.2928,
+      "step": 5693
+    },
+    {
+      "epoch": 0.23974736842105263,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0004404926711616167,
+      "loss": 2.7366,
+      "step": 5694
+    },
+    {
+      "epoch": 0.23978947368421052,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00044047081624078643,
+      "loss": 3.2447,
+      "step": 5695
+    },
+    {
+      "epoch": 0.23983157894736842,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00044044895784980167,
+      "loss": 3.202,
+      "step": 5696
+    },
+    {
+      "epoch": 0.23987368421052632,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00044042709598906074,
+      "loss": 3.3263,
+      "step": 5697
+    },
+    {
+      "epoch": 0.23991578947368422,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00044040523065896176,
+      "loss": 3.103,
+      "step": 5698
+    },
+    {
+      "epoch": 0.2399578947368421,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00044038336185990324,
+      "loss": 3.2397,
+      "step": 5699
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004403614895922836,
+      "loss": 3.2629,
+      "step": 5700
+    },
+    {
+      "epoch": 0.2400421052631579,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00044033961385650126,
+      "loss": 3.5444,
+      "step": 5701
+    },
+    {
+      "epoch": 0.2400842105263158,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00044031773465295476,
+      "loss": 2.864,
+      "step": 5702
+    },
+    {
+      "epoch": 0.2401263157894737,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004402958519820428,
+      "loss": 2.922,
+      "step": 5703
+    },
+    {
+      "epoch": 0.2401684210526316,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.000440273965844164,
+      "loss": 3.4326,
+      "step": 5704
+    },
+    {
+      "epoch": 0.24021052631578949,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004402520762397171,
+      "loss": 3.035,
+      "step": 5705
+    },
+    {
+      "epoch": 0.24025263157894736,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00044023018316910093,
+      "loss": 2.7112,
+      "step": 5706
+    },
+    {
+      "epoch": 0.24029473684210526,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004402082866327143,
+      "loss": 3.4546,
+      "step": 5707
+    },
+    {
+      "epoch": 0.24033684210526315,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00044018638663095616,
+      "loss": 3.508,
+      "step": 5708
+    },
+    {
+      "epoch": 0.24037894736842105,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00044016448316422553,
+      "loss": 3.0937,
+      "step": 5709
+    },
+    {
+      "epoch": 0.24042105263157895,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00044014257623292143,
+      "loss": 3.1282,
+      "step": 5710
+    },
+    {
+      "epoch": 0.24046315789473685,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.000440120665837443,
+      "loss": 3.3809,
+      "step": 5711
+    },
+    {
+      "epoch": 0.24050526315789475,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004400987519781895,
+      "loss": 3.7989,
+      "step": 5712
+    },
+    {
+      "epoch": 0.24054736842105262,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00044007683465556,
+      "loss": 3.4561,
+      "step": 5713
+    },
+    {
+      "epoch": 0.24058947368421052,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.000440054913869954,
+      "loss": 3.6463,
+      "step": 5714
+    },
+    {
+      "epoch": 0.24063157894736842,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004400329896217707,
+      "loss": 3.1636,
+      "step": 5715
+    },
+    {
+      "epoch": 0.24067368421052632,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004400110619114096,
+      "loss": 3.5094,
+      "step": 5716
+    },
+    {
+      "epoch": 0.24071578947368422,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00043998913073927027,
+      "loss": 3.3629,
+      "step": 5717
+    },
+    {
+      "epoch": 0.24075789473684212,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00043996719610575217,
+      "loss": 3.2755,
+      "step": 5718
+    },
+    {
+      "epoch": 0.2408,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00043994525801125496,
+      "loss": 3.3365,
+      "step": 5719
+    },
+    {
+      "epoch": 0.24084210526315789,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004399233164561783,
+      "loss": 3.2219,
+      "step": 5720
+    },
+    {
+      "epoch": 0.24088421052631578,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.000439901371440922,
+      "loss": 3.4116,
+      "step": 5721
+    },
+    {
+      "epoch": 0.24092631578947368,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0004398794229658858,
+      "loss": 3.0025,
+      "step": 5722
+    },
+    {
+      "epoch": 0.24096842105263158,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004398574710314697,
+      "loss": 3.6059,
+      "step": 5723
+    },
+    {
+      "epoch": 0.24101052631578948,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00043983551563807344,
+      "loss": 2.8624,
+      "step": 5724
+    },
+    {
+      "epoch": 0.24105263157894738,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004398135567860972,
+      "loss": 3.3028,
+      "step": 5725
+    },
+    {
+      "epoch": 0.24109473684210525,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00043979159447594094,
+      "loss": 2.9742,
+      "step": 5726
+    },
+    {
+      "epoch": 0.24113684210526315,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00043976962870800483,
+      "loss": 3.2551,
+      "step": 5727
+    },
+    {
+      "epoch": 0.24117894736842105,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004397476594826891,
+      "loss": 3.4834,
+      "step": 5728
+    },
+    {
+      "epoch": 0.24122105263157895,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0004397256868003939,
+      "loss": 3.0787,
+      "step": 5729
+    },
+    {
+      "epoch": 0.24126315789473685,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00043970371066151965,
+      "loss": 3.5561,
+      "step": 5730
+    },
+    {
+      "epoch": 0.24130526315789474,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004396817310664666,
+      "loss": 3.5463,
+      "step": 5731
+    },
+    {
+      "epoch": 0.24134736842105264,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00043965974801563534,
+      "loss": 3.3084,
+      "step": 5732
+    },
+    {
+      "epoch": 0.24138947368421051,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00043963776150942636,
+      "loss": 3.1025,
+      "step": 5733
+    },
+    {
+      "epoch": 0.2414315789473684,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004396157715482402,
+      "loss": 3.0449,
+      "step": 5734
+    },
+    {
+      "epoch": 0.2414736842105263,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00043959377813247735,
+      "loss": 3.1708,
+      "step": 5735
+    },
+    {
+      "epoch": 0.2415157894736842,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004395717812625387,
+      "loss": 3.8698,
+      "step": 5736
+    },
+    {
+      "epoch": 0.2415578947368421,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004395497809388249,
+      "loss": 3.2482,
+      "step": 5737
+    },
+    {
+      "epoch": 0.2416,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00043952777716173687,
+      "loss": 3.3577,
+      "step": 5738
+    },
+    {
+      "epoch": 0.2416421052631579,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00043950576993167536,
+      "loss": 3.3885,
+      "step": 5739
+    },
+    {
+      "epoch": 0.24168421052631578,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00043948375924904136,
+      "loss": 3.5907,
+      "step": 5740
+    },
+    {
+      "epoch": 0.24172631578947368,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.000439461745114236,
+      "loss": 3.4877,
+      "step": 5741
+    },
+    {
+      "epoch": 0.24176842105263158,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00043943972752766015,
+      "loss": 3.7744,
+      "step": 5742
+    },
+    {
+      "epoch": 0.24181052631578948,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004394177064897151,
+      "loss": 3.2848,
+      "step": 5743
+    },
+    {
+      "epoch": 0.24185263157894737,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00043939568200080195,
+      "loss": 3.1925,
+      "step": 5744
+    },
+    {
+      "epoch": 0.24189473684210527,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00043937365406132204,
+      "loss": 3.2824,
+      "step": 5745
+    },
+    {
+      "epoch": 0.24193684210526314,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004393516226716767,
+      "loss": 3.2666,
+      "step": 5746
+    },
+    {
+      "epoch": 0.24197894736842104,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00043932958783226715,
+      "loss": 3.4416,
+      "step": 5747
+    },
+    {
+      "epoch": 0.24202105263157894,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004393075495434951,
+      "loss": 3.4158,
+      "step": 5748
+    },
+    {
+      "epoch": 0.24206315789473684,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00043928550780576183,
+      "loss": 3.6666,
+      "step": 5749
+    },
+    {
+      "epoch": 0.24210526315789474,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00043926346261946906,
+      "loss": 2.8852,
+      "step": 5750
+    },
+    {
+      "epoch": 0.24214736842105264,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004392414139850184,
+      "loss": 2.9076,
+      "step": 5751
+    },
+    {
+      "epoch": 0.24218947368421054,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0004392193619028114,
+      "loss": 2.988,
+      "step": 5752
+    },
+    {
+      "epoch": 0.2422315789473684,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00043919730637325007,
+      "loss": 3.5439,
+      "step": 5753
+    },
+    {
+      "epoch": 0.2422736842105263,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004391752473967361,
+      "loss": 3.2195,
+      "step": 5754
+    },
+    {
+      "epoch": 0.2423157894736842,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0004391531849736714,
+      "loss": 3.0817,
+      "step": 5755
+    },
+    {
+      "epoch": 0.2423578947368421,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004391311191044579,
+      "loss": 3.3419,
+      "step": 5756
+    },
+    {
+      "epoch": 0.2424,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00043910904978949764,
+      "loss": 3.4587,
+      "step": 5757
+    },
+    {
+      "epoch": 0.2424421052631579,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004390869770291926,
+      "loss": 3.6578,
+      "step": 5758
+    },
+    {
+      "epoch": 0.2424842105263158,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00043906490082394514,
+      "loss": 3.1426,
+      "step": 5759
+    },
+    {
+      "epoch": 0.24252631578947367,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004390428211741572,
+      "loss": 3.2086,
+      "step": 5760
+    },
+    {
+      "epoch": 0.24256842105263157,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00043902073808023125,
+      "loss": 3.4118,
+      "step": 5761
+    },
+    {
+      "epoch": 0.24261052631578947,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004389986515425696,
+      "loss": 3.3434,
+      "step": 5762
+    },
+    {
+      "epoch": 0.24265263157894737,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00043897656156157447,
+      "loss": 3.5494,
+      "step": 5763
+    },
+    {
+      "epoch": 0.24269473684210527,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004389544681376485,
+      "loss": 3.5953,
+      "step": 5764
+    },
+    {
+      "epoch": 0.24273684210526317,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004389323712711941,
+      "loss": 3.1458,
+      "step": 5765
+    },
+    {
+      "epoch": 0.24277894736842107,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004389102709626138,
+      "loss": 3.3899,
+      "step": 5766
+    },
+    {
+      "epoch": 0.24282105263157894,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004388881672123105,
+      "loss": 3.423,
+      "step": 5767
+    },
+    {
+      "epoch": 0.24286315789473684,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004388660600206866,
+      "loss": 3.1283,
+      "step": 5768
+    },
+    {
+      "epoch": 0.24290526315789474,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00043884394938814504,
+      "loss": 4.1053,
+      "step": 5769
+    },
+    {
+      "epoch": 0.24294736842105263,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004388218353150886,
+      "loss": 3.3953,
+      "step": 5770
+    },
+    {
+      "epoch": 0.24298947368421053,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004387997178019202,
+      "loss": 3.7624,
+      "step": 5771
+    },
+    {
+      "epoch": 0.24303157894736843,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00043877759684904274,
+      "loss": 3.4884,
+      "step": 5772
+    },
+    {
+      "epoch": 0.2430736842105263,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00043875547245685926,
+      "loss": 3.4067,
+      "step": 5773
+    },
+    {
+      "epoch": 0.2431157894736842,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00043873334462577287,
+      "loss": 3.3116,
+      "step": 5774
+    },
+    {
+      "epoch": 0.2431578947368421,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00043871121335618666,
+      "loss": 3.3819,
+      "step": 5775
+    },
+    {
+      "epoch": 0.2432,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0004386890786485039,
+      "loss": 3.2145,
+      "step": 5776
+    },
+    {
+      "epoch": 0.2432421052631579,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00043866694050312784,
+      "loss": 3.0936,
+      "step": 5777
+    },
+    {
+      "epoch": 0.2432842105263158,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004386447989204618,
+      "loss": 2.9633,
+      "step": 5778
+    },
+    {
+      "epoch": 0.2433263157894737,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00043862265390090917,
+      "loss": 3.9013,
+      "step": 5779
+    },
+    {
+      "epoch": 0.24336842105263157,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004386005054448734,
+      "loss": 3.2631,
+      "step": 5780
+    },
+    {
+      "epoch": 0.24341052631578947,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.000438578353552758,
+      "loss": 3.4248,
+      "step": 5781
+    },
+    {
+      "epoch": 0.24345263157894736,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00043855619822496655,
+      "loss": 3.2423,
+      "step": 5782
+    },
+    {
+      "epoch": 0.24349473684210526,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00043853403946190274,
+      "loss": 3.1958,
+      "step": 5783
+    },
+    {
+      "epoch": 0.24353684210526316,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004385118772639703,
+      "loss": 3.4908,
+      "step": 5784
+    },
+    {
+      "epoch": 0.24357894736842106,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004384897116315729,
+      "loss": 3.2035,
+      "step": 5785
+    },
+    {
+      "epoch": 0.24362105263157896,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004384675425651144,
+      "loss": 3.335,
+      "step": 5786
+    },
+    {
+      "epoch": 0.24366315789473683,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00043844537006499875,
+      "loss": 3.5487,
+      "step": 5787
+    },
+    {
+      "epoch": 0.24370526315789473,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004384231941316299,
+      "loss": 3.3652,
+      "step": 5788
+    },
+    {
+      "epoch": 0.24374736842105263,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00043840101476541176,
+      "loss": 2.6983,
+      "step": 5789
+    },
+    {
+      "epoch": 0.24378947368421053,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00043837883196674854,
+      "loss": 3.6199,
+      "step": 5790
+    },
+    {
+      "epoch": 0.24383157894736843,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004383566457360443,
+      "loss": 3.2554,
+      "step": 5791
+    },
+    {
+      "epoch": 0.24387368421052633,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00043833445607370335,
+      "loss": 3.0794,
+      "step": 5792
+    },
+    {
+      "epoch": 0.2439157894736842,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0004383122629801298,
+      "loss": 2.9403,
+      "step": 5793
+    },
+    {
+      "epoch": 0.2439578947368421,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004382900664557282,
+      "loss": 3.3505,
+      "step": 5794
+    },
+    {
+      "epoch": 0.244,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00043826786650090276,
+      "loss": 3.1737,
+      "step": 5795
+    },
+    {
+      "epoch": 0.2440421052631579,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00043824566311605797,
+      "loss": 3.395,
+      "step": 5796
+    },
+    {
+      "epoch": 0.2440842105263158,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00043822345630159846,
+      "loss": 2.9132,
+      "step": 5797
+    },
+    {
+      "epoch": 0.2441263157894737,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004382012460579287,
+      "loss": 2.9797,
+      "step": 5798
+    },
+    {
+      "epoch": 0.2441684210526316,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0004381790323854533,
+      "loss": 3.3223,
+      "step": 5799
+    },
+    {
+      "epoch": 0.24421052631578946,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00043815681528457703,
+      "loss": 3.4199,
+      "step": 5800
+    },
+    {
+      "epoch": 0.24425263157894736,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004381345947557047,
+      "loss": 2.8119,
+      "step": 5801
+    },
+    {
+      "epoch": 0.24429473684210526,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00043811237079924113,
+      "loss": 3.2709,
+      "step": 5802
+    },
+    {
+      "epoch": 0.24433684210526316,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004380901434155911,
+      "loss": 3.1417,
+      "step": 5803
+    },
+    {
+      "epoch": 0.24437894736842106,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004380679126051598,
+      "loss": 3.3475,
+      "step": 5804
+    },
+    {
+      "epoch": 0.24442105263157896,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00043804567836835197,
+      "loss": 3.3515,
+      "step": 5805
+    },
+    {
+      "epoch": 0.24446315789473685,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004380234407055729,
+      "loss": 2.8609,
+      "step": 5806
+    },
+    {
+      "epoch": 0.24450526315789473,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004380011996172276,
+      "loss": 3.1671,
+      "step": 5807
+    },
+    {
+      "epoch": 0.24454736842105262,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004379789551037213,
+      "loss": 3.1463,
+      "step": 5808
+    },
+    {
+      "epoch": 0.24458947368421052,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004379567071654594,
+      "loss": 3.5578,
+      "step": 5809
+    },
+    {
+      "epoch": 0.24463157894736842,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00043793445580284714,
+      "loss": 3.4545,
+      "step": 5810
+    },
+    {
+      "epoch": 0.24467368421052632,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004379122010162898,
+      "loss": 3.3313,
+      "step": 5811
+    },
+    {
+      "epoch": 0.24471578947368422,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004378899428061931,
+      "loss": 2.9929,
+      "step": 5812
+    },
+    {
+      "epoch": 0.24475789473684212,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00043786768117296226,
+      "loss": 3.3065,
+      "step": 5813
+    },
+    {
+      "epoch": 0.2448,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0004378454161170031,
+      "loss": 3.2392,
+      "step": 5814
+    },
+    {
+      "epoch": 0.2448421052631579,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004378231476387211,
+      "loss": 3.2017,
+      "step": 5815
+    },
+    {
+      "epoch": 0.2448842105263158,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00043780087573852213,
+      "loss": 3.3878,
+      "step": 5816
+    },
+    {
+      "epoch": 0.2449263157894737,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004377786004168118,
+      "loss": 3.7428,
+      "step": 5817
+    },
+    {
+      "epoch": 0.24496842105263159,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.000437756321673996,
+      "loss": 3.2103,
+      "step": 5818
+    },
+    {
+      "epoch": 0.24501052631578948,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004377340395104806,
+      "loss": 3.3718,
+      "step": 5819
+    },
+    {
+      "epoch": 0.24505263157894736,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004377117539266716,
+      "loss": 2.4851,
+      "step": 5820
+    },
+    {
+      "epoch": 0.24509473684210525,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.000437689464922975,
+      "loss": 3.379,
+      "step": 5821
+    },
+    {
+      "epoch": 0.24513684210526315,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00043766717249979687,
+      "loss": 3.4554,
+      "step": 5822
+    },
+    {
+      "epoch": 0.24517894736842105,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004376448766575434,
+      "loss": 3.4001,
+      "step": 5823
+    },
+    {
+      "epoch": 0.24522105263157895,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0004376225773966206,
+      "loss": 3.4533,
+      "step": 5824
+    },
+    {
+      "epoch": 0.24526315789473685,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00043760027471743506,
+      "loss": 3.8316,
+      "step": 5825
+    },
+    {
+      "epoch": 0.24530526315789475,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004375779686203929,
+      "loss": 3.3495,
+      "step": 5826
+    },
+    {
+      "epoch": 0.24534736842105262,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004375556591059005,
+      "loss": 3.1279,
+      "step": 5827
+    },
+    {
+      "epoch": 0.24538947368421052,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00043753334617436433,
+      "loss": 3.8785,
+      "step": 5828
+    },
+    {
+      "epoch": 0.24543157894736842,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00043751102982619107,
+      "loss": 3.7162,
+      "step": 5829
+    },
+    {
+      "epoch": 0.24547368421052632,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.000437488710061787,
+      "loss": 3.359,
+      "step": 5830
+    },
+    {
+      "epoch": 0.24551578947368422,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00043746638688155903,
+      "loss": 3.3231,
+      "step": 5831
+    },
+    {
+      "epoch": 0.24555789473684211,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004374440602859138,
+      "loss": 2.9204,
+      "step": 5832
+    },
+    {
+      "epoch": 0.2456,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00043742173027525787,
+      "loss": 3.3773,
+      "step": 5833
+    },
+    {
+      "epoch": 0.24564210526315788,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004373993968499983,
+      "loss": 4.0035,
+      "step": 5834
+    },
+    {
+      "epoch": 0.24568421052631578,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00043737706001054197,
+      "loss": 3.3762,
+      "step": 5835
+    },
+    {
+      "epoch": 0.24572631578947368,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004373547197572957,
+      "loss": 3.3595,
+      "step": 5836
+    },
+    {
+      "epoch": 0.24576842105263158,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00043733237609066655,
+      "loss": 3.4537,
+      "step": 5837
+    },
+    {
+      "epoch": 0.24581052631578948,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00043731002901106167,
+      "loss": 2.9956,
+      "step": 5838
+    },
+    {
+      "epoch": 0.24585263157894738,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004372876785188881,
+      "loss": 3.4069,
+      "step": 5839
+    },
+    {
+      "epoch": 0.24589473684210525,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00043726532461455317,
+      "loss": 3.3592,
+      "step": 5840
+    },
+    {
+      "epoch": 0.24593684210526315,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0004372429672984639,
+      "loss": 3.2404,
+      "step": 5841
+    },
+    {
+      "epoch": 0.24597894736842105,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004372206065710279,
+      "loss": 3.1899,
+      "step": 5842
+    },
+    {
+      "epoch": 0.24602105263157895,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004371982424326523,
+      "loss": 3.5338,
+      "step": 5843
+    },
+    {
+      "epoch": 0.24606315789473684,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00043717587488374476,
+      "loss": 3.0906,
+      "step": 5844
+    },
+    {
+      "epoch": 0.24610526315789474,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004371535039247127,
+      "loss": 3.0455,
+      "step": 5845
+    },
+    {
+      "epoch": 0.24614736842105264,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00043713112955596367,
+      "loss": 3.4332,
+      "step": 5846
+    },
+    {
+      "epoch": 0.2461894736842105,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004371087517779053,
+      "loss": 2.9745,
+      "step": 5847
+    },
+    {
+      "epoch": 0.2462315789473684,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004370863705909454,
+      "loss": 3.3057,
+      "step": 5848
+    },
+    {
+      "epoch": 0.2462736842105263,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004370639859954916,
+      "loss": 3.385,
+      "step": 5849
+    },
+    {
+      "epoch": 0.2463157894736842,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00043704159799195174,
+      "loss": 3.521,
+      "step": 5850
+    },
+    {
+      "epoch": 0.2463578947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00043701920658073377,
+      "loss": 3.3314,
+      "step": 5851
+    },
+    {
+      "epoch": 0.2464,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00043699681176224557,
+      "loss": 2.798,
+      "step": 5852
+    },
+    {
+      "epoch": 0.2464421052631579,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00043697441353689523,
+      "loss": 3.4144,
+      "step": 5853
+    },
+    {
+      "epoch": 0.24648421052631578,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004369520119050907,
+      "loss": 3.4753,
+      "step": 5854
+    },
+    {
+      "epoch": 0.24652631578947368,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.0004369296068672402,
+      "loss": 3.4717,
+      "step": 5855
+    },
+    {
+      "epoch": 0.24656842105263158,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00043690719842375187,
+      "loss": 3.7164,
+      "step": 5856
+    },
+    {
+      "epoch": 0.24661052631578947,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.000436884786575034,
+      "loss": 2.9484,
+      "step": 5857
+    },
+    {
+      "epoch": 0.24665263157894737,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00043686237132149487,
+      "loss": 3.4486,
+      "step": 5858
+    },
+    {
+      "epoch": 0.24669473684210527,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.000436839952663543,
+      "loss": 3.3158,
+      "step": 5859
+    },
+    {
+      "epoch": 0.24673684210526317,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004368175306015866,
+      "loss": 2.6301,
+      "step": 5860
+    },
+    {
+      "epoch": 0.24677894736842104,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004367951051360344,
+      "loss": 3.6451,
+      "step": 5861
+    },
+    {
+      "epoch": 0.24682105263157894,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00043677267626729475,
+      "loss": 3.6841,
+      "step": 5862
+    },
+    {
+      "epoch": 0.24686315789473684,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00043675024399577646,
+      "loss": 3.0407,
+      "step": 5863
+    },
+    {
+      "epoch": 0.24690526315789474,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004367278083218881,
+      "loss": 3.1672,
+      "step": 5864
+    },
+    {
+      "epoch": 0.24694736842105264,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004367053692460385,
+      "loss": 3.4539,
+      "step": 5865
+    },
+    {
+      "epoch": 0.24698947368421054,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004366829267686364,
+      "loss": 3.6821,
+      "step": 5866
+    },
+    {
+      "epoch": 0.2470315789473684,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004366604808900907,
+      "loss": 3.3186,
+      "step": 5867
+    },
+    {
+      "epoch": 0.2470736842105263,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00043663803161081053,
+      "loss": 3.4977,
+      "step": 5868
+    },
+    {
+      "epoch": 0.2471157894736842,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004366155789312045,
+      "loss": 3.2481,
+      "step": 5869
+    },
+    {
+      "epoch": 0.2471578947368421,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00043659312285168194,
+      "loss": 3.4353,
+      "step": 5870
+    },
+    {
+      "epoch": 0.2472,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.000436570663372652,
+      "loss": 2.9201,
+      "step": 5871
+    },
+    {
+      "epoch": 0.2472421052631579,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00043654820049452374,
+      "loss": 3.2972,
+      "step": 5872
+    },
+    {
+      "epoch": 0.2472842105263158,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004365257342177065,
+      "loss": 3.0657,
+      "step": 5873
+    },
+    {
+      "epoch": 0.24732631578947367,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004365032645426095,
+      "loss": 3.4154,
+      "step": 5874
+    },
+    {
+      "epoch": 0.24736842105263157,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004364807914696421,
+      "loss": 3.1522,
+      "step": 5875
+    },
+    {
+      "epoch": 0.24741052631578947,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00043645831499921386,
+      "loss": 3.2581,
+      "step": 5876
+    },
+    {
+      "epoch": 0.24745263157894737,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004364358351317341,
+      "loss": 3.457,
+      "step": 5877
+    },
+    {
+      "epoch": 0.24749473684210527,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004364133518676126,
+      "loss": 3.3686,
+      "step": 5878
+    },
+    {
+      "epoch": 0.24753684210526317,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004363908652072588,
+      "loss": 3.2853,
+      "step": 5879
+    },
+    {
+      "epoch": 0.24757894736842107,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004363683751510824,
+      "loss": 3.3215,
+      "step": 5880
+    },
+    {
+      "epoch": 0.24762105263157894,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00043634588169949324,
+      "loss": 3.4715,
+      "step": 5881
+    },
+    {
+      "epoch": 0.24766315789473684,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004363233848529011,
+      "loss": 3.2916,
+      "step": 5882
+    },
+    {
+      "epoch": 0.24770526315789473,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004363008846117157,
+      "loss": 3.4356,
+      "step": 5883
+    },
+    {
+      "epoch": 0.24774736842105263,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004362783809763471,
+      "loss": 3.0956,
+      "step": 5884
+    },
+    {
+      "epoch": 0.24778947368421053,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004362558739472053,
+      "loss": 3.2828,
+      "step": 5885
+    },
+    {
+      "epoch": 0.24783157894736843,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004362333635247003,
+      "loss": 3.5293,
+      "step": 5886
+    },
+    {
+      "epoch": 0.24787368421052633,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004362108497092423,
+      "loss": 3.4688,
+      "step": 5887
+    },
+    {
+      "epoch": 0.2479157894736842,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00043618833250124134,
+      "loss": 3.1629,
+      "step": 5888
+    },
+    {
+      "epoch": 0.2479578947368421,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00043616581190110773,
+      "loss": 3.6413,
+      "step": 5889
+    },
+    {
+      "epoch": 0.248,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004361432879092518,
+      "loss": 3.1462,
+      "step": 5890
+    },
+    {
+      "epoch": 0.2480421052631579,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0004361207605260838,
+      "loss": 3.4458,
+      "step": 5891
+    },
+    {
+      "epoch": 0.2480842105263158,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004360982297520143,
+      "loss": 3.1227,
+      "step": 5892
+    },
+    {
+      "epoch": 0.2481263157894737,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00043607569558745365,
+      "loss": 3.321,
+      "step": 5893
+    },
+    {
+      "epoch": 0.24816842105263157,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00043605315803281247,
+      "loss": 3.7713,
+      "step": 5894
+    },
+    {
+      "epoch": 0.24821052631578946,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00043603061708850135,
+      "loss": 2.9461,
+      "step": 5895
+    },
+    {
+      "epoch": 0.24825263157894736,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00043600807275493103,
+      "loss": 3.2043,
+      "step": 5896
+    },
+    {
+      "epoch": 0.24829473684210526,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.0004359855250325121,
+      "loss": 3.1347,
+      "step": 5897
+    },
+    {
+      "epoch": 0.24833684210526316,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004359629739216555,
+      "loss": 3.2559,
+      "step": 5898
+    },
+    {
+      "epoch": 0.24837894736842106,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00043594041942277196,
+      "loss": 3.4053,
+      "step": 5899
+    },
+    {
+      "epoch": 0.24842105263157896,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00043591786153627247,
+      "loss": 3.2963,
+      "step": 5900
+    },
+    {
+      "epoch": 0.24846315789473683,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00043589530026256796,
+      "loss": 3.0448,
+      "step": 5901
+    },
+    {
+      "epoch": 0.24850526315789473,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0004358727356020695,
+      "loss": 3.1376,
+      "step": 5902
+    },
+    {
+      "epoch": 0.24854736842105263,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004358501675551882,
+      "loss": 3.4717,
+      "step": 5903
+    },
+    {
+      "epoch": 0.24858947368421053,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004358275961223352,
+      "loss": 3.207,
+      "step": 5904
+    },
+    {
+      "epoch": 0.24863157894736843,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004358050213039217,
+      "loss": 3.4567,
+      "step": 5905
+    },
+    {
+      "epoch": 0.24867368421052632,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00043578244310035915,
+      "loss": 2.851,
+      "step": 5906
+    },
+    {
+      "epoch": 0.24871578947368422,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00043575986151205866,
+      "loss": 3.0597,
+      "step": 5907
+    },
+    {
+      "epoch": 0.2487578947368421,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00043573727653943177,
+      "loss": 3.1925,
+      "step": 5908
+    },
+    {
+      "epoch": 0.2488,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00043571468818288993,
+      "loss": 3.4693,
+      "step": 5909
+    },
+    {
+      "epoch": 0.2488421052631579,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00043569209644284476,
+      "loss": 3.0362,
+      "step": 5910
+    },
+    {
+      "epoch": 0.2488842105263158,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004356695013197077,
+      "loss": 3.2697,
+      "step": 5911
+    },
+    {
+      "epoch": 0.2489263157894737,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00043564690281389047,
+      "loss": 3.5087,
+      "step": 5912
+    },
+    {
+      "epoch": 0.2489684210526316,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00043562430092580485,
+      "loss": 3.5611,
+      "step": 5913
+    },
+    {
+      "epoch": 0.24901052631578946,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004356016956558625,
+      "loss": 2.921,
+      "step": 5914
+    },
+    {
+      "epoch": 0.24905263157894736,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00043557908700447536,
+      "loss": 3.0882,
+      "step": 5915
+    },
+    {
+      "epoch": 0.24909473684210526,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004355564749720553,
+      "loss": 3.4369,
+      "step": 5916
+    },
+    {
+      "epoch": 0.24913684210526316,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00043553385955901424,
+      "loss": 3.4472,
+      "step": 5917
+    },
+    {
+      "epoch": 0.24917894736842106,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00043551124076576434,
+      "loss": 3.3542,
+      "step": 5918
+    },
+    {
+      "epoch": 0.24922105263157895,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004354886185927176,
+      "loss": 3.7573,
+      "step": 5919
+    },
+    {
+      "epoch": 0.24926315789473685,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00043546599304028613,
+      "loss": 3.6567,
+      "step": 5920
+    },
+    {
+      "epoch": 0.24930526315789472,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00043544336410888214,
+      "loss": 3.8174,
+      "step": 5921
+    },
+    {
+      "epoch": 0.24934736842105262,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.000435420731798918,
+      "loss": 3.4057,
+      "step": 5922
+    },
+    {
+      "epoch": 0.24938947368421052,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0004353980961108059,
+      "loss": 3.3162,
+      "step": 5923
+    },
+    {
+      "epoch": 0.24943157894736842,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00043537545704495846,
+      "loss": 3.4463,
+      "step": 5924
+    },
+    {
+      "epoch": 0.24947368421052632,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004353528146017879,
+      "loss": 3.4923,
+      "step": 5925
+    },
+    {
+      "epoch": 0.24951578947368422,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00043533016878170687,
+      "loss": 2.9915,
+      "step": 5926
+    },
+    {
+      "epoch": 0.24955789473684212,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004353075195851279,
+      "loss": 2.6857,
+      "step": 5927
+    },
+    {
+      "epoch": 0.2496,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00043528486701246375,
+      "loss": 3.4822,
+      "step": 5928
+    },
+    {
+      "epoch": 0.2496421052631579,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004352622110641269,
+      "loss": 2.8233,
+      "step": 5929
+    },
+    {
+      "epoch": 0.2496842105263158,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004352395517405303,
+      "loss": 3.4272,
+      "step": 5930
+    },
+    {
+      "epoch": 0.24972631578947369,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004352168890420867,
+      "loss": 3.3662,
+      "step": 5931
+    },
+    {
+      "epoch": 0.24976842105263158,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.000435194222969209,
+      "loss": 3.4941,
+      "step": 5932
+    },
+    {
+      "epoch": 0.24981052631578948,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004351715535223101,
+      "loss": 3.5766,
+      "step": 5933
+    },
+    {
+      "epoch": 0.24985263157894738,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00043514888070180315,
+      "loss": 3.2833,
+      "step": 5934
+    },
+    {
+      "epoch": 0.24989473684210525,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004351262045081011,
+      "loss": 3.081,
+      "step": 5935
+    },
+    {
+      "epoch": 0.24993684210526315,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004351035249416171,
+      "loss": 3.6775,
+      "step": 5936
+    },
+    {
+      "epoch": 0.24997894736842105,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00043508084200276446,
+      "loss": 3.1271,
+      "step": 5937
+    },
+    {
+      "epoch": 0.2500210526315789,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004350581556919562,
+      "loss": 2.9615,
+      "step": 5938
+    },
+    {
+      "epoch": 0.2500631578947368,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00043503546600960586,
+      "loss": 2.9622,
+      "step": 5939
+    },
+    {
+      "epoch": 0.2501052631578947,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00043501277295612676,
+      "loss": 3.4923,
+      "step": 5940
+    },
+    {
+      "epoch": 0.2501473684210526,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004349900765319322,
+      "loss": 2.946,
+      "step": 5941
+    },
+    {
+      "epoch": 0.2501894736842105,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00043496737673743587,
+      "loss": 2.9072,
+      "step": 5942
+    },
+    {
+      "epoch": 0.2502315789473684,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0004349446735730512,
+      "loss": 3.0582,
+      "step": 5943
+    },
+    {
+      "epoch": 0.2502736842105263,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.000434921967039192,
+      "loss": 3.1677,
+      "step": 5944
+    },
+    {
+      "epoch": 0.2503157894736842,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004348992571362718,
+      "loss": 3.1582,
+      "step": 5945
+    },
+    {
+      "epoch": 0.2503578947368421,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004348765438647043,
+      "loss": 3.1602,
+      "step": 5946
+    },
+    {
+      "epoch": 0.2504,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004348538272249035,
+      "loss": 2.8246,
+      "step": 5947
+    },
+    {
+      "epoch": 0.2504421052631579,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004348311072172831,
+      "loss": 3.3215,
+      "step": 5948
+    },
+    {
+      "epoch": 0.2504842105263158,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00043480838384225715,
+      "loss": 3.189,
+      "step": 5949
+    },
+    {
+      "epoch": 0.2505263157894737,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00043478565710023954,
+      "loss": 3.031,
+      "step": 5950
+    },
+    {
+      "epoch": 0.25056842105263155,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0004347629269916443,
+      "loss": 3.1738,
+      "step": 5951
+    },
+    {
+      "epoch": 0.25061052631578945,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00043474019351688576,
+      "loss": 3.4134,
+      "step": 5952
+    },
+    {
+      "epoch": 0.25065263157894735,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004347174566763779,
+      "loss": 3.1434,
+      "step": 5953
+    },
+    {
+      "epoch": 0.25069473684210525,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.000434694716470535,
+      "loss": 3.3328,
+      "step": 5954
+    },
+    {
+      "epoch": 0.25073684210526315,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00043467197289977135,
+      "loss": 3.3598,
+      "step": 5955
+    },
+    {
+      "epoch": 0.25077894736842105,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004346492259645014,
+      "loss": 2.6097,
+      "step": 5956
+    },
+    {
+      "epoch": 0.25082105263157894,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00043462647566513946,
+      "loss": 3.6658,
+      "step": 5957
+    },
+    {
+      "epoch": 0.25086315789473684,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004346037220021001,
+      "loss": 3.384,
+      "step": 5958
+    },
+    {
+      "epoch": 0.25090526315789474,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00043458096497579773,
+      "loss": 3.3693,
+      "step": 5959
+    },
+    {
+      "epoch": 0.25094736842105264,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004345582045866472,
+      "loss": 3.3107,
+      "step": 5960
+    },
+    {
+      "epoch": 0.25098947368421054,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004345354408350629,
+      "loss": 3.0628,
+      "step": 5961
+    },
+    {
+      "epoch": 0.25103157894736844,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00043451267372145977,
+      "loss": 3.6457,
+      "step": 5962
+    },
+    {
+      "epoch": 0.25107368421052634,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004344899032462524,
+      "loss": 3.1352,
+      "step": 5963
+    },
+    {
+      "epoch": 0.25111578947368424,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004344671294098559,
+      "loss": 3.4436,
+      "step": 5964
+    },
+    {
+      "epoch": 0.2511578947368421,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.000434444352212685,
+      "loss": 2.9168,
+      "step": 5965
+    },
+    {
+      "epoch": 0.2512,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004344215716551547,
+      "loss": 3.2977,
+      "step": 5966
+    },
+    {
+      "epoch": 0.2512421052631579,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00043439878773768006,
+      "loss": 3.3153,
+      "step": 5967
+    },
+    {
+      "epoch": 0.2512842105263158,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004343760004606761,
+      "loss": 3.1176,
+      "step": 5968
+    },
+    {
+      "epoch": 0.2513263157894737,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004343532098245581,
+      "loss": 3.4627,
+      "step": 5969
+    },
+    {
+      "epoch": 0.2513684210526316,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00043433041582974127,
+      "loss": 2.8761,
+      "step": 5970
+    },
+    {
+      "epoch": 0.2514105263157895,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0004343076184766408,
+      "loss": 3.4108,
+      "step": 5971
+    },
+    {
+      "epoch": 0.25145263157894737,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004342848177656721,
+      "loss": 3.311,
+      "step": 5972
+    },
+    {
+      "epoch": 0.25149473684210527,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004342620136972505,
+      "loss": 3.9995,
+      "step": 5973
+    },
+    {
+      "epoch": 0.25153684210526317,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004342392062717915,
+      "loss": 3.451,
+      "step": 5974
+    },
+    {
+      "epoch": 0.25157894736842107,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00043421639548971066,
+      "loss": 3.5722,
+      "step": 5975
+    },
+    {
+      "epoch": 0.25162105263157897,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004341935813514235,
+      "loss": 3.1107,
+      "step": 5976
+    },
+    {
+      "epoch": 0.25166315789473687,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00043417076385734576,
+      "loss": 3.5168,
+      "step": 5977
+    },
+    {
+      "epoch": 0.2517052631578947,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004341479430078931,
+      "loss": 3.1381,
+      "step": 5978
+    },
+    {
+      "epoch": 0.2517473684210526,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00043412511880348126,
+      "loss": 3.0685,
+      "step": 5979
+    },
+    {
+      "epoch": 0.2517894736842105,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004341022912445261,
+      "loss": 3.3139,
+      "step": 5980
+    },
+    {
+      "epoch": 0.2518315789473684,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0004340794603314435,
+      "loss": 3.2366,
+      "step": 5981
+    },
+    {
+      "epoch": 0.2518736842105263,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00043405662606464944,
+      "loss": 3.6517,
+      "step": 5982
+    },
+    {
+      "epoch": 0.2519157894736842,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00043403378844455987,
+      "loss": 2.9999,
+      "step": 5983
+    },
+    {
+      "epoch": 0.2519578947368421,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00043401094747159096,
+      "loss": 3.1826,
+      "step": 5984
+    },
+    {
+      "epoch": 0.252,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004339881031461588,
+      "loss": 3.0703,
+      "step": 5985
+    },
+    {
+      "epoch": 0.2520421052631579,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004339652554686795,
+      "loss": 2.9712,
+      "step": 5986
+    },
+    {
+      "epoch": 0.2520842105263158,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004339424044395694,
+      "loss": 3.3076,
+      "step": 5987
+    },
+    {
+      "epoch": 0.2521263157894737,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004339195500592449,
+      "loss": 3.553,
+      "step": 5988
+    },
+    {
+      "epoch": 0.2521684210526316,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004338966923281222,
+      "loss": 3.7331,
+      "step": 5989
+    },
+    {
+      "epoch": 0.2522105263157895,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004338738312466179,
+      "loss": 3.6566,
+      "step": 5990
+    },
+    {
+      "epoch": 0.2522526315789474,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00043385096681514834,
+      "loss": 3.4665,
+      "step": 5991
+    },
+    {
+      "epoch": 0.25229473684210524,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00043382809903413027,
+      "loss": 2.9238,
+      "step": 5992
+    },
+    {
+      "epoch": 0.25233684210526314,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00043380522790398025,
+      "loss": 3.1875,
+      "step": 5993
+    },
+    {
+      "epoch": 0.25237894736842104,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004337823534251149,
+      "loss": 3.0295,
+      "step": 5994
+    },
+    {
+      "epoch": 0.25242105263157893,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.000433759475597951,
+      "loss": 3.146,
+      "step": 5995
+    },
+    {
+      "epoch": 0.25246315789473683,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00043373659442290537,
+      "loss": 3.1638,
+      "step": 5996
+    },
+    {
+      "epoch": 0.25250526315789473,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00043371370990039483,
+      "loss": 3.3972,
+      "step": 5997
+    },
+    {
+      "epoch": 0.25254736842105263,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00043369082203083644,
+      "loss": 3.4852,
+      "step": 5998
+    },
+    {
+      "epoch": 0.25258947368421053,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004336679308146471,
+      "loss": 3.499,
+      "step": 5999
+    },
+    {
+      "epoch": 0.25263157894736843,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004336450362522437,
+      "loss": 2.9337,
+      "step": 6000
+    },
+    {
+      "epoch": 0.25263157894736843,
+      "eval_loss": 3.2944118976593018,
+      "eval_runtime": 335.4633,
+      "eval_samples_per_second": 44.714,
+      "eval_steps_per_second": 5.589,
+      "step": 6000
+    },
+    {
+      "epoch": 0.25267368421052633,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004336221383440437,
+      "loss": 3.6573,
+      "step": 6001
+    },
+    {
+      "epoch": 0.2527157894736842,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.000433599237090464,
+      "loss": 3.3607,
+      "step": 6002
+    },
+    {
+      "epoch": 0.2527578947368421,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00043357633249192193,
+      "loss": 2.9241,
+      "step": 6003
+    },
+    {
+      "epoch": 0.2528,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004335534245488347,
+      "loss": 3.3706,
+      "step": 6004
+    },
+    {
+      "epoch": 0.25284210526315787,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00043353051326161984,
+      "loss": 3.5743,
+      "step": 6005
+    },
+    {
+      "epoch": 0.25288421052631577,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004335075986306947,
+      "loss": 2.8625,
+      "step": 6006
+    },
+    {
+      "epoch": 0.25292631578947367,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004334846806564767,
+      "loss": 3.0568,
+      "step": 6007
+    },
+    {
+      "epoch": 0.25296842105263156,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00043346175933938337,
+      "loss": 3.0376,
+      "step": 6008
+    },
+    {
+      "epoch": 0.25301052631578946,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004334388346798323,
+      "loss": 2.6586,
+      "step": 6009
+    },
+    {
+      "epoch": 0.25305263157894736,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00043341590667824126,
+      "loss": 2.9894,
+      "step": 6010
+    },
+    {
+      "epoch": 0.25309473684210526,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00043339297533502786,
+      "loss": 2.773,
+      "step": 6011
+    },
+    {
+      "epoch": 0.25313684210526316,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00043337004065061,
+      "loss": 3.154,
+      "step": 6012
+    },
+    {
+      "epoch": 0.25317894736842106,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00043334710262540535,
+      "loss": 3.4374,
+      "step": 6013
+    },
+    {
+      "epoch": 0.25322105263157896,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.000433324161259832,
+      "loss": 3.3654,
+      "step": 6014
+    },
+    {
+      "epoch": 0.25326315789473686,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004333012165543078,
+      "loss": 3.6574,
+      "step": 6015
+    },
+    {
+      "epoch": 0.25330526315789476,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004332782685092508,
+      "loss": 3.2352,
+      "step": 6016
+    },
+    {
+      "epoch": 0.25334736842105265,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004332553171250791,
+      "loss": 3.056,
+      "step": 6017
+    },
+    {
+      "epoch": 0.25338947368421055,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004332323624022108,
+      "loss": 3.4178,
+      "step": 6018
+    },
+    {
+      "epoch": 0.2534315789473684,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0004332094043410642,
+      "loss": 3.4588,
+      "step": 6019
+    },
+    {
+      "epoch": 0.2534736842105263,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004331864429420575,
+      "loss": 3.3476,
+      "step": 6020
+    },
+    {
+      "epoch": 0.2535157894736842,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00043316347820560904,
+      "loss": 3.1959,
+      "step": 6021
+    },
+    {
+      "epoch": 0.2535578947368421,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00043314051013213715,
+      "loss": 2.885,
+      "step": 6022
+    },
+    {
+      "epoch": 0.2536,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004331175387220605,
+      "loss": 3.5337,
+      "step": 6023
+    },
+    {
+      "epoch": 0.2536421052631579,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00043309456397579735,
+      "loss": 3.0334,
+      "step": 6024
+    },
+    {
+      "epoch": 0.2536842105263158,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004330715858937664,
+      "loss": 3.5462,
+      "step": 6025
+    },
+    {
+      "epoch": 0.2537263157894737,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00043304860447638616,
+      "loss": 3.3193,
+      "step": 6026
+    },
+    {
+      "epoch": 0.2537684210526316,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004330256197240755,
+      "loss": 3.0765,
+      "step": 6027
+    },
+    {
+      "epoch": 0.2538105263157895,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004330026316372531,
+      "loss": 3.1876,
+      "step": 6028
+    },
+    {
+      "epoch": 0.2538526315789474,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0004329796402163378,
+      "loss": 3.8867,
+      "step": 6029
+    },
+    {
+      "epoch": 0.2538947368421053,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004329566454617484,
+      "loss": 3.5915,
+      "step": 6030
+    },
+    {
+      "epoch": 0.2539368421052632,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00043293364737390387,
+      "loss": 3.3869,
+      "step": 6031
+    },
+    {
+      "epoch": 0.253978947368421,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00043291064595322326,
+      "loss": 3.2226,
+      "step": 6032
+    },
+    {
+      "epoch": 0.2540210526315789,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00043288764120012557,
+      "loss": 3.3566,
+      "step": 6033
+    },
+    {
+      "epoch": 0.2540631578947368,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00043286463311503,
+      "loss": 3.1756,
+      "step": 6034
+    },
+    {
+      "epoch": 0.2541052631578947,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00043284162169835555,
+      "loss": 3.1058,
+      "step": 6035
+    },
+    {
+      "epoch": 0.2541473684210526,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00043281860695052165,
+      "loss": 3.1083,
+      "step": 6036
+    },
+    {
+      "epoch": 0.2541894736842105,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004327955888719476,
+      "loss": 2.6818,
+      "step": 6037
+    },
+    {
+      "epoch": 0.2542315789473684,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0004327725674630526,
+      "loss": 2.8582,
+      "step": 6038
+    },
+    {
+      "epoch": 0.2542736842105263,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00043274954272425623,
+      "loss": 3.0076,
+      "step": 6039
+    },
+    {
+      "epoch": 0.2543157894736842,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004327265146559779,
+      "loss": 3.3849,
+      "step": 6040
+    },
+    {
+      "epoch": 0.2543578947368421,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004327034832586372,
+      "loss": 3.204,
+      "step": 6041
+    },
+    {
+      "epoch": 0.2544,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004326804485326536,
+      "loss": 3.196,
+      "step": 6042
+    },
+    {
+      "epoch": 0.2544421052631579,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00043265741047844694,
+      "loss": 3.4673,
+      "step": 6043
+    },
+    {
+      "epoch": 0.2544842105263158,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00043263436909643684,
+      "loss": 2.9964,
+      "step": 6044
+    },
+    {
+      "epoch": 0.2545263157894737,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00043261132438704317,
+      "loss": 3.2746,
+      "step": 6045
+    },
+    {
+      "epoch": 0.25456842105263155,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00043258827635068563,
+      "loss": 3.3636,
+      "step": 6046
+    },
+    {
+      "epoch": 0.25461052631578945,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00043256522498778436,
+      "loss": 2.8073,
+      "step": 6047
+    },
+    {
+      "epoch": 0.25465263157894735,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004325421702987591,
+      "loss": 3.3582,
+      "step": 6048
+    },
+    {
+      "epoch": 0.25469473684210525,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004325191122840301,
+      "loss": 3.3618,
+      "step": 6049
+    },
+    {
+      "epoch": 0.25473684210526315,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00043249605094401717,
+      "loss": 3.2649,
+      "step": 6050
+    },
+    {
+      "epoch": 0.25477894736842105,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00043247298627914065,
+      "loss": 3.6377,
+      "step": 6051
+    },
+    {
+      "epoch": 0.25482105263157895,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00043244991828982076,
+      "loss": 3.243,
+      "step": 6052
+    },
+    {
+      "epoch": 0.25486315789473685,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004324268469764777,
+      "loss": 3.362,
+      "step": 6053
+    },
+    {
+      "epoch": 0.25490526315789475,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004324037723395319,
+      "loss": 3.6201,
+      "step": 6054
+    },
+    {
+      "epoch": 0.25494736842105264,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00043238069437940364,
+      "loss": 3.2052,
+      "step": 6055
+    },
+    {
+      "epoch": 0.25498947368421054,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004323576130965134,
+      "loss": 3.1994,
+      "step": 6056
+    },
+    {
+      "epoch": 0.25503157894736844,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00043233452849128175,
+      "loss": 3.3218,
+      "step": 6057
+    },
+    {
+      "epoch": 0.25507368421052634,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0004323114405641292,
+      "loss": 3.5504,
+      "step": 6058
+    },
+    {
+      "epoch": 0.2551157894736842,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00043228834931547636,
+      "loss": 3.4439,
+      "step": 6059
+    },
+    {
+      "epoch": 0.2551578947368421,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.000432265254745744,
+      "loss": 3.1921,
+      "step": 6060
+    },
+    {
+      "epoch": 0.2552,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00043224215685535287,
+      "loss": 3.0537,
+      "step": 6061
+    },
+    {
+      "epoch": 0.2552421052631579,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004322190556447238,
+      "loss": 2.8689,
+      "step": 6062
+    },
+    {
+      "epoch": 0.2552842105263158,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004321959511142776,
+      "loss": 3.309,
+      "step": 6063
+    },
+    {
+      "epoch": 0.2553263157894737,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00043217284326443526,
+      "loss": 3.1243,
+      "step": 6064
+    },
+    {
+      "epoch": 0.2553684210526316,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0004321497320956177,
+      "loss": 2.9133,
+      "step": 6065
+    },
+    {
+      "epoch": 0.2554105263157895,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00043212661760824614,
+      "loss": 3.1312,
+      "step": 6066
+    },
+    {
+      "epoch": 0.2554526315789474,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00043210349980274154,
+      "loss": 3.8386,
+      "step": 6067
+    },
+    {
+      "epoch": 0.2554947368421053,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0004320803786795251,
+      "loss": 3.4378,
+      "step": 6068
+    },
+    {
+      "epoch": 0.2555368421052632,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00043205725423901814,
+      "loss": 3.5953,
+      "step": 6069
+    },
+    {
+      "epoch": 0.25557894736842107,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0004320341264816419,
+      "loss": 4.0064,
+      "step": 6070
+    },
+    {
+      "epoch": 0.25562105263157897,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004320109954078178,
+      "loss": 3.8868,
+      "step": 6071
+    },
+    {
+      "epoch": 0.2556631578947368,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00043198786101796715,
+      "loss": 3.3687,
+      "step": 6072
+    },
+    {
+      "epoch": 0.2557052631578947,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00043196472331251157,
+      "loss": 3.568,
+      "step": 6073
+    },
+    {
+      "epoch": 0.2557473684210526,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0004319415822918725,
+      "loss": 3.7574,
+      "step": 6074
+    },
+    {
+      "epoch": 0.2557894736842105,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004319184379564716,
+      "loss": 3.2719,
+      "step": 6075
+    },
+    {
+      "epoch": 0.2558315789473684,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00043189529030673043,
+      "loss": 3.1474,
+      "step": 6076
+    },
+    {
+      "epoch": 0.2558736842105263,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00043187213934307086,
+      "loss": 3.1235,
+      "step": 6077
+    },
+    {
+      "epoch": 0.2559157894736842,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00043184898506591456,
+      "loss": 3.1134,
+      "step": 6078
+    },
+    {
+      "epoch": 0.2559578947368421,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00043182582747568345,
+      "loss": 3.2941,
+      "step": 6079
+    },
+    {
+      "epoch": 0.256,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004318026665727993,
+      "loss": 3.1172,
+      "step": 6080
+    },
+    {
+      "epoch": 0.2560421052631579,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00043177950235768434,
+      "loss": 3.325,
+      "step": 6081
+    },
+    {
+      "epoch": 0.2560842105263158,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00043175633483076033,
+      "loss": 3.1216,
+      "step": 6082
+    },
+    {
+      "epoch": 0.2561263157894737,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00043173316399244946,
+      "loss": 3.249,
+      "step": 6083
+    },
+    {
+      "epoch": 0.2561684210526316,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004317099898431739,
+      "loss": 3.6156,
+      "step": 6084
+    },
+    {
+      "epoch": 0.2562105263157895,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00043168681238335576,
+      "loss": 3.6679,
+      "step": 6085
+    },
+    {
+      "epoch": 0.25625263157894734,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00043166363161341743,
+      "loss": 3.1704,
+      "step": 6086
+    },
+    {
+      "epoch": 0.25629473684210524,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004316404475337812,
+      "loss": 3.4398,
+      "step": 6087
+    },
+    {
+      "epoch": 0.25633684210526314,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0004316172601448694,
+      "loss": 3.2026,
+      "step": 6088
+    },
+    {
+      "epoch": 0.25637894736842104,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00043159406944710447,
+      "loss": 3.4618,
+      "step": 6089
+    },
+    {
+      "epoch": 0.25642105263157894,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.000431570875440909,
+      "loss": 3.1465,
+      "step": 6090
+    },
+    {
+      "epoch": 0.25646315789473684,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004315476781267055,
+      "loss": 3.3156,
+      "step": 6091
+    },
+    {
+      "epoch": 0.25650526315789474,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00043152447750491665,
+      "loss": 3.6675,
+      "step": 6092
+    },
+    {
+      "epoch": 0.25654736842105264,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004315012735759651,
+      "loss": 3.5745,
+      "step": 6093
+    },
+    {
+      "epoch": 0.25658947368421053,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004314780663402736,
+      "loss": 3.1197,
+      "step": 6094
+    },
+    {
+      "epoch": 0.25663157894736843,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00043145485579826494,
+      "loss": 3.4775,
+      "step": 6095
+    },
+    {
+      "epoch": 0.25667368421052633,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.000431431641950362,
+      "loss": 3.2463,
+      "step": 6096
+    },
+    {
+      "epoch": 0.25671578947368423,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0004314084247969877,
+      "loss": 3.4397,
+      "step": 6097
+    },
+    {
+      "epoch": 0.25675789473684213,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0004313852043385651,
+      "loss": 3.1508,
+      "step": 6098
+    },
+    {
+      "epoch": 0.2568,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0004313619805755171,
+      "loss": 3.3843,
+      "step": 6099
+    },
+    {
+      "epoch": 0.25684210526315787,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00043133875350826704,
+      "loss": 2.8319,
+      "step": 6100
+    },
+    {
+      "epoch": 0.25688421052631577,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004313155231372379,
+      "loss": 3.1014,
+      "step": 6101
+    },
+    {
+      "epoch": 0.25692631578947367,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00043129228946285294,
+      "loss": 3.2822,
+      "step": 6102
+    },
+    {
+      "epoch": 0.25696842105263157,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00043126905248553547,
+      "loss": 3.3325,
+      "step": 6103
+    },
+    {
+      "epoch": 0.25701052631578947,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004312458122057089,
+      "loss": 3.2015,
+      "step": 6104
+    },
+    {
+      "epoch": 0.25705263157894737,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004312225686237965,
+      "loss": 3.5371,
+      "step": 6105
+    },
+    {
+      "epoch": 0.25709473684210526,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004311993217402219,
+      "loss": 3.0762,
+      "step": 6106
+    },
+    {
+      "epoch": 0.25713684210526316,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00043117607155540854,
+      "loss": 3.2927,
+      "step": 6107
+    },
+    {
+      "epoch": 0.25717894736842106,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00043115281806978007,
+      "loss": 3.385,
+      "step": 6108
+    },
+    {
+      "epoch": 0.25722105263157896,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00043112956128376,
+      "loss": 2.9517,
+      "step": 6109
+    },
+    {
+      "epoch": 0.25726315789473686,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0004311063011977723,
+      "loss": 3.3454,
+      "step": 6110
+    },
+    {
+      "epoch": 0.25730526315789476,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00043108303781224045,
+      "loss": 3.1351,
+      "step": 6111
+    },
+    {
+      "epoch": 0.25734736842105266,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00043105977112758845,
+      "loss": 3.2994,
+      "step": 6112
+    },
+    {
+      "epoch": 0.2573894736842105,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00043103650114424013,
+      "loss": 3.3109,
+      "step": 6113
+    },
+    {
+      "epoch": 0.2574315789473684,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00043101322786261956,
+      "loss": 3.4919,
+      "step": 6114
+    },
+    {
+      "epoch": 0.2574736842105263,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00043098995128315056,
+      "loss": 3.5706,
+      "step": 6115
+    },
+    {
+      "epoch": 0.2575157894736842,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004309666714062573,
+      "loss": 3.3651,
+      "step": 6116
+    },
+    {
+      "epoch": 0.2575578947368421,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.000430943388232364,
+      "loss": 3.1717,
+      "step": 6117
+    },
+    {
+      "epoch": 0.2576,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00043092010176189465,
+      "loss": 3.1678,
+      "step": 6118
+    },
+    {
+      "epoch": 0.2576421052631579,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00043089681199527376,
+      "loss": 2.6509,
+      "step": 6119
+    },
+    {
+      "epoch": 0.2576842105263158,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00043087351893292535,
+      "loss": 3.6603,
+      "step": 6120
+    },
+    {
+      "epoch": 0.2577263157894737,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00043085022257527406,
+      "loss": 3.7938,
+      "step": 6121
+    },
+    {
+      "epoch": 0.2577684210526316,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0004308269229227442,
+      "loss": 3.2138,
+      "step": 6122
+    },
+    {
+      "epoch": 0.2578105263157895,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004308036199757602,
+      "loss": 3.2045,
+      "step": 6123
+    },
+    {
+      "epoch": 0.2578526315789474,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004307803137347467,
+      "loss": 3.5277,
+      "step": 6124
+    },
+    {
+      "epoch": 0.2578947368421053,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0004307570042001283,
+      "loss": 2.8281,
+      "step": 6125
+    },
+    {
+      "epoch": 0.25793684210526313,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00043073369137232965,
+      "loss": 3.424,
+      "step": 6126
+    },
+    {
+      "epoch": 0.25797894736842103,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004307103752517755,
+      "loss": 3.6572,
+      "step": 6127
+    },
+    {
+      "epoch": 0.25802105263157893,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00043068705583889066,
+      "loss": 3.6863,
+      "step": 6128
+    },
+    {
+      "epoch": 0.2580631578947368,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004306637331341,
+      "loss": 3.4523,
+      "step": 6129
+    },
+    {
+      "epoch": 0.2581052631578947,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00043064040713782837,
+      "loss": 2.8686,
+      "step": 6130
+    },
+    {
+      "epoch": 0.2581473684210526,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004306170778505007,
+      "loss": 3.2661,
+      "step": 6131
+    },
+    {
+      "epoch": 0.2581894736842105,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00043059374527254214,
+      "loss": 3.1896,
+      "step": 6132
+    },
+    {
+      "epoch": 0.2582315789473684,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004305704094043777,
+      "loss": 2.9932,
+      "step": 6133
+    },
+    {
+      "epoch": 0.2582736842105263,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00043054707024643257,
+      "loss": 3.2916,
+      "step": 6134
+    },
+    {
+      "epoch": 0.2583157894736842,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.000430523727799132,
+      "loss": 2.804,
+      "step": 6135
+    },
+    {
+      "epoch": 0.2583578947368421,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004305003820629012,
+      "loss": 2.9091,
+      "step": 6136
+    },
+    {
+      "epoch": 0.2584,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00043047703303816543,
+      "loss": 3.3525,
+      "step": 6137
+    },
+    {
+      "epoch": 0.2584421052631579,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004304536807253503,
+      "loss": 3.1367,
+      "step": 6138
+    },
+    {
+      "epoch": 0.2584842105263158,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.000430430325124881,
+      "loss": 3.4791,
+      "step": 6139
+    },
+    {
+      "epoch": 0.25852631578947366,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004304069662371833,
+      "loss": 3.0645,
+      "step": 6140
+    },
+    {
+      "epoch": 0.25856842105263156,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00043038360406268264,
+      "loss": 3.1117,
+      "step": 6141
+    },
+    {
+      "epoch": 0.25861052631578946,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00043036023860180457,
+      "loss": 2.9278,
+      "step": 6142
+    },
+    {
+      "epoch": 0.25865263157894736,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0004303368698549749,
+      "loss": 3.2543,
+      "step": 6143
+    },
+    {
+      "epoch": 0.25869473684210526,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004303134978226193,
+      "loss": 3.1228,
+      "step": 6144
+    },
+    {
+      "epoch": 0.25873684210526315,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00043029012250516373,
+      "loss": 3.366,
+      "step": 6145
+    },
+    {
+      "epoch": 0.25877894736842105,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004302667439030339,
+      "loss": 3.0633,
+      "step": 6146
+    },
+    {
+      "epoch": 0.25882105263157895,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004302433620166558,
+      "loss": 3.3485,
+      "step": 6147
+    },
+    {
+      "epoch": 0.25886315789473685,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004302199768464554,
+      "loss": 3.2263,
+      "step": 6148
+    },
+    {
+      "epoch": 0.25890526315789475,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00043019658839285884,
+      "loss": 3.0154,
+      "step": 6149
+    },
+    {
+      "epoch": 0.25894736842105265,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0004301731966562921,
+      "loss": 2.9147,
+      "step": 6150
+    },
+    {
+      "epoch": 0.25898947368421055,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00043014980163718144,
+      "loss": 2.9014,
+      "step": 6151
+    },
+    {
+      "epoch": 0.25903157894736845,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004301264033359531,
+      "loss": 3.4199,
+      "step": 6152
+    },
+    {
+      "epoch": 0.2590736842105263,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0004301030017530333,
+      "loss": 2.9437,
+      "step": 6153
+    },
+    {
+      "epoch": 0.2591157894736842,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004300795968888484,
+      "loss": 3.4307,
+      "step": 6154
+    },
+    {
+      "epoch": 0.2591578947368421,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00043005618874382477,
+      "loss": 3.0463,
+      "step": 6155
+    },
+    {
+      "epoch": 0.2592,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004300327773183891,
+      "loss": 3.1884,
+      "step": 6156
+    },
+    {
+      "epoch": 0.2592421052631579,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004300093626129676,
+      "loss": 3.3494,
+      "step": 6157
+    },
+    {
+      "epoch": 0.2592842105263158,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00042998594462798713,
+      "loss": 3.5587,
+      "step": 6158
+    },
+    {
+      "epoch": 0.2593263157894737,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00042996252336387414,
+      "loss": 3.5775,
+      "step": 6159
+    },
+    {
+      "epoch": 0.2593684210526316,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00042993909882105546,
+      "loss": 3.1355,
+      "step": 6160
+    },
+    {
+      "epoch": 0.2594105263157895,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004299156709999578,
+      "loss": 3.2287,
+      "step": 6161
+    },
+    {
+      "epoch": 0.2594526315789474,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00042989223990100805,
+      "loss": 3.4734,
+      "step": 6162
+    },
+    {
+      "epoch": 0.2594947368421053,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.000429868805524633,
+      "loss": 3.2615,
+      "step": 6163
+    },
+    {
+      "epoch": 0.2595368421052632,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004298453678712597,
+      "loss": 3.6193,
+      "step": 6164
+    },
+    {
+      "epoch": 0.2595789473684211,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00042982192694131504,
+      "loss": 2.8476,
+      "step": 6165
+    },
+    {
+      "epoch": 0.259621052631579,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00042979848273522626,
+      "loss": 3.5556,
+      "step": 6166
+    },
+    {
+      "epoch": 0.2596631578947368,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004297750352534203,
+      "loss": 3.1545,
+      "step": 6167
+    },
+    {
+      "epoch": 0.2597052631578947,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00042975158449632447,
+      "loss": 2.9406,
+      "step": 6168
+    },
+    {
+      "epoch": 0.2597473684210526,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00042972813046436586,
+      "loss": 3.6385,
+      "step": 6169
+    },
+    {
+      "epoch": 0.2597894736842105,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.000429704673157972,
+      "loss": 3.2468,
+      "step": 6170
+    },
+    {
+      "epoch": 0.2598315789473684,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004296812125775701,
+      "loss": 3.3389,
+      "step": 6171
+    },
+    {
+      "epoch": 0.2598736842105263,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00042965774872358764,
+      "loss": 3.2557,
+      "step": 6172
+    },
+    {
+      "epoch": 0.2599157894736842,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004296342815964521,
+      "loss": 3.3789,
+      "step": 6173
+    },
+    {
+      "epoch": 0.2599578947368421,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00042961081119659094,
+      "loss": 3.179,
+      "step": 6174
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004295873375244319,
+      "loss": 2.9999,
+      "step": 6175
+    },
+    {
+      "epoch": 0.2600421052631579,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0004295638605804025,
+      "loss": 3.3216,
+      "step": 6176
+    },
+    {
+      "epoch": 0.2600842105263158,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00042954038036493064,
+      "loss": 3.7098,
+      "step": 6177
+    },
+    {
+      "epoch": 0.2601263157894737,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.000429516896878444,
+      "loss": 3.2123,
+      "step": 6178
+    },
+    {
+      "epoch": 0.2601684210526316,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004294934101213703,
+      "loss": 3.5476,
+      "step": 6179
+    },
+    {
+      "epoch": 0.26021052631578945,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00042946992009413774,
+      "loss": 3.0534,
+      "step": 6180
+    },
+    {
+      "epoch": 0.26025263157894735,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00042944642679717393,
+      "loss": 3.2793,
+      "step": 6181
+    },
+    {
+      "epoch": 0.26029473684210525,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00042942293023090716,
+      "loss": 3.2507,
+      "step": 6182
+    },
+    {
+      "epoch": 0.26033684210526314,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00042939943039576543,
+      "loss": 3.4143,
+      "step": 6183
+    },
+    {
+      "epoch": 0.26037894736842104,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00042937592729217676,
+      "loss": 3.0732,
+      "step": 6184
+    },
+    {
+      "epoch": 0.26042105263157894,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0004293524209205696,
+      "loss": 3.098,
+      "step": 6185
+    },
+    {
+      "epoch": 0.26046315789473684,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00042932891128137196,
+      "loss": 3.2063,
+      "step": 6186
+    },
+    {
+      "epoch": 0.26050526315789474,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004293053983750123,
+      "loss": 3.0457,
+      "step": 6187
+    },
+    {
+      "epoch": 0.26054736842105264,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00042928188220191894,
+      "loss": 3.2588,
+      "step": 6188
+    },
+    {
+      "epoch": 0.26058947368421054,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004292583627625203,
+      "loss": 3.6176,
+      "step": 6189
+    },
+    {
+      "epoch": 0.26063157894736844,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004292348400572449,
+      "loss": 3.1577,
+      "step": 6190
+    },
+    {
+      "epoch": 0.26067368421052634,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004292113140865214,
+      "loss": 3.1727,
+      "step": 6191
+    },
+    {
+      "epoch": 0.26071578947368423,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004291877848507783,
+      "loss": 3.2862,
+      "step": 6192
+    },
+    {
+      "epoch": 0.26075789473684213,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00042916425235044425,
+      "loss": 3.1198,
+      "step": 6193
+    },
+    {
+      "epoch": 0.2608,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004291407165859481,
+      "loss": 2.8052,
+      "step": 6194
+    },
+    {
+      "epoch": 0.2608421052631579,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004291171775577185,
+      "loss": 3.3822,
+      "step": 6195
+    },
+    {
+      "epoch": 0.2608842105263158,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004290936352661844,
+      "loss": 3.4492,
+      "step": 6196
+    },
+    {
+      "epoch": 0.2609263157894737,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00042907008971177467,
+      "loss": 3.4028,
+      "step": 6197
+    },
+    {
+      "epoch": 0.26096842105263157,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00042904654089491836,
+      "loss": 3.4107,
+      "step": 6198
+    },
+    {
+      "epoch": 0.26101052631578947,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004290229888160444,
+      "loss": 3.0222,
+      "step": 6199
+    },
+    {
+      "epoch": 0.26105263157894737,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.000428999433475582,
+      "loss": 3.7807,
+      "step": 6200
+    },
+    {
+      "epoch": 0.26109473684210527,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0004289758748739602,
+      "loss": 3.0948,
+      "step": 6201
+    },
+    {
+      "epoch": 0.26113684210526317,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004289523130116082,
+      "loss": 3.2291,
+      "step": 6202
+    },
+    {
+      "epoch": 0.26117894736842107,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004289287478889553,
+      "loss": 3.7891,
+      "step": 6203
+    },
+    {
+      "epoch": 0.26122105263157896,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004289051795064309,
+      "loss": 3.4318,
+      "step": 6204
+    },
+    {
+      "epoch": 0.26126315789473686,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004288816078644643,
+      "loss": 3.515,
+      "step": 6205
+    },
+    {
+      "epoch": 0.26130526315789476,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00042885803296348503,
+      "loss": 3.0889,
+      "step": 6206
+    },
+    {
+      "epoch": 0.2613473684210526,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00042883445480392256,
+      "loss": 3.4653,
+      "step": 6207
+    },
+    {
+      "epoch": 0.2613894736842105,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004288108733862064,
+      "loss": 3.3627,
+      "step": 6208
+    },
+    {
+      "epoch": 0.2614315789473684,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0004287872887107662,
+      "loss": 2.9652,
+      "step": 6209
+    },
+    {
+      "epoch": 0.2614736842105263,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004287637007780317,
+      "loss": 3.4843,
+      "step": 6210
+    },
+    {
+      "epoch": 0.2615157894736842,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00042874010958843267,
+      "loss": 3.1182,
+      "step": 6211
+    },
+    {
+      "epoch": 0.2615578947368421,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004287165151423988,
+      "loss": 2.8971,
+      "step": 6212
+    },
+    {
+      "epoch": 0.2616,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00042869291744036,
+      "loss": 3.6778,
+      "step": 6213
+    },
+    {
+      "epoch": 0.2616421052631579,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0004286693164827462,
+      "loss": 3.3359,
+      "step": 6214
+    },
+    {
+      "epoch": 0.2616842105263158,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00042864571226998736,
+      "loss": 3.2704,
+      "step": 6215
+    },
+    {
+      "epoch": 0.2617263157894737,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00042862210480251365,
+      "loss": 3.4575,
+      "step": 6216
+    },
+    {
+      "epoch": 0.2617684210526316,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004285984940807549,
+      "loss": 3.4357,
+      "step": 6217
+    },
+    {
+      "epoch": 0.2618105263157895,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004285748801051416,
+      "loss": 3.1995,
+      "step": 6218
+    },
+    {
+      "epoch": 0.2618526315789474,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00042855126287610364,
+      "loss": 3.3123,
+      "step": 6219
+    },
+    {
+      "epoch": 0.26189473684210524,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004285276423940715,
+      "loss": 3.3992,
+      "step": 6220
+    },
+    {
+      "epoch": 0.26193684210526313,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004285040186594755,
+      "loss": 3.5687,
+      "step": 6221
+    },
+    {
+      "epoch": 0.26197894736842103,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004284803916727461,
+      "loss": 3.6773,
+      "step": 6222
+    },
+    {
+      "epoch": 0.26202105263157893,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00042845676143431357,
+      "loss": 2.8695,
+      "step": 6223
+    },
+    {
+      "epoch": 0.26206315789473683,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00042843312794460856,
+      "loss": 3.3007,
+      "step": 6224
+    },
+    {
+      "epoch": 0.26210526315789473,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00042840949120406157,
+      "loss": 3.2453,
+      "step": 6225
+    },
+    {
+      "epoch": 0.26214736842105263,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004283858512131033,
+      "loss": 3.3979,
+      "step": 6226
+    },
+    {
+      "epoch": 0.26218947368421053,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004283622079721644,
+      "loss": 3.207,
+      "step": 6227
+    },
+    {
+      "epoch": 0.2622315789473684,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00042833856148167567,
+      "loss": 3.3125,
+      "step": 6228
+    },
+    {
+      "epoch": 0.2622736842105263,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00042831491174206786,
+      "loss": 3.0378,
+      "step": 6229
+    },
+    {
+      "epoch": 0.2623157894736842,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004282912587537719,
+      "loss": 3.5209,
+      "step": 6230
+    },
+    {
+      "epoch": 0.2623578947368421,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00042826760251721865,
+      "loss": 3.2132,
+      "step": 6231
+    },
+    {
+      "epoch": 0.2624,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004282439430328392,
+      "loss": 3.4563,
+      "step": 6232
+    },
+    {
+      "epoch": 0.2624421052631579,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00042822028030106447,
+      "loss": 3.312,
+      "step": 6233
+    },
+    {
+      "epoch": 0.26248421052631576,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00042819661432232564,
+      "loss": 3.0309,
+      "step": 6234
+    },
+    {
+      "epoch": 0.26252631578947366,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004281729450970539,
+      "loss": 3.2094,
+      "step": 6235
+    },
+    {
+      "epoch": 0.26256842105263156,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00042814927262568045,
+      "loss": 3.2903,
+      "step": 6236
+    },
+    {
+      "epoch": 0.26261052631578946,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004281255969086365,
+      "loss": 3.2936,
+      "step": 6237
+    },
+    {
+      "epoch": 0.26265263157894736,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00042810191794635355,
+      "loss": 2.7746,
+      "step": 6238
+    },
+    {
+      "epoch": 0.26269473684210526,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00042807823573926283,
+      "loss": 3.033,
+      "step": 6239
+    },
+    {
+      "epoch": 0.26273684210526316,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00042805455028779594,
+      "loss": 3.1633,
+      "step": 6240
+    },
+    {
+      "epoch": 0.26277894736842106,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00042803086159238444,
+      "loss": 3.1929,
+      "step": 6241
+    },
+    {
+      "epoch": 0.26282105263157896,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004280071696534597,
+      "loss": 3.1701,
+      "step": 6242
+    },
+    {
+      "epoch": 0.26286315789473685,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00042798347447145344,
+      "loss": 3.3468,
+      "step": 6243
+    },
+    {
+      "epoch": 0.26290526315789475,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00042795977604679746,
+      "loss": 3.1284,
+      "step": 6244
+    },
+    {
+      "epoch": 0.26294736842105265,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00042793607437992357,
+      "loss": 3.2611,
+      "step": 6245
+    },
+    {
+      "epoch": 0.26298947368421055,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00042791236947126334,
+      "loss": 3.4824,
+      "step": 6246
+    },
+    {
+      "epoch": 0.2630315789473684,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00042788866132124876,
+      "loss": 3.5099,
+      "step": 6247
+    },
+    {
+      "epoch": 0.2630736842105263,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00042786494993031184,
+      "loss": 3.0518,
+      "step": 6248
+    },
+    {
+      "epoch": 0.2631157894736842,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004278412352988845,
+      "loss": 3.2429,
+      "step": 6249
+    },
+    {
+      "epoch": 0.2631578947368421,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004278175174273988,
+      "loss": 3.5243,
+      "step": 6250
+    },
+    {
+      "epoch": 0.2632,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004277937963162868,
+      "loss": 3.257,
+      "step": 6251
+    },
+    {
+      "epoch": 0.2632421052631579,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004277700719659808,
+      "loss": 3.2812,
+      "step": 6252
+    },
+    {
+      "epoch": 0.2632842105263158,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00042774634437691295,
+      "loss": 3.097,
+      "step": 6253
+    },
+    {
+      "epoch": 0.2633263157894737,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00042772261354951556,
+      "loss": 3.5116,
+      "step": 6254
+    },
+    {
+      "epoch": 0.2633684210526316,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.000427698879484221,
+      "loss": 3.3276,
+      "step": 6255
+    },
+    {
+      "epoch": 0.2634105263157895,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004276751421814615,
+      "loss": 3.3316,
+      "step": 6256
+    },
+    {
+      "epoch": 0.2634526315789474,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004276514016416698,
+      "loss": 3.3754,
+      "step": 6257
+    },
+    {
+      "epoch": 0.2634947368421053,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004276276578652783,
+      "loss": 3.0314,
+      "step": 6258
+    },
+    {
+      "epoch": 0.2635368421052632,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00042760391085271956,
+      "loss": 3.4036,
+      "step": 6259
+    },
+    {
+      "epoch": 0.2635789473684211,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004275801606044262,
+      "loss": 3.1138,
+      "step": 6260
+    },
+    {
+      "epoch": 0.2636210526315789,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00042755640712083097,
+      "loss": 3.5388,
+      "step": 6261
+    },
+    {
+      "epoch": 0.2636631578947368,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0004275326504023667,
+      "loss": 3.8405,
+      "step": 6262
+    },
+    {
+      "epoch": 0.2637052631578947,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00042750889044946606,
+      "loss": 3.125,
+      "step": 6263
+    },
+    {
+      "epoch": 0.2637473684210526,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00042748512726256194,
+      "loss": 3.291,
+      "step": 6264
+    },
+    {
+      "epoch": 0.2637894736842105,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0004274613608420874,
+      "loss": 2.6533,
+      "step": 6265
+    },
+    {
+      "epoch": 0.2638315789473684,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004274375911884754,
+      "loss": 2.9643,
+      "step": 6266
+    },
+    {
+      "epoch": 0.2638736842105263,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00042741381830215897,
+      "loss": 3.3915,
+      "step": 6267
+    },
+    {
+      "epoch": 0.2639157894736842,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004273900421835712,
+      "loss": 3.024,
+      "step": 6268
+    },
+    {
+      "epoch": 0.2639578947368421,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00042736626283314526,
+      "loss": 3.1551,
+      "step": 6269
+    },
+    {
+      "epoch": 0.264,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00042734248025131445,
+      "loss": 3.1435,
+      "step": 6270
+    },
+    {
+      "epoch": 0.2640421052631579,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.000427318694438512,
+      "loss": 3.1671,
+      "step": 6271
+    },
+    {
+      "epoch": 0.2640842105263158,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00042729490539517126,
+      "loss": 3.1248,
+      "step": 6272
+    },
+    {
+      "epoch": 0.2641263157894737,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00042727111312172575,
+      "loss": 3.5478,
+      "step": 6273
+    },
+    {
+      "epoch": 0.26416842105263155,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00042724731761860867,
+      "loss": 3.694,
+      "step": 6274
+    },
+    {
+      "epoch": 0.26421052631578945,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00042722351888625386,
+      "loss": 2.9388,
+      "step": 6275
+    },
+    {
+      "epoch": 0.26425263157894735,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004271997169250947,
+      "loss": 2.8644,
+      "step": 6276
+    },
+    {
+      "epoch": 0.26429473684210525,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.000427175911735565,
+      "loss": 3.2576,
+      "step": 6277
+    },
+    {
+      "epoch": 0.26433684210526315,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00042715210331809825,
+      "loss": 3.0457,
+      "step": 6278
+    },
+    {
+      "epoch": 0.26437894736842105,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004271282916731284,
+      "loss": 2.9549,
+      "step": 6279
+    },
+    {
+      "epoch": 0.26442105263157895,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.0004271044768010891,
+      "loss": 2.9924,
+      "step": 6280
+    },
+    {
+      "epoch": 0.26446315789473684,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00042708065870241443,
+      "loss": 3.4148,
+      "step": 6281
+    },
+    {
+      "epoch": 0.26450526315789474,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00042705683737753806,
+      "loss": 3.0394,
+      "step": 6282
+    },
+    {
+      "epoch": 0.26454736842105264,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00042703301282689433,
+      "loss": 3.3899,
+      "step": 6283
+    },
+    {
+      "epoch": 0.26458947368421054,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00042700918505091703,
+      "loss": 3.3845,
+      "step": 6284
+    },
+    {
+      "epoch": 0.26463157894736844,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004269853540500404,
+      "loss": 3.3631,
+      "step": 6285
+    },
+    {
+      "epoch": 0.26467368421052634,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00042696151982469843,
+      "loss": 3.5155,
+      "step": 6286
+    },
+    {
+      "epoch": 0.26471578947368424,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004269376823753256,
+      "loss": 3.1301,
+      "step": 6287
+    },
+    {
+      "epoch": 0.2647578947368421,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004269138417023561,
+      "loss": 2.8259,
+      "step": 6288
+    },
+    {
+      "epoch": 0.2648,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00042688999780622427,
+      "loss": 2.852,
+      "step": 6289
+    },
+    {
+      "epoch": 0.2648421052631579,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00042686615068736447,
+      "loss": 3.1415,
+      "step": 6290
+    },
+    {
+      "epoch": 0.2648842105263158,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0004268423003462112,
+      "loss": 2.9191,
+      "step": 6291
+    },
+    {
+      "epoch": 0.2649263157894737,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.000426818446783199,
+      "loss": 3.1254,
+      "step": 6292
+    },
+    {
+      "epoch": 0.2649684210526316,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00042679458999876255,
+      "loss": 3.2031,
+      "step": 6293
+    },
+    {
+      "epoch": 0.2650105263157895,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004267707299933363,
+      "loss": 3.2816,
+      "step": 6294
+    },
+    {
+      "epoch": 0.2650526315789474,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004267468667673551,
+      "loss": 3.1382,
+      "step": 6295
+    },
+    {
+      "epoch": 0.26509473684210527,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00042672300032125366,
+      "loss": 3.3349,
+      "step": 6296
+    },
+    {
+      "epoch": 0.26513684210526317,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004266991306554668,
+      "loss": 3.6174,
+      "step": 6297
+    },
+    {
+      "epoch": 0.26517894736842107,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004266752577704293,
+      "loss": 3.3813,
+      "step": 6298
+    },
+    {
+      "epoch": 0.26522105263157897,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004266513816665762,
+      "loss": 3.0026,
+      "step": 6299
+    },
+    {
+      "epoch": 0.26526315789473687,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00042662750234434256,
+      "loss": 3.7354,
+      "step": 6300
+    },
+    {
+      "epoch": 0.2653052631578947,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0004266036198041634,
+      "loss": 3.3894,
+      "step": 6301
+    },
+    {
+      "epoch": 0.2653473684210526,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00042657973404647373,
+      "loss": 3.3925,
+      "step": 6302
+    },
+    {
+      "epoch": 0.2653894736842105,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00042655584507170875,
+      "loss": 3.3881,
+      "step": 6303
+    },
+    {
+      "epoch": 0.2654315789473684,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004265319528803037,
+      "loss": 3.0366,
+      "step": 6304
+    },
+    {
+      "epoch": 0.2654736842105263,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00042650805747269395,
+      "loss": 3.53,
+      "step": 6305
+    },
+    {
+      "epoch": 0.2655157894736842,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00042648415884931476,
+      "loss": 3.2141,
+      "step": 6306
+    },
+    {
+      "epoch": 0.2655578947368421,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00042646025701060153,
+      "loss": 3.2089,
+      "step": 6307
+    },
+    {
+      "epoch": 0.2656,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004264363519569898,
+      "loss": 3.1877,
+      "step": 6308
+    },
+    {
+      "epoch": 0.2656421052631579,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00042641244368891496,
+      "loss": 3.268,
+      "step": 6309
+    },
+    {
+      "epoch": 0.2656842105263158,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00042638853220681273,
+      "loss": 3.2334,
+      "step": 6310
+    },
+    {
+      "epoch": 0.2657263157894737,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00042636461751111867,
+      "loss": 2.8376,
+      "step": 6311
+    },
+    {
+      "epoch": 0.2657684210526316,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004263406996022685,
+      "loss": 2.9539,
+      "step": 6312
+    },
+    {
+      "epoch": 0.2658105263157895,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00042631677848069797,
+      "loss": 3.253,
+      "step": 6313
+    },
+    {
+      "epoch": 0.2658526315789474,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004262928541468429,
+      "loss": 3.5099,
+      "step": 6314
+    },
+    {
+      "epoch": 0.26589473684210524,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00042626892660113923,
+      "loss": 3.2963,
+      "step": 6315
+    },
+    {
+      "epoch": 0.26593684210526314,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00042624499584402274,
+      "loss": 2.8445,
+      "step": 6316
+    },
+    {
+      "epoch": 0.26597894736842104,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004262210618759296,
+      "loss": 3.2586,
+      "step": 6317
+    },
+    {
+      "epoch": 0.26602105263157894,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00042619712469729563,
+      "loss": 3.3481,
+      "step": 6318
+    },
+    {
+      "epoch": 0.26606315789473683,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00042617318430855715,
+      "loss": 3.3241,
+      "step": 6319
+    },
+    {
+      "epoch": 0.26610526315789473,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004261492407101503,
+      "loss": 3.1188,
+      "step": 6320
+    },
+    {
+      "epoch": 0.26614736842105263,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004261252939025111,
+      "loss": 3.4091,
+      "step": 6321
+    },
+    {
+      "epoch": 0.26618947368421053,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00042610134388607616,
+      "loss": 2.8455,
+      "step": 6322
+    },
+    {
+      "epoch": 0.26623157894736843,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00042607739066128156,
+      "loss": 2.8574,
+      "step": 6323
+    },
+    {
+      "epoch": 0.26627368421052633,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00042605343422856377,
+      "loss": 2.6856,
+      "step": 6324
+    },
+    {
+      "epoch": 0.26631578947368423,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004260294745883593,
+      "loss": 3.5231,
+      "step": 6325
+    },
+    {
+      "epoch": 0.2663578947368421,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004260055117411046,
+      "loss": 3.4352,
+      "step": 6326
+    },
+    {
+      "epoch": 0.2664,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00042598154568723626,
+      "loss": 3.0456,
+      "step": 6327
+    },
+    {
+      "epoch": 0.26644210526315787,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004259575764271909,
+      "loss": 3.0534,
+      "step": 6328
+    },
+    {
+      "epoch": 0.26648421052631577,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00042593360396140533,
+      "loss": 3.0856,
+      "step": 6329
+    },
+    {
+      "epoch": 0.26652631578947367,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004259096282903161,
+      "loss": 3.2412,
+      "step": 6330
+    },
+    {
+      "epoch": 0.26656842105263157,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00042588564941436027,
+      "loss": 3.2838,
+      "step": 6331
+    },
+    {
+      "epoch": 0.26661052631578946,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00042586166733397446,
+      "loss": 3.4664,
+      "step": 6332
+    },
+    {
+      "epoch": 0.26665263157894736,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004258376820495957,
+      "loss": 3.0562,
+      "step": 6333
+    },
+    {
+      "epoch": 0.26669473684210526,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.000425813693561661,
+      "loss": 3.3986,
+      "step": 6334
+    },
+    {
+      "epoch": 0.26673684210526316,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004257897018706074,
+      "loss": 3.2179,
+      "step": 6335
+    },
+    {
+      "epoch": 0.26677894736842106,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.000425765706976872,
+      "loss": 3.1342,
+      "step": 6336
+    },
+    {
+      "epoch": 0.26682105263157896,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00042574170888089185,
+      "loss": 3.2029,
+      "step": 6337
+    },
+    {
+      "epoch": 0.26686315789473686,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00042571770758310435,
+      "loss": 3.1142,
+      "step": 6338
+    },
+    {
+      "epoch": 0.26690526315789476,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00042569370308394663,
+      "loss": 3.2718,
+      "step": 6339
+    },
+    {
+      "epoch": 0.26694736842105266,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.000425669695383856,
+      "loss": 3.1049,
+      "step": 6340
+    },
+    {
+      "epoch": 0.2669894736842105,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00042564568448326997,
+      "loss": 3.4215,
+      "step": 6341
+    },
+    {
+      "epoch": 0.2670315789473684,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.000425621670382626,
+      "loss": 3.1289,
+      "step": 6342
+    },
+    {
+      "epoch": 0.2670736842105263,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00042559765308236146,
+      "loss": 3.2794,
+      "step": 6343
+    },
+    {
+      "epoch": 0.2671157894736842,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00042557363258291406,
+      "loss": 3.2385,
+      "step": 6344
+    },
+    {
+      "epoch": 0.2671578947368421,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004255496088847214,
+      "loss": 3.6462,
+      "step": 6345
+    },
+    {
+      "epoch": 0.2672,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.000425525581988221,
+      "loss": 3.5533,
+      "step": 6346
+    },
+    {
+      "epoch": 0.2672421052631579,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00042550155189385083,
+      "loss": 3.5518,
+      "step": 6347
+    },
+    {
+      "epoch": 0.2672842105263158,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0004254775186020485,
+      "loss": 3.0705,
+      "step": 6348
+    },
+    {
+      "epoch": 0.2673263157894737,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00042545348211325205,
+      "loss": 3.3356,
+      "step": 6349
+    },
+    {
+      "epoch": 0.2673684210526316,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004254294424278993,
+      "loss": 3.2119,
+      "step": 6350
+    },
+    {
+      "epoch": 0.2674105263157895,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004254053995464282,
+      "loss": 3.348,
+      "step": 6351
+    },
+    {
+      "epoch": 0.2674526315789474,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00042538135346927686,
+      "loss": 3.1821,
+      "step": 6352
+    },
+    {
+      "epoch": 0.2674947368421053,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00042535730419688325,
+      "loss": 3.0415,
+      "step": 6353
+    },
+    {
+      "epoch": 0.2675368421052632,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004253332517296856,
+      "loss": 3.2275,
+      "step": 6354
+    },
+    {
+      "epoch": 0.267578947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00042530919606812215,
+      "loss": 3.6709,
+      "step": 6355
+    },
+    {
+      "epoch": 0.2676210526315789,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0004252851372126312,
+      "loss": 3.2427,
+      "step": 6356
+    },
+    {
+      "epoch": 0.2676631578947368,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0004252610751636509,
+      "loss": 3.3825,
+      "step": 6357
+    },
+    {
+      "epoch": 0.2677052631578947,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00042523700992161975,
+      "loss": 3.0383,
+      "step": 6358
+    },
+    {
+      "epoch": 0.2677473684210526,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004252129414869762,
+      "loss": 3.279,
+      "step": 6359
+    },
+    {
+      "epoch": 0.2677894736842105,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00042518886986015863,
+      "loss": 3.3094,
+      "step": 6360
+    },
+    {
+      "epoch": 0.2678315789473684,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00042516479504160575,
+      "loss": 3.6452,
+      "step": 6361
+    },
+    {
+      "epoch": 0.2678736842105263,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00042514071703175614,
+      "loss": 3.4311,
+      "step": 6362
+    },
+    {
+      "epoch": 0.2679157894736842,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004251166358310484,
+      "loss": 2.8366,
+      "step": 6363
+    },
+    {
+      "epoch": 0.2679578947368421,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004250925514399214,
+      "loss": 2.9144,
+      "step": 6364
+    },
+    {
+      "epoch": 0.268,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00042506846385881374,
+      "loss": 3.5243,
+      "step": 6365
+    },
+    {
+      "epoch": 0.2680421052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00042504437308816444,
+      "loss": 3.1334,
+      "step": 6366
+    },
+    {
+      "epoch": 0.2680842105263158,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00042502027912841224,
+      "loss": 3.4249,
+      "step": 6367
+    },
+    {
+      "epoch": 0.26812631578947366,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004249961819799963,
+      "loss": 3.6981,
+      "step": 6368
+    },
+    {
+      "epoch": 0.26816842105263156,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004249720816433555,
+      "loss": 3.2701,
+      "step": 6369
+    },
+    {
+      "epoch": 0.26821052631578945,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00042494797811892896,
+      "loss": 3.3865,
+      "step": 6370
+    },
+    {
+      "epoch": 0.26825263157894735,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00042492387140715575,
+      "loss": 3.1325,
+      "step": 6371
+    },
+    {
+      "epoch": 0.26829473684210525,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004248997615084752,
+      "loss": 3.2783,
+      "step": 6372
+    },
+    {
+      "epoch": 0.26833684210526315,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00042487564842332644,
+      "loss": 3.6164,
+      "step": 6373
+    },
+    {
+      "epoch": 0.26837894736842105,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004248515321521489,
+      "loss": 2.3453,
+      "step": 6374
+    },
+    {
+      "epoch": 0.26842105263157895,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00042482741269538185,
+      "loss": 3.4442,
+      "step": 6375
+    },
+    {
+      "epoch": 0.26846315789473685,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004248032900534647,
+      "loss": 2.8978,
+      "step": 6376
+    },
+    {
+      "epoch": 0.26850526315789475,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0004247791642268371,
+      "loss": 3.0188,
+      "step": 6377
+    },
+    {
+      "epoch": 0.26854736842105265,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00042475503521593837,
+      "loss": 3.1182,
+      "step": 6378
+    },
+    {
+      "epoch": 0.26858947368421054,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0004247309030212084,
+      "loss": 3.238,
+      "step": 6379
+    },
+    {
+      "epoch": 0.26863157894736844,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004247067676430865,
+      "loss": 3.3033,
+      "step": 6380
+    },
+    {
+      "epoch": 0.26867368421052634,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0004246826290820126,
+      "loss": 3.1905,
+      "step": 6381
+    },
+    {
+      "epoch": 0.2687157894736842,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004246584873384265,
+      "loss": 2.9784,
+      "step": 6382
+    },
+    {
+      "epoch": 0.2687578947368421,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00042463434241276797,
+      "loss": 3.4526,
+      "step": 6383
+    },
+    {
+      "epoch": 0.2688,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004246101943054769,
+      "loss": 2.9956,
+      "step": 6384
+    },
+    {
+      "epoch": 0.2688421052631579,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0004245860430169932,
+      "loss": 2.9979,
+      "step": 6385
+    },
+    {
+      "epoch": 0.2688842105263158,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.000424561888547757,
+      "loss": 3.0014,
+      "step": 6386
+    },
+    {
+      "epoch": 0.2689263157894737,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004245377308982082,
+      "loss": 2.8607,
+      "step": 6387
+    },
+    {
+      "epoch": 0.2689684210526316,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004245135700687871,
+      "loss": 3.2286,
+      "step": 6388
+    },
+    {
+      "epoch": 0.2690105263157895,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004244894060599338,
+      "loss": 3.1448,
+      "step": 6389
+    },
+    {
+      "epoch": 0.2690526315789474,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00042446523887208857,
+      "loss": 3.2772,
+      "step": 6390
+    },
+    {
+      "epoch": 0.2690947368421053,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00042444106850569153,
+      "loss": 2.9565,
+      "step": 6391
+    },
+    {
+      "epoch": 0.2691368421052632,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004244168949611833,
+      "loss": 3.6742,
+      "step": 6392
+    },
+    {
+      "epoch": 0.2691789473684211,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004243927182390042,
+      "loss": 3.1222,
+      "step": 6393
+    },
+    {
+      "epoch": 0.26922105263157897,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00042436853833959453,
+      "loss": 2.6047,
+      "step": 6394
+    },
+    {
+      "epoch": 0.2692631578947368,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00042434435526339504,
+      "loss": 3.123,
+      "step": 6395
+    },
+    {
+      "epoch": 0.2693052631578947,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00042432016901084623,
+      "loss": 3.3929,
+      "step": 6396
+    },
+    {
+      "epoch": 0.2693473684210526,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004242959795823889,
+      "loss": 3.2623,
+      "step": 6397
+    },
+    {
+      "epoch": 0.2693894736842105,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00042427178697846345,
+      "loss": 2.6597,
+      "step": 6398
+    },
+    {
+      "epoch": 0.2694315789473684,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00042424759119951083,
+      "loss": 3.2708,
+      "step": 6399
+    },
+    {
+      "epoch": 0.2694736842105263,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00042422339224597186,
+      "loss": 2.5968,
+      "step": 6400
+    },
+    {
+      "epoch": 0.2695157894736842,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004241991901182873,
+      "loss": 3.6032,
+      "step": 6401
+    },
+    {
+      "epoch": 0.2695578947368421,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0004241749848168983,
+      "loss": 3.1867,
+      "step": 6402
+    },
+    {
+      "epoch": 0.2696,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004241507763422457,
+      "loss": 3.017,
+      "step": 6403
+    },
+    {
+      "epoch": 0.2696421052631579,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0004241265646947706,
+      "loss": 3.219,
+      "step": 6404
+    },
+    {
+      "epoch": 0.2696842105263158,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.000424102349874914,
+      "loss": 3.4647,
+      "step": 6405
+    },
+    {
+      "epoch": 0.2697263157894737,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00042407813188311715,
+      "loss": 2.9676,
+      "step": 6406
+    },
+    {
+      "epoch": 0.2697684210526316,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004240539107198213,
+      "loss": 3.4421,
+      "step": 6407
+    },
+    {
+      "epoch": 0.2698105263157895,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004240296863854677,
+      "loss": 3.0129,
+      "step": 6408
+    },
+    {
+      "epoch": 0.26985263157894734,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004240054588804977,
+      "loss": 3.1717,
+      "step": 6409
+    },
+    {
+      "epoch": 0.26989473684210524,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00042398122820535265,
+      "loss": 3.5012,
+      "step": 6410
+    },
+    {
+      "epoch": 0.26993684210526314,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004239569943604741,
+      "loss": 3.3762,
+      "step": 6411
+    },
+    {
+      "epoch": 0.26997894736842104,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004239327573463034,
+      "loss": 3.393,
+      "step": 6412
+    },
+    {
+      "epoch": 0.27002105263157894,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004239085171632824,
+      "loss": 3.2879,
+      "step": 6413
+    },
+    {
+      "epoch": 0.27006315789473684,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00042388427381185234,
+      "loss": 3.163,
+      "step": 6414
+    },
+    {
+      "epoch": 0.27010526315789474,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0004238600272924553,
+      "loss": 3.435,
+      "step": 6415
+    },
+    {
+      "epoch": 0.27014736842105264,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004238357776055327,
+      "loss": 3.2586,
+      "step": 6416
+    },
+    {
+      "epoch": 0.27018947368421053,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00042381152475152664,
+      "loss": 3.2817,
+      "step": 6417
+    },
+    {
+      "epoch": 0.27023157894736843,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004237872687308787,
+      "loss": 3.0169,
+      "step": 6418
+    },
+    {
+      "epoch": 0.27027368421052633,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004237630095440309,
+      "loss": 3.4522,
+      "step": 6419
+    },
+    {
+      "epoch": 0.27031578947368423,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00042373874719142527,
+      "loss": 3.6355,
+      "step": 6420
+    },
+    {
+      "epoch": 0.27035789473684213,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004237144816735038,
+      "loss": 3.6477,
+      "step": 6421
+    },
+    {
+      "epoch": 0.2704,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00042369021299070857,
+      "loss": 3.39,
+      "step": 6422
+    },
+    {
+      "epoch": 0.2704421052631579,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00042366594114348177,
+      "loss": 3.0867,
+      "step": 6423
+    },
+    {
+      "epoch": 0.27048421052631577,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004236416661322655,
+      "loss": 3.5644,
+      "step": 6424
+    },
+    {
+      "epoch": 0.27052631578947367,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00042361738795750216,
+      "loss": 3.6193,
+      "step": 6425
+    },
+    {
+      "epoch": 0.27056842105263157,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.000423593106619634,
+      "loss": 3.4706,
+      "step": 6426
+    },
+    {
+      "epoch": 0.27061052631578947,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00042356882211910335,
+      "loss": 3.6939,
+      "step": 6427
+    },
+    {
+      "epoch": 0.27065263157894737,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00042354453445635275,
+      "loss": 3.3058,
+      "step": 6428
+    },
+    {
+      "epoch": 0.27069473684210527,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004235202436318246,
+      "loss": 3.2101,
+      "step": 6429
+    },
+    {
+      "epoch": 0.27073684210526316,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004234959496459615,
+      "loss": 3.69,
+      "step": 6430
+    },
+    {
+      "epoch": 0.27077894736842106,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004234716524992061,
+      "loss": 2.8015,
+      "step": 6431
+    },
+    {
+      "epoch": 0.27082105263157896,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004234473521920009,
+      "loss": 3.2596,
+      "step": 6432
+    },
+    {
+      "epoch": 0.27086315789473686,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004234230487247889,
+      "loss": 2.9416,
+      "step": 6433
+    },
+    {
+      "epoch": 0.27090526315789476,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00042339874209801256,
+      "loss": 3.4046,
+      "step": 6434
+    },
+    {
+      "epoch": 0.27094736842105266,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00042337443231211494,
+      "loss": 3.3399,
+      "step": 6435
+    },
+    {
+      "epoch": 0.2709894736842105,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004233501193675389,
+      "loss": 3.8201,
+      "step": 6436
+    },
+    {
+      "epoch": 0.2710315789473684,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004233258032647273,
+      "loss": 3.6096,
+      "step": 6437
+    },
+    {
+      "epoch": 0.2710736842105263,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004233014840041233,
+      "loss": 3.1883,
+      "step": 6438
+    },
+    {
+      "epoch": 0.2711157894736842,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004232771615861698,
+      "loss": 3.4065,
+      "step": 6439
+    },
+    {
+      "epoch": 0.2711578947368421,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00042325283601131003,
+      "loss": 3.2293,
+      "step": 6440
+    },
+    {
+      "epoch": 0.2712,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00042322850727998705,
+      "loss": 3.0446,
+      "step": 6441
+    },
+    {
+      "epoch": 0.2712421052631579,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004232041753926443,
+      "loss": 3.4257,
+      "step": 6442
+    },
+    {
+      "epoch": 0.2712842105263158,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00042317984034972506,
+      "loss": 3.6864,
+      "step": 6443
+    },
+    {
+      "epoch": 0.2713263157894737,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00042315550215167255,
+      "loss": 3.0599,
+      "step": 6444
+    },
+    {
+      "epoch": 0.2713684210526316,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004231311607989302,
+      "loss": 3.3024,
+      "step": 6445
+    },
+    {
+      "epoch": 0.2714105263157895,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00042310681629194156,
+      "loss": 2.4417,
+      "step": 6446
+    },
+    {
+      "epoch": 0.2714526315789474,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004230824686311501,
+      "loss": 3.0827,
+      "step": 6447
+    },
+    {
+      "epoch": 0.2714947368421053,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00042305811781699946,
+      "loss": 3.2718,
+      "step": 6448
+    },
+    {
+      "epoch": 0.27153684210526313,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004230337638499332,
+      "loss": 3.2378,
+      "step": 6449
+    },
+    {
+      "epoch": 0.27157894736842103,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.000423009406730395,
+      "loss": 3.3503,
+      "step": 6450
+    },
+    {
+      "epoch": 0.27162105263157893,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004229850464588288,
+      "loss": 3.6389,
+      "step": 6451
+    },
+    {
+      "epoch": 0.27166315789473683,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0004229606830356783,
+      "loss": 3.2197,
+      "step": 6452
+    },
+    {
+      "epoch": 0.2717052631578947,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00042293631646138737,
+      "loss": 3.3798,
+      "step": 6453
+    },
+    {
+      "epoch": 0.2717473684210526,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004229119467363999,
+      "loss": 2.9206,
+      "step": 6454
+    },
+    {
+      "epoch": 0.2717894736842105,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00042288757386115997,
+      "loss": 3.2362,
+      "step": 6455
+    },
+    {
+      "epoch": 0.2718315789473684,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00042286319783611147,
+      "loss": 2.6049,
+      "step": 6456
+    },
+    {
+      "epoch": 0.2718736842105263,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00042283881866169874,
+      "loss": 3.3568,
+      "step": 6457
+    },
+    {
+      "epoch": 0.2719157894736842,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004228144363383658,
+      "loss": 3.5674,
+      "step": 6458
+    },
+    {
+      "epoch": 0.2719578947368421,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004227900508665569,
+      "loss": 3.2902,
+      "step": 6459
+    },
+    {
+      "epoch": 0.272,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004227656622467162,
+      "loss": 3.3236,
+      "step": 6460
+    },
+    {
+      "epoch": 0.2720421052631579,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00042274127047928813,
+      "loss": 3.5077,
+      "step": 6461
+    },
+    {
+      "epoch": 0.27208421052631576,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00042271687556471715,
+      "loss": 3.5554,
+      "step": 6462
+    },
+    {
+      "epoch": 0.27212631578947366,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00042269247750344755,
+      "loss": 2.8024,
+      "step": 6463
+    },
+    {
+      "epoch": 0.27216842105263156,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00042266807629592396,
+      "loss": 2.8508,
+      "step": 6464
+    },
+    {
+      "epoch": 0.27221052631578946,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004226436719425909,
+      "loss": 3.3113,
+      "step": 6465
+    },
+    {
+      "epoch": 0.27225263157894736,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.000422619264443893,
+      "loss": 3.0102,
+      "step": 6466
+    },
+    {
+      "epoch": 0.27229473684210526,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004225948538002749,
+      "loss": 3.4644,
+      "step": 6467
+    },
+    {
+      "epoch": 0.27233684210526315,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00042257044001218136,
+      "loss": 3.2368,
+      "step": 6468
+    },
+    {
+      "epoch": 0.27237894736842105,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00042254602308005717,
+      "loss": 2.8964,
+      "step": 6469
+    },
+    {
+      "epoch": 0.27242105263157895,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00042252160300434717,
+      "loss": 3.5766,
+      "step": 6470
+    },
+    {
+      "epoch": 0.27246315789473685,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00042249717978549625,
+      "loss": 3.3034,
+      "step": 6471
+    },
+    {
+      "epoch": 0.27250526315789475,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00042247275342394944,
+      "loss": 3.0497,
+      "step": 6472
+    },
+    {
+      "epoch": 0.27254736842105265,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004224483239201516,
+      "loss": 3.6123,
+      "step": 6473
+    },
+    {
+      "epoch": 0.27258947368421055,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.000422423891274548,
+      "loss": 3.3051,
+      "step": 6474
+    },
+    {
+      "epoch": 0.27263157894736845,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00042239945548758364,
+      "loss": 2.8966,
+      "step": 6475
+    },
+    {
+      "epoch": 0.2726736842105263,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00042237501655970387,
+      "loss": 3.1048,
+      "step": 6476
+    },
+    {
+      "epoch": 0.2727157894736842,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00042235057449135373,
+      "loss": 3.0976,
+      "step": 6477
+    },
+    {
+      "epoch": 0.2727578947368421,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00042232612928297863,
+      "loss": 3.3886,
+      "step": 6478
+    },
+    {
+      "epoch": 0.2728,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00042230168093502395,
+      "loss": 3.4853,
+      "step": 6479
+    },
+    {
+      "epoch": 0.2728421052631579,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00042227722944793513,
+      "loss": 3.2212,
+      "step": 6480
+    },
+    {
+      "epoch": 0.2728842105263158,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004222527748221575,
+      "loss": 2.7216,
+      "step": 6481
+    },
+    {
+      "epoch": 0.2729263157894737,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004222283170581368,
+      "loss": 3.4158,
+      "step": 6482
+    },
+    {
+      "epoch": 0.2729684210526316,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004222038561563185,
+      "loss": 3.5475,
+      "step": 6483
+    },
+    {
+      "epoch": 0.2730105263157895,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0004221793921171483,
+      "loss": 3.5681,
+      "step": 6484
+    },
+    {
+      "epoch": 0.2730526315789474,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00042215492494107186,
+      "loss": 3.8476,
+      "step": 6485
+    },
+    {
+      "epoch": 0.2730947368421053,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00042213045462853496,
+      "loss": 3.378,
+      "step": 6486
+    },
+    {
+      "epoch": 0.2731368421052632,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00042210598117998344,
+      "loss": 3.1606,
+      "step": 6487
+    },
+    {
+      "epoch": 0.2731789473684211,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.000422081504595863,
+      "loss": 3.1243,
+      "step": 6488
+    },
+    {
+      "epoch": 0.2732210526315789,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004220570248766199,
+      "loss": 3.2083,
+      "step": 6489
+    },
+    {
+      "epoch": 0.2732631578947368,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00042203254202269997,
+      "loss": 2.8056,
+      "step": 6490
+    },
+    {
+      "epoch": 0.2733052631578947,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004220080560345492,
+      "loss": 3.3898,
+      "step": 6491
+    },
+    {
+      "epoch": 0.2733473684210526,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00042198356691261376,
+      "loss": 2.8662,
+      "step": 6492
+    },
+    {
+      "epoch": 0.2733894736842105,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00042195907465733984,
+      "loss": 3.4138,
+      "step": 6493
+    },
+    {
+      "epoch": 0.2734315789473684,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00042193457926917354,
+      "loss": 3.4618,
+      "step": 6494
+    },
+    {
+      "epoch": 0.2734736842105263,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004219100807485613,
+      "loss": 3.1989,
+      "step": 6495
+    },
+    {
+      "epoch": 0.2735157894736842,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004218855790959493,
+      "loss": 2.4662,
+      "step": 6496
+    },
+    {
+      "epoch": 0.2735578947368421,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004218610743117841,
+      "loss": 3.424,
+      "step": 6497
+    },
+    {
+      "epoch": 0.2736,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.000421836566396512,
+      "loss": 3.268,
+      "step": 6498
+    },
+    {
+      "epoch": 0.2736421052631579,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00042181205535057957,
+      "loss": 3.4284,
+      "step": 6499
+    },
+    {
+      "epoch": 0.2736842105263158,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00042178754117443336,
+      "loss": 3.2268,
+      "step": 6500
+    },
+    {
+      "epoch": 0.2737263157894737,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004217630238685199,
+      "loss": 3.1248,
+      "step": 6501
+    },
+    {
+      "epoch": 0.2737684210526316,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004217385034332861,
+      "loss": 3.4265,
+      "step": 6502
+    },
+    {
+      "epoch": 0.27381052631578945,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004217139798691785,
+      "loss": 3.4103,
+      "step": 6503
+    },
+    {
+      "epoch": 0.27385263157894735,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.000421689453176644,
+      "loss": 3.1926,
+      "step": 6504
+    },
+    {
+      "epoch": 0.27389473684210525,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004216649233561293,
+      "loss": 3.3586,
+      "step": 6505
+    },
+    {
+      "epoch": 0.27393684210526315,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00042164039040808144,
+      "loss": 3.8455,
+      "step": 6506
+    },
+    {
+      "epoch": 0.27397894736842104,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00042161585433294734,
+      "loss": 3.4343,
+      "step": 6507
+    },
+    {
+      "epoch": 0.27402105263157894,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.000421591315131174,
+      "loss": 3.1182,
+      "step": 6508
+    },
+    {
+      "epoch": 0.27406315789473684,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0004215667728032084,
+      "loss": 3.1718,
+      "step": 6509
+    },
+    {
+      "epoch": 0.27410526315789474,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00042154222734949796,
+      "loss": 3.0234,
+      "step": 6510
+    },
+    {
+      "epoch": 0.27414736842105264,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004215176787704896,
+      "loss": 3.0397,
+      "step": 6511
+    },
+    {
+      "epoch": 0.27418947368421054,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00042149312706663065,
+      "loss": 3.5663,
+      "step": 6512
+    },
+    {
+      "epoch": 0.27423157894736844,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00042146857223836854,
+      "loss": 3.6205,
+      "step": 6513
+    },
+    {
+      "epoch": 0.27427368421052634,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00042144401428615043,
+      "loss": 3.2493,
+      "step": 6514
+    },
+    {
+      "epoch": 0.27431578947368424,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00042141945321042375,
+      "loss": 3.5602,
+      "step": 6515
+    },
+    {
+      "epoch": 0.2743578947368421,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00042139488901163613,
+      "loss": 3.4949,
+      "step": 6516
+    },
+    {
+      "epoch": 0.2744,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004213703216902349,
+      "loss": 3.2818,
+      "step": 6517
+    },
+    {
+      "epoch": 0.2744421052631579,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0004213457512466679,
+      "loss": 3.3253,
+      "step": 6518
+    },
+    {
+      "epoch": 0.2744842105263158,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004213211776813826,
+      "loss": 3.3786,
+      "step": 6519
+    },
+    {
+      "epoch": 0.2745263157894737,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004212966009948267,
+      "loss": 3.6269,
+      "step": 6520
+    },
+    {
+      "epoch": 0.2745684210526316,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.000421272021187448,
+      "loss": 3.4241,
+      "step": 6521
+    },
+    {
+      "epoch": 0.27461052631578947,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004212474382596943,
+      "loss": 3.2873,
+      "step": 6522
+    },
+    {
+      "epoch": 0.27465263157894737,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004212228522120135,
+      "loss": 3.386,
+      "step": 6523
+    },
+    {
+      "epoch": 0.27469473684210527,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00042119826304485355,
+      "loss": 2.8141,
+      "step": 6524
+    },
+    {
+      "epoch": 0.27473684210526317,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004211736707586624,
+      "loss": 3.4275,
+      "step": 6525
+    },
+    {
+      "epoch": 0.27477894736842107,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.000421149075353888,
+      "loss": 3.5436,
+      "step": 6526
+    },
+    {
+      "epoch": 0.27482105263157897,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00042112447683097866,
+      "loss": 3.1321,
+      "step": 6527
+    },
+    {
+      "epoch": 0.27486315789473686,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00042109987519038227,
+      "loss": 3.5082,
+      "step": 6528
+    },
+    {
+      "epoch": 0.27490526315789476,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00042107527043254726,
+      "loss": 3.2106,
+      "step": 6529
+    },
+    {
+      "epoch": 0.2749473684210526,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00042105066255792185,
+      "loss": 3.3604,
+      "step": 6530
+    },
+    {
+      "epoch": 0.2749894736842105,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00042102605156695434,
+      "loss": 3.373,
+      "step": 6531
+    },
+    {
+      "epoch": 0.2750315789473684,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004210014374600931,
+      "loss": 3.4307,
+      "step": 6532
+    },
+    {
+      "epoch": 0.2750736842105263,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004209768202377866,
+      "loss": 2.9288,
+      "step": 6533
+    },
+    {
+      "epoch": 0.2751157894736842,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00042095219990048326,
+      "loss": 3.4639,
+      "step": 6534
+    },
+    {
+      "epoch": 0.2751578947368421,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00042092757644863175,
+      "loss": 3.4761,
+      "step": 6535
+    },
+    {
+      "epoch": 0.2752,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004209029498826806,
+      "loss": 2.8818,
+      "step": 6536
+    },
+    {
+      "epoch": 0.2752421052631579,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004208783202030786,
+      "loss": 3.2147,
+      "step": 6537
+    },
+    {
+      "epoch": 0.2752842105263158,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0004208536874102743,
+      "loss": 3.3226,
+      "step": 6538
+    },
+    {
+      "epoch": 0.2753263157894737,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004208290515047166,
+      "loss": 3.3759,
+      "step": 6539
+    },
+    {
+      "epoch": 0.2753684210526316,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004208044124868543,
+      "loss": 2.9062,
+      "step": 6540
+    },
+    {
+      "epoch": 0.2754105263157895,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0004207797703571363,
+      "loss": 3.7027,
+      "step": 6541
+    },
+    {
+      "epoch": 0.2754526315789474,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00042075512511601144,
+      "loss": 3.3396,
+      "step": 6542
+    },
+    {
+      "epoch": 0.27549473684210524,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.000420730476763929,
+      "loss": 2.8917,
+      "step": 6543
+    },
+    {
+      "epoch": 0.27553684210526314,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00042070582530133775,
+      "loss": 3.5631,
+      "step": 6544
+    },
+    {
+      "epoch": 0.27557894736842103,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00042068117072868695,
+      "loss": 2.7966,
+      "step": 6545
+    },
+    {
+      "epoch": 0.27562105263157893,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004206565130464257,
+      "loss": 2.8822,
+      "step": 6546
+    },
+    {
+      "epoch": 0.27566315789473683,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00042063185225500336,
+      "loss": 2.9495,
+      "step": 6547
+    },
+    {
+      "epoch": 0.27570526315789473,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00042060718835486917,
+      "loss": 3.3082,
+      "step": 6548
+    },
+    {
+      "epoch": 0.27574736842105263,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00042058252134647247,
+      "loss": 3.1263,
+      "step": 6549
+    },
+    {
+      "epoch": 0.27578947368421053,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004205578512302626,
+      "loss": 3.4752,
+      "step": 6550
+    },
+    {
+      "epoch": 0.2758315789473684,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00042053317800668915,
+      "loss": 2.8122,
+      "step": 6551
+    },
+    {
+      "epoch": 0.2758736842105263,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004205085016762016,
+      "loss": 2.996,
+      "step": 6552
+    },
+    {
+      "epoch": 0.2759157894736842,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00042048382223924943,
+      "loss": 3.4017,
+      "step": 6553
+    },
+    {
+      "epoch": 0.2759578947368421,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00042045913969628225,
+      "loss": 3.131,
+      "step": 6554
+    },
+    {
+      "epoch": 0.276,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004204344540477499,
+      "loss": 3.304,
+      "step": 6555
+    },
+    {
+      "epoch": 0.2760421052631579,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.000420409765294102,
+      "loss": 3.2075,
+      "step": 6556
+    },
+    {
+      "epoch": 0.27608421052631577,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004203850734357885,
+      "loss": 2.8884,
+      "step": 6557
+    },
+    {
+      "epoch": 0.27612631578947366,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00042036037847325913,
+      "loss": 3.4836,
+      "step": 6558
+    },
+    {
+      "epoch": 0.27616842105263156,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00042033568040696377,
+      "loss": 3.1663,
+      "step": 6559
+    },
+    {
+      "epoch": 0.27621052631578946,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004203109792373525,
+      "loss": 3.242,
+      "step": 6560
+    },
+    {
+      "epoch": 0.27625263157894736,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00042028627496487527,
+      "loss": 3.6265,
+      "step": 6561
+    },
+    {
+      "epoch": 0.27629473684210526,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00042026156758998213,
+      "loss": 2.8325,
+      "step": 6562
+    },
+    {
+      "epoch": 0.27633684210526316,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00042023685711312336,
+      "loss": 3.2276,
+      "step": 6563
+    },
+    {
+      "epoch": 0.27637894736842106,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004202121435347491,
+      "loss": 2.9697,
+      "step": 6564
+    },
+    {
+      "epoch": 0.27642105263157896,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00042018742685530954,
+      "loss": 3.2601,
+      "step": 6565
+    },
+    {
+      "epoch": 0.27646315789473686,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.000420162707075255,
+      "loss": 3.1864,
+      "step": 6566
+    },
+    {
+      "epoch": 0.27650526315789475,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004201379841950359,
+      "loss": 3.1265,
+      "step": 6567
+    },
+    {
+      "epoch": 0.27654736842105265,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0004201132582151026,
+      "loss": 3.1121,
+      "step": 6568
+    },
+    {
+      "epoch": 0.27658947368421055,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004200885291359056,
+      "loss": 2.7161,
+      "step": 6569
+    },
+    {
+      "epoch": 0.2766315789473684,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004200637969578955,
+      "loss": 3.0314,
+      "step": 6570
+    },
+    {
+      "epoch": 0.2766736842105263,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0004200390616815228,
+      "loss": 3.2215,
+      "step": 6571
+    },
+    {
+      "epoch": 0.2767157894736842,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0004200143233072382,
+      "loss": 3.1522,
+      "step": 6572
+    },
+    {
+      "epoch": 0.2767578947368421,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004199895818354924,
+      "loss": 3.1174,
+      "step": 6573
+    },
+    {
+      "epoch": 0.2768,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00041996483726673613,
+      "loss": 3.0369,
+      "step": 6574
+    },
+    {
+      "epoch": 0.2768421052631579,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00041994008960142014,
+      "loss": 3.1888,
+      "step": 6575
+    },
+    {
+      "epoch": 0.2768842105263158,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00041991533883999554,
+      "loss": 3.5151,
+      "step": 6576
+    },
+    {
+      "epoch": 0.2769263157894737,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.000419890584982913,
+      "loss": 2.7248,
+      "step": 6577
+    },
+    {
+      "epoch": 0.2769684210526316,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004198658280306237,
+      "loss": 3.175,
+      "step": 6578
+    },
+    {
+      "epoch": 0.2770105263157895,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00041984106798357857,
+      "loss": 3.4148,
+      "step": 6579
+    },
+    {
+      "epoch": 0.2770526315789474,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004198163048422287,
+      "loss": 3.0725,
+      "step": 6580
+    },
+    {
+      "epoch": 0.2770947368421053,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00041979153860702533,
+      "loss": 3.5595,
+      "step": 6581
+    },
+    {
+      "epoch": 0.2771368421052632,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004197667692784196,
+      "loss": 2.9202,
+      "step": 6582
+    },
+    {
+      "epoch": 0.2771789473684211,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00041974199685686274,
+      "loss": 3.2668,
+      "step": 6583
+    },
+    {
+      "epoch": 0.2772210526315789,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00041971722134280623,
+      "loss": 3.5529,
+      "step": 6584
+    },
+    {
+      "epoch": 0.2772631578947368,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004196924427367013,
+      "loss": 3.6094,
+      "step": 6585
+    },
+    {
+      "epoch": 0.2773052631578947,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00041966766103899945,
+      "loss": 3.4502,
+      "step": 6586
+    },
+    {
+      "epoch": 0.2773473684210526,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004196428762501522,
+      "loss": 3.1369,
+      "step": 6587
+    },
+    {
+      "epoch": 0.2773894736842105,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00041961808837061103,
+      "loss": 3.4338,
+      "step": 6588
+    },
+    {
+      "epoch": 0.2774315789473684,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00041959329740082766,
+      "loss": 3.1948,
+      "step": 6589
+    },
+    {
+      "epoch": 0.2774736842105263,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004195685033412536,
+      "loss": 3.252,
+      "step": 6590
+    },
+    {
+      "epoch": 0.2775157894736842,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00041954370619234064,
+      "loss": 3.2458,
+      "step": 6591
+    },
+    {
+      "epoch": 0.2775578947368421,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00041951890595454057,
+      "loss": 3.4346,
+      "step": 6592
+    },
+    {
+      "epoch": 0.2776,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00041949410262830523,
+      "loss": 3.5173,
+      "step": 6593
+    },
+    {
+      "epoch": 0.2776421052631579,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00041946929621408647,
+      "loss": 3.2384,
+      "step": 6594
+    },
+    {
+      "epoch": 0.2776842105263158,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0004194444867123363,
+      "loss": 3.1336,
+      "step": 6595
+    },
+    {
+      "epoch": 0.2777263157894737,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00041941967412350656,
+      "loss": 3.3995,
+      "step": 6596
+    },
+    {
+      "epoch": 0.27776842105263155,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004193948584480495,
+      "loss": 3.5149,
+      "step": 6597
+    },
+    {
+      "epoch": 0.27781052631578945,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004193700396864172,
+      "loss": 3.7719,
+      "step": 6598
+    },
+    {
+      "epoch": 0.27785263157894735,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004193452178390617,
+      "loss": 3.4228,
+      "step": 6599
+    },
+    {
+      "epoch": 0.27789473684210525,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004193203929064353,
+      "loss": 2.8364,
+      "step": 6600
+    },
+    {
+      "epoch": 0.27793684210526315,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00041929556488899027,
+      "loss": 2.9872,
+      "step": 6601
+    },
+    {
+      "epoch": 0.27797894736842105,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00041927073378717893,
+      "loss": 3.131,
+      "step": 6602
+    },
+    {
+      "epoch": 0.27802105263157895,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004192458996014539,
+      "loss": 3.5979,
+      "step": 6603
+    },
+    {
+      "epoch": 0.27806315789473685,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004192210623322673,
+      "loss": 3.0426,
+      "step": 6604
+    },
+    {
+      "epoch": 0.27810526315789474,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00041919622198007174,
+      "loss": 3.1596,
+      "step": 6605
+    },
+    {
+      "epoch": 0.27814736842105264,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00041917137854531985,
+      "loss": 3.474,
+      "step": 6606
+    },
+    {
+      "epoch": 0.27818947368421054,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004191465320284642,
+      "loss": 3.2664,
+      "step": 6607
+    },
+    {
+      "epoch": 0.27823157894736844,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00041912168242995754,
+      "loss": 3.2065,
+      "step": 6608
+    },
+    {
+      "epoch": 0.27827368421052634,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00041909682975025243,
+      "loss": 3.314,
+      "step": 6609
+    },
+    {
+      "epoch": 0.2783157894736842,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004190719739898018,
+      "loss": 3.0488,
+      "step": 6610
+    },
+    {
+      "epoch": 0.2783578947368421,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00041904711514905845,
+      "loss": 3.3811,
+      "step": 6611
+    },
+    {
+      "epoch": 0.2784,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0004190222532284753,
+      "loss": 3.0741,
+      "step": 6612
+    },
+    {
+      "epoch": 0.2784421052631579,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004189973882285053,
+      "loss": 2.911,
+      "step": 6613
+    },
+    {
+      "epoch": 0.2784842105263158,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004189725201496015,
+      "loss": 2.9579,
+      "step": 6614
+    },
+    {
+      "epoch": 0.2785263157894737,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004189476489922168,
+      "loss": 3.0509,
+      "step": 6615
+    },
+    {
+      "epoch": 0.2785684210526316,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004189227747568045,
+      "loss": 3.468,
+      "step": 6616
+    },
+    {
+      "epoch": 0.2786105263157895,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0004188978974438177,
+      "loss": 3.5568,
+      "step": 6617
+    },
+    {
+      "epoch": 0.2786526315789474,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00041887301705370973,
+      "loss": 3.4185,
+      "step": 6618
+    },
+    {
+      "epoch": 0.2786947368421053,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00041884813358693376,
+      "loss": 3.3872,
+      "step": 6619
+    },
+    {
+      "epoch": 0.27873684210526317,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00041882324704394316,
+      "loss": 3.4621,
+      "step": 6620
+    },
+    {
+      "epoch": 0.27877894736842107,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00041879835742519134,
+      "loss": 2.8777,
+      "step": 6621
+    },
+    {
+      "epoch": 0.27882105263157897,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004187734647311319,
+      "loss": 3.6372,
+      "step": 6622
+    },
+    {
+      "epoch": 0.27886315789473687,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0004187485689622181,
+      "loss": 3.2462,
+      "step": 6623
+    },
+    {
+      "epoch": 0.2789052631578947,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00041872367011890364,
+      "loss": 3.2563,
+      "step": 6624
+    },
+    {
+      "epoch": 0.2789473684210526,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004186987682016422,
+      "loss": 3.0996,
+      "step": 6625
+    },
+    {
+      "epoch": 0.2789894736842105,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004186738632108875,
+      "loss": 3.1638,
+      "step": 6626
+    },
+    {
+      "epoch": 0.2790315789473684,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004186489551470931,
+      "loss": 2.9084,
+      "step": 6627
+    },
+    {
+      "epoch": 0.2790736842105263,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004186240440107129,
+      "loss": 3.2046,
+      "step": 6628
+    },
+    {
+      "epoch": 0.2791157894736842,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0004185991298022007,
+      "loss": 3.2739,
+      "step": 6629
+    },
+    {
+      "epoch": 0.2791578947368421,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004185742125220106,
+      "loss": 3.1638,
+      "step": 6630
+    },
+    {
+      "epoch": 0.2792,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00041854929217059625,
+      "loss": 3.2391,
+      "step": 6631
+    },
+    {
+      "epoch": 0.2792421052631579,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00041852436874841194,
+      "loss": 3.1495,
+      "step": 6632
+    },
+    {
+      "epoch": 0.2792842105263158,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004184994422559116,
+      "loss": 3.1134,
+      "step": 6633
+    },
+    {
+      "epoch": 0.2793263157894737,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00041847451269354935,
+      "loss": 3.655,
+      "step": 6634
+    },
+    {
+      "epoch": 0.2793684210526316,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00041844958006177947,
+      "loss": 3.4538,
+      "step": 6635
+    },
+    {
+      "epoch": 0.2794105263157895,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0004184246443610562,
+      "loss": 3.3199,
+      "step": 6636
+    },
+    {
+      "epoch": 0.27945263157894734,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00041839970559183373,
+      "loss": 3.0707,
+      "step": 6637
+    },
+    {
+      "epoch": 0.27949473684210524,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.0004183747637545666,
+      "loss": 2.748,
+      "step": 6638
+    },
+    {
+      "epoch": 0.27953684210526314,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.000418349818849709,
+      "loss": 3.04,
+      "step": 6639
+    },
+    {
+      "epoch": 0.27957894736842104,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004183248708777155,
+      "loss": 3.2245,
+      "step": 6640
+    },
+    {
+      "epoch": 0.27962105263157894,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004182999198390407,
+      "loss": 3.211,
+      "step": 6641
+    },
+    {
+      "epoch": 0.27966315789473684,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00041827496573413903,
+      "loss": 3.2164,
+      "step": 6642
+    },
+    {
+      "epoch": 0.27970526315789473,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00041825000856346526,
+      "loss": 3.3138,
+      "step": 6643
+    },
+    {
+      "epoch": 0.27974736842105263,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.000418225048327474,
+      "loss": 3.0779,
+      "step": 6644
+    },
+    {
+      "epoch": 0.27978947368421053,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00041820008502662,
+      "loss": 3.2993,
+      "step": 6645
+    },
+    {
+      "epoch": 0.27983157894736843,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004181751186613581,
+      "loss": 3.5527,
+      "step": 6646
+    },
+    {
+      "epoch": 0.27987368421052633,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00041815014923214314,
+      "loss": 3.0518,
+      "step": 6647
+    },
+    {
+      "epoch": 0.27991578947368423,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00041812517673942993,
+      "loss": 3.2773,
+      "step": 6648
+    },
+    {
+      "epoch": 0.27995789473684213,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00041810020118367364,
+      "loss": 3.6576,
+      "step": 6649
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004180752225653292,
+      "loss": 3.8121,
+      "step": 6650
+    },
+    {
+      "epoch": 0.28004210526315787,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00041805024088485167,
+      "loss": 3.587,
+      "step": 6651
+    },
+    {
+      "epoch": 0.28008421052631577,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004180252561426962,
+      "loss": 3.6611,
+      "step": 6652
+    },
+    {
+      "epoch": 0.28012631578947367,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00041800026833931797,
+      "loss": 3.2335,
+      "step": 6653
+    },
+    {
+      "epoch": 0.28016842105263157,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004179752774751723,
+      "loss": 3.3893,
+      "step": 6654
+    },
+    {
+      "epoch": 0.28021052631578947,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00041795028355071437,
+      "loss": 3.2301,
+      "step": 6655
+    },
+    {
+      "epoch": 0.28025263157894736,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004179252865663996,
+      "loss": 3.3729,
+      "step": 6656
+    },
+    {
+      "epoch": 0.28029473684210526,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004179002865226834,
+      "loss": 2.9862,
+      "step": 6657
+    },
+    {
+      "epoch": 0.28033684210526316,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00041787528342002135,
+      "loss": 3.2229,
+      "step": 6658
+    },
+    {
+      "epoch": 0.28037894736842106,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004178502772588688,
+      "loss": 3.3188,
+      "step": 6659
+    },
+    {
+      "epoch": 0.28042105263157896,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00041782526803968147,
+      "loss": 3.4063,
+      "step": 6660
+    },
+    {
+      "epoch": 0.28046315789473686,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00041780025576291493,
+      "loss": 3.6166,
+      "step": 6661
+    },
+    {
+      "epoch": 0.28050526315789476,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00041777524042902484,
+      "loss": 3.4047,
+      "step": 6662
+    },
+    {
+      "epoch": 0.28054736842105266,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00041775022203846703,
+      "loss": 3.3336,
+      "step": 6663
+    },
+    {
+      "epoch": 0.2805894736842105,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0004177252005916972,
+      "loss": 3.5159,
+      "step": 6664
+    },
+    {
+      "epoch": 0.2806315789473684,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004177001760891714,
+      "loss": 3.2001,
+      "step": 6665
+    },
+    {
+      "epoch": 0.2806736842105263,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004176751485313454,
+      "loss": 3.1684,
+      "step": 6666
+    },
+    {
+      "epoch": 0.2807157894736842,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004176501179186752,
+      "loss": 3.5735,
+      "step": 6667
+    },
+    {
+      "epoch": 0.2807578947368421,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00041762508425161683,
+      "loss": 3.0811,
+      "step": 6668
+    },
+    {
+      "epoch": 0.2808,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004176000475306263,
+      "loss": 3.5004,
+      "step": 6669
+    },
+    {
+      "epoch": 0.2808421052631579,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004175750077561599,
+      "loss": 3.5581,
+      "step": 6670
+    },
+    {
+      "epoch": 0.2808842105263158,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00041754996492867377,
+      "loss": 3.0918,
+      "step": 6671
+    },
+    {
+      "epoch": 0.2809263157894737,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.0004175249190486241,
+      "loss": 3.0459,
+      "step": 6672
+    },
+    {
+      "epoch": 0.2809684210526316,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00041749987011646726,
+      "loss": 3.1027,
+      "step": 6673
+    },
+    {
+      "epoch": 0.2810105263157895,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00041747481813265964,
+      "loss": 3.3682,
+      "step": 6674
+    },
+    {
+      "epoch": 0.2810526315789474,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004174497630976575,
+      "loss": 3.2241,
+      "step": 6675
+    },
+    {
+      "epoch": 0.2810947368421053,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004174247050119175,
+      "loss": 3.2557,
+      "step": 6676
+    },
+    {
+      "epoch": 0.2811368421052632,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.000417399643875896,
+      "loss": 3.0142,
+      "step": 6677
+    },
+    {
+      "epoch": 0.28117894736842103,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004173745796900498,
+      "loss": 3.4463,
+      "step": 6678
+    },
+    {
+      "epoch": 0.2812210526315789,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0004173495124548353,
+      "loss": 3.312,
+      "step": 6679
+    },
+    {
+      "epoch": 0.2812631578947368,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004173244421707094,
+      "loss": 3.6851,
+      "step": 6680
+    },
+    {
+      "epoch": 0.2813052631578947,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00041729936883812876,
+      "loss": 2.7317,
+      "step": 6681
+    },
+    {
+      "epoch": 0.2813473684210526,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004172742924575501,
+      "loss": 3.4062,
+      "step": 6682
+    },
+    {
+      "epoch": 0.2813894736842105,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00041724921302943043,
+      "loss": 3.6427,
+      "step": 6683
+    },
+    {
+      "epoch": 0.2814315789473684,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00041722413055422667,
+      "loss": 3.5665,
+      "step": 6684
+    },
+    {
+      "epoch": 0.2814736842105263,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004171990450323956,
+      "loss": 3.418,
+      "step": 6685
+    },
+    {
+      "epoch": 0.2815157894736842,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004171739564643945,
+      "loss": 3.4383,
+      "step": 6686
+    },
+    {
+      "epoch": 0.2815578947368421,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00041714886485068027,
+      "loss": 3.4488,
+      "step": 6687
+    },
+    {
+      "epoch": 0.2816,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004171237701917101,
+      "loss": 3.001,
+      "step": 6688
+    },
+    {
+      "epoch": 0.2816421052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004170986724879413,
+      "loss": 3.5818,
+      "step": 6689
+    },
+    {
+      "epoch": 0.2816842105263158,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00041707357173983084,
+      "loss": 3.2684,
+      "step": 6690
+    },
+    {
+      "epoch": 0.28172631578947366,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00041704846794783635,
+      "loss": 3.2806,
+      "step": 6691
+    },
+    {
+      "epoch": 0.28176842105263156,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.000417023361112415,
+      "loss": 3.0444,
+      "step": 6692
+    },
+    {
+      "epoch": 0.28181052631578946,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004169982512340242,
+      "loss": 3.3491,
+      "step": 6693
+    },
+    {
+      "epoch": 0.28185263157894735,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00041697313831312165,
+      "loss": 3.3791,
+      "step": 6694
+    },
+    {
+      "epoch": 0.28189473684210525,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004169480223501646,
+      "loss": 3.1435,
+      "step": 6695
+    },
+    {
+      "epoch": 0.28193684210526315,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004169229033456107,
+      "loss": 3.6401,
+      "step": 6696
+    },
+    {
+      "epoch": 0.28197894736842105,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00041689778129991767,
+      "loss": 3.3951,
+      "step": 6697
+    },
+    {
+      "epoch": 0.28202105263157895,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00041687265621354313,
+      "loss": 3.2654,
+      "step": 6698
+    },
+    {
+      "epoch": 0.28206315789473685,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004168475280869449,
+      "loss": 3.5863,
+      "step": 6699
+    },
+    {
+      "epoch": 0.28210526315789475,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0004168223969205808,
+      "loss": 2.6782,
+      "step": 6700
+    },
+    {
+      "epoch": 0.28214736842105265,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00041679726271490853,
+      "loss": 3.1562,
+      "step": 6701
+    },
+    {
+      "epoch": 0.28218947368421055,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00041677212547038613,
+      "loss": 3.4072,
+      "step": 6702
+    },
+    {
+      "epoch": 0.28223157894736844,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00041674698518747154,
+      "loss": 2.9622,
+      "step": 6703
+    },
+    {
+      "epoch": 0.28227368421052634,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00041672184186662285,
+      "loss": 3.2485,
+      "step": 6704
+    },
+    {
+      "epoch": 0.2823157894736842,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0004166966955082981,
+      "loss": 3.4762,
+      "step": 6705
+    },
+    {
+      "epoch": 0.2823578947368421,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004166715461129554,
+      "loss": 3.064,
+      "step": 6706
+    },
+    {
+      "epoch": 0.2824,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00041664639368105293,
+      "loss": 3.3304,
+      "step": 6707
+    },
+    {
+      "epoch": 0.2824421052631579,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00041662123821304905,
+      "loss": 3.1445,
+      "step": 6708
+    },
+    {
+      "epoch": 0.2824842105263158,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004165960797094019,
+      "loss": 2.8116,
+      "step": 6709
+    },
+    {
+      "epoch": 0.2825263157894737,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004165709181705699,
+      "loss": 3.5259,
+      "step": 6710
+    },
+    {
+      "epoch": 0.2825684210526316,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004165457535970115,
+      "loss": 3.1078,
+      "step": 6711
+    },
+    {
+      "epoch": 0.2826105263157895,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00041652058598918527,
+      "loss": 3.404,
+      "step": 6712
+    },
+    {
+      "epoch": 0.2826526315789474,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00041649541534754956,
+      "loss": 3.4321,
+      "step": 6713
+    },
+    {
+      "epoch": 0.2826947368421053,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0004164702416725629,
+      "loss": 3.1983,
+      "step": 6714
+    },
+    {
+      "epoch": 0.2827368421052632,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004164450649646842,
+      "loss": 3.6488,
+      "step": 6715
+    },
+    {
+      "epoch": 0.2827789473684211,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0004164198852243718,
+      "loss": 3.1648,
+      "step": 6716
+    },
+    {
+      "epoch": 0.282821052631579,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00041639470245208467,
+      "loss": 3.7616,
+      "step": 6717
+    },
+    {
+      "epoch": 0.2828631578947368,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00041636951664828165,
+      "loss": 3.5557,
+      "step": 6718
+    },
+    {
+      "epoch": 0.2829052631578947,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00041634432781342145,
+      "loss": 3.2865,
+      "step": 6719
+    },
+    {
+      "epoch": 0.2829473684210526,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00041631913594796313,
+      "loss": 3.4424,
+      "step": 6720
+    },
+    {
+      "epoch": 0.2829894736842105,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004162939410523655,
+      "loss": 3.6272,
+      "step": 6721
+    },
+    {
+      "epoch": 0.2830315789473684,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004162687431270876,
+      "loss": 2.9641,
+      "step": 6722
+    },
+    {
+      "epoch": 0.2830736842105263,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004162435421725886,
+      "loss": 3.3319,
+      "step": 6723
+    },
+    {
+      "epoch": 0.2831157894736842,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00041621833818932766,
+      "loss": 3.195,
+      "step": 6724
+    },
+    {
+      "epoch": 0.2831578947368421,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00041619313117776383,
+      "loss": 3.195,
+      "step": 6725
+    },
+    {
+      "epoch": 0.2832,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00041616792113835646,
+      "loss": 2.6857,
+      "step": 6726
+    },
+    {
+      "epoch": 0.2832421052631579,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00041614270807156475,
+      "loss": 3.6809,
+      "step": 6727
+    },
+    {
+      "epoch": 0.2832842105263158,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00041611749197784814,
+      "loss": 3.3442,
+      "step": 6728
+    },
+    {
+      "epoch": 0.2833263157894737,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00041609227285766605,
+      "loss": 3.6784,
+      "step": 6729
+    },
+    {
+      "epoch": 0.2833684210526316,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00041606705071147777,
+      "loss": 3.2473,
+      "step": 6730
+    },
+    {
+      "epoch": 0.28341052631578945,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00041604182553974303,
+      "loss": 3.4388,
+      "step": 6731
+    },
+    {
+      "epoch": 0.28345263157894734,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004160165973429213,
+      "loss": 3.1442,
+      "step": 6732
+    },
+    {
+      "epoch": 0.28349473684210524,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004159913661214723,
+      "loss": 3.254,
+      "step": 6733
+    },
+    {
+      "epoch": 0.28353684210526314,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004159661318758556,
+      "loss": 3.3356,
+      "step": 6734
+    },
+    {
+      "epoch": 0.28357894736842104,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00041594089460653095,
+      "loss": 3.3217,
+      "step": 6735
+    },
+    {
+      "epoch": 0.28362105263157894,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00041591565431395815,
+      "loss": 3.5002,
+      "step": 6736
+    },
+    {
+      "epoch": 0.28366315789473684,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00041589041099859717,
+      "loss": 3.1628,
+      "step": 6737
+    },
+    {
+      "epoch": 0.28370526315789474,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00041586516466090776,
+      "loss": 2.7701,
+      "step": 6738
+    },
+    {
+      "epoch": 0.28374736842105264,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004158399153013499,
+      "loss": 3.1499,
+      "step": 6739
+    },
+    {
+      "epoch": 0.28378947368421054,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004158146629203836,
+      "loss": 3.3434,
+      "step": 6740
+    },
+    {
+      "epoch": 0.28383157894736843,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00041578940751846907,
+      "loss": 3.0107,
+      "step": 6741
+    },
+    {
+      "epoch": 0.28387368421052633,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004157641490960663,
+      "loss": 3.0127,
+      "step": 6742
+    },
+    {
+      "epoch": 0.28391578947368423,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004157388876536355,
+      "loss": 3.1251,
+      "step": 6743
+    },
+    {
+      "epoch": 0.28395789473684213,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00041571362319163687,
+      "loss": 3.0175,
+      "step": 6744
+    },
+    {
+      "epoch": 0.284,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00041568835571053074,
+      "loss": 3.3101,
+      "step": 6745
+    },
+    {
+      "epoch": 0.2840421052631579,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00041566308521077743,
+      "loss": 3.0066,
+      "step": 6746
+    },
+    {
+      "epoch": 0.2840842105263158,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004156378116928374,
+      "loss": 3.0327,
+      "step": 6747
+    },
+    {
+      "epoch": 0.28412631578947367,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00041561253515717106,
+      "loss": 3.6476,
+      "step": 6748
+    },
+    {
+      "epoch": 0.28416842105263157,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0004155872556042389,
+      "loss": 3.353,
+      "step": 6749
+    },
+    {
+      "epoch": 0.28421052631578947,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004155619730345015,
+      "loss": 3.1492,
+      "step": 6750
+    },
+    {
+      "epoch": 0.28425263157894737,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004155366874484194,
+      "loss": 3.4902,
+      "step": 6751
+    },
+    {
+      "epoch": 0.28429473684210527,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004155113988464535,
+      "loss": 3.7805,
+      "step": 6752
+    },
+    {
+      "epoch": 0.28433684210526317,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00041548610722906435,
+      "loss": 3.4252,
+      "step": 6753
+    },
+    {
+      "epoch": 0.28437894736842106,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00041546081259671267,
+      "loss": 3.1236,
+      "step": 6754
+    },
+    {
+      "epoch": 0.28442105263157896,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004154355149498594,
+      "loss": 2.9728,
+      "step": 6755
+    },
+    {
+      "epoch": 0.28446315789473686,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00041541021428896554,
+      "loss": 3.3853,
+      "step": 6756
+    },
+    {
+      "epoch": 0.28450526315789476,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0004153849106144918,
+      "loss": 3.0118,
+      "step": 6757
+    },
+    {
+      "epoch": 0.2845473684210526,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00041535960392689934,
+      "loss": 2.996,
+      "step": 6758
+    },
+    {
+      "epoch": 0.2845894736842105,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00041533429422664923,
+      "loss": 2.8705,
+      "step": 6759
+    },
+    {
+      "epoch": 0.2846315789473684,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00041530898151420246,
+      "loss": 3.3637,
+      "step": 6760
+    },
+    {
+      "epoch": 0.2846736842105263,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004152836657900204,
+      "loss": 3.3062,
+      "step": 6761
+    },
+    {
+      "epoch": 0.2847157894736842,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0004152583470545641,
+      "loss": 3.4685,
+      "step": 6762
+    },
+    {
+      "epoch": 0.2847578947368421,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004152330253082949,
+      "loss": 3.2851,
+      "step": 6763
+    },
+    {
+      "epoch": 0.2848,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00041520770055167403,
+      "loss": 3.2389,
+      "step": 6764
+    },
+    {
+      "epoch": 0.2848421052631579,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00041518237278516303,
+      "loss": 3.4337,
+      "step": 6765
+    },
+    {
+      "epoch": 0.2848842105263158,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00041515704200922334,
+      "loss": 2.9614,
+      "step": 6766
+    },
+    {
+      "epoch": 0.2849263157894737,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004151317082243163,
+      "loss": 3.1639,
+      "step": 6767
+    },
+    {
+      "epoch": 0.2849684210526316,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00041510637143090355,
+      "loss": 2.8583,
+      "step": 6768
+    },
+    {
+      "epoch": 0.2850105263157895,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00041508103162944676,
+      "loss": 3.576,
+      "step": 6769
+    },
+    {
+      "epoch": 0.2850526315789474,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004150556888204075,
+      "loss": 3.0098,
+      "step": 6770
+    },
+    {
+      "epoch": 0.2850947368421053,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00041503034300424755,
+      "loss": 3.3303,
+      "step": 6771
+    },
+    {
+      "epoch": 0.28513684210526313,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00041500499418142866,
+      "loss": 3.5926,
+      "step": 6772
+    },
+    {
+      "epoch": 0.28517894736842103,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00041497964235241263,
+      "loss": 3.2274,
+      "step": 6773
+    },
+    {
+      "epoch": 0.28522105263157893,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004149542875176614,
+      "loss": 3.4937,
+      "step": 6774
+    },
+    {
+      "epoch": 0.28526315789473683,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00041492892967763686,
+      "loss": 3.526,
+      "step": 6775
+    },
+    {
+      "epoch": 0.28530526315789473,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004149035688328009,
+      "loss": 3.3338,
+      "step": 6776
+    },
+    {
+      "epoch": 0.2853473684210526,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00041487820498361573,
+      "loss": 3.4505,
+      "step": 6777
+    },
+    {
+      "epoch": 0.2853894736842105,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00041485283813054343,
+      "loss": 3.1628,
+      "step": 6778
+    },
+    {
+      "epoch": 0.2854315789473684,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00041482746827404605,
+      "loss": 3.3462,
+      "step": 6779
+    },
+    {
+      "epoch": 0.2854736842105263,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00041480209541458593,
+      "loss": 3.436,
+      "step": 6780
+    },
+    {
+      "epoch": 0.2855157894736842,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004147767195526252,
+      "loss": 3.7156,
+      "step": 6781
+    },
+    {
+      "epoch": 0.2855578947368421,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004147513406886263,
+      "loss": 3.022,
+      "step": 6782
+    },
+    {
+      "epoch": 0.2856,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004147259588230515,
+      "loss": 3.1655,
+      "step": 6783
+    },
+    {
+      "epoch": 0.2856421052631579,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00041470057395636334,
+      "loss": 3.5808,
+      "step": 6784
+    },
+    {
+      "epoch": 0.28568421052631576,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004146751860890242,
+      "loss": 3.3094,
+      "step": 6785
+    },
+    {
+      "epoch": 0.28572631578947366,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0004146497952214966,
+      "loss": 3.4299,
+      "step": 6786
+    },
+    {
+      "epoch": 0.28576842105263156,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004146244013542433,
+      "loss": 3.2844,
+      "step": 6787
+    },
+    {
+      "epoch": 0.28581052631578946,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004145990044877268,
+      "loss": 3.4305,
+      "step": 6788
+    },
+    {
+      "epoch": 0.28585263157894736,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00041457360462240976,
+      "loss": 3.4777,
+      "step": 6789
+    },
+    {
+      "epoch": 0.28589473684210526,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004145482017587551,
+      "loss": 3.348,
+      "step": 6790
+    },
+    {
+      "epoch": 0.28593684210526316,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00041452279589722547,
+      "loss": 3.1765,
+      "step": 6791
+    },
+    {
+      "epoch": 0.28597894736842105,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004144973870382838,
+      "loss": 3.4061,
+      "step": 6792
+    },
+    {
+      "epoch": 0.28602105263157895,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0004144719751823931,
+      "loss": 3.3084,
+      "step": 6793
+    },
+    {
+      "epoch": 0.28606315789473685,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004144465603300162,
+      "loss": 3.4772,
+      "step": 6794
+    },
+    {
+      "epoch": 0.28610526315789475,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00041442114248161624,
+      "loss": 3.1667,
+      "step": 6795
+    },
+    {
+      "epoch": 0.28614736842105265,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00041439572163765617,
+      "loss": 3.1921,
+      "step": 6796
+    },
+    {
+      "epoch": 0.28618947368421055,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00041437029779859926,
+      "loss": 3.3992,
+      "step": 6797
+    },
+    {
+      "epoch": 0.28623157894736845,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00041434487096490857,
+      "loss": 3.1245,
+      "step": 6798
+    },
+    {
+      "epoch": 0.2862736842105263,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004143194411370475,
+      "loss": 3.2488,
+      "step": 6799
+    },
+    {
+      "epoch": 0.2863157894736842,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004142940083154793,
+      "loss": 3.2958,
+      "step": 6800
+    },
+    {
+      "epoch": 0.2863578947368421,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004142685725006673,
+      "loss": 2.9586,
+      "step": 6801
+    },
+    {
+      "epoch": 0.2864,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00041424313369307485,
+      "loss": 3.1594,
+      "step": 6802
+    },
+    {
+      "epoch": 0.2864421052631579,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00041421769189316557,
+      "loss": 3.3788,
+      "step": 6803
+    },
+    {
+      "epoch": 0.2864842105263158,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004141922471014028,
+      "loss": 2.9266,
+      "step": 6804
+    },
+    {
+      "epoch": 0.2865263157894737,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004141667993182503,
+      "loss": 3.2032,
+      "step": 6805
+    },
+    {
+      "epoch": 0.2865684210526316,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00041414134854417153,
+      "loss": 3.2129,
+      "step": 6806
+    },
+    {
+      "epoch": 0.2866105263157895,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0004141158947796303,
+      "loss": 2.9108,
+      "step": 6807
+    },
+    {
+      "epoch": 0.2866526315789474,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00041409043802509025,
+      "loss": 3.3863,
+      "step": 6808
+    },
+    {
+      "epoch": 0.2866947368421053,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0004140649782810153,
+      "loss": 3.2424,
+      "step": 6809
+    },
+    {
+      "epoch": 0.2867368421052632,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004140395155478691,
+      "loss": 3.0535,
+      "step": 6810
+    },
+    {
+      "epoch": 0.2867789473684211,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00041401404982611566,
+      "loss": 3.386,
+      "step": 6811
+    },
+    {
+      "epoch": 0.2868210526315789,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00041398858111621906,
+      "loss": 3.2655,
+      "step": 6812
+    },
+    {
+      "epoch": 0.2868631578947368,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00041396310941864314,
+      "loss": 3.0905,
+      "step": 6813
+    },
+    {
+      "epoch": 0.2869052631578947,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00041393763473385203,
+      "loss": 2.9902,
+      "step": 6814
+    },
+    {
+      "epoch": 0.2869473684210526,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004139121570623098,
+      "loss": 3.0494,
+      "step": 6815
+    },
+    {
+      "epoch": 0.2869894736842105,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00041388667640448063,
+      "loss": 3.2856,
+      "step": 6816
+    },
+    {
+      "epoch": 0.2870315789473684,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004138611927608288,
+      "loss": 3.9361,
+      "step": 6817
+    },
+    {
+      "epoch": 0.2870736842105263,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004138357061318186,
+      "loss": 3.1721,
+      "step": 6818
+    },
+    {
+      "epoch": 0.2871157894736842,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004138102165179143,
+      "loss": 3.4302,
+      "step": 6819
+    },
+    {
+      "epoch": 0.2871578947368421,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0004137847239195803,
+      "loss": 3.3751,
+      "step": 6820
+    },
+    {
+      "epoch": 0.2872,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004137592283372812,
+      "loss": 3.5782,
+      "step": 6821
+    },
+    {
+      "epoch": 0.2872421052631579,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00041373372977148126,
+      "loss": 3.0822,
+      "step": 6822
+    },
+    {
+      "epoch": 0.2872842105263158,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0004137082282226451,
+      "loss": 3.1254,
+      "step": 6823
+    },
+    {
+      "epoch": 0.2873263157894737,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0004136827236912374,
+      "loss": 2.7952,
+      "step": 6824
+    },
+    {
+      "epoch": 0.2873684210526316,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004136572161777229,
+      "loss": 2.9388,
+      "step": 6825
+    },
+    {
+      "epoch": 0.28741052631578945,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00041363170568256606,
+      "loss": 2.7385,
+      "step": 6826
+    },
+    {
+      "epoch": 0.28745263157894735,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004136061922062319,
+      "loss": 3.0984,
+      "step": 6827
+    },
+    {
+      "epoch": 0.28749473684210525,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004135806757491851,
+      "loss": 3.4219,
+      "step": 6828
+    },
+    {
+      "epoch": 0.28753684210526315,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00041355515631189064,
+      "loss": 3.4787,
+      "step": 6829
+    },
+    {
+      "epoch": 0.28757894736842105,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004135296338948134,
+      "loss": 3.7318,
+      "step": 6830
+    },
+    {
+      "epoch": 0.28762105263157894,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004135041084984183,
+      "loss": 3.0566,
+      "step": 6831
+    },
+    {
+      "epoch": 0.28766315789473684,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00041347858012317053,
+      "loss": 3.1387,
+      "step": 6832
+    },
+    {
+      "epoch": 0.28770526315789474,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00041345304876953506,
+      "loss": 3.4598,
+      "step": 6833
+    },
+    {
+      "epoch": 0.28774736842105264,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004134275144379771,
+      "loss": 3.725,
+      "step": 6834
+    },
+    {
+      "epoch": 0.28778947368421054,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00041340197712896187,
+      "loss": 3.1959,
+      "step": 6835
+    },
+    {
+      "epoch": 0.28783157894736844,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00041337643684295457,
+      "loss": 3.2991,
+      "step": 6836
+    },
+    {
+      "epoch": 0.28787368421052634,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00041335089358042053,
+      "loss": 2.7578,
+      "step": 6837
+    },
+    {
+      "epoch": 0.28791578947368424,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00041332534734182515,
+      "loss": 3.3701,
+      "step": 6838
+    },
+    {
+      "epoch": 0.2879578947368421,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004132997981276339,
+      "loss": 3.5056,
+      "step": 6839
+    },
+    {
+      "epoch": 0.288,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00041327424593831213,
+      "loss": 3.2825,
+      "step": 6840
+    },
+    {
+      "epoch": 0.2880421052631579,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004132486907743255,
+      "loss": 2.7677,
+      "step": 6841
+    },
+    {
+      "epoch": 0.2880842105263158,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00041322313263613954,
+      "loss": 3.0814,
+      "step": 6842
+    },
+    {
+      "epoch": 0.2881263157894737,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00041319757152421984,
+      "loss": 3.4905,
+      "step": 6843
+    },
+    {
+      "epoch": 0.2881684210526316,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004131720074390321,
+      "loss": 2.9604,
+      "step": 6844
+    },
+    {
+      "epoch": 0.2882105263157895,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00041314644038104216,
+      "loss": 3.2489,
+      "step": 6845
+    },
+    {
+      "epoch": 0.28825263157894737,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00041312087035071576,
+      "loss": 3.0028,
+      "step": 6846
+    },
+    {
+      "epoch": 0.28829473684210527,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00041309529734851873,
+      "loss": 3.5113,
+      "step": 6847
+    },
+    {
+      "epoch": 0.28833684210526317,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00041306972137491706,
+      "loss": 3.3556,
+      "step": 6848
+    },
+    {
+      "epoch": 0.28837894736842107,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00041304414243037663,
+      "loss": 3.4655,
+      "step": 6849
+    },
+    {
+      "epoch": 0.28842105263157897,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00041301856051536353,
+      "loss": 3.2174,
+      "step": 6850
+    },
+    {
+      "epoch": 0.28846315789473687,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0004129929756303437,
+      "loss": 3.2303,
+      "step": 6851
+    },
+    {
+      "epoch": 0.2885052631578947,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00041296738777578347,
+      "loss": 3.1464,
+      "step": 6852
+    },
+    {
+      "epoch": 0.2885473684210526,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00041294179695214886,
+      "loss": 3.3757,
+      "step": 6853
+    },
+    {
+      "epoch": 0.2885894736842105,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004129162031599062,
+      "loss": 3.5455,
+      "step": 6854
+    },
+    {
+      "epoch": 0.2886315789473684,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004128906063995217,
+      "loss": 3.6344,
+      "step": 6855
+    },
+    {
+      "epoch": 0.2886736842105263,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0004128650066714618,
+      "loss": 3.5887,
+      "step": 6856
+    },
+    {
+      "epoch": 0.2887157894736842,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004128394039761928,
+      "loss": 3.4138,
+      "step": 6857
+    },
+    {
+      "epoch": 0.2887578947368421,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004128137983141811,
+      "loss": 3.4601,
+      "step": 6858
+    },
+    {
+      "epoch": 0.2888,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004127881896858934,
+      "loss": 2.9784,
+      "step": 6859
+    },
+    {
+      "epoch": 0.2888421052631579,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004127625780917961,
+      "loss": 3.5096,
+      "step": 6860
+    },
+    {
+      "epoch": 0.2888842105263158,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0004127369635323559,
+      "loss": 3.2296,
+      "step": 6861
+    },
+    {
+      "epoch": 0.2889263157894737,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004127113460080394,
+      "loss": 3.0172,
+      "step": 6862
+    },
+    {
+      "epoch": 0.2889684210526316,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004126857255193134,
+      "loss": 3.122,
+      "step": 6863
+    },
+    {
+      "epoch": 0.2890105263157895,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004126601020666446,
+      "loss": 3.637,
+      "step": 6864
+    },
+    {
+      "epoch": 0.2890526315789474,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00041263447565049983,
+      "loss": 2.765,
+      "step": 6865
+    },
+    {
+      "epoch": 0.28909473684210524,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00041260884627134604,
+      "loss": 3.2371,
+      "step": 6866
+    },
+    {
+      "epoch": 0.28913684210526314,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00041258321392965015,
+      "loss": 3.334,
+      "step": 6867
+    },
+    {
+      "epoch": 0.28917894736842104,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00041255757862587906,
+      "loss": 3.5513,
+      "step": 6868
+    },
+    {
+      "epoch": 0.28922105263157893,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00041253194036049997,
+      "loss": 3.3936,
+      "step": 6869
+    },
+    {
+      "epoch": 0.28926315789473683,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00041250629913397983,
+      "loss": 3.5201,
+      "step": 6870
+    },
+    {
+      "epoch": 0.28930526315789473,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004124806549467859,
+      "loss": 3.4831,
+      "step": 6871
+    },
+    {
+      "epoch": 0.28934736842105263,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00041245500779938526,
+      "loss": 3.0633,
+      "step": 6872
+    },
+    {
+      "epoch": 0.28938947368421053,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004124293576922453,
+      "loss": 3.3907,
+      "step": 6873
+    },
+    {
+      "epoch": 0.28943157894736843,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0004124037046258333,
+      "loss": 3.2553,
+      "step": 6874
+    },
+    {
+      "epoch": 0.2894736842105263,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00041237804860061667,
+      "loss": 3.5275,
+      "step": 6875
+    },
+    {
+      "epoch": 0.2895157894736842,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00041235238961706273,
+      "loss": 3.2615,
+      "step": 6876
+    },
+    {
+      "epoch": 0.2895578947368421,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.000412326727675639,
+      "loss": 3.2144,
+      "step": 6877
+    },
+    {
+      "epoch": 0.2896,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.000412301062776813,
+      "loss": 3.5277,
+      "step": 6878
+    },
+    {
+      "epoch": 0.28964210526315787,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004122753949210524,
+      "loss": 3.5069,
+      "step": 6879
+    },
+    {
+      "epoch": 0.28968421052631577,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004122497241088247,
+      "loss": 3.2184,
+      "step": 6880
+    },
+    {
+      "epoch": 0.28972631578947367,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00041222405034059773,
+      "loss": 3.0129,
+      "step": 6881
+    },
+    {
+      "epoch": 0.28976842105263156,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004121983736168391,
+      "loss": 3.505,
+      "step": 6882
+    },
+    {
+      "epoch": 0.28981052631578946,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00041217269393801673,
+      "loss": 3.3256,
+      "step": 6883
+    },
+    {
+      "epoch": 0.28985263157894736,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004121470113045984,
+      "loss": 3.0588,
+      "step": 6884
+    },
+    {
+      "epoch": 0.28989473684210526,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.000412121325717052,
+      "loss": 3.4451,
+      "step": 6885
+    },
+    {
+      "epoch": 0.28993684210526316,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004120956371758456,
+      "loss": 3.3772,
+      "step": 6886
+    },
+    {
+      "epoch": 0.28997894736842106,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0004120699456814471,
+      "loss": 2.8542,
+      "step": 6887
+    },
+    {
+      "epoch": 0.29002105263157896,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00041204425123432466,
+      "loss": 3.1662,
+      "step": 6888
+    },
+    {
+      "epoch": 0.29006315789473686,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004120185538349463,
+      "loss": 2.5155,
+      "step": 6889
+    },
+    {
+      "epoch": 0.29010526315789475,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004119928534837803,
+      "loss": 3.1446,
+      "step": 6890
+    },
+    {
+      "epoch": 0.29014736842105265,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00041196715018129476,
+      "loss": 2.975,
+      "step": 6891
+    },
+    {
+      "epoch": 0.29018947368421055,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004119414439279581,
+      "loss": 2.6824,
+      "step": 6892
+    },
+    {
+      "epoch": 0.2902315789473684,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00041191573472423865,
+      "loss": 3.1621,
+      "step": 6893
+    },
+    {
+      "epoch": 0.2902736842105263,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004118900225706047,
+      "loss": 3.1146,
+      "step": 6894
+    },
+    {
+      "epoch": 0.2903157894736842,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004118643074675248,
+      "loss": 3.0639,
+      "step": 6895
+    },
+    {
+      "epoch": 0.2903578947368421,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00041183858941546737,
+      "loss": 3.3838,
+      "step": 6896
+    },
+    {
+      "epoch": 0.2904,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00041181286841490103,
+      "loss": 3.3239,
+      "step": 6897
+    },
+    {
+      "epoch": 0.2904421052631579,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004117871444662943,
+      "loss": 3.047,
+      "step": 6898
+    },
+    {
+      "epoch": 0.2904842105263158,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00041176141757011587,
+      "loss": 3.2523,
+      "step": 6899
+    },
+    {
+      "epoch": 0.2905263157894737,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0004117356877268345,
+      "loss": 3.1833,
+      "step": 6900
+    },
+    {
+      "epoch": 0.2905684210526316,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.000411709954936919,
+      "loss": 3.4644,
+      "step": 6901
+    },
+    {
+      "epoch": 0.2906105263157895,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00041168421920083803,
+      "loss": 3.1531,
+      "step": 6902
+    },
+    {
+      "epoch": 0.2906526315789474,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00041165848051906065,
+      "loss": 3.4381,
+      "step": 6903
+    },
+    {
+      "epoch": 0.2906947368421053,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00041163273889205555,
+      "loss": 3.4376,
+      "step": 6904
+    },
+    {
+      "epoch": 0.2907368421052632,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004116069943202919,
+      "loss": 3.5508,
+      "step": 6905
+    },
+    {
+      "epoch": 0.290778947368421,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00041158124680423883,
+      "loss": 3.1528,
+      "step": 6906
+    },
+    {
+      "epoch": 0.2908210526315789,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00041155549634436516,
+      "loss": 3.0612,
+      "step": 6907
+    },
+    {
+      "epoch": 0.2908631578947368,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004115297429411402,
+      "loss": 3.3424,
+      "step": 6908
+    },
+    {
+      "epoch": 0.2909052631578947,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004115039865950332,
+      "loss": 3.5114,
+      "step": 6909
+    },
+    {
+      "epoch": 0.2909473684210526,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004114782273065132,
+      "loss": 2.8298,
+      "step": 6910
+    },
+    {
+      "epoch": 0.2909894736842105,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00041145246507604964,
+      "loss": 2.9669,
+      "step": 6911
+    },
+    {
+      "epoch": 0.2910315789473684,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004114266999041119,
+      "loss": 3.4259,
+      "step": 6912
+    },
+    {
+      "epoch": 0.2910736842105263,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004114009317911693,
+      "loss": 3.1891,
+      "step": 6913
+    },
+    {
+      "epoch": 0.2911157894736842,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004113751607376915,
+      "loss": 3.2577,
+      "step": 6914
+    },
+    {
+      "epoch": 0.2911578947368421,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004113493867441478,
+      "loss": 2.7257,
+      "step": 6915
+    },
+    {
+      "epoch": 0.2912,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004113236098110078,
+      "loss": 3.3934,
+      "step": 6916
+    },
+    {
+      "epoch": 0.2912421052631579,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00041129782993874124,
+      "loss": 3.0108,
+      "step": 6917
+    },
+    {
+      "epoch": 0.2912842105263158,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00041127204712781767,
+      "loss": 2.7261,
+      "step": 6918
+    },
+    {
+      "epoch": 0.2913263157894737,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00041124626137870697,
+      "loss": 3.3889,
+      "step": 6919
+    },
+    {
+      "epoch": 0.29136842105263155,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004112204726918788,
+      "loss": 3.0211,
+      "step": 6920
+    },
+    {
+      "epoch": 0.29141052631578945,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00041119468106780305,
+      "loss": 3.3265,
+      "step": 6921
+    },
+    {
+      "epoch": 0.29145263157894735,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004111688865069496,
+      "loss": 3.4631,
+      "step": 6922
+    },
+    {
+      "epoch": 0.29149473684210525,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00041114308900978846,
+      "loss": 3.4626,
+      "step": 6923
+    },
+    {
+      "epoch": 0.29153684210526315,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00041111728857678954,
+      "loss": 3.1028,
+      "step": 6924
+    },
+    {
+      "epoch": 0.29157894736842105,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004110914852084229,
+      "loss": 3.1169,
+      "step": 6925
+    },
+    {
+      "epoch": 0.29162105263157895,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00041106567890515866,
+      "loss": 2.7875,
+      "step": 6926
+    },
+    {
+      "epoch": 0.29166315789473685,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00041103986966746703,
+      "loss": 3.6257,
+      "step": 6927
+    },
+    {
+      "epoch": 0.29170526315789475,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00041101405749581824,
+      "loss": 3.2642,
+      "step": 6928
+    },
+    {
+      "epoch": 0.29174736842105264,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004109882423906824,
+      "loss": 3.2128,
+      "step": 6929
+    },
+    {
+      "epoch": 0.29178947368421054,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00041096242435253,
+      "loss": 3.1603,
+      "step": 6930
+    },
+    {
+      "epoch": 0.29183157894736844,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00041093660338183135,
+      "loss": 3.3577,
+      "step": 6931
+    },
+    {
+      "epoch": 0.29187368421052634,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004109107794790569,
+      "loss": 3.1983,
+      "step": 6932
+    },
+    {
+      "epoch": 0.2919157894736842,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00041088495264467693,
+      "loss": 3.1853,
+      "step": 6933
+    },
+    {
+      "epoch": 0.2919578947368421,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004108591228791624,
+      "loss": 3.4953,
+      "step": 6934
+    },
+    {
+      "epoch": 0.292,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00041083329018298353,
+      "loss": 3.0693,
+      "step": 6935
+    },
+    {
+      "epoch": 0.2920421052631579,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004108074545566111,
+      "loss": 3.6212,
+      "step": 6936
+    },
+    {
+      "epoch": 0.2920842105263158,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00041078161600051577,
+      "loss": 3.2413,
+      "step": 6937
+    },
+    {
+      "epoch": 0.2921263157894737,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00041075577451516837,
+      "loss": 2.9161,
+      "step": 6938
+    },
+    {
+      "epoch": 0.2921684210526316,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0004107299301010396,
+      "loss": 3.2237,
+      "step": 6939
+    },
+    {
+      "epoch": 0.2922105263157895,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004107040827586004,
+      "loss": 3.4368,
+      "step": 6940
+    },
+    {
+      "epoch": 0.2922526315789474,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00041067823248832155,
+      "loss": 2.7992,
+      "step": 6941
+    },
+    {
+      "epoch": 0.2922947368421053,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00041065237929067413,
+      "loss": 3.1245,
+      "step": 6942
+    },
+    {
+      "epoch": 0.2923368421052632,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00041062652316612916,
+      "loss": 2.8013,
+      "step": 6943
+    },
+    {
+      "epoch": 0.29237894736842107,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004106006641151576,
+      "loss": 3.4383,
+      "step": 6944
+    },
+    {
+      "epoch": 0.29242105263157897,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004105748021382306,
+      "loss": 3.3552,
+      "step": 6945
+    },
+    {
+      "epoch": 0.29246315789473687,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00041054893723581944,
+      "loss": 3.3643,
+      "step": 6946
+    },
+    {
+      "epoch": 0.2925052631578947,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00041052306940839536,
+      "loss": 3.4502,
+      "step": 6947
+    },
+    {
+      "epoch": 0.2925473684210526,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0004104971986564294,
+      "loss": 3.549,
+      "step": 6948
+    },
+    {
+      "epoch": 0.2925894736842105,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004104713249803932,
+      "loss": 3.3577,
+      "step": 6949
+    },
+    {
+      "epoch": 0.2926315789473684,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00041044544838075793,
+      "loss": 3.1391,
+      "step": 6950
+    },
+    {
+      "epoch": 0.2926736842105263,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00041041956885799513,
+      "loss": 3.2707,
+      "step": 6951
+    },
+    {
+      "epoch": 0.2927157894736842,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00041039368641257625,
+      "loss": 3.512,
+      "step": 6952
+    },
+    {
+      "epoch": 0.2927578947368421,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00041036780104497284,
+      "loss": 3.5691,
+      "step": 6953
+    },
+    {
+      "epoch": 0.2928,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00041034191275565653,
+      "loss": 2.9653,
+      "step": 6954
+    },
+    {
+      "epoch": 0.2928421052631579,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.000410316021545099,
+      "loss": 3.4059,
+      "step": 6955
+    },
+    {
+      "epoch": 0.2928842105263158,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0004102901274137719,
+      "loss": 3.0426,
+      "step": 6956
+    },
+    {
+      "epoch": 0.2929263157894737,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.000410264230362147,
+      "loss": 3.3646,
+      "step": 6957
+    },
+    {
+      "epoch": 0.2929684210526316,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00041023833039069614,
+      "loss": 3.5184,
+      "step": 6958
+    },
+    {
+      "epoch": 0.2930105263157895,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0004102124274998912,
+      "loss": 3.3137,
+      "step": 6959
+    },
+    {
+      "epoch": 0.29305263157894734,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.000410186521690204,
+      "loss": 3.3484,
+      "step": 6960
+    },
+    {
+      "epoch": 0.29309473684210524,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004101606129621066,
+      "loss": 3.2754,
+      "step": 6961
+    },
+    {
+      "epoch": 0.29313684210526314,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004101347013160711,
+      "loss": 3.5101,
+      "step": 6962
+    },
+    {
+      "epoch": 0.29317894736842104,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0004101087867525694,
+      "loss": 3.5125,
+      "step": 6963
+    },
+    {
+      "epoch": 0.29322105263157894,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004100828692720738,
+      "loss": 3.3988,
+      "step": 6964
+    },
+    {
+      "epoch": 0.29326315789473684,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004100569488750563,
+      "loss": 3.069,
+      "step": 6965
+    },
+    {
+      "epoch": 0.29330526315789474,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00041003102556198937,
+      "loss": 2.9953,
+      "step": 6966
+    },
+    {
+      "epoch": 0.29334736842105263,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004100050993333451,
+      "loss": 3.5592,
+      "step": 6967
+    },
+    {
+      "epoch": 0.29338947368421053,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00040997917018959596,
+      "loss": 3.2103,
+      "step": 6968
+    },
+    {
+      "epoch": 0.29343157894736843,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0004099532381312143,
+      "loss": 3.0738,
+      "step": 6969
+    },
+    {
+      "epoch": 0.29347368421052633,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004099273031586726,
+      "loss": 3.1552,
+      "step": 6970
+    },
+    {
+      "epoch": 0.29351578947368423,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004099013652724434,
+      "loss": 3.158,
+      "step": 6971
+    },
+    {
+      "epoch": 0.29355789473684213,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00040987542447299904,
+      "loss": 3.0497,
+      "step": 6972
+    },
+    {
+      "epoch": 0.2936,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00040984948076081234,
+      "loss": 3.5721,
+      "step": 6973
+    },
+    {
+      "epoch": 0.29364210526315787,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.000409823534136356,
+      "loss": 3.2918,
+      "step": 6974
+    },
+    {
+      "epoch": 0.29368421052631577,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004097975846001026,
+      "loss": 3.4771,
+      "step": 6975
+    },
+    {
+      "epoch": 0.29372631578947367,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.000409771632152525,
+      "loss": 2.9753,
+      "step": 6976
+    },
+    {
+      "epoch": 0.29376842105263157,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004097456767940959,
+      "loss": 3.3356,
+      "step": 6977
+    },
+    {
+      "epoch": 0.29381052631578947,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004097197185252883,
+      "loss": 2.9741,
+      "step": 6978
+    },
+    {
+      "epoch": 0.29385263157894737,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0004096937573465751,
+      "loss": 3.524,
+      "step": 6979
+    },
+    {
+      "epoch": 0.29389473684210526,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0004096677932584293,
+      "loss": 2.8723,
+      "step": 6980
+    },
+    {
+      "epoch": 0.29393684210526316,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004096418262613239,
+      "loss": 3.4511,
+      "step": 6981
+    },
+    {
+      "epoch": 0.29397894736842106,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00040961585635573193,
+      "loss": 3.2945,
+      "step": 6982
+    },
+    {
+      "epoch": 0.29402105263157896,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004095898835421267,
+      "loss": 3.1488,
+      "step": 6983
+    },
+    {
+      "epoch": 0.29406315789473686,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0004095639078209813,
+      "loss": 2.9756,
+      "step": 6984
+    },
+    {
+      "epoch": 0.29410526315789476,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00040953792919276886,
+      "loss": 3.2762,
+      "step": 6985
+    },
+    {
+      "epoch": 0.29414736842105266,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004095119476579629,
+      "loss": 3.2267,
+      "step": 6986
+    },
+    {
+      "epoch": 0.2941894736842105,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004094859632170367,
+      "loss": 3.1944,
+      "step": 6987
+    },
+    {
+      "epoch": 0.2942315789473684,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0004094599758704636,
+      "loss": 3.3935,
+      "step": 6988
+    },
+    {
+      "epoch": 0.2942736842105263,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.000409433985618717,
+      "loss": 3.4182,
+      "step": 6989
+    },
+    {
+      "epoch": 0.2943157894736842,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00040940799246227066,
+      "loss": 3.3397,
+      "step": 6990
+    },
+    {
+      "epoch": 0.2943578947368421,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004093819964015979,
+      "loss": 3.0362,
+      "step": 6991
+    },
+    {
+      "epoch": 0.2944,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00040935599743717243,
+      "loss": 3.0901,
+      "step": 6992
+    },
+    {
+      "epoch": 0.2944421052631579,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00040932999556946795,
+      "loss": 3.0273,
+      "step": 6993
+    },
+    {
+      "epoch": 0.2944842105263158,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004093039907989582,
+      "loss": 3.5781,
+      "step": 6994
+    },
+    {
+      "epoch": 0.2945263157894737,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00040927798312611686,
+      "loss": 3.161,
+      "step": 6995
+    },
+    {
+      "epoch": 0.2945684210526316,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004092519725514178,
+      "loss": 3.2262,
+      "step": 6996
+    },
+    {
+      "epoch": 0.2946105263157895,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0004092259590753349,
+      "loss": 2.9021,
+      "step": 6997
+    },
+    {
+      "epoch": 0.2946526315789474,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00040919994269834214,
+      "loss": 3.24,
+      "step": 6998
+    },
+    {
+      "epoch": 0.2946947368421053,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00040917392342091355,
+      "loss": 2.8868,
+      "step": 6999
+    },
+    {
+      "epoch": 0.29473684210526313,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00040914790124352296,
+      "loss": 3.4138,
+      "step": 7000
+    },
+    {
+      "epoch": 0.29477894736842103,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00040912187616664464,
+      "loss": 3.1524,
+      "step": 7001
+    },
+    {
+      "epoch": 0.29482105263157893,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004090958481907527,
+      "loss": 3.3403,
+      "step": 7002
+    },
+    {
+      "epoch": 0.2948631578947368,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004090698173163214,
+      "loss": 3.0669,
+      "step": 7003
+    },
+    {
+      "epoch": 0.2949052631578947,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00040904378354382487,
+      "loss": 3.5379,
+      "step": 7004
+    },
+    {
+      "epoch": 0.2949473684210526,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00040901774687373757,
+      "loss": 3.3879,
+      "step": 7005
+    },
+    {
+      "epoch": 0.2949894736842105,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00040899170730653364,
+      "loss": 2.6395,
+      "step": 7006
+    },
+    {
+      "epoch": 0.2950315789473684,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0004089656648426876,
+      "loss": 3.1169,
+      "step": 7007
+    },
+    {
+      "epoch": 0.2950736842105263,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.000408939619482674,
+      "loss": 3.0944,
+      "step": 7008
+    },
+    {
+      "epoch": 0.2951157894736842,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00040891357122696726,
+      "loss": 3.2274,
+      "step": 7009
+    },
+    {
+      "epoch": 0.2951578947368421,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.000408887520076042,
+      "loss": 3.4985,
+      "step": 7010
+    },
+    {
+      "epoch": 0.2952,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00040886146603037277,
+      "loss": 3.0975,
+      "step": 7011
+    },
+    {
+      "epoch": 0.2952421052631579,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00040883540909043424,
+      "loss": 3.0399,
+      "step": 7012
+    },
+    {
+      "epoch": 0.2952842105263158,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004088093492567012,
+      "loss": 3.4971,
+      "step": 7013
+    },
+    {
+      "epoch": 0.29532631578947366,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00040878328652964844,
+      "loss": 2.8487,
+      "step": 7014
+    },
+    {
+      "epoch": 0.29536842105263156,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00040875722090975077,
+      "loss": 3.0139,
+      "step": 7015
+    },
+    {
+      "epoch": 0.29541052631578946,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00040873115239748305,
+      "loss": 3.2148,
+      "step": 7016
+    },
+    {
+      "epoch": 0.29545263157894736,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00040870508099332027,
+      "loss": 3.2863,
+      "step": 7017
+    },
+    {
+      "epoch": 0.29549473684210525,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004086790066977373,
+      "loss": 3.0548,
+      "step": 7018
+    },
+    {
+      "epoch": 0.29553684210526315,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00040865292951120934,
+      "loss": 3.419,
+      "step": 7019
+    },
+    {
+      "epoch": 0.29557894736842105,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004086268494342114,
+      "loss": 3.4219,
+      "step": 7020
+    },
+    {
+      "epoch": 0.29562105263157895,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004086007664672186,
+      "loss": 3.6585,
+      "step": 7021
+    },
+    {
+      "epoch": 0.29566315789473685,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00040857468061070623,
+      "loss": 3.6051,
+      "step": 7022
+    },
+    {
+      "epoch": 0.29570526315789475,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004085485918651495,
+      "loss": 3.0094,
+      "step": 7023
+    },
+    {
+      "epoch": 0.29574736842105265,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00040852250023102373,
+      "loss": 2.9432,
+      "step": 7024
+    },
+    {
+      "epoch": 0.29578947368421055,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00040849640570880417,
+      "loss": 3.7083,
+      "step": 7025
+    },
+    {
+      "epoch": 0.29583157894736845,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00040847030829896635,
+      "loss": 3.3362,
+      "step": 7026
+    },
+    {
+      "epoch": 0.2958736842105263,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004084442080019858,
+      "loss": 3.6235,
+      "step": 7027
+    },
+    {
+      "epoch": 0.2959157894736842,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004084181048183379,
+      "loss": 3.5923,
+      "step": 7028
+    },
+    {
+      "epoch": 0.2959578947368421,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00040839199874849827,
+      "loss": 3.1826,
+      "step": 7029
+    },
+    {
+      "epoch": 0.296,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00040836588979294256,
+      "loss": 3.3164,
+      "step": 7030
+    },
+    {
+      "epoch": 0.2960421052631579,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004083397779521463,
+      "loss": 3.3661,
+      "step": 7031
+    },
+    {
+      "epoch": 0.2960842105263158,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00040831366322658535,
+      "loss": 3.2896,
+      "step": 7032
+    },
+    {
+      "epoch": 0.2961263157894737,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004082875456167355,
+      "loss": 2.8008,
+      "step": 7033
+    },
+    {
+      "epoch": 0.2961684210526316,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00040826142512307253,
+      "loss": 3.3985,
+      "step": 7034
+    },
+    {
+      "epoch": 0.2962105263157895,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00040823530174607237,
+      "loss": 3.4451,
+      "step": 7035
+    },
+    {
+      "epoch": 0.2962526315789474,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00040820917548621094,
+      "loss": 3.4909,
+      "step": 7036
+    },
+    {
+      "epoch": 0.2962947368421053,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0004081830463439642,
+      "loss": 3.4369,
+      "step": 7037
+    },
+    {
+      "epoch": 0.2963368421052632,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004081569143198083,
+      "loss": 2.883,
+      "step": 7038
+    },
+    {
+      "epoch": 0.2963789473684211,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004081307794142191,
+      "loss": 3.1317,
+      "step": 7039
+    },
+    {
+      "epoch": 0.296421052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.000408104641627673,
+      "loss": 3.0456,
+      "step": 7040
+    },
+    {
+      "epoch": 0.2964631578947368,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00040807850096064605,
+      "loss": 3.5733,
+      "step": 7041
+    },
+    {
+      "epoch": 0.2965052631578947,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004080523574136145,
+      "loss": 3.3384,
+      "step": 7042
+    },
+    {
+      "epoch": 0.2965473684210526,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004080262109870548,
+      "loss": 3.4938,
+      "step": 7043
+    },
+    {
+      "epoch": 0.2965894736842105,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00040800006168144315,
+      "loss": 3.2096,
+      "step": 7044
+    },
+    {
+      "epoch": 0.2966315789473684,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00040797390949725605,
+      "loss": 3.1908,
+      "step": 7045
+    },
+    {
+      "epoch": 0.2966736842105263,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004079477544349699,
+      "loss": 3.1697,
+      "step": 7046
+    },
+    {
+      "epoch": 0.2967157894736842,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00040792159649506124,
+      "loss": 3.4531,
+      "step": 7047
+    },
+    {
+      "epoch": 0.2967578947368421,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00040789543567800666,
+      "loss": 3.2872,
+      "step": 7048
+    },
+    {
+      "epoch": 0.2968,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004078692719842828,
+      "loss": 3.3932,
+      "step": 7049
+    },
+    {
+      "epoch": 0.2968421052631579,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004078431054143662,
+      "loss": 3.1714,
+      "step": 7050
+    },
+    {
+      "epoch": 0.2968842105263158,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00040781693596873375,
+      "loss": 3.0404,
+      "step": 7051
+    },
+    {
+      "epoch": 0.2969263157894737,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004077907636478622,
+      "loss": 3.2849,
+      "step": 7052
+    },
+    {
+      "epoch": 0.2969684210526316,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004077645884522282,
+      "loss": 3.2119,
+      "step": 7053
+    },
+    {
+      "epoch": 0.29701052631578945,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00040773841038230886,
+      "loss": 2.911,
+      "step": 7054
+    },
+    {
+      "epoch": 0.29705263157894735,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.000407712229438581,
+      "loss": 3.5003,
+      "step": 7055
+    },
+    {
+      "epoch": 0.29709473684210524,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004076860456215216,
+      "loss": 2.9564,
+      "step": 7056
+    },
+    {
+      "epoch": 0.29713684210526314,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00040765985893160775,
+      "loss": 3.0703,
+      "step": 7057
+    },
+    {
+      "epoch": 0.29717894736842104,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00040763366936931645,
+      "loss": 2.8627,
+      "step": 7058
+    },
+    {
+      "epoch": 0.29722105263157894,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.000407607476935125,
+      "loss": 3.1157,
+      "step": 7059
+    },
+    {
+      "epoch": 0.29726315789473684,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00040758128162951047,
+      "loss": 3.4056,
+      "step": 7060
+    },
+    {
+      "epoch": 0.29730526315789474,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004075550834529501,
+      "loss": 2.9523,
+      "step": 7061
+    },
+    {
+      "epoch": 0.29734736842105264,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004075288824059212,
+      "loss": 3.1129,
+      "step": 7062
+    },
+    {
+      "epoch": 0.29738947368421054,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00040750267848890125,
+      "loss": 3.1112,
+      "step": 7063
+    },
+    {
+      "epoch": 0.29743157894736844,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0004074764717023676,
+      "loss": 3.4455,
+      "step": 7064
+    },
+    {
+      "epoch": 0.29747368421052633,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004074502620467975,
+      "loss": 2.9379,
+      "step": 7065
+    },
+    {
+      "epoch": 0.29751578947368423,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00040742404952266866,
+      "loss": 3.4671,
+      "step": 7066
+    },
+    {
+      "epoch": 0.29755789473684213,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004073978341304587,
+      "loss": 3.0391,
+      "step": 7067
+    },
+    {
+      "epoch": 0.2976,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004073716158706451,
+      "loss": 3.2764,
+      "step": 7068
+    },
+    {
+      "epoch": 0.2976421052631579,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004073453947437055,
+      "loss": 3.1283,
+      "step": 7069
+    },
+    {
+      "epoch": 0.2976842105263158,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004073191707501177,
+      "loss": 3.2943,
+      "step": 7070
+    },
+    {
+      "epoch": 0.2977263157894737,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00040729294389035955,
+      "loss": 3.3016,
+      "step": 7071
+    },
+    {
+      "epoch": 0.29776842105263157,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004072667141649087,
+      "loss": 3.4313,
+      "step": 7072
+    },
+    {
+      "epoch": 0.29781052631578947,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00040724048157424306,
+      "loss": 3.7441,
+      "step": 7073
+    },
+    {
+      "epoch": 0.29785263157894737,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004072142461188406,
+      "loss": 2.9675,
+      "step": 7074
+    },
+    {
+      "epoch": 0.29789473684210527,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00040718800779917933,
+      "loss": 3.2042,
+      "step": 7075
+    },
+    {
+      "epoch": 0.29793684210526317,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0004071617666157372,
+      "loss": 3.3818,
+      "step": 7076
+    },
+    {
+      "epoch": 0.29797894736842107,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00040713552256899246,
+      "loss": 3.4673,
+      "step": 7077
+    },
+    {
+      "epoch": 0.29802105263157896,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00040710927565942303,
+      "loss": 3.3942,
+      "step": 7078
+    },
+    {
+      "epoch": 0.29806315789473686,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00040708302588750723,
+      "loss": 3.5458,
+      "step": 7079
+    },
+    {
+      "epoch": 0.29810526315789476,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004070567732537232,
+      "loss": 2.9831,
+      "step": 7080
+    },
+    {
+      "epoch": 0.2981473684210526,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00040703051775854937,
+      "loss": 3.6183,
+      "step": 7081
+    },
+    {
+      "epoch": 0.2981894736842105,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.000407004259402464,
+      "loss": 3.3972,
+      "step": 7082
+    },
+    {
+      "epoch": 0.2982315789473684,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004069779981859455,
+      "loss": 3.2861,
+      "step": 7083
+    },
+    {
+      "epoch": 0.2982736842105263,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00040695173410947223,
+      "loss": 3.3951,
+      "step": 7084
+    },
+    {
+      "epoch": 0.2983157894736842,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00040692546717352283,
+      "loss": 3.6335,
+      "step": 7085
+    },
+    {
+      "epoch": 0.2983578947368421,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00040689919737857584,
+      "loss": 3.4825,
+      "step": 7086
+    },
+    {
+      "epoch": 0.2984,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004068729247251097,
+      "loss": 3.3585,
+      "step": 7087
+    },
+    {
+      "epoch": 0.2984421052631579,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00040684664921360333,
+      "loss": 3.1765,
+      "step": 7088
+    },
+    {
+      "epoch": 0.2984842105263158,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00040682037084453527,
+      "loss": 3.5421,
+      "step": 7089
+    },
+    {
+      "epoch": 0.2985263157894737,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00040679408961838426,
+      "loss": 3.2434,
+      "step": 7090
+    },
+    {
+      "epoch": 0.2985684210526316,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0004067678055356292,
+      "loss": 3.2472,
+      "step": 7091
+    },
+    {
+      "epoch": 0.2986105263157895,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004067415185967489,
+      "loss": 3.1952,
+      "step": 7092
+    },
+    {
+      "epoch": 0.2986526315789474,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00040671522880222227,
+      "loss": 3.2267,
+      "step": 7093
+    },
+    {
+      "epoch": 0.2986947368421053,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004066889361525283,
+      "loss": 3.7341,
+      "step": 7094
+    },
+    {
+      "epoch": 0.29873684210526313,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00040666264064814604,
+      "loss": 3.3532,
+      "step": 7095
+    },
+    {
+      "epoch": 0.29877894736842103,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004066363422895546,
+      "loss": 3.6474,
+      "step": 7096
+    },
+    {
+      "epoch": 0.29882105263157893,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.000406610041077233,
+      "loss": 3.3863,
+      "step": 7097
+    },
+    {
+      "epoch": 0.29886315789473683,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00040658373701166043,
+      "loss": 3.2587,
+      "step": 7098
+    },
+    {
+      "epoch": 0.29890526315789473,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00040655743009331626,
+      "loss": 3.4349,
+      "step": 7099
+    },
+    {
+      "epoch": 0.29894736842105263,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00040653112032267946,
+      "loss": 2.6972,
+      "step": 7100
+    },
+    {
+      "epoch": 0.2989894736842105,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004065048077002297,
+      "loss": 3.3591,
+      "step": 7101
+    },
+    {
+      "epoch": 0.2990315789473684,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00040647849222644623,
+      "loss": 3.4455,
+      "step": 7102
+    },
+    {
+      "epoch": 0.2990736842105263,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00040645217390180844,
+      "loss": 2.7638,
+      "step": 7103
+    },
+    {
+      "epoch": 0.2991157894736842,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0004064258527267959,
+      "loss": 3.0703,
+      "step": 7104
+    },
+    {
+      "epoch": 0.2991578947368421,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00040639952870188815,
+      "loss": 3.5365,
+      "step": 7105
+    },
+    {
+      "epoch": 0.2992,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00040637320182756466,
+      "loss": 3.0702,
+      "step": 7106
+    },
+    {
+      "epoch": 0.2992421052631579,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004063468721043052,
+      "loss": 3.2313,
+      "step": 7107
+    },
+    {
+      "epoch": 0.29928421052631576,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004063205395325894,
+      "loss": 3.6548,
+      "step": 7108
+    },
+    {
+      "epoch": 0.29932631578947366,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00040629420411289707,
+      "loss": 3.2564,
+      "step": 7109
+    },
+    {
+      "epoch": 0.29936842105263156,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004062678658457081,
+      "loss": 3.2673,
+      "step": 7110
+    },
+    {
+      "epoch": 0.29941052631578946,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00040624152473150207,
+      "loss": 3.1549,
+      "step": 7111
+    },
+    {
+      "epoch": 0.29945263157894736,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.000406215180770759,
+      "loss": 3.2754,
+      "step": 7112
+    },
+    {
+      "epoch": 0.29949473684210526,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.000406188833963959,
+      "loss": 3.4655,
+      "step": 7113
+    },
+    {
+      "epoch": 0.29953684210526316,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004061624843115819,
+      "loss": 3.1522,
+      "step": 7114
+    },
+    {
+      "epoch": 0.29957894736842106,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0004061361318141078,
+      "loss": 3.1621,
+      "step": 7115
+    },
+    {
+      "epoch": 0.29962105263157895,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00040610977647201693,
+      "loss": 3.0234,
+      "step": 7116
+    },
+    {
+      "epoch": 0.29966315789473685,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00040608341828578926,
+      "loss": 3.4838,
+      "step": 7117
+    },
+    {
+      "epoch": 0.29970526315789475,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00040605705725590513,
+      "loss": 3.2417,
+      "step": 7118
+    },
+    {
+      "epoch": 0.29974736842105265,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00040603069338284483,
+      "loss": 3.6645,
+      "step": 7119
+    },
+    {
+      "epoch": 0.29978947368421055,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00040600432666708846,
+      "loss": 3.2095,
+      "step": 7120
+    },
+    {
+      "epoch": 0.2998315789473684,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00040597795710911664,
+      "loss": 3.4361,
+      "step": 7121
+    },
+    {
+      "epoch": 0.2998736842105263,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004059515847094098,
+      "loss": 3.1715,
+      "step": 7122
+    },
+    {
+      "epoch": 0.2999157894736842,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00040592520946844824,
+      "loss": 3.204,
+      "step": 7123
+    },
+    {
+      "epoch": 0.2999578947368421,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004058988313867126,
+      "loss": 3.4369,
+      "step": 7124
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0004058724504646834,
+      "loss": 2.9328,
+      "step": 7125
+    },
+    {
+      "epoch": 0.3000421052631579,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0004058460667028413,
+      "loss": 3.117,
+      "step": 7126
+    },
+    {
+      "epoch": 0.3000842105263158,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004058196801016669,
+      "loss": 3.5972,
+      "step": 7127
+    },
+    {
+      "epoch": 0.3001263157894737,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004057932906616412,
+      "loss": 3.9995,
+      "step": 7128
+    },
+    {
+      "epoch": 0.3001684210526316,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00040576689838324465,
+      "loss": 3.2179,
+      "step": 7129
+    },
+    {
+      "epoch": 0.3002105263157895,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0004057405032669582,
+      "loss": 2.4477,
+      "step": 7130
+    },
+    {
+      "epoch": 0.3002526315789474,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00040571410531326294,
+      "loss": 3.2257,
+      "step": 7131
+    },
+    {
+      "epoch": 0.3002947368421053,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00040568770452263945,
+      "loss": 3.3107,
+      "step": 7132
+    },
+    {
+      "epoch": 0.3003368421052632,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.000405661300895569,
+      "loss": 3.4885,
+      "step": 7133
+    },
+    {
+      "epoch": 0.3003789473684211,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00040563489443253253,
+      "loss": 2.8229,
+      "step": 7134
+    },
+    {
+      "epoch": 0.3004210526315789,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00040560848513401115,
+      "loss": 3.3754,
+      "step": 7135
+    },
+    {
+      "epoch": 0.3004631578947368,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.000405582073000486,
+      "loss": 3.1276,
+      "step": 7136
+    },
+    {
+      "epoch": 0.3005052631578947,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004055556580324383,
+      "loss": 2.8902,
+      "step": 7137
+    },
+    {
+      "epoch": 0.3005473684210526,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004055292402303492,
+      "loss": 3.3738,
+      "step": 7138
+    },
+    {
+      "epoch": 0.3005894736842105,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004055028195947002,
+      "loss": 3.2514,
+      "step": 7139
+    },
+    {
+      "epoch": 0.3006315789473684,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004054763961259724,
+      "loss": 2.9669,
+      "step": 7140
+    },
+    {
+      "epoch": 0.3006736842105263,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00040544996982464746,
+      "loss": 2.9713,
+      "step": 7141
+    },
+    {
+      "epoch": 0.3007157894736842,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004054235406912067,
+      "loss": 3.2531,
+      "step": 7142
+    },
+    {
+      "epoch": 0.3007578947368421,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004053971087261316,
+      "loss": 3.147,
+      "step": 7143
+    },
+    {
+      "epoch": 0.3008,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004053706739299038,
+      "loss": 3.5368,
+      "step": 7144
+    },
+    {
+      "epoch": 0.3008421052631579,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00040534423630300486,
+      "loss": 3.1637,
+      "step": 7145
+    },
+    {
+      "epoch": 0.3008842105263158,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00040531779584591644,
+      "loss": 3.4304,
+      "step": 7146
+    },
+    {
+      "epoch": 0.3009263157894737,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004052913525591203,
+      "loss": 3.2401,
+      "step": 7147
+    },
+    {
+      "epoch": 0.30096842105263155,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004052649064430981,
+      "loss": 3.2303,
+      "step": 7148
+    },
+    {
+      "epoch": 0.30101052631578945,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00040523845749833185,
+      "loss": 3.2858,
+      "step": 7149
+    },
+    {
+      "epoch": 0.30105263157894735,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00040521200572530327,
+      "loss": 3.3665,
+      "step": 7150
+    },
+    {
+      "epoch": 0.30109473684210525,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00040518555112449426,
+      "loss": 3.6882,
+      "step": 7151
+    },
+    {
+      "epoch": 0.30113684210526315,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00040515909369638685,
+      "loss": 3.1602,
+      "step": 7152
+    },
+    {
+      "epoch": 0.30117894736842105,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004051326334414631,
+      "loss": 3.0372,
+      "step": 7153
+    },
+    {
+      "epoch": 0.30122105263157894,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00040510617036020497,
+      "loss": 3.3583,
+      "step": 7154
+    },
+    {
+      "epoch": 0.30126315789473684,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00040507970445309474,
+      "loss": 3.4743,
+      "step": 7155
+    },
+    {
+      "epoch": 0.30130526315789474,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00040505323572061447,
+      "loss": 2.8635,
+      "step": 7156
+    },
+    {
+      "epoch": 0.30134736842105264,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00040502676416324647,
+      "loss": 3.1104,
+      "step": 7157
+    },
+    {
+      "epoch": 0.30138947368421054,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004050002897814729,
+      "loss": 3.6362,
+      "step": 7158
+    },
+    {
+      "epoch": 0.30143157894736844,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004049738125757762,
+      "loss": 3.1378,
+      "step": 7159
+    },
+    {
+      "epoch": 0.30147368421052634,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00040494733254663873,
+      "loss": 3.0104,
+      "step": 7160
+    },
+    {
+      "epoch": 0.30151578947368424,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004049208496945429,
+      "loss": 3.0925,
+      "step": 7161
+    },
+    {
+      "epoch": 0.3015578947368421,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004048943640199712,
+      "loss": 3.2423,
+      "step": 7162
+    },
+    {
+      "epoch": 0.3016,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004048678755234062,
+      "loss": 3.5937,
+      "step": 7163
+    },
+    {
+      "epoch": 0.3016421052631579,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004048413842053305,
+      "loss": 3.3515,
+      "step": 7164
+    },
+    {
+      "epoch": 0.3016842105263158,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00040481489006622675,
+      "loss": 3.0265,
+      "step": 7165
+    },
+    {
+      "epoch": 0.3017263157894737,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004047883931065775,
+      "loss": 3.1981,
+      "step": 7166
+    },
+    {
+      "epoch": 0.3017684210526316,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004047618933268656,
+      "loss": 2.9761,
+      "step": 7167
+    },
+    {
+      "epoch": 0.3018105263157895,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00040473539072757386,
+      "loss": 3.0301,
+      "step": 7168
+    },
+    {
+      "epoch": 0.3018526315789474,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00040470888530918516,
+      "loss": 2.8747,
+      "step": 7169
+    },
+    {
+      "epoch": 0.30189473684210527,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004046823770721823,
+      "loss": 3.4779,
+      "step": 7170
+    },
+    {
+      "epoch": 0.30193684210526317,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004046558660170483,
+      "loss": 2.9559,
+      "step": 7171
+    },
+    {
+      "epoch": 0.30197894736842107,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004046293521442661,
+      "loss": 3.3375,
+      "step": 7172
+    },
+    {
+      "epoch": 0.30202105263157897,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00040460283545431884,
+      "loss": 3.4059,
+      "step": 7173
+    },
+    {
+      "epoch": 0.30206315789473687,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00040457631594768953,
+      "loss": 2.9437,
+      "step": 7174
+    },
+    {
+      "epoch": 0.3021052631578947,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00040454979362486134,
+      "loss": 3.3757,
+      "step": 7175
+    },
+    {
+      "epoch": 0.3021473684210526,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0004045232684863175,
+      "loss": 3.1857,
+      "step": 7176
+    },
+    {
+      "epoch": 0.3021894736842105,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00040449674053254133,
+      "loss": 3.1325,
+      "step": 7177
+    },
+    {
+      "epoch": 0.3022315789473684,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00040447020976401603,
+      "loss": 3.4661,
+      "step": 7178
+    },
+    {
+      "epoch": 0.3022736842105263,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00040444367618122504,
+      "loss": 3.0325,
+      "step": 7179
+    },
+    {
+      "epoch": 0.3023157894736842,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00040441713978465164,
+      "loss": 3.2169,
+      "step": 7180
+    },
+    {
+      "epoch": 0.3023578947368421,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00040439060057477944,
+      "loss": 3.0947,
+      "step": 7181
+    },
+    {
+      "epoch": 0.3024,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00040436405855209193,
+      "loss": 3.5381,
+      "step": 7182
+    },
+    {
+      "epoch": 0.3024421052631579,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0004043375137170726,
+      "loss": 2.8737,
+      "step": 7183
+    },
+    {
+      "epoch": 0.3024842105263158,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0004043109660702051,
+      "loss": 2.9908,
+      "step": 7184
+    },
+    {
+      "epoch": 0.3025263157894737,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00040428441561197315,
+      "loss": 3.5554,
+      "step": 7185
+    },
+    {
+      "epoch": 0.3025684210526316,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00040425786234286047,
+      "loss": 3.256,
+      "step": 7186
+    },
+    {
+      "epoch": 0.3026105263157895,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004042313062633507,
+      "loss": 3.2908,
+      "step": 7187
+    },
+    {
+      "epoch": 0.3026526315789474,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00040420474737392774,
+      "loss": 2.9989,
+      "step": 7188
+    },
+    {
+      "epoch": 0.30269473684210524,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00040417818567507546,
+      "loss": 3.3374,
+      "step": 7189
+    },
+    {
+      "epoch": 0.30273684210526314,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0004041516211672779,
+      "loss": 3.3622,
+      "step": 7190
+    },
+    {
+      "epoch": 0.30277894736842104,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00040412505385101883,
+      "loss": 3.219,
+      "step": 7191
+    },
+    {
+      "epoch": 0.30282105263157894,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00040409848372678235,
+      "loss": 2.3298,
+      "step": 7192
+    },
+    {
+      "epoch": 0.30286315789473683,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00040407191079505265,
+      "loss": 3.6835,
+      "step": 7193
+    },
+    {
+      "epoch": 0.30290526315789473,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00040404533505631365,
+      "loss": 3.3307,
+      "step": 7194
+    },
+    {
+      "epoch": 0.30294736842105263,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004040187565110497,
+      "loss": 3.0551,
+      "step": 7195
+    },
+    {
+      "epoch": 0.30298947368421053,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00040399217515974493,
+      "loss": 3.2392,
+      "step": 7196
+    },
+    {
+      "epoch": 0.30303157894736843,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00040396559100288373,
+      "loss": 3.4133,
+      "step": 7197
+    },
+    {
+      "epoch": 0.30307368421052633,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004039390040409503,
+      "loss": 3.6795,
+      "step": 7198
+    },
+    {
+      "epoch": 0.3031157894736842,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00040391241427442914,
+      "loss": 3.3116,
+      "step": 7199
+    },
+    {
+      "epoch": 0.3031578947368421,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004038858217038046,
+      "loss": 3.4817,
+      "step": 7200
+    },
+    {
+      "epoch": 0.3032,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00040385922632956117,
+      "loss": 3.3277,
+      "step": 7201
+    },
+    {
+      "epoch": 0.30324210526315787,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00040383262815218345,
+      "loss": 3.8246,
+      "step": 7202
+    },
+    {
+      "epoch": 0.30328421052631577,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00040380602717215595,
+      "loss": 3.267,
+      "step": 7203
+    },
+    {
+      "epoch": 0.30332631578947367,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00040377942338996344,
+      "loss": 2.916,
+      "step": 7204
+    },
+    {
+      "epoch": 0.30336842105263156,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00040375281680609044,
+      "loss": 2.9805,
+      "step": 7205
+    },
+    {
+      "epoch": 0.30341052631578946,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00040372620742102185,
+      "loss": 3.043,
+      "step": 7206
+    },
+    {
+      "epoch": 0.30345263157894736,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004036995952352424,
+      "loss": 3.1854,
+      "step": 7207
+    },
+    {
+      "epoch": 0.30349473684210526,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0004036729802492367,
+      "loss": 2.9862,
+      "step": 7208
+    },
+    {
+      "epoch": 0.30353684210526316,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00040364636246349007,
+      "loss": 3.6425,
+      "step": 7209
+    },
+    {
+      "epoch": 0.30357894736842106,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00040361974187848716,
+      "loss": 3.3296,
+      "step": 7210
+    },
+    {
+      "epoch": 0.30362105263157896,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.000403593118494713,
+      "loss": 3.2004,
+      "step": 7211
+    },
+    {
+      "epoch": 0.30366315789473686,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00040356649231265283,
+      "loss": 2.9588,
+      "step": 7212
+    },
+    {
+      "epoch": 0.30370526315789476,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00040353986333279145,
+      "loss": 3.1981,
+      "step": 7213
+    },
+    {
+      "epoch": 0.30374736842105265,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00040351323155561424,
+      "loss": 3.204,
+      "step": 7214
+    },
+    {
+      "epoch": 0.30378947368421055,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00040348659698160634,
+      "loss": 3.2757,
+      "step": 7215
+    },
+    {
+      "epoch": 0.3038315789473684,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.000403459959611253,
+      "loss": 3.4009,
+      "step": 7216
+    },
+    {
+      "epoch": 0.3038736842105263,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00040343331944503946,
+      "loss": 3.3942,
+      "step": 7217
+    },
+    {
+      "epoch": 0.3039157894736842,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004034066764834511,
+      "loss": 3.4926,
+      "step": 7218
+    },
+    {
+      "epoch": 0.3039578947368421,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0004033800307269734,
+      "loss": 3.1336,
+      "step": 7219
+    },
+    {
+      "epoch": 0.304,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0004033533821760917,
+      "loss": 3.2857,
+      "step": 7220
+    },
+    {
+      "epoch": 0.3040421052631579,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00040332673083129156,
+      "loss": 2.7415,
+      "step": 7221
+    },
+    {
+      "epoch": 0.3040842105263158,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004033000766930585,
+      "loss": 3.2251,
+      "step": 7222
+    },
+    {
+      "epoch": 0.3041263157894737,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00040327341976187825,
+      "loss": 3.1824,
+      "step": 7223
+    },
+    {
+      "epoch": 0.3041684210526316,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00040324676003823635,
+      "loss": 3.2979,
+      "step": 7224
+    },
+    {
+      "epoch": 0.3042105263157895,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004032200975226185,
+      "loss": 3.3774,
+      "step": 7225
+    },
+    {
+      "epoch": 0.3042526315789474,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00040319343221551055,
+      "loss": 3.4369,
+      "step": 7226
+    },
+    {
+      "epoch": 0.3042947368421053,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00040316676411739826,
+      "loss": 2.9629,
+      "step": 7227
+    },
+    {
+      "epoch": 0.3043368421052632,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00040314009322876744,
+      "loss": 3.4383,
+      "step": 7228
+    },
+    {
+      "epoch": 0.304378947368421,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00040311341955010405,
+      "loss": 3.3273,
+      "step": 7229
+    },
+    {
+      "epoch": 0.3044210526315789,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004030867430818941,
+      "loss": 3.1951,
+      "step": 7230
+    },
+    {
+      "epoch": 0.3044631578947368,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00040306006382462354,
+      "loss": 3.4609,
+      "step": 7231
+    },
+    {
+      "epoch": 0.3045052631578947,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00040303338177877853,
+      "loss": 3.4164,
+      "step": 7232
+    },
+    {
+      "epoch": 0.3045473684210526,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00040300669694484507,
+      "loss": 2.8759,
+      "step": 7233
+    },
+    {
+      "epoch": 0.3045894736842105,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004029800093233093,
+      "loss": 3.2779,
+      "step": 7234
+    },
+    {
+      "epoch": 0.3046315789473684,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00040295331891465754,
+      "loss": 3.1125,
+      "step": 7235
+    },
+    {
+      "epoch": 0.3046736842105263,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.000402926625719376,
+      "loss": 3.0441,
+      "step": 7236
+    },
+    {
+      "epoch": 0.3047157894736842,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004028999297379511,
+      "loss": 3.3236,
+      "step": 7237
+    },
+    {
+      "epoch": 0.3047578947368421,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004028732309708691,
+      "loss": 3.4255,
+      "step": 7238
+    },
+    {
+      "epoch": 0.3048,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004028465294186164,
+      "loss": 3.0669,
+      "step": 7239
+    },
+    {
+      "epoch": 0.3048421052631579,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004028198250816796,
+      "loss": 3.6884,
+      "step": 7240
+    },
+    {
+      "epoch": 0.3048842105263158,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.000402793117960545,
+      "loss": 3.4626,
+      "step": 7241
+    },
+    {
+      "epoch": 0.3049263157894737,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004027664080556994,
+      "loss": 3.215,
+      "step": 7242
+    },
+    {
+      "epoch": 0.30496842105263156,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00040273969536762947,
+      "loss": 3.3726,
+      "step": 7243
+    },
+    {
+      "epoch": 0.30501052631578945,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00040271297989682156,
+      "loss": 3.3472,
+      "step": 7244
+    },
+    {
+      "epoch": 0.30505263157894735,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00040268626164376267,
+      "loss": 3.553,
+      "step": 7245
+    },
+    {
+      "epoch": 0.30509473684210525,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0004026595406089395,
+      "loss": 3.4587,
+      "step": 7246
+    },
+    {
+      "epoch": 0.30513684210526315,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004026328167928388,
+      "loss": 2.9574,
+      "step": 7247
+    },
+    {
+      "epoch": 0.30517894736842105,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00040260609019594757,
+      "loss": 3.205,
+      "step": 7248
+    },
+    {
+      "epoch": 0.30522105263157895,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00040257936081875267,
+      "loss": 2.8901,
+      "step": 7249
+    },
+    {
+      "epoch": 0.30526315789473685,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00040255262866174113,
+      "loss": 3.8291,
+      "step": 7250
+    },
+    {
+      "epoch": 0.30530526315789475,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00040252589372539993,
+      "loss": 3.3451,
+      "step": 7251
+    },
+    {
+      "epoch": 0.30534736842105265,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0004024991560102161,
+      "loss": 2.8464,
+      "step": 7252
+    },
+    {
+      "epoch": 0.30538947368421054,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00040247241551667686,
+      "loss": 3.1878,
+      "step": 7253
+    },
+    {
+      "epoch": 0.30543157894736844,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004024456722452694,
+      "loss": 3.3616,
+      "step": 7254
+    },
+    {
+      "epoch": 0.30547368421052634,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004024189261964808,
+      "loss": 3.2905,
+      "step": 7255
+    },
+    {
+      "epoch": 0.3055157894736842,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00040239217737079845,
+      "loss": 3.0141,
+      "step": 7256
+    },
+    {
+      "epoch": 0.3055578947368421,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0004023654257687098,
+      "loss": 3.0116,
+      "step": 7257
+    },
+    {
+      "epoch": 0.3056,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00040233867139070206,
+      "loss": 2.9015,
+      "step": 7258
+    },
+    {
+      "epoch": 0.3056421052631579,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00040231191423726275,
+      "loss": 3.3728,
+      "step": 7259
+    },
+    {
+      "epoch": 0.3056842105263158,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00040228515430887926,
+      "loss": 3.584,
+      "step": 7260
+    },
+    {
+      "epoch": 0.3057263157894737,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004022583916060392,
+      "loss": 3.2033,
+      "step": 7261
+    },
+    {
+      "epoch": 0.3057684210526316,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00040223162612923013,
+      "loss": 3.2497,
+      "step": 7262
+    },
+    {
+      "epoch": 0.3058105263157895,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00040220485787893973,
+      "loss": 3.3595,
+      "step": 7263
+    },
+    {
+      "epoch": 0.3058526315789474,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0004021780868556556,
+      "loss": 3.1713,
+      "step": 7264
+    },
+    {
+      "epoch": 0.3058947368421053,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00040215131305986563,
+      "loss": 3.533,
+      "step": 7265
+    },
+    {
+      "epoch": 0.3059368421052632,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00040212453649205745,
+      "loss": 3.5698,
+      "step": 7266
+    },
+    {
+      "epoch": 0.3059789473684211,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00040209775715271896,
+      "loss": 3.1792,
+      "step": 7267
+    },
+    {
+      "epoch": 0.30602105263157897,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.000402070975042338,
+      "loss": 3.3752,
+      "step": 7268
+    },
+    {
+      "epoch": 0.3060631578947368,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0004020441901614026,
+      "loss": 3.1684,
+      "step": 7269
+    },
+    {
+      "epoch": 0.3061052631578947,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0004020174025104006,
+      "loss": 3.2718,
+      "step": 7270
+    },
+    {
+      "epoch": 0.3061473684210526,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00040199061208982036,
+      "loss": 2.7433,
+      "step": 7271
+    },
+    {
+      "epoch": 0.3061894736842105,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00040196381890014954,
+      "loss": 3.2036,
+      "step": 7272
+    },
+    {
+      "epoch": 0.3062315789473684,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00040193702294187664,
+      "loss": 3.2174,
+      "step": 7273
+    },
+    {
+      "epoch": 0.3062736842105263,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0004019102242154896,
+      "loss": 3.0607,
+      "step": 7274
+    },
+    {
+      "epoch": 0.3063157894736842,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004018834227214767,
+      "loss": 3.1852,
+      "step": 7275
+    },
+    {
+      "epoch": 0.3063578947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004018566184603264,
+      "loss": 2.9697,
+      "step": 7276
+    },
+    {
+      "epoch": 0.3064,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00040182981143252694,
+      "loss": 2.9302,
+      "step": 7277
+    },
+    {
+      "epoch": 0.3064421052631579,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0004018030016385667,
+      "loss": 3.4449,
+      "step": 7278
+    },
+    {
+      "epoch": 0.3064842105263158,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0004017761890789341,
+      "loss": 3.0799,
+      "step": 7279
+    },
+    {
+      "epoch": 0.3065263157894737,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0004017493737541177,
+      "loss": 2.6804,
+      "step": 7280
+    },
+    {
+      "epoch": 0.3065684210526316,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00040172255566460604,
+      "loss": 2.8409,
+      "step": 7281
+    },
+    {
+      "epoch": 0.3066105263157895,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0004016957348108875,
+      "loss": 3.3747,
+      "step": 7282
+    },
+    {
+      "epoch": 0.30665263157894734,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00040166891119345103,
+      "loss": 3.7477,
+      "step": 7283
+    },
+    {
+      "epoch": 0.30669473684210524,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0004016420848127853,
+      "loss": 2.9036,
+      "step": 7284
+    },
+    {
+      "epoch": 0.30673684210526314,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00040161525566937887,
+      "loss": 3.4695,
+      "step": 7285
+    },
+    {
+      "epoch": 0.30677894736842104,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0004015884237637206,
+      "loss": 3.512,
+      "step": 7286
+    },
+    {
+      "epoch": 0.30682105263157894,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00040156158909629934,
+      "loss": 3.3636,
+      "step": 7287
+    },
+    {
+      "epoch": 0.30686315789473684,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004015347516676041,
+      "loss": 3.222,
+      "step": 7288
+    },
+    {
+      "epoch": 0.30690526315789474,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0004015079114781236,
+      "loss": 3.3031,
+      "step": 7289
+    },
+    {
+      "epoch": 0.30694736842105264,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0004014810685283471,
+      "loss": 3.0917,
+      "step": 7290
+    },
+    {
+      "epoch": 0.30698947368421053,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004014542228187634,
+      "loss": 3.6075,
+      "step": 7291
+    },
+    {
+      "epoch": 0.30703157894736843,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004014273743498618,
+      "loss": 2.7728,
+      "step": 7292
+    },
+    {
+      "epoch": 0.30707368421052633,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00040140052312213133,
+      "loss": 3.0342,
+      "step": 7293
+    },
+    {
+      "epoch": 0.30711578947368423,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00040137366913606124,
+      "loss": 3.2024,
+      "step": 7294
+    },
+    {
+      "epoch": 0.30715789473684213,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00040134681239214066,
+      "loss": 3.0137,
+      "step": 7295
+    },
+    {
+      "epoch": 0.3072,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00040131995289085907,
+      "loss": 3.5906,
+      "step": 7296
+    },
+    {
+      "epoch": 0.30724210526315787,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0004012930906327057,
+      "loss": 3.3314,
+      "step": 7297
+    },
+    {
+      "epoch": 0.30728421052631577,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00040126622561817,
+      "loss": 3.1571,
+      "step": 7298
+    },
+    {
+      "epoch": 0.30732631578947367,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004012393578477414,
+      "loss": 2.8867,
+      "step": 7299
+    },
+    {
+      "epoch": 0.30736842105263157,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004012124873219094,
+      "loss": 3.1355,
+      "step": 7300
+    },
+    {
+      "epoch": 0.30741052631578947,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004011856140411635,
+      "loss": 3.3088,
+      "step": 7301
+    },
+    {
+      "epoch": 0.30745263157894737,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00040115873800599343,
+      "loss": 3.3906,
+      "step": 7302
+    },
+    {
+      "epoch": 0.30749473684210527,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00040113185921688866,
+      "loss": 3.3675,
+      "step": 7303
+    },
+    {
+      "epoch": 0.30753684210526316,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00040110497767433906,
+      "loss": 3.2245,
+      "step": 7304
+    },
+    {
+      "epoch": 0.30757894736842106,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004010780933788343,
+      "loss": 3.3727,
+      "step": 7305
+    },
+    {
+      "epoch": 0.30762105263157896,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004010512063308642,
+      "loss": 2.8313,
+      "step": 7306
+    },
+    {
+      "epoch": 0.30766315789473686,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0004010243165309186,
+      "loss": 3.0548,
+      "step": 7307
+    },
+    {
+      "epoch": 0.30770526315789476,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00040099742397948737,
+      "loss": 2.9205,
+      "step": 7308
+    },
+    {
+      "epoch": 0.30774736842105266,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00040097052867706047,
+      "loss": 2.7559,
+      "step": 7309
+    },
+    {
+      "epoch": 0.3077894736842105,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00040094363062412796,
+      "loss": 3.5637,
+      "step": 7310
+    },
+    {
+      "epoch": 0.3078315789473684,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00040091672982117975,
+      "loss": 3.051,
+      "step": 7311
+    },
+    {
+      "epoch": 0.3078736842105263,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00040088982626870616,
+      "loss": 2.5523,
+      "step": 7312
+    },
+    {
+      "epoch": 0.3079157894736842,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00040086291996719724,
+      "loss": 3.0535,
+      "step": 7313
+    },
+    {
+      "epoch": 0.3079578947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0004008360109171432,
+      "loss": 3.2792,
+      "step": 7314
+    },
+    {
+      "epoch": 0.308,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00040080909911903407,
+      "loss": 2.986,
+      "step": 7315
+    },
+    {
+      "epoch": 0.3080421052631579,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0004007821845733605,
+      "loss": 3.0661,
+      "step": 7316
+    },
+    {
+      "epoch": 0.3080842105263158,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00040075526728061266,
+      "loss": 2.871,
+      "step": 7317
+    },
+    {
+      "epoch": 0.3081263157894737,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.000400728347241281,
+      "loss": 3.1113,
+      "step": 7318
+    },
+    {
+      "epoch": 0.3081684210526316,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004007014244558559,
+      "loss": 3.4014,
+      "step": 7319
+    },
+    {
+      "epoch": 0.3082105263157895,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.000400674498924828,
+      "loss": 3.5296,
+      "step": 7320
+    },
+    {
+      "epoch": 0.3082526315789474,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0004006475706486877,
+      "loss": 2.8545,
+      "step": 7321
+    },
+    {
+      "epoch": 0.3082947368421053,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0004006206396279257,
+      "loss": 3.2672,
+      "step": 7322
+    },
+    {
+      "epoch": 0.30833684210526313,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00040059370586303255,
+      "loss": 3.1239,
+      "step": 7323
+    },
+    {
+      "epoch": 0.30837894736842103,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0004005667693544991,
+      "loss": 3.2319,
+      "step": 7324
+    },
+    {
+      "epoch": 0.30842105263157893,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00040053983010281596,
+      "loss": 3.4342,
+      "step": 7325
+    },
+    {
+      "epoch": 0.30846315789473683,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.000400512888108474,
+      "loss": 3.5286,
+      "step": 7326
+    },
+    {
+      "epoch": 0.3085052631578947,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004004859433719641,
+      "loss": 3.2997,
+      "step": 7327
+    },
+    {
+      "epoch": 0.3085473684210526,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0004004589958937771,
+      "loss": 3.1088,
+      "step": 7328
+    },
+    {
+      "epoch": 0.3085894736842105,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.0004004320456744039,
+      "loss": 3.0759,
+      "step": 7329
+    },
+    {
+      "epoch": 0.3086315789473684,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0004004050927143358,
+      "loss": 2.9664,
+      "step": 7330
+    },
+    {
+      "epoch": 0.3086736842105263,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00040037813701406346,
+      "loss": 3.0985,
+      "step": 7331
+    },
+    {
+      "epoch": 0.3087157894736842,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00040035117857407823,
+      "loss": 3.1973,
+      "step": 7332
+    },
+    {
+      "epoch": 0.3087578947368421,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0004003242173948711,
+      "loss": 2.8775,
+      "step": 7333
+    },
+    {
+      "epoch": 0.3088,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0004002972534769334,
+      "loss": 3.2265,
+      "step": 7334
+    },
+    {
+      "epoch": 0.3088421052631579,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00040027028682075626,
+      "loss": 3.2896,
+      "step": 7335
+    },
+    {
+      "epoch": 0.3088842105263158,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00040024331742683113,
+      "loss": 3.3278,
+      "step": 7336
+    },
+    {
+      "epoch": 0.30892631578947366,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004002163452956493,
+      "loss": 3.1903,
+      "step": 7337
+    },
+    {
+      "epoch": 0.30896842105263156,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0004001893704277021,
+      "loss": 3.2786,
+      "step": 7338
+    },
+    {
+      "epoch": 0.30901052631578946,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0004001623928234811,
+      "loss": 3.4059,
+      "step": 7339
+    },
+    {
+      "epoch": 0.30905263157894736,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0004001354124834777,
+      "loss": 3.2033,
+      "step": 7340
+    },
+    {
+      "epoch": 0.30909473684210526,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0004001084294081835,
+      "loss": 3.5038,
+      "step": 7341
+    },
+    {
+      "epoch": 0.30913684210526315,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00040008144359809,
+      "loss": 3.5789,
+      "step": 7342
+    },
+    {
+      "epoch": 0.30917894736842105,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.000400054455053689,
+      "loss": 3.352,
+      "step": 7343
+    },
+    {
+      "epoch": 0.30922105263157895,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004000274637754722,
+      "loss": 2.9182,
+      "step": 7344
+    },
+    {
+      "epoch": 0.30926315789473685,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0004000004697639312,
+      "loss": 2.9284,
+      "step": 7345
+    },
+    {
+      "epoch": 0.30930526315789475,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0003999734730195579,
+      "loss": 3.6692,
+      "step": 7346
+    },
+    {
+      "epoch": 0.30934736842105265,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0003999464735428442,
+      "loss": 3.4437,
+      "step": 7347
+    },
+    {
+      "epoch": 0.30938947368421055,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00039991947133428187,
+      "loss": 3.6116,
+      "step": 7348
+    },
+    {
+      "epoch": 0.30943157894736845,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003998924663943629,
+      "loss": 3.4609,
+      "step": 7349
+    },
+    {
+      "epoch": 0.3094736842105263,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0003998654587235793,
+      "loss": 3.1128,
+      "step": 7350
+    },
+    {
+      "epoch": 0.3095157894736842,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0003998384483224232,
+      "loss": 3.2029,
+      "step": 7351
+    },
+    {
+      "epoch": 0.3095578947368421,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00039981143519138664,
+      "loss": 3.0567,
+      "step": 7352
+    },
+    {
+      "epoch": 0.3096,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00039978441933096164,
+      "loss": 3.2865,
+      "step": 7353
+    },
+    {
+      "epoch": 0.3096421052631579,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003997574007416406,
+      "loss": 3.7253,
+      "step": 7354
+    },
+    {
+      "epoch": 0.3096842105263158,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003997303794239157,
+      "loss": 2.9425,
+      "step": 7355
+    },
+    {
+      "epoch": 0.3097263157894737,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00039970335537827906,
+      "loss": 3.2557,
+      "step": 7356
+    },
+    {
+      "epoch": 0.3097684210526316,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0003996763286052233,
+      "loss": 3.7003,
+      "step": 7357
+    },
+    {
+      "epoch": 0.3098105263157895,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00039964929910524065,
+      "loss": 3.0516,
+      "step": 7358
+    },
+    {
+      "epoch": 0.3098526315789474,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00039962226687882365,
+      "loss": 2.9853,
+      "step": 7359
+    },
+    {
+      "epoch": 0.3098947368421053,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003995952319264647,
+      "loss": 3.5054,
+      "step": 7360
+    },
+    {
+      "epoch": 0.3099368421052632,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00039956819424865643,
+      "loss": 3.1239,
+      "step": 7361
+    },
+    {
+      "epoch": 0.3099789473684211,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003995411538458913,
+      "loss": 3.2333,
+      "step": 7362
+    },
+    {
+      "epoch": 0.310021052631579,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00039951411071866217,
+      "loss": 3.22,
+      "step": 7363
+    },
+    {
+      "epoch": 0.3100631578947368,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0003994870648674616,
+      "loss": 2.7908,
+      "step": 7364
+    },
+    {
+      "epoch": 0.3101052631578947,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00039946001629278226,
+      "loss": 3.411,
+      "step": 7365
+    },
+    {
+      "epoch": 0.3101473684210526,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00039943296499511715,
+      "loss": 3.3301,
+      "step": 7366
+    },
+    {
+      "epoch": 0.3101894736842105,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003994059109749589,
+      "loss": 2.9942,
+      "step": 7367
+    },
+    {
+      "epoch": 0.3102315789473684,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00039937885423280064,
+      "loss": 3.3597,
+      "step": 7368
+    },
+    {
+      "epoch": 0.3102736842105263,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.000399351794769135,
+      "loss": 3.2265,
+      "step": 7369
+    },
+    {
+      "epoch": 0.3103157894736842,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003993247325844552,
+      "loss": 3.1726,
+      "step": 7370
+    },
+    {
+      "epoch": 0.3103578947368421,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00039929766767925423,
+      "loss": 3.0721,
+      "step": 7371
+    },
+    {
+      "epoch": 0.3104,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00039927060005402515,
+      "loss": 3.5639,
+      "step": 7372
+    },
+    {
+      "epoch": 0.3104421052631579,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003992435297092611,
+      "loss": 3.814,
+      "step": 7373
+    },
+    {
+      "epoch": 0.3104842105263158,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00039921645664545537,
+      "loss": 2.988,
+      "step": 7374
+    },
+    {
+      "epoch": 0.3105263157894737,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003991893808631011,
+      "loss": 3.612,
+      "step": 7375
+    },
+    {
+      "epoch": 0.3105684210526316,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0003991623023626916,
+      "loss": 3.248,
+      "step": 7376
+    },
+    {
+      "epoch": 0.31061052631578945,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00039913522114472016,
+      "loss": 3.1642,
+      "step": 7377
+    },
+    {
+      "epoch": 0.31065263157894735,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00039910813720968026,
+      "loss": 3.2375,
+      "step": 7378
+    },
+    {
+      "epoch": 0.31069473684210525,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003990810505580653,
+      "loss": 3.0129,
+      "step": 7379
+    },
+    {
+      "epoch": 0.31073684210526314,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00039905396119036874,
+      "loss": 2.8426,
+      "step": 7380
+    },
+    {
+      "epoch": 0.31077894736842104,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003990268691070841,
+      "loss": 3.0598,
+      "step": 7381
+    },
+    {
+      "epoch": 0.31082105263157894,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003989997743087051,
+      "loss": 3.1521,
+      "step": 7382
+    },
+    {
+      "epoch": 0.31086315789473684,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00039897267679572524,
+      "loss": 2.8502,
+      "step": 7383
+    },
+    {
+      "epoch": 0.31090526315789474,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003989455765686382,
+      "loss": 3.3394,
+      "step": 7384
+    },
+    {
+      "epoch": 0.31094736842105264,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0003989184736279378,
+      "loss": 2.8559,
+      "step": 7385
+    },
+    {
+      "epoch": 0.31098947368421054,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00039889136797411777,
+      "loss": 3.401,
+      "step": 7386
+    },
+    {
+      "epoch": 0.31103157894736844,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00039886425960767197,
+      "loss": 3.4051,
+      "step": 7387
+    },
+    {
+      "epoch": 0.31107368421052634,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003988371485290943,
+      "loss": 3.0646,
+      "step": 7388
+    },
+    {
+      "epoch": 0.31111578947368423,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00039881003473887853,
+      "loss": 3.3619,
+      "step": 7389
+    },
+    {
+      "epoch": 0.3111578947368421,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003987829182375189,
+      "loss": 2.5672,
+      "step": 7390
+    },
+    {
+      "epoch": 0.3112,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0003987557990255093,
+      "loss": 3.7055,
+      "step": 7391
+    },
+    {
+      "epoch": 0.3112421052631579,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003987286771033438,
+      "loss": 3.2788,
+      "step": 7392
+    },
+    {
+      "epoch": 0.3112842105263158,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00039870155247151656,
+      "loss": 3.3632,
+      "step": 7393
+    },
+    {
+      "epoch": 0.3113263157894737,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0003986744251305218,
+      "loss": 3.2673,
+      "step": 7394
+    },
+    {
+      "epoch": 0.31136842105263157,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00039864729508085354,
+      "loss": 3.1612,
+      "step": 7395
+    },
+    {
+      "epoch": 0.31141052631578947,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00039862016232300636,
+      "loss": 3.0306,
+      "step": 7396
+    },
+    {
+      "epoch": 0.31145263157894737,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00039859302685747434,
+      "loss": 3.0856,
+      "step": 7397
+    },
+    {
+      "epoch": 0.31149473684210527,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00039856588868475207,
+      "loss": 3.2754,
+      "step": 7398
+    },
+    {
+      "epoch": 0.31153684210526317,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0003985387478053338,
+      "loss": 3.7474,
+      "step": 7399
+    },
+    {
+      "epoch": 0.31157894736842107,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00039851160421971413,
+      "loss": 3.342,
+      "step": 7400
+    },
+    {
+      "epoch": 0.31162105263157897,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003984844579283875,
+      "loss": 3.2166,
+      "step": 7401
+    },
+    {
+      "epoch": 0.31166315789473686,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00039845730893184843,
+      "loss": 3.2499,
+      "step": 7402
+    },
+    {
+      "epoch": 0.31170526315789476,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00039843015723059174,
+      "loss": 3.4656,
+      "step": 7403
+    },
+    {
+      "epoch": 0.3117473684210526,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00039840300282511186,
+      "loss": 3.3947,
+      "step": 7404
+    },
+    {
+      "epoch": 0.3117894736842105,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0003983758457159037,
+      "loss": 3.3461,
+      "step": 7405
+    },
+    {
+      "epoch": 0.3118315789473684,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00039834868590346204,
+      "loss": 3.441,
+      "step": 7406
+    },
+    {
+      "epoch": 0.3118736842105263,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00039832152338828154,
+      "loss": 3.3021,
+      "step": 7407
+    },
+    {
+      "epoch": 0.3119157894736842,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0003982943581708571,
+      "loss": 3.3159,
+      "step": 7408
+    },
+    {
+      "epoch": 0.3119578947368421,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00039826719025168377,
+      "loss": 2.7332,
+      "step": 7409
+    },
+    {
+      "epoch": 0.312,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00039824001963125643,
+      "loss": 3.6423,
+      "step": 7410
+    },
+    {
+      "epoch": 0.3120421052631579,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00039821284631007015,
+      "loss": 3.083,
+      "step": 7411
+    },
+    {
+      "epoch": 0.3120842105263158,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00039818567028861987,
+      "loss": 2.8152,
+      "step": 7412
+    },
+    {
+      "epoch": 0.3121263157894737,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00039815849156740084,
+      "loss": 3.3076,
+      "step": 7413
+    },
+    {
+      "epoch": 0.3121684210526316,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003981313101469082,
+      "loss": 3.4314,
+      "step": 7414
+    },
+    {
+      "epoch": 0.3122105263157895,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003981041260276371,
+      "loss": 3.4,
+      "step": 7415
+    },
+    {
+      "epoch": 0.3122526315789474,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00039807693921008284,
+      "loss": 3.4535,
+      "step": 7416
+    },
+    {
+      "epoch": 0.31229473684210524,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003980497496947407,
+      "loss": 3.1372,
+      "step": 7417
+    },
+    {
+      "epoch": 0.31233684210526313,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003980225574821061,
+      "loss": 3.1125,
+      "step": 7418
+    },
+    {
+      "epoch": 0.31237894736842103,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003979953625726744,
+      "loss": 2.9954,
+      "step": 7419
+    },
+    {
+      "epoch": 0.31242105263157893,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003979681649669411,
+      "loss": 3.1749,
+      "step": 7420
+    },
+    {
+      "epoch": 0.31246315789473683,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00039794096466540176,
+      "loss": 3.4737,
+      "step": 7421
+    },
+    {
+      "epoch": 0.31250526315789473,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00039791376166855175,
+      "loss": 3.7119,
+      "step": 7422
+    },
+    {
+      "epoch": 0.31254736842105263,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00039788655597688684,
+      "loss": 3.6328,
+      "step": 7423
+    },
+    {
+      "epoch": 0.31258947368421053,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003978593475909026,
+      "loss": 2.9337,
+      "step": 7424
+    },
+    {
+      "epoch": 0.3126315789473684,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00039783213651109484,
+      "loss": 3.0813,
+      "step": 7425
+    },
+    {
+      "epoch": 0.3126736842105263,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0003978049227379592,
+      "loss": 3.4933,
+      "step": 7426
+    },
+    {
+      "epoch": 0.3127157894736842,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003977777062719916,
+      "loss": 3.1789,
+      "step": 7427
+    },
+    {
+      "epoch": 0.3127578947368421,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00039775048711368776,
+      "loss": 3.3626,
+      "step": 7428
+    },
+    {
+      "epoch": 0.3128,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00039772326526354366,
+      "loss": 3.0785,
+      "step": 7429
+    },
+    {
+      "epoch": 0.3128421052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00039769604072205524,
+      "loss": 3.2577,
+      "step": 7430
+    },
+    {
+      "epoch": 0.31288421052631576,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0003976688134897185,
+      "loss": 3.2389,
+      "step": 7431
+    },
+    {
+      "epoch": 0.31292631578947366,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003976415835670295,
+      "loss": 3.4281,
+      "step": 7432
+    },
+    {
+      "epoch": 0.31296842105263156,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003976143509544843,
+      "loss": 3.4499,
+      "step": 7433
+    },
+    {
+      "epoch": 0.31301052631578946,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00039758711565257907,
+      "loss": 3.5175,
+      "step": 7434
+    },
+    {
+      "epoch": 0.31305263157894736,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00039755987766181,
+      "loss": 3.2529,
+      "step": 7435
+    },
+    {
+      "epoch": 0.31309473684210526,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003975326369826733,
+      "loss": 3.315,
+      "step": 7436
+    },
+    {
+      "epoch": 0.31313684210526316,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00039750539361566527,
+      "loss": 3.2597,
+      "step": 7437
+    },
+    {
+      "epoch": 0.31317894736842106,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00039747814756128236,
+      "loss": 3.2667,
+      "step": 7438
+    },
+    {
+      "epoch": 0.31322105263157896,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00039745089882002084,
+      "loss": 3.2693,
+      "step": 7439
+    },
+    {
+      "epoch": 0.31326315789473685,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0003974236473923772,
+      "loss": 2.9445,
+      "step": 7440
+    },
+    {
+      "epoch": 0.31330526315789475,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00039739639327884797,
+      "loss": 3.3583,
+      "step": 7441
+    },
+    {
+      "epoch": 0.31334736842105265,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0003973691364799296,
+      "loss": 2.86,
+      "step": 7442
+    },
+    {
+      "epoch": 0.31338947368421055,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003973418769961187,
+      "loss": 3.4056,
+      "step": 7443
+    },
+    {
+      "epoch": 0.3134315789473684,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003973146148279119,
+      "loss": 3.0453,
+      "step": 7444
+    },
+    {
+      "epoch": 0.3134736842105263,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00039728734997580595,
+      "loss": 3.0676,
+      "step": 7445
+    },
+    {
+      "epoch": 0.3135157894736842,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00039726008244029756,
+      "loss": 3.1237,
+      "step": 7446
+    },
+    {
+      "epoch": 0.3135578947368421,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00039723281222188347,
+      "loss": 3.267,
+      "step": 7447
+    },
+    {
+      "epoch": 0.3136,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003972055393210605,
+      "loss": 3.3411,
+      "step": 7448
+    },
+    {
+      "epoch": 0.3136421052631579,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00039717826373832554,
+      "loss": 3.1061,
+      "step": 7449
+    },
+    {
+      "epoch": 0.3136842105263158,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003971509854741756,
+      "loss": 3.0621,
+      "step": 7450
+    },
+    {
+      "epoch": 0.3137263157894737,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00039712370452910757,
+      "loss": 3.3853,
+      "step": 7451
+    },
+    {
+      "epoch": 0.3137684210526316,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00039709642090361854,
+      "loss": 3.3365,
+      "step": 7452
+    },
+    {
+      "epoch": 0.3138105263157895,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00039706913459820547,
+      "loss": 2.7026,
+      "step": 7453
+    },
+    {
+      "epoch": 0.3138526315789474,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003970418456133657,
+      "loss": 3.0366,
+      "step": 7454
+    },
+    {
+      "epoch": 0.3138947368421053,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00039701455394959616,
+      "loss": 3.2023,
+      "step": 7455
+    },
+    {
+      "epoch": 0.3139368421052632,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00039698725960739424,
+      "loss": 3.1295,
+      "step": 7456
+    },
+    {
+      "epoch": 0.3139789473684211,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003969599625872571,
+      "loss": 3.4282,
+      "step": 7457
+    },
+    {
+      "epoch": 0.3140210526315789,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003969326628896822,
+      "loss": 3.2976,
+      "step": 7458
+    },
+    {
+      "epoch": 0.3140631578947368,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00039690536051516673,
+      "loss": 3.347,
+      "step": 7459
+    },
+    {
+      "epoch": 0.3141052631578947,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00039687805546420824,
+      "loss": 3.0647,
+      "step": 7460
+    },
+    {
+      "epoch": 0.3141473684210526,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0003968507477373041,
+      "loss": 3.1937,
+      "step": 7461
+    },
+    {
+      "epoch": 0.3141894736842105,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.000396823437334952,
+      "loss": 3.225,
+      "step": 7462
+    },
+    {
+      "epoch": 0.3142315789473684,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003967961242576493,
+      "loss": 3.1896,
+      "step": 7463
+    },
+    {
+      "epoch": 0.3142736842105263,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00039676880850589367,
+      "loss": 3.4536,
+      "step": 7464
+    },
+    {
+      "epoch": 0.3143157894736842,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00039674149008018286,
+      "loss": 2.8948,
+      "step": 7465
+    },
+    {
+      "epoch": 0.3143578947368421,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00039671416898101445,
+      "loss": 3.2017,
+      "step": 7466
+    },
+    {
+      "epoch": 0.3144,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003966868452088863,
+      "loss": 3.6467,
+      "step": 7467
+    },
+    {
+      "epoch": 0.3144421052631579,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00039665951876429617,
+      "loss": 3.3365,
+      "step": 7468
+    },
+    {
+      "epoch": 0.3144842105263158,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00039663218964774196,
+      "loss": 3.6083,
+      "step": 7469
+    },
+    {
+      "epoch": 0.3145263157894737,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003966048578597215,
+      "loss": 3.2057,
+      "step": 7470
+    },
+    {
+      "epoch": 0.31456842105263155,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003965775234007328,
+      "loss": 3.3493,
+      "step": 7471
+    },
+    {
+      "epoch": 0.31461052631578945,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0003965501862712738,
+      "loss": 3.1587,
+      "step": 7472
+    },
+    {
+      "epoch": 0.31465263157894735,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00039652284647184266,
+      "loss": 3.6313,
+      "step": 7473
+    },
+    {
+      "epoch": 0.31469473684210525,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00039649550400293744,
+      "loss": 2.87,
+      "step": 7474
+    },
+    {
+      "epoch": 0.31473684210526315,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00039646815886505623,
+      "loss": 3.2297,
+      "step": 7475
+    },
+    {
+      "epoch": 0.31477894736842105,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003964408110586972,
+      "loss": 3.3323,
+      "step": 7476
+    },
+    {
+      "epoch": 0.31482105263157895,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00039641346058435866,
+      "loss": 2.9341,
+      "step": 7477
+    },
+    {
+      "epoch": 0.31486315789473684,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.000396386107442539,
+      "loss": 2.8989,
+      "step": 7478
+    },
+    {
+      "epoch": 0.31490526315789474,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00039635875163373635,
+      "loss": 3.4092,
+      "step": 7479
+    },
+    {
+      "epoch": 0.31494736842105264,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0003963313931584493,
+      "loss": 2.8028,
+      "step": 7480
+    },
+    {
+      "epoch": 0.31498947368421054,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003963040320171761,
+      "loss": 3.0914,
+      "step": 7481
+    },
+    {
+      "epoch": 0.31503157894736844,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0003962766682104154,
+      "loss": 2.9201,
+      "step": 7482
+    },
+    {
+      "epoch": 0.31507368421052634,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0003962493017386657,
+      "loss": 3.1492,
+      "step": 7483
+    },
+    {
+      "epoch": 0.31511578947368424,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00039622193260242545,
+      "loss": 3.2404,
+      "step": 7484
+    },
+    {
+      "epoch": 0.3151578947368421,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003961945608021935,
+      "loss": 3.5027,
+      "step": 7485
+    },
+    {
+      "epoch": 0.3152,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00039616718633846837,
+      "loss": 3.4893,
+      "step": 7486
+    },
+    {
+      "epoch": 0.3152421052631579,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0003961398092117489,
+      "loss": 3.7984,
+      "step": 7487
+    },
+    {
+      "epoch": 0.3152842105263158,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003961124294225338,
+      "loss": 3.0178,
+      "step": 7488
+    },
+    {
+      "epoch": 0.3153263157894737,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00039608504697132187,
+      "loss": 3.2586,
+      "step": 7489
+    },
+    {
+      "epoch": 0.3153684210526316,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00039605766185861203,
+      "loss": 3.5667,
+      "step": 7490
+    },
+    {
+      "epoch": 0.3154105263157895,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0003960302740849032,
+      "loss": 3.3659,
+      "step": 7491
+    },
+    {
+      "epoch": 0.3154526315789474,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00039600288365069435,
+      "loss": 2.9802,
+      "step": 7492
+    },
+    {
+      "epoch": 0.3154947368421053,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0003959754905564845,
+      "loss": 2.848,
+      "step": 7493
+    },
+    {
+      "epoch": 0.31553684210526317,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003959480948027728,
+      "loss": 3.2274,
+      "step": 7494
+    },
+    {
+      "epoch": 0.31557894736842107,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003959206963900582,
+      "loss": 3.1172,
+      "step": 7495
+    },
+    {
+      "epoch": 0.31562105263157897,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00039589329531884005,
+      "loss": 2.912,
+      "step": 7496
+    },
+    {
+      "epoch": 0.31566315789473687,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00039586589158961737,
+      "loss": 2.893,
+      "step": 7497
+    },
+    {
+      "epoch": 0.3157052631578947,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003958384852028896,
+      "loss": 3.0973,
+      "step": 7498
+    },
+    {
+      "epoch": 0.3157473684210526,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.000395811076159156,
+      "loss": 3.3533,
+      "step": 7499
+    },
+    {
+      "epoch": 0.3157894736842105,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00039578366445891587,
+      "loss": 3.2017,
+      "step": 7500
+    },
+    {
+      "epoch": 0.3158315789473684,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003957562501026687,
+      "loss": 2.975,
+      "step": 7501
+    },
+    {
+      "epoch": 0.3158736842105263,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003957288330909139,
+      "loss": 3.3691,
+      "step": 7502
+    },
+    {
+      "epoch": 0.3159157894736842,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.000395701413424151,
+      "loss": 3.3199,
+      "step": 7503
+    },
+    {
+      "epoch": 0.3159578947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003956739911028795,
+      "loss": 3.4767,
+      "step": 7504
+    },
+    {
+      "epoch": 0.316,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00039564656612759905,
+      "loss": 2.9177,
+      "step": 7505
+    },
+    {
+      "epoch": 0.3160421052631579,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00039561913849880926,
+      "loss": 3.4656,
+      "step": 7506
+    },
+    {
+      "epoch": 0.3160842105263158,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00039559170821701,
+      "loss": 3.5141,
+      "step": 7507
+    },
+    {
+      "epoch": 0.3161263157894737,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003955642752827007,
+      "loss": 3.2666,
+      "step": 7508
+    },
+    {
+      "epoch": 0.3161684210526316,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003955368396963814,
+      "loss": 3.5638,
+      "step": 7509
+    },
+    {
+      "epoch": 0.3162105263157895,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003955094014585519,
+      "loss": 3.117,
+      "step": 7510
+    },
+    {
+      "epoch": 0.31625263157894734,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0003954819605697121,
+      "loss": 3.1942,
+      "step": 7511
+    },
+    {
+      "epoch": 0.31629473684210524,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0003954545170303619,
+      "loss": 3.2696,
+      "step": 7512
+    },
+    {
+      "epoch": 0.31633684210526314,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00039542707084100126,
+      "loss": 3.4786,
+      "step": 7513
+    },
+    {
+      "epoch": 0.31637894736842104,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0003953996220021303,
+      "loss": 2.8893,
+      "step": 7514
+    },
+    {
+      "epoch": 0.31642105263157894,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00039537217051424905,
+      "loss": 3.3121,
+      "step": 7515
+    },
+    {
+      "epoch": 0.31646315789473684,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00039534471637785764,
+      "loss": 2.9904,
+      "step": 7516
+    },
+    {
+      "epoch": 0.31650526315789473,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003953172595934563,
+      "loss": 3.3074,
+      "step": 7517
+    },
+    {
+      "epoch": 0.31654736842105263,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0003952898001615452,
+      "loss": 3.2015,
+      "step": 7518
+    },
+    {
+      "epoch": 0.31658947368421053,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0003952623380826247,
+      "loss": 3.3429,
+      "step": 7519
+    },
+    {
+      "epoch": 0.31663157894736843,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.000395234873357195,
+      "loss": 3.2838,
+      "step": 7520
+    },
+    {
+      "epoch": 0.31667368421052633,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003952074059857566,
+      "loss": 3.5184,
+      "step": 7521
+    },
+    {
+      "epoch": 0.31671578947368423,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00039517993596880984,
+      "loss": 3.2192,
+      "step": 7522
+    },
+    {
+      "epoch": 0.3167578947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003951524633068553,
+      "loss": 3.4803,
+      "step": 7523
+    },
+    {
+      "epoch": 0.3168,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00039512498800039336,
+      "loss": 3.1692,
+      "step": 7524
+    },
+    {
+      "epoch": 0.31684210526315787,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003950975100499247,
+      "loss": 3.5659,
+      "step": 7525
+    },
+    {
+      "epoch": 0.31688421052631577,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00039507002945594986,
+      "loss": 3.2774,
+      "step": 7526
+    },
+    {
+      "epoch": 0.31692631578947367,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0003950425462189695,
+      "loss": 3.4328,
+      "step": 7527
+    },
+    {
+      "epoch": 0.31696842105263157,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003950150603394844,
+      "loss": 3.3204,
+      "step": 7528
+    },
+    {
+      "epoch": 0.31701052631578946,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0003949875718179953,
+      "loss": 3.3494,
+      "step": 7529
+    },
+    {
+      "epoch": 0.31705263157894736,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.000394960080655003,
+      "loss": 3.1209,
+      "step": 7530
+    },
+    {
+      "epoch": 0.31709473684210526,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003949325868510083,
+      "loss": 3.1695,
+      "step": 7531
+    },
+    {
+      "epoch": 0.31713684210526316,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003949050904065122,
+      "loss": 3.3405,
+      "step": 7532
+    },
+    {
+      "epoch": 0.31717894736842106,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0003948775913220156,
+      "loss": 3.1241,
+      "step": 7533
+    },
+    {
+      "epoch": 0.31722105263157896,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00039485008959801957,
+      "loss": 2.4633,
+      "step": 7534
+    },
+    {
+      "epoch": 0.31726315789473686,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.000394822585235025,
+      "loss": 3.1037,
+      "step": 7535
+    },
+    {
+      "epoch": 0.31730526315789476,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.0003947950782335332,
+      "loss": 3.2732,
+      "step": 7536
+    },
+    {
+      "epoch": 0.31734736842105266,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003947675685940451,
+      "loss": 3.2676,
+      "step": 7537
+    },
+    {
+      "epoch": 0.3173894736842105,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.000394740056317062,
+      "loss": 3.2375,
+      "step": 7538
+    },
+    {
+      "epoch": 0.3174315789473684,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00039471254140308515,
+      "loss": 3.2616,
+      "step": 7539
+    },
+    {
+      "epoch": 0.3174736842105263,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0003946850238526158,
+      "loss": 3.5061,
+      "step": 7540
+    },
+    {
+      "epoch": 0.3175157894736842,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00039465750366615536,
+      "loss": 2.7974,
+      "step": 7541
+    },
+    {
+      "epoch": 0.3175578947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003946299808442052,
+      "loss": 3.2561,
+      "step": 7542
+    },
+    {
+      "epoch": 0.3176,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003946024553872667,
+      "loss": 3.3596,
+      "step": 7543
+    },
+    {
+      "epoch": 0.3176421052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003945749272958413,
+      "loss": 3.1859,
+      "step": 7544
+    },
+    {
+      "epoch": 0.3176842105263158,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00039454739657043066,
+      "loss": 3.5208,
+      "step": 7545
+    },
+    {
+      "epoch": 0.3177263157894737,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00039451986321153624,
+      "loss": 2.8656,
+      "step": 7546
+    },
+    {
+      "epoch": 0.3177684210526316,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00039449232721965976,
+      "loss": 2.42,
+      "step": 7547
+    },
+    {
+      "epoch": 0.3178105263157895,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0003944647885953029,
+      "loss": 3.3134,
+      "step": 7548
+    },
+    {
+      "epoch": 0.3178526315789474,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00039443724733896726,
+      "loss": 3.2705,
+      "step": 7549
+    },
+    {
+      "epoch": 0.3178947368421053,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00039440970345115466,
+      "loss": 3.0717,
+      "step": 7550
+    },
+    {
+      "epoch": 0.3179368421052632,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003943821569323669,
+      "loss": 3.0024,
+      "step": 7551
+    },
+    {
+      "epoch": 0.317978947368421,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00039435460778310593,
+      "loss": 3.1211,
+      "step": 7552
+    },
+    {
+      "epoch": 0.3180210526315789,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0003943270560038737,
+      "loss": 3.5537,
+      "step": 7553
+    },
+    {
+      "epoch": 0.3180631578947368,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.000394299501595172,
+      "loss": 2.7536,
+      "step": 7554
+    },
+    {
+      "epoch": 0.3181052631578947,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003942719445575029,
+      "loss": 3.2498,
+      "step": 7555
+    },
+    {
+      "epoch": 0.3181473684210526,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.0003942443848913686,
+      "loss": 3.2498,
+      "step": 7556
+    },
+    {
+      "epoch": 0.3181894736842105,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.000394216822597271,
+      "loss": 3.1386,
+      "step": 7557
+    },
+    {
+      "epoch": 0.3182315789473684,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00039418925767571234,
+      "loss": 3.1969,
+      "step": 7558
+    },
+    {
+      "epoch": 0.3182736842105263,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00039416169012719483,
+      "loss": 3.5012,
+      "step": 7559
+    },
+    {
+      "epoch": 0.3183157894736842,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0003941341199522207,
+      "loss": 3.3976,
+      "step": 7560
+    },
+    {
+      "epoch": 0.3183578947368421,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0003941065471512923,
+      "loss": 2.9703,
+      "step": 7561
+    },
+    {
+      "epoch": 0.3184,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003940789717249119,
+      "loss": 3.506,
+      "step": 7562
+    },
+    {
+      "epoch": 0.3184421052631579,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003940513936735819,
+      "loss": 3.428,
+      "step": 7563
+    },
+    {
+      "epoch": 0.3184842105263158,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003940238129978048,
+      "loss": 3.6986,
+      "step": 7564
+    },
+    {
+      "epoch": 0.31852631578947366,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.000393996229698083,
+      "loss": 3.0628,
+      "step": 7565
+    },
+    {
+      "epoch": 0.31856842105263156,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00039396864377491914,
+      "loss": 3.2429,
+      "step": 7566
+    },
+    {
+      "epoch": 0.31861052631578946,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00039394105522881574,
+      "loss": 2.7927,
+      "step": 7567
+    },
+    {
+      "epoch": 0.31865263157894735,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003939134640602754,
+      "loss": 3.3551,
+      "step": 7568
+    },
+    {
+      "epoch": 0.31869473684210525,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00039388587026980084,
+      "loss": 2.9891,
+      "step": 7569
+    },
+    {
+      "epoch": 0.31873684210526315,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0003938582738578948,
+      "loss": 3.0966,
+      "step": 7570
+    },
+    {
+      "epoch": 0.31877894736842105,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00039383067482506,
+      "loss": 3.1353,
+      "step": 7571
+    },
+    {
+      "epoch": 0.31882105263157895,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0003938030731717993,
+      "loss": 2.9529,
+      "step": 7572
+    },
+    {
+      "epoch": 0.31886315789473685,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00039377546889861564,
+      "loss": 3.1503,
+      "step": 7573
+    },
+    {
+      "epoch": 0.31890526315789475,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0003937478620060119,
+      "loss": 2.9756,
+      "step": 7574
+    },
+    {
+      "epoch": 0.31894736842105265,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00039372025249449083,
+      "loss": 3.0516,
+      "step": 7575
+    },
+    {
+      "epoch": 0.31898947368421054,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00039369264036455577,
+      "loss": 3.0545,
+      "step": 7576
+    },
+    {
+      "epoch": 0.31903157894736844,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00039366502561670957,
+      "loss": 3.3041,
+      "step": 7577
+    },
+    {
+      "epoch": 0.31907368421052634,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00039363740825145544,
+      "loss": 3.0833,
+      "step": 7578
+    },
+    {
+      "epoch": 0.3191157894736842,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0003936097882692965,
+      "loss": 2.8939,
+      "step": 7579
+    },
+    {
+      "epoch": 0.3191578947368421,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003935821656707359,
+      "loss": 3.4183,
+      "step": 7580
+    },
+    {
+      "epoch": 0.3192,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.000393554540456277,
+      "loss": 2.8895,
+      "step": 7581
+    },
+    {
+      "epoch": 0.3192421052631579,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00039352691262642305,
+      "loss": 3.1806,
+      "step": 7582
+    },
+    {
+      "epoch": 0.3192842105263158,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00039349928218167734,
+      "loss": 3.1961,
+      "step": 7583
+    },
+    {
+      "epoch": 0.3193263157894737,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00039347164912254337,
+      "loss": 3.0577,
+      "step": 7584
+    },
+    {
+      "epoch": 0.3193684210526316,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003934440134495244,
+      "loss": 3.0253,
+      "step": 7585
+    },
+    {
+      "epoch": 0.3194105263157895,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00039341637516312426,
+      "loss": 3.0755,
+      "step": 7586
+    },
+    {
+      "epoch": 0.3194526315789474,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003933887342638461,
+      "loss": 2.944,
+      "step": 7587
+    },
+    {
+      "epoch": 0.3194947368421053,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00039336109075219374,
+      "loss": 3.5343,
+      "step": 7588
+    },
+    {
+      "epoch": 0.3195368421052632,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00039333344462867074,
+      "loss": 3.4872,
+      "step": 7589
+    },
+    {
+      "epoch": 0.3195789473684211,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003933057958937808,
+      "loss": 2.8422,
+      "step": 7590
+    },
+    {
+      "epoch": 0.319621052631579,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00039327814454802757,
+      "loss": 3.1027,
+      "step": 7591
+    },
+    {
+      "epoch": 0.3196631578947368,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00039325049059191496,
+      "loss": 3.4847,
+      "step": 7592
+    },
+    {
+      "epoch": 0.3197052631578947,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00039322283402594665,
+      "loss": 3.0165,
+      "step": 7593
+    },
+    {
+      "epoch": 0.3197473684210526,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00039319517485062665,
+      "loss": 3.0557,
+      "step": 7594
+    },
+    {
+      "epoch": 0.3197894736842105,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003931675130664588,
+      "loss": 3.4603,
+      "step": 7595
+    },
+    {
+      "epoch": 0.3198315789473684,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00039313984867394707,
+      "loss": 3.3797,
+      "step": 7596
+    },
+    {
+      "epoch": 0.3198736842105263,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00039311218167359553,
+      "loss": 3.4195,
+      "step": 7597
+    },
+    {
+      "epoch": 0.3199157894736842,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003930845120659081,
+      "loss": 3.2018,
+      "step": 7598
+    },
+    {
+      "epoch": 0.3199578947368421,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003930568398513891,
+      "loss": 3.0816,
+      "step": 7599
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00039302916503054243,
+      "loss": 3.8771,
+      "step": 7600
+    },
+    {
+      "epoch": 0.3200421052631579,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00039300148760387245,
+      "loss": 3.2862,
+      "step": 7601
+    },
+    {
+      "epoch": 0.3200842105263158,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00039297380757188346,
+      "loss": 3.3038,
+      "step": 7602
+    },
+    {
+      "epoch": 0.3201263157894737,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003929461249350796,
+      "loss": 3.0297,
+      "step": 7603
+    },
+    {
+      "epoch": 0.3201684210526316,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00039291843969396534,
+      "loss": 2.7953,
+      "step": 7604
+    },
+    {
+      "epoch": 0.3202105263157895,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.000392890751849045,
+      "loss": 3.5401,
+      "step": 7605
+    },
+    {
+      "epoch": 0.32025263157894734,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00039286306140082314,
+      "loss": 2.865,
+      "step": 7606
+    },
+    {
+      "epoch": 0.32029473684210524,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003928353683498041,
+      "loss": 3.4754,
+      "step": 7607
+    },
+    {
+      "epoch": 0.32033684210526314,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0003928076726964925,
+      "loss": 2.7858,
+      "step": 7608
+    },
+    {
+      "epoch": 0.32037894736842104,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0003927799744413928,
+      "loss": 3.1655,
+      "step": 7609
+    },
+    {
+      "epoch": 0.32042105263157894,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00039275227358500985,
+      "loss": 3.4781,
+      "step": 7610
+    },
+    {
+      "epoch": 0.32046315789473684,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00039272457012784815,
+      "loss": 3.0972,
+      "step": 7611
+    },
+    {
+      "epoch": 0.32050526315789474,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00039269686407041244,
+      "loss": 2.9018,
+      "step": 7612
+    },
+    {
+      "epoch": 0.32054736842105264,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0003926691554132076,
+      "loss": 3.0494,
+      "step": 7613
+    },
+    {
+      "epoch": 0.32058947368421054,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003926414441567383,
+      "loss": 2.8116,
+      "step": 7614
+    },
+    {
+      "epoch": 0.32063157894736843,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0003926137303015095,
+      "loss": 2.6566,
+      "step": 7615
+    },
+    {
+      "epoch": 0.32067368421052633,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003925860138480262,
+      "loss": 3.2842,
+      "step": 7616
+    },
+    {
+      "epoch": 0.32071578947368423,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0003925582947967932,
+      "loss": 2.9873,
+      "step": 7617
+    },
+    {
+      "epoch": 0.32075789473684213,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0003925305731483154,
+      "loss": 3.2454,
+      "step": 7618
+    },
+    {
+      "epoch": 0.3208,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00039250284890309824,
+      "loss": 2.9796,
+      "step": 7619
+    },
+    {
+      "epoch": 0.3208421052631579,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00039247512206164653,
+      "loss": 3.0595,
+      "step": 7620
+    },
+    {
+      "epoch": 0.32088421052631577,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00039244739262446547,
+      "loss": 3.2905,
+      "step": 7621
+    },
+    {
+      "epoch": 0.32092631578947367,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00039241966059206035,
+      "loss": 2.975,
+      "step": 7622
+    },
+    {
+      "epoch": 0.32096842105263157,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003923919259649363,
+      "loss": 2.9943,
+      "step": 7623
+    },
+    {
+      "epoch": 0.32101052631578947,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00039236418874359874,
+      "loss": 3.705,
+      "step": 7624
+    },
+    {
+      "epoch": 0.32105263157894737,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003923364489285529,
+      "loss": 3.6752,
+      "step": 7625
+    },
+    {
+      "epoch": 0.32109473684210527,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00039230870652030413,
+      "loss": 3.4704,
+      "step": 7626
+    },
+    {
+      "epoch": 0.32113684210526316,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.000392280961519358,
+      "loss": 3.2319,
+      "step": 7627
+    },
+    {
+      "epoch": 0.32117894736842106,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00039225321392621985,
+      "loss": 3.4493,
+      "step": 7628
+    },
+    {
+      "epoch": 0.32122105263157896,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00039222546374139533,
+      "loss": 3.6995,
+      "step": 7629
+    },
+    {
+      "epoch": 0.32126315789473686,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00039219771096539,
+      "loss": 3.4226,
+      "step": 7630
+    },
+    {
+      "epoch": 0.32130526315789476,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003921699555987095,
+      "loss": 3.0865,
+      "step": 7631
+    },
+    {
+      "epoch": 0.32134736842105266,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003921421976418593,
+      "loss": 2.878,
+      "step": 7632
+    },
+    {
+      "epoch": 0.3213894736842105,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0003921144370953453,
+      "loss": 3.0306,
+      "step": 7633
+    },
+    {
+      "epoch": 0.3214315789473684,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003920866739596733,
+      "loss": 3.1394,
+      "step": 7634
+    },
+    {
+      "epoch": 0.3214736842105263,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00039205890823534905,
+      "loss": 2.985,
+      "step": 7635
+    },
+    {
+      "epoch": 0.3215157894736842,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00039203113992287845,
+      "loss": 3.3118,
+      "step": 7636
+    },
+    {
+      "epoch": 0.3215578947368421,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00039200336902276724,
+      "loss": 2.7652,
+      "step": 7637
+    },
+    {
+      "epoch": 0.3216,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00039197559553552153,
+      "loss": 3.1406,
+      "step": 7638
+    },
+    {
+      "epoch": 0.3216421052631579,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00039194781946164727,
+      "loss": 3.4233,
+      "step": 7639
+    },
+    {
+      "epoch": 0.3216842105263158,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00039192004080165055,
+      "loss": 3.0441,
+      "step": 7640
+    },
+    {
+      "epoch": 0.3217263157894737,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003918922595560375,
+      "loss": 3.2039,
+      "step": 7641
+    },
+    {
+      "epoch": 0.3217684210526316,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00039186447572531414,
+      "loss": 3.0943,
+      "step": 7642
+    },
+    {
+      "epoch": 0.3218105263157895,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00039183668930998665,
+      "loss": 3.3078,
+      "step": 7643
+    },
+    {
+      "epoch": 0.3218526315789474,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00039180890031056145,
+      "loss": 2.9149,
+      "step": 7644
+    },
+    {
+      "epoch": 0.3218947368421053,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00039178110872754455,
+      "loss": 3.505,
+      "step": 7645
+    },
+    {
+      "epoch": 0.32193684210526313,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003917533145614426,
+      "loss": 3.4221,
+      "step": 7646
+    },
+    {
+      "epoch": 0.32197894736842103,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00039172551781276174,
+      "loss": 3.1722,
+      "step": 7647
+    },
+    {
+      "epoch": 0.32202105263157893,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00039169771848200847,
+      "loss": 3.4358,
+      "step": 7648
+    },
+    {
+      "epoch": 0.32206315789473683,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00039166991656968927,
+      "loss": 3.3386,
+      "step": 7649
+    },
+    {
+      "epoch": 0.32210526315789473,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0003916421120763106,
+      "loss": 2.9449,
+      "step": 7650
+    },
+    {
+      "epoch": 0.3221473684210526,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0003916143050023792,
+      "loss": 2.8812,
+      "step": 7651
+    },
+    {
+      "epoch": 0.3221894736842105,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003915864953484014,
+      "loss": 2.9171,
+      "step": 7652
+    },
+    {
+      "epoch": 0.3222315789473684,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.000391558683114884,
+      "loss": 2.9394,
+      "step": 7653
+    },
+    {
+      "epoch": 0.3222736842105263,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0003915308683023338,
+      "loss": 3.077,
+      "step": 7654
+    },
+    {
+      "epoch": 0.3223157894736842,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0003915030509112575,
+      "loss": 3.0394,
+      "step": 7655
+    },
+    {
+      "epoch": 0.3223578947368421,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0003914752309421618,
+      "loss": 3.1122,
+      "step": 7656
+    },
+    {
+      "epoch": 0.3224,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00039144740839555366,
+      "loss": 3.0605,
+      "step": 7657
+    },
+    {
+      "epoch": 0.3224421052631579,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003914195832719399,
+      "loss": 2.955,
+      "step": 7658
+    },
+    {
+      "epoch": 0.32248421052631576,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0003913917555718275,
+      "loss": 3.3906,
+      "step": 7659
+    },
+    {
+      "epoch": 0.32252631578947366,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00039136392529572344,
+      "loss": 3.1478,
+      "step": 7660
+    },
+    {
+      "epoch": 0.32256842105263156,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003913360924441348,
+      "loss": 3.1572,
+      "step": 7661
+    },
+    {
+      "epoch": 0.32261052631578946,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003913082570175686,
+      "loss": 3.3406,
+      "step": 7662
+    },
+    {
+      "epoch": 0.32265263157894736,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.000391280419016532,
+      "loss": 2.9968,
+      "step": 7663
+    },
+    {
+      "epoch": 0.32269473684210526,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0003912525784415321,
+      "loss": 2.9329,
+      "step": 7664
+    },
+    {
+      "epoch": 0.32273684210526316,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00039122473529307623,
+      "loss": 3.1511,
+      "step": 7665
+    },
+    {
+      "epoch": 0.32277894736842105,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003911968895716716,
+      "loss": 3.4112,
+      "step": 7666
+    },
+    {
+      "epoch": 0.32282105263157895,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0003911690412778256,
+      "loss": 2.9805,
+      "step": 7667
+    },
+    {
+      "epoch": 0.32286315789473685,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0003911411904120454,
+      "loss": 2.9821,
+      "step": 7668
+    },
+    {
+      "epoch": 0.32290526315789475,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0003911133369748387,
+      "loss": 3.7423,
+      "step": 7669
+    },
+    {
+      "epoch": 0.32294736842105265,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00039108548096671275,
+      "loss": 3.0658,
+      "step": 7670
+    },
+    {
+      "epoch": 0.32298947368421055,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003910576223881751,
+      "loss": 3.228,
+      "step": 7671
+    },
+    {
+      "epoch": 0.32303157894736845,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003910297612397333,
+      "loss": 3.0798,
+      "step": 7672
+    },
+    {
+      "epoch": 0.3230736842105263,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0003910018975218949,
+      "loss": 3.1546,
+      "step": 7673
+    },
+    {
+      "epoch": 0.3231157894736842,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00039097403123516765,
+      "loss": 2.8845,
+      "step": 7674
+    },
+    {
+      "epoch": 0.3231578947368421,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003909461623800592,
+      "loss": 3.659,
+      "step": 7675
+    },
+    {
+      "epoch": 0.3232,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00039091829095707725,
+      "loss": 3.4917,
+      "step": 7676
+    },
+    {
+      "epoch": 0.3232421052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00039089041696672966,
+      "loss": 2.8891,
+      "step": 7677
+    },
+    {
+      "epoch": 0.3232842105263158,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00039086254040952416,
+      "loss": 2.9514,
+      "step": 7678
+    },
+    {
+      "epoch": 0.3233263157894737,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0003908346612859687,
+      "loss": 3.2723,
+      "step": 7679
+    },
+    {
+      "epoch": 0.3233684210526316,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00039080677959657116,
+      "loss": 3.4022,
+      "step": 7680
+    },
+    {
+      "epoch": 0.3234105263157895,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003907788953418395,
+      "loss": 3.0917,
+      "step": 7681
+    },
+    {
+      "epoch": 0.3234526315789474,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0003907510085222819,
+      "loss": 3.415,
+      "step": 7682
+    },
+    {
+      "epoch": 0.3234947368421053,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00039072311913840627,
+      "loss": 3.4375,
+      "step": 7683
+    },
+    {
+      "epoch": 0.3235368421052632,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0003906952271907207,
+      "loss": 2.8949,
+      "step": 7684
+    },
+    {
+      "epoch": 0.3235789473684211,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00039066733267973335,
+      "loss": 3.3628,
+      "step": 7685
+    },
+    {
+      "epoch": 0.3236210526315789,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003906394356059526,
+      "loss": 3.5041,
+      "step": 7686
+    },
+    {
+      "epoch": 0.3236631578947368,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00039061153596988655,
+      "loss": 3.4094,
+      "step": 7687
+    },
+    {
+      "epoch": 0.3237052631578947,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00039058363377204343,
+      "loss": 3.0015,
+      "step": 7688
+    },
+    {
+      "epoch": 0.3237473684210526,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003905557290129318,
+      "loss": 3.2188,
+      "step": 7689
+    },
+    {
+      "epoch": 0.3237894736842105,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0003905278216930599,
+      "loss": 2.8507,
+      "step": 7690
+    },
+    {
+      "epoch": 0.3238315789473684,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003904999118129361,
+      "loss": 3.6317,
+      "step": 7691
+    },
+    {
+      "epoch": 0.3238736842105263,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003904719993730691,
+      "loss": 3.0733,
+      "step": 7692
+    },
+    {
+      "epoch": 0.3239157894736842,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003904440843739673,
+      "loss": 3.3828,
+      "step": 7693
+    },
+    {
+      "epoch": 0.3239578947368421,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003904161668161393,
+      "loss": 3.2669,
+      "step": 7694
+    },
+    {
+      "epoch": 0.324,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0003903882467000937,
+      "loss": 3.0747,
+      "step": 7695
+    },
+    {
+      "epoch": 0.3240421052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00039036032402633926,
+      "loss": 2.8318,
+      "step": 7696
+    },
+    {
+      "epoch": 0.3240842105263158,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00039033239879538454,
+      "loss": 3.4097,
+      "step": 7697
+    },
+    {
+      "epoch": 0.3241263157894737,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00039030447100773846,
+      "loss": 3.3199,
+      "step": 7698
+    },
+    {
+      "epoch": 0.3241684210526316,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0003902765406639097,
+      "loss": 3.1288,
+      "step": 7699
+    },
+    {
+      "epoch": 0.32421052631578945,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00039024860776440726,
+      "loss": 3.034,
+      "step": 7700
+    },
+    {
+      "epoch": 0.32425263157894735,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003902206723097399,
+      "loss": 3.1077,
+      "step": 7701
+    },
+    {
+      "epoch": 0.32429473684210525,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003901927343004167,
+      "loss": 3.368,
+      "step": 7702
+    },
+    {
+      "epoch": 0.32433684210526315,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003901647937369466,
+      "loss": 3.3997,
+      "step": 7703
+    },
+    {
+      "epoch": 0.32437894736842104,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00039013685061983864,
+      "loss": 3.368,
+      "step": 7704
+    },
+    {
+      "epoch": 0.32442105263157894,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0003901089049496019,
+      "loss": 3.0184,
+      "step": 7705
+    },
+    {
+      "epoch": 0.32446315789473684,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00039008095672674547,
+      "loss": 3.3627,
+      "step": 7706
+    },
+    {
+      "epoch": 0.32450526315789474,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00039005300595177874,
+      "loss": 3.0908,
+      "step": 7707
+    },
+    {
+      "epoch": 0.32454736842105264,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00039002505262521073,
+      "loss": 2.915,
+      "step": 7708
+    },
+    {
+      "epoch": 0.32458947368421054,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0003899970967475508,
+      "loss": 3.5016,
+      "step": 7709
+    },
+    {
+      "epoch": 0.32463157894736844,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00038996913831930826,
+      "loss": 2.8844,
+      "step": 7710
+    },
+    {
+      "epoch": 0.32467368421052634,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00038994117734099246,
+      "loss": 3.3524,
+      "step": 7711
+    },
+    {
+      "epoch": 0.32471578947368424,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00038991321381311276,
+      "loss": 3.0339,
+      "step": 7712
+    },
+    {
+      "epoch": 0.3247578947368421,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003898852477361787,
+      "loss": 3.4029,
+      "step": 7713
+    },
+    {
+      "epoch": 0.3248,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003898572791106999,
+      "loss": 3.4315,
+      "step": 7714
+    },
+    {
+      "epoch": 0.3248421052631579,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00038982930793718574,
+      "loss": 3.4829,
+      "step": 7715
+    },
+    {
+      "epoch": 0.3248842105263158,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003898013342161459,
+      "loss": 2.808,
+      "step": 7716
+    },
+    {
+      "epoch": 0.3249263157894737,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00038977335794809,
+      "loss": 3.1473,
+      "step": 7717
+    },
+    {
+      "epoch": 0.3249684210526316,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00038974537913352773,
+      "loss": 3.7466,
+      "step": 7718
+    },
+    {
+      "epoch": 0.32501052631578947,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003897173977729689,
+      "loss": 3.0478,
+      "step": 7719
+    },
+    {
+      "epoch": 0.32505263157894737,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00038968941386692313,
+      "loss": 2.9991,
+      "step": 7720
+    },
+    {
+      "epoch": 0.32509473684210527,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003896614274159005,
+      "loss": 3.4582,
+      "step": 7721
+    },
+    {
+      "epoch": 0.32513684210526317,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00038963343842041067,
+      "loss": 3.3717,
+      "step": 7722
+    },
+    {
+      "epoch": 0.32517894736842107,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003896054468809637,
+      "loss": 3.4539,
+      "step": 7723
+    },
+    {
+      "epoch": 0.32522105263157897,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00038957745279806955,
+      "loss": 3.4444,
+      "step": 7724
+    },
+    {
+      "epoch": 0.32526315789473687,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00038954945617223816,
+      "loss": 3.6441,
+      "step": 7725
+    },
+    {
+      "epoch": 0.32530526315789476,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00038952145700397957,
+      "loss": 3.6744,
+      "step": 7726
+    },
+    {
+      "epoch": 0.3253473684210526,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003894934552938041,
+      "loss": 3.1167,
+      "step": 7727
+    },
+    {
+      "epoch": 0.3253894736842105,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00038946545104222166,
+      "loss": 3.1487,
+      "step": 7728
+    },
+    {
+      "epoch": 0.3254315789473684,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00038943744424974264,
+      "loss": 3.4988,
+      "step": 7729
+    },
+    {
+      "epoch": 0.3254736842105263,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003894094349168772,
+      "loss": 3.2542,
+      "step": 7730
+    },
+    {
+      "epoch": 0.3255157894736842,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00038938142304413564,
+      "loss": 3.2545,
+      "step": 7731
+    },
+    {
+      "epoch": 0.3255578947368421,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00038935340863202833,
+      "loss": 2.8935,
+      "step": 7732
+    },
+    {
+      "epoch": 0.3256,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0003893253916810656,
+      "loss": 2.9665,
+      "step": 7733
+    },
+    {
+      "epoch": 0.3256421052631579,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.000389297372191758,
+      "loss": 3.6583,
+      "step": 7734
+    },
+    {
+      "epoch": 0.3256842105263158,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.000389269350164616,
+      "loss": 3.5827,
+      "step": 7735
+    },
+    {
+      "epoch": 0.3257263157894737,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003892413256001499,
+      "loss": 3.5032,
+      "step": 7736
+    },
+    {
+      "epoch": 0.3257684210526316,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003892132984988706,
+      "loss": 3.0255,
+      "step": 7737
+    },
+    {
+      "epoch": 0.3258105263157895,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00038918526886128855,
+      "loss": 2.721,
+      "step": 7738
+    },
+    {
+      "epoch": 0.3258526315789474,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003891572366879144,
+      "loss": 3.2272,
+      "step": 7739
+    },
+    {
+      "epoch": 0.32589473684210524,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00038912920197925883,
+      "loss": 3.1767,
+      "step": 7740
+    },
+    {
+      "epoch": 0.32593684210526314,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003891011647358328,
+      "loss": 3.3291,
+      "step": 7741
+    },
+    {
+      "epoch": 0.32597894736842103,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00038907312495814695,
+      "loss": 3.6203,
+      "step": 7742
+    },
+    {
+      "epoch": 0.32602105263157893,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00038904508264671215,
+      "loss": 3.0731,
+      "step": 7743
+    },
+    {
+      "epoch": 0.32606315789473683,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003890170378020393,
+      "loss": 3.066,
+      "step": 7744
+    },
+    {
+      "epoch": 0.32610526315789473,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00038898899042463943,
+      "loss": 3.0193,
+      "step": 7745
+    },
+    {
+      "epoch": 0.32614736842105263,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00038896094051502333,
+      "loss": 3.1845,
+      "step": 7746
+    },
+    {
+      "epoch": 0.32618947368421053,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00038893288807370223,
+      "loss": 3.029,
+      "step": 7747
+    },
+    {
+      "epoch": 0.32623157894736843,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0003889048331011872,
+      "loss": 3.0007,
+      "step": 7748
+    },
+    {
+      "epoch": 0.3262736842105263,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0003888767755979893,
+      "loss": 3.3586,
+      "step": 7749
+    },
+    {
+      "epoch": 0.3263157894736842,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00038884871556461967,
+      "loss": 3.8711,
+      "step": 7750
+    },
+    {
+      "epoch": 0.3263578947368421,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00038882065300158966,
+      "loss": 3.376,
+      "step": 7751
+    },
+    {
+      "epoch": 0.3264,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0003887925879094104,
+      "loss": 2.9198,
+      "step": 7752
+    },
+    {
+      "epoch": 0.3264421052631579,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00038876452028859323,
+      "loss": 3.5429,
+      "step": 7753
+    },
+    {
+      "epoch": 0.32648421052631577,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003887364501396496,
+      "loss": 3.2168,
+      "step": 7754
+    },
+    {
+      "epoch": 0.32652631578947366,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003887083774630908,
+      "loss": 3.1433,
+      "step": 7755
+    },
+    {
+      "epoch": 0.32656842105263156,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0003886803022594284,
+      "loss": 3.4417,
+      "step": 7756
+    },
+    {
+      "epoch": 0.32661052631578946,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003886522245291738,
+      "loss": 3.4743,
+      "step": 7757
+    },
+    {
+      "epoch": 0.32665263157894736,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00038862414427283864,
+      "loss": 3.2501,
+      "step": 7758
+    },
+    {
+      "epoch": 0.32669473684210526,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003885960614909344,
+      "loss": 3.3404,
+      "step": 7759
+    },
+    {
+      "epoch": 0.32673684210526316,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003885679761839728,
+      "loss": 3.5173,
+      "step": 7760
+    },
+    {
+      "epoch": 0.32677894736842106,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0003885398883524654,
+      "loss": 3.2097,
+      "step": 7761
+    },
+    {
+      "epoch": 0.32682105263157896,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0003885117979969241,
+      "loss": 3.3873,
+      "step": 7762
+    },
+    {
+      "epoch": 0.32686315789473686,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003884837051178606,
+      "loss": 3.3623,
+      "step": 7763
+    },
+    {
+      "epoch": 0.32690526315789475,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0003884556097157866,
+      "loss": 3.2641,
+      "step": 7764
+    },
+    {
+      "epoch": 0.32694736842105265,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00038842751179121417,
+      "loss": 3.0428,
+      "step": 7765
+    },
+    {
+      "epoch": 0.32698947368421055,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.000388399411344655,
+      "loss": 3.6526,
+      "step": 7766
+    },
+    {
+      "epoch": 0.3270315789473684,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00038837130837662126,
+      "loss": 3.047,
+      "step": 7767
+    },
+    {
+      "epoch": 0.3270736842105263,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0003883432028876248,
+      "loss": 3.466,
+      "step": 7768
+    },
+    {
+      "epoch": 0.3271157894736842,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0003883150948781778,
+      "loss": 3.0885,
+      "step": 7769
+    },
+    {
+      "epoch": 0.3271578947368421,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00038828698434879226,
+      "loss": 3.2987,
+      "step": 7770
+    },
+    {
+      "epoch": 0.3272,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0003882588712999804,
+      "loss": 3.1211,
+      "step": 7771
+    },
+    {
+      "epoch": 0.3272421052631579,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0003882307557322543,
+      "loss": 3.4063,
+      "step": 7772
+    },
+    {
+      "epoch": 0.3272842105263158,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0003882026376461262,
+      "loss": 3.3777,
+      "step": 7773
+    },
+    {
+      "epoch": 0.3273263157894737,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003881745170421085,
+      "loss": 3.039,
+      "step": 7774
+    },
+    {
+      "epoch": 0.3273684210526316,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0003881463939207134,
+      "loss": 3.3054,
+      "step": 7775
+    },
+    {
+      "epoch": 0.3274105263157895,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00038811826828245334,
+      "loss": 2.9524,
+      "step": 7776
+    },
+    {
+      "epoch": 0.3274526315789474,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00038809014012784065,
+      "loss": 3.3036,
+      "step": 7777
+    },
+    {
+      "epoch": 0.3274947368421053,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00038806200945738794,
+      "loss": 3.5261,
+      "step": 7778
+    },
+    {
+      "epoch": 0.3275368421052632,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00038803387627160766,
+      "loss": 3.4628,
+      "step": 7779
+    },
+    {
+      "epoch": 0.327578947368421,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00038800574057101225,
+      "loss": 3.4118,
+      "step": 7780
+    },
+    {
+      "epoch": 0.3276210526315789,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00038797760235611433,
+      "loss": 3.1013,
+      "step": 7781
+    },
+    {
+      "epoch": 0.3276631578947368,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003879494616274268,
+      "loss": 3.0355,
+      "step": 7782
+    },
+    {
+      "epoch": 0.3277052631578947,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0003879213183854621,
+      "loss": 3.2491,
+      "step": 7783
+    },
+    {
+      "epoch": 0.3277473684210526,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.000387893172630733,
+      "loss": 2.858,
+      "step": 7784
+    },
+    {
+      "epoch": 0.3277894736842105,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00038786502436375227,
+      "loss": 3.2516,
+      "step": 7785
+    },
+    {
+      "epoch": 0.3278315789473684,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00038783687358503284,
+      "loss": 3.4861,
+      "step": 7786
+    },
+    {
+      "epoch": 0.3278736842105263,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00038780872029508753,
+      "loss": 3.0299,
+      "step": 7787
+    },
+    {
+      "epoch": 0.3279157894736842,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00038778056449442923,
+      "loss": 3.409,
+      "step": 7788
+    },
+    {
+      "epoch": 0.3279578947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003877524061835709,
+      "loss": 3.2539,
+      "step": 7789
+    },
+    {
+      "epoch": 0.328,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0003877242453630256,
+      "loss": 3.2238,
+      "step": 7790
+    },
+    {
+      "epoch": 0.3280421052631579,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0003876960820333064,
+      "loss": 3.291,
+      "step": 7791
+    },
+    {
+      "epoch": 0.3280842105263158,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0003876679161949264,
+      "loss": 3.3712,
+      "step": 7792
+    },
+    {
+      "epoch": 0.3281263157894737,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00038763974784839867,
+      "loss": 3.2013,
+      "step": 7793
+    },
+    {
+      "epoch": 0.32816842105263155,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00038761157699423646,
+      "loss": 3.6054,
+      "step": 7794
+    },
+    {
+      "epoch": 0.32821052631578945,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.000387583403632953,
+      "loss": 3.2329,
+      "step": 7795
+    },
+    {
+      "epoch": 0.32825263157894735,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00038755522776506157,
+      "loss": 3.1909,
+      "step": 7796
+    },
+    {
+      "epoch": 0.32829473684210525,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0003875270493910755,
+      "loss": 2.8601,
+      "step": 7797
+    },
+    {
+      "epoch": 0.32833684210526315,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00038749886851150827,
+      "loss": 3.0517,
+      "step": 7798
+    },
+    {
+      "epoch": 0.32837894736842105,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00038747068512687315,
+      "loss": 2.7676,
+      "step": 7799
+    },
+    {
+      "epoch": 0.32842105263157895,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00038744249923768363,
+      "loss": 2.859,
+      "step": 7800
+    },
+    {
+      "epoch": 0.32846315789473685,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00038741431084445335,
+      "loss": 3.3355,
+      "step": 7801
+    },
+    {
+      "epoch": 0.32850526315789474,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003873861199476957,
+      "loss": 3.6095,
+      "step": 7802
+    },
+    {
+      "epoch": 0.32854736842105264,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0003873579265479244,
+      "loss": 3.6201,
+      "step": 7803
+    },
+    {
+      "epoch": 0.32858947368421054,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0003873297306456531,
+      "loss": 3.3249,
+      "step": 7804
+    },
+    {
+      "epoch": 0.32863157894736844,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00038730153224139543,
+      "loss": 3.1805,
+      "step": 7805
+    },
+    {
+      "epoch": 0.32867368421052634,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003872733313356651,
+      "loss": 3.3706,
+      "step": 7806
+    },
+    {
+      "epoch": 0.3287157894736842,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00038724512792897604,
+      "loss": 3.2081,
+      "step": 7807
+    },
+    {
+      "epoch": 0.3287578947368421,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.000387216922021842,
+      "loss": 3.0704,
+      "step": 7808
+    },
+    {
+      "epoch": 0.3288,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00038718871361477687,
+      "loss": 3.2854,
+      "step": 7809
+    },
+    {
+      "epoch": 0.3288421052631579,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0003871605027082945,
+      "loss": 3.2346,
+      "step": 7810
+    },
+    {
+      "epoch": 0.3288842105263158,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.000387132289302909,
+      "loss": 3.0638,
+      "step": 7811
+    },
+    {
+      "epoch": 0.3289263157894737,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00038710407339913436,
+      "loss": 3.1617,
+      "step": 7812
+    },
+    {
+      "epoch": 0.3289684210526316,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00038707585499748445,
+      "loss": 3.3601,
+      "step": 7813
+    },
+    {
+      "epoch": 0.3290105263157895,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00038704763409847366,
+      "loss": 2.9455,
+      "step": 7814
+    },
+    {
+      "epoch": 0.3290526315789474,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00038701941070261585,
+      "loss": 3.2572,
+      "step": 7815
+    },
+    {
+      "epoch": 0.3290947368421053,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003869911848104255,
+      "loss": 2.8311,
+      "step": 7816
+    },
+    {
+      "epoch": 0.32913684210526317,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003869629564224167,
+      "loss": 3.1114,
+      "step": 7817
+    },
+    {
+      "epoch": 0.32917894736842107,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003869347255391037,
+      "loss": 3.3113,
+      "step": 7818
+    },
+    {
+      "epoch": 0.32922105263157897,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00038690649216100093,
+      "loss": 3.0194,
+      "step": 7819
+    },
+    {
+      "epoch": 0.32926315789473687,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0003868782562886227,
+      "loss": 3.262,
+      "step": 7820
+    },
+    {
+      "epoch": 0.3293052631578947,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00038685001792248354,
+      "loss": 3.1593,
+      "step": 7821
+    },
+    {
+      "epoch": 0.3293473684210526,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003868217770630978,
+      "loss": 3.2642,
+      "step": 7822
+    },
+    {
+      "epoch": 0.3293894736842105,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00038679353371098005,
+      "loss": 3.112,
+      "step": 7823
+    },
+    {
+      "epoch": 0.3294315789473684,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00038676528786664476,
+      "loss": 3.4191,
+      "step": 7824
+    },
+    {
+      "epoch": 0.3294736842105263,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00038673703953060677,
+      "loss": 3.7351,
+      "step": 7825
+    },
+    {
+      "epoch": 0.3295157894736842,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003867087887033804,
+      "loss": 2.8767,
+      "step": 7826
+    },
+    {
+      "epoch": 0.3295578947368421,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00038668053538548066,
+      "loss": 3.0665,
+      "step": 7827
+    },
+    {
+      "epoch": 0.3296,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003866522795774221,
+      "loss": 3.1165,
+      "step": 7828
+    },
+    {
+      "epoch": 0.3296421052631579,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00038662402127971965,
+      "loss": 3.0275,
+      "step": 7829
+    },
+    {
+      "epoch": 0.3296842105263158,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.000386595760492888,
+      "loss": 2.9383,
+      "step": 7830
+    },
+    {
+      "epoch": 0.3297263157894737,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00038656749721744207,
+      "loss": 3.2514,
+      "step": 7831
+    },
+    {
+      "epoch": 0.3297684210526316,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0003865392314538968,
+      "loss": 3.1872,
+      "step": 7832
+    },
+    {
+      "epoch": 0.3298105263157895,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003865109632027671,
+      "loss": 3.6949,
+      "step": 7833
+    },
+    {
+      "epoch": 0.32985263157894734,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00038648269246456817,
+      "loss": 3.4407,
+      "step": 7834
+    },
+    {
+      "epoch": 0.32989473684210524,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00038645441923981495,
+      "loss": 3.2819,
+      "step": 7835
+    },
+    {
+      "epoch": 0.32993684210526314,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003864261435290224,
+      "loss": 3.0382,
+      "step": 7836
+    },
+    {
+      "epoch": 0.32997894736842104,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00038639786533270596,
+      "loss": 3.3755,
+      "step": 7837
+    },
+    {
+      "epoch": 0.33002105263157894,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003863695846513806,
+      "loss": 3.015,
+      "step": 7838
+    },
+    {
+      "epoch": 0.33006315789473684,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0003863413014855617,
+      "loss": 3.1719,
+      "step": 7839
+    },
+    {
+      "epoch": 0.33010526315789473,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003863130158357644,
+      "loss": 3.3593,
+      "step": 7840
+    },
+    {
+      "epoch": 0.33014736842105263,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00038628472770250413,
+      "loss": 3.3057,
+      "step": 7841
+    },
+    {
+      "epoch": 0.33018947368421053,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003862564370862962,
+      "loss": 3.3171,
+      "step": 7842
+    },
+    {
+      "epoch": 0.33023157894736843,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003862281439876562,
+      "loss": 3.0689,
+      "step": 7843
+    },
+    {
+      "epoch": 0.33027368421052633,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003861998484070994,
+      "loss": 3.2629,
+      "step": 7844
+    },
+    {
+      "epoch": 0.33031578947368423,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003861715503451414,
+      "loss": 3.1689,
+      "step": 7845
+    },
+    {
+      "epoch": 0.33035789473684213,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00038614324980229775,
+      "loss": 3.2722,
+      "step": 7846
+    },
+    {
+      "epoch": 0.3304,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00038611494677908396,
+      "loss": 3.3857,
+      "step": 7847
+    },
+    {
+      "epoch": 0.33044210526315787,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003860866412760159,
+      "loss": 3.5801,
+      "step": 7848
+    },
+    {
+      "epoch": 0.33048421052631577,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003860583332936091,
+      "loss": 2.9785,
+      "step": 7849
+    },
+    {
+      "epoch": 0.33052631578947367,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0003860300228323793,
+      "loss": 3.1409,
+      "step": 7850
+    },
+    {
+      "epoch": 0.33056842105263157,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003860017098928424,
+      "loss": 3.2595,
+      "step": 7851
+    },
+    {
+      "epoch": 0.33061052631578947,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00038597339447551406,
+      "loss": 3.5387,
+      "step": 7852
+    },
+    {
+      "epoch": 0.33065263157894736,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003859450765809103,
+      "loss": 3.525,
+      "step": 7853
+    },
+    {
+      "epoch": 0.33069473684210526,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003859167562095468,
+      "loss": 3.3199,
+      "step": 7854
+    },
+    {
+      "epoch": 0.33073684210526316,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003858884333619399,
+      "loss": 3.624,
+      "step": 7855
+    },
+    {
+      "epoch": 0.33077894736842106,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00038586010803860527,
+      "loss": 3.1144,
+      "step": 7856
+    },
+    {
+      "epoch": 0.33082105263157896,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00038583178024005914,
+      "loss": 3.1542,
+      "step": 7857
+    },
+    {
+      "epoch": 0.33086315789473686,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00038580344996681765,
+      "loss": 3.1113,
+      "step": 7858
+    },
+    {
+      "epoch": 0.33090526315789476,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003857751172193968,
+      "loss": 3.1701,
+      "step": 7859
+    },
+    {
+      "epoch": 0.33094736842105266,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003857467819983128,
+      "loss": 3.322,
+      "step": 7860
+    },
+    {
+      "epoch": 0.3309894736842105,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.000385718444304082,
+      "loss": 3.1112,
+      "step": 7861
+    },
+    {
+      "epoch": 0.3310315789473684,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0003856901041372206,
+      "loss": 3.3565,
+      "step": 7862
+    },
+    {
+      "epoch": 0.3310736842105263,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00038566176149824487,
+      "loss": 3.5305,
+      "step": 7863
+    },
+    {
+      "epoch": 0.3311157894736842,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00038563341638767133,
+      "loss": 3.2326,
+      "step": 7864
+    },
+    {
+      "epoch": 0.3311578947368421,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003856050688060162,
+      "loss": 3.0828,
+      "step": 7865
+    },
+    {
+      "epoch": 0.3312,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00038557671875379617,
+      "loss": 3.2994,
+      "step": 7866
+    },
+    {
+      "epoch": 0.3312421052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003855483662315275,
+      "loss": 3.2313,
+      "step": 7867
+    },
+    {
+      "epoch": 0.3312842105263158,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003855200112397269,
+      "loss": 3.2541,
+      "step": 7868
+    },
+    {
+      "epoch": 0.3313263157894737,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00038549165377891096,
+      "loss": 3.1876,
+      "step": 7869
+    },
+    {
+      "epoch": 0.3313684210526316,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003854632938495963,
+      "loss": 3.2066,
+      "step": 7870
+    },
+    {
+      "epoch": 0.3314105263157895,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00038543493145229955,
+      "loss": 3.2407,
+      "step": 7871
+    },
+    {
+      "epoch": 0.3314526315789474,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003854065665875375,
+      "loss": 3.6896,
+      "step": 7872
+    },
+    {
+      "epoch": 0.3314947368421053,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00038537819925582687,
+      "loss": 3.7763,
+      "step": 7873
+    },
+    {
+      "epoch": 0.3315368421052632,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003853498294576845,
+      "loss": 3.0685,
+      "step": 7874
+    },
+    {
+      "epoch": 0.33157894736842103,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003853214571936273,
+      "loss": 3.2299,
+      "step": 7875
+    },
+    {
+      "epoch": 0.3316210526315789,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00038529308246417214,
+      "loss": 3.1194,
+      "step": 7876
+    },
+    {
+      "epoch": 0.3316631578947368,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.000385264705269836,
+      "loss": 3.1279,
+      "step": 7877
+    },
+    {
+      "epoch": 0.3317052631578947,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003852363256111359,
+      "loss": 3.0307,
+      "step": 7878
+    },
+    {
+      "epoch": 0.3317473684210526,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00038520794348858873,
+      "loss": 2.9669,
+      "step": 7879
+    },
+    {
+      "epoch": 0.3317894736842105,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00038517955890271174,
+      "loss": 3.2877,
+      "step": 7880
+    },
+    {
+      "epoch": 0.3318315789473684,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.000385151171854022,
+      "loss": 2.7874,
+      "step": 7881
+    },
+    {
+      "epoch": 0.3318736842105263,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0003851227823430367,
+      "loss": 3.2639,
+      "step": 7882
+    },
+    {
+      "epoch": 0.3319157894736842,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00038509439037027316,
+      "loss": 3.5954,
+      "step": 7883
+    },
+    {
+      "epoch": 0.3319578947368421,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0003850659959362485,
+      "loss": 3.3279,
+      "step": 7884
+    },
+    {
+      "epoch": 0.332,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00038503759904148007,
+      "loss": 3.2124,
+      "step": 7885
+    },
+    {
+      "epoch": 0.3320421052631579,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003850091996864853,
+      "loss": 2.8286,
+      "step": 7886
+    },
+    {
+      "epoch": 0.3320842105263158,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00038498079787178144,
+      "loss": 2.8199,
+      "step": 7887
+    },
+    {
+      "epoch": 0.33212631578947366,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00038495239359788615,
+      "loss": 2.9207,
+      "step": 7888
+    },
+    {
+      "epoch": 0.33216842105263156,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003849239868653167,
+      "loss": 3.2967,
+      "step": 7889
+    },
+    {
+      "epoch": 0.33221052631578946,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00038489557767459084,
+      "loss": 3.1647,
+      "step": 7890
+    },
+    {
+      "epoch": 0.33225263157894735,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00038486716602622597,
+      "loss": 3.1773,
+      "step": 7891
+    },
+    {
+      "epoch": 0.33229473684210525,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00038483875192073984,
+      "loss": 3.3369,
+      "step": 7892
+    },
+    {
+      "epoch": 0.33233684210526315,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00038481033535865005,
+      "loss": 3.281,
+      "step": 7893
+    },
+    {
+      "epoch": 0.33237894736842105,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00038478191634047437,
+      "loss": 2.9946,
+      "step": 7894
+    },
+    {
+      "epoch": 0.33242105263157895,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0003847534948667306,
+      "loss": 3.1804,
+      "step": 7895
+    },
+    {
+      "epoch": 0.33246315789473685,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003847250709379364,
+      "loss": 3.1189,
+      "step": 7896
+    },
+    {
+      "epoch": 0.33250526315789475,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00038469664455460974,
+      "loss": 2.8684,
+      "step": 7897
+    },
+    {
+      "epoch": 0.33254736842105265,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003846682157172684,
+      "loss": 3.0656,
+      "step": 7898
+    },
+    {
+      "epoch": 0.33258947368421055,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003846397844264305,
+      "loss": 3.344,
+      "step": 7899
+    },
+    {
+      "epoch": 0.33263157894736844,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003846113506826139,
+      "loss": 3.1041,
+      "step": 7900
+    },
+    {
+      "epoch": 0.3326736842105263,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0003845829144863366,
+      "loss": 3.076,
+      "step": 7901
+    },
+    {
+      "epoch": 0.3327157894736842,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00038455447583811674,
+      "loss": 3.3739,
+      "step": 7902
+    },
+    {
+      "epoch": 0.3327578947368421,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00038452603473847246,
+      "loss": 2.945,
+      "step": 7903
+    },
+    {
+      "epoch": 0.3328,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00038449759118792194,
+      "loss": 3.4231,
+      "step": 7904
+    },
+    {
+      "epoch": 0.3328421052631579,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003844691451869832,
+      "loss": 3.5847,
+      "step": 7905
+    },
+    {
+      "epoch": 0.3328842105263158,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003844406967361748,
+      "loss": 3.1225,
+      "step": 7906
+    },
+    {
+      "epoch": 0.3329263157894737,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0003844122458360146,
+      "loss": 2.8601,
+      "step": 7907
+    },
+    {
+      "epoch": 0.3329684210526316,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003843837924870215,
+      "loss": 3.444,
+      "step": 7908
+    },
+    {
+      "epoch": 0.3330105263157895,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00038435533668971345,
+      "loss": 2.9577,
+      "step": 7909
+    },
+    {
+      "epoch": 0.3330526315789474,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00038432687844460904,
+      "loss": 3.0916,
+      "step": 7910
+    },
+    {
+      "epoch": 0.3330947368421053,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00038429841775222673,
+      "loss": 3.3113,
+      "step": 7911
+    },
+    {
+      "epoch": 0.3331368421052632,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003842699546130851,
+      "loss": 3.1777,
+      "step": 7912
+    },
+    {
+      "epoch": 0.3331789473684211,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003842414890277026,
+      "loss": 3.0431,
+      "step": 7913
+    },
+    {
+      "epoch": 0.333221052631579,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00038421302099659785,
+      "loss": 2.9283,
+      "step": 7914
+    },
+    {
+      "epoch": 0.3332631578947368,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00038418455052028965,
+      "loss": 2.7833,
+      "step": 7915
+    },
+    {
+      "epoch": 0.3333052631578947,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00038415607759929654,
+      "loss": 3.1989,
+      "step": 7916
+    },
+    {
+      "epoch": 0.3333473684210526,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00038412760223413735,
+      "loss": 2.9709,
+      "step": 7917
+    },
+    {
+      "epoch": 0.3333894736842105,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00038409912442533093,
+      "loss": 3.1927,
+      "step": 7918
+    },
+    {
+      "epoch": 0.3334315789473684,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00038407064417339586,
+      "loss": 3.1928,
+      "step": 7919
+    },
+    {
+      "epoch": 0.3334736842105263,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0003840421614788513,
+      "loss": 3.7893,
+      "step": 7920
+    },
+    {
+      "epoch": 0.3335157894736842,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00038401367634221595,
+      "loss": 3.0991,
+      "step": 7921
+    },
+    {
+      "epoch": 0.3335578947368421,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003839851887640089,
+      "loss": 3.0931,
+      "step": 7922
+    },
+    {
+      "epoch": 0.3336,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00038395669874474915,
+      "loss": 3.2374,
+      "step": 7923
+    },
+    {
+      "epoch": 0.3336421052631579,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0003839282062849557,
+      "loss": 2.9823,
+      "step": 7924
+    },
+    {
+      "epoch": 0.3336842105263158,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003838997113851478,
+      "loss": 3.1536,
+      "step": 7925
+    },
+    {
+      "epoch": 0.3337263157894737,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00038387121404584446,
+      "loss": 3.3822,
+      "step": 7926
+    },
+    {
+      "epoch": 0.3337684210526316,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003838427142675649,
+      "loss": 3.5139,
+      "step": 7927
+    },
+    {
+      "epoch": 0.33381052631578945,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00038381421205082824,
+      "loss": 3.0201,
+      "step": 7928
+    },
+    {
+      "epoch": 0.33385263157894735,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003837857073961539,
+      "loss": 2.9081,
+      "step": 7929
+    },
+    {
+      "epoch": 0.33389473684210524,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0003837572003040612,
+      "loss": 3.28,
+      "step": 7930
+    },
+    {
+      "epoch": 0.33393684210526314,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003837286907750694,
+      "loss": 3.0508,
+      "step": 7931
+    },
+    {
+      "epoch": 0.33397894736842104,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003837001788096981,
+      "loss": 3.2872,
+      "step": 7932
+    },
+    {
+      "epoch": 0.33402105263157894,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003836716644084666,
+      "loss": 3.4463,
+      "step": 7933
+    },
+    {
+      "epoch": 0.33406315789473684,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00038364314757189435,
+      "loss": 3.0854,
+      "step": 7934
+    },
+    {
+      "epoch": 0.33410526315789474,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003836146283005011,
+      "loss": 3.3409,
+      "step": 7935
+    },
+    {
+      "epoch": 0.33414736842105264,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0003835861065948062,
+      "loss": 3.2731,
+      "step": 7936
+    },
+    {
+      "epoch": 0.33418947368421054,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00038355758245532944,
+      "loss": 3.2119,
+      "step": 7937
+    },
+    {
+      "epoch": 0.33423157894736844,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00038352905588259046,
+      "loss": 3.1405,
+      "step": 7938
+    },
+    {
+      "epoch": 0.33427368421052633,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00038350052687710894,
+      "loss": 3.2394,
+      "step": 7939
+    },
+    {
+      "epoch": 0.33431578947368423,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003834719954394047,
+      "loss": 3.677,
+      "step": 7940
+    },
+    {
+      "epoch": 0.33435789473684213,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003834434615699975,
+      "loss": 3.4682,
+      "step": 7941
+    },
+    {
+      "epoch": 0.3344,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00038341492526940726,
+      "loss": 2.7045,
+      "step": 7942
+    },
+    {
+      "epoch": 0.3344421052631579,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00038338638653815385,
+      "loss": 2.6402,
+      "step": 7943
+    },
+    {
+      "epoch": 0.3344842105263158,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00038335784537675717,
+      "loss": 3.0987,
+      "step": 7944
+    },
+    {
+      "epoch": 0.33452631578947367,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0003833293017857372,
+      "loss": 2.8756,
+      "step": 7945
+    },
+    {
+      "epoch": 0.33456842105263157,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0003833007557656141,
+      "loss": 2.7931,
+      "step": 7946
+    },
+    {
+      "epoch": 0.33461052631578947,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00038327220731690777,
+      "loss": 3.2178,
+      "step": 7947
+    },
+    {
+      "epoch": 0.33465263157894737,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003832436564401384,
+      "loss": 3.0454,
+      "step": 7948
+    },
+    {
+      "epoch": 0.33469473684210527,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00038321510313582617,
+      "loss": 3.3407,
+      "step": 7949
+    },
+    {
+      "epoch": 0.33473684210526317,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00038318654740449135,
+      "loss": 2.9371,
+      "step": 7950
+    },
+    {
+      "epoch": 0.33477894736842106,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0003831579892466541,
+      "loss": 3.4842,
+      "step": 7951
+    },
+    {
+      "epoch": 0.33482105263157896,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00038312942866283473,
+      "loss": 2.8166,
+      "step": 7952
+    },
+    {
+      "epoch": 0.33486315789473686,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003831008656535536,
+      "loss": 3.7003,
+      "step": 7953
+    },
+    {
+      "epoch": 0.33490526315789476,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00038307230021933104,
+      "loss": 3.7203,
+      "step": 7954
+    },
+    {
+      "epoch": 0.3349473684210526,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00038304373236068757,
+      "loss": 3.1975,
+      "step": 7955
+    },
+    {
+      "epoch": 0.3349894736842105,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00038301516207814355,
+      "loss": 3.1756,
+      "step": 7956
+    },
+    {
+      "epoch": 0.3350315789473684,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00038298658937221963,
+      "loss": 3.4375,
+      "step": 7957
+    },
+    {
+      "epoch": 0.3350736842105263,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0003829580142434364,
+      "loss": 3.2679,
+      "step": 7958
+    },
+    {
+      "epoch": 0.3351157894736842,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00038292943669231415,
+      "loss": 3.1619,
+      "step": 7959
+    },
+    {
+      "epoch": 0.3351578947368421,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0003829008567193739,
+      "loss": 3.1673,
+      "step": 7960
+    },
+    {
+      "epoch": 0.3352,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00038287227432513615,
+      "loss": 3.1471,
+      "step": 7961
+    },
+    {
+      "epoch": 0.3352421052631579,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003828436895101217,
+      "loss": 3.4299,
+      "step": 7962
+    },
+    {
+      "epoch": 0.3352842105263158,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003828151022748513,
+      "loss": 2.8468,
+      "step": 7963
+    },
+    {
+      "epoch": 0.3353263157894737,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00038278651261984575,
+      "loss": 3.1144,
+      "step": 7964
+    },
+    {
+      "epoch": 0.3353684210526316,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00038275792054562595,
+      "loss": 3.4049,
+      "step": 7965
+    },
+    {
+      "epoch": 0.3354105263157895,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00038272932605271293,
+      "loss": 3.4588,
+      "step": 7966
+    },
+    {
+      "epoch": 0.3354526315789474,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003827007291416275,
+      "loss": 2.7715,
+      "step": 7967
+    },
+    {
+      "epoch": 0.3354947368421053,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00038267212981289055,
+      "loss": 3.1617,
+      "step": 7968
+    },
+    {
+      "epoch": 0.33553684210526313,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003826435280670235,
+      "loss": 3.3855,
+      "step": 7969
+    },
+    {
+      "epoch": 0.33557894736842103,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0003826149239045471,
+      "loss": 2.8377,
+      "step": 7970
+    },
+    {
+      "epoch": 0.33562105263157893,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003825863173259826,
+      "loss": 3.4018,
+      "step": 7971
+    },
+    {
+      "epoch": 0.33566315789473683,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00038255770833185124,
+      "loss": 3.1767,
+      "step": 7972
+    },
+    {
+      "epoch": 0.33570526315789473,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00038252909692267415,
+      "loss": 3.5168,
+      "step": 7973
+    },
+    {
+      "epoch": 0.3357473684210526,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0003825004830989726,
+      "loss": 2.9353,
+      "step": 7974
+    },
+    {
+      "epoch": 0.3357894736842105,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003824718668612679,
+      "loss": 3.5127,
+      "step": 7975
+    },
+    {
+      "epoch": 0.3358315789473684,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00038244324821008155,
+      "loss": 2.5587,
+      "step": 7976
+    },
+    {
+      "epoch": 0.3358736842105263,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00038241462714593477,
+      "loss": 3.0238,
+      "step": 7977
+    },
+    {
+      "epoch": 0.3359157894736842,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0003823860036693491,
+      "loss": 2.9694,
+      "step": 7978
+    },
+    {
+      "epoch": 0.3359578947368421,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003823573777808459,
+      "loss": 3.1992,
+      "step": 7979
+    },
+    {
+      "epoch": 0.336,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00038232874948094684,
+      "loss": 2.9677,
+      "step": 7980
+    },
+    {
+      "epoch": 0.3360421052631579,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00038230011877017346,
+      "loss": 3.1137,
+      "step": 7981
+    },
+    {
+      "epoch": 0.33608421052631576,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00038227148564904734,
+      "loss": 2.6302,
+      "step": 7982
+    },
+    {
+      "epoch": 0.33612631578947366,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00038224285011809016,
+      "loss": 3.0678,
+      "step": 7983
+    },
+    {
+      "epoch": 0.33616842105263156,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00038221421217782373,
+      "loss": 3.454,
+      "step": 7984
+    },
+    {
+      "epoch": 0.33621052631578946,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00038218557182876956,
+      "loss": 3.6481,
+      "step": 7985
+    },
+    {
+      "epoch": 0.33625263157894736,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003821569290714497,
+      "loss": 2.7855,
+      "step": 7986
+    },
+    {
+      "epoch": 0.33629473684210526,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003821282839063858,
+      "loss": 3.2587,
+      "step": 7987
+    },
+    {
+      "epoch": 0.33633684210526316,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003820996363340997,
+      "loss": 3.5867,
+      "step": 7988
+    },
+    {
+      "epoch": 0.33637894736842106,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0003820709863551136,
+      "loss": 3.2944,
+      "step": 7989
+    },
+    {
+      "epoch": 0.33642105263157895,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00038204233396994927,
+      "loss": 3.0407,
+      "step": 7990
+    },
+    {
+      "epoch": 0.33646315789473685,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0003820136791791288,
+      "loss": 3.3878,
+      "step": 7991
+    },
+    {
+      "epoch": 0.33650526315789475,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00038198502198317407,
+      "loss": 3.6311,
+      "step": 7992
+    },
+    {
+      "epoch": 0.33654736842105265,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00038195636238260746,
+      "loss": 3.1132,
+      "step": 7993
+    },
+    {
+      "epoch": 0.33658947368421055,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00038192770037795095,
+      "loss": 3.2524,
+      "step": 7994
+    },
+    {
+      "epoch": 0.33663157894736845,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00038189903596972665,
+      "loss": 3.2628,
+      "step": 7995
+    },
+    {
+      "epoch": 0.3366736842105263,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.000381870369158457,
+      "loss": 2.9233,
+      "step": 7996
+    },
+    {
+      "epoch": 0.3367157894736842,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.000381841699944664,
+      "loss": 3.0799,
+      "step": 7997
+    },
+    {
+      "epoch": 0.3367578947368421,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00038181302832887034,
+      "loss": 2.9335,
+      "step": 7998
+    },
+    {
+      "epoch": 0.3368,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003817843543115981,
+      "loss": 3.471,
+      "step": 7999
+    },
+    {
+      "epoch": 0.3368421052631579,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00038175567789336975,
+      "loss": 3.5798,
+      "step": 8000
+    },
+    {
+      "epoch": 0.3368842105263158,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00038172699907470777,
+      "loss": 3.1683,
+      "step": 8001
+    },
+    {
+      "epoch": 0.3369263157894737,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00038169831785613456,
+      "loss": 3.1716,
+      "step": 8002
+    },
+    {
+      "epoch": 0.3369684210526316,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0003816696342381728,
+      "loss": 2.9892,
+      "step": 8003
+    },
+    {
+      "epoch": 0.3370105263157895,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.000381640948221345,
+      "loss": 3.5864,
+      "step": 8004
+    },
+    {
+      "epoch": 0.3370526315789474,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003816122598061739,
+      "loss": 3.3006,
+      "step": 8005
+    },
+    {
+      "epoch": 0.3370947368421053,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003815835689931819,
+      "loss": 3.3315,
+      "step": 8006
+    },
+    {
+      "epoch": 0.3371368421052632,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00038155487578289194,
+      "loss": 2.9677,
+      "step": 8007
+    },
+    {
+      "epoch": 0.3371789473684211,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00038152618017582677,
+      "loss": 3.1989,
+      "step": 8008
+    },
+    {
+      "epoch": 0.3372210526315789,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.000381497482172509,
+      "loss": 3.5271,
+      "step": 8009
+    },
+    {
+      "epoch": 0.3372631578947368,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00038146878177346176,
+      "loss": 3.646,
+      "step": 8010
+    },
+    {
+      "epoch": 0.3373052631578947,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00038144007897920776,
+      "loss": 3.4276,
+      "step": 8011
+    },
+    {
+      "epoch": 0.3373473684210526,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003814113737902699,
+      "loss": 3.2396,
+      "step": 8012
+    },
+    {
+      "epoch": 0.3373894736842105,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00038138266620717123,
+      "loss": 3.2712,
+      "step": 8013
+    },
+    {
+      "epoch": 0.3374315789473684,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00038135395623043475,
+      "loss": 3.5348,
+      "step": 8014
+    },
+    {
+      "epoch": 0.3374736842105263,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00038132524386058343,
+      "loss": 3.2445,
+      "step": 8015
+    },
+    {
+      "epoch": 0.3375157894736842,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0003812965290981406,
+      "loss": 3.8382,
+      "step": 8016
+    },
+    {
+      "epoch": 0.3375578947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00038126781194362927,
+      "loss": 3.4179,
+      "step": 8017
+    },
+    {
+      "epoch": 0.3376,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00038123909239757257,
+      "loss": 3.0286,
+      "step": 8018
+    },
+    {
+      "epoch": 0.3376421052631579,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00038121037046049387,
+      "loss": 3.3881,
+      "step": 8019
+    },
+    {
+      "epoch": 0.3376842105263158,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003811816461329164,
+      "loss": 3.3889,
+      "step": 8020
+    },
+    {
+      "epoch": 0.3377263157894737,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00038115291941536345,
+      "loss": 3.3465,
+      "step": 8021
+    },
+    {
+      "epoch": 0.3377684210526316,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003811241903083583,
+      "loss": 3.3145,
+      "step": 8022
+    },
+    {
+      "epoch": 0.33781052631578945,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0003810954588124246,
+      "loss": 3.1014,
+      "step": 8023
+    },
+    {
+      "epoch": 0.33785263157894735,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003810667249280856,
+      "loss": 3.0904,
+      "step": 8024
+    },
+    {
+      "epoch": 0.33789473684210525,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00038103798865586494,
+      "loss": 3.2609,
+      "step": 8025
+    },
+    {
+      "epoch": 0.33793684210526315,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.000381009249996286,
+      "loss": 3.4993,
+      "step": 8026
+    },
+    {
+      "epoch": 0.33797894736842105,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00038098050894987254,
+      "loss": 3.0005,
+      "step": 8027
+    },
+    {
+      "epoch": 0.33802105263157894,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003809517655171481,
+      "loss": 3.5184,
+      "step": 8028
+    },
+    {
+      "epoch": 0.33806315789473684,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0003809230196986362,
+      "loss": 3.2835,
+      "step": 8029
+    },
+    {
+      "epoch": 0.33810526315789474,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0003808942714948609,
+      "loss": 3.1913,
+      "step": 8030
+    },
+    {
+      "epoch": 0.33814736842105264,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00038086552090634565,
+      "loss": 2.9306,
+      "step": 8031
+    },
+    {
+      "epoch": 0.33818947368421054,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003808367679336144,
+      "loss": 3.0143,
+      "step": 8032
+    },
+    {
+      "epoch": 0.33823157894736844,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.000380808012577191,
+      "loss": 3.4699,
+      "step": 8033
+    },
+    {
+      "epoch": 0.33827368421052634,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00038077925483759925,
+      "loss": 3.6256,
+      "step": 8034
+    },
+    {
+      "epoch": 0.33831578947368424,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00038075049471536314,
+      "loss": 3.0378,
+      "step": 8035
+    },
+    {
+      "epoch": 0.3383578947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003807217322110067,
+      "loss": 3.5336,
+      "step": 8036
+    },
+    {
+      "epoch": 0.3384,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003806929673250539,
+      "loss": 3.3073,
+      "step": 8037
+    },
+    {
+      "epoch": 0.3384421052631579,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003806642000580287,
+      "loss": 3.0693,
+      "step": 8038
+    },
+    {
+      "epoch": 0.3384842105263158,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00038063543041045537,
+      "loss": 3.4275,
+      "step": 8039
+    },
+    {
+      "epoch": 0.3385263157894737,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003806066583828579,
+      "loss": 3.2804,
+      "step": 8040
+    },
+    {
+      "epoch": 0.3385684210526316,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00038057788397576064,
+      "loss": 3.4944,
+      "step": 8041
+    },
+    {
+      "epoch": 0.3386105263157895,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0003805491071896878,
+      "loss": 3.4908,
+      "step": 8042
+    },
+    {
+      "epoch": 0.33865263157894737,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0003805203280251636,
+      "loss": 3.0786,
+      "step": 8043
+    },
+    {
+      "epoch": 0.33869473684210527,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0003804915464827123,
+      "loss": 2.5287,
+      "step": 8044
+    },
+    {
+      "epoch": 0.33873684210526317,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0003804627625628584,
+      "loss": 3.5034,
+      "step": 8045
+    },
+    {
+      "epoch": 0.33877894736842107,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003804339762661263,
+      "loss": 3.3386,
+      "step": 8046
+    },
+    {
+      "epoch": 0.33882105263157897,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00038040518759304033,
+      "loss": 2.8087,
+      "step": 8047
+    },
+    {
+      "epoch": 0.33886315789473687,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00038037639654412504,
+      "loss": 2.4735,
+      "step": 8048
+    },
+    {
+      "epoch": 0.3389052631578947,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.000380347603119905,
+      "loss": 3.1707,
+      "step": 8049
+    },
+    {
+      "epoch": 0.3389473684210526,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00038031880732090483,
+      "loss": 3.3938,
+      "step": 8050
+    },
+    {
+      "epoch": 0.3389894736842105,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0003802900091476491,
+      "loss": 3.4361,
+      "step": 8051
+    },
+    {
+      "epoch": 0.3390315789473684,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003802612086006625,
+      "loss": 3.3268,
+      "step": 8052
+    },
+    {
+      "epoch": 0.3390736842105263,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003802324056804698,
+      "loss": 3.0745,
+      "step": 8053
+    },
+    {
+      "epoch": 0.3391157894736842,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003802036003875956,
+      "loss": 2.6664,
+      "step": 8054
+    },
+    {
+      "epoch": 0.3391578947368421,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0003801747927225647,
+      "loss": 3.253,
+      "step": 8055
+    },
+    {
+      "epoch": 0.3392,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0003801459826859022,
+      "loss": 3.0784,
+      "step": 8056
+    },
+    {
+      "epoch": 0.3392421052631579,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00038011717027813265,
+      "loss": 3.2767,
+      "step": 8057
+    },
+    {
+      "epoch": 0.3392842105263158,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00038008835549978125,
+      "loss": 2.9422,
+      "step": 8058
+    },
+    {
+      "epoch": 0.3393263157894737,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003800595383513728,
+      "loss": 3.1485,
+      "step": 8059
+    },
+    {
+      "epoch": 0.3393684210526316,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00038003071883343245,
+      "loss": 3.4343,
+      "step": 8060
+    },
+    {
+      "epoch": 0.3394105263157895,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003800018969464851,
+      "loss": 3.411,
+      "step": 8061
+    },
+    {
+      "epoch": 0.3394526315789474,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00037997307269105595,
+      "loss": 3.0111,
+      "step": 8062
+    },
+    {
+      "epoch": 0.33949473684210524,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00037994424606767015,
+      "loss": 3.5503,
+      "step": 8063
+    },
+    {
+      "epoch": 0.33953684210526314,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00037991541707685286,
+      "loss": 3.5176,
+      "step": 8064
+    },
+    {
+      "epoch": 0.33957894736842104,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00037988658571912925,
+      "loss": 3.3148,
+      "step": 8065
+    },
+    {
+      "epoch": 0.33962105263157893,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003798577519950248,
+      "loss": 3.5518,
+      "step": 8066
+    },
+    {
+      "epoch": 0.33966315789473683,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003798289159050646,
+      "loss": 3.29,
+      "step": 8067
+    },
+    {
+      "epoch": 0.33970526315789473,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003798000774497741,
+      "loss": 3.1019,
+      "step": 8068
+    },
+    {
+      "epoch": 0.33974736842105263,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0003797712366296786,
+      "loss": 3.7446,
+      "step": 8069
+    },
+    {
+      "epoch": 0.33978947368421053,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003797423934453038,
+      "loss": 3.6007,
+      "step": 8070
+    },
+    {
+      "epoch": 0.33983157894736843,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.000379713547897175,
+      "loss": 2.9921,
+      "step": 8071
+    },
+    {
+      "epoch": 0.33987368421052633,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0003796846999858177,
+      "loss": 3.3085,
+      "step": 8072
+    },
+    {
+      "epoch": 0.3399157894736842,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0003796558497117576,
+      "loss": 3.713,
+      "step": 8073
+    },
+    {
+      "epoch": 0.3399578947368421,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003796269970755202,
+      "loss": 3.4899,
+      "step": 8074
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003795981420776313,
+      "loss": 3.4302,
+      "step": 8075
+    },
+    {
+      "epoch": 0.34004210526315787,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003795692847186164,
+      "loss": 3.4794,
+      "step": 8076
+    },
+    {
+      "epoch": 0.34008421052631577,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0003795404249990015,
+      "loss": 2.8114,
+      "step": 8077
+    },
+    {
+      "epoch": 0.34012631578947367,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003795115629193122,
+      "loss": 3.7891,
+      "step": 8078
+    },
+    {
+      "epoch": 0.34016842105263156,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00037948269848007445,
+      "loss": 2.8361,
+      "step": 8079
+    },
+    {
+      "epoch": 0.34021052631578946,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.000379453831681814,
+      "loss": 3.2785,
+      "step": 8080
+    },
+    {
+      "epoch": 0.34025263157894736,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00037942496252505684,
+      "loss": 3.0125,
+      "step": 8081
+    },
+    {
+      "epoch": 0.34029473684210526,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0003793960910103289,
+      "loss": 3.2825,
+      "step": 8082
+    },
+    {
+      "epoch": 0.34033684210526316,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0003793672171381562,
+      "loss": 3.2323,
+      "step": 8083
+    },
+    {
+      "epoch": 0.34037894736842106,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0003793383409090649,
+      "loss": 3.3577,
+      "step": 8084
+    },
+    {
+      "epoch": 0.34042105263157896,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00037930946232358104,
+      "loss": 2.7474,
+      "step": 8085
+    },
+    {
+      "epoch": 0.34046315789473686,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003792805813822307,
+      "loss": 3.2967,
+      "step": 8086
+    },
+    {
+      "epoch": 0.34050526315789476,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00037925169808553994,
+      "loss": 3.592,
+      "step": 8087
+    },
+    {
+      "epoch": 0.34054736842105265,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0003792228124340352,
+      "loss": 3.4521,
+      "step": 8088
+    },
+    {
+      "epoch": 0.34058947368421055,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0003791939244282425,
+      "loss": 2.6262,
+      "step": 8089
+    },
+    {
+      "epoch": 0.3406315789473684,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0003791650340686884,
+      "loss": 3.4331,
+      "step": 8090
+    },
+    {
+      "epoch": 0.3406736842105263,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0003791361413558992,
+      "loss": 3.0454,
+      "step": 8091
+    },
+    {
+      "epoch": 0.3407157894736842,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00037910724629040116,
+      "loss": 3.0273,
+      "step": 8092
+    },
+    {
+      "epoch": 0.3407578947368421,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00037907834887272084,
+      "loss": 3.1323,
+      "step": 8093
+    },
+    {
+      "epoch": 0.3408,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003790494491033847,
+      "loss": 3.2584,
+      "step": 8094
+    },
+    {
+      "epoch": 0.3408421052631579,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0003790205469829191,
+      "loss": 3.1852,
+      "step": 8095
+    },
+    {
+      "epoch": 0.3408842105263158,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00037899164251185076,
+      "loss": 3.5723,
+      "step": 8096
+    },
+    {
+      "epoch": 0.3409263157894737,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00037896273569070627,
+      "loss": 3.3689,
+      "step": 8097
+    },
+    {
+      "epoch": 0.3409684210526316,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003789338265200123,
+      "loss": 3.3757,
+      "step": 8098
+    },
+    {
+      "epoch": 0.3410105263157895,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003789049150002954,
+      "loss": 2.8035,
+      "step": 8099
+    },
+    {
+      "epoch": 0.3410526315789474,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00037887600113208254,
+      "loss": 3.5801,
+      "step": 8100
+    },
+    {
+      "epoch": 0.3410947368421053,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00037884708491590024,
+      "loss": 3.2076,
+      "step": 8101
+    },
+    {
+      "epoch": 0.3411368421052632,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00037881816635227555,
+      "loss": 3.1833,
+      "step": 8102
+    },
+    {
+      "epoch": 0.341178947368421,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00037878924544173517,
+      "loss": 3.5282,
+      "step": 8103
+    },
+    {
+      "epoch": 0.3412210526315789,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00037876032218480603,
+      "loss": 3.2198,
+      "step": 8104
+    },
+    {
+      "epoch": 0.3412631578947368,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00037873139658201516,
+      "loss": 3.7004,
+      "step": 8105
+    },
+    {
+      "epoch": 0.3413052631578947,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003787024686338895,
+      "loss": 3.5955,
+      "step": 8106
+    },
+    {
+      "epoch": 0.3413473684210526,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.000378673538340956,
+      "loss": 3.4038,
+      "step": 8107
+    },
+    {
+      "epoch": 0.3413894736842105,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0003786446057037419,
+      "loss": 3.0353,
+      "step": 8108
+    },
+    {
+      "epoch": 0.3414315789473684,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00037861567072277415,
+      "loss": 3.0447,
+      "step": 8109
+    },
+    {
+      "epoch": 0.3414736842105263,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003785867333985801,
+      "loss": 2.8875,
+      "step": 8110
+    },
+    {
+      "epoch": 0.3415157894736842,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003785577937316868,
+      "loss": 3.1532,
+      "step": 8111
+    },
+    {
+      "epoch": 0.3415578947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00037852885172262157,
+      "loss": 3.4033,
+      "step": 8112
+    },
+    {
+      "epoch": 0.3416,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0003784999073719116,
+      "loss": 3.2983,
+      "step": 8113
+    },
+    {
+      "epoch": 0.3416421052631579,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003784709606800844,
+      "loss": 3.2931,
+      "step": 8114
+    },
+    {
+      "epoch": 0.3416842105263158,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00037844201164766725,
+      "loss": 3.1403,
+      "step": 8115
+    },
+    {
+      "epoch": 0.3417263157894737,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003784130602751875,
+      "loss": 3.439,
+      "step": 8116
+    },
+    {
+      "epoch": 0.34176842105263155,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003783841065631727,
+      "loss": 3.2401,
+      "step": 8117
+    },
+    {
+      "epoch": 0.34181052631578945,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0003783551505121503,
+      "loss": 2.7191,
+      "step": 8118
+    },
+    {
+      "epoch": 0.34185263157894735,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0003783261921226478,
+      "loss": 2.6075,
+      "step": 8119
+    },
+    {
+      "epoch": 0.34189473684210525,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00037829723139519303,
+      "loss": 3.1234,
+      "step": 8120
+    },
+    {
+      "epoch": 0.34193684210526315,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00037826826833031334,
+      "loss": 3.4139,
+      "step": 8121
+    },
+    {
+      "epoch": 0.34197894736842105,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00037823930292853647,
+      "loss": 2.9458,
+      "step": 8122
+    },
+    {
+      "epoch": 0.34202105263157895,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0003782103351903903,
+      "loss": 3.4771,
+      "step": 8123
+    },
+    {
+      "epoch": 0.34206315789473685,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00037818136511640234,
+      "loss": 3.4509,
+      "step": 8124
+    },
+    {
+      "epoch": 0.34210526315789475,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00037815239270710054,
+      "loss": 2.9436,
+      "step": 8125
+    },
+    {
+      "epoch": 0.34214736842105264,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00037812341796301274,
+      "loss": 3.1038,
+      "step": 8126
+    },
+    {
+      "epoch": 0.34218947368421054,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0003780944408846668,
+      "loss": 3.1535,
+      "step": 8127
+    },
+    {
+      "epoch": 0.34223157894736844,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003780654614725907,
+      "loss": 2.9944,
+      "step": 8128
+    },
+    {
+      "epoch": 0.34227368421052634,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00037803647972731234,
+      "loss": 3.2479,
+      "step": 8129
+    },
+    {
+      "epoch": 0.3423157894736842,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00037800749564935964,
+      "loss": 3.2011,
+      "step": 8130
+    },
+    {
+      "epoch": 0.3423578947368421,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003779785092392609,
+      "loss": 3.1058,
+      "step": 8131
+    },
+    {
+      "epoch": 0.3424,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00037794952049754403,
+      "loss": 3.0025,
+      "step": 8132
+    },
+    {
+      "epoch": 0.3424421052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003779205294247373,
+      "loss": 3.4586,
+      "step": 8133
+    },
+    {
+      "epoch": 0.3424842105263158,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003778915360213688,
+      "loss": 3.331,
+      "step": 8134
+    },
+    {
+      "epoch": 0.3425263157894737,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00037786254028796673,
+      "loss": 3.1038,
+      "step": 8135
+    },
+    {
+      "epoch": 0.3425684210526316,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00037783354222505945,
+      "loss": 3.464,
+      "step": 8136
+    },
+    {
+      "epoch": 0.3426105263157895,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00037780454183317525,
+      "loss": 3.7711,
+      "step": 8137
+    },
+    {
+      "epoch": 0.3426526315789474,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00037777553911284243,
+      "loss": 2.8272,
+      "step": 8138
+    },
+    {
+      "epoch": 0.3426947368421053,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0003777465340645894,
+      "loss": 3.0339,
+      "step": 8139
+    },
+    {
+      "epoch": 0.3427368421052632,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00037771752668894464,
+      "loss": 3.0848,
+      "step": 8140
+    },
+    {
+      "epoch": 0.34277894736842107,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00037768851698643656,
+      "loss": 3.2503,
+      "step": 8141
+    },
+    {
+      "epoch": 0.34282105263157897,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00037765950495759383,
+      "loss": 3.013,
+      "step": 8142
+    },
+    {
+      "epoch": 0.34286315789473687,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00037763049060294477,
+      "loss": 3.1282,
+      "step": 8143
+    },
+    {
+      "epoch": 0.3429052631578947,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00037760147392301827,
+      "loss": 3.4879,
+      "step": 8144
+    },
+    {
+      "epoch": 0.3429473684210526,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003775724549183428,
+      "loss": 3.2463,
+      "step": 8145
+    },
+    {
+      "epoch": 0.3429894736842105,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003775434335894471,
+      "loss": 3.3159,
+      "step": 8146
+    },
+    {
+      "epoch": 0.3430315789473684,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00037751440993685987,
+      "loss": 3.196,
+      "step": 8147
+    },
+    {
+      "epoch": 0.3430736842105263,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00037748538396110986,
+      "loss": 3.4593,
+      "step": 8148
+    },
+    {
+      "epoch": 0.3431157894736842,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.000377456355662726,
+      "loss": 3.2124,
+      "step": 8149
+    },
+    {
+      "epoch": 0.3431578947368421,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003774273250422371,
+      "loss": 3.2354,
+      "step": 8150
+    },
+    {
+      "epoch": 0.3432,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00037739829210017205,
+      "loss": 2.7043,
+      "step": 8151
+    },
+    {
+      "epoch": 0.3432421052631579,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003773692568370598,
+      "loss": 3.1422,
+      "step": 8152
+    },
+    {
+      "epoch": 0.3432842105263158,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0003773402192534294,
+      "loss": 3.5105,
+      "step": 8153
+    },
+    {
+      "epoch": 0.3433263157894737,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00037731117934980976,
+      "loss": 3.1692,
+      "step": 8154
+    },
+    {
+      "epoch": 0.3433684210526316,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0003772821371267301,
+      "loss": 3.3828,
+      "step": 8155
+    },
+    {
+      "epoch": 0.3434105263157895,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00037725309258471927,
+      "loss": 3.2335,
+      "step": 8156
+    },
+    {
+      "epoch": 0.34345263157894734,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0003772240457243067,
+      "loss": 3.0447,
+      "step": 8157
+    },
+    {
+      "epoch": 0.34349473684210524,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00037719499654602156,
+      "loss": 3.0578,
+      "step": 8158
+    },
+    {
+      "epoch": 0.34353684210526314,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00037716594505039303,
+      "loss": 3.1863,
+      "step": 8159
+    },
+    {
+      "epoch": 0.34357894736842104,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003771368912379504,
+      "loss": 2.9676,
+      "step": 8160
+    },
+    {
+      "epoch": 0.34362105263157894,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00037710783510922294,
+      "loss": 3.1447,
+      "step": 8161
+    },
+    {
+      "epoch": 0.34366315789473684,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003770787766647401,
+      "loss": 3.7328,
+      "step": 8162
+    },
+    {
+      "epoch": 0.34370526315789474,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0003770497159050312,
+      "loss": 3.2592,
+      "step": 8163
+    },
+    {
+      "epoch": 0.34374736842105263,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0003770206528306258,
+      "loss": 3.2207,
+      "step": 8164
+    },
+    {
+      "epoch": 0.34378947368421053,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00037699158744205336,
+      "loss": 3.0336,
+      "step": 8165
+    },
+    {
+      "epoch": 0.34383157894736843,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00037696251973984334,
+      "loss": 3.2014,
+      "step": 8166
+    },
+    {
+      "epoch": 0.34387368421052633,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0003769334497245255,
+      "loss": 3.1273,
+      "step": 8167
+    },
+    {
+      "epoch": 0.34391578947368423,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00037690437739662926,
+      "loss": 2.9255,
+      "step": 8168
+    },
+    {
+      "epoch": 0.34395789473684213,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00037687530275668435,
+      "loss": 3.6337,
+      "step": 8169
+    },
+    {
+      "epoch": 0.344,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00037684622580522057,
+      "loss": 3.3003,
+      "step": 8170
+    },
+    {
+      "epoch": 0.34404210526315787,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00037681714654276755,
+      "loss": 2.8783,
+      "step": 8171
+    },
+    {
+      "epoch": 0.34408421052631577,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003767880649698552,
+      "loss": 3.4545,
+      "step": 8172
+    },
+    {
+      "epoch": 0.34412631578947367,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00037675898108701316,
+      "loss": 3.3453,
+      "step": 8173
+    },
+    {
+      "epoch": 0.34416842105263157,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003767298948947714,
+      "loss": 3.3267,
+      "step": 8174
+    },
+    {
+      "epoch": 0.34421052631578947,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00037670080639366,
+      "loss": 3.0539,
+      "step": 8175
+    },
+    {
+      "epoch": 0.34425263157894737,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003766717155842087,
+      "loss": 3.6165,
+      "step": 8176
+    },
+    {
+      "epoch": 0.34429473684210526,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00037664262246694756,
+      "loss": 3.092,
+      "step": 8177
+    },
+    {
+      "epoch": 0.34433684210526316,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00037661352704240666,
+      "loss": 3.4342,
+      "step": 8178
+    },
+    {
+      "epoch": 0.34437894736842106,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.000376584429311116,
+      "loss": 3.1952,
+      "step": 8179
+    },
+    {
+      "epoch": 0.34442105263157896,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00037655532927360593,
+      "loss": 3.1271,
+      "step": 8180
+    },
+    {
+      "epoch": 0.34446315789473686,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00037652622693040634,
+      "loss": 3.3971,
+      "step": 8181
+    },
+    {
+      "epoch": 0.34450526315789476,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003764971222820476,
+      "loss": 3.4204,
+      "step": 8182
+    },
+    {
+      "epoch": 0.34454736842105266,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00037646801532905983,
+      "loss": 2.9938,
+      "step": 8183
+    },
+    {
+      "epoch": 0.3445894736842105,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00037643890607197346,
+      "loss": 3.0937,
+      "step": 8184
+    },
+    {
+      "epoch": 0.3446315789473684,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003764097945113188,
+      "loss": 3.317,
+      "step": 8185
+    },
+    {
+      "epoch": 0.3446736842105263,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003763806806476262,
+      "loss": 2.7222,
+      "step": 8186
+    },
+    {
+      "epoch": 0.3447157894736842,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00037635156448142615,
+      "loss": 3.2029,
+      "step": 8187
+    },
+    {
+      "epoch": 0.3447578947368421,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.000376322446013249,
+      "loss": 3.2374,
+      "step": 8188
+    },
+    {
+      "epoch": 0.3448,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003762933252436253,
+      "loss": 3.0058,
+      "step": 8189
+    },
+    {
+      "epoch": 0.3448421052631579,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0003762642021730855,
+      "loss": 3.2653,
+      "step": 8190
+    },
+    {
+      "epoch": 0.3448842105263158,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003762350768021604,
+      "loss": 3.2787,
+      "step": 8191
+    },
+    {
+      "epoch": 0.3449263157894737,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00037620594913138054,
+      "loss": 3.2269,
+      "step": 8192
+    },
+    {
+      "epoch": 0.3449684210526316,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0003761768191612766,
+      "loss": 3.5187,
+      "step": 8193
+    },
+    {
+      "epoch": 0.3450105263157895,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0003761476868923792,
+      "loss": 3.3041,
+      "step": 8194
+    },
+    {
+      "epoch": 0.3450526315789474,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003761185523252192,
+      "loss": 3.3635,
+      "step": 8195
+    },
+    {
+      "epoch": 0.3450947368421053,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00037608941546032735,
+      "loss": 3.1972,
+      "step": 8196
+    },
+    {
+      "epoch": 0.34513684210526313,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00037606027629823446,
+      "loss": 3.2351,
+      "step": 8197
+    },
+    {
+      "epoch": 0.34517894736842103,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003760311348394715,
+      "loss": 3.1247,
+      "step": 8198
+    },
+    {
+      "epoch": 0.34522105263157893,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00037600199108456933,
+      "loss": 3.2359,
+      "step": 8199
+    },
+    {
+      "epoch": 0.3452631578947368,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0003759728450340589,
+      "loss": 3.2204,
+      "step": 8200
+    },
+    {
+      "epoch": 0.3453052631578947,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003759436966884713,
+      "loss": 3.2993,
+      "step": 8201
+    },
+    {
+      "epoch": 0.3453473684210526,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00037591454604833755,
+      "loss": 2.8041,
+      "step": 8202
+    },
+    {
+      "epoch": 0.3453894736842105,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00037588539311418856,
+      "loss": 2.9183,
+      "step": 8203
+    },
+    {
+      "epoch": 0.3454315789473684,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00037585623788655576,
+      "loss": 3.321,
+      "step": 8204
+    },
+    {
+      "epoch": 0.3454736842105263,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00037582708036597014,
+      "loss": 3.1191,
+      "step": 8205
+    },
+    {
+      "epoch": 0.3455157894736842,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00037579792055296294,
+      "loss": 3.0857,
+      "step": 8206
+    },
+    {
+      "epoch": 0.3455578947368421,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00037576875844806543,
+      "loss": 3.4273,
+      "step": 8207
+    },
+    {
+      "epoch": 0.3456,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0003757395940518089,
+      "loss": 2.9021,
+      "step": 8208
+    },
+    {
+      "epoch": 0.3456421052631579,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00037571042736472474,
+      "loss": 2.8855,
+      "step": 8209
+    },
+    {
+      "epoch": 0.3456842105263158,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00037568125838734423,
+      "loss": 3.1106,
+      "step": 8210
+    },
+    {
+      "epoch": 0.34572631578947366,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00037565208712019894,
+      "loss": 3.9363,
+      "step": 8211
+    },
+    {
+      "epoch": 0.34576842105263156,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003756229135638203,
+      "loss": 3.0832,
+      "step": 8212
+    },
+    {
+      "epoch": 0.34581052631578946,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00037559373771873974,
+      "loss": 3.4189,
+      "step": 8213
+    },
+    {
+      "epoch": 0.34585263157894736,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003755645595854888,
+      "loss": 3.1372,
+      "step": 8214
+    },
+    {
+      "epoch": 0.34589473684210525,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003755353791645991,
+      "loss": 3.4186,
+      "step": 8215
+    },
+    {
+      "epoch": 0.34593684210526315,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00037550619645660235,
+      "loss": 3.3381,
+      "step": 8216
+    },
+    {
+      "epoch": 0.34597894736842105,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00037547701146203004,
+      "loss": 3.512,
+      "step": 8217
+    },
+    {
+      "epoch": 0.34602105263157895,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003754478241814142,
+      "loss": 3.0702,
+      "step": 8218
+    },
+    {
+      "epoch": 0.34606315789473685,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003754186346152863,
+      "loss": 3.1449,
+      "step": 8219
+    },
+    {
+      "epoch": 0.34610526315789475,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003753894427641783,
+      "loss": 2.7087,
+      "step": 8220
+    },
+    {
+      "epoch": 0.34614736842105265,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0003753602486286219,
+      "loss": 3.1418,
+      "step": 8221
+    },
+    {
+      "epoch": 0.34618947368421055,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003753310522091491,
+      "loss": 3.4326,
+      "step": 8222
+    },
+    {
+      "epoch": 0.34623157894736845,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00037530185350629177,
+      "loss": 3.5793,
+      "step": 8223
+    },
+    {
+      "epoch": 0.3462736842105263,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00037527265252058185,
+      "loss": 3.2095,
+      "step": 8224
+    },
+    {
+      "epoch": 0.3463157894736842,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00037524344925255143,
+      "loss": 3.3471,
+      "step": 8225
+    },
+    {
+      "epoch": 0.3463578947368421,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00037521424370273254,
+      "loss": 3.1191,
+      "step": 8226
+    },
+    {
+      "epoch": 0.3464,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00037518503587165725,
+      "loss": 3.2251,
+      "step": 8227
+    },
+    {
+      "epoch": 0.3464421052631579,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00037515582575985774,
+      "loss": 3.1622,
+      "step": 8228
+    },
+    {
+      "epoch": 0.3464842105263158,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00037512661336786603,
+      "loss": 3.0325,
+      "step": 8229
+    },
+    {
+      "epoch": 0.3465263157894737,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003750973986962145,
+      "loss": 3.5389,
+      "step": 8230
+    },
+    {
+      "epoch": 0.3465684210526316,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003750681817454353,
+      "loss": 3.4414,
+      "step": 8231
+    },
+    {
+      "epoch": 0.3466105263157895,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00037503896251606084,
+      "loss": 3.6645,
+      "step": 8232
+    },
+    {
+      "epoch": 0.3466526315789474,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00037500974100862337,
+      "loss": 3.123,
+      "step": 8233
+    },
+    {
+      "epoch": 0.3466947368421053,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0003749805172236553,
+      "loss": 3.0322,
+      "step": 8234
+    },
+    {
+      "epoch": 0.3467368421052632,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003749512911616891,
+      "loss": 3.1078,
+      "step": 8235
+    },
+    {
+      "epoch": 0.3467789473684211,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00037492206282325715,
+      "loss": 2.5913,
+      "step": 8236
+    },
+    {
+      "epoch": 0.346821052631579,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003748928322088919,
+      "loss": 3.2595,
+      "step": 8237
+    },
+    {
+      "epoch": 0.3468631578947368,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00037486359931912614,
+      "loss": 3.1914,
+      "step": 8238
+    },
+    {
+      "epoch": 0.3469052631578947,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00037483436415449226,
+      "loss": 3.6113,
+      "step": 8239
+    },
+    {
+      "epoch": 0.3469473684210526,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0003748051267155229,
+      "loss": 3.1891,
+      "step": 8240
+    },
+    {
+      "epoch": 0.3469894736842105,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0003747758870027508,
+      "loss": 3.5032,
+      "step": 8241
+    },
+    {
+      "epoch": 0.3470315789473684,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00037474664501670863,
+      "loss": 3.4103,
+      "step": 8242
+    },
+    {
+      "epoch": 0.3470736842105263,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00037471740075792917,
+      "loss": 3.3254,
+      "step": 8243
+    },
+    {
+      "epoch": 0.3471157894736842,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0003746881542269452,
+      "loss": 3.7788,
+      "step": 8244
+    },
+    {
+      "epoch": 0.3471578947368421,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003746589054242895,
+      "loss": 2.7412,
+      "step": 8245
+    },
+    {
+      "epoch": 0.3472,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00037462965435049514,
+      "loss": 3.1441,
+      "step": 8246
+    },
+    {
+      "epoch": 0.3472421052631579,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00037460040100609486,
+      "loss": 3.5463,
+      "step": 8247
+    },
+    {
+      "epoch": 0.3472842105263158,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0003745711453916217,
+      "loss": 2.8335,
+      "step": 8248
+    },
+    {
+      "epoch": 0.3473263157894737,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0003745418875076085,
+      "loss": 3.301,
+      "step": 8249
+    },
+    {
+      "epoch": 0.3473684210526316,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003745126273545886,
+      "loss": 3.4761,
+      "step": 8250
+    },
+    {
+      "epoch": 0.34741052631578945,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003744833649330948,
+      "loss": 3.5165,
+      "step": 8251
+    },
+    {
+      "epoch": 0.34745263157894735,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00037445410024366046,
+      "loss": 3.5215,
+      "step": 8252
+    },
+    {
+      "epoch": 0.34749473684210525,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0003744248332868186,
+      "loss": 3.6126,
+      "step": 8253
+    },
+    {
+      "epoch": 0.34753684210526314,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003743955640631025,
+      "loss": 3.192,
+      "step": 8254
+    },
+    {
+      "epoch": 0.34757894736842104,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003743662925730453,
+      "loss": 3.1128,
+      "step": 8255
+    },
+    {
+      "epoch": 0.34762105263157894,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00037433701881718054,
+      "loss": 3.1386,
+      "step": 8256
+    },
+    {
+      "epoch": 0.34766315789473684,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0003743077427960412,
+      "loss": 2.9618,
+      "step": 8257
+    },
+    {
+      "epoch": 0.34770526315789474,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00037427846451016083,
+      "loss": 2.8275,
+      "step": 8258
+    },
+    {
+      "epoch": 0.34774736842105264,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00037424918396007303,
+      "loss": 3.6072,
+      "step": 8259
+    },
+    {
+      "epoch": 0.34778947368421054,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.000374219901146311,
+      "loss": 3.3121,
+      "step": 8260
+    },
+    {
+      "epoch": 0.34783157894736844,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003741906160694083,
+      "loss": 3.3312,
+      "step": 8261
+    },
+    {
+      "epoch": 0.34787368421052633,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003741613287298985,
+      "loss": 3.446,
+      "step": 8262
+    },
+    {
+      "epoch": 0.34791578947368423,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00037413203912831523,
+      "loss": 3.1421,
+      "step": 8263
+    },
+    {
+      "epoch": 0.34795789473684213,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.000374102747265192,
+      "loss": 3.2528,
+      "step": 8264
+    },
+    {
+      "epoch": 0.348,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00037407345314106254,
+      "loss": 3.2314,
+      "step": 8265
+    },
+    {
+      "epoch": 0.3480421052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00037404415675646054,
+      "loss": 3.4725,
+      "step": 8266
+    },
+    {
+      "epoch": 0.3480842105263158,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003740148581119197,
+      "loss": 3.4624,
+      "step": 8267
+    },
+    {
+      "epoch": 0.3481263157894737,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00037398555720797395,
+      "loss": 3.3025,
+      "step": 8268
+    },
+    {
+      "epoch": 0.34816842105263157,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.000373956254045157,
+      "loss": 3.2174,
+      "step": 8269
+    },
+    {
+      "epoch": 0.34821052631578947,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0003739269486240027,
+      "loss": 3.4145,
+      "step": 8270
+    },
+    {
+      "epoch": 0.34825263157894737,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0003738976409450451,
+      "loss": 3.0555,
+      "step": 8271
+    },
+    {
+      "epoch": 0.34829473684210527,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.000373868331008818,
+      "loss": 3.0283,
+      "step": 8272
+    },
+    {
+      "epoch": 0.34833684210526317,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003738390188158554,
+      "loss": 2.9652,
+      "step": 8273
+    },
+    {
+      "epoch": 0.34837894736842107,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0003738097043666914,
+      "loss": 3.2567,
+      "step": 8274
+    },
+    {
+      "epoch": 0.34842105263157896,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00037378038766186007,
+      "loss": 3.0372,
+      "step": 8275
+    },
+    {
+      "epoch": 0.34846315789473686,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.0003737510687018956,
+      "loss": 3.2202,
+      "step": 8276
+    },
+    {
+      "epoch": 0.34850526315789476,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00037372174748733194,
+      "loss": 3.2648,
+      "step": 8277
+    },
+    {
+      "epoch": 0.3485473684210526,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0003736924240187034,
+      "loss": 3.5579,
+      "step": 8278
+    },
+    {
+      "epoch": 0.3485894736842105,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00037366309829654434,
+      "loss": 3.5527,
+      "step": 8279
+    },
+    {
+      "epoch": 0.3486315789473684,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0003736337703213888,
+      "loss": 3.4104,
+      "step": 8280
+    },
+    {
+      "epoch": 0.3486736842105263,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00037360444009377136,
+      "loss": 2.9484,
+      "step": 8281
+    },
+    {
+      "epoch": 0.3487157894736842,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00037357510761422614,
+      "loss": 2.8831,
+      "step": 8282
+    },
+    {
+      "epoch": 0.3487578947368421,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003735457728832877,
+      "loss": 3.7802,
+      "step": 8283
+    },
+    {
+      "epoch": 0.3488,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003735164359014904,
+      "loss": 3.0854,
+      "step": 8284
+    },
+    {
+      "epoch": 0.3488421052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003734870966693688,
+      "loss": 3.2821,
+      "step": 8285
+    },
+    {
+      "epoch": 0.3488842105263158,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0003734577551874574,
+      "loss": 2.9781,
+      "step": 8286
+    },
+    {
+      "epoch": 0.3489263157894737,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0003734284114562908,
+      "loss": 3.2406,
+      "step": 8287
+    },
+    {
+      "epoch": 0.3489684210526316,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0003733990654764035,
+      "loss": 3.2524,
+      "step": 8288
+    },
+    {
+      "epoch": 0.3490105263157895,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00037336971724833025,
+      "loss": 3.0693,
+      "step": 8289
+    },
+    {
+      "epoch": 0.3490526315789474,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003733403667726056,
+      "loss": 2.9165,
+      "step": 8290
+    },
+    {
+      "epoch": 0.34909473684210524,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00037331101404976453,
+      "loss": 3.3348,
+      "step": 8291
+    },
+    {
+      "epoch": 0.34913684210526313,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00037328165908034157,
+      "loss": 3.3673,
+      "step": 8292
+    },
+    {
+      "epoch": 0.34917894736842103,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00037325230186487166,
+      "loss": 3.8094,
+      "step": 8293
+    },
+    {
+      "epoch": 0.34922105263157893,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00037322294240388967,
+      "loss": 3.2573,
+      "step": 8294
+    },
+    {
+      "epoch": 0.34926315789473683,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0003731935806979304,
+      "loss": 3.4634,
+      "step": 8295
+    },
+    {
+      "epoch": 0.34930526315789473,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003731642167475289,
+      "loss": 3.2582,
+      "step": 8296
+    },
+    {
+      "epoch": 0.34934736842105263,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00037313485055322004,
+      "loss": 3.0152,
+      "step": 8297
+    },
+    {
+      "epoch": 0.3493894736842105,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00037310548211553883,
+      "loss": 3.5773,
+      "step": 8298
+    },
+    {
+      "epoch": 0.3494315789473684,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003730761114350204,
+      "loss": 3.3615,
+      "step": 8299
+    },
+    {
+      "epoch": 0.3494736842105263,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00037304673851219983,
+      "loss": 3.5529,
+      "step": 8300
+    },
+    {
+      "epoch": 0.3495157894736842,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003730173633476123,
+      "loss": 3.2005,
+      "step": 8301
+    },
+    {
+      "epoch": 0.3495578947368421,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00037298798594179297,
+      "loss": 2.7605,
+      "step": 8302
+    },
+    {
+      "epoch": 0.3496,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.000372958606295277,
+      "loss": 2.8665,
+      "step": 8303
+    },
+    {
+      "epoch": 0.3496421052631579,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00037292922440859965,
+      "loss": 2.8548,
+      "step": 8304
+    },
+    {
+      "epoch": 0.34968421052631576,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003728998402822963,
+      "loss": 3.2386,
+      "step": 8305
+    },
+    {
+      "epoch": 0.34972631578947366,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00037287045391690236,
+      "loss": 3.6093,
+      "step": 8306
+    },
+    {
+      "epoch": 0.34976842105263156,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00037284106531295297,
+      "loss": 3.1839,
+      "step": 8307
+    },
+    {
+      "epoch": 0.34981052631578946,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00037281167447098375,
+      "loss": 3.069,
+      "step": 8308
+    },
+    {
+      "epoch": 0.34985263157894736,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00037278228139153014,
+      "loss": 3.0544,
+      "step": 8309
+    },
+    {
+      "epoch": 0.34989473684210526,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003727528860751276,
+      "loss": 3.4984,
+      "step": 8310
+    },
+    {
+      "epoch": 0.34993684210526316,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003727234885223117,
+      "loss": 3.7331,
+      "step": 8311
+    },
+    {
+      "epoch": 0.34997894736842106,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0003726940887336181,
+      "loss": 3.3205,
+      "step": 8312
+    },
+    {
+      "epoch": 0.35002105263157895,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0003726646867095823,
+      "loss": 2.9202,
+      "step": 8313
+    },
+    {
+      "epoch": 0.35006315789473685,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00037263528245074006,
+      "loss": 3.1256,
+      "step": 8314
+    },
+    {
+      "epoch": 0.35010526315789475,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00037260587595762705,
+      "loss": 3.3829,
+      "step": 8315
+    },
+    {
+      "epoch": 0.35014736842105265,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00037257646723077903,
+      "loss": 3.1338,
+      "step": 8316
+    },
+    {
+      "epoch": 0.35018947368421055,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003725470562707318,
+      "loss": 3.2075,
+      "step": 8317
+    },
+    {
+      "epoch": 0.3502315789473684,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00037251764307802117,
+      "loss": 3.2942,
+      "step": 8318
+    },
+    {
+      "epoch": 0.3502736842105263,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00037248822765318306,
+      "loss": 3.216,
+      "step": 8319
+    },
+    {
+      "epoch": 0.3503157894736842,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0003724588099967534,
+      "loss": 3.4296,
+      "step": 8320
+    },
+    {
+      "epoch": 0.3503578947368421,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00037242939010926815,
+      "loss": 3.5463,
+      "step": 8321
+    },
+    {
+      "epoch": 0.3504,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00037239996799126314,
+      "loss": 3.132,
+      "step": 8322
+    },
+    {
+      "epoch": 0.3504421052631579,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003723705436432746,
+      "loss": 2.8653,
+      "step": 8323
+    },
+    {
+      "epoch": 0.3504842105263158,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00037234111706583846,
+      "loss": 3.2856,
+      "step": 8324
+    },
+    {
+      "epoch": 0.3505263157894737,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00037231168825949096,
+      "loss": 3.4909,
+      "step": 8325
+    },
+    {
+      "epoch": 0.3505684210526316,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00037228225722476816,
+      "loss": 3.0142,
+      "step": 8326
+    },
+    {
+      "epoch": 0.3506105263157895,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0003722528239622064,
+      "loss": 3.0661,
+      "step": 8327
+    },
+    {
+      "epoch": 0.3506526315789474,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0003722233884723418,
+      "loss": 3.415,
+      "step": 8328
+    },
+    {
+      "epoch": 0.3506947368421053,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0003721939507557106,
+      "loss": 2.7923,
+      "step": 8329
+    },
+    {
+      "epoch": 0.3507368421052632,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0003721645108128493,
+      "loss": 3.3168,
+      "step": 8330
+    },
+    {
+      "epoch": 0.3507789473684211,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.000372135068644294,
+      "loss": 3.3632,
+      "step": 8331
+    },
+    {
+      "epoch": 0.3508210526315789,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00037210562425058136,
+      "loss": 3.4724,
+      "step": 8332
+    },
+    {
+      "epoch": 0.3508631578947368,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00037207617763224777,
+      "loss": 3.2023,
+      "step": 8333
+    },
+    {
+      "epoch": 0.3509052631578947,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00037204672878982955,
+      "loss": 3.1863,
+      "step": 8334
+    },
+    {
+      "epoch": 0.3509473684210526,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00037201727772386337,
+      "loss": 3.0408,
+      "step": 8335
+    },
+    {
+      "epoch": 0.3509894736842105,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00037198782443488576,
+      "loss": 3.4997,
+      "step": 8336
+    },
+    {
+      "epoch": 0.3510315789473684,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00037195836892343334,
+      "loss": 3.1214,
+      "step": 8337
+    },
+    {
+      "epoch": 0.3510736842105263,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00037192891119004263,
+      "loss": 3.0433,
+      "step": 8338
+    },
+    {
+      "epoch": 0.3511157894736842,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00037189945123525047,
+      "loss": 3.0337,
+      "step": 8339
+    },
+    {
+      "epoch": 0.3511578947368421,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003718699890595936,
+      "loss": 3.0498,
+      "step": 8340
+    },
+    {
+      "epoch": 0.3512,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0003718405246636087,
+      "loss": 3.2395,
+      "step": 8341
+    },
+    {
+      "epoch": 0.3512421052631579,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00037181105804783254,
+      "loss": 3.5704,
+      "step": 8342
+    },
+    {
+      "epoch": 0.3512842105263158,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00037178158921280204,
+      "loss": 3.1893,
+      "step": 8343
+    },
+    {
+      "epoch": 0.3513263157894737,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00037175211815905414,
+      "loss": 3.1545,
+      "step": 8344
+    },
+    {
+      "epoch": 0.35136842105263155,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003717226448871256,
+      "loss": 2.911,
+      "step": 8345
+    },
+    {
+      "epoch": 0.35141052631578945,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00037169316939755353,
+      "loss": 3.1644,
+      "step": 8346
+    },
+    {
+      "epoch": 0.35145263157894735,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00037166369169087493,
+      "loss": 3.1562,
+      "step": 8347
+    },
+    {
+      "epoch": 0.35149473684210525,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003716342117676268,
+      "loss": 3.1829,
+      "step": 8348
+    },
+    {
+      "epoch": 0.35153684210526315,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00037160472962834623,
+      "loss": 2.7158,
+      "step": 8349
+    },
+    {
+      "epoch": 0.35157894736842105,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00037157524527357035,
+      "loss": 3.2903,
+      "step": 8350
+    },
+    {
+      "epoch": 0.35162105263157895,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003715457587038364,
+      "loss": 2.6682,
+      "step": 8351
+    },
+    {
+      "epoch": 0.35166315789473684,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00037151626991968145,
+      "loss": 3.49,
+      "step": 8352
+    },
+    {
+      "epoch": 0.35170526315789474,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003714867789216429,
+      "loss": 3.1354,
+      "step": 8353
+    },
+    {
+      "epoch": 0.35174736842105264,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.000371457285710258,
+      "loss": 3.2262,
+      "step": 8354
+    },
+    {
+      "epoch": 0.35178947368421054,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00037142779028606404,
+      "loss": 3.6194,
+      "step": 8355
+    },
+    {
+      "epoch": 0.35183157894736844,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0003713982926495984,
+      "loss": 3.1915,
+      "step": 8356
+    },
+    {
+      "epoch": 0.35187368421052634,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00037136879280139854,
+      "loss": 3.194,
+      "step": 8357
+    },
+    {
+      "epoch": 0.35191578947368424,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0003713392907420018,
+      "loss": 3.0096,
+      "step": 8358
+    },
+    {
+      "epoch": 0.3519578947368421,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003713097864719458,
+      "loss": 3.6196,
+      "step": 8359
+    },
+    {
+      "epoch": 0.352,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.000371280279991768,
+      "loss": 3.5267,
+      "step": 8360
+    },
+    {
+      "epoch": 0.3520421052631579,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00037125077130200604,
+      "loss": 2.8173,
+      "step": 8361
+    },
+    {
+      "epoch": 0.3520842105263158,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00037122126040319746,
+      "loss": 3.1809,
+      "step": 8362
+    },
+    {
+      "epoch": 0.3521263157894737,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00037119174729588,
+      "loss": 3.4177,
+      "step": 8363
+    },
+    {
+      "epoch": 0.3521684210526316,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0003711622319805913,
+      "loss": 3.1437,
+      "step": 8364
+    },
+    {
+      "epoch": 0.3522105263157895,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.000371132714457869,
+      "loss": 3.3474,
+      "step": 8365
+    },
+    {
+      "epoch": 0.3522526315789474,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.000371103194728251,
+      "loss": 3.1933,
+      "step": 8366
+    },
+    {
+      "epoch": 0.35229473684210527,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00037107367279227505,
+      "loss": 3.6506,
+      "step": 8367
+    },
+    {
+      "epoch": 0.35233684210526317,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003710441486504791,
+      "loss": 3.3752,
+      "step": 8368
+    },
+    {
+      "epoch": 0.35237894736842107,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.000371014622303401,
+      "loss": 3.3134,
+      "step": 8369
+    },
+    {
+      "epoch": 0.35242105263157897,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003709850937515786,
+      "loss": 2.9721,
+      "step": 8370
+    },
+    {
+      "epoch": 0.35246315789473687,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00037095556299555,
+      "loss": 3.2375,
+      "step": 8371
+    },
+    {
+      "epoch": 0.3525052631578947,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00037092603003585314,
+      "loss": 3.3239,
+      "step": 8372
+    },
+    {
+      "epoch": 0.3525473684210526,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00037089649487302614,
+      "loss": 3.3319,
+      "step": 8373
+    },
+    {
+      "epoch": 0.3525894736842105,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00037086695750760697,
+      "loss": 3.2538,
+      "step": 8374
+    },
+    {
+      "epoch": 0.3526315789473684,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00037083741794013384,
+      "loss": 3.1107,
+      "step": 8375
+    },
+    {
+      "epoch": 0.3526736842105263,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.000370807876171145,
+      "loss": 3.5319,
+      "step": 8376
+    },
+    {
+      "epoch": 0.3527157894736842,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00037077833220117856,
+      "loss": 3.1483,
+      "step": 8377
+    },
+    {
+      "epoch": 0.3527578947368421,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00037074878603077287,
+      "loss": 3.3937,
+      "step": 8378
+    },
+    {
+      "epoch": 0.3528,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0003707192376604661,
+      "loss": 3.1436,
+      "step": 8379
+    },
+    {
+      "epoch": 0.3528421052631579,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0003706896870907967,
+      "loss": 3.5844,
+      "step": 8380
+    },
+    {
+      "epoch": 0.3528842105263158,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.000370660134322303,
+      "loss": 2.9009,
+      "step": 8381
+    },
+    {
+      "epoch": 0.3529263157894737,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003706305793555234,
+      "loss": 3.2249,
+      "step": 8382
+    },
+    {
+      "epoch": 0.3529684210526316,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003706010221909964,
+      "loss": 3.1035,
+      "step": 8383
+    },
+    {
+      "epoch": 0.3530105263157895,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00037057146282926047,
+      "loss": 3.6417,
+      "step": 8384
+    },
+    {
+      "epoch": 0.3530526315789474,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003705419012708541,
+      "loss": 3.0527,
+      "step": 8385
+    },
+    {
+      "epoch": 0.35309473684210524,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00037051233751631604,
+      "loss": 3.4006,
+      "step": 8386
+    },
+    {
+      "epoch": 0.35313684210526314,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003704827715661847,
+      "loss": 3.4922,
+      "step": 8387
+    },
+    {
+      "epoch": 0.35317894736842104,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003704532034209989,
+      "loss": 3.0543,
+      "step": 8388
+    },
+    {
+      "epoch": 0.35322105263157894,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003704236330812972,
+      "loss": 3.1797,
+      "step": 8389
+    },
+    {
+      "epoch": 0.35326315789473683,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003703940605476184,
+      "loss": 3.3814,
+      "step": 8390
+    },
+    {
+      "epoch": 0.35330526315789473,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00037036448582050127,
+      "loss": 3.3005,
+      "step": 8391
+    },
+    {
+      "epoch": 0.35334736842105263,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00037033490890048466,
+      "loss": 3.2047,
+      "step": 8392
+    },
+    {
+      "epoch": 0.35338947368421053,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00037030532978810733,
+      "loss": 3.2599,
+      "step": 8393
+    },
+    {
+      "epoch": 0.35343157894736843,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003702757484839083,
+      "loss": 3.3488,
+      "step": 8394
+    },
+    {
+      "epoch": 0.35347368421052633,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003702461649884264,
+      "loss": 3.0243,
+      "step": 8395
+    },
+    {
+      "epoch": 0.3535157894736842,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00037021657930220075,
+      "loss": 2.9283,
+      "step": 8396
+    },
+    {
+      "epoch": 0.3535578947368421,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00037018699142577025,
+      "loss": 2.8585,
+      "step": 8397
+    },
+    {
+      "epoch": 0.3536,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.000370157401359674,
+      "loss": 3.5259,
+      "step": 8398
+    },
+    {
+      "epoch": 0.35364210526315787,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00037012780910445096,
+      "loss": 2.9404,
+      "step": 8399
+    },
+    {
+      "epoch": 0.35368421052631577,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00037009821466064047,
+      "loss": 3.357,
+      "step": 8400
+    },
+    {
+      "epoch": 0.35372631578947367,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00037006861802878156,
+      "loss": 3.5274,
+      "step": 8401
+    },
+    {
+      "epoch": 0.35376842105263157,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00037003901920941353,
+      "loss": 3.3709,
+      "step": 8402
+    },
+    {
+      "epoch": 0.35381052631578946,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00037000941820307566,
+      "loss": 3.5608,
+      "step": 8403
+    },
+    {
+      "epoch": 0.35385263157894736,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00036997981501030717,
+      "loss": 3.244,
+      "step": 8404
+    },
+    {
+      "epoch": 0.35389473684210526,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003699502096316474,
+      "loss": 3.2515,
+      "step": 8405
+    },
+    {
+      "epoch": 0.35393684210526316,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003699206020676358,
+      "loss": 2.8718,
+      "step": 8406
+    },
+    {
+      "epoch": 0.35397894736842106,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00036989099231881165,
+      "loss": 3.0641,
+      "step": 8407
+    },
+    {
+      "epoch": 0.35402105263157896,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00036986138038571455,
+      "loss": 3.5158,
+      "step": 8408
+    },
+    {
+      "epoch": 0.35406315789473686,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0003698317662688839,
+      "loss": 3.3388,
+      "step": 8409
+    },
+    {
+      "epoch": 0.35410526315789476,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00036980214996885924,
+      "loss": 3.3412,
+      "step": 8410
+    },
+    {
+      "epoch": 0.35414736842105266,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00036977253148618024,
+      "loss": 2.7781,
+      "step": 8411
+    },
+    {
+      "epoch": 0.35418947368421055,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00036974291082138636,
+      "loss": 3.35,
+      "step": 8412
+    },
+    {
+      "epoch": 0.3542315789473684,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00036971328797501734,
+      "loss": 3.5997,
+      "step": 8413
+    },
+    {
+      "epoch": 0.3542736842105263,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.000369683662947613,
+      "loss": 3.3039,
+      "step": 8414
+    },
+    {
+      "epoch": 0.3543157894736842,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003696540357397128,
+      "loss": 2.9617,
+      "step": 8415
+    },
+    {
+      "epoch": 0.3543578947368421,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0003696244063518567,
+      "loss": 3.252,
+      "step": 8416
+    },
+    {
+      "epoch": 0.3544,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00036959477478458447,
+      "loss": 2.8422,
+      "step": 8417
+    },
+    {
+      "epoch": 0.3544421052631579,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00036956514103843595,
+      "loss": 3.2049,
+      "step": 8418
+    },
+    {
+      "epoch": 0.3544842105263158,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003695355051139511,
+      "loss": 3.1824,
+      "step": 8419
+    },
+    {
+      "epoch": 0.3545263157894737,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00036950586701166977,
+      "loss": 3.3621,
+      "step": 8420
+    },
+    {
+      "epoch": 0.3545684210526316,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003694762267321319,
+      "loss": 2.9948,
+      "step": 8421
+    },
+    {
+      "epoch": 0.3546105263157895,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0003694465842758777,
+      "loss": 3.5632,
+      "step": 8422
+    },
+    {
+      "epoch": 0.3546526315789474,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.000369416939643447,
+      "loss": 3.1322,
+      "step": 8423
+    },
+    {
+      "epoch": 0.3546947368421053,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003693872928353799,
+      "loss": 3.2696,
+      "step": 8424
+    },
+    {
+      "epoch": 0.3547368421052632,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00036935764385221667,
+      "loss": 3.1635,
+      "step": 8425
+    },
+    {
+      "epoch": 0.354778947368421,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003693279926944974,
+      "loss": 3.3804,
+      "step": 8426
+    },
+    {
+      "epoch": 0.3548210526315789,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00036929833936276223,
+      "loss": 3.3277,
+      "step": 8427
+    },
+    {
+      "epoch": 0.3548631578947368,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003692686838575516,
+      "loss": 3.0877,
+      "step": 8428
+    },
+    {
+      "epoch": 0.3549052631578947,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00036923902617940564,
+      "loss": 3.3003,
+      "step": 8429
+    },
+    {
+      "epoch": 0.3549473684210526,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0003692093663288647,
+      "loss": 3.5651,
+      "step": 8430
+    },
+    {
+      "epoch": 0.3549894736842105,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003691797043064692,
+      "loss": 3.1963,
+      "step": 8431
+    },
+    {
+      "epoch": 0.3550315789473684,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00036915004011275947,
+      "loss": 3.2812,
+      "step": 8432
+    },
+    {
+      "epoch": 0.3550736842105263,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00036912037374827605,
+      "loss": 3.3763,
+      "step": 8433
+    },
+    {
+      "epoch": 0.3551157894736842,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00036909070521355935,
+      "loss": 3.1488,
+      "step": 8434
+    },
+    {
+      "epoch": 0.3551578947368421,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00036906103450914994,
+      "loss": 3.2133,
+      "step": 8435
+    },
+    {
+      "epoch": 0.3552,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003690313616355884,
+      "loss": 2.9893,
+      "step": 8436
+    },
+    {
+      "epoch": 0.3552421052631579,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00036900168659341535,
+      "loss": 3.2021,
+      "step": 8437
+    },
+    {
+      "epoch": 0.3552842105263158,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00036897200938317134,
+      "loss": 3.0011,
+      "step": 8438
+    },
+    {
+      "epoch": 0.35532631578947366,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00036894233000539707,
+      "loss": 3.5855,
+      "step": 8439
+    },
+    {
+      "epoch": 0.35536842105263156,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00036891264846063333,
+      "loss": 3.3401,
+      "step": 8440
+    },
+    {
+      "epoch": 0.35541052631578945,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0003688829647494208,
+      "loss": 3.3315,
+      "step": 8441
+    },
+    {
+      "epoch": 0.35545263157894735,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0003688532788723004,
+      "loss": 3.0723,
+      "step": 8442
+    },
+    {
+      "epoch": 0.35549473684210525,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0003688235908298129,
+      "loss": 3.2475,
+      "step": 8443
+    },
+    {
+      "epoch": 0.35553684210526315,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003687939006224992,
+      "loss": 3.7703,
+      "step": 8444
+    },
+    {
+      "epoch": 0.35557894736842105,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0003687642082509002,
+      "loss": 3.2372,
+      "step": 8445
+    },
+    {
+      "epoch": 0.35562105263157895,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003687345137155568,
+      "loss": 3.3663,
+      "step": 8446
+    },
+    {
+      "epoch": 0.35566315789473685,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003687048170170101,
+      "loss": 3.4374,
+      "step": 8447
+    },
+    {
+      "epoch": 0.35570526315789475,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00036867511815580117,
+      "loss": 2.7673,
+      "step": 8448
+    },
+    {
+      "epoch": 0.35574736842105265,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.000368645417132471,
+      "loss": 3.7187,
+      "step": 8449
+    },
+    {
+      "epoch": 0.35578947368421054,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00036861571394756066,
+      "loss": 3.5667,
+      "step": 8450
+    },
+    {
+      "epoch": 0.35583157894736844,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0003685860086016115,
+      "loss": 3.3765,
+      "step": 8451
+    },
+    {
+      "epoch": 0.35587368421052634,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0003685563010951645,
+      "loss": 3.4038,
+      "step": 8452
+    },
+    {
+      "epoch": 0.3559157894736842,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.000368526591428761,
+      "loss": 3.4084,
+      "step": 8453
+    },
+    {
+      "epoch": 0.3559578947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003684968796029423,
+      "loss": 3.5268,
+      "step": 8454
+    },
+    {
+      "epoch": 0.356,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00036846716561824967,
+      "loss": 3.0478,
+      "step": 8455
+    },
+    {
+      "epoch": 0.3560421052631579,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0003684374494752245,
+      "loss": 2.9799,
+      "step": 8456
+    },
+    {
+      "epoch": 0.3560842105263158,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003684077311744082,
+      "loss": 3.1679,
+      "step": 8457
+    },
+    {
+      "epoch": 0.3561263157894737,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00036837801071634203,
+      "loss": 2.7548,
+      "step": 8458
+    },
+    {
+      "epoch": 0.3561684210526316,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0003683482881015676,
+      "loss": 3.3332,
+      "step": 8459
+    },
+    {
+      "epoch": 0.3562105263157895,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00036831856333062647,
+      "loss": 3.5327,
+      "step": 8460
+    },
+    {
+      "epoch": 0.3562526315789474,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0003682888364040602,
+      "loss": 3.1463,
+      "step": 8461
+    },
+    {
+      "epoch": 0.3562947368421053,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00036825910732241025,
+      "loss": 3.1507,
+      "step": 8462
+    },
+    {
+      "epoch": 0.3563368421052632,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003682293760862183,
+      "loss": 3.032,
+      "step": 8463
+    },
+    {
+      "epoch": 0.3563789473684211,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00036819964269602605,
+      "loss": 3.0423,
+      "step": 8464
+    },
+    {
+      "epoch": 0.35642105263157897,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003681699071523752,
+      "loss": 3.1754,
+      "step": 8465
+    },
+    {
+      "epoch": 0.3564631578947368,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00036814016945580744,
+      "loss": 2.7483,
+      "step": 8466
+    },
+    {
+      "epoch": 0.3565052631578947,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0003681104296068646,
+      "loss": 3.6548,
+      "step": 8467
+    },
+    {
+      "epoch": 0.3565473684210526,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0003680806876060886,
+      "loss": 3.4123,
+      "step": 8468
+    },
+    {
+      "epoch": 0.3565894736842105,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0003680509434540211,
+      "loss": 2.6754,
+      "step": 8469
+    },
+    {
+      "epoch": 0.3566315789473684,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00036802119715120417,
+      "loss": 3.3347,
+      "step": 8470
+    },
+    {
+      "epoch": 0.3566736842105263,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00036799144869817976,
+      "loss": 3.4992,
+      "step": 8471
+    },
+    {
+      "epoch": 0.3567157894736842,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00036796169809548976,
+      "loss": 3.6774,
+      "step": 8472
+    },
+    {
+      "epoch": 0.3567578947368421,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00036793194534367606,
+      "loss": 3.2564,
+      "step": 8473
+    },
+    {
+      "epoch": 0.3568,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0003679021904432811,
+      "loss": 3.3799,
+      "step": 8474
+    },
+    {
+      "epoch": 0.3568421052631579,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003678724333948466,
+      "loss": 2.9776,
+      "step": 8475
+    },
+    {
+      "epoch": 0.3568842105263158,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0003678426741989149,
+      "loss": 3.2577,
+      "step": 8476
+    },
+    {
+      "epoch": 0.3569263157894737,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00036781291285602815,
+      "loss": 3.5254,
+      "step": 8477
+    },
+    {
+      "epoch": 0.3569684210526316,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00036778314936672857,
+      "loss": 2.9002,
+      "step": 8478
+    },
+    {
+      "epoch": 0.3570105263157895,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00036775338373155837,
+      "loss": 3.5185,
+      "step": 8479
+    },
+    {
+      "epoch": 0.35705263157894734,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00036772361595105994,
+      "loss": 3.5643,
+      "step": 8480
+    },
+    {
+      "epoch": 0.35709473684210524,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00036769384602577547,
+      "loss": 3.0884,
+      "step": 8481
+    },
+    {
+      "epoch": 0.35713684210526314,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00036766407395624745,
+      "loss": 3.3613,
+      "step": 8482
+    },
+    {
+      "epoch": 0.35717894736842104,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003676342997430182,
+      "loss": 3.1784,
+      "step": 8483
+    },
+    {
+      "epoch": 0.35722105263157894,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00036760452338663026,
+      "loss": 3.2515,
+      "step": 8484
+    },
+    {
+      "epoch": 0.35726315789473684,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0003675747448876261,
+      "loss": 2.7707,
+      "step": 8485
+    },
+    {
+      "epoch": 0.35730526315789474,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00036754496424654814,
+      "loss": 3.3514,
+      "step": 8486
+    },
+    {
+      "epoch": 0.35734736842105264,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00036751518146393915,
+      "loss": 3.2161,
+      "step": 8487
+    },
+    {
+      "epoch": 0.35738947368421053,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00036748539654034166,
+      "loss": 3.5402,
+      "step": 8488
+    },
+    {
+      "epoch": 0.35743157894736843,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003674556094762982,
+      "loss": 2.8124,
+      "step": 8489
+    },
+    {
+      "epoch": 0.35747368421052633,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00036742582027235156,
+      "loss": 3.3315,
+      "step": 8490
+    },
+    {
+      "epoch": 0.35751578947368423,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00036739602892904447,
+      "loss": 3.1274,
+      "step": 8491
+    },
+    {
+      "epoch": 0.35755789473684213,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0003673662354469196,
+      "loss": 3.6463,
+      "step": 8492
+    },
+    {
+      "epoch": 0.3576,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0003673364398265199,
+      "loss": 3.4198,
+      "step": 8493
+    },
+    {
+      "epoch": 0.3576421052631579,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00036730664206838806,
+      "loss": 3.3445,
+      "step": 8494
+    },
+    {
+      "epoch": 0.35768421052631577,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00036727684217306705,
+      "loss": 3.308,
+      "step": 8495
+    },
+    {
+      "epoch": 0.35772631578947367,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00036724704014109977,
+      "loss": 3.0323,
+      "step": 8496
+    },
+    {
+      "epoch": 0.35776842105263157,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00036721723597302923,
+      "loss": 3.1548,
+      "step": 8497
+    },
+    {
+      "epoch": 0.35781052631578947,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0003671874296693983,
+      "loss": 3.1609,
+      "step": 8498
+    },
+    {
+      "epoch": 0.35785263157894737,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00036715762123075017,
+      "loss": 3.5152,
+      "step": 8499
+    },
+    {
+      "epoch": 0.35789473684210527,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003671278106576278,
+      "loss": 3.2645,
+      "step": 8500
+    },
+    {
+      "epoch": 0.35793684210526316,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003670979979505743,
+      "loss": 2.6878,
+      "step": 8501
+    },
+    {
+      "epoch": 0.35797894736842106,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003670681831101329,
+      "loss": 2.7624,
+      "step": 8502
+    },
+    {
+      "epoch": 0.35802105263157896,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00036703836613684673,
+      "loss": 3.2892,
+      "step": 8503
+    },
+    {
+      "epoch": 0.35806315789473686,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0003670085470312591,
+      "loss": 3.0266,
+      "step": 8504
+    },
+    {
+      "epoch": 0.35810526315789476,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00036697872579391315,
+      "loss": 3.1614,
+      "step": 8505
+    },
+    {
+      "epoch": 0.35814736842105266,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0003669489024253523,
+      "loss": 3.4023,
+      "step": 8506
+    },
+    {
+      "epoch": 0.3581894736842105,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0003669190769261199,
+      "loss": 3.4111,
+      "step": 8507
+    },
+    {
+      "epoch": 0.3582315789473684,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003668892492967592,
+      "loss": 2.6688,
+      "step": 8508
+    },
+    {
+      "epoch": 0.3582736842105263,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00036685941953781376,
+      "loss": 3.307,
+      "step": 8509
+    },
+    {
+      "epoch": 0.3583157894736842,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.000366829587649827,
+      "loss": 3.0182,
+      "step": 8510
+    },
+    {
+      "epoch": 0.3583578947368421,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0003667997536333424,
+      "loss": 2.8687,
+      "step": 8511
+    },
+    {
+      "epoch": 0.3584,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003667699174889036,
+      "loss": 3.5888,
+      "step": 8512
+    },
+    {
+      "epoch": 0.3584421052631579,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.000366740079217054,
+      "loss": 3.4365,
+      "step": 8513
+    },
+    {
+      "epoch": 0.3584842105263158,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00036671023881833736,
+      "loss": 3.4626,
+      "step": 8514
+    },
+    {
+      "epoch": 0.3585263157894737,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00036668039629329736,
+      "loss": 3.0771,
+      "step": 8515
+    },
+    {
+      "epoch": 0.3585684210526316,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00036665055164247757,
+      "loss": 3.2052,
+      "step": 8516
+    },
+    {
+      "epoch": 0.3586105263157895,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0003666207048664218,
+      "loss": 3.2578,
+      "step": 8517
+    },
+    {
+      "epoch": 0.3586526315789474,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00036659085596567375,
+      "loss": 3.449,
+      "step": 8518
+    },
+    {
+      "epoch": 0.3586947368421053,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003665610049407774,
+      "loss": 3.1248,
+      "step": 8519
+    },
+    {
+      "epoch": 0.35873684210526313,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003665311517922764,
+      "loss": 2.9644,
+      "step": 8520
+    },
+    {
+      "epoch": 0.35877894736842103,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003665012965207148,
+      "loss": 3.2378,
+      "step": 8521
+    },
+    {
+      "epoch": 0.35882105263157893,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003664714391266364,
+      "loss": 3.2095,
+      "step": 8522
+    },
+    {
+      "epoch": 0.35886315789473683,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00036644157961058534,
+      "loss": 3.2669,
+      "step": 8523
+    },
+    {
+      "epoch": 0.3589052631578947,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003664117179731055,
+      "loss": 3.2812,
+      "step": 8524
+    },
+    {
+      "epoch": 0.3589473684210526,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00036638185421474087,
+      "loss": 3.2559,
+      "step": 8525
+    },
+    {
+      "epoch": 0.3589894736842105,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0003663519883360356,
+      "loss": 3.3712,
+      "step": 8526
+    },
+    {
+      "epoch": 0.3590315789473684,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0003663221203375337,
+      "loss": 3.3206,
+      "step": 8527
+    },
+    {
+      "epoch": 0.3590736842105263,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00036629225021977957,
+      "loss": 3.0226,
+      "step": 8528
+    },
+    {
+      "epoch": 0.3591157894736842,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003662623779833173,
+      "loss": 3.3128,
+      "step": 8529
+    },
+    {
+      "epoch": 0.3591578947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003662325036286911,
+      "loss": 3.3421,
+      "step": 8530
+    },
+    {
+      "epoch": 0.3592,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003662026271564452,
+      "loss": 3.3016,
+      "step": 8531
+    },
+    {
+      "epoch": 0.3592421052631579,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00036617274856712403,
+      "loss": 3.5453,
+      "step": 8532
+    },
+    {
+      "epoch": 0.3592842105263158,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00036614286786127174,
+      "loss": 3.1034,
+      "step": 8533
+    },
+    {
+      "epoch": 0.35932631578947366,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00036611298503943303,
+      "loss": 3.2479,
+      "step": 8534
+    },
+    {
+      "epoch": 0.35936842105263156,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.000366083100102152,
+      "loss": 2.9535,
+      "step": 8535
+    },
+    {
+      "epoch": 0.35941052631578946,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0003660532130499734,
+      "loss": 3.0495,
+      "step": 8536
+    },
+    {
+      "epoch": 0.35945263157894736,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003660233238834416,
+      "loss": 3.1214,
+      "step": 8537
+    },
+    {
+      "epoch": 0.35949473684210526,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00036599343260310113,
+      "loss": 3.5953,
+      "step": 8538
+    },
+    {
+      "epoch": 0.35953684210526315,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0003659635392094967,
+      "loss": 3.1289,
+      "step": 8539
+    },
+    {
+      "epoch": 0.35957894736842105,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00036593364370317267,
+      "loss": 2.8361,
+      "step": 8540
+    },
+    {
+      "epoch": 0.35962105263157895,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00036590374608467396,
+      "loss": 2.8669,
+      "step": 8541
+    },
+    {
+      "epoch": 0.35966315789473685,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00036587384635454516,
+      "loss": 3.1655,
+      "step": 8542
+    },
+    {
+      "epoch": 0.35970526315789475,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.000365843944513331,
+      "loss": 3.2912,
+      "step": 8543
+    },
+    {
+      "epoch": 0.35974736842105265,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00036581404056157635,
+      "loss": 3.2754,
+      "step": 8544
+    },
+    {
+      "epoch": 0.35978947368421055,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00036578413449982596,
+      "loss": 2.8622,
+      "step": 8545
+    },
+    {
+      "epoch": 0.35983157894736845,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0003657542263286246,
+      "loss": 3.07,
+      "step": 8546
+    },
+    {
+      "epoch": 0.3598736842105263,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00036572431604851727,
+      "loss": 3.2908,
+      "step": 8547
+    },
+    {
+      "epoch": 0.3599157894736842,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0003656944036600489,
+      "loss": 3.5344,
+      "step": 8548
+    },
+    {
+      "epoch": 0.3599578947368421,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003656644891637644,
+      "loss": 3.3022,
+      "step": 8549
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00036563457256020887,
+      "loss": 3.5644,
+      "step": 8550
+    },
+    {
+      "epoch": 0.3600421052631579,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00036560465384992717,
+      "loss": 3.1371,
+      "step": 8551
+    },
+    {
+      "epoch": 0.3600842105263158,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003655747330334646,
+      "loss": 3.301,
+      "step": 8552
+    },
+    {
+      "epoch": 0.3601263157894737,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00036554481011136615,
+      "loss": 3.2776,
+      "step": 8553
+    },
+    {
+      "epoch": 0.3601684210526316,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00036551488508417706,
+      "loss": 3.3209,
+      "step": 8554
+    },
+    {
+      "epoch": 0.3602105263157895,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00036548495795244254,
+      "loss": 3.0257,
+      "step": 8555
+    },
+    {
+      "epoch": 0.3602526315789474,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00036545502871670766,
+      "loss": 3.1573,
+      "step": 8556
+    },
+    {
+      "epoch": 0.3602947368421053,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0003654250973775179,
+      "loss": 3.4982,
+      "step": 8557
+    },
+    {
+      "epoch": 0.3603368421052632,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00036539516393541843,
+      "loss": 2.9723,
+      "step": 8558
+    },
+    {
+      "epoch": 0.3603789473684211,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003653652283909547,
+      "loss": 3.3682,
+      "step": 8559
+    },
+    {
+      "epoch": 0.3604210526315789,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.000365335290744672,
+      "loss": 3.3689,
+      "step": 8560
+    },
+    {
+      "epoch": 0.3604631578947368,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00036530535099711577,
+      "loss": 3.1767,
+      "step": 8561
+    },
+    {
+      "epoch": 0.3605052631578947,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0003652754091488316,
+      "loss": 3.1023,
+      "step": 8562
+    },
+    {
+      "epoch": 0.3605473684210526,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00036524546520036493,
+      "loss": 3.2475,
+      "step": 8563
+    },
+    {
+      "epoch": 0.3605894736842105,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00036521551915226133,
+      "loss": 2.9145,
+      "step": 8564
+    },
+    {
+      "epoch": 0.3606315789473684,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003651855710050663,
+      "loss": 3.2941,
+      "step": 8565
+    },
+    {
+      "epoch": 0.3606736842105263,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00036515562075932544,
+      "loss": 3.3197,
+      "step": 8566
+    },
+    {
+      "epoch": 0.3607157894736842,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00036512566841558454,
+      "loss": 3.1512,
+      "step": 8567
+    },
+    {
+      "epoch": 0.3607578947368421,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0003650957139743892,
+      "loss": 3.0299,
+      "step": 8568
+    },
+    {
+      "epoch": 0.3608,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00036506575743628516,
+      "loss": 3.2048,
+      "step": 8569
+    },
+    {
+      "epoch": 0.3608421052631579,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00036503579880181825,
+      "loss": 3.1144,
+      "step": 8570
+    },
+    {
+      "epoch": 0.3608842105263158,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003650058380715343,
+      "loss": 3.009,
+      "step": 8571
+    },
+    {
+      "epoch": 0.3609263157894737,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.000364975875245979,
+      "loss": 3.3897,
+      "step": 8572
+    },
+    {
+      "epoch": 0.3609684210526316,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0003649459103256985,
+      "loss": 2.8768,
+      "step": 8573
+    },
+    {
+      "epoch": 0.36101052631578945,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003649159433112384,
+      "loss": 3.0635,
+      "step": 8574
+    },
+    {
+      "epoch": 0.36105263157894735,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0003648859742031449,
+      "loss": 2.724,
+      "step": 8575
+    },
+    {
+      "epoch": 0.36109473684210525,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00036485600300196396,
+      "loss": 3.3569,
+      "step": 8576
+    },
+    {
+      "epoch": 0.36113684210526314,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003648260297082415,
+      "loss": 3.0637,
+      "step": 8577
+    },
+    {
+      "epoch": 0.36117894736842104,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0003647960543225238,
+      "loss": 3.51,
+      "step": 8578
+    },
+    {
+      "epoch": 0.36122105263157894,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003647660768453569,
+      "loss": 3.3347,
+      "step": 8579
+    },
+    {
+      "epoch": 0.36126315789473684,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0003647360972772868,
+      "loss": 3.1537,
+      "step": 8580
+    },
+    {
+      "epoch": 0.36130526315789474,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00036470611561885993,
+      "loss": 3.0917,
+      "step": 8581
+    },
+    {
+      "epoch": 0.36134736842105264,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003646761318706223,
+      "loss": 3.0826,
+      "step": 8582
+    },
+    {
+      "epoch": 0.36138947368421054,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00036464614603312043,
+      "loss": 3.2376,
+      "step": 8583
+    },
+    {
+      "epoch": 0.36143157894736844,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00036461615810690035,
+      "loss": 3.056,
+      "step": 8584
+    },
+    {
+      "epoch": 0.36147368421052634,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0003645861680925086,
+      "loss": 3.3235,
+      "step": 8585
+    },
+    {
+      "epoch": 0.36151578947368423,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003645561759904915,
+      "loss": 3.1803,
+      "step": 8586
+    },
+    {
+      "epoch": 0.3615578947368421,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0003645261818013954,
+      "loss": 2.6219,
+      "step": 8587
+    },
+    {
+      "epoch": 0.3616,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0003644961855257669,
+      "loss": 3.3653,
+      "step": 8588
+    },
+    {
+      "epoch": 0.3616421052631579,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003644661871641524,
+      "loss": 3.1121,
+      "step": 8589
+    },
+    {
+      "epoch": 0.3616842105263158,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003644361867170985,
+      "loss": 3.4479,
+      "step": 8590
+    },
+    {
+      "epoch": 0.3617263157894737,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0003644061841851517,
+      "loss": 3.5041,
+      "step": 8591
+    },
+    {
+      "epoch": 0.3617684210526316,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003643761795688587,
+      "loss": 3.4588,
+      "step": 8592
+    },
+    {
+      "epoch": 0.36181052631578947,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00036434617286876605,
+      "loss": 3.0729,
+      "step": 8593
+    },
+    {
+      "epoch": 0.36185263157894737,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003643161640854204,
+      "loss": 2.9089,
+      "step": 8594
+    },
+    {
+      "epoch": 0.36189473684210527,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00036428615321936863,
+      "loss": 3.2881,
+      "step": 8595
+    },
+    {
+      "epoch": 0.36193684210526317,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0003642561402711575,
+      "loss": 3.2396,
+      "step": 8596
+    },
+    {
+      "epoch": 0.36197894736842107,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003642261252413337,
+      "loss": 3.0016,
+      "step": 8597
+    },
+    {
+      "epoch": 0.36202105263157897,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00036419610813044406,
+      "loss": 3.0481,
+      "step": 8598
+    },
+    {
+      "epoch": 0.36206315789473686,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0003641660889390356,
+      "loss": 3.0796,
+      "step": 8599
+    },
+    {
+      "epoch": 0.36210526315789476,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00036413606766765505,
+      "loss": 3.2641,
+      "step": 8600
+    },
+    {
+      "epoch": 0.3621473684210526,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0003641060443168494,
+      "loss": 3.4444,
+      "step": 8601
+    },
+    {
+      "epoch": 0.3621894736842105,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003640760188871657,
+      "loss": 3.1555,
+      "step": 8602
+    },
+    {
+      "epoch": 0.3622315789473684,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00036404599137915104,
+      "loss": 3.2479,
+      "step": 8603
+    },
+    {
+      "epoch": 0.3622736842105263,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00036401596179335246,
+      "loss": 3.2287,
+      "step": 8604
+    },
+    {
+      "epoch": 0.3623157894736842,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.000363985930130317,
+      "loss": 3.3387,
+      "step": 8605
+    },
+    {
+      "epoch": 0.3623578947368421,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0003639558963905918,
+      "loss": 3.393,
+      "step": 8606
+    },
+    {
+      "epoch": 0.3624,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00036392586057472397,
+      "loss": 2.7533,
+      "step": 8607
+    },
+    {
+      "epoch": 0.3624421052631579,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003638958226832609,
+      "loss": 2.9388,
+      "step": 8608
+    },
+    {
+      "epoch": 0.3624842105263158,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003638657827167498,
+      "loss": 3.2171,
+      "step": 8609
+    },
+    {
+      "epoch": 0.3625263157894737,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0003638357406757379,
+      "loss": 2.8198,
+      "step": 8610
+    },
+    {
+      "epoch": 0.3625684210526316,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0003638056965607725,
+      "loss": 2.9649,
+      "step": 8611
+    },
+    {
+      "epoch": 0.3626105263157895,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003637756503724011,
+      "loss": 3.4412,
+      "step": 8612
+    },
+    {
+      "epoch": 0.3626526315789474,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00036374560211117104,
+      "loss": 2.9699,
+      "step": 8613
+    },
+    {
+      "epoch": 0.36269473684210524,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0003637155517776297,
+      "loss": 3.3216,
+      "step": 8614
+    },
+    {
+      "epoch": 0.36273684210526314,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00036368549937232456,
+      "loss": 3.3272,
+      "step": 8615
+    },
+    {
+      "epoch": 0.36277894736842103,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003636554448958033,
+      "loss": 2.8462,
+      "step": 8616
+    },
+    {
+      "epoch": 0.36282105263157893,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0003636253883486134,
+      "loss": 3.3656,
+      "step": 8617
+    },
+    {
+      "epoch": 0.36286315789473683,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00036359532973130235,
+      "loss": 3.0729,
+      "step": 8618
+    },
+    {
+      "epoch": 0.36290526315789473,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0003635652690444178,
+      "loss": 3.1648,
+      "step": 8619
+    },
+    {
+      "epoch": 0.36294736842105263,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0003635352062885076,
+      "loss": 2.1842,
+      "step": 8620
+    },
+    {
+      "epoch": 0.36298947368421053,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003635051414641193,
+      "loss": 3.2677,
+      "step": 8621
+    },
+    {
+      "epoch": 0.3630315789473684,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003634750745718007,
+      "loss": 3.2,
+      "step": 8622
+    },
+    {
+      "epoch": 0.3630736842105263,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00036344500561209954,
+      "loss": 3.2294,
+      "step": 8623
+    },
+    {
+      "epoch": 0.3631157894736842,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00036341493458556367,
+      "loss": 3.1805,
+      "step": 8624
+    },
+    {
+      "epoch": 0.3631578947368421,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00036338486149274104,
+      "loss": 3.5624,
+      "step": 8625
+    },
+    {
+      "epoch": 0.3632,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00036335478633417933,
+      "loss": 3.3642,
+      "step": 8626
+    },
+    {
+      "epoch": 0.3632421052631579,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003633247091104266,
+      "loss": 2.8678,
+      "step": 8627
+    },
+    {
+      "epoch": 0.36328421052631577,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00036329462982203085,
+      "loss": 3.5098,
+      "step": 8628
+    },
+    {
+      "epoch": 0.36332631578947366,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003632645484695401,
+      "loss": 2.7913,
+      "step": 8629
+    },
+    {
+      "epoch": 0.36336842105263156,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00036323446505350237,
+      "loss": 3.3404,
+      "step": 8630
+    },
+    {
+      "epoch": 0.36341052631578946,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0003632043795744657,
+      "loss": 2.7892,
+      "step": 8631
+    },
+    {
+      "epoch": 0.36345263157894736,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003631742920329783,
+      "loss": 3.4458,
+      "step": 8632
+    },
+    {
+      "epoch": 0.36349473684210526,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003631442024295882,
+      "loss": 2.7049,
+      "step": 8633
+    },
+    {
+      "epoch": 0.36353684210526316,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00036311411076484363,
+      "loss": 2.7395,
+      "step": 8634
+    },
+    {
+      "epoch": 0.36357894736842106,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00036308401703929295,
+      "loss": 2.9941,
+      "step": 8635
+    },
+    {
+      "epoch": 0.36362105263157896,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003630539212534843,
+      "loss": 3.2984,
+      "step": 8636
+    },
+    {
+      "epoch": 0.36366315789473685,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0003630238234079661,
+      "loss": 2.9861,
+      "step": 8637
+    },
+    {
+      "epoch": 0.36370526315789475,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00036299372350328656,
+      "loss": 2.9727,
+      "step": 8638
+    },
+    {
+      "epoch": 0.36374736842105265,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00036296362153999417,
+      "loss": 2.8809,
+      "step": 8639
+    },
+    {
+      "epoch": 0.36378947368421055,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00036293351751863734,
+      "loss": 2.8221,
+      "step": 8640
+    },
+    {
+      "epoch": 0.3638315789473684,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00036290341143976445,
+      "loss": 3.0221,
+      "step": 8641
+    },
+    {
+      "epoch": 0.3638736842105263,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003628733033039241,
+      "loss": 3.253,
+      "step": 8642
+    },
+    {
+      "epoch": 0.3639157894736842,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00036284319311166485,
+      "loss": 3.3358,
+      "step": 8643
+    },
+    {
+      "epoch": 0.3639578947368421,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00036281308086353513,
+      "loss": 3.2004,
+      "step": 8644
+    },
+    {
+      "epoch": 0.364,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0003627829665600836,
+      "loss": 3.321,
+      "step": 8645
+    },
+    {
+      "epoch": 0.3640421052631579,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00036275285020185904,
+      "loss": 3.2841,
+      "step": 8646
+    },
+    {
+      "epoch": 0.3640842105263158,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00036272273178940994,
+      "loss": 2.6696,
+      "step": 8647
+    },
+    {
+      "epoch": 0.3641263157894737,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00036269261132328514,
+      "loss": 2.9027,
+      "step": 8648
+    },
+    {
+      "epoch": 0.3641684210526316,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003626624888040334,
+      "loss": 3.3783,
+      "step": 8649
+    },
+    {
+      "epoch": 0.3642105263157895,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003626323642322035,
+      "loss": 3.3822,
+      "step": 8650
+    },
+    {
+      "epoch": 0.3642526315789474,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003626022376083442,
+      "loss": 2.758,
+      "step": 8651
+    },
+    {
+      "epoch": 0.3642947368421053,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003625721089330044,
+      "loss": 3.4959,
+      "step": 8652
+    },
+    {
+      "epoch": 0.3643368421052632,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00036254197820673314,
+      "loss": 3.3613,
+      "step": 8653
+    },
+    {
+      "epoch": 0.3643789473684211,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0003625118454300791,
+      "loss": 3.1931,
+      "step": 8654
+    },
+    {
+      "epoch": 0.3644210526315789,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003624817106035916,
+      "loss": 3.2664,
+      "step": 8655
+    },
+    {
+      "epoch": 0.3644631578947368,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00036245157372781945,
+      "loss": 3.4686,
+      "step": 8656
+    },
+    {
+      "epoch": 0.3645052631578947,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003624214348033118,
+      "loss": 2.9575,
+      "step": 8657
+    },
+    {
+      "epoch": 0.3645473684210526,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00036239129383061765,
+      "loss": 3.1362,
+      "step": 8658
+    },
+    {
+      "epoch": 0.3645894736842105,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003623611508102862,
+      "loss": 3.1704,
+      "step": 8659
+    },
+    {
+      "epoch": 0.3646315789473684,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00036233100574286655,
+      "loss": 3.2556,
+      "step": 8660
+    },
+    {
+      "epoch": 0.3646736842105263,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.000362300858628908,
+      "loss": 3.0411,
+      "step": 8661
+    },
+    {
+      "epoch": 0.3647157894736842,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0003622707094689597,
+      "loss": 3.3357,
+      "step": 8662
+    },
+    {
+      "epoch": 0.3647578947368421,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.000362240558263571,
+      "loss": 2.7502,
+      "step": 8663
+    },
+    {
+      "epoch": 0.3648,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00036221040501329126,
+      "loss": 3.0725,
+      "step": 8664
+    },
+    {
+      "epoch": 0.3648421052631579,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00036218024971866977,
+      "loss": 3.4346,
+      "step": 8665
+    },
+    {
+      "epoch": 0.3648842105263158,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00036215009238025595,
+      "loss": 3.086,
+      "step": 8666
+    },
+    {
+      "epoch": 0.3649263157894737,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00036211993299859916,
+      "loss": 2.894,
+      "step": 8667
+    },
+    {
+      "epoch": 0.36496842105263155,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00036208977157424895,
+      "loss": 3.3428,
+      "step": 8668
+    },
+    {
+      "epoch": 0.36501052631578945,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003620596081077548,
+      "loss": 3.0775,
+      "step": 8669
+    },
+    {
+      "epoch": 0.36505263157894735,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00036202944259966627,
+      "loss": 3.5118,
+      "step": 8670
+    },
+    {
+      "epoch": 0.36509473684210525,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00036199927505053286,
+      "loss": 3.1989,
+      "step": 8671
+    },
+    {
+      "epoch": 0.36513684210526315,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00036196910546090435,
+      "loss": 3.3286,
+      "step": 8672
+    },
+    {
+      "epoch": 0.36517894736842105,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00036193893383133026,
+      "loss": 3.5424,
+      "step": 8673
+    },
+    {
+      "epoch": 0.36522105263157895,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00036190876016236034,
+      "loss": 3.1608,
+      "step": 8674
+    },
+    {
+      "epoch": 0.36526315789473685,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00036187858445454427,
+      "loss": 3.3762,
+      "step": 8675
+    },
+    {
+      "epoch": 0.36530526315789474,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003618484067084318,
+      "loss": 3.5689,
+      "step": 8676
+    },
+    {
+      "epoch": 0.36534736842105264,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003618182269245729,
+      "loss": 3.0055,
+      "step": 8677
+    },
+    {
+      "epoch": 0.36538947368421054,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00036178804510351715,
+      "loss": 3.6636,
+      "step": 8678
+    },
+    {
+      "epoch": 0.36543157894736844,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003617578612458146,
+      "loss": 2.9881,
+      "step": 8679
+    },
+    {
+      "epoch": 0.36547368421052634,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003617276753520152,
+      "loss": 3.3805,
+      "step": 8680
+    },
+    {
+      "epoch": 0.3655157894736842,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0003616974874226687,
+      "loss": 3.2967,
+      "step": 8681
+    },
+    {
+      "epoch": 0.3655578947368421,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0003616672974583253,
+      "loss": 3.2124,
+      "step": 8682
+    },
+    {
+      "epoch": 0.3656,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.000361637105459535,
+      "loss": 3.0296,
+      "step": 8683
+    },
+    {
+      "epoch": 0.3656421052631579,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003616069114268478,
+      "loss": 3.2064,
+      "step": 8684
+    },
+    {
+      "epoch": 0.3656842105263158,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0003615767153608137,
+      "loss": 3.2912,
+      "step": 8685
+    },
+    {
+      "epoch": 0.3657263157894737,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00036154651726198295,
+      "loss": 3.4403,
+      "step": 8686
+    },
+    {
+      "epoch": 0.3657684210526316,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00036151631713090583,
+      "loss": 3.2468,
+      "step": 8687
+    },
+    {
+      "epoch": 0.3658105263157895,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00036148611496813235,
+      "loss": 3.5002,
+      "step": 8688
+    },
+    {
+      "epoch": 0.3658526315789474,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00036145591077421283,
+      "loss": 3.6422,
+      "step": 8689
+    },
+    {
+      "epoch": 0.3658947368421053,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003614257045496976,
+      "loss": 3.3135,
+      "step": 8690
+    },
+    {
+      "epoch": 0.36593684210526317,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.000361395496295137,
+      "loss": 3.0861,
+      "step": 8691
+    },
+    {
+      "epoch": 0.36597894736842107,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003613652860110813,
+      "loss": 2.8707,
+      "step": 8692
+    },
+    {
+      "epoch": 0.36602105263157897,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.000361335073698081,
+      "loss": 3.4729,
+      "step": 8693
+    },
+    {
+      "epoch": 0.36606315789473687,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00036130485935668635,
+      "loss": 2.9446,
+      "step": 8694
+    },
+    {
+      "epoch": 0.3661052631578947,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003612746429874479,
+      "loss": 3.2595,
+      "step": 8695
+    },
+    {
+      "epoch": 0.3661473684210526,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003612444245909164,
+      "loss": 3.2128,
+      "step": 8696
+    },
+    {
+      "epoch": 0.3661894736842105,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00036121420416764205,
+      "loss": 3.0934,
+      "step": 8697
+    },
+    {
+      "epoch": 0.3662315789473684,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00036118398171817567,
+      "loss": 3.0088,
+      "step": 8698
+    },
+    {
+      "epoch": 0.3662736842105263,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00036115375724306775,
+      "loss": 3.1936,
+      "step": 8699
+    },
+    {
+      "epoch": 0.3663157894736842,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00036112353074286896,
+      "loss": 3.3108,
+      "step": 8700
+    },
+    {
+      "epoch": 0.3663578947368421,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003610933022181299,
+      "loss": 2.928,
+      "step": 8701
+    },
+    {
+      "epoch": 0.3664,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00036106307166940154,
+      "loss": 3.4794,
+      "step": 8702
+    },
+    {
+      "epoch": 0.3664421052631579,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003610328390972344,
+      "loss": 3.6009,
+      "step": 8703
+    },
+    {
+      "epoch": 0.3664842105263158,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00036100260450217946,
+      "loss": 3.2725,
+      "step": 8704
+    },
+    {
+      "epoch": 0.3665263157894737,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003609723678847875,
+      "loss": 3.0208,
+      "step": 8705
+    },
+    {
+      "epoch": 0.3665684210526316,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003609421292456094,
+      "loss": 3.5445,
+      "step": 8706
+    },
+    {
+      "epoch": 0.3666105263157895,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00036091188858519604,
+      "loss": 3.4127,
+      "step": 8707
+    },
+    {
+      "epoch": 0.36665263157894734,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00036088164590409834,
+      "loss": 3.227,
+      "step": 8708
+    },
+    {
+      "epoch": 0.36669473684210524,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00036085140120286735,
+      "loss": 3.3638,
+      "step": 8709
+    },
+    {
+      "epoch": 0.36673684210526314,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003608211544820541,
+      "loss": 3.5464,
+      "step": 8710
+    },
+    {
+      "epoch": 0.36677894736842104,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0003607909057422096,
+      "loss": 3.2785,
+      "step": 8711
+    },
+    {
+      "epoch": 0.36682105263157894,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00036076065498388495,
+      "loss": 2.791,
+      "step": 8712
+    },
+    {
+      "epoch": 0.36686315789473684,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0003607304022076314,
+      "loss": 3.5848,
+      "step": 8713
+    },
+    {
+      "epoch": 0.36690526315789473,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00036070014741399994,
+      "loss": 3.0988,
+      "step": 8714
+    },
+    {
+      "epoch": 0.36694736842105263,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00036066989060354185,
+      "loss": 3.0988,
+      "step": 8715
+    },
+    {
+      "epoch": 0.36698947368421053,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00036063963177680844,
+      "loss": 3.0324,
+      "step": 8716
+    },
+    {
+      "epoch": 0.36703157894736843,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00036060937093435085,
+      "loss": 3.4222,
+      "step": 8717
+    },
+    {
+      "epoch": 0.36707368421052633,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00036057910807672056,
+      "loss": 3.2787,
+      "step": 8718
+    },
+    {
+      "epoch": 0.36711578947368423,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00036054884320446884,
+      "loss": 3.0296,
+      "step": 8719
+    },
+    {
+      "epoch": 0.3671578947368421,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0003605185763181471,
+      "loss": 2.952,
+      "step": 8720
+    },
+    {
+      "epoch": 0.3672,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00036048830741830675,
+      "loss": 3.1313,
+      "step": 8721
+    },
+    {
+      "epoch": 0.36724210526315787,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003604580365054991,
+      "loss": 3.4035,
+      "step": 8722
+    },
+    {
+      "epoch": 0.36728421052631577,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.000360427763580276,
+      "loss": 3.351,
+      "step": 8723
+    },
+    {
+      "epoch": 0.36732631578947367,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003603974886431888,
+      "loss": 3.2386,
+      "step": 8724
+    },
+    {
+      "epoch": 0.36736842105263157,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.000360367211694789,
+      "loss": 3.19,
+      "step": 8725
+    },
+    {
+      "epoch": 0.36741052631578947,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00036033693273562826,
+      "loss": 3.3061,
+      "step": 8726
+    },
+    {
+      "epoch": 0.36745263157894736,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0003603066517662583,
+      "loss": 2.9887,
+      "step": 8727
+    },
+    {
+      "epoch": 0.36749473684210526,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003602763687872307,
+      "loss": 3.2417,
+      "step": 8728
+    },
+    {
+      "epoch": 0.36753684210526316,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003602460837990972,
+      "loss": 2.8331,
+      "step": 8729
+    },
+    {
+      "epoch": 0.36757894736842106,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003602157968024097,
+      "loss": 3.1964,
+      "step": 8730
+    },
+    {
+      "epoch": 0.36762105263157896,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003601855077977198,
+      "loss": 3.0692,
+      "step": 8731
+    },
+    {
+      "epoch": 0.36766315789473686,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003601552167855794,
+      "loss": 3.0501,
+      "step": 8732
+    },
+    {
+      "epoch": 0.36770526315789476,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003601249237665405,
+      "loss": 3.4019,
+      "step": 8733
+    },
+    {
+      "epoch": 0.36774736842105266,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003600946287411548,
+      "loss": 3.2833,
+      "step": 8734
+    },
+    {
+      "epoch": 0.3677894736842105,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003600643317099742,
+      "loss": 3.4603,
+      "step": 8735
+    },
+    {
+      "epoch": 0.3678315789473684,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00036003403267355095,
+      "loss": 3.5506,
+      "step": 8736
+    },
+    {
+      "epoch": 0.3678736842105263,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00036000373163243683,
+      "loss": 3.0973,
+      "step": 8737
+    },
+    {
+      "epoch": 0.3679157894736842,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.000359973428587184,
+      "loss": 2.8814,
+      "step": 8738
+    },
+    {
+      "epoch": 0.3679578947368421,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0003599431235383446,
+      "loss": 3.447,
+      "step": 8739
+    },
+    {
+      "epoch": 0.368,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00035991281648647055,
+      "loss": 3.1102,
+      "step": 8740
+    },
+    {
+      "epoch": 0.3680421052631579,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00035988250743211413,
+      "loss": 3.0448,
+      "step": 8741
+    },
+    {
+      "epoch": 0.3680842105263158,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00035985219637582757,
+      "loss": 3.1204,
+      "step": 8742
+    },
+    {
+      "epoch": 0.3681263157894737,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00035982188331816305,
+      "loss": 2.7306,
+      "step": 8743
+    },
+    {
+      "epoch": 0.3681684210526316,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00035979156825967286,
+      "loss": 3.6199,
+      "step": 8744
+    },
+    {
+      "epoch": 0.3682105263157895,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003597612512009093,
+      "loss": 3.3402,
+      "step": 8745
+    },
+    {
+      "epoch": 0.3682526315789474,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003597309321424247,
+      "loss": 3.5146,
+      "step": 8746
+    },
+    {
+      "epoch": 0.3682947368421053,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003597006110847714,
+      "loss": 3.0812,
+      "step": 8747
+    },
+    {
+      "epoch": 0.3683368421052632,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00035967028802850196,
+      "loss": 3.102,
+      "step": 8748
+    },
+    {
+      "epoch": 0.36837894736842103,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00035963996297416864,
+      "loss": 3.3346,
+      "step": 8749
+    },
+    {
+      "epoch": 0.3684210526315789,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.000359609635922324,
+      "loss": 2.6832,
+      "step": 8750
+    },
+    {
+      "epoch": 0.3684631578947368,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0003595793068735207,
+      "loss": 2.7353,
+      "step": 8751
+    },
+    {
+      "epoch": 0.3685052631578947,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003595489758283111,
+      "loss": 3.399,
+      "step": 8752
+    },
+    {
+      "epoch": 0.3685473684210526,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003595186427872479,
+      "loss": 3.1513,
+      "step": 8753
+    },
+    {
+      "epoch": 0.3685894736842105,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003594883077508836,
+      "loss": 3.0264,
+      "step": 8754
+    },
+    {
+      "epoch": 0.3686315789473684,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003594579707197711,
+      "loss": 3.2465,
+      "step": 8755
+    },
+    {
+      "epoch": 0.3686736842105263,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00035942763169446296,
+      "loss": 3.7956,
+      "step": 8756
+    },
+    {
+      "epoch": 0.3687157894736842,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003593972906755119,
+      "loss": 2.888,
+      "step": 8757
+    },
+    {
+      "epoch": 0.3687578947368421,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003593669476634708,
+      "loss": 3.3127,
+      "step": 8758
+    },
+    {
+      "epoch": 0.3688,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0003593366026588924,
+      "loss": 3.0411,
+      "step": 8759
+    },
+    {
+      "epoch": 0.3688421052631579,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00035930625566232957,
+      "loss": 3.2916,
+      "step": 8760
+    },
+    {
+      "epoch": 0.3688842105263158,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0003592759066743351,
+      "loss": 2.8771,
+      "step": 8761
+    },
+    {
+      "epoch": 0.36892631578947366,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003592455556954621,
+      "loss": 3.7729,
+      "step": 8762
+    },
+    {
+      "epoch": 0.36896842105263156,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003592152027262634,
+      "loss": 3.3187,
+      "step": 8763
+    },
+    {
+      "epoch": 0.36901052631578946,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00035918484776729204,
+      "loss": 3.5149,
+      "step": 8764
+    },
+    {
+      "epoch": 0.36905263157894735,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00035915449081910106,
+      "loss": 3.5412,
+      "step": 8765
+    },
+    {
+      "epoch": 0.36909473684210525,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00035912413188224346,
+      "loss": 3.0853,
+      "step": 8766
+    },
+    {
+      "epoch": 0.36913684210526315,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00035909377095727247,
+      "loss": 3.4828,
+      "step": 8767
+    },
+    {
+      "epoch": 0.36917894736842105,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00035906340804474113,
+      "loss": 3.1695,
+      "step": 8768
+    },
+    {
+      "epoch": 0.36922105263157895,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00035903304314520256,
+      "loss": 2.5928,
+      "step": 8769
+    },
+    {
+      "epoch": 0.36926315789473685,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00035900267625921006,
+      "loss": 3.471,
+      "step": 8770
+    },
+    {
+      "epoch": 0.36930526315789475,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.000358972307387317,
+      "loss": 2.885,
+      "step": 8771
+    },
+    {
+      "epoch": 0.36934736842105265,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0003589419365300764,
+      "loss": 3.4296,
+      "step": 8772
+    },
+    {
+      "epoch": 0.36938947368421055,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0003589115636880418,
+      "loss": 3.275,
+      "step": 8773
+    },
+    {
+      "epoch": 0.36943157894736844,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00035888118886176647,
+      "loss": 3.2415,
+      "step": 8774
+    },
+    {
+      "epoch": 0.36947368421052634,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00035885081205180375,
+      "loss": 2.7681,
+      "step": 8775
+    },
+    {
+      "epoch": 0.3695157894736842,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00035882043325870716,
+      "loss": 3.1312,
+      "step": 8776
+    },
+    {
+      "epoch": 0.3695578947368421,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00035879005248303013,
+      "loss": 3.0068,
+      "step": 8777
+    },
+    {
+      "epoch": 0.3696,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00035875966972532614,
+      "loss": 2.8719,
+      "step": 8778
+    },
+    {
+      "epoch": 0.3696421052631579,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00035872928498614884,
+      "loss": 3.0963,
+      "step": 8779
+    },
+    {
+      "epoch": 0.3696842105263158,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00035869889826605167,
+      "loss": 2.9659,
+      "step": 8780
+    },
+    {
+      "epoch": 0.3697263157894737,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00035866850956558827,
+      "loss": 3.1153,
+      "step": 8781
+    },
+    {
+      "epoch": 0.3697684210526316,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003586381188853123,
+      "loss": 2.7501,
+      "step": 8782
+    },
+    {
+      "epoch": 0.3698105263157895,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0003586077262257775,
+      "loss": 2.8393,
+      "step": 8783
+    },
+    {
+      "epoch": 0.3698526315789474,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00035857733158753754,
+      "loss": 2.7448,
+      "step": 8784
+    },
+    {
+      "epoch": 0.3698947368421053,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0003585469349711461,
+      "loss": 3.0044,
+      "step": 8785
+    },
+    {
+      "epoch": 0.3699368421052632,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003585165363771571,
+      "loss": 3.1144,
+      "step": 8786
+    },
+    {
+      "epoch": 0.3699789473684211,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00035848613580612425,
+      "loss": 2.8756,
+      "step": 8787
+    },
+    {
+      "epoch": 0.370021052631579,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0003584557332586016,
+      "loss": 3.1545,
+      "step": 8788
+    },
+    {
+      "epoch": 0.3700631578947368,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0003584253287351428,
+      "loss": 3.4403,
+      "step": 8789
+    },
+    {
+      "epoch": 0.3701052631578947,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0003583949222363019,
+      "loss": 3.062,
+      "step": 8790
+    },
+    {
+      "epoch": 0.3701473684210526,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.000358364513762633,
+      "loss": 3.5378,
+      "step": 8791
+    },
+    {
+      "epoch": 0.3701894736842105,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00035833410331468984,
+      "loss": 2.6829,
+      "step": 8792
+    },
+    {
+      "epoch": 0.3702315789473684,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00035830369089302674,
+      "loss": 3.058,
+      "step": 8793
+    },
+    {
+      "epoch": 0.3702736842105263,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00035827327649819754,
+      "loss": 2.8752,
+      "step": 8794
+    },
+    {
+      "epoch": 0.3703157894736842,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003582428601307565,
+      "loss": 3.3184,
+      "step": 8795
+    },
+    {
+      "epoch": 0.3703578947368421,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003582124417912576,
+      "loss": 3.3844,
+      "step": 8796
+    },
+    {
+      "epoch": 0.3704,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0003581820214802553,
+      "loss": 2.8874,
+      "step": 8797
+    },
+    {
+      "epoch": 0.3704421052631579,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003581515991983036,
+      "loss": 2.8709,
+      "step": 8798
+    },
+    {
+      "epoch": 0.3704842105263158,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00035812117494595695,
+      "loss": 2.9632,
+      "step": 8799
+    },
+    {
+      "epoch": 0.3705263157894737,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00035809074872376943,
+      "loss": 3.374,
+      "step": 8800
+    },
+    {
+      "epoch": 0.3705684210526316,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00035806032053229553,
+      "loss": 3.2579,
+      "step": 8801
+    },
+    {
+      "epoch": 0.3706105263157895,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003580298903720895,
+      "loss": 3.5015,
+      "step": 8802
+    },
+    {
+      "epoch": 0.37065263157894734,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00035799945824370585,
+      "loss": 3.026,
+      "step": 8803
+    },
+    {
+      "epoch": 0.37069473684210524,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003579690241476989,
+      "loss": 3.1322,
+      "step": 8804
+    },
+    {
+      "epoch": 0.37073684210526314,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003579385880846232,
+      "loss": 3.4497,
+      "step": 8805
+    },
+    {
+      "epoch": 0.37077894736842104,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003579081500550333,
+      "loss": 2.8693,
+      "step": 8806
+    },
+    {
+      "epoch": 0.37082105263157894,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0003578777100594837,
+      "loss": 3.2827,
+      "step": 8807
+    },
+    {
+      "epoch": 0.37086315789473684,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00035784726809852895,
+      "loss": 2.7358,
+      "step": 8808
+    },
+    {
+      "epoch": 0.37090526315789474,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00035781682417272367,
+      "loss": 3.3884,
+      "step": 8809
+    },
+    {
+      "epoch": 0.37094736842105264,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00035778637828262253,
+      "loss": 3.4693,
+      "step": 8810
+    },
+    {
+      "epoch": 0.37098947368421054,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00035775593042878026,
+      "loss": 3.2111,
+      "step": 8811
+    },
+    {
+      "epoch": 0.37103157894736843,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003577254806117515,
+      "loss": 3.6081,
+      "step": 8812
+    },
+    {
+      "epoch": 0.37107368421052633,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00035769502883209106,
+      "loss": 3.1627,
+      "step": 8813
+    },
+    {
+      "epoch": 0.37111578947368423,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00035766457509035375,
+      "loss": 3.1394,
+      "step": 8814
+    },
+    {
+      "epoch": 0.37115789473684213,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00035763411938709446,
+      "loss": 3.4173,
+      "step": 8815
+    },
+    {
+      "epoch": 0.3712,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00035760366172286786,
+      "loss": 3.6149,
+      "step": 8816
+    },
+    {
+      "epoch": 0.3712421052631579,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00035757320209822897,
+      "loss": 3.2791,
+      "step": 8817
+    },
+    {
+      "epoch": 0.37128421052631577,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0003575427405137328,
+      "loss": 3.0291,
+      "step": 8818
+    },
+    {
+      "epoch": 0.37132631578947367,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0003575122769699343,
+      "loss": 2.8175,
+      "step": 8819
+    },
+    {
+      "epoch": 0.37136842105263157,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0003574818114673883,
+      "loss": 3.1578,
+      "step": 8820
+    },
+    {
+      "epoch": 0.37141052631578947,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00035745134400665005,
+      "loss": 3.0817,
+      "step": 8821
+    },
+    {
+      "epoch": 0.37145263157894737,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003574208745882745,
+      "loss": 3.3079,
+      "step": 8822
+    },
+    {
+      "epoch": 0.37149473684210527,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003573904032128169,
+      "loss": 3.479,
+      "step": 8823
+    },
+    {
+      "epoch": 0.37153684210526317,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00035735992988083235,
+      "loss": 3.4404,
+      "step": 8824
+    },
+    {
+      "epoch": 0.37157894736842106,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00035732945459287596,
+      "loss": 2.8156,
+      "step": 8825
+    },
+    {
+      "epoch": 0.37162105263157896,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00035729897734950307,
+      "loss": 2.9612,
+      "step": 8826
+    },
+    {
+      "epoch": 0.37166315789473686,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003572684981512689,
+      "loss": 3.1853,
+      "step": 8827
+    },
+    {
+      "epoch": 0.37170526315789476,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003572380169987286,
+      "loss": 3.2269,
+      "step": 8828
+    },
+    {
+      "epoch": 0.3717473684210526,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00035720753389243776,
+      "loss": 2.9896,
+      "step": 8829
+    },
+    {
+      "epoch": 0.3717894736842105,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00035717704883295154,
+      "loss": 2.8069,
+      "step": 8830
+    },
+    {
+      "epoch": 0.3718315789473684,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00035714656182082546,
+      "loss": 3.3079,
+      "step": 8831
+    },
+    {
+      "epoch": 0.3718736842105263,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003571160728566149,
+      "loss": 3.0712,
+      "step": 8832
+    },
+    {
+      "epoch": 0.3719157894736842,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00035708558194087536,
+      "loss": 2.9352,
+      "step": 8833
+    },
+    {
+      "epoch": 0.3719578947368421,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003570550890741624,
+      "loss": 3.1976,
+      "step": 8834
+    },
+    {
+      "epoch": 0.372,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00035702459425703144,
+      "loss": 3.6019,
+      "step": 8835
+    },
+    {
+      "epoch": 0.3720421052631579,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0003569940974900381,
+      "loss": 2.9547,
+      "step": 8836
+    },
+    {
+      "epoch": 0.3720842105263158,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.000356963598773738,
+      "loss": 3.3466,
+      "step": 8837
+    },
+    {
+      "epoch": 0.3721263157894737,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003569330981086868,
+      "loss": 3.5103,
+      "step": 8838
+    },
+    {
+      "epoch": 0.3721684210526316,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00035690259549544026,
+      "loss": 3.7032,
+      "step": 8839
+    },
+    {
+      "epoch": 0.3722105263157895,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00035687209093455404,
+      "loss": 3.299,
+      "step": 8840
+    },
+    {
+      "epoch": 0.3722526315789474,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.0003568415844265839,
+      "loss": 3.2712,
+      "step": 8841
+    },
+    {
+      "epoch": 0.3722947368421053,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003568110759720855,
+      "loss": 3.3655,
+      "step": 8842
+    },
+    {
+      "epoch": 0.37233684210526313,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003567805655716149,
+      "loss": 3.0251,
+      "step": 8843
+    },
+    {
+      "epoch": 0.37237894736842103,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00035675005322572783,
+      "loss": 3.1441,
+      "step": 8844
+    },
+    {
+      "epoch": 0.37242105263157893,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0003567195389349802,
+      "loss": 3.2199,
+      "step": 8845
+    },
+    {
+      "epoch": 0.37246315789473683,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00035668902269992797,
+      "loss": 3.4215,
+      "step": 8846
+    },
+    {
+      "epoch": 0.37250526315789473,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003566585045211271,
+      "loss": 2.9532,
+      "step": 8847
+    },
+    {
+      "epoch": 0.3725473684210526,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0003566279843991337,
+      "loss": 3.8134,
+      "step": 8848
+    },
+    {
+      "epoch": 0.3725894736842105,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00035659746233450364,
+      "loss": 3.0871,
+      "step": 8849
+    },
+    {
+      "epoch": 0.3726315789473684,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00035656693832779296,
+      "loss": 3.2368,
+      "step": 8850
+    },
+    {
+      "epoch": 0.3726736842105263,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.000356536412379558,
+      "loss": 3.1628,
+      "step": 8851
+    },
+    {
+      "epoch": 0.3727157894736842,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003565058844903547,
+      "loss": 2.5716,
+      "step": 8852
+    },
+    {
+      "epoch": 0.3727578947368421,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00035647535466073945,
+      "loss": 3.3633,
+      "step": 8853
+    },
+    {
+      "epoch": 0.3728,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0003564448228912682,
+      "loss": 3.1276,
+      "step": 8854
+    },
+    {
+      "epoch": 0.3728421052631579,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003564142891824974,
+      "loss": 3.1487,
+      "step": 8855
+    },
+    {
+      "epoch": 0.37288421052631576,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0003563837535349832,
+      "loss": 3.1549,
+      "step": 8856
+    },
+    {
+      "epoch": 0.37292631578947366,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003563532159492821,
+      "loss": 3.3066,
+      "step": 8857
+    },
+    {
+      "epoch": 0.37296842105263156,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0003563226764259504,
+      "loss": 3.5371,
+      "step": 8858
+    },
+    {
+      "epoch": 0.37301052631578946,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003562921349655444,
+      "loss": 3.4258,
+      "step": 8859
+    },
+    {
+      "epoch": 0.37305263157894736,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00035626159156862066,
+      "loss": 3.2932,
+      "step": 8860
+    },
+    {
+      "epoch": 0.37309473684210526,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0003562310462357355,
+      "loss": 2.9712,
+      "step": 8861
+    },
+    {
+      "epoch": 0.37313684210526316,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003562004989674454,
+      "loss": 2.8239,
+      "step": 8862
+    },
+    {
+      "epoch": 0.37317894736842105,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003561699497643071,
+      "loss": 3.736,
+      "step": 8863
+    },
+    {
+      "epoch": 0.37322105263157895,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0003561393986268771,
+      "loss": 2.8324,
+      "step": 8864
+    },
+    {
+      "epoch": 0.37326315789473685,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00035610884555571193,
+      "loss": 3.107,
+      "step": 8865
+    },
+    {
+      "epoch": 0.37330526315789475,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00035607829055136824,
+      "loss": 3.4908,
+      "step": 8866
+    },
+    {
+      "epoch": 0.37334736842105265,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0003560477336144028,
+      "loss": 3.3465,
+      "step": 8867
+    },
+    {
+      "epoch": 0.37338947368421055,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00035601717474537223,
+      "loss": 3.2578,
+      "step": 8868
+    },
+    {
+      "epoch": 0.37343157894736845,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00035598661394483326,
+      "loss": 2.8672,
+      "step": 8869
+    },
+    {
+      "epoch": 0.3734736842105263,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00035595605121334275,
+      "loss": 3.0722,
+      "step": 8870
+    },
+    {
+      "epoch": 0.3735157894736842,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003559254865514574,
+      "loss": 3.2482,
+      "step": 8871
+    },
+    {
+      "epoch": 0.3735578947368421,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0003558949199597342,
+      "loss": 3.1095,
+      "step": 8872
+    },
+    {
+      "epoch": 0.3736,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00035586435143873,
+      "loss": 3.4075,
+      "step": 8873
+    },
+    {
+      "epoch": 0.3736421052631579,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0003558337809890017,
+      "loss": 2.998,
+      "step": 8874
+    },
+    {
+      "epoch": 0.3736842105263158,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0003558032086111063,
+      "loss": 2.9599,
+      "step": 8875
+    },
+    {
+      "epoch": 0.3737263157894737,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00035577263430560063,
+      "loss": 3.0893,
+      "step": 8876
+    },
+    {
+      "epoch": 0.3737684210526316,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00035574205807304196,
+      "loss": 3.311,
+      "step": 8877
+    },
+    {
+      "epoch": 0.3738105263157895,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00035571147991398717,
+      "loss": 3.6897,
+      "step": 8878
+    },
+    {
+      "epoch": 0.3738526315789474,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00035568089982899336,
+      "loss": 3.2627,
+      "step": 8879
+    },
+    {
+      "epoch": 0.3738947368421053,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00035565031781861783,
+      "loss": 3.6754,
+      "step": 8880
+    },
+    {
+      "epoch": 0.3739368421052632,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00035561973388341763,
+      "loss": 2.9702,
+      "step": 8881
+    },
+    {
+      "epoch": 0.3739789473684211,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00035558914802395,
+      "loss": 2.9022,
+      "step": 8882
+    },
+    {
+      "epoch": 0.3740210526315789,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00035555856024077203,
+      "loss": 3.0538,
+      "step": 8883
+    },
+    {
+      "epoch": 0.3740631578947368,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00035552797053444124,
+      "loss": 2.8616,
+      "step": 8884
+    },
+    {
+      "epoch": 0.3741052631578947,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003554973789055148,
+      "loss": 3.7329,
+      "step": 8885
+    },
+    {
+      "epoch": 0.3741473684210526,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003554667853545501,
+      "loss": 3.1809,
+      "step": 8886
+    },
+    {
+      "epoch": 0.3741894736842105,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00035543618988210433,
+      "loss": 3.1128,
+      "step": 8887
+    },
+    {
+      "epoch": 0.3742315789473684,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003554055924887352,
+      "loss": 3.4302,
+      "step": 8888
+    },
+    {
+      "epoch": 0.3742736842105263,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003553749931750001,
+      "loss": 3.4367,
+      "step": 8889
+    },
+    {
+      "epoch": 0.3743157894736842,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00035534439194145624,
+      "loss": 3.357,
+      "step": 8890
+    },
+    {
+      "epoch": 0.3743578947368421,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00035531378878866145,
+      "loss": 3.2225,
+      "step": 8891
+    },
+    {
+      "epoch": 0.3744,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0003552831837171731,
+      "loss": 2.9735,
+      "step": 8892
+    },
+    {
+      "epoch": 0.3744421052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00035525257672754893,
+      "loss": 3.0731,
+      "step": 8893
+    },
+    {
+      "epoch": 0.3744842105263158,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00035522196782034645,
+      "loss": 3.1161,
+      "step": 8894
+    },
+    {
+      "epoch": 0.3745263157894737,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00035519135699612324,
+      "loss": 3.1231,
+      "step": 8895
+    },
+    {
+      "epoch": 0.3745684210526316,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00035516074425543726,
+      "loss": 3.0028,
+      "step": 8896
+    },
+    {
+      "epoch": 0.37461052631578945,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003551301295988459,
+      "loss": 3.2266,
+      "step": 8897
+    },
+    {
+      "epoch": 0.37465263157894735,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003550995130269073,
+      "loss": 3.2177,
+      "step": 8898
+    },
+    {
+      "epoch": 0.37469473684210525,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.000355068894540179,
+      "loss": 2.9098,
+      "step": 8899
+    },
+    {
+      "epoch": 0.37473684210526315,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003550382741392189,
+      "loss": 2.9361,
+      "step": 8900
+    },
+    {
+      "epoch": 0.37477894736842104,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00035500765182458477,
+      "loss": 3.4568,
+      "step": 8901
+    },
+    {
+      "epoch": 0.37482105263157894,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0003549770275968347,
+      "loss": 3.3822,
+      "step": 8902
+    },
+    {
+      "epoch": 0.37486315789473684,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00035494640145652647,
+      "loss": 3.5677,
+      "step": 8903
+    },
+    {
+      "epoch": 0.37490526315789474,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0003549157734042181,
+      "loss": 3.1161,
+      "step": 8904
+    },
+    {
+      "epoch": 0.37494736842105264,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003548851434404676,
+      "loss": 2.6879,
+      "step": 8905
+    },
+    {
+      "epoch": 0.37498947368421054,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003548545115658331,
+      "loss": 2.9278,
+      "step": 8906
+    },
+    {
+      "epoch": 0.37503157894736844,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00035482387778087255,
+      "loss": 3.1649,
+      "step": 8907
+    },
+    {
+      "epoch": 0.37507368421052634,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00035479324208614416,
+      "loss": 3.5147,
+      "step": 8908
+    },
+    {
+      "epoch": 0.37511578947368424,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.000354762604482206,
+      "loss": 3.3122,
+      "step": 8909
+    },
+    {
+      "epoch": 0.3751578947368421,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0003547319649696162,
+      "loss": 2.9256,
+      "step": 8910
+    },
+    {
+      "epoch": 0.3752,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00035470132354893316,
+      "loss": 2.842,
+      "step": 8911
+    },
+    {
+      "epoch": 0.3752421052631579,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00035467068022071495,
+      "loss": 3.192,
+      "step": 8912
+    },
+    {
+      "epoch": 0.3752842105263158,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0003546400349855199,
+      "loss": 3.364,
+      "step": 8913
+    },
+    {
+      "epoch": 0.3753263157894737,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00035460938784390644,
+      "loss": 3.4077,
+      "step": 8914
+    },
+    {
+      "epoch": 0.3753684210526316,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00035457873879643286,
+      "loss": 3.2066,
+      "step": 8915
+    },
+    {
+      "epoch": 0.3754105263157895,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00035454808784365747,
+      "loss": 2.8645,
+      "step": 8916
+    },
+    {
+      "epoch": 0.37545263157894737,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00035451743498613874,
+      "loss": 3.0542,
+      "step": 8917
+    },
+    {
+      "epoch": 0.37549473684210527,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00035448678022443524,
+      "loss": 3.0786,
+      "step": 8918
+    },
+    {
+      "epoch": 0.37553684210526317,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00035445612355910534,
+      "loss": 2.6355,
+      "step": 8919
+    },
+    {
+      "epoch": 0.37557894736842107,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0003544254649907076,
+      "loss": 3.3552,
+      "step": 8920
+    },
+    {
+      "epoch": 0.37562105263157897,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00035439480451980056,
+      "loss": 3.0728,
+      "step": 8921
+    },
+    {
+      "epoch": 0.37566315789473687,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0003543641421469428,
+      "loss": 2.9536,
+      "step": 8922
+    },
+    {
+      "epoch": 0.37570526315789476,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003543334778726931,
+      "loss": 3.5309,
+      "step": 8923
+    },
+    {
+      "epoch": 0.3757473684210526,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0003543028116976099,
+      "loss": 3.4626,
+      "step": 8924
+    },
+    {
+      "epoch": 0.3757894736842105,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0003542721436222521,
+      "loss": 3.2538,
+      "step": 8925
+    },
+    {
+      "epoch": 0.3758315789473684,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003542414736471783,
+      "loss": 3.2549,
+      "step": 8926
+    },
+    {
+      "epoch": 0.3758736842105263,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0003542108017729474,
+      "loss": 3.0534,
+      "step": 8927
+    },
+    {
+      "epoch": 0.3759157894736842,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0003541801280001181,
+      "loss": 3.0592,
+      "step": 8928
+    },
+    {
+      "epoch": 0.3759578947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00035414945232924925,
+      "loss": 3.0419,
+      "step": 8929
+    },
+    {
+      "epoch": 0.376,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00035411877476089975,
+      "loss": 3.4069,
+      "step": 8930
+    },
+    {
+      "epoch": 0.3760421052631579,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00035408809529562855,
+      "loss": 3.0436,
+      "step": 8931
+    },
+    {
+      "epoch": 0.3760842105263158,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003540574139339945,
+      "loss": 3.5367,
+      "step": 8932
+    },
+    {
+      "epoch": 0.3761263157894737,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0003540267306765567,
+      "loss": 3.0923,
+      "step": 8933
+    },
+    {
+      "epoch": 0.3761684210526316,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003539960455238741,
+      "loss": 3.1992,
+      "step": 8934
+    },
+    {
+      "epoch": 0.3762105263157895,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00035396535847650567,
+      "loss": 3.3281,
+      "step": 8935
+    },
+    {
+      "epoch": 0.3762526315789474,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00035393466953501063,
+      "loss": 3.6561,
+      "step": 8936
+    },
+    {
+      "epoch": 0.37629473684210524,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00035390397869994794,
+      "loss": 3.1228,
+      "step": 8937
+    },
+    {
+      "epoch": 0.37633684210526314,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0003538732859718769,
+      "loss": 3.327,
+      "step": 8938
+    },
+    {
+      "epoch": 0.37637894736842104,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00035384259135135665,
+      "loss": 3.2079,
+      "step": 8939
+    },
+    {
+      "epoch": 0.37642105263157893,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00035381189483894637,
+      "loss": 3.4769,
+      "step": 8940
+    },
+    {
+      "epoch": 0.37646315789473683,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00035378119643520535,
+      "loss": 3.3818,
+      "step": 8941
+    },
+    {
+      "epoch": 0.37650526315789473,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0003537504961406929,
+      "loss": 3.0938,
+      "step": 8942
+    },
+    {
+      "epoch": 0.37654736842105263,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00035371979395596825,
+      "loss": 3.1622,
+      "step": 8943
+    },
+    {
+      "epoch": 0.37658947368421053,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0003536890898815908,
+      "loss": 3.1466,
+      "step": 8944
+    },
+    {
+      "epoch": 0.37663157894736843,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00035365838391812,
+      "loss": 3.024,
+      "step": 8945
+    },
+    {
+      "epoch": 0.3766736842105263,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00035362767606611533,
+      "loss": 3.4037,
+      "step": 8946
+    },
+    {
+      "epoch": 0.3767157894736842,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00035359696632613605,
+      "loss": 2.6377,
+      "step": 8947
+    },
+    {
+      "epoch": 0.3767578947368421,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003535662546987418,
+      "loss": 3.0471,
+      "step": 8948
+    },
+    {
+      "epoch": 0.3768,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00035353554118449206,
+      "loss": 3.1841,
+      "step": 8949
+    },
+    {
+      "epoch": 0.37684210526315787,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00035350482578394636,
+      "loss": 3.2254,
+      "step": 8950
+    },
+    {
+      "epoch": 0.37688421052631577,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0003534741084976644,
+      "loss": 3.1085,
+      "step": 8951
+    },
+    {
+      "epoch": 0.37692631578947366,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00035344338932620577,
+      "loss": 3.1488,
+      "step": 8952
+    },
+    {
+      "epoch": 0.37696842105263156,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0003534126682701302,
+      "loss": 2.7812,
+      "step": 8953
+    },
+    {
+      "epoch": 0.37701052631578946,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003533819453299972,
+      "loss": 3.2079,
+      "step": 8954
+    },
+    {
+      "epoch": 0.37705263157894736,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00035335122050636665,
+      "loss": 3.3812,
+      "step": 8955
+    },
+    {
+      "epoch": 0.37709473684210526,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0003533204937997983,
+      "loss": 3.0381,
+      "step": 8956
+    },
+    {
+      "epoch": 0.37713684210526316,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00035328976521085194,
+      "loss": 3.0779,
+      "step": 8957
+    },
+    {
+      "epoch": 0.37717894736842106,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003532590347400874,
+      "loss": 3.3667,
+      "step": 8958
+    },
+    {
+      "epoch": 0.37722105263157896,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00035322830238806465,
+      "loss": 3.2379,
+      "step": 8959
+    },
+    {
+      "epoch": 0.37726315789473686,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0003531975681553435,
+      "loss": 3.5144,
+      "step": 8960
+    },
+    {
+      "epoch": 0.37730526315789475,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0003531668320424839,
+      "loss": 3.6362,
+      "step": 8961
+    },
+    {
+      "epoch": 0.37734736842105265,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00035313609405004584,
+      "loss": 2.9789,
+      "step": 8962
+    },
+    {
+      "epoch": 0.37738947368421055,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003531053541785892,
+      "loss": 3.0765,
+      "step": 8963
+    },
+    {
+      "epoch": 0.3774315789473684,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003530746124286743,
+      "loss": 3.0762,
+      "step": 8964
+    },
+    {
+      "epoch": 0.3774736842105263,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00035304386880086105,
+      "loss": 3.2973,
+      "step": 8965
+    },
+    {
+      "epoch": 0.3775157894736842,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003530131232957096,
+      "loss": 2.9986,
+      "step": 8966
+    },
+    {
+      "epoch": 0.3775578947368421,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00035298237591378003,
+      "loss": 2.6886,
+      "step": 8967
+    },
+    {
+      "epoch": 0.3776,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00035295162665563255,
+      "loss": 3.7233,
+      "step": 8968
+    },
+    {
+      "epoch": 0.3776421052631579,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00035292087552182746,
+      "loss": 2.8755,
+      "step": 8969
+    },
+    {
+      "epoch": 0.3776842105263158,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003528901225129249,
+      "loss": 3.265,
+      "step": 8970
+    },
+    {
+      "epoch": 0.3777263157894737,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0003528593676294852,
+      "loss": 3.3299,
+      "step": 8971
+    },
+    {
+      "epoch": 0.3777684210526316,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00035282861087206866,
+      "loss": 3.2227,
+      "step": 8972
+    },
+    {
+      "epoch": 0.3778105263157895,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00035279785224123566,
+      "loss": 3.2662,
+      "step": 8973
+    },
+    {
+      "epoch": 0.3778526315789474,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00035276709173754663,
+      "loss": 2.9549,
+      "step": 8974
+    },
+    {
+      "epoch": 0.3778947368421053,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003527363293615619,
+      "loss": 3.0989,
+      "step": 8975
+    },
+    {
+      "epoch": 0.3779368421052632,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00035270556511384193,
+      "loss": 3.15,
+      "step": 8976
+    },
+    {
+      "epoch": 0.377978947368421,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0003526747989949473,
+      "loss": 3.1652,
+      "step": 8977
+    },
+    {
+      "epoch": 0.3780210526315789,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0003526440310054384,
+      "loss": 2.8999,
+      "step": 8978
+    },
+    {
+      "epoch": 0.3780631578947368,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00035261326114587586,
+      "loss": 3.601,
+      "step": 8979
+    },
+    {
+      "epoch": 0.3781052631578947,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00035258248941682027,
+      "loss": 2.7751,
+      "step": 8980
+    },
+    {
+      "epoch": 0.3781473684210526,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00035255171581883226,
+      "loss": 3.3878,
+      "step": 8981
+    },
+    {
+      "epoch": 0.3781894736842105,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003525209403524726,
+      "loss": 3.5145,
+      "step": 8982
+    },
+    {
+      "epoch": 0.3782315789473684,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0003524901630183017,
+      "loss": 3.0004,
+      "step": 8983
+    },
+    {
+      "epoch": 0.3782736842105263,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00035245938381688047,
+      "loss": 2.7245,
+      "step": 8984
+    },
+    {
+      "epoch": 0.3783157894736842,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00035242860274876974,
+      "loss": 3.0773,
+      "step": 8985
+    },
+    {
+      "epoch": 0.3783578947368421,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003523978198145302,
+      "loss": 3.2519,
+      "step": 8986
+    },
+    {
+      "epoch": 0.3784,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00035236703501472264,
+      "loss": 3.3628,
+      "step": 8987
+    },
+    {
+      "epoch": 0.3784421052631579,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00035233624834990794,
+      "loss": 3.2578,
+      "step": 8988
+    },
+    {
+      "epoch": 0.3784842105263158,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00035230545982064713,
+      "loss": 3.214,
+      "step": 8989
+    },
+    {
+      "epoch": 0.3785263157894737,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0003522746694275011,
+      "loss": 3.3579,
+      "step": 8990
+    },
+    {
+      "epoch": 0.37856842105263155,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003522438771710306,
+      "loss": 2.9936,
+      "step": 8991
+    },
+    {
+      "epoch": 0.37861052631578945,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00035221308305179687,
+      "loss": 3.5256,
+      "step": 8992
+    },
+    {
+      "epoch": 0.37865263157894735,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00035218228707036085,
+      "loss": 2.9523,
+      "step": 8993
+    },
+    {
+      "epoch": 0.37869473684210525,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00035215148922728367,
+      "loss": 3.5945,
+      "step": 8994
+    },
+    {
+      "epoch": 0.37873684210526315,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00035212068952312635,
+      "loss": 2.8541,
+      "step": 8995
+    },
+    {
+      "epoch": 0.37877894736842105,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00035208988795844997,
+      "loss": 2.8665,
+      "step": 8996
+    },
+    {
+      "epoch": 0.37882105263157895,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003520590845338159,
+      "loss": 3.259,
+      "step": 8997
+    },
+    {
+      "epoch": 0.37886315789473685,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00035202827924978513,
+      "loss": 3.2115,
+      "step": 8998
+    },
+    {
+      "epoch": 0.37890526315789474,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0003519974721069191,
+      "loss": 3.0475,
+      "step": 8999
+    },
+    {
+      "epoch": 0.37894736842105264,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0003519666631057789,
+      "loss": 2.898,
+      "step": 9000
+    },
+    {
+      "epoch": 0.37894736842105264,
+      "eval_loss": 3.1893510818481445,
+      "eval_runtime": 335.4664,
+      "eval_samples_per_second": 44.714,
+      "eval_steps_per_second": 5.589,
+      "step": 9000
+    },
+    {
+      "epoch": 0.37898947368421054,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003519358522469259,
+      "loss": 3.2336,
+      "step": 9001
+    },
+    {
+      "epoch": 0.37903157894736844,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00035190503953092146,
+      "loss": 3.3474,
+      "step": 9002
+    },
+    {
+      "epoch": 0.37907368421052634,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.000351874224958327,
+      "loss": 3.1011,
+      "step": 9003
+    },
+    {
+      "epoch": 0.3791157894736842,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0003518434085297037,
+      "loss": 2.8481,
+      "step": 9004
+    },
+    {
+      "epoch": 0.3791578947368421,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0003518125902456133,
+      "loss": 3.0454,
+      "step": 9005
+    },
+    {
+      "epoch": 0.3792,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.000351781770106617,
+      "loss": 3.114,
+      "step": 9006
+    },
+    {
+      "epoch": 0.3792421052631579,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00035175094811327654,
+      "loss": 3.1363,
+      "step": 9007
+    },
+    {
+      "epoch": 0.3792842105263158,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00035172012426615333,
+      "loss": 3.0914,
+      "step": 9008
+    },
+    {
+      "epoch": 0.3793263157894737,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.000351689298565809,
+      "loss": 3.4058,
+      "step": 9009
+    },
+    {
+      "epoch": 0.3793684210526316,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003516584710128051,
+      "loss": 3.2226,
+      "step": 9010
+    },
+    {
+      "epoch": 0.3794105263157895,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003516276416077033,
+      "loss": 3.2155,
+      "step": 9011
+    },
+    {
+      "epoch": 0.3794526315789474,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003515968103510653,
+      "loss": 3.1663,
+      "step": 9012
+    },
+    {
+      "epoch": 0.3794947368421053,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00035156597724345274,
+      "loss": 2.9861,
+      "step": 9013
+    },
+    {
+      "epoch": 0.3795368421052632,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00035153514228542735,
+      "loss": 3.4498,
+      "step": 9014
+    },
+    {
+      "epoch": 0.37957894736842107,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003515043054775511,
+      "loss": 3.0106,
+      "step": 9015
+    },
+    {
+      "epoch": 0.37962105263157897,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00035147346682038567,
+      "loss": 3.119,
+      "step": 9016
+    },
+    {
+      "epoch": 0.37966315789473687,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003514426263144929,
+      "loss": 3.455,
+      "step": 9017
+    },
+    {
+      "epoch": 0.3797052631578947,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00035141178396043453,
+      "loss": 2.9395,
+      "step": 9018
+    },
+    {
+      "epoch": 0.3797473684210526,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00035138093975877274,
+      "loss": 3.0246,
+      "step": 9019
+    },
+    {
+      "epoch": 0.3797894736842105,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00035135009371006933,
+      "loss": 3.1106,
+      "step": 9020
+    },
+    {
+      "epoch": 0.3798315789473684,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003513192458148863,
+      "loss": 3.452,
+      "step": 9021
+    },
+    {
+      "epoch": 0.3798736842105263,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00035128839607378553,
+      "loss": 3.1534,
+      "step": 9022
+    },
+    {
+      "epoch": 0.3799157894736842,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003512575444873293,
+      "loss": 2.9802,
+      "step": 9023
+    },
+    {
+      "epoch": 0.3799578947368421,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003512266910560795,
+      "loss": 3.1454,
+      "step": 9024
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00035119583578059843,
+      "loss": 3.2772,
+      "step": 9025
+    },
+    {
+      "epoch": 0.3800421052631579,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00035116497866144816,
+      "loss": 3.2466,
+      "step": 9026
+    },
+    {
+      "epoch": 0.3800842105263158,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00035113411969919073,
+      "loss": 3.2412,
+      "step": 9027
+    },
+    {
+      "epoch": 0.3801263157894737,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00035110325889438853,
+      "loss": 3.1463,
+      "step": 9028
+    },
+    {
+      "epoch": 0.3801684210526316,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003510723962476038,
+      "loss": 3.0792,
+      "step": 9029
+    },
+    {
+      "epoch": 0.3802105263157895,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00035104153175939863,
+      "loss": 3.2869,
+      "step": 9030
+    },
+    {
+      "epoch": 0.38025263157894734,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00035101066543033547,
+      "loss": 3.1292,
+      "step": 9031
+    },
+    {
+      "epoch": 0.38029473684210524,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00035097979726097676,
+      "loss": 3.319,
+      "step": 9032
+    },
+    {
+      "epoch": 0.38033684210526314,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00035094892725188483,
+      "loss": 3.203,
+      "step": 9033
+    },
+    {
+      "epoch": 0.38037894736842104,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.000350918055403622,
+      "loss": 2.8066,
+      "step": 9034
+    },
+    {
+      "epoch": 0.38042105263157894,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003508871817167508,
+      "loss": 3.0446,
+      "step": 9035
+    },
+    {
+      "epoch": 0.38046315789473684,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00035085630619183364,
+      "loss": 3.2608,
+      "step": 9036
+    },
+    {
+      "epoch": 0.38050526315789474,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00035082542882943315,
+      "loss": 2.8483,
+      "step": 9037
+    },
+    {
+      "epoch": 0.38054736842105263,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0003507945496301117,
+      "loss": 3.5058,
+      "step": 9038
+    },
+    {
+      "epoch": 0.38058947368421053,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00035076366859443196,
+      "loss": 3.3024,
+      "step": 9039
+    },
+    {
+      "epoch": 0.38063157894736843,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00035073278572295667,
+      "loss": 3.0329,
+      "step": 9040
+    },
+    {
+      "epoch": 0.38067368421052633,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003507019010162484,
+      "loss": 3.5822,
+      "step": 9041
+    },
+    {
+      "epoch": 0.38071578947368423,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00035067101447486975,
+      "loss": 3.3362,
+      "step": 9042
+    },
+    {
+      "epoch": 0.38075789473684213,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003506401260993836,
+      "loss": 3.1626,
+      "step": 9043
+    },
+    {
+      "epoch": 0.3808,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00035060923589035254,
+      "loss": 3.3958,
+      "step": 9044
+    },
+    {
+      "epoch": 0.38084210526315787,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00035057834384833933,
+      "loss": 2.8822,
+      "step": 9045
+    },
+    {
+      "epoch": 0.38088421052631577,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.000350547449973907,
+      "loss": 3.2318,
+      "step": 9046
+    },
+    {
+      "epoch": 0.38092631578947367,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0003505165542676181,
+      "loss": 3.1886,
+      "step": 9047
+    },
+    {
+      "epoch": 0.38096842105263157,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00035048565673003585,
+      "loss": 3.1682,
+      "step": 9048
+    },
+    {
+      "epoch": 0.38101052631578947,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00035045475736172297,
+      "loss": 2.963,
+      "step": 9049
+    },
+    {
+      "epoch": 0.38105263157894737,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003504238561632424,
+      "loss": 3.1009,
+      "step": 9050
+    },
+    {
+      "epoch": 0.38109473684210526,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003503929531351571,
+      "loss": 2.9971,
+      "step": 9051
+    },
+    {
+      "epoch": 0.38113684210526316,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00035036204827803023,
+      "loss": 3.1567,
+      "step": 9052
+    },
+    {
+      "epoch": 0.38117894736842106,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003503311415924248,
+      "loss": 3.1449,
+      "step": 9053
+    },
+    {
+      "epoch": 0.38122105263157896,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00035030023307890384,
+      "loss": 3.3972,
+      "step": 9054
+    },
+    {
+      "epoch": 0.38126315789473686,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003502693227380304,
+      "loss": 2.9572,
+      "step": 9055
+    },
+    {
+      "epoch": 0.38130526315789476,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0003502384105703678,
+      "loss": 3.136,
+      "step": 9056
+    },
+    {
+      "epoch": 0.38134736842105266,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003502074965764791,
+      "loss": 3.204,
+      "step": 9057
+    },
+    {
+      "epoch": 0.3813894736842105,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003501765807569275,
+      "loss": 3.1345,
+      "step": 9058
+    },
+    {
+      "epoch": 0.3814315789473684,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00035014566311227635,
+      "loss": 3.192,
+      "step": 9059
+    },
+    {
+      "epoch": 0.3814736842105263,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00035011474364308895,
+      "loss": 3.2295,
+      "step": 9060
+    },
+    {
+      "epoch": 0.3815157894736842,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003500838223499285,
+      "loss": 3.1093,
+      "step": 9061
+    },
+    {
+      "epoch": 0.3815578947368421,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00035005289923335835,
+      "loss": 3.4212,
+      "step": 9062
+    },
+    {
+      "epoch": 0.3816,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.000350021974293942,
+      "loss": 2.9013,
+      "step": 9063
+    },
+    {
+      "epoch": 0.3816421052631579,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00034999104753224275,
+      "loss": 2.9509,
+      "step": 9064
+    },
+    {
+      "epoch": 0.3816842105263158,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00034996011894882406,
+      "loss": 3.2612,
+      "step": 9065
+    },
+    {
+      "epoch": 0.3817263157894737,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00034992918854424957,
+      "loss": 3.2508,
+      "step": 9066
+    },
+    {
+      "epoch": 0.3817684210526316,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003498982563190826,
+      "loss": 3.2259,
+      "step": 9067
+    },
+    {
+      "epoch": 0.3818105263157895,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00034986732227388685,
+      "loss": 3.4245,
+      "step": 9068
+    },
+    {
+      "epoch": 0.3818526315789474,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00034983638640922576,
+      "loss": 2.9004,
+      "step": 9069
+    },
+    {
+      "epoch": 0.3818947368421053,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.000349805448725663,
+      "loss": 2.8261,
+      "step": 9070
+    },
+    {
+      "epoch": 0.38193684210526313,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003497745092237623,
+      "loss": 2.9816,
+      "step": 9071
+    },
+    {
+      "epoch": 0.38197894736842103,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00034974356790408716,
+      "loss": 2.7078,
+      "step": 9072
+    },
+    {
+      "epoch": 0.38202105263157893,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003497126247672014,
+      "loss": 3.0877,
+      "step": 9073
+    },
+    {
+      "epoch": 0.3820631578947368,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003496816798136689,
+      "loss": 3.2347,
+      "step": 9074
+    },
+    {
+      "epoch": 0.3821052631578947,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00034965073304405326,
+      "loss": 3.5157,
+      "step": 9075
+    },
+    {
+      "epoch": 0.3821473684210526,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003496197844589184,
+      "loss": 3.0099,
+      "step": 9076
+    },
+    {
+      "epoch": 0.3821894736842105,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003495888340588281,
+      "loss": 3.3364,
+      "step": 9077
+    },
+    {
+      "epoch": 0.3822315789473684,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00034955788184434615,
+      "loss": 2.9762,
+      "step": 9078
+    },
+    {
+      "epoch": 0.3822736842105263,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003495269278160367,
+      "loss": 3.3684,
+      "step": 9079
+    },
+    {
+      "epoch": 0.3823157894736842,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0003494959719744635,
+      "loss": 2.8853,
+      "step": 9080
+    },
+    {
+      "epoch": 0.3823578947368421,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003494650143201905,
+      "loss": 3.4443,
+      "step": 9081
+    },
+    {
+      "epoch": 0.3824,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.000349434054853782,
+      "loss": 2.6348,
+      "step": 9082
+    },
+    {
+      "epoch": 0.3824421052631579,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003494030935758017,
+      "loss": 2.9665,
+      "step": 9083
+    },
+    {
+      "epoch": 0.3824842105263158,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00034937213048681393,
+      "loss": 3.072,
+      "step": 9084
+    },
+    {
+      "epoch": 0.38252631578947366,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0003493411655873826,
+      "loss": 3.4029,
+      "step": 9085
+    },
+    {
+      "epoch": 0.38256842105263156,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.000349310198878072,
+      "loss": 2.9327,
+      "step": 9086
+    },
+    {
+      "epoch": 0.38261052631578946,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00034927923035944634,
+      "loss": 2.8851,
+      "step": 9087
+    },
+    {
+      "epoch": 0.38265263157894736,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003492482600320697,
+      "loss": 3.0079,
+      "step": 9088
+    },
+    {
+      "epoch": 0.38269473684210525,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0003492172878965063,
+      "loss": 3.5401,
+      "step": 9089
+    },
+    {
+      "epoch": 0.38273684210526315,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00034918631395332056,
+      "loss": 3.0147,
+      "step": 9090
+    },
+    {
+      "epoch": 0.38277894736842105,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003491553382030767,
+      "loss": 3.0488,
+      "step": 9091
+    },
+    {
+      "epoch": 0.38282105263157895,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.000349124360646339,
+      "loss": 2.9225,
+      "step": 9092
+    },
+    {
+      "epoch": 0.38286315789473685,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003490933812836721,
+      "loss": 3.4416,
+      "step": 9093
+    },
+    {
+      "epoch": 0.38290526315789475,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00034906240011564003,
+      "loss": 3.678,
+      "step": 9094
+    },
+    {
+      "epoch": 0.38294736842105265,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003490314171428076,
+      "loss": 2.9728,
+      "step": 9095
+    },
+    {
+      "epoch": 0.38298947368421055,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00034900043236573905,
+      "loss": 2.9566,
+      "step": 9096
+    },
+    {
+      "epoch": 0.38303157894736845,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00034896944578499884,
+      "loss": 3.2297,
+      "step": 9097
+    },
+    {
+      "epoch": 0.3830736842105263,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00034893845740115165,
+      "loss": 3.1214,
+      "step": 9098
+    },
+    {
+      "epoch": 0.3831157894736842,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0003489074672147621,
+      "loss": 3.4199,
+      "step": 9099
+    },
+    {
+      "epoch": 0.3831578947368421,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00034887647522639467,
+      "loss": 3.1668,
+      "step": 9100
+    },
+    {
+      "epoch": 0.3832,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00034884548143661396,
+      "loss": 3.2169,
+      "step": 9101
+    },
+    {
+      "epoch": 0.3832421052631579,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00034881448584598483,
+      "loss": 3.3098,
+      "step": 9102
+    },
+    {
+      "epoch": 0.3832842105263158,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003487834884550718,
+      "loss": 2.8624,
+      "step": 9103
+    },
+    {
+      "epoch": 0.3833263157894737,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00034875248926443975,
+      "loss": 2.9868,
+      "step": 9104
+    },
+    {
+      "epoch": 0.3833684210526316,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00034872148827465326,
+      "loss": 3.3611,
+      "step": 9105
+    },
+    {
+      "epoch": 0.3834105263157895,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00034869048548627734,
+      "loss": 3.4054,
+      "step": 9106
+    },
+    {
+      "epoch": 0.3834526315789474,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003486594808998767,
+      "loss": 2.9441,
+      "step": 9107
+    },
+    {
+      "epoch": 0.3834947368421053,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003486284745160162,
+      "loss": 3.1384,
+      "step": 9108
+    },
+    {
+      "epoch": 0.3835368421052632,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00034859746633526085,
+      "loss": 3.2456,
+      "step": 9109
+    },
+    {
+      "epoch": 0.3835789473684211,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003485664563581755,
+      "loss": 3.0276,
+      "step": 9110
+    },
+    {
+      "epoch": 0.383621052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003485354445853252,
+      "loss": 3.4695,
+      "step": 9111
+    },
+    {
+      "epoch": 0.3836631578947368,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003485044310172747,
+      "loss": 3.3797,
+      "step": 9112
+    },
+    {
+      "epoch": 0.3837052631578947,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0003484734156545892,
+      "loss": 2.8459,
+      "step": 9113
+    },
+    {
+      "epoch": 0.3837473684210526,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00034844239849783393,
+      "loss": 3.267,
+      "step": 9114
+    },
+    {
+      "epoch": 0.3837894736842105,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00034841137954757373,
+      "loss": 3.0776,
+      "step": 9115
+    },
+    {
+      "epoch": 0.3838315789473684,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003483803588043738,
+      "loss": 2.7924,
+      "step": 9116
+    },
+    {
+      "epoch": 0.3838736842105263,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003483493362687994,
+      "loss": 3.2586,
+      "step": 9117
+    },
+    {
+      "epoch": 0.3839157894736842,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003483183119414156,
+      "loss": 2.9108,
+      "step": 9118
+    },
+    {
+      "epoch": 0.3839578947368421,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00034828728582278766,
+      "loss": 3.0227,
+      "step": 9119
+    },
+    {
+      "epoch": 0.384,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003482562579134809,
+      "loss": 3.0682,
+      "step": 9120
+    },
+    {
+      "epoch": 0.3840421052631579,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00034822522821406057,
+      "loss": 3.1137,
+      "step": 9121
+    },
+    {
+      "epoch": 0.3840842105263158,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.000348194196725092,
+      "loss": 3.1703,
+      "step": 9122
+    },
+    {
+      "epoch": 0.3841263157894737,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003481631634471405,
+      "loss": 3.2669,
+      "step": 9123
+    },
+    {
+      "epoch": 0.3841684210526316,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00034813212838077153,
+      "loss": 3.1843,
+      "step": 9124
+    },
+    {
+      "epoch": 0.38421052631578945,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003481010915265504,
+      "loss": 2.879,
+      "step": 9125
+    },
+    {
+      "epoch": 0.38425263157894735,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0003480700528850427,
+      "loss": 3.1436,
+      "step": 9126
+    },
+    {
+      "epoch": 0.38429473684210524,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0003480390124568139,
+      "loss": 3.0416,
+      "step": 9127
+    },
+    {
+      "epoch": 0.38433684210526314,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003480079702424295,
+      "loss": 3.1986,
+      "step": 9128
+    },
+    {
+      "epoch": 0.38437894736842104,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00034797692624245497,
+      "loss": 3.4086,
+      "step": 9129
+    },
+    {
+      "epoch": 0.38442105263157894,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.000347945880457456,
+      "loss": 3.3974,
+      "step": 9130
+    },
+    {
+      "epoch": 0.38446315789473684,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00034791483288799816,
+      "loss": 3.5608,
+      "step": 9131
+    },
+    {
+      "epoch": 0.38450526315789474,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.000347883783534647,
+      "loss": 3.2104,
+      "step": 9132
+    },
+    {
+      "epoch": 0.38454736842105264,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00034785273239796845,
+      "loss": 3.3573,
+      "step": 9133
+    },
+    {
+      "epoch": 0.38458947368421054,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003478216794785281,
+      "loss": 3.2512,
+      "step": 9134
+    },
+    {
+      "epoch": 0.38463157894736844,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00034779062477689163,
+      "loss": 2.8528,
+      "step": 9135
+    },
+    {
+      "epoch": 0.38467368421052633,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00034775956829362494,
+      "loss": 2.7154,
+      "step": 9136
+    },
+    {
+      "epoch": 0.38471578947368423,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00034772851002929376,
+      "loss": 3.0953,
+      "step": 9137
+    },
+    {
+      "epoch": 0.38475789473684213,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003476974499844639,
+      "loss": 3.2482,
+      "step": 9138
+    },
+    {
+      "epoch": 0.3848,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00034766638815970127,
+      "loss": 3.3156,
+      "step": 9139
+    },
+    {
+      "epoch": 0.3848421052631579,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00034763532455557176,
+      "loss": 3.4069,
+      "step": 9140
+    },
+    {
+      "epoch": 0.3848842105263158,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00034760425917264147,
+      "loss": 2.835,
+      "step": 9141
+    },
+    {
+      "epoch": 0.38492631578947367,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00034757319201147627,
+      "loss": 2.9664,
+      "step": 9142
+    },
+    {
+      "epoch": 0.38496842105263157,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003475421230726421,
+      "loss": 3.5689,
+      "step": 9143
+    },
+    {
+      "epoch": 0.38501052631578947,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0003475110523567051,
+      "loss": 3.6589,
+      "step": 9144
+    },
+    {
+      "epoch": 0.38505263157894737,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00034747997986423124,
+      "loss": 3.1623,
+      "step": 9145
+    },
+    {
+      "epoch": 0.38509473684210527,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00034744890559578666,
+      "loss": 3.3893,
+      "step": 9146
+    },
+    {
+      "epoch": 0.38513684210526317,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00034741782955193754,
+      "loss": 3.1992,
+      "step": 9147
+    },
+    {
+      "epoch": 0.38517894736842107,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00034738675173325007,
+      "loss": 3.1352,
+      "step": 9148
+    },
+    {
+      "epoch": 0.38522105263157896,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003473556721402904,
+      "loss": 2.9454,
+      "step": 9149
+    },
+    {
+      "epoch": 0.38526315789473686,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003473245907736248,
+      "loss": 2.888,
+      "step": 9150
+    },
+    {
+      "epoch": 0.38530526315789476,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003472935076338194,
+      "loss": 3.086,
+      "step": 9151
+    },
+    {
+      "epoch": 0.3853473684210526,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0003472624227214406,
+      "loss": 3.3419,
+      "step": 9152
+    },
+    {
+      "epoch": 0.3853894736842105,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003472313360370549,
+      "loss": 3.1707,
+      "step": 9153
+    },
+    {
+      "epoch": 0.3854315789473684,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00034720024758122834,
+      "loss": 3.6091,
+      "step": 9154
+    },
+    {
+      "epoch": 0.3854736842105263,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0003471691573545276,
+      "loss": 3.4001,
+      "step": 9155
+    },
+    {
+      "epoch": 0.3855157894736842,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003471380653575189,
+      "loss": 2.9759,
+      "step": 9156
+    },
+    {
+      "epoch": 0.3855578947368421,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003471069715907688,
+      "loss": 3.2656,
+      "step": 9157
+    },
+    {
+      "epoch": 0.3856,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003470758760548438,
+      "loss": 3.1051,
+      "step": 9158
+    },
+    {
+      "epoch": 0.3856421052631579,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0003470447787503103,
+      "loss": 3.3771,
+      "step": 9159
+    },
+    {
+      "epoch": 0.3856842105263158,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00034701367967773507,
+      "loss": 3.4899,
+      "step": 9160
+    },
+    {
+      "epoch": 0.3857263157894737,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00034698257883768456,
+      "loss": 3.2793,
+      "step": 9161
+    },
+    {
+      "epoch": 0.3857684210526316,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00034695147623072545,
+      "loss": 3.2151,
+      "step": 9162
+    },
+    {
+      "epoch": 0.3858105263157895,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003469203718574243,
+      "loss": 2.8305,
+      "step": 9163
+    },
+    {
+      "epoch": 0.3858526315789474,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003468892657183478,
+      "loss": 3.3651,
+      "step": 9164
+    },
+    {
+      "epoch": 0.3858947368421053,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0003468581578140628,
+      "loss": 3.3477,
+      "step": 9165
+    },
+    {
+      "epoch": 0.38593684210526313,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003468270481451359,
+      "loss": 3.3132,
+      "step": 9166
+    },
+    {
+      "epoch": 0.38597894736842103,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.000346795936712134,
+      "loss": 3.0948,
+      "step": 9167
+    },
+    {
+      "epoch": 0.38602105263157893,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00034676482351562386,
+      "loss": 3.6712,
+      "step": 9168
+    },
+    {
+      "epoch": 0.38606315789473683,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0003467337085561723,
+      "loss": 3.3806,
+      "step": 9169
+    },
+    {
+      "epoch": 0.38610526315789473,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0003467025918343463,
+      "loss": 2.8861,
+      "step": 9170
+    },
+    {
+      "epoch": 0.38614736842105263,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00034667147335071263,
+      "loss": 3.2543,
+      "step": 9171
+    },
+    {
+      "epoch": 0.3861894736842105,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0003466403531058383,
+      "loss": 3.1865,
+      "step": 9172
+    },
+    {
+      "epoch": 0.3862315789473684,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003466092311002903,
+      "loss": 2.9101,
+      "step": 9173
+    },
+    {
+      "epoch": 0.3862736842105263,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00034657810733463565,
+      "loss": 3.0891,
+      "step": 9174
+    },
+    {
+      "epoch": 0.3863157894736842,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00034654698180944135,
+      "loss": 2.9474,
+      "step": 9175
+    },
+    {
+      "epoch": 0.3863578947368421,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003465158545252745,
+      "loss": 3.3993,
+      "step": 9176
+    },
+    {
+      "epoch": 0.3864,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003464847254827021,
+      "loss": 2.9952,
+      "step": 9177
+    },
+    {
+      "epoch": 0.3864421052631579,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003464535946822914,
+      "loss": 3.1781,
+      "step": 9178
+    },
+    {
+      "epoch": 0.38648421052631576,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003464224621246095,
+      "loss": 3.3455,
+      "step": 9179
+    },
+    {
+      "epoch": 0.38652631578947366,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003463913278102237,
+      "loss": 3.1198,
+      "step": 9180
+    },
+    {
+      "epoch": 0.38656842105263156,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00034636019173970114,
+      "loss": 3.5939,
+      "step": 9181
+    },
+    {
+      "epoch": 0.38661052631578946,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003463290539136091,
+      "loss": 2.9391,
+      "step": 9182
+    },
+    {
+      "epoch": 0.38665263157894736,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0003462979143325149,
+      "loss": 3.5022,
+      "step": 9183
+    },
+    {
+      "epoch": 0.38669473684210526,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00034626677299698583,
+      "loss": 3.0262,
+      "step": 9184
+    },
+    {
+      "epoch": 0.38673684210526316,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00034623562990758927,
+      "loss": 3.2477,
+      "step": 9185
+    },
+    {
+      "epoch": 0.38677894736842106,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00034620448506489257,
+      "loss": 3.3257,
+      "step": 9186
+    },
+    {
+      "epoch": 0.38682105263157895,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0003461733384694632,
+      "loss": 3.1583,
+      "step": 9187
+    },
+    {
+      "epoch": 0.38686315789473685,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00034614219012186863,
+      "loss": 3.101,
+      "step": 9188
+    },
+    {
+      "epoch": 0.38690526315789475,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00034611104002267633,
+      "loss": 3.1601,
+      "step": 9189
+    },
+    {
+      "epoch": 0.38694736842105265,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003460798881724537,
+      "loss": 3.0,
+      "step": 9190
+    },
+    {
+      "epoch": 0.38698947368421055,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00034604873457176854,
+      "loss": 3.0523,
+      "step": 9191
+    },
+    {
+      "epoch": 0.38703157894736845,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00034601757922118827,
+      "loss": 3.3357,
+      "step": 9192
+    },
+    {
+      "epoch": 0.3870736842105263,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00034598642212128037,
+      "loss": 3.2214,
+      "step": 9193
+    },
+    {
+      "epoch": 0.3871157894736842,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00034595526327261286,
+      "loss": 3.1781,
+      "step": 9194
+    },
+    {
+      "epoch": 0.3871578947368421,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00034592410267575305,
+      "loss": 2.7701,
+      "step": 9195
+    },
+    {
+      "epoch": 0.3872,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00034589294033126884,
+      "loss": 3.1996,
+      "step": 9196
+    },
+    {
+      "epoch": 0.3872421052631579,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003458617762397279,
+      "loss": 3.3618,
+      "step": 9197
+    },
+    {
+      "epoch": 0.3872842105263158,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00034583061040169803,
+      "loss": 3.1964,
+      "step": 9198
+    },
+    {
+      "epoch": 0.3873263157894737,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00034579944281774703,
+      "loss": 3.1649,
+      "step": 9199
+    },
+    {
+      "epoch": 0.3873684210526316,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0003457682734884428,
+      "loss": 3.0159,
+      "step": 9200
+    },
+    {
+      "epoch": 0.3874105263157895,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00034573710241435313,
+      "loss": 3.1175,
+      "step": 9201
+    },
+    {
+      "epoch": 0.3874526315789474,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003457059295960459,
+      "loss": 3.1072,
+      "step": 9202
+    },
+    {
+      "epoch": 0.3874947368421053,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00034567475503408915,
+      "loss": 2.8826,
+      "step": 9203
+    },
+    {
+      "epoch": 0.3875368421052632,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003456435787290507,
+      "loss": 3.1208,
+      "step": 9204
+    },
+    {
+      "epoch": 0.3875789473684211,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0003456124006814987,
+      "loss": 2.8116,
+      "step": 9205
+    },
+    {
+      "epoch": 0.3876210526315789,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00034558122089200106,
+      "loss": 2.973,
+      "step": 9206
+    },
+    {
+      "epoch": 0.3876631578947368,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0003455500393611258,
+      "loss": 3.7385,
+      "step": 9207
+    },
+    {
+      "epoch": 0.3877052631578947,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00034551885608944115,
+      "loss": 2.984,
+      "step": 9208
+    },
+    {
+      "epoch": 0.3877473684210526,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00034548767107751523,
+      "loss": 2.9402,
+      "step": 9209
+    },
+    {
+      "epoch": 0.3877894736842105,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00034545648432591613,
+      "loss": 3.1153,
+      "step": 9210
+    },
+    {
+      "epoch": 0.3878315789473684,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00034542529583521197,
+      "loss": 3.331,
+      "step": 9211
+    },
+    {
+      "epoch": 0.3878736842105263,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00034539410560597113,
+      "loss": 2.6639,
+      "step": 9212
+    },
+    {
+      "epoch": 0.3879157894736842,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00034536291363876164,
+      "loss": 2.9007,
+      "step": 9213
+    },
+    {
+      "epoch": 0.3879578947368421,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.000345331719934152,
+      "loss": 3.1507,
+      "step": 9214
+    },
+    {
+      "epoch": 0.388,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00034530052449271043,
+      "loss": 2.8033,
+      "step": 9215
+    },
+    {
+      "epoch": 0.3880421052631579,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003452693273150053,
+      "loss": 3.2249,
+      "step": 9216
+    },
+    {
+      "epoch": 0.3880842105263158,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003452381284016049,
+      "loss": 2.801,
+      "step": 9217
+    },
+    {
+      "epoch": 0.3881263157894737,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00034520692775307773,
+      "loss": 3.317,
+      "step": 9218
+    },
+    {
+      "epoch": 0.38816842105263155,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003451757253699922,
+      "loss": 3.1291,
+      "step": 9219
+    },
+    {
+      "epoch": 0.38821052631578945,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0003451445212529167,
+      "loss": 3.0008,
+      "step": 9220
+    },
+    {
+      "epoch": 0.38825263157894735,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00034511331540242,
+      "loss": 3.1768,
+      "step": 9221
+    },
+    {
+      "epoch": 0.38829473684210525,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003450821078190703,
+      "loss": 3.0742,
+      "step": 9222
+    },
+    {
+      "epoch": 0.38833684210526315,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00034505089850343627,
+      "loss": 3.1913,
+      "step": 9223
+    },
+    {
+      "epoch": 0.38837894736842105,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003450196874560867,
+      "loss": 3.0506,
+      "step": 9224
+    },
+    {
+      "epoch": 0.38842105263157894,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00034498847467759,
+      "loss": 3.1226,
+      "step": 9225
+    },
+    {
+      "epoch": 0.38846315789473684,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003449572601685149,
+      "loss": 3.5846,
+      "step": 9226
+    },
+    {
+      "epoch": 0.38850526315789474,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00034492604392943004,
+      "loss": 3.1976,
+      "step": 9227
+    },
+    {
+      "epoch": 0.38854736842105264,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00034489482596090427,
+      "loss": 2.8722,
+      "step": 9228
+    },
+    {
+      "epoch": 0.38858947368421054,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00034486360626350624,
+      "loss": 3.0238,
+      "step": 9229
+    },
+    {
+      "epoch": 0.38863157894736844,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003448323848378048,
+      "loss": 3.2199,
+      "step": 9230
+    },
+    {
+      "epoch": 0.38867368421052634,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00034480116168436863,
+      "loss": 3.2655,
+      "step": 9231
+    },
+    {
+      "epoch": 0.38871578947368424,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0003447699368037668,
+      "loss": 3.3257,
+      "step": 9232
+    },
+    {
+      "epoch": 0.3887578947368421,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.000344738710196568,
+      "loss": 2.9313,
+      "step": 9233
+    },
+    {
+      "epoch": 0.3888,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003447074818633412,
+      "loss": 3.5461,
+      "step": 9234
+    },
+    {
+      "epoch": 0.3888421052631579,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003446762518046554,
+      "loss": 3.5364,
+      "step": 9235
+    },
+    {
+      "epoch": 0.3888842105263158,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00034464502002107955,
+      "loss": 3.3248,
+      "step": 9236
+    },
+    {
+      "epoch": 0.3889263157894737,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003446137865131826,
+      "loss": 3.2285,
+      "step": 9237
+    },
+    {
+      "epoch": 0.3889684210526316,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0003445825512815337,
+      "loss": 3.5039,
+      "step": 9238
+    },
+    {
+      "epoch": 0.3890105263157895,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003445513143267017,
+      "loss": 3.5397,
+      "step": 9239
+    },
+    {
+      "epoch": 0.38905263157894737,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0003445200756492559,
+      "loss": 2.9977,
+      "step": 9240
+    },
+    {
+      "epoch": 0.38909473684210527,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003444888352497654,
+      "loss": 2.7339,
+      "step": 9241
+    },
+    {
+      "epoch": 0.38913684210526317,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0003444575931287994,
+      "loss": 3.4114,
+      "step": 9242
+    },
+    {
+      "epoch": 0.38917894736842107,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00034442634928692696,
+      "loss": 3.2276,
+      "step": 9243
+    },
+    {
+      "epoch": 0.38922105263157897,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003443951037247174,
+      "loss": 3.2348,
+      "step": 9244
+    },
+    {
+      "epoch": 0.38926315789473687,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00034436385644274,
+      "loss": 3.0852,
+      "step": 9245
+    },
+    {
+      "epoch": 0.3893052631578947,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.000344332607441564,
+      "loss": 3.6413,
+      "step": 9246
+    },
+    {
+      "epoch": 0.3893473684210526,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00034430135672175867,
+      "loss": 3.2108,
+      "step": 9247
+    },
+    {
+      "epoch": 0.3893894736842105,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00034427010428389346,
+      "loss": 3.6164,
+      "step": 9248
+    },
+    {
+      "epoch": 0.3894315789473684,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00034423885012853763,
+      "loss": 2.682,
+      "step": 9249
+    },
+    {
+      "epoch": 0.3894736842105263,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003442075942562608,
+      "loss": 3.3274,
+      "step": 9250
+    },
+    {
+      "epoch": 0.3895157894736842,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00034417633666763227,
+      "loss": 3.354,
+      "step": 9251
+    },
+    {
+      "epoch": 0.3895578947368421,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003441450773632215,
+      "loss": 3.0121,
+      "step": 9252
+    },
+    {
+      "epoch": 0.3896,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00034411381634359803,
+      "loss": 3.0092,
+      "step": 9253
+    },
+    {
+      "epoch": 0.3896421052631579,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003440825536093314,
+      "loss": 3.2775,
+      "step": 9254
+    },
+    {
+      "epoch": 0.3896842105263158,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0003440512891609912,
+      "loss": 3.6086,
+      "step": 9255
+    },
+    {
+      "epoch": 0.3897263157894737,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.000344020022999147,
+      "loss": 3.575,
+      "step": 9256
+    },
+    {
+      "epoch": 0.3897684210526316,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0003439887551243685,
+      "loss": 3.3348,
+      "step": 9257
+    },
+    {
+      "epoch": 0.3898105263157895,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00034395748553722527,
+      "loss": 3.5396,
+      "step": 9258
+    },
+    {
+      "epoch": 0.3898526315789474,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00034392621423828714,
+      "loss": 3.886,
+      "step": 9259
+    },
+    {
+      "epoch": 0.38989473684210524,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00034389494122812357,
+      "loss": 3.2329,
+      "step": 9260
+    },
+    {
+      "epoch": 0.38993684210526314,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00034386366650730467,
+      "loss": 2.9259,
+      "step": 9261
+    },
+    {
+      "epoch": 0.38997894736842104,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00034383239007639997,
+      "loss": 3.1411,
+      "step": 9262
+    },
+    {
+      "epoch": 0.39002105263157893,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0003438011119359794,
+      "loss": 2.7093,
+      "step": 9263
+    },
+    {
+      "epoch": 0.39006315789473683,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0003437698320866128,
+      "loss": 3.5848,
+      "step": 9264
+    },
+    {
+      "epoch": 0.39010526315789473,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00034373855052886995,
+      "loss": 3.3434,
+      "step": 9265
+    },
+    {
+      "epoch": 0.39014736842105263,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0003437072672633209,
+      "loss": 2.8476,
+      "step": 9266
+    },
+    {
+      "epoch": 0.39018947368421053,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00034367598229053547,
+      "loss": 3.2013,
+      "step": 9267
+    },
+    {
+      "epoch": 0.39023157894736843,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003436446956110838,
+      "loss": 3.2493,
+      "step": 9268
+    },
+    {
+      "epoch": 0.39027368421052633,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003436134072255358,
+      "loss": 3.2342,
+      "step": 9269
+    },
+    {
+      "epoch": 0.3903157894736842,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003435821171344615,
+      "loss": 3.1419,
+      "step": 9270
+    },
+    {
+      "epoch": 0.3903578947368421,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.000343550825338431,
+      "loss": 3.486,
+      "step": 9271
+    },
+    {
+      "epoch": 0.3904,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003435195318380143,
+      "loss": 2.7725,
+      "step": 9272
+    },
+    {
+      "epoch": 0.39044210526315787,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00034348823663378166,
+      "loss": 3.297,
+      "step": 9273
+    },
+    {
+      "epoch": 0.39048421052631577,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003434569397263032,
+      "loss": 3.2614,
+      "step": 9274
+    },
+    {
+      "epoch": 0.39052631578947367,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00034342564111614906,
+      "loss": 3.2127,
+      "step": 9275
+    },
+    {
+      "epoch": 0.39056842105263156,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003433943408038895,
+      "loss": 3.1767,
+      "step": 9276
+    },
+    {
+      "epoch": 0.39061052631578946,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003433630387900949,
+      "loss": 2.7615,
+      "step": 9277
+    },
+    {
+      "epoch": 0.39065263157894736,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003433317350753354,
+      "loss": 3.0941,
+      "step": 9278
+    },
+    {
+      "epoch": 0.39069473684210526,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00034330042966018134,
+      "loss": 3.3639,
+      "step": 9279
+    },
+    {
+      "epoch": 0.39073684210526316,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.000343269122545203,
+      "loss": 3.1707,
+      "step": 9280
+    },
+    {
+      "epoch": 0.39077894736842106,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003432378137309708,
+      "loss": 3.16,
+      "step": 9281
+    },
+    {
+      "epoch": 0.39082105263157896,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00034320650321805537,
+      "loss": 3.2619,
+      "step": 9282
+    },
+    {
+      "epoch": 0.39086315789473686,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00034317519100702674,
+      "loss": 3.4493,
+      "step": 9283
+    },
+    {
+      "epoch": 0.39090526315789476,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0003431438770984558,
+      "loss": 3.0853,
+      "step": 9284
+    },
+    {
+      "epoch": 0.39094736842105265,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00034311256149291284,
+      "loss": 2.9384,
+      "step": 9285
+    },
+    {
+      "epoch": 0.39098947368421055,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003430812441909683,
+      "loss": 3.2078,
+      "step": 9286
+    },
+    {
+      "epoch": 0.3910315789473684,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0003430499251931929,
+      "loss": 3.2524,
+      "step": 9287
+    },
+    {
+      "epoch": 0.3910736842105263,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003430186045001572,
+      "loss": 3.5292,
+      "step": 9288
+    },
+    {
+      "epoch": 0.3911157894736842,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00034298728211243183,
+      "loss": 2.9912,
+      "step": 9289
+    },
+    {
+      "epoch": 0.3911578947368421,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003429559580305874,
+      "loss": 3.3431,
+      "step": 9290
+    },
+    {
+      "epoch": 0.3912,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00034292463225519464,
+      "loss": 3.2726,
+      "step": 9291
+    },
+    {
+      "epoch": 0.3912421052631579,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00034289330478682434,
+      "loss": 3.3585,
+      "step": 9292
+    },
+    {
+      "epoch": 0.3912842105263158,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00034286197562604715,
+      "loss": 3.4024,
+      "step": 9293
+    },
+    {
+      "epoch": 0.3913263157894737,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003428306447734338,
+      "loss": 3.4011,
+      "step": 9294
+    },
+    {
+      "epoch": 0.3913684210526316,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00034279931222955517,
+      "loss": 3.1112,
+      "step": 9295
+    },
+    {
+      "epoch": 0.3914105263157895,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00034276797799498215,
+      "loss": 2.9329,
+      "step": 9296
+    },
+    {
+      "epoch": 0.3914526315789474,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003427366420702856,
+      "loss": 3.0345,
+      "step": 9297
+    },
+    {
+      "epoch": 0.3914947368421053,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003427053044560364,
+      "loss": 3.1107,
+      "step": 9298
+    },
+    {
+      "epoch": 0.3915368421052632,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0003426739651528054,
+      "loss": 3.0583,
+      "step": 9299
+    },
+    {
+      "epoch": 0.391578947368421,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00034264262416116366,
+      "loss": 2.8402,
+      "step": 9300
+    },
+    {
+      "epoch": 0.3916210526315789,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00034261128148168214,
+      "loss": 3.214,
+      "step": 9301
+    },
+    {
+      "epoch": 0.3916631578947368,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00034257993711493196,
+      "loss": 3.7449,
+      "step": 9302
+    },
+    {
+      "epoch": 0.3917052631578947,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00034254859106148407,
+      "loss": 3.5061,
+      "step": 9303
+    },
+    {
+      "epoch": 0.3917473684210526,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003425172433219096,
+      "loss": 3.0373,
+      "step": 9304
+    },
+    {
+      "epoch": 0.3917894736842105,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00034248589389677963,
+      "loss": 3.5142,
+      "step": 9305
+    },
+    {
+      "epoch": 0.3918315789473684,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0003424545427866654,
+      "loss": 3.4156,
+      "step": 9306
+    },
+    {
+      "epoch": 0.3918736842105263,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.000342423189992138,
+      "loss": 2.8161,
+      "step": 9307
+    },
+    {
+      "epoch": 0.3919157894736842,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00034239183551376857,
+      "loss": 2.9438,
+      "step": 9308
+    },
+    {
+      "epoch": 0.3919578947368421,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00034236047935212857,
+      "loss": 2.8596,
+      "step": 9309
+    },
+    {
+      "epoch": 0.392,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0003423291215077891,
+      "loss": 3.431,
+      "step": 9310
+    },
+    {
+      "epoch": 0.3920421052631579,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00034229776198132155,
+      "loss": 3.4723,
+      "step": 9311
+    },
+    {
+      "epoch": 0.3920842105263158,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00034226640077329723,
+      "loss": 3.3518,
+      "step": 9312
+    },
+    {
+      "epoch": 0.3921263157894737,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003422350378842875,
+      "loss": 3.1741,
+      "step": 9313
+    },
+    {
+      "epoch": 0.39216842105263156,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003422036733148637,
+      "loss": 3.3177,
+      "step": 9314
+    },
+    {
+      "epoch": 0.39221052631578945,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003421723070655973,
+      "loss": 3.8976,
+      "step": 9315
+    },
+    {
+      "epoch": 0.39225263157894735,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0003421409391370598,
+      "loss": 2.6924,
+      "step": 9316
+    },
+    {
+      "epoch": 0.39229473684210525,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00034210956952982266,
+      "loss": 3.2678,
+      "step": 9317
+    },
+    {
+      "epoch": 0.39233684210526315,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0003420781982444574,
+      "loss": 2.7106,
+      "step": 9318
+    },
+    {
+      "epoch": 0.39237894736842105,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00034204682528153553,
+      "loss": 3.5089,
+      "step": 9319
+    },
+    {
+      "epoch": 0.39242105263157895,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003420154506416287,
+      "loss": 3.1609,
+      "step": 9320
+    },
+    {
+      "epoch": 0.39246315789473685,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0003419840743253084,
+      "loss": 2.9988,
+      "step": 9321
+    },
+    {
+      "epoch": 0.39250526315789475,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003419526963331464,
+      "loss": 3.2808,
+      "step": 9322
+    },
+    {
+      "epoch": 0.39254736842105264,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003419213166657143,
+      "loss": 3.0894,
+      "step": 9323
+    },
+    {
+      "epoch": 0.39258947368421054,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003418899353235838,
+      "loss": 3.499,
+      "step": 9324
+    },
+    {
+      "epoch": 0.39263157894736844,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003418585523073266,
+      "loss": 3.5154,
+      "step": 9325
+    },
+    {
+      "epoch": 0.39267368421052634,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0003418271676175146,
+      "loss": 3.0868,
+      "step": 9326
+    },
+    {
+      "epoch": 0.3927157894736842,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003417957812547194,
+      "loss": 3.2105,
+      "step": 9327
+    },
+    {
+      "epoch": 0.3927578947368421,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.000341764393219513,
+      "loss": 3.2384,
+      "step": 9328
+    },
+    {
+      "epoch": 0.3928,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0003417330035124671,
+      "loss": 2.7008,
+      "step": 9329
+    },
+    {
+      "epoch": 0.3928421052631579,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0003417016121341537,
+      "loss": 3.0429,
+      "step": 9330
+    },
+    {
+      "epoch": 0.3928842105263158,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00034167021908514475,
+      "loss": 3.0118,
+      "step": 9331
+    },
+    {
+      "epoch": 0.3929263157894737,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.000341638824366012,
+      "loss": 3.5525,
+      "step": 9332
+    },
+    {
+      "epoch": 0.3929684210526316,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003416074279773276,
+      "loss": 2.8257,
+      "step": 9333
+    },
+    {
+      "epoch": 0.3930105263157895,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00034157602991966344,
+      "loss": 3.1373,
+      "step": 9334
+    },
+    {
+      "epoch": 0.3930526315789474,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00034154463019359164,
+      "loss": 3.2645,
+      "step": 9335
+    },
+    {
+      "epoch": 0.3930947368421053,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00034151322879968427,
+      "loss": 2.8609,
+      "step": 9336
+    },
+    {
+      "epoch": 0.3931368421052632,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0003414818257385134,
+      "loss": 3.4161,
+      "step": 9337
+    },
+    {
+      "epoch": 0.3931789473684211,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003414504210106512,
+      "loss": 3.0461,
+      "step": 9338
+    },
+    {
+      "epoch": 0.39322105263157897,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00034141901461666967,
+      "loss": 3.3991,
+      "step": 9339
+    },
+    {
+      "epoch": 0.3932631578947368,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00034138760655714106,
+      "loss": 3.3579,
+      "step": 9340
+    },
+    {
+      "epoch": 0.3933052631578947,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00034135619683263777,
+      "loss": 2.8748,
+      "step": 9341
+    },
+    {
+      "epoch": 0.3933473684210526,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003413247854437318,
+      "loss": 2.9787,
+      "step": 9342
+    },
+    {
+      "epoch": 0.3933894736842105,
+      "grad_norm": 3.609375,
+      "learning_rate": 0.0003412933723909956,
+      "loss": 3.2526,
+      "step": 9343
+    },
+    {
+      "epoch": 0.3934315789473684,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00034126195767500146,
+      "loss": 3.4516,
+      "step": 9344
+    },
+    {
+      "epoch": 0.3934736842105263,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00034123054129632166,
+      "loss": 3.3662,
+      "step": 9345
+    },
+    {
+      "epoch": 0.3935157894736842,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003411991232555286,
+      "loss": 3.488,
+      "step": 9346
+    },
+    {
+      "epoch": 0.3935578947368421,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003411677035531946,
+      "loss": 3.3121,
+      "step": 9347
+    },
+    {
+      "epoch": 0.3936,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0003411362821898922,
+      "loss": 3.1821,
+      "step": 9348
+    },
+    {
+      "epoch": 0.3936421052631579,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003411048591661938,
+      "loss": 2.9684,
+      "step": 9349
+    },
+    {
+      "epoch": 0.3936842105263158,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003410734344826719,
+      "loss": 2.8663,
+      "step": 9350
+    },
+    {
+      "epoch": 0.3937263157894737,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.000341042008139899,
+      "loss": 3.4578,
+      "step": 9351
+    },
+    {
+      "epoch": 0.3937684210526316,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00034101058013844785,
+      "loss": 3.1262,
+      "step": 9352
+    },
+    {
+      "epoch": 0.3938105263157895,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00034097915047889073,
+      "loss": 2.8811,
+      "step": 9353
+    },
+    {
+      "epoch": 0.39385263157894734,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003409477191618004,
+      "loss": 3.334,
+      "step": 9354
+    },
+    {
+      "epoch": 0.39389473684210524,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0003409162861877495,
+      "loss": 3.3667,
+      "step": 9355
+    },
+    {
+      "epoch": 0.39393684210526314,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0003408848515573107,
+      "loss": 3.3957,
+      "step": 9356
+    },
+    {
+      "epoch": 0.39397894736842104,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0003408534152710567,
+      "loss": 3.4571,
+      "step": 9357
+    },
+    {
+      "epoch": 0.39402105263157894,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003408219773295601,
+      "loss": 3.084,
+      "step": 9358
+    },
+    {
+      "epoch": 0.39406315789473684,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003407905377333939,
+      "loss": 3.1206,
+      "step": 9359
+    },
+    {
+      "epoch": 0.39410526315789474,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003407590964831308,
+      "loss": 3.3634,
+      "step": 9360
+    },
+    {
+      "epoch": 0.39414736842105264,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003407276535793435,
+      "loss": 3.0231,
+      "step": 9361
+    },
+    {
+      "epoch": 0.39418947368421053,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00034069620902260503,
+      "loss": 3.1187,
+      "step": 9362
+    },
+    {
+      "epoch": 0.39423157894736843,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0003406647628134882,
+      "loss": 3.3034,
+      "step": 9363
+    },
+    {
+      "epoch": 0.39427368421052633,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0003406333149525659,
+      "loss": 3.0686,
+      "step": 9364
+    },
+    {
+      "epoch": 0.39431578947368423,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00034060186544041107,
+      "loss": 3.0862,
+      "step": 9365
+    },
+    {
+      "epoch": 0.39435789473684213,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003405704142775967,
+      "loss": 2.8707,
+      "step": 9366
+    },
+    {
+      "epoch": 0.3944,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003405389614646958,
+      "loss": 3.5068,
+      "step": 9367
+    },
+    {
+      "epoch": 0.39444210526315787,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0003405075070022814,
+      "loss": 2.9891,
+      "step": 9368
+    },
+    {
+      "epoch": 0.39448421052631577,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003404760508909266,
+      "loss": 3.341,
+      "step": 9369
+    },
+    {
+      "epoch": 0.39452631578947367,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00034044459313120444,
+      "loss": 3.3758,
+      "step": 9370
+    },
+    {
+      "epoch": 0.39456842105263157,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.000340413133723688,
+      "loss": 3.2097,
+      "step": 9371
+    },
+    {
+      "epoch": 0.39461052631578947,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003403816726689506,
+      "loss": 3.3644,
+      "step": 9372
+    },
+    {
+      "epoch": 0.39465263157894737,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003403502099675653,
+      "loss": 3.1723,
+      "step": 9373
+    },
+    {
+      "epoch": 0.39469473684210526,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003403187456201053,
+      "loss": 3.0189,
+      "step": 9374
+    },
+    {
+      "epoch": 0.39473684210526316,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0003402872796271438,
+      "loss": 3.0011,
+      "step": 9375
+    },
+    {
+      "epoch": 0.39477894736842106,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00034025581198925424,
+      "loss": 3.3418,
+      "step": 9376
+    },
+    {
+      "epoch": 0.39482105263157896,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003402243427070099,
+      "loss": 3.0275,
+      "step": 9377
+    },
+    {
+      "epoch": 0.39486315789473686,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003401928717809839,
+      "loss": 3.2846,
+      "step": 9378
+    },
+    {
+      "epoch": 0.39490526315789476,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003401613992117498,
+      "loss": 3.0887,
+      "step": 9379
+    },
+    {
+      "epoch": 0.39494736842105266,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.000340129924999881,
+      "loss": 3.2651,
+      "step": 9380
+    },
+    {
+      "epoch": 0.3949894736842105,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003400984491459508,
+      "loss": 3.2532,
+      "step": 9381
+    },
+    {
+      "epoch": 0.3950315789473684,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00034006697165053274,
+      "loss": 3.5388,
+      "step": 9382
+    },
+    {
+      "epoch": 0.3950736842105263,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00034003549251420025,
+      "loss": 3.6687,
+      "step": 9383
+    },
+    {
+      "epoch": 0.3951157894736842,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003400040117375269,
+      "loss": 2.8632,
+      "step": 9384
+    },
+    {
+      "epoch": 0.3951578947368421,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003399725293210862,
+      "loss": 3.1374,
+      "step": 9385
+    },
+    {
+      "epoch": 0.3952,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0003399410452654518,
+      "loss": 3.1815,
+      "step": 9386
+    },
+    {
+      "epoch": 0.3952421052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003399095595711971,
+      "loss": 3.5786,
+      "step": 9387
+    },
+    {
+      "epoch": 0.3952842105263158,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00033987807223889596,
+      "loss": 3.2187,
+      "step": 9388
+    },
+    {
+      "epoch": 0.3953263157894737,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0003398465832691219,
+      "loss": 3.2285,
+      "step": 9389
+    },
+    {
+      "epoch": 0.3953684210526316,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003398150926624487,
+      "loss": 3.3207,
+      "step": 9390
+    },
+    {
+      "epoch": 0.3954105263157895,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00033978360041945,
+      "loss": 3.351,
+      "step": 9391
+    },
+    {
+      "epoch": 0.3954526315789474,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0003397521065406996,
+      "loss": 3.1505,
+      "step": 9392
+    },
+    {
+      "epoch": 0.3954947368421053,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0003397206110267713,
+      "loss": 3.4979,
+      "step": 9393
+    },
+    {
+      "epoch": 0.39553684210526313,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003396891138782389,
+      "loss": 2.8629,
+      "step": 9394
+    },
+    {
+      "epoch": 0.39557894736842103,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003396576150956761,
+      "loss": 3.326,
+      "step": 9395
+    },
+    {
+      "epoch": 0.39562105263157893,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.000339626114679657,
+      "loss": 3.1833,
+      "step": 9396
+    },
+    {
+      "epoch": 0.3956631578947368,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003395946126307554,
+      "loss": 3.1944,
+      "step": 9397
+    },
+    {
+      "epoch": 0.3957052631578947,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0003395631089495452,
+      "loss": 3.4424,
+      "step": 9398
+    },
+    {
+      "epoch": 0.3957473684210526,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00033953160363660046,
+      "loss": 3.2839,
+      "step": 9399
+    },
+    {
+      "epoch": 0.3957894736842105,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.000339500096692495,
+      "loss": 2.7449,
+      "step": 9400
+    },
+    {
+      "epoch": 0.3958315789473684,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0003394685881178029,
+      "loss": 2.9628,
+      "step": 9401
+    },
+    {
+      "epoch": 0.3958736842105263,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00033943707791309833,
+      "loss": 3.2985,
+      "step": 9402
+    },
+    {
+      "epoch": 0.3959157894736842,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0003394055660789553,
+      "loss": 3.1842,
+      "step": 9403
+    },
+    {
+      "epoch": 0.3959578947368421,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0003393740526159479,
+      "loss": 2.9547,
+      "step": 9404
+    },
+    {
+      "epoch": 0.396,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003393425375246503,
+      "loss": 3.5107,
+      "step": 9405
+    },
+    {
+      "epoch": 0.3960421052631579,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00033931102080563653,
+      "loss": 2.876,
+      "step": 9406
+    },
+    {
+      "epoch": 0.3960842105263158,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.000339279502459481,
+      "loss": 3.3906,
+      "step": 9407
+    },
+    {
+      "epoch": 0.39612631578947366,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0003392479824867577,
+      "loss": 3.4675,
+      "step": 9408
+    },
+    {
+      "epoch": 0.39616842105263156,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0003392164608880411,
+      "loss": 2.8985,
+      "step": 9409
+    },
+    {
+      "epoch": 0.39621052631578946,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0003391849376639055,
+      "loss": 2.9495,
+      "step": 9410
+    },
+    {
+      "epoch": 0.39625263157894736,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00033915341281492495,
+      "loss": 3.4323,
+      "step": 9411
+    },
+    {
+      "epoch": 0.39629473684210526,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003391218863416741,
+      "loss": 3.0022,
+      "step": 9412
+    },
+    {
+      "epoch": 0.39633684210526315,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003390903582447272,
+      "loss": 3.087,
+      "step": 9413
+    },
+    {
+      "epoch": 0.39637894736842105,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003390588285246586,
+      "loss": 3.0441,
+      "step": 9414
+    },
+    {
+      "epoch": 0.39642105263157895,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00033902729718204275,
+      "loss": 3.4206,
+      "step": 9415
+    },
+    {
+      "epoch": 0.39646315789473685,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0003389957642174542,
+      "loss": 3.0876,
+      "step": 9416
+    },
+    {
+      "epoch": 0.39650526315789475,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00033896422963146735,
+      "loss": 3.0009,
+      "step": 9417
+    },
+    {
+      "epoch": 0.39654736842105265,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0003389326934246568,
+      "loss": 3.3044,
+      "step": 9418
+    },
+    {
+      "epoch": 0.39658947368421055,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00033890115559759705,
+      "loss": 3.1981,
+      "step": 9419
+    },
+    {
+      "epoch": 0.39663157894736845,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00033886961615086275,
+      "loss": 3.0344,
+      "step": 9420
+    },
+    {
+      "epoch": 0.3966736842105263,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00033883807508502836,
+      "loss": 3.456,
+      "step": 9421
+    },
+    {
+      "epoch": 0.3967157894736842,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003388065324006687,
+      "loss": 3.3726,
+      "step": 9422
+    },
+    {
+      "epoch": 0.3967578947368421,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00033877498809835836,
+      "loss": 3.187,
+      "step": 9423
+    },
+    {
+      "epoch": 0.3968,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00033874344217867206,
+      "loss": 3.365,
+      "step": 9424
+    },
+    {
+      "epoch": 0.3968421052631579,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00033871189464218445,
+      "loss": 3.4772,
+      "step": 9425
+    },
+    {
+      "epoch": 0.3968842105263158,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00033868034548947037,
+      "loss": 3.1937,
+      "step": 9426
+    },
+    {
+      "epoch": 0.3969263157894737,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0003386487947211046,
+      "loss": 3.2586,
+      "step": 9427
+    },
+    {
+      "epoch": 0.3969684210526316,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00033861724233766197,
+      "loss": 3.0654,
+      "step": 9428
+    },
+    {
+      "epoch": 0.3970105263157895,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0003385856883397173,
+      "loss": 3.6218,
+      "step": 9429
+    },
+    {
+      "epoch": 0.3970526315789474,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003385541327278454,
+      "loss": 3.4129,
+      "step": 9430
+    },
+    {
+      "epoch": 0.3970947368421053,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0003385225755026213,
+      "loss": 2.9773,
+      "step": 9431
+    },
+    {
+      "epoch": 0.3971368421052632,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00033849101666461994,
+      "loss": 3.5603,
+      "step": 9432
+    },
+    {
+      "epoch": 0.3971789473684211,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00033845945621441617,
+      "loss": 3.0165,
+      "step": 9433
+    },
+    {
+      "epoch": 0.397221052631579,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00033842789415258507,
+      "loss": 3.4725,
+      "step": 9434
+    },
+    {
+      "epoch": 0.3972631578947368,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003383963304797016,
+      "loss": 3.0359,
+      "step": 9435
+    },
+    {
+      "epoch": 0.3973052631578947,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0003383647651963409,
+      "loss": 3.3101,
+      "step": 9436
+    },
+    {
+      "epoch": 0.3973473684210526,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00033833319830307797,
+      "loss": 3.3951,
+      "step": 9437
+    },
+    {
+      "epoch": 0.3973894736842105,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00033830162980048796,
+      "loss": 3.1684,
+      "step": 9438
+    },
+    {
+      "epoch": 0.3974315789473684,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00033827005968914604,
+      "loss": 3.0398,
+      "step": 9439
+    },
+    {
+      "epoch": 0.3974736842105263,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003382384879696273,
+      "loss": 3.1277,
+      "step": 9440
+    },
+    {
+      "epoch": 0.3975157894736842,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.000338206914642507,
+      "loss": 2.8026,
+      "step": 9441
+    },
+    {
+      "epoch": 0.3975578947368421,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00033817533970836037,
+      "loss": 2.9444,
+      "step": 9442
+    },
+    {
+      "epoch": 0.3976,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0003381437631677625,
+      "loss": 3.4598,
+      "step": 9443
+    },
+    {
+      "epoch": 0.3976421052631579,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00033811218502128904,
+      "loss": 3.1898,
+      "step": 9444
+    },
+    {
+      "epoch": 0.3976842105263158,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.000338080605269515,
+      "loss": 3.2844,
+      "step": 9445
+    },
+    {
+      "epoch": 0.3977263157894737,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0003380490239130158,
+      "loss": 3.1321,
+      "step": 9446
+    },
+    {
+      "epoch": 0.3977684210526316,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.000338017440952367,
+      "loss": 3.1293,
+      "step": 9447
+    },
+    {
+      "epoch": 0.39781052631578945,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003379858563881436,
+      "loss": 3.3363,
+      "step": 9448
+    },
+    {
+      "epoch": 0.39785263157894735,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00033795427022092144,
+      "loss": 3.4328,
+      "step": 9449
+    },
+    {
+      "epoch": 0.39789473684210525,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00033792268245127577,
+      "loss": 3.5661,
+      "step": 9450
+    },
+    {
+      "epoch": 0.39793684210526314,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003378910930797821,
+      "loss": 3.0374,
+      "step": 9451
+    },
+    {
+      "epoch": 0.39797894736842104,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00033785950210701605,
+      "loss": 3.1964,
+      "step": 9452
+    },
+    {
+      "epoch": 0.39802105263157894,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003378279095335531,
+      "loss": 3.1254,
+      "step": 9453
+    },
+    {
+      "epoch": 0.39806315789473684,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003377963153599688,
+      "loss": 3.0871,
+      "step": 9454
+    },
+    {
+      "epoch": 0.39810526315789474,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00033776471958683874,
+      "loss": 3.2675,
+      "step": 9455
+    },
+    {
+      "epoch": 0.39814736842105264,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00033773312221473865,
+      "loss": 3.302,
+      "step": 9456
+    },
+    {
+      "epoch": 0.39818947368421054,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0003377015232442442,
+      "loss": 3.0438,
+      "step": 9457
+    },
+    {
+      "epoch": 0.39823157894736844,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.000337669922675931,
+      "loss": 3.2306,
+      "step": 9458
+    },
+    {
+      "epoch": 0.39827368421052634,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003376383205103748,
+      "loss": 3.0962,
+      "step": 9459
+    },
+    {
+      "epoch": 0.39831578947368423,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003376067167481514,
+      "loss": 3.5966,
+      "step": 9460
+    },
+    {
+      "epoch": 0.3983578947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00033757511138983654,
+      "loss": 3.5273,
+      "step": 9461
+    },
+    {
+      "epoch": 0.3984,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003375435044360061,
+      "loss": 3.3945,
+      "step": 9462
+    },
+    {
+      "epoch": 0.3984421052631579,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003375118958872358,
+      "loss": 3.7384,
+      "step": 9463
+    },
+    {
+      "epoch": 0.3984842105263158,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003374802857441016,
+      "loss": 3.5527,
+      "step": 9464
+    },
+    {
+      "epoch": 0.3985263157894737,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0003374486740071794,
+      "loss": 3.0801,
+      "step": 9465
+    },
+    {
+      "epoch": 0.39856842105263157,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003374170606770451,
+      "loss": 3.1779,
+      "step": 9466
+    },
+    {
+      "epoch": 0.39861052631578947,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003373854457542746,
+      "loss": 3.4532,
+      "step": 9467
+    },
+    {
+      "epoch": 0.39865263157894737,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.000337353829239444,
+      "loss": 3.077,
+      "step": 9468
+    },
+    {
+      "epoch": 0.39869473684210527,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003373222111331293,
+      "loss": 2.9798,
+      "step": 9469
+    },
+    {
+      "epoch": 0.39873684210526317,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00033729059143590645,
+      "loss": 3.1474,
+      "step": 9470
+    },
+    {
+      "epoch": 0.39877894736842107,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003372589701483517,
+      "loss": 2.9066,
+      "step": 9471
+    },
+    {
+      "epoch": 0.39882105263157897,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003372273472710409,
+      "loss": 3.2572,
+      "step": 9472
+    },
+    {
+      "epoch": 0.39886315789473686,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00033719572280455047,
+      "loss": 3.2899,
+      "step": 9473
+    },
+    {
+      "epoch": 0.39890526315789476,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003371640967494563,
+      "loss": 3.5375,
+      "step": 9474
+    },
+    {
+      "epoch": 0.3989473684210526,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003371324691063347,
+      "loss": 2.7358,
+      "step": 9475
+    },
+    {
+      "epoch": 0.3989894736842105,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003371008398757619,
+      "loss": 3.2017,
+      "step": 9476
+    },
+    {
+      "epoch": 0.3990315789473684,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003370692090583142,
+      "loss": 2.9936,
+      "step": 9477
+    },
+    {
+      "epoch": 0.3990736842105263,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00033703757665456773,
+      "loss": 3.5172,
+      "step": 9478
+    },
+    {
+      "epoch": 0.3991157894736842,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.000337005942665099,
+      "loss": 3.5242,
+      "step": 9479
+    },
+    {
+      "epoch": 0.3991578947368421,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00033697430709048415,
+      "loss": 3.3897,
+      "step": 9480
+    },
+    {
+      "epoch": 0.3992,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00033694266993129963,
+      "loss": 2.9709,
+      "step": 9481
+    },
+    {
+      "epoch": 0.3992421052631579,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003369110311881217,
+      "loss": 3.229,
+      "step": 9482
+    },
+    {
+      "epoch": 0.3992842105263158,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.000336879390861527,
+      "loss": 2.7598,
+      "step": 9483
+    },
+    {
+      "epoch": 0.3993263157894737,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00033684774895209184,
+      "loss": 3.261,
+      "step": 9484
+    },
+    {
+      "epoch": 0.3993684210526316,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00033681610546039284,
+      "loss": 3.0724,
+      "step": 9485
+    },
+    {
+      "epoch": 0.3994105263157895,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003367844603870063,
+      "loss": 3.4202,
+      "step": 9486
+    },
+    {
+      "epoch": 0.3994526315789474,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00033675281373250895,
+      "loss": 3.1828,
+      "step": 9487
+    },
+    {
+      "epoch": 0.39949473684210524,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003367211654974772,
+      "loss": 3.0022,
+      "step": 9488
+    },
+    {
+      "epoch": 0.39953684210526313,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003366895156824877,
+      "loss": 3.2549,
+      "step": 9489
+    },
+    {
+      "epoch": 0.39957894736842103,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00033665786428811707,
+      "loss": 3.2343,
+      "step": 9490
+    },
+    {
+      "epoch": 0.39962105263157893,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.000336626211314942,
+      "loss": 3.3211,
+      "step": 9491
+    },
+    {
+      "epoch": 0.39966315789473683,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003365945567635391,
+      "loss": 3.2794,
+      "step": 9492
+    },
+    {
+      "epoch": 0.39970526315789473,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00033656290063448517,
+      "loss": 2.8306,
+      "step": 9493
+    },
+    {
+      "epoch": 0.39974736842105263,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003365312429283569,
+      "loss": 2.3731,
+      "step": 9494
+    },
+    {
+      "epoch": 0.39978947368421053,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.000336499583645731,
+      "loss": 2.9537,
+      "step": 9495
+    },
+    {
+      "epoch": 0.3998315789473684,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003364679227871844,
+      "loss": 2.9668,
+      "step": 9496
+    },
+    {
+      "epoch": 0.3998736842105263,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003364362603532938,
+      "loss": 2.8881,
+      "step": 9497
+    },
+    {
+      "epoch": 0.3999157894736842,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003364045963446361,
+      "loss": 3.194,
+      "step": 9498
+    },
+    {
+      "epoch": 0.3999578947368421,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0003363729307617882,
+      "loss": 3.1098,
+      "step": 9499
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003363412636053269,
+      "loss": 3.5797,
+      "step": 9500
+    },
+    {
+      "epoch": 0.4000421052631579,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0003363095948758293,
+      "loss": 3.3311,
+      "step": 9501
+    },
+    {
+      "epoch": 0.40008421052631576,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003362779245738722,
+      "loss": 3.476,
+      "step": 9502
+    },
+    {
+      "epoch": 0.40012631578947366,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00033624625270003283,
+      "loss": 3.073,
+      "step": 9503
+    },
+    {
+      "epoch": 0.40016842105263156,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00033621457925488797,
+      "loss": 3.0007,
+      "step": 9504
+    },
+    {
+      "epoch": 0.40021052631578946,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003361829042390148,
+      "loss": 3.6113,
+      "step": 9505
+    },
+    {
+      "epoch": 0.40025263157894736,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00033615122765299045,
+      "loss": 3.2049,
+      "step": 9506
+    },
+    {
+      "epoch": 0.40029473684210526,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003361195494973919,
+      "loss": 3.5253,
+      "step": 9507
+    },
+    {
+      "epoch": 0.40033684210526316,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00033608786977279626,
+      "loss": 3.234,
+      "step": 9508
+    },
+    {
+      "epoch": 0.40037894736842106,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003360561884797809,
+      "loss": 3.1876,
+      "step": 9509
+    },
+    {
+      "epoch": 0.40042105263157896,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003360245056189228,
+      "loss": 3.2033,
+      "step": 9510
+    },
+    {
+      "epoch": 0.40046315789473685,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003359928211907993,
+      "loss": 3.1127,
+      "step": 9511
+    },
+    {
+      "epoch": 0.40050526315789475,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003359611351959877,
+      "loss": 3.5139,
+      "step": 9512
+    },
+    {
+      "epoch": 0.40054736842105265,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003359294476350652,
+      "loss": 3.2633,
+      "step": 9513
+    },
+    {
+      "epoch": 0.40058947368421055,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00033589775850860915,
+      "loss": 3.2925,
+      "step": 9514
+    },
+    {
+      "epoch": 0.4006315789473684,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00033586606781719683,
+      "loss": 3.0153,
+      "step": 9515
+    },
+    {
+      "epoch": 0.4006736842105263,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00033583437556140557,
+      "loss": 2.9299,
+      "step": 9516
+    },
+    {
+      "epoch": 0.4007157894736842,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00033580268174181297,
+      "loss": 3.1478,
+      "step": 9517
+    },
+    {
+      "epoch": 0.4007578947368421,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00033577098635899615,
+      "loss": 3.4643,
+      "step": 9518
+    },
+    {
+      "epoch": 0.4008,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0003357392894135329,
+      "loss": 3.1587,
+      "step": 9519
+    },
+    {
+      "epoch": 0.4008421052631579,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00033570759090600043,
+      "loss": 3.1527,
+      "step": 9520
+    },
+    {
+      "epoch": 0.4008842105263158,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00033567589083697647,
+      "loss": 3.3283,
+      "step": 9521
+    },
+    {
+      "epoch": 0.4009263157894737,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00033564418920703834,
+      "loss": 3.4517,
+      "step": 9522
+    },
+    {
+      "epoch": 0.4009684210526316,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.0003356124860167637,
+      "loss": 3.0725,
+      "step": 9523
+    },
+    {
+      "epoch": 0.4010105263157895,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00033558078126673033,
+      "loss": 3.3593,
+      "step": 9524
+    },
+    {
+      "epoch": 0.4010526315789474,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00033554907495751555,
+      "loss": 3.31,
+      "step": 9525
+    },
+    {
+      "epoch": 0.4010947368421053,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0003355173670896971,
+      "loss": 2.5479,
+      "step": 9526
+    },
+    {
+      "epoch": 0.4011368421052632,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00033548565766385274,
+      "loss": 3.721,
+      "step": 9527
+    },
+    {
+      "epoch": 0.4011789473684211,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00033545394668056016,
+      "loss": 2.8006,
+      "step": 9528
+    },
+    {
+      "epoch": 0.4012210526315789,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.000335422234140397,
+      "loss": 3.1794,
+      "step": 9529
+    },
+    {
+      "epoch": 0.4012631578947368,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0003353905200439412,
+      "loss": 3.2252,
+      "step": 9530
+    },
+    {
+      "epoch": 0.4013052631578947,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00033535880439177043,
+      "loss": 3.2743,
+      "step": 9531
+    },
+    {
+      "epoch": 0.4013473684210526,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0003353270871844625,
+      "loss": 2.9152,
+      "step": 9532
+    },
+    {
+      "epoch": 0.4013894736842105,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003352953684225953,
+      "loss": 3.2095,
+      "step": 9533
+    },
+    {
+      "epoch": 0.4014315789473684,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00033526364810674674,
+      "loss": 3.0716,
+      "step": 9534
+    },
+    {
+      "epoch": 0.4014736842105263,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0003352319262374947,
+      "loss": 3.0851,
+      "step": 9535
+    },
+    {
+      "epoch": 0.4015157894736842,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003352002028154171,
+      "loss": 2.9024,
+      "step": 9536
+    },
+    {
+      "epoch": 0.4015578947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00033516847784109184,
+      "loss": 3.2751,
+      "step": 9537
+    },
+    {
+      "epoch": 0.4016,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003351367513150971,
+      "loss": 2.552,
+      "step": 9538
+    },
+    {
+      "epoch": 0.4016421052631579,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003351050232380107,
+      "loss": 3.4615,
+      "step": 9539
+    },
+    {
+      "epoch": 0.4016842105263158,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0003350732936104108,
+      "loss": 3.3881,
+      "step": 9540
+    },
+    {
+      "epoch": 0.4017263157894737,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0003350415624328754,
+      "loss": 2.6621,
+      "step": 9541
+    },
+    {
+      "epoch": 0.40176842105263155,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00033500982970598273,
+      "loss": 3.0586,
+      "step": 9542
+    },
+    {
+      "epoch": 0.40181052631578945,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0003349780954303108,
+      "loss": 3.2474,
+      "step": 9543
+    },
+    {
+      "epoch": 0.40185263157894735,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00033494635960643775,
+      "loss": 3.3123,
+      "step": 9544
+    },
+    {
+      "epoch": 0.40189473684210525,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00033491462223494197,
+      "loss": 2.974,
+      "step": 9545
+    },
+    {
+      "epoch": 0.40193684210526315,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00033488288331640147,
+      "loss": 3.0403,
+      "step": 9546
+    },
+    {
+      "epoch": 0.40197894736842105,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003348511428513946,
+      "loss": 3.1178,
+      "step": 9547
+    },
+    {
+      "epoch": 0.40202105263157895,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003348194008404996,
+      "loss": 3.0713,
+      "step": 9548
+    },
+    {
+      "epoch": 0.40206315789473684,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003347876572842947,
+      "loss": 3.0525,
+      "step": 9549
+    },
+    {
+      "epoch": 0.40210526315789474,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00033475591218335836,
+      "loss": 3.2282,
+      "step": 9550
+    },
+    {
+      "epoch": 0.40214736842105264,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003347241655382689,
+      "loss": 3.1419,
+      "step": 9551
+    },
+    {
+      "epoch": 0.40218947368421054,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003346924173496046,
+      "loss": 3.5581,
+      "step": 9552
+    },
+    {
+      "epoch": 0.40223157894736844,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00033466066761794413,
+      "loss": 2.9949,
+      "step": 9553
+    },
+    {
+      "epoch": 0.40227368421052634,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003346289163438657,
+      "loss": 3.4723,
+      "step": 9554
+    },
+    {
+      "epoch": 0.40231578947368424,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003345971635279478,
+      "loss": 3.3126,
+      "step": 9555
+    },
+    {
+      "epoch": 0.4023578947368421,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.000334565409170769,
+      "loss": 3.5224,
+      "step": 9556
+    },
+    {
+      "epoch": 0.4024,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00033453365327290786,
+      "loss": 3.1665,
+      "step": 9557
+    },
+    {
+      "epoch": 0.4024421052631579,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00033450189583494286,
+      "loss": 3.6346,
+      "step": 9558
+    },
+    {
+      "epoch": 0.4024842105263158,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003344701368574525,
+      "loss": 2.9938,
+      "step": 9559
+    },
+    {
+      "epoch": 0.4025263157894737,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003344383763410156,
+      "loss": 3.3557,
+      "step": 9560
+    },
+    {
+      "epoch": 0.4025684210526316,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00033440661428621064,
+      "loss": 2.9688,
+      "step": 9561
+    },
+    {
+      "epoch": 0.4026105263157895,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003343748506936164,
+      "loss": 2.9146,
+      "step": 9562
+    },
+    {
+      "epoch": 0.4026526315789474,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003343430855638115,
+      "loss": 3.3913,
+      "step": 9563
+    },
+    {
+      "epoch": 0.40269473684210527,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00033431131889737466,
+      "loss": 3.3432,
+      "step": 9564
+    },
+    {
+      "epoch": 0.40273684210526317,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003342795506948847,
+      "loss": 3.1368,
+      "step": 9565
+    },
+    {
+      "epoch": 0.40277894736842107,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00033424778095692025,
+      "loss": 3.4826,
+      "step": 9566
+    },
+    {
+      "epoch": 0.40282105263157897,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003342160096840603,
+      "loss": 3.2695,
+      "step": 9567
+    },
+    {
+      "epoch": 0.40286315789473687,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00033418423687688357,
+      "loss": 3.2549,
+      "step": 9568
+    },
+    {
+      "epoch": 0.4029052631578947,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.000334152462535969,
+      "loss": 3.4559,
+      "step": 9569
+    },
+    {
+      "epoch": 0.4029473684210526,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00033412068666189535,
+      "loss": 3.5343,
+      "step": 9570
+    },
+    {
+      "epoch": 0.4029894736842105,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003340889092552417,
+      "loss": 3.3089,
+      "step": 9571
+    },
+    {
+      "epoch": 0.4030315789473684,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003340571303165869,
+      "loss": 2.8076,
+      "step": 9572
+    },
+    {
+      "epoch": 0.4030736842105263,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00033402534984650994,
+      "loss": 3.341,
+      "step": 9573
+    },
+    {
+      "epoch": 0.4031157894736842,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00033399356784558987,
+      "loss": 3.4694,
+      "step": 9574
+    },
+    {
+      "epoch": 0.4031578947368421,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003339617843144057,
+      "loss": 3.0503,
+      "step": 9575
+    },
+    {
+      "epoch": 0.4032,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003339299992535364,
+      "loss": 3.153,
+      "step": 9576
+    },
+    {
+      "epoch": 0.4032421052631579,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0003338982126635611,
+      "loss": 3.301,
+      "step": 9577
+    },
+    {
+      "epoch": 0.4032842105263158,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.000333866424545059,
+      "loss": 3.4547,
+      "step": 9578
+    },
+    {
+      "epoch": 0.4033263157894737,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003338346348986092,
+      "loss": 3.3145,
+      "step": 9579
+    },
+    {
+      "epoch": 0.4033684210526316,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00033380284372479084,
+      "loss": 2.9829,
+      "step": 9580
+    },
+    {
+      "epoch": 0.4034105263157895,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00033377105102418316,
+      "loss": 2.9806,
+      "step": 9581
+    },
+    {
+      "epoch": 0.4034526315789474,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003337392567973653,
+      "loss": 3.078,
+      "step": 9582
+    },
+    {
+      "epoch": 0.40349473684210524,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0003337074610449166,
+      "loss": 3.2246,
+      "step": 9583
+    },
+    {
+      "epoch": 0.40353684210526314,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003336756637674163,
+      "loss": 2.7454,
+      "step": 9584
+    },
+    {
+      "epoch": 0.40357894736842104,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003336438649654437,
+      "loss": 3.0879,
+      "step": 9585
+    },
+    {
+      "epoch": 0.40362105263157894,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00033361206463957814,
+      "loss": 3.4993,
+      "step": 9586
+    },
+    {
+      "epoch": 0.40366315789473683,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00033358026279039913,
+      "loss": 2.9112,
+      "step": 9587
+    },
+    {
+      "epoch": 0.40370526315789473,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003335484594184858,
+      "loss": 3.2015,
+      "step": 9588
+    },
+    {
+      "epoch": 0.40374736842105263,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003335166545244178,
+      "loss": 3.2426,
+      "step": 9589
+    },
+    {
+      "epoch": 0.40378947368421053,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00033348484810877445,
+      "loss": 3.646,
+      "step": 9590
+    },
+    {
+      "epoch": 0.40383157894736843,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00033345304017213525,
+      "loss": 3.3021,
+      "step": 9591
+    },
+    {
+      "epoch": 0.40387368421052633,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00033342123071507967,
+      "loss": 3.0405,
+      "step": 9592
+    },
+    {
+      "epoch": 0.40391578947368423,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003333894197381873,
+      "loss": 3.0886,
+      "step": 9593
+    },
+    {
+      "epoch": 0.4039578947368421,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00033335760724203774,
+      "loss": 2.9195,
+      "step": 9594
+    },
+    {
+      "epoch": 0.404,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00033332579322721046,
+      "loss": 3.2085,
+      "step": 9595
+    },
+    {
+      "epoch": 0.40404210526315787,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00033329397769428515,
+      "loss": 3.441,
+      "step": 9596
+    },
+    {
+      "epoch": 0.40408421052631577,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003332621606438414,
+      "loss": 3.1667,
+      "step": 9597
+    },
+    {
+      "epoch": 0.40412631578947367,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00033323034207645896,
+      "loss": 3.125,
+      "step": 9598
+    },
+    {
+      "epoch": 0.40416842105263157,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0003331985219927175,
+      "loss": 3.143,
+      "step": 9599
+    },
+    {
+      "epoch": 0.40421052631578946,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003331667003931967,
+      "loss": 3.0768,
+      "step": 9600
+    },
+    {
+      "epoch": 0.40425263157894736,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0003331348772784762,
+      "loss": 2.6815,
+      "step": 9601
+    },
+    {
+      "epoch": 0.40429473684210526,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00033310305264913615,
+      "loss": 2.9858,
+      "step": 9602
+    },
+    {
+      "epoch": 0.40433684210526316,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00033307122650575593,
+      "loss": 3.0674,
+      "step": 9603
+    },
+    {
+      "epoch": 0.40437894736842106,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0003330393988489156,
+      "loss": 3.166,
+      "step": 9604
+    },
+    {
+      "epoch": 0.40442105263157896,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00033300756967919517,
+      "loss": 3.4479,
+      "step": 9605
+    },
+    {
+      "epoch": 0.40446315789473686,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0003329757389971742,
+      "loss": 3.4272,
+      "step": 9606
+    },
+    {
+      "epoch": 0.40450526315789476,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003329439068034328,
+      "loss": 3.4884,
+      "step": 9607
+    },
+    {
+      "epoch": 0.40454736842105266,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00033291207309855085,
+      "loss": 3.1125,
+      "step": 9608
+    },
+    {
+      "epoch": 0.4045894736842105,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003328802378831083,
+      "loss": 3.3973,
+      "step": 9609
+    },
+    {
+      "epoch": 0.4046315789473684,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00033284840115768526,
+      "loss": 3.26,
+      "step": 9610
+    },
+    {
+      "epoch": 0.4046736842105263,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003328165629228617,
+      "loss": 3.6006,
+      "step": 9611
+    },
+    {
+      "epoch": 0.4047157894736842,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003327847231792176,
+      "loss": 3.4155,
+      "step": 9612
+    },
+    {
+      "epoch": 0.4047578947368421,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00033275288192733315,
+      "loss": 2.8228,
+      "step": 9613
+    },
+    {
+      "epoch": 0.4048,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00033272103916778844,
+      "loss": 3.4277,
+      "step": 9614
+    },
+    {
+      "epoch": 0.4048421052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003326891949011636,
+      "loss": 3.0916,
+      "step": 9615
+    },
+    {
+      "epoch": 0.4048842105263158,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003326573491280388,
+      "loss": 3.5064,
+      "step": 9616
+    },
+    {
+      "epoch": 0.4049263157894737,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00033262550184899413,
+      "loss": 3.0846,
+      "step": 9617
+    },
+    {
+      "epoch": 0.4049684210526316,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00033259365306461,
+      "loss": 2.8935,
+      "step": 9618
+    },
+    {
+      "epoch": 0.4050105263157895,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003325618027754664,
+      "loss": 3.2446,
+      "step": 9619
+    },
+    {
+      "epoch": 0.4050526315789474,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0003325299509821439,
+      "loss": 3.2455,
+      "step": 9620
+    },
+    {
+      "epoch": 0.4050947368421053,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0003324980976852227,
+      "loss": 3.3241,
+      "step": 9621
+    },
+    {
+      "epoch": 0.4051368421052632,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00033246624288528297,
+      "loss": 3.1677,
+      "step": 9622
+    },
+    {
+      "epoch": 0.405178947368421,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0003324343865829052,
+      "loss": 3.3264,
+      "step": 9623
+    },
+    {
+      "epoch": 0.4052210526315789,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003324025287786698,
+      "loss": 3.0443,
+      "step": 9624
+    },
+    {
+      "epoch": 0.4052631578947368,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003323706694731572,
+      "loss": 3.5171,
+      "step": 9625
+    },
+    {
+      "epoch": 0.4053052631578947,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003323388086669477,
+      "loss": 3.3871,
+      "step": 9626
+    },
+    {
+      "epoch": 0.4053473684210526,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00033230694636062185,
+      "loss": 2.7277,
+      "step": 9627
+    },
+    {
+      "epoch": 0.4053894736842105,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0003322750825547602,
+      "loss": 3.2167,
+      "step": 9628
+    },
+    {
+      "epoch": 0.4054315789473684,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00033224321724994316,
+      "loss": 3.1942,
+      "step": 9629
+    },
+    {
+      "epoch": 0.4054736842105263,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003322113504467513,
+      "loss": 2.9969,
+      "step": 9630
+    },
+    {
+      "epoch": 0.4055157894736842,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003321794821457653,
+      "loss": 3.6783,
+      "step": 9631
+    },
+    {
+      "epoch": 0.4055578947368421,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00033214761234756575,
+      "loss": 3.2277,
+      "step": 9632
+    },
+    {
+      "epoch": 0.4056,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00033211574105273314,
+      "loss": 3.4145,
+      "step": 9633
+    },
+    {
+      "epoch": 0.4056421052631579,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003320838682618482,
+      "loss": 3.4892,
+      "step": 9634
+    },
+    {
+      "epoch": 0.4056842105263158,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0003320519939754917,
+      "loss": 2.8889,
+      "step": 9635
+    },
+    {
+      "epoch": 0.40572631578947366,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00033202011819424414,
+      "loss": 3.0783,
+      "step": 9636
+    },
+    {
+      "epoch": 0.40576842105263156,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003319882409186865,
+      "loss": 3.6964,
+      "step": 9637
+    },
+    {
+      "epoch": 0.40581052631578945,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003319563621493994,
+      "loss": 3.2361,
+      "step": 9638
+    },
+    {
+      "epoch": 0.40585263157894735,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00033192448188696375,
+      "loss": 3.5128,
+      "step": 9639
+    },
+    {
+      "epoch": 0.40589473684210525,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00033189260013196027,
+      "loss": 3.2741,
+      "step": 9640
+    },
+    {
+      "epoch": 0.40593684210526315,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00033186071688496975,
+      "loss": 3.055,
+      "step": 9641
+    },
+    {
+      "epoch": 0.40597894736842105,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00033182883214657327,
+      "loss": 3.0144,
+      "step": 9642
+    },
+    {
+      "epoch": 0.40602105263157895,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00033179694591735153,
+      "loss": 3.4203,
+      "step": 9643
+    },
+    {
+      "epoch": 0.40606315789473685,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0003317650581978855,
+      "loss": 2.8581,
+      "step": 9644
+    },
+    {
+      "epoch": 0.40610526315789475,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003317331689887563,
+      "loss": 3.7203,
+      "step": 9645
+    },
+    {
+      "epoch": 0.40614736842105265,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00033170127829054474,
+      "loss": 3.0822,
+      "step": 9646
+    },
+    {
+      "epoch": 0.40618947368421054,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0003316693861038319,
+      "loss": 3.0134,
+      "step": 9647
+    },
+    {
+      "epoch": 0.40623157894736844,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00033163749242919884,
+      "loss": 2.9565,
+      "step": 9648
+    },
+    {
+      "epoch": 0.40627368421052634,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0003316055972672265,
+      "loss": 3.0561,
+      "step": 9649
+    },
+    {
+      "epoch": 0.4063157894736842,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00033157370061849604,
+      "loss": 3.2031,
+      "step": 9650
+    },
+    {
+      "epoch": 0.4063578947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00033154180248358867,
+      "loss": 3.0986,
+      "step": 9651
+    },
+    {
+      "epoch": 0.4064,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00033150990286308547,
+      "loss": 3.2941,
+      "step": 9652
+    },
+    {
+      "epoch": 0.4064421052631579,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003314780017575676,
+      "loss": 3.1769,
+      "step": 9653
+    },
+    {
+      "epoch": 0.4064842105263158,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003314460991676163,
+      "loss": 3.2008,
+      "step": 9654
+    },
+    {
+      "epoch": 0.4065263157894737,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003314141950938127,
+      "loss": 3.3516,
+      "step": 9655
+    },
+    {
+      "epoch": 0.4065684210526316,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003313822895367382,
+      "loss": 3.0113,
+      "step": 9656
+    },
+    {
+      "epoch": 0.4066105263157895,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00033135038249697393,
+      "loss": 3.2706,
+      "step": 9657
+    },
+    {
+      "epoch": 0.4066526315789474,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0003313184739751013,
+      "loss": 3.1267,
+      "step": 9658
+    },
+    {
+      "epoch": 0.4066947368421053,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00033128656397170167,
+      "loss": 2.7876,
+      "step": 9659
+    },
+    {
+      "epoch": 0.4067368421052632,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003312546524873562,
+      "loss": 3.3756,
+      "step": 9660
+    },
+    {
+      "epoch": 0.4067789473684211,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00033122273952264656,
+      "loss": 3.3481,
+      "step": 9661
+    },
+    {
+      "epoch": 0.40682105263157897,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.000331190825078154,
+      "loss": 3.3248,
+      "step": 9662
+    },
+    {
+      "epoch": 0.4068631578947368,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00033115890915446007,
+      "loss": 2.9776,
+      "step": 9663
+    },
+    {
+      "epoch": 0.4069052631578947,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003311269917521461,
+      "loss": 2.8864,
+      "step": 9664
+    },
+    {
+      "epoch": 0.4069473684210526,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00033109507287179367,
+      "loss": 3.0867,
+      "step": 9665
+    },
+    {
+      "epoch": 0.4069894736842105,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00033106315251398434,
+      "loss": 3.1672,
+      "step": 9666
+    },
+    {
+      "epoch": 0.4070315789473684,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0003310312306792996,
+      "loss": 3.0602,
+      "step": 9667
+    },
+    {
+      "epoch": 0.4070736842105263,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00033099930736832096,
+      "loss": 2.9953,
+      "step": 9668
+    },
+    {
+      "epoch": 0.4071157894736842,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0003309673825816302,
+      "loss": 3.3506,
+      "step": 9669
+    },
+    {
+      "epoch": 0.4071578947368421,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0003309354563198088,
+      "loss": 2.8665,
+      "step": 9670
+    },
+    {
+      "epoch": 0.4072,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00033090352858343853,
+      "loss": 2.6563,
+      "step": 9671
+    },
+    {
+      "epoch": 0.4072421052631579,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00033087159937310106,
+      "loss": 2.9886,
+      "step": 9672
+    },
+    {
+      "epoch": 0.4072842105263158,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00033083966868937805,
+      "loss": 3.4338,
+      "step": 9673
+    },
+    {
+      "epoch": 0.4073263157894737,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003308077365328512,
+      "loss": 3.3274,
+      "step": 9674
+    },
+    {
+      "epoch": 0.4073684210526316,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003307758029041024,
+      "loss": 3.4929,
+      "step": 9675
+    },
+    {
+      "epoch": 0.4074105263157895,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00033074386780371337,
+      "loss": 3.0753,
+      "step": 9676
+    },
+    {
+      "epoch": 0.40745263157894734,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0003307119312322659,
+      "loss": 3.265,
+      "step": 9677
+    },
+    {
+      "epoch": 0.40749473684210524,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003306799931903419,
+      "loss": 3.2584,
+      "step": 9678
+    },
+    {
+      "epoch": 0.40753684210526314,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00033064805367852317,
+      "loss": 3.1428,
+      "step": 9679
+    },
+    {
+      "epoch": 0.40757894736842104,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003306161126973918,
+      "loss": 3.2893,
+      "step": 9680
+    },
+    {
+      "epoch": 0.40762105263157894,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003305841702475295,
+      "loss": 3.3859,
+      "step": 9681
+    },
+    {
+      "epoch": 0.40766315789473684,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00033055222632951824,
+      "loss": 3.5214,
+      "step": 9682
+    },
+    {
+      "epoch": 0.40770526315789474,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003305202809439401,
+      "loss": 3.074,
+      "step": 9683
+    },
+    {
+      "epoch": 0.40774736842105264,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.000330488334091377,
+      "loss": 2.8221,
+      "step": 9684
+    },
+    {
+      "epoch": 0.40778947368421053,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003304563857724111,
+      "loss": 3.2497,
+      "step": 9685
+    },
+    {
+      "epoch": 0.40783157894736843,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0003304244359876243,
+      "loss": 3.4423,
+      "step": 9686
+    },
+    {
+      "epoch": 0.40787368421052633,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00033039248473759885,
+      "loss": 3.5976,
+      "step": 9687
+    },
+    {
+      "epoch": 0.40791578947368423,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00033036053202291685,
+      "loss": 3.0927,
+      "step": 9688
+    },
+    {
+      "epoch": 0.40795789473684213,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003303285778441603,
+      "loss": 3.2877,
+      "step": 9689
+    },
+    {
+      "epoch": 0.408,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003302966222019114,
+      "loss": 3.482,
+      "step": 9690
+    },
+    {
+      "epoch": 0.4080421052631579,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003302646650967524,
+      "loss": 3.1242,
+      "step": 9691
+    },
+    {
+      "epoch": 0.40808421052631577,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003302327065292655,
+      "loss": 2.9952,
+      "step": 9692
+    },
+    {
+      "epoch": 0.40812631578947367,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00033020074650003307,
+      "loss": 3.0478,
+      "step": 9693
+    },
+    {
+      "epoch": 0.40816842105263157,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00033016878500963714,
+      "loss": 2.9878,
+      "step": 9694
+    },
+    {
+      "epoch": 0.40821052631578947,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0003301368220586602,
+      "loss": 3.0597,
+      "step": 9695
+    },
+    {
+      "epoch": 0.40825263157894737,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003301048576476845,
+      "loss": 3.2524,
+      "step": 9696
+    },
+    {
+      "epoch": 0.40829473684210527,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00033007289177729236,
+      "loss": 3.2568,
+      "step": 9697
+    },
+    {
+      "epoch": 0.40833684210526316,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0003300409244480663,
+      "loss": 3.2716,
+      "step": 9698
+    },
+    {
+      "epoch": 0.40837894736842106,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00033000895566058866,
+      "loss": 3.1387,
+      "step": 9699
+    },
+    {
+      "epoch": 0.40842105263157896,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00032997698541544184,
+      "loss": 3.0485,
+      "step": 9700
+    },
+    {
+      "epoch": 0.40846315789473686,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0003299450137132083,
+      "loss": 3.2091,
+      "step": 9701
+    },
+    {
+      "epoch": 0.40850526315789476,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00032991304055447046,
+      "loss": 3.1921,
+      "step": 9702
+    },
+    {
+      "epoch": 0.40854736842105266,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00032988106593981103,
+      "loss": 3.1941,
+      "step": 9703
+    },
+    {
+      "epoch": 0.4085894736842105,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00032984908986981234,
+      "loss": 2.709,
+      "step": 9704
+    },
+    {
+      "epoch": 0.4086315789473684,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00032981711234505713,
+      "loss": 3.2018,
+      "step": 9705
+    },
+    {
+      "epoch": 0.4086736842105263,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00032978513336612785,
+      "loss": 3.3015,
+      "step": 9706
+    },
+    {
+      "epoch": 0.4087157894736842,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00032975315293360725,
+      "loss": 3.074,
+      "step": 9707
+    },
+    {
+      "epoch": 0.4087578947368421,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00032972117104807793,
+      "loss": 3.4442,
+      "step": 9708
+    },
+    {
+      "epoch": 0.4088,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00032968918771012244,
+      "loss": 3.0092,
+      "step": 9709
+    },
+    {
+      "epoch": 0.4088421052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00032965720292032363,
+      "loss": 2.7784,
+      "step": 9710
+    },
+    {
+      "epoch": 0.4088842105263158,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00032962521667926416,
+      "loss": 3.1199,
+      "step": 9711
+    },
+    {
+      "epoch": 0.4089263157894737,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003295932289875268,
+      "loss": 3.5696,
+      "step": 9712
+    },
+    {
+      "epoch": 0.4089684210526316,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003295612398456944,
+      "loss": 3.0235,
+      "step": 9713
+    },
+    {
+      "epoch": 0.4090105263157895,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00032952924925434965,
+      "loss": 3.1547,
+      "step": 9714
+    },
+    {
+      "epoch": 0.4090526315789474,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0003294972572140754,
+      "loss": 3.38,
+      "step": 9715
+    },
+    {
+      "epoch": 0.4090947368421053,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00032946526372545454,
+      "loss": 3.6133,
+      "step": 9716
+    },
+    {
+      "epoch": 0.40913684210526313,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00032943326878906996,
+      "loss": 2.7362,
+      "step": 9717
+    },
+    {
+      "epoch": 0.40917894736842103,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00032940127240550446,
+      "loss": 3.6014,
+      "step": 9718
+    },
+    {
+      "epoch": 0.40922105263157893,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003293692745753412,
+      "loss": 2.9918,
+      "step": 9719
+    },
+    {
+      "epoch": 0.40926315789473683,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00032933727529916283,
+      "loss": 2.5423,
+      "step": 9720
+    },
+    {
+      "epoch": 0.4093052631578947,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00032930527457755265,
+      "loss": 3.0446,
+      "step": 9721
+    },
+    {
+      "epoch": 0.4093473684210526,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003292732724110936,
+      "loss": 3.2719,
+      "step": 9722
+    },
+    {
+      "epoch": 0.4093894736842105,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00032924126880036866,
+      "loss": 3.3202,
+      "step": 9723
+    },
+    {
+      "epoch": 0.4094315789473684,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0003292092637459608,
+      "loss": 2.8208,
+      "step": 9724
+    },
+    {
+      "epoch": 0.4094736842105263,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003291772572484533,
+      "loss": 3.4516,
+      "step": 9725
+    },
+    {
+      "epoch": 0.4095157894736842,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00032914524930842913,
+      "loss": 3.4475,
+      "step": 9726
+    },
+    {
+      "epoch": 0.4095578947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00032911323992647156,
+      "loss": 2.9029,
+      "step": 9727
+    },
+    {
+      "epoch": 0.4096,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0003290812291031637,
+      "loss": 3.0591,
+      "step": 9728
+    },
+    {
+      "epoch": 0.4096421052631579,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00032904921683908877,
+      "loss": 2.5879,
+      "step": 9729
+    },
+    {
+      "epoch": 0.40968421052631576,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00032901720313483,
+      "loss": 3.1034,
+      "step": 9730
+    },
+    {
+      "epoch": 0.40972631578947366,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0003289851879909706,
+      "loss": 3.3222,
+      "step": 9731
+    },
+    {
+      "epoch": 0.40976842105263156,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0003289531714080939,
+      "loss": 3.222,
+      "step": 9732
+    },
+    {
+      "epoch": 0.40981052631578946,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003289211533867832,
+      "loss": 3.4566,
+      "step": 9733
+    },
+    {
+      "epoch": 0.40985263157894736,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00032888913392762183,
+      "loss": 2.7881,
+      "step": 9734
+    },
+    {
+      "epoch": 0.40989473684210526,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003288571130311931,
+      "loss": 3.228,
+      "step": 9735
+    },
+    {
+      "epoch": 0.40993684210526316,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00032882509069808043,
+      "loss": 3.2224,
+      "step": 9736
+    },
+    {
+      "epoch": 0.40997894736842105,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0003287930669288672,
+      "loss": 2.8991,
+      "step": 9737
+    },
+    {
+      "epoch": 0.41002105263157895,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00032876104172413685,
+      "loss": 3.139,
+      "step": 9738
+    },
+    {
+      "epoch": 0.41006315789473685,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003287290150844729,
+      "loss": 3.0915,
+      "step": 9739
+    },
+    {
+      "epoch": 0.41010526315789475,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003286969870104588,
+      "loss": 2.923,
+      "step": 9740
+    },
+    {
+      "epoch": 0.41014736842105265,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00032866495750267805,
+      "loss": 3.3063,
+      "step": 9741
+    },
+    {
+      "epoch": 0.41018947368421055,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003286329265617142,
+      "loss": 3.5232,
+      "step": 9742
+    },
+    {
+      "epoch": 0.41023157894736845,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00032860089418815084,
+      "loss": 2.81,
+      "step": 9743
+    },
+    {
+      "epoch": 0.4102736842105263,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003285688603825715,
+      "loss": 3.4731,
+      "step": 9744
+    },
+    {
+      "epoch": 0.4103157894736842,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00032853682514555983,
+      "loss": 3.2572,
+      "step": 9745
+    },
+    {
+      "epoch": 0.4103578947368421,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00032850478847769955,
+      "loss": 3.3471,
+      "step": 9746
+    },
+    {
+      "epoch": 0.4104,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00032847275037957424,
+      "loss": 2.5486,
+      "step": 9747
+    },
+    {
+      "epoch": 0.4104421052631579,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00032844071085176763,
+      "loss": 3.3354,
+      "step": 9748
+    },
+    {
+      "epoch": 0.4104842105263158,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003284086698948634,
+      "loss": 3.1434,
+      "step": 9749
+    },
+    {
+      "epoch": 0.4105263157894737,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00032837662750944535,
+      "loss": 3.2549,
+      "step": 9750
+    },
+    {
+      "epoch": 0.4105684210526316,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0003283445836960972,
+      "loss": 2.9685,
+      "step": 9751
+    },
+    {
+      "epoch": 0.4106105263157895,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003283125384554028,
+      "loss": 3.1856,
+      "step": 9752
+    },
+    {
+      "epoch": 0.4106526315789474,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00032828049178794595,
+      "loss": 2.879,
+      "step": 9753
+    },
+    {
+      "epoch": 0.4106947368421053,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003282484436943106,
+      "loss": 3.3675,
+      "step": 9754
+    },
+    {
+      "epoch": 0.4107368421052632,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003282163941750804,
+      "loss": 2.9492,
+      "step": 9755
+    },
+    {
+      "epoch": 0.4107789473684211,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00032818434323083956,
+      "loss": 3.3967,
+      "step": 9756
+    },
+    {
+      "epoch": 0.4108210526315789,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003281522908621718,
+      "loss": 3.2263,
+      "step": 9757
+    },
+    {
+      "epoch": 0.4108631578947368,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00032812023706966104,
+      "loss": 3.2892,
+      "step": 9758
+    },
+    {
+      "epoch": 0.4109052631578947,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003280881818538914,
+      "loss": 3.0762,
+      "step": 9759
+    },
+    {
+      "epoch": 0.4109473684210526,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00032805612521544686,
+      "loss": 3.3244,
+      "step": 9760
+    },
+    {
+      "epoch": 0.4109894736842105,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003280240671549114,
+      "loss": 3.7043,
+      "step": 9761
+    },
+    {
+      "epoch": 0.4110315789473684,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003279920076728691,
+      "loss": 3.1436,
+      "step": 9762
+    },
+    {
+      "epoch": 0.4110736842105263,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003279599467699041,
+      "loss": 3.0941,
+      "step": 9763
+    },
+    {
+      "epoch": 0.4111157894736842,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003279278844466005,
+      "loss": 3.4766,
+      "step": 9764
+    },
+    {
+      "epoch": 0.4111578947368421,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00032789582070354227,
+      "loss": 2.8604,
+      "step": 9765
+    },
+    {
+      "epoch": 0.4112,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0003278637555413138,
+      "loss": 2.9306,
+      "step": 9766
+    },
+    {
+      "epoch": 0.4112421052631579,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003278316889604993,
+      "loss": 3.516,
+      "step": 9767
+    },
+    {
+      "epoch": 0.4112842105263158,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0003277996209616827,
+      "loss": 3.2826,
+      "step": 9768
+    },
+    {
+      "epoch": 0.4113263157894737,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00032776755154544844,
+      "loss": 3.0299,
+      "step": 9769
+    },
+    {
+      "epoch": 0.4113684210526316,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003277354807123808,
+      "loss": 3.1031,
+      "step": 9770
+    },
+    {
+      "epoch": 0.41141052631578945,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00032770340846306403,
+      "loss": 3.3292,
+      "step": 9771
+    },
+    {
+      "epoch": 0.41145263157894735,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00032767133479808246,
+      "loss": 2.9134,
+      "step": 9772
+    },
+    {
+      "epoch": 0.41149473684210525,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00032763925971802043,
+      "loss": 2.5583,
+      "step": 9773
+    },
+    {
+      "epoch": 0.41153684210526315,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003276071832234624,
+      "loss": 3.3809,
+      "step": 9774
+    },
+    {
+      "epoch": 0.41157894736842104,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00032757510531499255,
+      "loss": 3.1292,
+      "step": 9775
+    },
+    {
+      "epoch": 0.41162105263157894,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003275430259931955,
+      "loss": 3.599,
+      "step": 9776
+    },
+    {
+      "epoch": 0.41166315789473684,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00032751094525865557,
+      "loss": 2.9445,
+      "step": 9777
+    },
+    {
+      "epoch": 0.41170526315789474,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00032747886311195736,
+      "loss": 3.0075,
+      "step": 9778
+    },
+    {
+      "epoch": 0.41174736842105264,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003274467795536852,
+      "loss": 3.2097,
+      "step": 9779
+    },
+    {
+      "epoch": 0.41178947368421054,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00032741469458442377,
+      "loss": 3.0515,
+      "step": 9780
+    },
+    {
+      "epoch": 0.41183157894736844,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003273826082047576,
+      "loss": 3.2655,
+      "step": 9781
+    },
+    {
+      "epoch": 0.41187368421052634,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00032735052041527117,
+      "loss": 3.4306,
+      "step": 9782
+    },
+    {
+      "epoch": 0.41191578947368424,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003273184312165492,
+      "loss": 2.6865,
+      "step": 9783
+    },
+    {
+      "epoch": 0.4119578947368421,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0003272863406091762,
+      "loss": 3.5431,
+      "step": 9784
+    },
+    {
+      "epoch": 0.412,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00032725424859373687,
+      "loss": 2.9599,
+      "step": 9785
+    },
+    {
+      "epoch": 0.4120421052631579,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003272221551708159,
+      "loss": 3.4338,
+      "step": 9786
+    },
+    {
+      "epoch": 0.4120842105263158,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00032719006034099803,
+      "loss": 3.3016,
+      "step": 9787
+    },
+    {
+      "epoch": 0.4121263157894737,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.000327157964104868,
+      "loss": 3.2408,
+      "step": 9788
+    },
+    {
+      "epoch": 0.4121684210526316,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00032712586646301044,
+      "loss": 3.2086,
+      "step": 9789
+    },
+    {
+      "epoch": 0.41221052631578947,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0003270937674160103,
+      "loss": 2.9384,
+      "step": 9790
+    },
+    {
+      "epoch": 0.41225263157894737,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003270616669644522,
+      "loss": 3.3137,
+      "step": 9791
+    },
+    {
+      "epoch": 0.41229473684210527,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00032702956510892114,
+      "loss": 2.9307,
+      "step": 9792
+    },
+    {
+      "epoch": 0.41233684210526317,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003269974618500019,
+      "loss": 3.3631,
+      "step": 9793
+    },
+    {
+      "epoch": 0.41237894736842107,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00032696535718827936,
+      "loss": 3.4386,
+      "step": 9794
+    },
+    {
+      "epoch": 0.41242105263157897,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00032693325112433835,
+      "loss": 2.9951,
+      "step": 9795
+    },
+    {
+      "epoch": 0.41246315789473686,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003269011436587641,
+      "loss": 3.3136,
+      "step": 9796
+    },
+    {
+      "epoch": 0.41250526315789476,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003268690347921412,
+      "loss": 3.2238,
+      "step": 9797
+    },
+    {
+      "epoch": 0.4125473684210526,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003268369245250548,
+      "loss": 2.987,
+      "step": 9798
+    },
+    {
+      "epoch": 0.4125894736842105,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00032680481285809007,
+      "loss": 2.9411,
+      "step": 9799
+    },
+    {
+      "epoch": 0.4126315789473684,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00032677269979183176,
+      "loss": 3.2473,
+      "step": 9800
+    },
+    {
+      "epoch": 0.4126736842105263,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003267405853268651,
+      "loss": 3.0204,
+      "step": 9801
+    },
+    {
+      "epoch": 0.4127157894736842,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0003267084694637752,
+      "loss": 3.4134,
+      "step": 9802
+    },
+    {
+      "epoch": 0.4127578947368421,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00032667635220314694,
+      "loss": 3.6243,
+      "step": 9803
+    },
+    {
+      "epoch": 0.4128,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003266442335455658,
+      "loss": 3.7101,
+      "step": 9804
+    },
+    {
+      "epoch": 0.4128421052631579,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00032661211349161665,
+      "loss": 3.4791,
+      "step": 9805
+    },
+    {
+      "epoch": 0.4128842105263158,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003265799920418849,
+      "loss": 3.1093,
+      "step": 9806
+    },
+    {
+      "epoch": 0.4129263157894737,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00032654786919695573,
+      "loss": 3.4331,
+      "step": 9807
+    },
+    {
+      "epoch": 0.4129684210526316,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00032651574495741426,
+      "loss": 3.4692,
+      "step": 9808
+    },
+    {
+      "epoch": 0.4130105263157895,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00032648361932384587,
+      "loss": 3.1984,
+      "step": 9809
+    },
+    {
+      "epoch": 0.4130526315789474,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003264514922968357,
+      "loss": 2.9987,
+      "step": 9810
+    },
+    {
+      "epoch": 0.41309473684210524,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0003264193638769692,
+      "loss": 3.7,
+      "step": 9811
+    },
+    {
+      "epoch": 0.41313684210526314,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003263872340648316,
+      "loss": 3.186,
+      "step": 9812
+    },
+    {
+      "epoch": 0.41317894736842103,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0003263551028610085,
+      "loss": 3.2846,
+      "step": 9813
+    },
+    {
+      "epoch": 0.41322105263157893,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.000326322970266085,
+      "loss": 3.1472,
+      "step": 9814
+    },
+    {
+      "epoch": 0.41326315789473683,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003262908362806468,
+      "loss": 3.083,
+      "step": 9815
+    },
+    {
+      "epoch": 0.41330526315789473,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00032625870090527913,
+      "loss": 2.8058,
+      "step": 9816
+    },
+    {
+      "epoch": 0.41334736842105263,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00032622656414056764,
+      "loss": 3.3321,
+      "step": 9817
+    },
+    {
+      "epoch": 0.41338947368421053,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003261944259870975,
+      "loss": 3.3838,
+      "step": 9818
+    },
+    {
+      "epoch": 0.4134315789473684,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003261622864454546,
+      "loss": 3.2222,
+      "step": 9819
+    },
+    {
+      "epoch": 0.4134736842105263,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00032613014551622423,
+      "loss": 2.8741,
+      "step": 9820
+    },
+    {
+      "epoch": 0.4135157894736842,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003260980031999921,
+      "loss": 3.0814,
+      "step": 9821
+    },
+    {
+      "epoch": 0.4135578947368421,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0003260658594973438,
+      "loss": 3.2751,
+      "step": 9822
+    },
+    {
+      "epoch": 0.4136,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00032603371440886484,
+      "loss": 3.2581,
+      "step": 9823
+    },
+    {
+      "epoch": 0.4136421052631579,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.000326001567935141,
+      "loss": 3.6344,
+      "step": 9824
+    },
+    {
+      "epoch": 0.41368421052631577,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0003259694200767579,
+      "loss": 2.9319,
+      "step": 9825
+    },
+    {
+      "epoch": 0.41372631578947366,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003259372708343012,
+      "loss": 3.1244,
+      "step": 9826
+    },
+    {
+      "epoch": 0.41376842105263156,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00032590512020835665,
+      "loss": 3.1019,
+      "step": 9827
+    },
+    {
+      "epoch": 0.41381052631578946,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00032587296819950993,
+      "loss": 3.2263,
+      "step": 9828
+    },
+    {
+      "epoch": 0.41385263157894736,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00032584081480834697,
+      "loss": 3.2399,
+      "step": 9829
+    },
+    {
+      "epoch": 0.41389473684210526,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00032580866003545345,
+      "loss": 3.2034,
+      "step": 9830
+    },
+    {
+      "epoch": 0.41393684210526316,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0003257765038814152,
+      "loss": 2.8272,
+      "step": 9831
+    },
+    {
+      "epoch": 0.41397894736842106,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003257443463468181,
+      "loss": 3.2751,
+      "step": 9832
+    },
+    {
+      "epoch": 0.41402105263157896,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00032571218743224803,
+      "loss": 3.1496,
+      "step": 9833
+    },
+    {
+      "epoch": 0.41406315789473686,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0003256800271382908,
+      "loss": 3.0077,
+      "step": 9834
+    },
+    {
+      "epoch": 0.41410526315789475,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00032564786546553245,
+      "loss": 3.0288,
+      "step": 9835
+    },
+    {
+      "epoch": 0.41414736842105265,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00032561570241455883,
+      "loss": 3.1003,
+      "step": 9836
+    },
+    {
+      "epoch": 0.41418947368421055,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00032558353798595597,
+      "loss": 3.0699,
+      "step": 9837
+    },
+    {
+      "epoch": 0.4142315789473684,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003255513721803099,
+      "loss": 2.9111,
+      "step": 9838
+    },
+    {
+      "epoch": 0.4142736842105263,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00032551920499820653,
+      "loss": 2.688,
+      "step": 9839
+    },
+    {
+      "epoch": 0.4143157894736842,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00032548703644023197,
+      "loss": 3.4043,
+      "step": 9840
+    },
+    {
+      "epoch": 0.4143578947368421,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003254548665069724,
+      "loss": 2.8924,
+      "step": 9841
+    },
+    {
+      "epoch": 0.4144,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0003254226951990137,
+      "loss": 3.2164,
+      "step": 9842
+    },
+    {
+      "epoch": 0.4144421052631579,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00032539052251694226,
+      "loss": 2.7516,
+      "step": 9843
+    },
+    {
+      "epoch": 0.4144842105263158,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00032535834846134386,
+      "loss": 3.383,
+      "step": 9844
+    },
+    {
+      "epoch": 0.4145263157894737,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003253261730328051,
+      "loss": 3.272,
+      "step": 9845
+    },
+    {
+      "epoch": 0.4145684210526316,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0003252939962319118,
+      "loss": 2.9747,
+      "step": 9846
+    },
+    {
+      "epoch": 0.4146105263157895,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00032526181805925047,
+      "loss": 3.0692,
+      "step": 9847
+    },
+    {
+      "epoch": 0.4146526315789474,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003252296385154072,
+      "loss": 2.8142,
+      "step": 9848
+    },
+    {
+      "epoch": 0.4146947368421053,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0003251974576009684,
+      "loss": 2.9298,
+      "step": 9849
+    },
+    {
+      "epoch": 0.4147368421052632,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00032516527531652016,
+      "loss": 3.1044,
+      "step": 9850
+    },
+    {
+      "epoch": 0.4147789473684211,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00032513309166264906,
+      "loss": 3.221,
+      "step": 9851
+    },
+    {
+      "epoch": 0.4148210526315789,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003251009066399411,
+      "loss": 3.048,
+      "step": 9852
+    },
+    {
+      "epoch": 0.4148631578947368,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00032506872024898304,
+      "loss": 3.2,
+      "step": 9853
+    },
+    {
+      "epoch": 0.4149052631578947,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00032503653249036095,
+      "loss": 3.5713,
+      "step": 9854
+    },
+    {
+      "epoch": 0.4149473684210526,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003250043433646615,
+      "loss": 3.3172,
+      "step": 9855
+    },
+    {
+      "epoch": 0.4149894736842105,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00032497215287247106,
+      "loss": 3.0392,
+      "step": 9856
+    },
+    {
+      "epoch": 0.4150315789473684,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00032493996101437604,
+      "loss": 3.0969,
+      "step": 9857
+    },
+    {
+      "epoch": 0.4150736842105263,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.000324907767790963,
+      "loss": 3.0283,
+      "step": 9858
+    },
+    {
+      "epoch": 0.4151157894736842,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003248755732028184,
+      "loss": 3.2518,
+      "step": 9859
+    },
+    {
+      "epoch": 0.4151578947368421,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0003248433772505289,
+      "loss": 3.4062,
+      "step": 9860
+    },
+    {
+      "epoch": 0.4152,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00032481117993468097,
+      "loss": 3.0963,
+      "step": 9861
+    },
+    {
+      "epoch": 0.4152421052631579,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0003247789812558612,
+      "loss": 3.5872,
+      "step": 9862
+    },
+    {
+      "epoch": 0.4152842105263158,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003247467812146563,
+      "loss": 3.0918,
+      "step": 9863
+    },
+    {
+      "epoch": 0.4153263157894737,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003247145798116529,
+      "loss": 3.3447,
+      "step": 9864
+    },
+    {
+      "epoch": 0.41536842105263155,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00032468237704743756,
+      "loss": 3.0961,
+      "step": 9865
+    },
+    {
+      "epoch": 0.41541052631578945,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0003246501729225971,
+      "loss": 3.3209,
+      "step": 9866
+    },
+    {
+      "epoch": 0.41545263157894735,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00032461796743771823,
+      "loss": 3.4434,
+      "step": 9867
+    },
+    {
+      "epoch": 0.41549473684210525,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0003245857605933876,
+      "loss": 3.2151,
+      "step": 9868
+    },
+    {
+      "epoch": 0.41553684210526315,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0003245535523901921,
+      "loss": 2.6901,
+      "step": 9869
+    },
+    {
+      "epoch": 0.41557894736842105,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003245213428287184,
+      "loss": 2.5731,
+      "step": 9870
+    },
+    {
+      "epoch": 0.41562105263157895,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003244891319095535,
+      "loss": 3.1945,
+      "step": 9871
+    },
+    {
+      "epoch": 0.41566315789473685,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00032445691963328406,
+      "loss": 3.3934,
+      "step": 9872
+    },
+    {
+      "epoch": 0.41570526315789474,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.000324424706000497,
+      "loss": 3.3875,
+      "step": 9873
+    },
+    {
+      "epoch": 0.41574736842105264,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00032439249101177937,
+      "loss": 2.7071,
+      "step": 9874
+    },
+    {
+      "epoch": 0.41578947368421054,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00032436027466771787,
+      "loss": 2.7539,
+      "step": 9875
+    },
+    {
+      "epoch": 0.41583157894736844,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00032432805696889957,
+      "loss": 3.2034,
+      "step": 9876
+    },
+    {
+      "epoch": 0.41587368421052634,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003242958379159114,
+      "loss": 3.351,
+      "step": 9877
+    },
+    {
+      "epoch": 0.4159157894736842,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0003242636175093403,
+      "loss": 3.3835,
+      "step": 9878
+    },
+    {
+      "epoch": 0.4159578947368421,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0003242313957497734,
+      "loss": 3.292,
+      "step": 9879
+    },
+    {
+      "epoch": 0.416,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0003241991726377976,
+      "loss": 3.1166,
+      "step": 9880
+    },
+    {
+      "epoch": 0.4160421052631579,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00032416694817400017,
+      "loss": 3.1668,
+      "step": 9881
+    },
+    {
+      "epoch": 0.4160842105263158,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00032413472235896806,
+      "loss": 3.1273,
+      "step": 9882
+    },
+    {
+      "epoch": 0.4161263157894737,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00032410249519328844,
+      "loss": 3.1253,
+      "step": 9883
+    },
+    {
+      "epoch": 0.4161684210526316,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0003240702666775484,
+      "loss": 3.2969,
+      "step": 9884
+    },
+    {
+      "epoch": 0.4162105263157895,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003240380368123351,
+      "loss": 3.4129,
+      "step": 9885
+    },
+    {
+      "epoch": 0.4162526315789474,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0003240058055982358,
+      "loss": 2.9068,
+      "step": 9886
+    },
+    {
+      "epoch": 0.4162947368421053,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0003239735730358376,
+      "loss": 2.9856,
+      "step": 9887
+    },
+    {
+      "epoch": 0.41633684210526317,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003239413391257279,
+      "loss": 3.0167,
+      "step": 9888
+    },
+    {
+      "epoch": 0.41637894736842107,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003239091038684938,
+      "loss": 3.2524,
+      "step": 9889
+    },
+    {
+      "epoch": 0.41642105263157897,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003238768672647228,
+      "loss": 3.3682,
+      "step": 9890
+    },
+    {
+      "epoch": 0.41646315789473687,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003238446293150019,
+      "loss": 3.452,
+      "step": 9891
+    },
+    {
+      "epoch": 0.4165052631578947,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0003238123900199187,
+      "loss": 2.7072,
+      "step": 9892
+    },
+    {
+      "epoch": 0.4165473684210526,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00032378014938006053,
+      "loss": 3.0442,
+      "step": 9893
+    },
+    {
+      "epoch": 0.4165894736842105,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00032374790739601465,
+      "loss": 3.1754,
+      "step": 9894
+    },
+    {
+      "epoch": 0.4166315789473684,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0003237156640683686,
+      "loss": 3.2779,
+      "step": 9895
+    },
+    {
+      "epoch": 0.4166736842105263,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003236834193977097,
+      "loss": 3.4479,
+      "step": 9896
+    },
+    {
+      "epoch": 0.4167157894736842,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00032365117338462555,
+      "loss": 3.2864,
+      "step": 9897
+    },
+    {
+      "epoch": 0.4167578947368421,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00032361892602970357,
+      "loss": 3.3573,
+      "step": 9898
+    },
+    {
+      "epoch": 0.4168,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00032358667733353117,
+      "loss": 3.2354,
+      "step": 9899
+    },
+    {
+      "epoch": 0.4168421052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00032355442729669606,
+      "loss": 3.3138,
+      "step": 9900
+    },
+    {
+      "epoch": 0.4168842105263158,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003235221759197856,
+      "loss": 3.2333,
+      "step": 9901
+    },
+    {
+      "epoch": 0.4169263157894737,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0003234899232033876,
+      "loss": 3.158,
+      "step": 9902
+    },
+    {
+      "epoch": 0.4169684210526316,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003234576691480895,
+      "loss": 3.172,
+      "step": 9903
+    },
+    {
+      "epoch": 0.4170105263157895,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00032342541375447894,
+      "loss": 3.1702,
+      "step": 9904
+    },
+    {
+      "epoch": 0.41705263157894734,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00032339315702314367,
+      "loss": 3.4364,
+      "step": 9905
+    },
+    {
+      "epoch": 0.41709473684210524,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003233608989546712,
+      "loss": 2.6916,
+      "step": 9906
+    },
+    {
+      "epoch": 0.41713684210526314,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0003233286395496494,
+      "loss": 2.665,
+      "step": 9907
+    },
+    {
+      "epoch": 0.41717894736842104,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.000323296378808666,
+      "loss": 3.1891,
+      "step": 9908
+    },
+    {
+      "epoch": 0.41722105263157894,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003232641167323087,
+      "loss": 3.3153,
+      "step": 9909
+    },
+    {
+      "epoch": 0.41726315789473684,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00032323185332116524,
+      "loss": 3.2422,
+      "step": 9910
+    },
+    {
+      "epoch": 0.41730526315789473,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003231995885758235,
+      "loss": 3.4683,
+      "step": 9911
+    },
+    {
+      "epoch": 0.41734736842105263,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0003231673224968712,
+      "loss": 3.0819,
+      "step": 9912
+    },
+    {
+      "epoch": 0.41738947368421053,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00032313505508489623,
+      "loss": 2.9248,
+      "step": 9913
+    },
+    {
+      "epoch": 0.41743157894736843,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003231027863404865,
+      "loss": 3.0531,
+      "step": 9914
+    },
+    {
+      "epoch": 0.41747368421052633,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00032307051626422994,
+      "loss": 3.2802,
+      "step": 9915
+    },
+    {
+      "epoch": 0.41751578947368423,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00032303824485671443,
+      "loss": 3.2528,
+      "step": 9916
+    },
+    {
+      "epoch": 0.41755789473684213,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00032300597211852794,
+      "loss": 2.6933,
+      "step": 9917
+    },
+    {
+      "epoch": 0.4176,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003229736980502584,
+      "loss": 3.1386,
+      "step": 9918
+    },
+    {
+      "epoch": 0.41764210526315787,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003229414226524937,
+      "loss": 3.5457,
+      "step": 9919
+    },
+    {
+      "epoch": 0.41768421052631577,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0003229091459258221,
+      "loss": 3.6531,
+      "step": 9920
+    },
+    {
+      "epoch": 0.41772631578947367,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0003228768678708315,
+      "loss": 3.0543,
+      "step": 9921
+    },
+    {
+      "epoch": 0.41776842105263157,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00032284458848811,
+      "loss": 2.7756,
+      "step": 9922
+    },
+    {
+      "epoch": 0.41781052631578947,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0003228123077782457,
+      "loss": 2.7894,
+      "step": 9923
+    },
+    {
+      "epoch": 0.41785263157894736,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00032278002574182674,
+      "loss": 3.4973,
+      "step": 9924
+    },
+    {
+      "epoch": 0.41789473684210526,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00032274774237944114,
+      "loss": 3.3233,
+      "step": 9925
+    },
+    {
+      "epoch": 0.41793684210526316,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00032271545769167715,
+      "loss": 3.3199,
+      "step": 9926
+    },
+    {
+      "epoch": 0.41797894736842106,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00032268317167912305,
+      "loss": 3.5432,
+      "step": 9927
+    },
+    {
+      "epoch": 0.41802105263157896,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00032265088434236687,
+      "loss": 3.522,
+      "step": 9928
+    },
+    {
+      "epoch": 0.41806315789473686,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00032261859568199694,
+      "loss": 3.1878,
+      "step": 9929
+    },
+    {
+      "epoch": 0.41810526315789476,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0003225863056986016,
+      "loss": 3.0819,
+      "step": 9930
+    },
+    {
+      "epoch": 0.41814736842105266,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.000322554014392769,
+      "loss": 3.25,
+      "step": 9931
+    },
+    {
+      "epoch": 0.4181894736842105,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0003225217217650875,
+      "loss": 3.5337,
+      "step": 9932
+    },
+    {
+      "epoch": 0.4182315789473684,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00032248942781614543,
+      "loss": 3.3138,
+      "step": 9933
+    },
+    {
+      "epoch": 0.4182736842105263,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003224571325465312,
+      "loss": 2.7497,
+      "step": 9934
+    },
+    {
+      "epoch": 0.4183157894736842,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00032242483595683314,
+      "loss": 3.1407,
+      "step": 9935
+    },
+    {
+      "epoch": 0.4183578947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003223925380476396,
+      "loss": 2.8838,
+      "step": 9936
+    },
+    {
+      "epoch": 0.4184,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0003223602388195391,
+      "loss": 3.2789,
+      "step": 9937
+    },
+    {
+      "epoch": 0.4184421052631579,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00032232793827312013,
+      "loss": 3.4798,
+      "step": 9938
+    },
+    {
+      "epoch": 0.4184842105263158,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0003222956364089711,
+      "loss": 3.4487,
+      "step": 9939
+    },
+    {
+      "epoch": 0.4185263157894737,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0003222633332276804,
+      "loss": 3.1943,
+      "step": 9940
+    },
+    {
+      "epoch": 0.4185684210526316,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003222310287298368,
+      "loss": 3.0457,
+      "step": 9941
+    },
+    {
+      "epoch": 0.4186105263157895,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0003221987229160287,
+      "loss": 2.9015,
+      "step": 9942
+    },
+    {
+      "epoch": 0.4186526315789474,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003221664157868447,
+      "loss": 3.4956,
+      "step": 9943
+    },
+    {
+      "epoch": 0.4186947368421053,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00032213410734287343,
+      "loss": 3.1667,
+      "step": 9944
+    },
+    {
+      "epoch": 0.4187368421052632,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00032210179758470335,
+      "loss": 2.9965,
+      "step": 9945
+    },
+    {
+      "epoch": 0.41877894736842103,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00032206948651292334,
+      "loss": 3.3478,
+      "step": 9946
+    },
+    {
+      "epoch": 0.4188210526315789,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00032203717412812187,
+      "loss": 3.1169,
+      "step": 9947
+    },
+    {
+      "epoch": 0.4188631578947368,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003220048604308878,
+      "loss": 3.5785,
+      "step": 9948
+    },
+    {
+      "epoch": 0.4189052631578947,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00032197254542180984,
+      "loss": 3.2262,
+      "step": 9949
+    },
+    {
+      "epoch": 0.4189473684210526,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00032194022910147656,
+      "loss": 3.0814,
+      "step": 9950
+    },
+    {
+      "epoch": 0.4189894736842105,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0003219079114704769,
+      "loss": 2.812,
+      "step": 9951
+    },
+    {
+      "epoch": 0.4190315789473684,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003218755925293996,
+      "loss": 2.9703,
+      "step": 9952
+    },
+    {
+      "epoch": 0.4190736842105263,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00032184327227883337,
+      "loss": 3.2367,
+      "step": 9953
+    },
+    {
+      "epoch": 0.4191157894736842,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0003218109507193672,
+      "loss": 2.7638,
+      "step": 9954
+    },
+    {
+      "epoch": 0.4191578947368421,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003217786278515897,
+      "loss": 2.9502,
+      "step": 9955
+    },
+    {
+      "epoch": 0.4192,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00032174630367609017,
+      "loss": 2.4855,
+      "step": 9956
+    },
+    {
+      "epoch": 0.4192421052631579,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003217139781934572,
+      "loss": 3.0994,
+      "step": 9957
+    },
+    {
+      "epoch": 0.4192842105263158,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00032168165140427984,
+      "loss": 3.0774,
+      "step": 9958
+    },
+    {
+      "epoch": 0.41932631578947366,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.000321649323309147,
+      "loss": 3.0791,
+      "step": 9959
+    },
+    {
+      "epoch": 0.41936842105263156,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00032161699390864765,
+      "loss": 2.8809,
+      "step": 9960
+    },
+    {
+      "epoch": 0.41941052631578946,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003215846632033708,
+      "loss": 3.0215,
+      "step": 9961
+    },
+    {
+      "epoch": 0.41945263157894735,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00032155233119390554,
+      "loss": 3.2099,
+      "step": 9962
+    },
+    {
+      "epoch": 0.41949473684210525,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0003215199978808408,
+      "loss": 3.2015,
+      "step": 9963
+    },
+    {
+      "epoch": 0.41953684210526315,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00032148766326476583,
+      "loss": 2.9152,
+      "step": 9964
+    },
+    {
+      "epoch": 0.41957894736842105,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00032145532734626956,
+      "loss": 3.0224,
+      "step": 9965
+    },
+    {
+      "epoch": 0.41962105263157895,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003214229901259412,
+      "loss": 3.5088,
+      "step": 9966
+    },
+    {
+      "epoch": 0.41966315789473685,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00032139065160436986,
+      "loss": 3.5367,
+      "step": 9967
+    },
+    {
+      "epoch": 0.41970526315789475,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003213583117821448,
+      "loss": 3.3454,
+      "step": 9968
+    },
+    {
+      "epoch": 0.41974736842105265,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003213259706598551,
+      "loss": 3.0399,
+      "step": 9969
+    },
+    {
+      "epoch": 0.41978947368421055,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003212936282380901,
+      "loss": 3.3484,
+      "step": 9970
+    },
+    {
+      "epoch": 0.41983157894736844,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00032126128451743873,
+      "loss": 3.0956,
+      "step": 9971
+    },
+    {
+      "epoch": 0.41987368421052634,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003212289394984907,
+      "loss": 3.203,
+      "step": 9972
+    },
+    {
+      "epoch": 0.4199157894736842,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00032119659318183497,
+      "loss": 3.4919,
+      "step": 9973
+    },
+    {
+      "epoch": 0.4199578947368421,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00032116424556806103,
+      "loss": 3.4529,
+      "step": 9974
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0003211318966577581,
+      "loss": 3.2855,
+      "step": 9975
+    },
+    {
+      "epoch": 0.4200421052631579,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00032109954645151564,
+      "loss": 3.2054,
+      "step": 9976
+    },
+    {
+      "epoch": 0.4200842105263158,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.000321067194949923,
+      "loss": 2.8804,
+      "step": 9977
+    },
+    {
+      "epoch": 0.4201263157894737,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0003210348421535695,
+      "loss": 2.9569,
+      "step": 9978
+    },
+    {
+      "epoch": 0.4201684210526316,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00032100248806304465,
+      "loss": 3.225,
+      "step": 9979
+    },
+    {
+      "epoch": 0.4202105263157895,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0003209701326789379,
+      "loss": 3.4321,
+      "step": 9980
+    },
+    {
+      "epoch": 0.4202526315789474,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0003209377760018387,
+      "loss": 3.2132,
+      "step": 9981
+    },
+    {
+      "epoch": 0.4202947368421053,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00032090541803233663,
+      "loss": 3.298,
+      "step": 9982
+    },
+    {
+      "epoch": 0.4203368421052632,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00032087305877102115,
+      "loss": 3.2368,
+      "step": 9983
+    },
+    {
+      "epoch": 0.4203789473684211,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00032084069821848175,
+      "loss": 3.4467,
+      "step": 9984
+    },
+    {
+      "epoch": 0.420421052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00032080833637530815,
+      "loss": 2.8654,
+      "step": 9985
+    },
+    {
+      "epoch": 0.4204631578947368,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00032077597324208987,
+      "loss": 3.1979,
+      "step": 9986
+    },
+    {
+      "epoch": 0.4205052631578947,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00032074360881941634,
+      "loss": 3.414,
+      "step": 9987
+    },
+    {
+      "epoch": 0.4205473684210526,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00032071124310787747,
+      "loss": 3.3114,
+      "step": 9988
+    },
+    {
+      "epoch": 0.4205894736842105,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00032067887610806286,
+      "loss": 3.0939,
+      "step": 9989
+    },
+    {
+      "epoch": 0.4206315789473684,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00032064650782056214,
+      "loss": 2.9733,
+      "step": 9990
+    },
+    {
+      "epoch": 0.4206736842105263,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0003206141382459651,
+      "loss": 2.7988,
+      "step": 9991
+    },
+    {
+      "epoch": 0.4207157894736842,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00032058176738486145,
+      "loss": 3.3119,
+      "step": 9992
+    },
+    {
+      "epoch": 0.4207578947368421,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0003205493952378408,
+      "loss": 3.3106,
+      "step": 9993
+    },
+    {
+      "epoch": 0.4208,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003205170218054932,
+      "loss": 2.9681,
+      "step": 9994
+    },
+    {
+      "epoch": 0.4208421052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003204846470884083,
+      "loss": 3.2842,
+      "step": 9995
+    },
+    {
+      "epoch": 0.4208842105263158,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00032045227108717586,
+      "loss": 3.0096,
+      "step": 9996
+    },
+    {
+      "epoch": 0.4209263157894737,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00032041989380238585,
+      "loss": 3.1242,
+      "step": 9997
+    },
+    {
+      "epoch": 0.4209684210526316,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003203875152346282,
+      "loss": 3.1947,
+      "step": 9998
+    },
+    {
+      "epoch": 0.42101052631578945,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003203551353844926,
+      "loss": 2.913,
+      "step": 9999
+    },
+    {
+      "epoch": 0.42105263157894735,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00032032275425256915,
+      "loss": 3.3064,
+      "step": 10000
+    },
+    {
+      "epoch": 0.42109473684210524,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003202903718394478,
+      "loss": 3.4375,
+      "step": 10001
+    },
+    {
+      "epoch": 0.42113684210526314,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00032025798814571836,
+      "loss": 3.0528,
+      "step": 10002
+    },
+    {
+      "epoch": 0.42117894736842104,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.000320225603171971,
+      "loss": 3.5987,
+      "step": 10003
+    },
+    {
+      "epoch": 0.42122105263157894,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003201932169187956,
+      "loss": 3.3923,
+      "step": 10004
+    },
+    {
+      "epoch": 0.42126315789473684,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00032016082938678234,
+      "loss": 2.849,
+      "step": 10005
+    },
+    {
+      "epoch": 0.42130526315789474,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003201284405765211,
+      "loss": 3.0765,
+      "step": 10006
+    },
+    {
+      "epoch": 0.42134736842105264,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00032009605048860214,
+      "loss": 3.1943,
+      "step": 10007
+    },
+    {
+      "epoch": 0.42138947368421054,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00032006365912361554,
+      "loss": 3.1438,
+      "step": 10008
+    },
+    {
+      "epoch": 0.42143157894736843,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003200312664821513,
+      "loss": 3.4497,
+      "step": 10009
+    },
+    {
+      "epoch": 0.42147368421052633,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00031999887256479976,
+      "loss": 3.0464,
+      "step": 10010
+    },
+    {
+      "epoch": 0.42151578947368423,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00031996647737215096,
+      "loss": 2.8291,
+      "step": 10011
+    },
+    {
+      "epoch": 0.42155789473684213,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00031993408090479514,
+      "loss": 3.3749,
+      "step": 10012
+    },
+    {
+      "epoch": 0.4216,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003199016831633225,
+      "loss": 3.1971,
+      "step": 10013
+    },
+    {
+      "epoch": 0.4216421052631579,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00031986928414832333,
+      "loss": 2.7438,
+      "step": 10014
+    },
+    {
+      "epoch": 0.4216842105263158,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00031983688386038803,
+      "loss": 3.3877,
+      "step": 10015
+    },
+    {
+      "epoch": 0.42172631578947367,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0003198044823001066,
+      "loss": 3.3923,
+      "step": 10016
+    },
+    {
+      "epoch": 0.42176842105263157,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003197720794680696,
+      "loss": 3.4403,
+      "step": 10017
+    },
+    {
+      "epoch": 0.42181052631578947,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003197396753648673,
+      "loss": 3.7153,
+      "step": 10018
+    },
+    {
+      "epoch": 0.42185263157894737,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00031970726999109005,
+      "loss": 3.5582,
+      "step": 10019
+    },
+    {
+      "epoch": 0.42189473684210527,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0003196748633473282,
+      "loss": 3.1249,
+      "step": 10020
+    },
+    {
+      "epoch": 0.42193684210526317,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00031964245543417223,
+      "loss": 3.2499,
+      "step": 10021
+    },
+    {
+      "epoch": 0.42197894736842106,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003196100462522125,
+      "loss": 3.3743,
+      "step": 10022
+    },
+    {
+      "epoch": 0.42202105263157896,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0003195776358020396,
+      "loss": 3.2409,
+      "step": 10023
+    },
+    {
+      "epoch": 0.42206315789473686,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0003195452240842439,
+      "loss": 3.7971,
+      "step": 10024
+    },
+    {
+      "epoch": 0.42210526315789476,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003195128110994159,
+      "loss": 2.9724,
+      "step": 10025
+    },
+    {
+      "epoch": 0.4221473684210526,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003194803968481462,
+      "loss": 2.9188,
+      "step": 10026
+    },
+    {
+      "epoch": 0.4221894736842105,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003194479813310253,
+      "loss": 3.1805,
+      "step": 10027
+    },
+    {
+      "epoch": 0.4222315789473684,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00031941556454864376,
+      "loss": 3.4091,
+      "step": 10028
+    },
+    {
+      "epoch": 0.4222736842105263,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003193831465015922,
+      "loss": 3.1602,
+      "step": 10029
+    },
+    {
+      "epoch": 0.4223157894736842,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00031935072719046115,
+      "loss": 3.1512,
+      "step": 10030
+    },
+    {
+      "epoch": 0.4223578947368421,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00031931830661584146,
+      "loss": 3.0206,
+      "step": 10031
+    },
+    {
+      "epoch": 0.4224,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003192858847783236,
+      "loss": 3.1342,
+      "step": 10032
+    },
+    {
+      "epoch": 0.4224421052631579,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00031925346167849836,
+      "loss": 2.843,
+      "step": 10033
+    },
+    {
+      "epoch": 0.4224842105263158,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00031922103731695634,
+      "loss": 3.0924,
+      "step": 10034
+    },
+    {
+      "epoch": 0.4225263157894737,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003191886116942884,
+      "loss": 3.3818,
+      "step": 10035
+    },
+    {
+      "epoch": 0.4225684210526316,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00031915618481108535,
+      "loss": 3.005,
+      "step": 10036
+    },
+    {
+      "epoch": 0.4226105263157895,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00031912375666793777,
+      "loss": 3.1221,
+      "step": 10037
+    },
+    {
+      "epoch": 0.4226526315789474,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00031909132726543656,
+      "loss": 2.8993,
+      "step": 10038
+    },
+    {
+      "epoch": 0.4226947368421053,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0003190588966041725,
+      "loss": 3.076,
+      "step": 10039
+    },
+    {
+      "epoch": 0.42273684210526313,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00031902646468473663,
+      "loss": 3.0939,
+      "step": 10040
+    },
+    {
+      "epoch": 0.42277894736842103,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0003189940315077195,
+      "loss": 2.8906,
+      "step": 10041
+    },
+    {
+      "epoch": 0.42282105263157893,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003189615970737123,
+      "loss": 2.5777,
+      "step": 10042
+    },
+    {
+      "epoch": 0.42286315789473683,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0003189291613833058,
+      "loss": 3.0734,
+      "step": 10043
+    },
+    {
+      "epoch": 0.42290526315789473,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003188967244370909,
+      "loss": 3.3007,
+      "step": 10044
+    },
+    {
+      "epoch": 0.4229473684210526,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003188642862356588,
+      "loss": 2.9522,
+      "step": 10045
+    },
+    {
+      "epoch": 0.4229894736842105,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00031883184677960016,
+      "loss": 2.8573,
+      "step": 10046
+    },
+    {
+      "epoch": 0.4230315789473684,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003187994060695062,
+      "loss": 2.7364,
+      "step": 10047
+    },
+    {
+      "epoch": 0.4230736842105263,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00031876696410596786,
+      "loss": 2.9881,
+      "step": 10048
+    },
+    {
+      "epoch": 0.4231157894736842,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00031873452088957625,
+      "loss": 3.117,
+      "step": 10049
+    },
+    {
+      "epoch": 0.4231578947368421,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00031870207642092246,
+      "loss": 3.3258,
+      "step": 10050
+    },
+    {
+      "epoch": 0.4232,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003186696307005976,
+      "loss": 3.1575,
+      "step": 10051
+    },
+    {
+      "epoch": 0.4232421052631579,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003186371837291927,
+      "loss": 3.0875,
+      "step": 10052
+    },
+    {
+      "epoch": 0.42328421052631576,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00031860473550729895,
+      "loss": 3.2147,
+      "step": 10053
+    },
+    {
+      "epoch": 0.42332631578947366,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00031857228603550755,
+      "loss": 3.1935,
+      "step": 10054
+    },
+    {
+      "epoch": 0.42336842105263156,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003185398353144096,
+      "loss": 3.0529,
+      "step": 10055
+    },
+    {
+      "epoch": 0.42341052631578946,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0003185073833445964,
+      "loss": 3.0873,
+      "step": 10056
+    },
+    {
+      "epoch": 0.42345263157894736,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00031847493012665925,
+      "loss": 3.2185,
+      "step": 10057
+    },
+    {
+      "epoch": 0.42349473684210526,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00031844247566118925,
+      "loss": 3.2159,
+      "step": 10058
+    },
+    {
+      "epoch": 0.42353684210526316,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00031841001994877785,
+      "loss": 2.7102,
+      "step": 10059
+    },
+    {
+      "epoch": 0.42357894736842105,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0003183775629900162,
+      "loss": 3.0398,
+      "step": 10060
+    },
+    {
+      "epoch": 0.42362105263157895,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003183451047854956,
+      "loss": 3.238,
+      "step": 10061
+    },
+    {
+      "epoch": 0.42366315789473685,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003183126453358076,
+      "loss": 3.0174,
+      "step": 10062
+    },
+    {
+      "epoch": 0.42370526315789475,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00031828018464154353,
+      "loss": 3.2731,
+      "step": 10063
+    },
+    {
+      "epoch": 0.42374736842105265,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00031824772270329463,
+      "loss": 2.6642,
+      "step": 10064
+    },
+    {
+      "epoch": 0.42378947368421055,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00031821525952165243,
+      "loss": 2.9698,
+      "step": 10065
+    },
+    {
+      "epoch": 0.42383157894736845,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0003181827950972083,
+      "loss": 2.9055,
+      "step": 10066
+    },
+    {
+      "epoch": 0.4238736842105263,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00031815032943055375,
+      "loss": 3.081,
+      "step": 10067
+    },
+    {
+      "epoch": 0.4239157894736842,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00031811786252228035,
+      "loss": 3.4177,
+      "step": 10068
+    },
+    {
+      "epoch": 0.4239578947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00031808539437297955,
+      "loss": 2.746,
+      "step": 10069
+    },
+    {
+      "epoch": 0.424,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00031805292498324276,
+      "loss": 2.7964,
+      "step": 10070
+    },
+    {
+      "epoch": 0.4240421052631579,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00031802045435366176,
+      "loss": 3.0636,
+      "step": 10071
+    },
+    {
+      "epoch": 0.4240842105263158,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003179879824848279,
+      "loss": 2.5872,
+      "step": 10072
+    },
+    {
+      "epoch": 0.4241263157894737,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00031795550937733293,
+      "loss": 3.0958,
+      "step": 10073
+    },
+    {
+      "epoch": 0.4241684210526316,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00031792303503176835,
+      "loss": 2.901,
+      "step": 10074
+    },
+    {
+      "epoch": 0.4242105263157895,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.000317890559448726,
+      "loss": 3.2005,
+      "step": 10075
+    },
+    {
+      "epoch": 0.4242526315789474,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003178580826287973,
+      "loss": 2.8681,
+      "step": 10076
+    },
+    {
+      "epoch": 0.4242947368421053,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003178256045725741,
+      "loss": 3.1837,
+      "step": 10077
+    },
+    {
+      "epoch": 0.4243368421052632,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0003177931252806482,
+      "loss": 3.1584,
+      "step": 10078
+    },
+    {
+      "epoch": 0.4243789473684211,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003177606447536111,
+      "loss": 3.3915,
+      "step": 10079
+    },
+    {
+      "epoch": 0.4244210526315789,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00031772816299205467,
+      "loss": 3.5969,
+      "step": 10080
+    },
+    {
+      "epoch": 0.4244631578947368,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00031769567999657066,
+      "loss": 3.2616,
+      "step": 10081
+    },
+    {
+      "epoch": 0.4245052631578947,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00031766319576775095,
+      "loss": 3.1007,
+      "step": 10082
+    },
+    {
+      "epoch": 0.4245473684210526,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003176307103061873,
+      "loss": 3.4711,
+      "step": 10083
+    },
+    {
+      "epoch": 0.4245894736842105,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003175982236124716,
+      "loss": 3.1098,
+      "step": 10084
+    },
+    {
+      "epoch": 0.4246315789473684,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00031756573568719564,
+      "loss": 3.2801,
+      "step": 10085
+    },
+    {
+      "epoch": 0.4246736842105263,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0003175332465309514,
+      "loss": 3.1742,
+      "step": 10086
+    },
+    {
+      "epoch": 0.4247157894736842,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00031750075614433076,
+      "loss": 3.5646,
+      "step": 10087
+    },
+    {
+      "epoch": 0.4247578947368421,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0003174682645279256,
+      "loss": 3.4282,
+      "step": 10088
+    },
+    {
+      "epoch": 0.4248,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.000317435771682328,
+      "loss": 3.2263,
+      "step": 10089
+    },
+    {
+      "epoch": 0.4248421052631579,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003174032776081298,
+      "loss": 2.8816,
+      "step": 10090
+    },
+    {
+      "epoch": 0.4248842105263158,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00031737078230592315,
+      "loss": 2.8598,
+      "step": 10091
+    },
+    {
+      "epoch": 0.4249263157894737,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0003173382857763,
+      "loss": 3.3723,
+      "step": 10092
+    },
+    {
+      "epoch": 0.4249684210526316,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0003173057880198523,
+      "loss": 3.5783,
+      "step": 10093
+    },
+    {
+      "epoch": 0.42501052631578945,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00031727328903717237,
+      "loss": 3.4208,
+      "step": 10094
+    },
+    {
+      "epoch": 0.42505263157894735,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00031724078882885206,
+      "loss": 2.985,
+      "step": 10095
+    },
+    {
+      "epoch": 0.42509473684210525,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0003172082873954836,
+      "loss": 3.4192,
+      "step": 10096
+    },
+    {
+      "epoch": 0.42513684210526315,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003171757847376591,
+      "loss": 3.0772,
+      "step": 10097
+    },
+    {
+      "epoch": 0.42517894736842105,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003171432808559708,
+      "loss": 3.1269,
+      "step": 10098
+    },
+    {
+      "epoch": 0.42522105263157894,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00031711077575101074,
+      "loss": 3.5053,
+      "step": 10099
+    },
+    {
+      "epoch": 0.42526315789473684,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0003170782694233712,
+      "loss": 2.44,
+      "step": 10100
+    },
+    {
+      "epoch": 0.42530526315789474,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0003170457618736444,
+      "loss": 3.053,
+      "step": 10101
+    },
+    {
+      "epoch": 0.42534736842105264,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00031701325310242264,
+      "loss": 2.965,
+      "step": 10102
+    },
+    {
+      "epoch": 0.42538947368421054,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00031698074311029814,
+      "loss": 3.5375,
+      "step": 10103
+    },
+    {
+      "epoch": 0.42543157894736844,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00031694823189786316,
+      "loss": 3.0056,
+      "step": 10104
+    },
+    {
+      "epoch": 0.42547368421052634,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00031691571946571007,
+      "loss": 3.4757,
+      "step": 10105
+    },
+    {
+      "epoch": 0.42551578947368424,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0003168832058144312,
+      "loss": 3.0534,
+      "step": 10106
+    },
+    {
+      "epoch": 0.4255578947368421,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00031685069094461893,
+      "loss": 3.2086,
+      "step": 10107
+    },
+    {
+      "epoch": 0.4256,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00031681817485686555,
+      "loss": 3.4178,
+      "step": 10108
+    },
+    {
+      "epoch": 0.4256421052631579,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00031678565755176356,
+      "loss": 3.7112,
+      "step": 10109
+    },
+    {
+      "epoch": 0.4256842105263158,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00031675313902990545,
+      "loss": 2.7696,
+      "step": 10110
+    },
+    {
+      "epoch": 0.4257263157894737,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00031672061929188344,
+      "loss": 3.2333,
+      "step": 10111
+    },
+    {
+      "epoch": 0.4257684210526316,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0003166880983382903,
+      "loss": 3.198,
+      "step": 10112
+    },
+    {
+      "epoch": 0.4258105263157895,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00031665557616971816,
+      "loss": 2.88,
+      "step": 10113
+    },
+    {
+      "epoch": 0.42585263157894737,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003166230527867599,
+      "loss": 2.7536,
+      "step": 10114
+    },
+    {
+      "epoch": 0.42589473684210527,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003165905281900078,
+      "loss": 3.2176,
+      "step": 10115
+    },
+    {
+      "epoch": 0.42593684210526317,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00031655800238005453,
+      "loss": 3.5902,
+      "step": 10116
+    },
+    {
+      "epoch": 0.42597894736842107,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00031652547535749274,
+      "loss": 3.0308,
+      "step": 10117
+    },
+    {
+      "epoch": 0.42602105263157897,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0003164929471229149,
+      "loss": 2.6443,
+      "step": 10118
+    },
+    {
+      "epoch": 0.42606315789473687,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00031646041767691374,
+      "loss": 2.9164,
+      "step": 10119
+    },
+    {
+      "epoch": 0.4261052631578947,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0003164278870200818,
+      "loss": 3.0683,
+      "step": 10120
+    },
+    {
+      "epoch": 0.4261473684210526,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003163953551530118,
+      "loss": 2.7613,
+      "step": 10121
+    },
+    {
+      "epoch": 0.4261894736842105,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00031636282207629646,
+      "loss": 3.0128,
+      "step": 10122
+    },
+    {
+      "epoch": 0.4262315789473684,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0003163302877905284,
+      "loss": 2.9972,
+      "step": 10123
+    },
+    {
+      "epoch": 0.4262736842105263,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00031629775229630056,
+      "loss": 3.4395,
+      "step": 10124
+    },
+    {
+      "epoch": 0.4263157894736842,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003162652155942055,
+      "loss": 3.3965,
+      "step": 10125
+    },
+    {
+      "epoch": 0.4263578947368421,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0003162326776848361,
+      "loss": 2.8352,
+      "step": 10126
+    },
+    {
+      "epoch": 0.4264,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003162001385687852,
+      "loss": 3.0641,
+      "step": 10127
+    },
+    {
+      "epoch": 0.4264421052631579,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00031616759824664543,
+      "loss": 3.4249,
+      "step": 10128
+    },
+    {
+      "epoch": 0.4264842105263158,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00031613505671900975,
+      "loss": 2.8441,
+      "step": 10129
+    },
+    {
+      "epoch": 0.4265263157894737,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.0003161025139864712,
+      "loss": 3.1708,
+      "step": 10130
+    },
+    {
+      "epoch": 0.4265684210526316,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003160699700496223,
+      "loss": 3.5297,
+      "step": 10131
+    },
+    {
+      "epoch": 0.4266105263157895,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00031603742490905637,
+      "loss": 2.8646,
+      "step": 10132
+    },
+    {
+      "epoch": 0.4266526315789474,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.000316004878565366,
+      "loss": 2.7702,
+      "step": 10133
+    },
+    {
+      "epoch": 0.42669473684210524,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003159723310191444,
+      "loss": 3.4487,
+      "step": 10134
+    },
+    {
+      "epoch": 0.42673684210526314,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003159397822709844,
+      "loss": 3.2339,
+      "step": 10135
+    },
+    {
+      "epoch": 0.42677894736842104,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00031590723232147904,
+      "loss": 3.0922,
+      "step": 10136
+    },
+    {
+      "epoch": 0.42682105263157893,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003158746811712214,
+      "loss": 3.4962,
+      "step": 10137
+    },
+    {
+      "epoch": 0.42686315789473683,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00031584212882080444,
+      "loss": 3.1091,
+      "step": 10138
+    },
+    {
+      "epoch": 0.42690526315789473,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00031580957527082117,
+      "loss": 3.1597,
+      "step": 10139
+    },
+    {
+      "epoch": 0.42694736842105263,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00031577702052186486,
+      "loss": 3.2436,
+      "step": 10140
+    },
+    {
+      "epoch": 0.42698947368421053,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00031574446457452844,
+      "loss": 2.8126,
+      "step": 10141
+    },
+    {
+      "epoch": 0.42703157894736843,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00031571190742940515,
+      "loss": 2.7193,
+      "step": 10142
+    },
+    {
+      "epoch": 0.4270736842105263,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003156793490870882,
+      "loss": 3.528,
+      "step": 10143
+    },
+    {
+      "epoch": 0.4271157894736842,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003156467895481706,
+      "loss": 3.0729,
+      "step": 10144
+    },
+    {
+      "epoch": 0.4271578947368421,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0003156142288132456,
+      "loss": 3.0067,
+      "step": 10145
+    },
+    {
+      "epoch": 0.4272,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0003155816668829066,
+      "loss": 3.3205,
+      "step": 10146
+    },
+    {
+      "epoch": 0.42724210526315787,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00031554910375774646,
+      "loss": 3.2467,
+      "step": 10147
+    },
+    {
+      "epoch": 0.42728421052631577,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0003155165394383588,
+      "loss": 3.3178,
+      "step": 10148
+    },
+    {
+      "epoch": 0.42732631578947367,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00031548397392533673,
+      "loss": 3.2583,
+      "step": 10149
+    },
+    {
+      "epoch": 0.42736842105263156,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0003154514072192736,
+      "loss": 3.5147,
+      "step": 10150
+    },
+    {
+      "epoch": 0.42741052631578946,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0003154188393207628,
+      "loss": 2.9281,
+      "step": 10151
+    },
+    {
+      "epoch": 0.42745263157894736,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00031538627023039754,
+      "loss": 2.5242,
+      "step": 10152
+    },
+    {
+      "epoch": 0.42749473684210526,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00031535369994877135,
+      "loss": 3.312,
+      "step": 10153
+    },
+    {
+      "epoch": 0.42753684210526316,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003153211284764774,
+      "loss": 3.21,
+      "step": 10154
+    },
+    {
+      "epoch": 0.42757894736842106,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003152885558141093,
+      "loss": 3.2412,
+      "step": 10155
+    },
+    {
+      "epoch": 0.42762105263157896,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0003152559819622604,
+      "loss": 2.7756,
+      "step": 10156
+    },
+    {
+      "epoch": 0.42766315789473686,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00031522340692152425,
+      "loss": 3.2571,
+      "step": 10157
+    },
+    {
+      "epoch": 0.42770526315789476,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003151908306924942,
+      "loss": 2.7833,
+      "step": 10158
+    },
+    {
+      "epoch": 0.42774736842105265,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00031515825327576387,
+      "loss": 3.3814,
+      "step": 10159
+    },
+    {
+      "epoch": 0.42778947368421055,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00031512567467192665,
+      "loss": 2.9172,
+      "step": 10160
+    },
+    {
+      "epoch": 0.4278315789473684,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00031509309488157624,
+      "loss": 3.4523,
+      "step": 10161
+    },
+    {
+      "epoch": 0.4278736842105263,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.000315060513905306,
+      "loss": 2.9854,
+      "step": 10162
+    },
+    {
+      "epoch": 0.4279157894736842,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00031502793174370975,
+      "loss": 3.2274,
+      "step": 10163
+    },
+    {
+      "epoch": 0.4279578947368421,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00031499534839738087,
+      "loss": 3.1963,
+      "step": 10164
+    },
+    {
+      "epoch": 0.428,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00031496276386691326,
+      "loss": 3.1917,
+      "step": 10165
+    },
+    {
+      "epoch": 0.4280421052631579,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003149301781529003,
+      "loss": 3.0492,
+      "step": 10166
+    },
+    {
+      "epoch": 0.4280842105263158,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0003148975912559359,
+      "loss": 3.1299,
+      "step": 10167
+    },
+    {
+      "epoch": 0.4281263157894737,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003148650031766135,
+      "loss": 3.2426,
+      "step": 10168
+    },
+    {
+      "epoch": 0.4281684210526316,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003148324139155271,
+      "loss": 3.2413,
+      "step": 10169
+    },
+    {
+      "epoch": 0.4282105263157895,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0003147998234732702,
+      "loss": 3.2582,
+      "step": 10170
+    },
+    {
+      "epoch": 0.4282526315789474,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00031476723185043667,
+      "loss": 3.3921,
+      "step": 10171
+    },
+    {
+      "epoch": 0.4282947368421053,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00031473463904762027,
+      "loss": 2.907,
+      "step": 10172
+    },
+    {
+      "epoch": 0.4283368421052632,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00031470204506541476,
+      "loss": 3.0413,
+      "step": 10173
+    },
+    {
+      "epoch": 0.428378947368421,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0003146694499044141,
+      "loss": 3.0903,
+      "step": 10174
+    },
+    {
+      "epoch": 0.4284210526315789,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.000314636853565212,
+      "loss": 3.4594,
+      "step": 10175
+    },
+    {
+      "epoch": 0.4284631578947368,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003146042560484024,
+      "loss": 3.4647,
+      "step": 10176
+    },
+    {
+      "epoch": 0.4285052631578947,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00031457165735457917,
+      "loss": 3.3174,
+      "step": 10177
+    },
+    {
+      "epoch": 0.4285473684210526,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003145390574843362,
+      "loss": 3.3223,
+      "step": 10178
+    },
+    {
+      "epoch": 0.4285894736842105,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00031450645643826743,
+      "loss": 3.0605,
+      "step": 10179
+    },
+    {
+      "epoch": 0.4286315789473684,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00031447385421696676,
+      "loss": 3.0823,
+      "step": 10180
+    },
+    {
+      "epoch": 0.4286736842105263,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003144412508210283,
+      "loss": 3.0949,
+      "step": 10181
+    },
+    {
+      "epoch": 0.4287157894736842,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003144086462510458,
+      "loss": 3.2474,
+      "step": 10182
+    },
+    {
+      "epoch": 0.4287578947368421,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00031437604050761365,
+      "loss": 2.6926,
+      "step": 10183
+    },
+    {
+      "epoch": 0.4288,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003143434335913256,
+      "loss": 3.4124,
+      "step": 10184
+    },
+    {
+      "epoch": 0.4288421052631579,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003143108255027758,
+      "loss": 3.0115,
+      "step": 10185
+    },
+    {
+      "epoch": 0.4288842105263158,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0003142782162425584,
+      "loss": 3.1265,
+      "step": 10186
+    },
+    {
+      "epoch": 0.4289263157894737,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00031424560581126725,
+      "loss": 3.0554,
+      "step": 10187
+    },
+    {
+      "epoch": 0.42896842105263155,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00031421299420949676,
+      "loss": 3.2927,
+      "step": 10188
+    },
+    {
+      "epoch": 0.42901052631578945,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00031418038143784095,
+      "loss": 3.0832,
+      "step": 10189
+    },
+    {
+      "epoch": 0.42905263157894735,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0003141477674968939,
+      "loss": 3.2295,
+      "step": 10190
+    },
+    {
+      "epoch": 0.42909473684210525,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00031411515238725,
+      "loss": 3.5265,
+      "step": 10191
+    },
+    {
+      "epoch": 0.42913684210526315,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0003140825361095034,
+      "loss": 2.8361,
+      "step": 10192
+    },
+    {
+      "epoch": 0.42917894736842105,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00031404991866424824,
+      "loss": 3.2419,
+      "step": 10193
+    },
+    {
+      "epoch": 0.42922105263157895,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003140173000520788,
+      "loss": 3.0404,
+      "step": 10194
+    },
+    {
+      "epoch": 0.42926315789473685,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00031398468027358937,
+      "loss": 2.4601,
+      "step": 10195
+    },
+    {
+      "epoch": 0.42930526315789475,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0003139520593293742,
+      "loss": 3.4189,
+      "step": 10196
+    },
+    {
+      "epoch": 0.42934736842105264,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003139194372200278,
+      "loss": 3.3392,
+      "step": 10197
+    },
+    {
+      "epoch": 0.42938947368421054,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00031388681394614414,
+      "loss": 3.2365,
+      "step": 10198
+    },
+    {
+      "epoch": 0.42943157894736844,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00031385418950831797,
+      "loss": 2.9478,
+      "step": 10199
+    },
+    {
+      "epoch": 0.42947368421052634,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00031382156390714346,
+      "loss": 2.7601,
+      "step": 10200
+    },
+    {
+      "epoch": 0.4295157894736842,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.000313788937143215,
+      "loss": 2.9426,
+      "step": 10201
+    },
+    {
+      "epoch": 0.4295578947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.000313756309217127,
+      "loss": 3.3544,
+      "step": 10202
+    },
+    {
+      "epoch": 0.4296,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003137236801294741,
+      "loss": 3.1918,
+      "step": 10203
+    },
+    {
+      "epoch": 0.4296421052631579,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003136910498808505,
+      "loss": 3.1501,
+      "step": 10204
+    },
+    {
+      "epoch": 0.4296842105263158,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003136584184718509,
+      "loss": 3.2413,
+      "step": 10205
+    },
+    {
+      "epoch": 0.4297263157894737,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003136257859030696,
+      "loss": 3.1982,
+      "step": 10206
+    },
+    {
+      "epoch": 0.4297684210526316,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00031359315217510133,
+      "loss": 2.9473,
+      "step": 10207
+    },
+    {
+      "epoch": 0.4298105263157895,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003135605172885405,
+      "loss": 2.6487,
+      "step": 10208
+    },
+    {
+      "epoch": 0.4298526315789474,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003135278812439817,
+      "loss": 3.5155,
+      "step": 10209
+    },
+    {
+      "epoch": 0.4298947368421053,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0003134952440420196,
+      "loss": 3.0402,
+      "step": 10210
+    },
+    {
+      "epoch": 0.4299368421052632,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0003134626056832488,
+      "loss": 3.1165,
+      "step": 10211
+    },
+    {
+      "epoch": 0.42997894736842107,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0003134299661682638,
+      "loss": 3.1961,
+      "step": 10212
+    },
+    {
+      "epoch": 0.43002105263157897,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0003133973254976594,
+      "loss": 2.9106,
+      "step": 10213
+    },
+    {
+      "epoch": 0.43006315789473687,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003133646836720302,
+      "loss": 2.8769,
+      "step": 10214
+    },
+    {
+      "epoch": 0.4301052631578947,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0003133320406919709,
+      "loss": 2.9815,
+      "step": 10215
+    },
+    {
+      "epoch": 0.4301473684210526,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0003132993965580762,
+      "loss": 3.2554,
+      "step": 10216
+    },
+    {
+      "epoch": 0.4301894736842105,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00031326675127094094,
+      "loss": 3.8991,
+      "step": 10217
+    },
+    {
+      "epoch": 0.4302315789473684,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003132341048311598,
+      "loss": 3.1336,
+      "step": 10218
+    },
+    {
+      "epoch": 0.4302736842105263,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00031320145723932755,
+      "loss": 3.1203,
+      "step": 10219
+    },
+    {
+      "epoch": 0.4303157894736842,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.000313168808496039,
+      "loss": 3.2426,
+      "step": 10220
+    },
+    {
+      "epoch": 0.4303578947368421,
+      "grad_norm": 0.5,
+      "learning_rate": 0.000313136158601889,
+      "loss": 3.2841,
+      "step": 10221
+    },
+    {
+      "epoch": 0.4304,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003131035075574723,
+      "loss": 2.8845,
+      "step": 10222
+    },
+    {
+      "epoch": 0.4304421052631579,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00031307085536338395,
+      "loss": 3.3125,
+      "step": 10223
+    },
+    {
+      "epoch": 0.4304842105263158,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003130382020202186,
+      "loss": 2.7131,
+      "step": 10224
+    },
+    {
+      "epoch": 0.4305263157894737,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00031300554752857135,
+      "loss": 3.3339,
+      "step": 10225
+    },
+    {
+      "epoch": 0.4305684210526316,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00031297289188903703,
+      "loss": 3.2773,
+      "step": 10226
+    },
+    {
+      "epoch": 0.4306105263157895,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00031294023510221064,
+      "loss": 3.1077,
+      "step": 10227
+    },
+    {
+      "epoch": 0.43065263157894734,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0003129075771686871,
+      "loss": 2.9743,
+      "step": 10228
+    },
+    {
+      "epoch": 0.43069473684210524,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00031287491808906136,
+      "loss": 3.3877,
+      "step": 10229
+    },
+    {
+      "epoch": 0.43073684210526314,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0003128422578639286,
+      "loss": 3.1964,
+      "step": 10230
+    },
+    {
+      "epoch": 0.43077894736842104,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00031280959649388363,
+      "loss": 3.3281,
+      "step": 10231
+    },
+    {
+      "epoch": 0.43082105263157894,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00031277693397952167,
+      "loss": 3.4426,
+      "step": 10232
+    },
+    {
+      "epoch": 0.43086315789473684,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0003127442703214377,
+      "loss": 2.839,
+      "step": 10233
+    },
+    {
+      "epoch": 0.43090526315789474,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003127116055202269,
+      "loss": 3.0567,
+      "step": 10234
+    },
+    {
+      "epoch": 0.43094736842105263,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0003126789395764843,
+      "loss": 3.3066,
+      "step": 10235
+    },
+    {
+      "epoch": 0.43098947368421053,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.000312646272490805,
+      "loss": 3.0747,
+      "step": 10236
+    },
+    {
+      "epoch": 0.43103157894736843,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0003126136042637843,
+      "loss": 3.4677,
+      "step": 10237
+    },
+    {
+      "epoch": 0.43107368421052633,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00031258093489601734,
+      "loss": 2.7609,
+      "step": 10238
+    },
+    {
+      "epoch": 0.43111578947368423,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00031254826438809925,
+      "loss": 2.8618,
+      "step": 10239
+    },
+    {
+      "epoch": 0.43115789473684213,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00031251559274062516,
+      "loss": 3.1001,
+      "step": 10240
+    },
+    {
+      "epoch": 0.4312,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0003124829199541905,
+      "loss": 3.1229,
+      "step": 10241
+    },
+    {
+      "epoch": 0.43124210526315787,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00031245024602939047,
+      "loss": 3.2554,
+      "step": 10242
+    },
+    {
+      "epoch": 0.43128421052631577,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003124175709668203,
+      "loss": 2.7909,
+      "step": 10243
+    },
+    {
+      "epoch": 0.43132631578947367,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0003123848947670754,
+      "loss": 3.6928,
+      "step": 10244
+    },
+    {
+      "epoch": 0.43136842105263157,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.000312352217430751,
+      "loss": 3.2343,
+      "step": 10245
+    },
+    {
+      "epoch": 0.43141052631578947,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0003123195389584424,
+      "loss": 3.2263,
+      "step": 10246
+    },
+    {
+      "epoch": 0.43145263157894737,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003122868593507451,
+      "loss": 3.085,
+      "step": 10247
+    },
+    {
+      "epoch": 0.43149473684210526,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00031225417860825425,
+      "loss": 3.21,
+      "step": 10248
+    },
+    {
+      "epoch": 0.43153684210526316,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003122214967315655,
+      "loss": 3.1692,
+      "step": 10249
+    },
+    {
+      "epoch": 0.43157894736842106,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003121888137212742,
+      "loss": 3.3765,
+      "step": 10250
+    },
+    {
+      "epoch": 0.43162105263157896,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003121561295779758,
+      "loss": 2.9009,
+      "step": 10251
+    },
+    {
+      "epoch": 0.43166315789473686,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00031212344430226575,
+      "loss": 3.5694,
+      "step": 10252
+    },
+    {
+      "epoch": 0.43170526315789476,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00031209075789473954,
+      "loss": 3.0949,
+      "step": 10253
+    },
+    {
+      "epoch": 0.43174736842105266,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00031205807035599266,
+      "loss": 2.7531,
+      "step": 10254
+    },
+    {
+      "epoch": 0.4317894736842105,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0003120253816866207,
+      "loss": 2.9114,
+      "step": 10255
+    },
+    {
+      "epoch": 0.4318315789473684,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0003119926918872191,
+      "loss": 3.1408,
+      "step": 10256
+    },
+    {
+      "epoch": 0.4318736842105263,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003119600009583834,
+      "loss": 3.3425,
+      "step": 10257
+    },
+    {
+      "epoch": 0.4319157894736842,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003119273089007094,
+      "loss": 2.8746,
+      "step": 10258
+    },
+    {
+      "epoch": 0.4319578947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00031189461571479263,
+      "loss": 3.3955,
+      "step": 10259
+    },
+    {
+      "epoch": 0.432,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003118619214012286,
+      "loss": 3.4417,
+      "step": 10260
+    },
+    {
+      "epoch": 0.4320421052631579,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00031182922596061313,
+      "loss": 3.2029,
+      "step": 10261
+    },
+    {
+      "epoch": 0.4320842105263158,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00031179652939354173,
+      "loss": 3.3395,
+      "step": 10262
+    },
+    {
+      "epoch": 0.4321263157894737,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003117638317006102,
+      "loss": 3.1245,
+      "step": 10263
+    },
+    {
+      "epoch": 0.4321684210526316,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0003117311328824142,
+      "loss": 2.8827,
+      "step": 10264
+    },
+    {
+      "epoch": 0.4322105263157895,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0003116984329395495,
+      "loss": 3.1062,
+      "step": 10265
+    },
+    {
+      "epoch": 0.4322526315789474,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00031166573187261186,
+      "loss": 3.5417,
+      "step": 10266
+    },
+    {
+      "epoch": 0.4322947368421053,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0003116330296821971,
+      "loss": 3.1474,
+      "step": 10267
+    },
+    {
+      "epoch": 0.43233684210526313,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003116003263689009,
+      "loss": 3.2847,
+      "step": 10268
+    },
+    {
+      "epoch": 0.43237894736842103,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003115676219333191,
+      "loss": 3.2136,
+      "step": 10269
+    },
+    {
+      "epoch": 0.43242105263157893,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003115349163760476,
+      "loss": 3.3208,
+      "step": 10270
+    },
+    {
+      "epoch": 0.4324631578947368,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00031150220969768225,
+      "loss": 3.2305,
+      "step": 10271
+    },
+    {
+      "epoch": 0.4325052631578947,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003114695018988189,
+      "loss": 3.2357,
+      "step": 10272
+    },
+    {
+      "epoch": 0.4325473684210526,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003114367929800533,
+      "loss": 3.1619,
+      "step": 10273
+    },
+    {
+      "epoch": 0.4325894736842105,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00031140408294198167,
+      "loss": 2.7291,
+      "step": 10274
+    },
+    {
+      "epoch": 0.4326315789473684,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0003113713717851998,
+      "loss": 3.0375,
+      "step": 10275
+    },
+    {
+      "epoch": 0.4326736842105263,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0003113386595103036,
+      "loss": 2.9779,
+      "step": 10276
+    },
+    {
+      "epoch": 0.4327157894736842,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003113059461178891,
+      "loss": 2.8716,
+      "step": 10277
+    },
+    {
+      "epoch": 0.4327578947368421,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0003112732316085523,
+      "loss": 2.5239,
+      "step": 10278
+    },
+    {
+      "epoch": 0.4328,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00031124051598288924,
+      "loss": 3.071,
+      "step": 10279
+    },
+    {
+      "epoch": 0.4328421052631579,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003112077992414959,
+      "loss": 3.494,
+      "step": 10280
+    },
+    {
+      "epoch": 0.4328842105263158,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003111750813849684,
+      "loss": 3.0461,
+      "step": 10281
+    },
+    {
+      "epoch": 0.43292631578947366,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0003111423624139027,
+      "loss": 2.9781,
+      "step": 10282
+    },
+    {
+      "epoch": 0.43296842105263156,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00031110964232889505,
+      "loss": 2.8699,
+      "step": 10283
+    },
+    {
+      "epoch": 0.43301052631578946,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00031107692113054155,
+      "loss": 3.0424,
+      "step": 10284
+    },
+    {
+      "epoch": 0.43305263157894736,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003110441988194383,
+      "loss": 3.3475,
+      "step": 10285
+    },
+    {
+      "epoch": 0.43309473684210525,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00031101147539618144,
+      "loss": 3.6058,
+      "step": 10286
+    },
+    {
+      "epoch": 0.43313684210526315,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003109787508613673,
+      "loss": 3.3069,
+      "step": 10287
+    },
+    {
+      "epoch": 0.43317894736842105,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00031094602521559186,
+      "loss": 3.0923,
+      "step": 10288
+    },
+    {
+      "epoch": 0.43322105263157895,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00031091329845945144,
+      "loss": 3.4624,
+      "step": 10289
+    },
+    {
+      "epoch": 0.43326315789473685,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003108805705935423,
+      "loss": 3.3341,
+      "step": 10290
+    },
+    {
+      "epoch": 0.43330526315789475,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003108478416184607,
+      "loss": 3.1184,
+      "step": 10291
+    },
+    {
+      "epoch": 0.43334736842105265,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0003108151115348029,
+      "loss": 3.4846,
+      "step": 10292
+    },
+    {
+      "epoch": 0.43338947368421055,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003107823803431653,
+      "loss": 2.9394,
+      "step": 10293
+    },
+    {
+      "epoch": 0.43343157894736845,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00031074964804414415,
+      "loss": 3.4557,
+      "step": 10294
+    },
+    {
+      "epoch": 0.4334736842105263,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003107169146383357,
+      "loss": 3.0079,
+      "step": 10295
+    },
+    {
+      "epoch": 0.4335157894736842,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0003106841801263364,
+      "loss": 3.6496,
+      "step": 10296
+    },
+    {
+      "epoch": 0.4335578947368421,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0003106514445087427,
+      "loss": 3.2014,
+      "step": 10297
+    },
+    {
+      "epoch": 0.4336,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.000310618707786151,
+      "loss": 3.4901,
+      "step": 10298
+    },
+    {
+      "epoch": 0.4336421052631579,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0003105859699591575,
+      "loss": 3.03,
+      "step": 10299
+    },
+    {
+      "epoch": 0.4336842105263158,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00031055323102835897,
+      "loss": 3.119,
+      "step": 10300
+    },
+    {
+      "epoch": 0.4337263157894737,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003105204909943516,
+      "loss": 3.2656,
+      "step": 10301
+    },
+    {
+      "epoch": 0.4337684210526316,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0003104877498577321,
+      "loss": 3.2354,
+      "step": 10302
+    },
+    {
+      "epoch": 0.4338105263157895,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00031045500761909677,
+      "loss": 3.2186,
+      "step": 10303
+    },
+    {
+      "epoch": 0.4338526315789474,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0003104222642790423,
+      "loss": 3.1294,
+      "step": 10304
+    },
+    {
+      "epoch": 0.4338947368421053,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00031038951983816513,
+      "loss": 3.2637,
+      "step": 10305
+    },
+    {
+      "epoch": 0.4339368421052632,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00031035677429706193,
+      "loss": 3.295,
+      "step": 10306
+    },
+    {
+      "epoch": 0.4339789473684211,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00031032402765632916,
+      "loss": 3.2949,
+      "step": 10307
+    },
+    {
+      "epoch": 0.434021052631579,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0003102912799165635,
+      "loss": 3.3416,
+      "step": 10308
+    },
+    {
+      "epoch": 0.4340631578947368,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00031025853107836157,
+      "loss": 3.2374,
+      "step": 10309
+    },
+    {
+      "epoch": 0.4341052631578947,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003102257811423199,
+      "loss": 3.2134,
+      "step": 10310
+    },
+    {
+      "epoch": 0.4341473684210526,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0003101930301090354,
+      "loss": 2.6182,
+      "step": 10311
+    },
+    {
+      "epoch": 0.4341894736842105,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00031016027797910455,
+      "loss": 3.136,
+      "step": 10312
+    },
+    {
+      "epoch": 0.4342315789473684,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003101275247531242,
+      "loss": 3.8209,
+      "step": 10313
+    },
+    {
+      "epoch": 0.4342736842105263,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00031009477043169095,
+      "loss": 2.9243,
+      "step": 10314
+    },
+    {
+      "epoch": 0.4343157894736842,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00031006201501540147,
+      "loss": 3.4454,
+      "step": 10315
+    },
+    {
+      "epoch": 0.4343578947368421,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00031002925850485275,
+      "loss": 3.1152,
+      "step": 10316
+    },
+    {
+      "epoch": 0.4344,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003099965009006415,
+      "loss": 3.2999,
+      "step": 10317
+    },
+    {
+      "epoch": 0.4344421052631579,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00030996374220336443,
+      "loss": 3.3634,
+      "step": 10318
+    },
+    {
+      "epoch": 0.4344842105263158,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0003099309824136185,
+      "loss": 3.1904,
+      "step": 10319
+    },
+    {
+      "epoch": 0.4345263157894737,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0003098982215320005,
+      "loss": 3.1303,
+      "step": 10320
+    },
+    {
+      "epoch": 0.4345684210526316,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00030986545955910723,
+      "loss": 3.2103,
+      "step": 10321
+    },
+    {
+      "epoch": 0.43461052631578945,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00030983269649553563,
+      "loss": 3.0465,
+      "step": 10322
+    },
+    {
+      "epoch": 0.43465263157894735,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0003097999323418826,
+      "loss": 2.7737,
+      "step": 10323
+    },
+    {
+      "epoch": 0.43469473684210524,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00030976716709874496,
+      "loss": 3.4548,
+      "step": 10324
+    },
+    {
+      "epoch": 0.43473684210526314,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00030973440076671987,
+      "loss": 2.8545,
+      "step": 10325
+    },
+    {
+      "epoch": 0.43477894736842104,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00030970163334640425,
+      "loss": 2.8709,
+      "step": 10326
+    },
+    {
+      "epoch": 0.43482105263157894,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00030966886483839494,
+      "loss": 3.0409,
+      "step": 10327
+    },
+    {
+      "epoch": 0.43486315789473684,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0003096360952432891,
+      "loss": 3.4493,
+      "step": 10328
+    },
+    {
+      "epoch": 0.43490526315789474,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00030960332456168356,
+      "loss": 3.5565,
+      "step": 10329
+    },
+    {
+      "epoch": 0.43494736842105264,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00030957055279417546,
+      "loss": 3.0773,
+      "step": 10330
+    },
+    {
+      "epoch": 0.43498947368421054,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00030953777994136196,
+      "loss": 3.2632,
+      "step": 10331
+    },
+    {
+      "epoch": 0.43503157894736844,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00030950500600384007,
+      "loss": 3.2477,
+      "step": 10332
+    },
+    {
+      "epoch": 0.43507368421052633,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00030947223098220676,
+      "loss": 3.0646,
+      "step": 10333
+    },
+    {
+      "epoch": 0.43511578947368423,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0003094394548770594,
+      "loss": 2.9225,
+      "step": 10334
+    },
+    {
+      "epoch": 0.43515789473684213,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00030940667768899506,
+      "loss": 3.1923,
+      "step": 10335
+    },
+    {
+      "epoch": 0.4352,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00030937389941861076,
+      "loss": 3.1587,
+      "step": 10336
+    },
+    {
+      "epoch": 0.4352421052631579,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00030934112006650375,
+      "loss": 2.8878,
+      "step": 10337
+    },
+    {
+      "epoch": 0.4352842105263158,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003093083396332713,
+      "loss": 3.4308,
+      "step": 10338
+    },
+    {
+      "epoch": 0.4353263157894737,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00030927555811951065,
+      "loss": 2.5809,
+      "step": 10339
+    },
+    {
+      "epoch": 0.43536842105263157,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0003092427755258189,
+      "loss": 3.1825,
+      "step": 10340
+    },
+    {
+      "epoch": 0.43541052631578947,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003092099918527934,
+      "loss": 3.2957,
+      "step": 10341
+    },
+    {
+      "epoch": 0.43545263157894737,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00030917720710103136,
+      "loss": 3.017,
+      "step": 10342
+    },
+    {
+      "epoch": 0.43549473684210527,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003091444212711302,
+      "loss": 3.2214,
+      "step": 10343
+    },
+    {
+      "epoch": 0.43553684210526317,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00030911163436368717,
+      "loss": 3.4679,
+      "step": 10344
+    },
+    {
+      "epoch": 0.43557894736842107,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0003090788463792996,
+      "loss": 2.9486,
+      "step": 10345
+    },
+    {
+      "epoch": 0.43562105263157896,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00030904605731856493,
+      "loss": 3.0537,
+      "step": 10346
+    },
+    {
+      "epoch": 0.43566315789473686,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00030901326718208046,
+      "loss": 3.5625,
+      "step": 10347
+    },
+    {
+      "epoch": 0.43570526315789476,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00030898047597044355,
+      "loss": 3.3916,
+      "step": 10348
+    },
+    {
+      "epoch": 0.4357473684210526,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00030894768368425164,
+      "loss": 3.4189,
+      "step": 10349
+    },
+    {
+      "epoch": 0.4357894736842105,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003089148903241022,
+      "loss": 3.3901,
+      "step": 10350
+    },
+    {
+      "epoch": 0.4358315789473684,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003088820958905927,
+      "loss": 3.1573,
+      "step": 10351
+    },
+    {
+      "epoch": 0.4358736842105263,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00030884930038432064,
+      "loss": 2.9614,
+      "step": 10352
+    },
+    {
+      "epoch": 0.4359157894736842,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00030881650380588347,
+      "loss": 2.9518,
+      "step": 10353
+    },
+    {
+      "epoch": 0.4359578947368421,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00030878370615587865,
+      "loss": 3.0883,
+      "step": 10354
+    },
+    {
+      "epoch": 0.436,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0003087509074349038,
+      "loss": 3.3926,
+      "step": 10355
+    },
+    {
+      "epoch": 0.4360421052631579,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00030871810764355643,
+      "loss": 2.8601,
+      "step": 10356
+    },
+    {
+      "epoch": 0.4360842105263158,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00030868530678243415,
+      "loss": 3.0935,
+      "step": 10357
+    },
+    {
+      "epoch": 0.4361263157894737,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00030865250485213446,
+      "loss": 3.3136,
+      "step": 10358
+    },
+    {
+      "epoch": 0.4361684210526316,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00030861970185325514,
+      "loss": 3.2019,
+      "step": 10359
+    },
+    {
+      "epoch": 0.4362105263157895,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0003085868977863937,
+      "loss": 2.9428,
+      "step": 10360
+    },
+    {
+      "epoch": 0.4362526315789474,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00030855409265214784,
+      "loss": 3.0771,
+      "step": 10361
+    },
+    {
+      "epoch": 0.4362947368421053,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00030852128645111524,
+      "loss": 3.284,
+      "step": 10362
+    },
+    {
+      "epoch": 0.43633684210526313,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003084884791838935,
+      "loss": 3.0459,
+      "step": 10363
+    },
+    {
+      "epoch": 0.43637894736842103,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0003084556708510804,
+      "loss": 3.1154,
+      "step": 10364
+    },
+    {
+      "epoch": 0.43642105263157893,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00030842286145327366,
+      "loss": 3.1394,
+      "step": 10365
+    },
+    {
+      "epoch": 0.43646315789473683,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.000308390050991071,
+      "loss": 3.3277,
+      "step": 10366
+    },
+    {
+      "epoch": 0.43650526315789473,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0003083572394650702,
+      "loss": 3.0497,
+      "step": 10367
+    },
+    {
+      "epoch": 0.43654736842105263,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00030832442687586914,
+      "loss": 2.8033,
+      "step": 10368
+    },
+    {
+      "epoch": 0.4365894736842105,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0003082916132240656,
+      "loss": 3.2446,
+      "step": 10369
+    },
+    {
+      "epoch": 0.4366315789473684,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00030825879851025716,
+      "loss": 3.4533,
+      "step": 10370
+    },
+    {
+      "epoch": 0.4366736842105263,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.000308225982735042,
+      "loss": 3.0644,
+      "step": 10371
+    },
+    {
+      "epoch": 0.4367157894736842,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00030819316589901784,
+      "loss": 2.8828,
+      "step": 10372
+    },
+    {
+      "epoch": 0.4367578947368421,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003081603480027826,
+      "loss": 3.1742,
+      "step": 10373
+    },
+    {
+      "epoch": 0.4368,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003081275290469341,
+      "loss": 3.4915,
+      "step": 10374
+    },
+    {
+      "epoch": 0.4368421052631579,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00030809470903207033,
+      "loss": 3.4146,
+      "step": 10375
+    },
+    {
+      "epoch": 0.43688421052631576,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003080618879587892,
+      "loss": 2.8529,
+      "step": 10376
+    },
+    {
+      "epoch": 0.43692631578947366,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0003080290658276887,
+      "loss": 3.1891,
+      "step": 10377
+    },
+    {
+      "epoch": 0.43696842105263156,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0003079962426393668,
+      "loss": 3.2417,
+      "step": 10378
+    },
+    {
+      "epoch": 0.43701052631578946,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00030796341839442156,
+      "loss": 3.0707,
+      "step": 10379
+    },
+    {
+      "epoch": 0.43705263157894736,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003079305930934509,
+      "loss": 2.8132,
+      "step": 10380
+    },
+    {
+      "epoch": 0.43709473684210526,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003078977667370529,
+      "loss": 3.6715,
+      "step": 10381
+    },
+    {
+      "epoch": 0.43713684210526316,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0003078649393258256,
+      "loss": 2.6121,
+      "step": 10382
+    },
+    {
+      "epoch": 0.43717894736842106,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003078321108603671,
+      "loss": 2.9402,
+      "step": 10383
+    },
+    {
+      "epoch": 0.43722105263157895,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00030779928134127546,
+      "loss": 2.992,
+      "step": 10384
+    },
+    {
+      "epoch": 0.43726315789473685,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003077664507691489,
+      "loss": 3.2131,
+      "step": 10385
+    },
+    {
+      "epoch": 0.43730526315789475,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003077336191445854,
+      "loss": 3.1198,
+      "step": 10386
+    },
+    {
+      "epoch": 0.43734736842105265,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003077007864681833,
+      "loss": 3.1842,
+      "step": 10387
+    },
+    {
+      "epoch": 0.43738947368421055,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0003076679527405406,
+      "loss": 3.1635,
+      "step": 10388
+    },
+    {
+      "epoch": 0.4374315789473684,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003076351179622556,
+      "loss": 2.8915,
+      "step": 10389
+    },
+    {
+      "epoch": 0.4374736842105263,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003076022821339264,
+      "loss": 3.1661,
+      "step": 10390
+    },
+    {
+      "epoch": 0.4375157894736842,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00030756944525615133,
+      "loss": 2.6321,
+      "step": 10391
+    },
+    {
+      "epoch": 0.4375578947368421,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00030753660732952854,
+      "loss": 3.4021,
+      "step": 10392
+    },
+    {
+      "epoch": 0.4376,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003075037683546564,
+      "loss": 3.1174,
+      "step": 10393
+    },
+    {
+      "epoch": 0.4376421052631579,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00030747092833213323,
+      "loss": 3.4748,
+      "step": 10394
+    },
+    {
+      "epoch": 0.4376842105263158,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003074380872625573,
+      "loss": 3.3705,
+      "step": 10395
+    },
+    {
+      "epoch": 0.4377263157894737,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00030740524514652685,
+      "loss": 3.1499,
+      "step": 10396
+    },
+    {
+      "epoch": 0.4377684210526316,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003073724019846402,
+      "loss": 3.216,
+      "step": 10397
+    },
+    {
+      "epoch": 0.4378105263157895,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00030733955777749596,
+      "loss": 3.2239,
+      "step": 10398
+    },
+    {
+      "epoch": 0.4378526315789474,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00030730671252569223,
+      "loss": 3.4278,
+      "step": 10399
+    },
+    {
+      "epoch": 0.4378947368421053,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00030727386622982756,
+      "loss": 3.167,
+      "step": 10400
+    },
+    {
+      "epoch": 0.4379368421052632,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00030724101889050036,
+      "loss": 2.7846,
+      "step": 10401
+    },
+    {
+      "epoch": 0.4379789473684211,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00030720817050830903,
+      "loss": 3.8017,
+      "step": 10402
+    },
+    {
+      "epoch": 0.4380210526315789,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0003071753210838521,
+      "loss": 2.8716,
+      "step": 10403
+    },
+    {
+      "epoch": 0.4380631578947368,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0003071424706177279,
+      "loss": 3.3399,
+      "step": 10404
+    },
+    {
+      "epoch": 0.4381052631578947,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003071096191105351,
+      "loss": 3.0516,
+      "step": 10405
+    },
+    {
+      "epoch": 0.4381473684210526,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00030707676656287214,
+      "loss": 3.1168,
+      "step": 10406
+    },
+    {
+      "epoch": 0.4381894736842105,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00030704391297533754,
+      "loss": 3.2377,
+      "step": 10407
+    },
+    {
+      "epoch": 0.4382315789473684,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00030701105834852983,
+      "loss": 3.2906,
+      "step": 10408
+    },
+    {
+      "epoch": 0.4382736842105263,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00030697820268304766,
+      "loss": 3.0429,
+      "step": 10409
+    },
+    {
+      "epoch": 0.4383157894736842,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00030694534597948956,
+      "loss": 3.0752,
+      "step": 10410
+    },
+    {
+      "epoch": 0.4383578947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003069124882384542,
+      "loss": 2.8575,
+      "step": 10411
+    },
+    {
+      "epoch": 0.4384,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00030687962946054014,
+      "loss": 2.9696,
+      "step": 10412
+    },
+    {
+      "epoch": 0.4384421052631579,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003068467696463461,
+      "loss": 3.0476,
+      "step": 10413
+    },
+    {
+      "epoch": 0.4384842105263158,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003068139087964707,
+      "loss": 3.0782,
+      "step": 10414
+    },
+    {
+      "epoch": 0.4385263157894737,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00030678104691151265,
+      "loss": 3.5278,
+      "step": 10415
+    },
+    {
+      "epoch": 0.43856842105263155,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00030674818399207063,
+      "loss": 3.4843,
+      "step": 10416
+    },
+    {
+      "epoch": 0.43861052631578945,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003067153200387433,
+      "loss": 3.2554,
+      "step": 10417
+    },
+    {
+      "epoch": 0.43865263157894735,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00030668245505212957,
+      "loss": 3.2461,
+      "step": 10418
+    },
+    {
+      "epoch": 0.43869473684210525,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00030664958903282803,
+      "loss": 3.4586,
+      "step": 10419
+    },
+    {
+      "epoch": 0.43873684210526315,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00030661672198143764,
+      "loss": 3.0523,
+      "step": 10420
+    },
+    {
+      "epoch": 0.43877894736842105,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00030658385389855704,
+      "loss": 2.9867,
+      "step": 10421
+    },
+    {
+      "epoch": 0.43882105263157895,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00030655098478478513,
+      "loss": 2.6462,
+      "step": 10422
+    },
+    {
+      "epoch": 0.43886315789473684,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003065181146407207,
+      "loss": 3.4885,
+      "step": 10423
+    },
+    {
+      "epoch": 0.43890526315789474,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003064852434669626,
+      "loss": 3.1478,
+      "step": 10424
+    },
+    {
+      "epoch": 0.43894736842105264,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0003064523712641097,
+      "loss": 3.1935,
+      "step": 10425
+    },
+    {
+      "epoch": 0.43898947368421054,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00030641949803276097,
+      "loss": 3.5379,
+      "step": 10426
+    },
+    {
+      "epoch": 0.43903157894736844,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00030638662377351533,
+      "loss": 3.1917,
+      "step": 10427
+    },
+    {
+      "epoch": 0.43907368421052634,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003063537484869716,
+      "loss": 3.2609,
+      "step": 10428
+    },
+    {
+      "epoch": 0.43911578947368424,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00030632087217372886,
+      "loss": 2.9715,
+      "step": 10429
+    },
+    {
+      "epoch": 0.4391578947368421,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00030628799483438596,
+      "loss": 3.0849,
+      "step": 10430
+    },
+    {
+      "epoch": 0.4392,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00030625511646954187,
+      "loss": 3.5909,
+      "step": 10431
+    },
+    {
+      "epoch": 0.4392421052631579,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00030622223707979567,
+      "loss": 2.7738,
+      "step": 10432
+    },
+    {
+      "epoch": 0.4392842105263158,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00030618935666574637,
+      "loss": 3.2497,
+      "step": 10433
+    },
+    {
+      "epoch": 0.4393263157894737,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00030615647522799305,
+      "loss": 3.0866,
+      "step": 10434
+    },
+    {
+      "epoch": 0.4393684210526316,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00030612359276713474,
+      "loss": 3.3239,
+      "step": 10435
+    },
+    {
+      "epoch": 0.4394105263157895,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003060907092837705,
+      "loss": 2.9103,
+      "step": 10436
+    },
+    {
+      "epoch": 0.4394526315789474,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00030605782477849945,
+      "loss": 2.983,
+      "step": 10437
+    },
+    {
+      "epoch": 0.43949473684210527,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00030602493925192066,
+      "loss": 3.053,
+      "step": 10438
+    },
+    {
+      "epoch": 0.43953684210526317,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003059920527046333,
+      "loss": 2.7399,
+      "step": 10439
+    },
+    {
+      "epoch": 0.43957894736842107,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003059591651372366,
+      "loss": 3.0214,
+      "step": 10440
+    },
+    {
+      "epoch": 0.43962105263157897,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00030592627655032955,
+      "loss": 2.9279,
+      "step": 10441
+    },
+    {
+      "epoch": 0.43966315789473687,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00030589338694451154,
+      "loss": 3.4296,
+      "step": 10442
+    },
+    {
+      "epoch": 0.4397052631578947,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003058604963203817,
+      "loss": 2.9419,
+      "step": 10443
+    },
+    {
+      "epoch": 0.4397473684210526,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00030582760467853917,
+      "loss": 3.1846,
+      "step": 10444
+    },
+    {
+      "epoch": 0.4397894736842105,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0003057947120195833,
+      "loss": 3.07,
+      "step": 10445
+    },
+    {
+      "epoch": 0.4398315789473684,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0003057618183441134,
+      "loss": 3.4921,
+      "step": 10446
+    },
+    {
+      "epoch": 0.4398736842105263,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00030572892365272864,
+      "loss": 3.2273,
+      "step": 10447
+    },
+    {
+      "epoch": 0.4399157894736842,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0003056960279460284,
+      "loss": 3.3551,
+      "step": 10448
+    },
+    {
+      "epoch": 0.4399578947368421,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0003056631312246119,
+      "loss": 3.1826,
+      "step": 10449
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003056302334890786,
+      "loss": 2.9921,
+      "step": 10450
+    },
+    {
+      "epoch": 0.4400421052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00030559733474002775,
+      "loss": 3.1729,
+      "step": 10451
+    },
+    {
+      "epoch": 0.4400842105263158,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00030556443497805887,
+      "loss": 2.9209,
+      "step": 10452
+    },
+    {
+      "epoch": 0.4401263157894737,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0003055315342037713,
+      "loss": 3.624,
+      "step": 10453
+    },
+    {
+      "epoch": 0.4401684210526316,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00030549863241776434,
+      "loss": 3.2372,
+      "step": 10454
+    },
+    {
+      "epoch": 0.4402105263157895,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00030546572962063755,
+      "loss": 3.3846,
+      "step": 10455
+    },
+    {
+      "epoch": 0.4402526315789474,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003054328258129903,
+      "loss": 3.2372,
+      "step": 10456
+    },
+    {
+      "epoch": 0.44029473684210524,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003053999209954221,
+      "loss": 3.3276,
+      "step": 10457
+    },
+    {
+      "epoch": 0.44033684210526314,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003053670151685324,
+      "loss": 2.5826,
+      "step": 10458
+    },
+    {
+      "epoch": 0.44037894736842104,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0003053341083329208,
+      "loss": 2.9943,
+      "step": 10459
+    },
+    {
+      "epoch": 0.44042105263157894,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00030530120048918673,
+      "loss": 3.1917,
+      "step": 10460
+    },
+    {
+      "epoch": 0.44046315789473683,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0003052682916379298,
+      "loss": 3.0646,
+      "step": 10461
+    },
+    {
+      "epoch": 0.44050526315789473,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003052353817797495,
+      "loss": 3.172,
+      "step": 10462
+    },
+    {
+      "epoch": 0.44054736842105263,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0003052024709152455,
+      "loss": 3.1724,
+      "step": 10463
+    },
+    {
+      "epoch": 0.44058947368421053,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00030516955904501725,
+      "loss": 2.8461,
+      "step": 10464
+    },
+    {
+      "epoch": 0.44063157894736843,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003051366461696645,
+      "loss": 3.3684,
+      "step": 10465
+    },
+    {
+      "epoch": 0.44067368421052633,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00030510373228978684,
+      "loss": 3.4937,
+      "step": 10466
+    },
+    {
+      "epoch": 0.4407157894736842,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00030507081740598383,
+      "loss": 2.8291,
+      "step": 10467
+    },
+    {
+      "epoch": 0.4407578947368421,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0003050379015188554,
+      "loss": 3.2368,
+      "step": 10468
+    },
+    {
+      "epoch": 0.4408,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.000305004984629001,
+      "loss": 3.3286,
+      "step": 10469
+    },
+    {
+      "epoch": 0.44084210526315787,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0003049720667370204,
+      "loss": 3.53,
+      "step": 10470
+    },
+    {
+      "epoch": 0.44088421052631577,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0003049391478435133,
+      "loss": 3.0017,
+      "step": 10471
+    },
+    {
+      "epoch": 0.44092631578947367,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003049062279490795,
+      "loss": 2.8856,
+      "step": 10472
+    },
+    {
+      "epoch": 0.44096842105263157,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0003048733070543188,
+      "loss": 3.7096,
+      "step": 10473
+    },
+    {
+      "epoch": 0.44101052631578946,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00030484038515983086,
+      "loss": 3.3101,
+      "step": 10474
+    },
+    {
+      "epoch": 0.44105263157894736,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0003048074622662155,
+      "loss": 3.4182,
+      "step": 10475
+    },
+    {
+      "epoch": 0.44109473684210526,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00030477453837407265,
+      "loss": 2.9342,
+      "step": 10476
+    },
+    {
+      "epoch": 0.44113684210526316,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.000304741613484002,
+      "loss": 3.3998,
+      "step": 10477
+    },
+    {
+      "epoch": 0.44117894736842106,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003047086875966035,
+      "loss": 3.5417,
+      "step": 10478
+    },
+    {
+      "epoch": 0.44122105263157896,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.000304675760712477,
+      "loss": 3.2562,
+      "step": 10479
+    },
+    {
+      "epoch": 0.44126315789473686,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00030464283283222244,
+      "loss": 3.4119,
+      "step": 10480
+    },
+    {
+      "epoch": 0.44130526315789476,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003046099039564396,
+      "loss": 3.4781,
+      "step": 10481
+    },
+    {
+      "epoch": 0.44134736842105265,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003045769740857285,
+      "loss": 2.7657,
+      "step": 10482
+    },
+    {
+      "epoch": 0.44138947368421055,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00030454404322068903,
+      "loss": 3.1645,
+      "step": 10483
+    },
+    {
+      "epoch": 0.4414315789473684,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0003045111113619212,
+      "loss": 3.3728,
+      "step": 10484
+    },
+    {
+      "epoch": 0.4414736842105263,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003044781785100249,
+      "loss": 2.675,
+      "step": 10485
+    },
+    {
+      "epoch": 0.4415157894736842,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00030444524466560026,
+      "loss": 3.2005,
+      "step": 10486
+    },
+    {
+      "epoch": 0.4415578947368421,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00030441230982924727,
+      "loss": 3.0171,
+      "step": 10487
+    },
+    {
+      "epoch": 0.4416,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003043793740015659,
+      "loss": 3.5587,
+      "step": 10488
+    },
+    {
+      "epoch": 0.4416421052631579,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003043464371831562,
+      "loss": 2.7387,
+      "step": 10489
+    },
+    {
+      "epoch": 0.4416842105263158,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00030431349937461833,
+      "loss": 3.2309,
+      "step": 10490
+    },
+    {
+      "epoch": 0.4417263157894737,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00030428056057655217,
+      "loss": 3.0492,
+      "step": 10491
+    },
+    {
+      "epoch": 0.4417684210526316,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003042476207895581,
+      "loss": 3.1302,
+      "step": 10492
+    },
+    {
+      "epoch": 0.4418105263157895,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0003042146800142361,
+      "loss": 3.1055,
+      "step": 10493
+    },
+    {
+      "epoch": 0.4418526315789474,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00030418173825118636,
+      "loss": 3.2716,
+      "step": 10494
+    },
+    {
+      "epoch": 0.4418947368421053,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00030414879550100894,
+      "loss": 3.227,
+      "step": 10495
+    },
+    {
+      "epoch": 0.4419368421052632,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003041158517643041,
+      "loss": 2.7149,
+      "step": 10496
+    },
+    {
+      "epoch": 0.441978947368421,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003040829070416721,
+      "loss": 3.0994,
+      "step": 10497
+    },
+    {
+      "epoch": 0.4420210526315789,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00030404996133371297,
+      "loss": 2.8689,
+      "step": 10498
+    },
+    {
+      "epoch": 0.4420631578947368,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003040170146410271,
+      "loss": 3.4841,
+      "step": 10499
+    },
+    {
+      "epoch": 0.4421052631578947,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003039840669642147,
+      "loss": 3.2117,
+      "step": 10500
+    },
+    {
+      "epoch": 0.4421473684210526,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00030395111830387603,
+      "loss": 3.2922,
+      "step": 10501
+    },
+    {
+      "epoch": 0.4421894736842105,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00030391816866061133,
+      "loss": 3.6952,
+      "step": 10502
+    },
+    {
+      "epoch": 0.4422315789473684,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.000303885218035021,
+      "loss": 2.9978,
+      "step": 10503
+    },
+    {
+      "epoch": 0.4422736842105263,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003038522664277054,
+      "loss": 2.8767,
+      "step": 10504
+    },
+    {
+      "epoch": 0.4423157894736842,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00030381931383926457,
+      "loss": 3.0088,
+      "step": 10505
+    },
+    {
+      "epoch": 0.4423578947368421,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0003037863602702992,
+      "loss": 3.1206,
+      "step": 10506
+    },
+    {
+      "epoch": 0.4424,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003037534057214095,
+      "loss": 3.1215,
+      "step": 10507
+    },
+    {
+      "epoch": 0.4424421052631579,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003037204501931959,
+      "loss": 2.668,
+      "step": 10508
+    },
+    {
+      "epoch": 0.4424842105263158,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0003036874936862589,
+      "loss": 3.3978,
+      "step": 10509
+    },
+    {
+      "epoch": 0.44252631578947366,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0003036545362011988,
+      "loss": 3.4685,
+      "step": 10510
+    },
+    {
+      "epoch": 0.44256842105263156,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003036215777386161,
+      "loss": 3.304,
+      "step": 10511
+    },
+    {
+      "epoch": 0.44261052631578945,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00030358861829911116,
+      "loss": 3.2584,
+      "step": 10512
+    },
+    {
+      "epoch": 0.44265263157894735,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00030355565788328464,
+      "loss": 2.985,
+      "step": 10513
+    },
+    {
+      "epoch": 0.44269473684210525,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.000303522696491737,
+      "loss": 2.9167,
+      "step": 10514
+    },
+    {
+      "epoch": 0.44273684210526315,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00030348973412506867,
+      "loss": 2.9929,
+      "step": 10515
+    },
+    {
+      "epoch": 0.44277894736842105,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003034567707838802,
+      "loss": 3.148,
+      "step": 10516
+    },
+    {
+      "epoch": 0.44282105263157895,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00030342380646877223,
+      "loss": 3.4823,
+      "step": 10517
+    },
+    {
+      "epoch": 0.44286315789473685,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0003033908411803452,
+      "loss": 2.7543,
+      "step": 10518
+    },
+    {
+      "epoch": 0.44290526315789475,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00030335787491919984,
+      "loss": 2.9379,
+      "step": 10519
+    },
+    {
+      "epoch": 0.44294736842105265,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003033249076859367,
+      "loss": 3.2164,
+      "step": 10520
+    },
+    {
+      "epoch": 0.44298947368421054,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003032919394811564,
+      "loss": 3.3342,
+      "step": 10521
+    },
+    {
+      "epoch": 0.44303157894736844,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00030325897030545956,
+      "loss": 3.1205,
+      "step": 10522
+    },
+    {
+      "epoch": 0.44307368421052634,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003032260001594469,
+      "loss": 3.2605,
+      "step": 10523
+    },
+    {
+      "epoch": 0.4431157894736842,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00030319302904371903,
+      "loss": 3.0136,
+      "step": 10524
+    },
+    {
+      "epoch": 0.4431578947368421,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003031600569588766,
+      "loss": 3.4244,
+      "step": 10525
+    },
+    {
+      "epoch": 0.4432,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00030312708390552043,
+      "loss": 2.7623,
+      "step": 10526
+    },
+    {
+      "epoch": 0.4432421052631579,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00030309410988425123,
+      "loss": 3.2121,
+      "step": 10527
+    },
+    {
+      "epoch": 0.4432842105263158,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003030611348956697,
+      "loss": 2.9423,
+      "step": 10528
+    },
+    {
+      "epoch": 0.4433263157894737,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003030281589403767,
+      "loss": 3.3202,
+      "step": 10529
+    },
+    {
+      "epoch": 0.4433684210526316,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00030299518201897294,
+      "loss": 3.3964,
+      "step": 10530
+    },
+    {
+      "epoch": 0.4434105263157895,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00030296220413205923,
+      "loss": 3.1689,
+      "step": 10531
+    },
+    {
+      "epoch": 0.4434526315789474,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0003029292252802363,
+      "loss": 2.8913,
+      "step": 10532
+    },
+    {
+      "epoch": 0.4434947368421053,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00030289624546410513,
+      "loss": 2.8485,
+      "step": 10533
+    },
+    {
+      "epoch": 0.4435368421052632,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003028632646842665,
+      "loss": 3.3953,
+      "step": 10534
+    },
+    {
+      "epoch": 0.4435789473684211,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00030283028294132133,
+      "loss": 3.6729,
+      "step": 10535
+    },
+    {
+      "epoch": 0.44362105263157897,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0003027973002358705,
+      "loss": 3.0697,
+      "step": 10536
+    },
+    {
+      "epoch": 0.4436631578947368,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003027643165685149,
+      "loss": 3.0972,
+      "step": 10537
+    },
+    {
+      "epoch": 0.4437052631578947,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00030273133193985536,
+      "loss": 3.3721,
+      "step": 10538
+    },
+    {
+      "epoch": 0.4437473684210526,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.000302698346350493,
+      "loss": 3.105,
+      "step": 10539
+    },
+    {
+      "epoch": 0.4437894736842105,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00030266535980102863,
+      "loss": 3.3508,
+      "step": 10540
+    },
+    {
+      "epoch": 0.4438315789473684,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00030263237229206333,
+      "loss": 3.4048,
+      "step": 10541
+    },
+    {
+      "epoch": 0.4438736842105263,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003025993838241979,
+      "loss": 3.4544,
+      "step": 10542
+    },
+    {
+      "epoch": 0.4439157894736842,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0003025663943980337,
+      "loss": 2.9386,
+      "step": 10543
+    },
+    {
+      "epoch": 0.4439578947368421,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003025334040141715,
+      "loss": 3.4232,
+      "step": 10544
+    },
+    {
+      "epoch": 0.444,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0003025004126732123,
+      "loss": 3.1231,
+      "step": 10545
+    },
+    {
+      "epoch": 0.4440421052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003024674203757574,
+      "loss": 3.5238,
+      "step": 10546
+    },
+    {
+      "epoch": 0.4440842105263158,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003024344271224077,
+      "loss": 3.3495,
+      "step": 10547
+    },
+    {
+      "epoch": 0.4441263157894737,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0003024014329137644,
+      "loss": 2.7878,
+      "step": 10548
+    },
+    {
+      "epoch": 0.4441684210526316,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003023684377504284,
+      "loss": 3.5051,
+      "step": 10549
+    },
+    {
+      "epoch": 0.4442105263157895,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00030233544163300115,
+      "loss": 3.1932,
+      "step": 10550
+    },
+    {
+      "epoch": 0.44425263157894734,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00030230244456208355,
+      "loss": 3.1534,
+      "step": 10551
+    },
+    {
+      "epoch": 0.44429473684210524,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00030226944653827687,
+      "loss": 3.0033,
+      "step": 10552
+    },
+    {
+      "epoch": 0.44433684210526314,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0003022364475621824,
+      "loss": 2.9797,
+      "step": 10553
+    },
+    {
+      "epoch": 0.44437894736842104,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003022034476344011,
+      "loss": 3.2166,
+      "step": 10554
+    },
+    {
+      "epoch": 0.44442105263157894,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0003021704467555344,
+      "loss": 2.9338,
+      "step": 10555
+    },
+    {
+      "epoch": 0.44446315789473684,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00030213744492618347,
+      "loss": 3.3273,
+      "step": 10556
+    },
+    {
+      "epoch": 0.44450526315789474,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003021044421469495,
+      "loss": 3.1894,
+      "step": 10557
+    },
+    {
+      "epoch": 0.44454736842105264,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00030207143841843383,
+      "loss": 2.7169,
+      "step": 10558
+    },
+    {
+      "epoch": 0.44458947368421053,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003020384337412377,
+      "loss": 3.4001,
+      "step": 10559
+    },
+    {
+      "epoch": 0.44463157894736843,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00030200542811596243,
+      "loss": 2.7693,
+      "step": 10560
+    },
+    {
+      "epoch": 0.44467368421052633,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003019724215432095,
+      "loss": 2.4381,
+      "step": 10561
+    },
+    {
+      "epoch": 0.44471578947368423,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00030193941402358,
+      "loss": 3.337,
+      "step": 10562
+    },
+    {
+      "epoch": 0.44475789473684213,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00030190640555767545,
+      "loss": 3.365,
+      "step": 10563
+    },
+    {
+      "epoch": 0.4448,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003018733961460972,
+      "loss": 3.04,
+      "step": 10564
+    },
+    {
+      "epoch": 0.44484210526315787,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0003018403857894465,
+      "loss": 3.2255,
+      "step": 10565
+    },
+    {
+      "epoch": 0.44488421052631577,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00030180737448832497,
+      "loss": 2.8701,
+      "step": 10566
+    },
+    {
+      "epoch": 0.44492631578947367,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.000301774362243334,
+      "loss": 3.2432,
+      "step": 10567
+    },
+    {
+      "epoch": 0.44496842105263157,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00030174134905507484,
+      "loss": 3.2066,
+      "step": 10568
+    },
+    {
+      "epoch": 0.44501052631578947,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00030170833492414916,
+      "loss": 2.7211,
+      "step": 10569
+    },
+    {
+      "epoch": 0.44505263157894737,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003016753198511584,
+      "loss": 3.3144,
+      "step": 10570
+    },
+    {
+      "epoch": 0.44509473684210527,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003016423038367041,
+      "loss": 3.4753,
+      "step": 10571
+    },
+    {
+      "epoch": 0.44513684210526316,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00030160928688138756,
+      "loss": 3.5059,
+      "step": 10572
+    },
+    {
+      "epoch": 0.44517894736842106,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00030157626898581053,
+      "loss": 2.8016,
+      "step": 10573
+    },
+    {
+      "epoch": 0.44522105263157896,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00030154325015057444,
+      "loss": 2.814,
+      "step": 10574
+    },
+    {
+      "epoch": 0.44526315789473686,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00030151023037628094,
+      "loss": 2.9643,
+      "step": 10575
+    },
+    {
+      "epoch": 0.44530526315789476,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00030147720966353146,
+      "loss": 2.9215,
+      "step": 10576
+    },
+    {
+      "epoch": 0.44534736842105266,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003014441880129278,
+      "loss": 3.2176,
+      "step": 10577
+    },
+    {
+      "epoch": 0.4453894736842105,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003014111654250715,
+      "loss": 3.0316,
+      "step": 10578
+    },
+    {
+      "epoch": 0.4454315789473684,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003013781419005641,
+      "loss": 3.4477,
+      "step": 10579
+    },
+    {
+      "epoch": 0.4454736842105263,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003013451174400074,
+      "loss": 2.8588,
+      "step": 10580
+    },
+    {
+      "epoch": 0.4455157894736842,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003013120920440029,
+      "loss": 3.1239,
+      "step": 10581
+    },
+    {
+      "epoch": 0.4455578947368421,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003012790657131524,
+      "loss": 3.2689,
+      "step": 10582
+    },
+    {
+      "epoch": 0.4456,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00030124603844805764,
+      "loss": 3.191,
+      "step": 10583
+    },
+    {
+      "epoch": 0.4456421052631579,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003012130102493202,
+      "loss": 3.0597,
+      "step": 10584
+    },
+    {
+      "epoch": 0.4456842105263158,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00030117998111754186,
+      "loss": 3.0704,
+      "step": 10585
+    },
+    {
+      "epoch": 0.4457263157894737,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003011469510533244,
+      "loss": 3.1581,
+      "step": 10586
+    },
+    {
+      "epoch": 0.4457684210526316,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00030111392005726967,
+      "loss": 3.2039,
+      "step": 10587
+    },
+    {
+      "epoch": 0.4458105263157895,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00030108088812997935,
+      "loss": 3.1708,
+      "step": 10588
+    },
+    {
+      "epoch": 0.4458526315789474,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00030104785527205525,
+      "loss": 3.0177,
+      "step": 10589
+    },
+    {
+      "epoch": 0.4458947368421053,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003010148214840992,
+      "loss": 3.4261,
+      "step": 10590
+    },
+    {
+      "epoch": 0.44593684210526313,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00030098178676671306,
+      "loss": 2.9311,
+      "step": 10591
+    },
+    {
+      "epoch": 0.44597894736842103,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00030094875112049864,
+      "loss": 2.8207,
+      "step": 10592
+    },
+    {
+      "epoch": 0.44602105263157893,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00030091571454605775,
+      "loss": 3.2118,
+      "step": 10593
+    },
+    {
+      "epoch": 0.44606315789473683,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0003008826770439925,
+      "loss": 3.1249,
+      "step": 10594
+    },
+    {
+      "epoch": 0.4461052631578947,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00030084963861490467,
+      "loss": 2.8836,
+      "step": 10595
+    },
+    {
+      "epoch": 0.4461473684210526,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003008165992593961,
+      "loss": 3.1556,
+      "step": 10596
+    },
+    {
+      "epoch": 0.4461894736842105,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00030078355897806885,
+      "loss": 3.0289,
+      "step": 10597
+    },
+    {
+      "epoch": 0.4462315789473684,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003007505177715248,
+      "loss": 2.9401,
+      "step": 10598
+    },
+    {
+      "epoch": 0.4462736842105263,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003007174756403659,
+      "loss": 2.8606,
+      "step": 10599
+    },
+    {
+      "epoch": 0.4463157894736842,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00030068443258519424,
+      "loss": 3.4791,
+      "step": 10600
+    },
+    {
+      "epoch": 0.4463578947368421,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00030065138860661166,
+      "loss": 2.4884,
+      "step": 10601
+    },
+    {
+      "epoch": 0.4464,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003006183437052204,
+      "loss": 2.8169,
+      "step": 10602
+    },
+    {
+      "epoch": 0.4464421052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00030058529788162244,
+      "loss": 3.2385,
+      "step": 10603
+    },
+    {
+      "epoch": 0.4464842105263158,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0003005522511364197,
+      "loss": 2.9201,
+      "step": 10604
+    },
+    {
+      "epoch": 0.44652631578947366,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00030051920347021443,
+      "loss": 3.504,
+      "step": 10605
+    },
+    {
+      "epoch": 0.44656842105263156,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003004861548836085,
+      "loss": 2.9129,
+      "step": 10606
+    },
+    {
+      "epoch": 0.44661052631578946,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0003004531053772042,
+      "loss": 3.5747,
+      "step": 10607
+    },
+    {
+      "epoch": 0.44665263157894736,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0003004200549516037,
+      "loss": 2.9357,
+      "step": 10608
+    },
+    {
+      "epoch": 0.44669473684210526,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003003870036074089,
+      "loss": 3.269,
+      "step": 10609
+    },
+    {
+      "epoch": 0.44673684210526315,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00030035395134522217,
+      "loss": 2.9713,
+      "step": 10610
+    },
+    {
+      "epoch": 0.44677894736842105,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00030032089816564556,
+      "loss": 2.8832,
+      "step": 10611
+    },
+    {
+      "epoch": 0.44682105263157895,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0003002878440692813,
+      "loss": 2.9304,
+      "step": 10612
+    },
+    {
+      "epoch": 0.44686315789473685,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00030025478905673164,
+      "loss": 2.9484,
+      "step": 10613
+    },
+    {
+      "epoch": 0.44690526315789475,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003002217331285988,
+      "loss": 2.9904,
+      "step": 10614
+    },
+    {
+      "epoch": 0.44694736842105265,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00030018867628548503,
+      "loss": 2.9531,
+      "step": 10615
+    },
+    {
+      "epoch": 0.44698947368421055,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00030015561852799247,
+      "loss": 2.8714,
+      "step": 10616
+    },
+    {
+      "epoch": 0.44703157894736845,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003001225598567234,
+      "loss": 3.3532,
+      "step": 10617
+    },
+    {
+      "epoch": 0.4470736842105263,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0003000895002722803,
+      "loss": 3.5382,
+      "step": 10618
+    },
+    {
+      "epoch": 0.4471157894736842,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003000564397752653,
+      "loss": 3.2694,
+      "step": 10619
+    },
+    {
+      "epoch": 0.4471578947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003000233783662808,
+      "loss": 2.919,
+      "step": 10620
+    },
+    {
+      "epoch": 0.4472,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002999903160459292,
+      "loss": 3.5945,
+      "step": 10621
+    },
+    {
+      "epoch": 0.4472421052631579,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00029995725281481265,
+      "loss": 3.4309,
+      "step": 10622
+    },
+    {
+      "epoch": 0.4472842105263158,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002999241886735338,
+      "loss": 3.058,
+      "step": 10623
+    },
+    {
+      "epoch": 0.4473263157894737,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00029989112362269476,
+      "loss": 3.1303,
+      "step": 10624
+    },
+    {
+      "epoch": 0.4473684210526316,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002998580576628981,
+      "loss": 3.2856,
+      "step": 10625
+    },
+    {
+      "epoch": 0.4474105263157895,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002998249907947462,
+      "loss": 2.9981,
+      "step": 10626
+    },
+    {
+      "epoch": 0.4474526315789474,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002997919230188415,
+      "loss": 3.149,
+      "step": 10627
+    },
+    {
+      "epoch": 0.4474947368421053,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002997588543357865,
+      "loss": 3.0835,
+      "step": 10628
+    },
+    {
+      "epoch": 0.4475368421052632,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002997257847461836,
+      "loss": 2.7491,
+      "step": 10629
+    },
+    {
+      "epoch": 0.4475789473684211,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00029969271425063536,
+      "loss": 2.8589,
+      "step": 10630
+    },
+    {
+      "epoch": 0.447621052631579,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002996596428497443,
+      "loss": 3.4192,
+      "step": 10631
+    },
+    {
+      "epoch": 0.4476631578947368,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00029962657054411287,
+      "loss": 3.8253,
+      "step": 10632
+    },
+    {
+      "epoch": 0.4477052631578947,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00029959349733434354,
+      "loss": 2.9984,
+      "step": 10633
+    },
+    {
+      "epoch": 0.4477473684210526,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00029956042322103904,
+      "loss": 3.6453,
+      "step": 10634
+    },
+    {
+      "epoch": 0.4477894736842105,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002995273482048018,
+      "loss": 2.9385,
+      "step": 10635
+    },
+    {
+      "epoch": 0.4478315789473684,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00029949427228623455,
+      "loss": 3.4606,
+      "step": 10636
+    },
+    {
+      "epoch": 0.4478736842105263,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00029946119546593976,
+      "loss": 3.5396,
+      "step": 10637
+    },
+    {
+      "epoch": 0.4479157894736842,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00029942811774452015,
+      "loss": 3.3423,
+      "step": 10638
+    },
+    {
+      "epoch": 0.4479578947368421,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002993950391225782,
+      "loss": 3.1435,
+      "step": 10639
+    },
+    {
+      "epoch": 0.448,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00029936195960071675,
+      "loss": 3.3942,
+      "step": 10640
+    },
+    {
+      "epoch": 0.4480421052631579,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00029932887917953843,
+      "loss": 3.5683,
+      "step": 10641
+    },
+    {
+      "epoch": 0.4480842105263158,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002992957978596459,
+      "loss": 3.2275,
+      "step": 10642
+    },
+    {
+      "epoch": 0.4481263157894737,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00029926271564164177,
+      "loss": 3.4429,
+      "step": 10643
+    },
+    {
+      "epoch": 0.4481684210526316,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002992296325261289,
+      "loss": 3.3622,
+      "step": 10644
+    },
+    {
+      "epoch": 0.44821052631578945,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00029919654851371,
+      "loss": 2.8307,
+      "step": 10645
+    },
+    {
+      "epoch": 0.44825263157894735,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00029916346360498765,
+      "loss": 3.1204,
+      "step": 10646
+    },
+    {
+      "epoch": 0.44829473684210525,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00029913037780056486,
+      "loss": 2.8019,
+      "step": 10647
+    },
+    {
+      "epoch": 0.44833684210526314,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002990972911010443,
+      "loss": 2.9142,
+      "step": 10648
+    },
+    {
+      "epoch": 0.44837894736842104,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002990642035070288,
+      "loss": 2.553,
+      "step": 10649
+    },
+    {
+      "epoch": 0.44842105263157894,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002990311150191212,
+      "loss": 3.2851,
+      "step": 10650
+    },
+    {
+      "epoch": 0.44846315789473684,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00029899802563792416,
+      "loss": 2.9604,
+      "step": 10651
+    },
+    {
+      "epoch": 0.44850526315789474,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002989649353640408,
+      "loss": 3.1207,
+      "step": 10652
+    },
+    {
+      "epoch": 0.44854736842105264,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00029893184419807375,
+      "loss": 3.0314,
+      "step": 10653
+    },
+    {
+      "epoch": 0.44858947368421054,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.000298898752140626,
+      "loss": 3.2684,
+      "step": 10654
+    },
+    {
+      "epoch": 0.44863157894736844,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00029886565919230046,
+      "loss": 3.284,
+      "step": 10655
+    },
+    {
+      "epoch": 0.44867368421052634,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00029883256535370007,
+      "loss": 3.0025,
+      "step": 10656
+    },
+    {
+      "epoch": 0.44871578947368423,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00029879947062542766,
+      "loss": 3.0816,
+      "step": 10657
+    },
+    {
+      "epoch": 0.4487578947368421,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002987663750080863,
+      "loss": 3.3353,
+      "step": 10658
+    },
+    {
+      "epoch": 0.4488,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00029873327850227883,
+      "loss": 3.4437,
+      "step": 10659
+    },
+    {
+      "epoch": 0.4488421052631579,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002987001811086083,
+      "loss": 3.45,
+      "step": 10660
+    },
+    {
+      "epoch": 0.4488842105263158,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002986670828276776,
+      "loss": 3.1667,
+      "step": 10661
+    },
+    {
+      "epoch": 0.4489263157894737,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00029863398366008993,
+      "loss": 2.6959,
+      "step": 10662
+    },
+    {
+      "epoch": 0.44896842105263157,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002986008836064482,
+      "loss": 3.1509,
+      "step": 10663
+    },
+    {
+      "epoch": 0.44901052631578947,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00029856778266735553,
+      "loss": 3.3621,
+      "step": 10664
+    },
+    {
+      "epoch": 0.44905263157894737,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002985346808434149,
+      "loss": 2.951,
+      "step": 10665
+    },
+    {
+      "epoch": 0.44909473684210527,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002985015781352293,
+      "loss": 2.5951,
+      "step": 10666
+    },
+    {
+      "epoch": 0.44913684210526317,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002984684745434021,
+      "loss": 3.3283,
+      "step": 10667
+    },
+    {
+      "epoch": 0.44917894736842107,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00029843537006853614,
+      "loss": 3.091,
+      "step": 10668
+    },
+    {
+      "epoch": 0.44922105263157897,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002984022647112346,
+      "loss": 3.0417,
+      "step": 10669
+    },
+    {
+      "epoch": 0.44926315789473686,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002983691584721008,
+      "loss": 2.6233,
+      "step": 10670
+    },
+    {
+      "epoch": 0.44930526315789476,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002983360513517377,
+      "loss": 3.0433,
+      "step": 10671
+    },
+    {
+      "epoch": 0.4493473684210526,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00029830294335074855,
+      "loss": 3.4448,
+      "step": 10672
+    },
+    {
+      "epoch": 0.4493894736842105,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00029826983446973646,
+      "loss": 2.9305,
+      "step": 10673
+    },
+    {
+      "epoch": 0.4494315789473684,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00029823672470930473,
+      "loss": 3.1799,
+      "step": 10674
+    },
+    {
+      "epoch": 0.4494736842105263,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002982036140700566,
+      "loss": 3.7231,
+      "step": 10675
+    },
+    {
+      "epoch": 0.4495157894736842,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00029817050255259524,
+      "loss": 3.5226,
+      "step": 10676
+    },
+    {
+      "epoch": 0.4495578947368421,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0002981373901575239,
+      "loss": 2.7976,
+      "step": 10677
+    },
+    {
+      "epoch": 0.4496,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002981042768854459,
+      "loss": 3.3463,
+      "step": 10678
+    },
+    {
+      "epoch": 0.4496421052631579,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002980711627369645,
+      "loss": 3.0226,
+      "step": 10679
+    },
+    {
+      "epoch": 0.4496842105263158,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002980380477126829,
+      "loss": 3.0923,
+      "step": 10680
+    },
+    {
+      "epoch": 0.4497263157894737,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002980049318132046,
+      "loss": 2.811,
+      "step": 10681
+    },
+    {
+      "epoch": 0.4497684210526316,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002979718150391328,
+      "loss": 2.9547,
+      "step": 10682
+    },
+    {
+      "epoch": 0.4498105263157895,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.000297938697391071,
+      "loss": 3.3961,
+      "step": 10683
+    },
+    {
+      "epoch": 0.4498526315789474,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00029790557886962243,
+      "loss": 3.0942,
+      "step": 10684
+    },
+    {
+      "epoch": 0.44989473684210524,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002978724594753904,
+      "loss": 2.9861,
+      "step": 10685
+    },
+    {
+      "epoch": 0.44993684210526314,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00029783933920897844,
+      "loss": 2.6441,
+      "step": 10686
+    },
+    {
+      "epoch": 0.44997894736842103,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00029780621807098994,
+      "loss": 3.4641,
+      "step": 10687
+    },
+    {
+      "epoch": 0.45002105263157893,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002977730960620283,
+      "loss": 3.2242,
+      "step": 10688
+    },
+    {
+      "epoch": 0.45006315789473683,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00029773997318269707,
+      "loss": 3.0055,
+      "step": 10689
+    },
+    {
+      "epoch": 0.45010526315789473,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002977068494335996,
+      "loss": 2.9299,
+      "step": 10690
+    },
+    {
+      "epoch": 0.45014736842105263,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00029767372481533935,
+      "loss": 3.3921,
+      "step": 10691
+    },
+    {
+      "epoch": 0.45018947368421053,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002976405993285198,
+      "loss": 3.4299,
+      "step": 10692
+    },
+    {
+      "epoch": 0.4502315789473684,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00029760747297374456,
+      "loss": 2.9678,
+      "step": 10693
+    },
+    {
+      "epoch": 0.4502736842105263,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.000297574345751617,
+      "loss": 3.2219,
+      "step": 10694
+    },
+    {
+      "epoch": 0.4503157894736842,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002975412176627409,
+      "loss": 3.0976,
+      "step": 10695
+    },
+    {
+      "epoch": 0.4503578947368421,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002975080887077196,
+      "loss": 3.122,
+      "step": 10696
+    },
+    {
+      "epoch": 0.4504,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002974749588871568,
+      "loss": 3.5867,
+      "step": 10697
+    },
+    {
+      "epoch": 0.4504421052631579,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.000297441828201656,
+      "loss": 3.2347,
+      "step": 10698
+    },
+    {
+      "epoch": 0.45048421052631576,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00029740869665182085,
+      "loss": 3.4858,
+      "step": 10699
+    },
+    {
+      "epoch": 0.45052631578947366,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002973755642382549,
+      "loss": 3.23,
+      "step": 10700
+    },
+    {
+      "epoch": 0.45056842105263156,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002973424309615619,
+      "loss": 2.8876,
+      "step": 10701
+    },
+    {
+      "epoch": 0.45061052631578946,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002973092968223453,
+      "loss": 2.9289,
+      "step": 10702
+    },
+    {
+      "epoch": 0.45065263157894736,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.000297276161821209,
+      "loss": 3.0171,
+      "step": 10703
+    },
+    {
+      "epoch": 0.45069473684210526,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002972430259587566,
+      "loss": 3.3216,
+      "step": 10704
+    },
+    {
+      "epoch": 0.45073684210526316,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002972098892355918,
+      "loss": 3.217,
+      "step": 10705
+    },
+    {
+      "epoch": 0.45077894736842106,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002971767516523182,
+      "loss": 2.7713,
+      "step": 10706
+    },
+    {
+      "epoch": 0.45082105263157896,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00029714361320953966,
+      "loss": 2.9664,
+      "step": 10707
+    },
+    {
+      "epoch": 0.45086315789473685,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00029711047390785985,
+      "loss": 3.1795,
+      "step": 10708
+    },
+    {
+      "epoch": 0.45090526315789475,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002970773337478826,
+      "loss": 3.132,
+      "step": 10709
+    },
+    {
+      "epoch": 0.45094736842105265,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00029704419273021156,
+      "loss": 3.5035,
+      "step": 10710
+    },
+    {
+      "epoch": 0.45098947368421055,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002970110508554507,
+      "loss": 2.9548,
+      "step": 10711
+    },
+    {
+      "epoch": 0.4510315789473684,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002969779081242037,
+      "loss": 3.4058,
+      "step": 10712
+    },
+    {
+      "epoch": 0.4510736842105263,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002969447645370743,
+      "loss": 3.236,
+      "step": 10713
+    },
+    {
+      "epoch": 0.4511157894736842,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002969116200946666,
+      "loss": 3.6311,
+      "step": 10714
+    },
+    {
+      "epoch": 0.4511578947368421,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002968784747975842,
+      "loss": 3.3382,
+      "step": 10715
+    },
+    {
+      "epoch": 0.4512,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0002968453286464312,
+      "loss": 3.4923,
+      "step": 10716
+    },
+    {
+      "epoch": 0.4512421052631579,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00029681218164181127,
+      "loss": 3.1058,
+      "step": 10717
+    },
+    {
+      "epoch": 0.4512842105263158,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00029677903378432836,
+      "loss": 2.9032,
+      "step": 10718
+    },
+    {
+      "epoch": 0.4513263157894737,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00029674588507458643,
+      "loss": 3.0587,
+      "step": 10719
+    },
+    {
+      "epoch": 0.4513684210526316,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002967127355131894,
+      "loss": 3.0987,
+      "step": 10720
+    },
+    {
+      "epoch": 0.4514105263157895,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00029667958510074124,
+      "loss": 3.0001,
+      "step": 10721
+    },
+    {
+      "epoch": 0.4514526315789474,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002966464338378459,
+      "loss": 2.9063,
+      "step": 10722
+    },
+    {
+      "epoch": 0.4514947368421053,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002966132817251073,
+      "loss": 3.4609,
+      "step": 10723
+    },
+    {
+      "epoch": 0.4515368421052632,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00029658012876312956,
+      "loss": 3.0503,
+      "step": 10724
+    },
+    {
+      "epoch": 0.4515789473684211,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002965469749525165,
+      "loss": 2.9545,
+      "step": 10725
+    },
+    {
+      "epoch": 0.4516210526315789,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002965138202938723,
+      "loss": 2.8021,
+      "step": 10726
+    },
+    {
+      "epoch": 0.4516631578947368,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00029648066478780093,
+      "loss": 3.1825,
+      "step": 10727
+    },
+    {
+      "epoch": 0.4517052631578947,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002964475084349064,
+      "loss": 3.0208,
+      "step": 10728
+    },
+    {
+      "epoch": 0.4517473684210526,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00029641435123579294,
+      "loss": 3.0671,
+      "step": 10729
+    },
+    {
+      "epoch": 0.4517894736842105,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002963811931910645,
+      "loss": 3.1906,
+      "step": 10730
+    },
+    {
+      "epoch": 0.4518315789473684,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002963480343013252,
+      "loss": 2.9931,
+      "step": 10731
+    },
+    {
+      "epoch": 0.4518736842105263,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00029631487456717925,
+      "loss": 3.2474,
+      "step": 10732
+    },
+    {
+      "epoch": 0.4519157894736842,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00029628171398923066,
+      "loss": 3.5211,
+      "step": 10733
+    },
+    {
+      "epoch": 0.4519578947368421,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002962485525680835,
+      "loss": 2.8547,
+      "step": 10734
+    },
+    {
+      "epoch": 0.452,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002962153903043422,
+      "loss": 3.0589,
+      "step": 10735
+    },
+    {
+      "epoch": 0.4520421052631579,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002961822271986107,
+      "loss": 3.3236,
+      "step": 10736
+    },
+    {
+      "epoch": 0.4520842105263158,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00029614906325149334,
+      "loss": 3.3026,
+      "step": 10737
+    },
+    {
+      "epoch": 0.4521263157894737,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002961158984635943,
+      "loss": 2.8896,
+      "step": 10738
+    },
+    {
+      "epoch": 0.45216842105263155,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002960827328355177,
+      "loss": 2.9734,
+      "step": 10739
+    },
+    {
+      "epoch": 0.45221052631578945,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00029604956636786786,
+      "loss": 3.1898,
+      "step": 10740
+    },
+    {
+      "epoch": 0.45225263157894735,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00029601639906124907,
+      "loss": 2.9725,
+      "step": 10741
+    },
+    {
+      "epoch": 0.45229473684210525,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002959832309162655,
+      "loss": 3.4522,
+      "step": 10742
+    },
+    {
+      "epoch": 0.45233684210526315,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002959500619335216,
+      "loss": 3.2405,
+      "step": 10743
+    },
+    {
+      "epoch": 0.45237894736842105,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002959168921136214,
+      "loss": 3.1098,
+      "step": 10744
+    },
+    {
+      "epoch": 0.45242105263157895,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00029588372145716944,
+      "loss": 2.8774,
+      "step": 10745
+    },
+    {
+      "epoch": 0.45246315789473684,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002958505499647701,
+      "loss": 3.1624,
+      "step": 10746
+    },
+    {
+      "epoch": 0.45250526315789474,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002958173776370274,
+      "loss": 2.9587,
+      "step": 10747
+    },
+    {
+      "epoch": 0.45254736842105264,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00029578420447454607,
+      "loss": 2.8961,
+      "step": 10748
+    },
+    {
+      "epoch": 0.45258947368421054,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00029575103047793026,
+      "loss": 3.0483,
+      "step": 10749
+    },
+    {
+      "epoch": 0.45263157894736844,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00029571785564778453,
+      "loss": 3.0328,
+      "step": 10750
+    },
+    {
+      "epoch": 0.45267368421052634,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002956846799847131,
+      "loss": 3.2593,
+      "step": 10751
+    },
+    {
+      "epoch": 0.45271578947368424,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00029565150348932046,
+      "loss": 3.3335,
+      "step": 10752
+    },
+    {
+      "epoch": 0.4527578947368421,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002956183261622111,
+      "loss": 3.399,
+      "step": 10753
+    },
+    {
+      "epoch": 0.4528,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002955851480039894,
+      "loss": 2.9145,
+      "step": 10754
+    },
+    {
+      "epoch": 0.4528421052631579,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002955519690152599,
+      "loss": 3.2448,
+      "step": 10755
+    },
+    {
+      "epoch": 0.4528842105263158,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002955187891966271,
+      "loss": 2.6433,
+      "step": 10756
+    },
+    {
+      "epoch": 0.4529263157894737,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002954856085486954,
+      "loss": 3.5114,
+      "step": 10757
+    },
+    {
+      "epoch": 0.4529684210526316,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002954524270720693,
+      "loss": 3.2221,
+      "step": 10758
+    },
+    {
+      "epoch": 0.4530105263157895,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00029541924476735347,
+      "loss": 3.3745,
+      "step": 10759
+    },
+    {
+      "epoch": 0.4530526315789474,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002953860616351523,
+      "loss": 3.1892,
+      "step": 10760
+    },
+    {
+      "epoch": 0.4530947368421053,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002953528776760704,
+      "loss": 2.9072,
+      "step": 10761
+    },
+    {
+      "epoch": 0.45313684210526317,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00029531969289071236,
+      "loss": 3.0702,
+      "step": 10762
+    },
+    {
+      "epoch": 0.45317894736842107,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00029528650727968285,
+      "loss": 3.0727,
+      "step": 10763
+    },
+    {
+      "epoch": 0.45322105263157897,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002952533208435864,
+      "loss": 3.2286,
+      "step": 10764
+    },
+    {
+      "epoch": 0.45326315789473687,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002952201335830275,
+      "loss": 3.2382,
+      "step": 10765
+    },
+    {
+      "epoch": 0.4533052631578947,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00029518694549861095,
+      "loss": 3.169,
+      "step": 10766
+    },
+    {
+      "epoch": 0.4533473684210526,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002951537565909413,
+      "loss": 2.8632,
+      "step": 10767
+    },
+    {
+      "epoch": 0.4533894736842105,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00029512056686062334,
+      "loss": 2.9703,
+      "step": 10768
+    },
+    {
+      "epoch": 0.4534315789473684,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00029508737630826154,
+      "loss": 3.3141,
+      "step": 10769
+    },
+    {
+      "epoch": 0.4534736842105263,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00029505418493446086,
+      "loss": 3.3536,
+      "step": 10770
+    },
+    {
+      "epoch": 0.4535157894736842,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002950209927398258,
+      "loss": 3.3407,
+      "step": 10771
+    },
+    {
+      "epoch": 0.4535578947368421,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002949877997249612,
+      "loss": 3.0141,
+      "step": 10772
+    },
+    {
+      "epoch": 0.4536,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002949546058904717,
+      "loss": 3.0799,
+      "step": 10773
+    },
+    {
+      "epoch": 0.4536421052631579,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002949214112369621,
+      "loss": 3.19,
+      "step": 10774
+    },
+    {
+      "epoch": 0.4536842105263158,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00029488821576503717,
+      "loss": 2.6722,
+      "step": 10775
+    },
+    {
+      "epoch": 0.4537263157894737,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002948550194753017,
+      "loss": 3.5367,
+      "step": 10776
+    },
+    {
+      "epoch": 0.4537684210526316,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00029482182236836044,
+      "loss": 3.2778,
+      "step": 10777
+    },
+    {
+      "epoch": 0.4538105263157895,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002947886244448183,
+      "loss": 3.0304,
+      "step": 10778
+    },
+    {
+      "epoch": 0.45385263157894734,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00029475542570528,
+      "loss": 3.087,
+      "step": 10779
+    },
+    {
+      "epoch": 0.45389473684210524,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002947222261503505,
+      "loss": 2.9305,
+      "step": 10780
+    },
+    {
+      "epoch": 0.45393684210526314,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0002946890257806345,
+      "loss": 3.4819,
+      "step": 10781
+    },
+    {
+      "epoch": 0.45397894736842104,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00029465582459673703,
+      "loss": 3.2831,
+      "step": 10782
+    },
+    {
+      "epoch": 0.45402105263157894,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0002946226225992628,
+      "loss": 3.0029,
+      "step": 10783
+    },
+    {
+      "epoch": 0.45406315789473684,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00029458941978881693,
+      "loss": 3.108,
+      "step": 10784
+    },
+    {
+      "epoch": 0.45410526315789473,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00029455621616600417,
+      "loss": 2.8524,
+      "step": 10785
+    },
+    {
+      "epoch": 0.45414736842105263,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002945230117314295,
+      "loss": 3.2725,
+      "step": 10786
+    },
+    {
+      "epoch": 0.45418947368421053,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002944898064856978,
+      "loss": 3.5465,
+      "step": 10787
+    },
+    {
+      "epoch": 0.45423157894736843,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002944566004294142,
+      "loss": 3.2026,
+      "step": 10788
+    },
+    {
+      "epoch": 0.45427368421052633,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002944233935631836,
+      "loss": 3.1862,
+      "step": 10789
+    },
+    {
+      "epoch": 0.45431578947368423,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002943901858876109,
+      "loss": 3.1205,
+      "step": 10790
+    },
+    {
+      "epoch": 0.4543578947368421,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0002943569774033012,
+      "loss": 3.0645,
+      "step": 10791
+    },
+    {
+      "epoch": 0.4544,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00029432376811085946,
+      "loss": 2.8163,
+      "step": 10792
+    },
+    {
+      "epoch": 0.45444210526315787,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00029429055801089075,
+      "loss": 2.9481,
+      "step": 10793
+    },
+    {
+      "epoch": 0.45448421052631577,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00029425734710400014,
+      "loss": 2.7775,
+      "step": 10794
+    },
+    {
+      "epoch": 0.45452631578947367,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00029422413539079265,
+      "loss": 3.406,
+      "step": 10795
+    },
+    {
+      "epoch": 0.45456842105263157,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002941909228718733,
+      "loss": 3.2039,
+      "step": 10796
+    },
+    {
+      "epoch": 0.45461052631578946,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00029415770954784737,
+      "loss": 3.3115,
+      "step": 10797
+    },
+    {
+      "epoch": 0.45465263157894736,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00029412449541931987,
+      "loss": 3.129,
+      "step": 10798
+    },
+    {
+      "epoch": 0.45469473684210526,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002940912804868958,
+      "loss": 3.3204,
+      "step": 10799
+    },
+    {
+      "epoch": 0.45473684210526316,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002940580647511805,
+      "loss": 2.9686,
+      "step": 10800
+    },
+    {
+      "epoch": 0.45477894736842106,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002940248482127789,
+      "loss": 3.7095,
+      "step": 10801
+    },
+    {
+      "epoch": 0.45482105263157896,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002939916308722964,
+      "loss": 3.1674,
+      "step": 10802
+    },
+    {
+      "epoch": 0.45486315789473686,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.000293958412730338,
+      "loss": 3.158,
+      "step": 10803
+    },
+    {
+      "epoch": 0.45490526315789476,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00029392519378750905,
+      "loss": 2.4594,
+      "step": 10804
+    },
+    {
+      "epoch": 0.45494736842105266,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00029389197404441463,
+      "loss": 2.8781,
+      "step": 10805
+    },
+    {
+      "epoch": 0.4549894736842105,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002938587535016601,
+      "loss": 3.3049,
+      "step": 10806
+    },
+    {
+      "epoch": 0.4550315789473684,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00029382553215985053,
+      "loss": 3.5205,
+      "step": 10807
+    },
+    {
+      "epoch": 0.4550736842105263,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00029379231001959124,
+      "loss": 3.0637,
+      "step": 10808
+    },
+    {
+      "epoch": 0.4551157894736842,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002937590870814876,
+      "loss": 2.7098,
+      "step": 10809
+    },
+    {
+      "epoch": 0.4551578947368421,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002937258633461447,
+      "loss": 3.6539,
+      "step": 10810
+    },
+    {
+      "epoch": 0.4552,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.000293692638814168,
+      "loss": 3.2849,
+      "step": 10811
+    },
+    {
+      "epoch": 0.4552421052631579,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002936594134861628,
+      "loss": 2.8887,
+      "step": 10812
+    },
+    {
+      "epoch": 0.4552842105263158,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002936261873627343,
+      "loss": 2.8937,
+      "step": 10813
+    },
+    {
+      "epoch": 0.4553263157894737,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002935929604444879,
+      "loss": 3.2617,
+      "step": 10814
+    },
+    {
+      "epoch": 0.4553684210526316,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00029355973273202906,
+      "loss": 3.758,
+      "step": 10815
+    },
+    {
+      "epoch": 0.4554105263157895,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00029352650422596313,
+      "loss": 3.0081,
+      "step": 10816
+    },
+    {
+      "epoch": 0.4554526315789474,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002934932749268953,
+      "loss": 3.3197,
+      "step": 10817
+    },
+    {
+      "epoch": 0.4554947368421053,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002934600448354312,
+      "loss": 3.046,
+      "step": 10818
+    },
+    {
+      "epoch": 0.4555368421052632,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00029342681395217606,
+      "loss": 3.0923,
+      "step": 10819
+    },
+    {
+      "epoch": 0.45557894736842103,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002933935822777355,
+      "loss": 2.9966,
+      "step": 10820
+    },
+    {
+      "epoch": 0.4556210526315789,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00029336034981271476,
+      "loss": 2.9183,
+      "step": 10821
+    },
+    {
+      "epoch": 0.4556631578947368,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00029332711655771944,
+      "loss": 3.259,
+      "step": 10822
+    },
+    {
+      "epoch": 0.4557052631578947,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002932938825133549,
+      "loss": 3.1181,
+      "step": 10823
+    },
+    {
+      "epoch": 0.4557473684210526,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00029326064768022683,
+      "loss": 3.3532,
+      "step": 10824
+    },
+    {
+      "epoch": 0.4557894736842105,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002932274120589405,
+      "loss": 2.6507,
+      "step": 10825
+    },
+    {
+      "epoch": 0.4558315789473684,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00029319417565010155,
+      "loss": 3.4675,
+      "step": 10826
+    },
+    {
+      "epoch": 0.4558736842105263,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00029316093845431533,
+      "loss": 2.7558,
+      "step": 10827
+    },
+    {
+      "epoch": 0.4559157894736842,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002931277004721877,
+      "loss": 2.8448,
+      "step": 10828
+    },
+    {
+      "epoch": 0.4559578947368421,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002930944617043239,
+      "loss": 2.8134,
+      "step": 10829
+    },
+    {
+      "epoch": 0.456,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002930612221513297,
+      "loss": 3.2255,
+      "step": 10830
+    },
+    {
+      "epoch": 0.4560421052631579,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002930279818138107,
+      "loss": 3.111,
+      "step": 10831
+    },
+    {
+      "epoch": 0.4560842105263158,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002929947406923724,
+      "loss": 2.8973,
+      "step": 10832
+    },
+    {
+      "epoch": 0.45612631578947366,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002929614987876204,
+      "loss": 3.1223,
+      "step": 10833
+    },
+    {
+      "epoch": 0.45616842105263156,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00029292825610016036,
+      "loss": 2.9167,
+      "step": 10834
+    },
+    {
+      "epoch": 0.45621052631578946,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00029289501263059795,
+      "loss": 3.2726,
+      "step": 10835
+    },
+    {
+      "epoch": 0.45625263157894735,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002928617683795388,
+      "loss": 3.4003,
+      "step": 10836
+    },
+    {
+      "epoch": 0.45629473684210525,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002928285233475886,
+      "loss": 3.4195,
+      "step": 10837
+    },
+    {
+      "epoch": 0.45633684210526315,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00029279527753535303,
+      "loss": 2.9525,
+      "step": 10838
+    },
+    {
+      "epoch": 0.45637894736842105,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00029276203094343777,
+      "loss": 3.2492,
+      "step": 10839
+    },
+    {
+      "epoch": 0.45642105263157895,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002927287835724486,
+      "loss": 3.3463,
+      "step": 10840
+    },
+    {
+      "epoch": 0.45646315789473685,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002926955354229911,
+      "loss": 3.2141,
+      "step": 10841
+    },
+    {
+      "epoch": 0.45650526315789475,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00029266228649567116,
+      "loss": 3.1454,
+      "step": 10842
+    },
+    {
+      "epoch": 0.45654736842105265,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00029262903679109456,
+      "loss": 3.17,
+      "step": 10843
+    },
+    {
+      "epoch": 0.45658947368421055,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002925957863098669,
+      "loss": 3.4798,
+      "step": 10844
+    },
+    {
+      "epoch": 0.45663157894736844,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00029256253505259404,
+      "loss": 3.1126,
+      "step": 10845
+    },
+    {
+      "epoch": 0.45667368421052634,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002925292830198819,
+      "loss": 3.2892,
+      "step": 10846
+    },
+    {
+      "epoch": 0.4567157894736842,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00029249603021233616,
+      "loss": 3.2025,
+      "step": 10847
+    },
+    {
+      "epoch": 0.4567578947368421,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00029246277663056265,
+      "loss": 2.9418,
+      "step": 10848
+    },
+    {
+      "epoch": 0.4568,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002924295222751672,
+      "loss": 3.6867,
+      "step": 10849
+    },
+    {
+      "epoch": 0.4568421052631579,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00029239626714675583,
+      "loss": 3.2373,
+      "step": 10850
+    },
+    {
+      "epoch": 0.4568842105263158,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002923630112459342,
+      "loss": 2.6255,
+      "step": 10851
+    },
+    {
+      "epoch": 0.4569263157894737,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002923297545733083,
+      "loss": 3.1834,
+      "step": 10852
+    },
+    {
+      "epoch": 0.4569684210526316,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.000292296497129484,
+      "loss": 2.9693,
+      "step": 10853
+    },
+    {
+      "epoch": 0.4570105263157895,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002922632389150672,
+      "loss": 3.2842,
+      "step": 10854
+    },
+    {
+      "epoch": 0.4570526315789474,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002922299799306639,
+      "loss": 3.5767,
+      "step": 10855
+    },
+    {
+      "epoch": 0.4570947368421053,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00029219672017687993,
+      "loss": 3.406,
+      "step": 10856
+    },
+    {
+      "epoch": 0.4571368421052632,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00029216345965432135,
+      "loss": 3.2245,
+      "step": 10857
+    },
+    {
+      "epoch": 0.4571789473684211,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00029213019836359403,
+      "loss": 3.284,
+      "step": 10858
+    },
+    {
+      "epoch": 0.457221052631579,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002920969363053041,
+      "loss": 3.5865,
+      "step": 10859
+    },
+    {
+      "epoch": 0.4572631578947368,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002920636734800574,
+      "loss": 2.8177,
+      "step": 10860
+    },
+    {
+      "epoch": 0.4573052631578947,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00029203040988845985,
+      "loss": 3.277,
+      "step": 10861
+    },
+    {
+      "epoch": 0.4573473684210526,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00029199714553111765,
+      "loss": 3.2967,
+      "step": 10862
+    },
+    {
+      "epoch": 0.4573894736842105,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002919638804086369,
+      "loss": 2.8957,
+      "step": 10863
+    },
+    {
+      "epoch": 0.4574315789473684,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002919306145216236,
+      "loss": 3.1055,
+      "step": 10864
+    },
+    {
+      "epoch": 0.4574736842105263,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00029189734787068365,
+      "loss": 3.307,
+      "step": 10865
+    },
+    {
+      "epoch": 0.4575157894736842,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002918640804564234,
+      "loss": 2.779,
+      "step": 10866
+    },
+    {
+      "epoch": 0.4575578947368421,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0002918308122794487,
+      "loss": 3.2026,
+      "step": 10867
+    },
+    {
+      "epoch": 0.4576,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002917975433403657,
+      "loss": 3.1163,
+      "step": 10868
+    },
+    {
+      "epoch": 0.4576421052631579,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00029176427363978056,
+      "loss": 2.857,
+      "step": 10869
+    },
+    {
+      "epoch": 0.4576842105263158,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00029173100317829946,
+      "loss": 3.3748,
+      "step": 10870
+    },
+    {
+      "epoch": 0.4577263157894737,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00029169773195652853,
+      "loss": 3.3672,
+      "step": 10871
+    },
+    {
+      "epoch": 0.4577684210526316,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00029166445997507394,
+      "loss": 3.3622,
+      "step": 10872
+    },
+    {
+      "epoch": 0.4578105263157895,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002916311872345419,
+      "loss": 3.1364,
+      "step": 10873
+    },
+    {
+      "epoch": 0.45785263157894734,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00029159791373553845,
+      "loss": 3.3991,
+      "step": 10874
+    },
+    {
+      "epoch": 0.45789473684210524,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00029156463947866985,
+      "loss": 3.5751,
+      "step": 10875
+    },
+    {
+      "epoch": 0.45793684210526314,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002915313644645424,
+      "loss": 3.1791,
+      "step": 10876
+    },
+    {
+      "epoch": 0.45797894736842104,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002914980886937623,
+      "loss": 2.9953,
+      "step": 10877
+    },
+    {
+      "epoch": 0.45802105263157894,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00029146481216693577,
+      "loss": 3.1552,
+      "step": 10878
+    },
+    {
+      "epoch": 0.45806315789473684,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0002914315348846691,
+      "loss": 2.9996,
+      "step": 10879
+    },
+    {
+      "epoch": 0.45810526315789474,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00029139825684756855,
+      "loss": 3.187,
+      "step": 10880
+    },
+    {
+      "epoch": 0.45814736842105264,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002913649780562404,
+      "loss": 3.2066,
+      "step": 10881
+    },
+    {
+      "epoch": 0.45818947368421054,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002913316985112909,
+      "loss": 2.8667,
+      "step": 10882
+    },
+    {
+      "epoch": 0.45823157894736843,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002912984182133265,
+      "loss": 2.8634,
+      "step": 10883
+    },
+    {
+      "epoch": 0.45827368421052633,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00029126513716295333,
+      "loss": 3.4343,
+      "step": 10884
+    },
+    {
+      "epoch": 0.45831578947368423,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00029123185536077795,
+      "loss": 3.296,
+      "step": 10885
+    },
+    {
+      "epoch": 0.45835789473684213,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00029119857280740654,
+      "loss": 2.8707,
+      "step": 10886
+    },
+    {
+      "epoch": 0.4584,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00029116528950344554,
+      "loss": 3.2486,
+      "step": 10887
+    },
+    {
+      "epoch": 0.4584421052631579,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00029113200544950135,
+      "loss": 2.8539,
+      "step": 10888
+    },
+    {
+      "epoch": 0.45848421052631577,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00029109872064618037,
+      "loss": 3.1036,
+      "step": 10889
+    },
+    {
+      "epoch": 0.45852631578947367,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.000291065435094089,
+      "loss": 3.2735,
+      "step": 10890
+    },
+    {
+      "epoch": 0.45856842105263157,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002910321487938337,
+      "loss": 3.2538,
+      "step": 10891
+    },
+    {
+      "epoch": 0.45861052631578947,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00029099886174602073,
+      "loss": 3.0726,
+      "step": 10892
+    },
+    {
+      "epoch": 0.45865263157894737,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00029096557395125676,
+      "loss": 3.3193,
+      "step": 10893
+    },
+    {
+      "epoch": 0.45869473684210527,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002909322854101481,
+      "loss": 3.2963,
+      "step": 10894
+    },
+    {
+      "epoch": 0.45873684210526317,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002908989961233014,
+      "loss": 3.0883,
+      "step": 10895
+    },
+    {
+      "epoch": 0.45877894736842106,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002908657060913229,
+      "loss": 3.082,
+      "step": 10896
+    },
+    {
+      "epoch": 0.45882105263157896,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00029083241531481936,
+      "loss": 3.1598,
+      "step": 10897
+    },
+    {
+      "epoch": 0.45886315789473686,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00029079912379439716,
+      "loss": 3.3134,
+      "step": 10898
+    },
+    {
+      "epoch": 0.45890526315789476,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00029076583153066286,
+      "loss": 3.0177,
+      "step": 10899
+    },
+    {
+      "epoch": 0.4589473684210526,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.000290732538524223,
+      "loss": 3.5606,
+      "step": 10900
+    },
+    {
+      "epoch": 0.4589894736842105,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002906992447756842,
+      "loss": 3.1346,
+      "step": 10901
+    },
+    {
+      "epoch": 0.4590315789473684,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002906659502856529,
+      "loss": 3.073,
+      "step": 10902
+    },
+    {
+      "epoch": 0.4590736842105263,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002906326550547358,
+      "loss": 2.9789,
+      "step": 10903
+    },
+    {
+      "epoch": 0.4591157894736842,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002905993590835394,
+      "loss": 3.4766,
+      "step": 10904
+    },
+    {
+      "epoch": 0.4591578947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002905660623726705,
+      "loss": 3.1613,
+      "step": 10905
+    },
+    {
+      "epoch": 0.4592,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00029053276492273553,
+      "loss": 3.3436,
+      "step": 10906
+    },
+    {
+      "epoch": 0.4592421052631579,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002904994667343412,
+      "loss": 2.7727,
+      "step": 10907
+    },
+    {
+      "epoch": 0.4592842105263158,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00029046616780809426,
+      "loss": 2.9928,
+      "step": 10908
+    },
+    {
+      "epoch": 0.4593263157894737,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00029043286814460116,
+      "loss": 3.3697,
+      "step": 10909
+    },
+    {
+      "epoch": 0.4593684210526316,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00029039956774446884,
+      "loss": 3.0619,
+      "step": 10910
+    },
+    {
+      "epoch": 0.4594105263157895,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002903662666083038,
+      "loss": 3.3863,
+      "step": 10911
+    },
+    {
+      "epoch": 0.4594526315789474,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002903329647367128,
+      "loss": 3.1996,
+      "step": 10912
+    },
+    {
+      "epoch": 0.4594947368421053,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00029029966213030257,
+      "loss": 3.2989,
+      "step": 10913
+    },
+    {
+      "epoch": 0.45953684210526313,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002902663587896799,
+      "loss": 3.2251,
+      "step": 10914
+    },
+    {
+      "epoch": 0.45957894736842103,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00029023305471545134,
+      "loss": 3.2306,
+      "step": 10915
+    },
+    {
+      "epoch": 0.45962105263157893,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00029019974990822394,
+      "loss": 2.8315,
+      "step": 10916
+    },
+    {
+      "epoch": 0.45966315789473683,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00029016644436860426,
+      "loss": 2.8982,
+      "step": 10917
+    },
+    {
+      "epoch": 0.45970526315789473,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002901331380971992,
+      "loss": 3.0095,
+      "step": 10918
+    },
+    {
+      "epoch": 0.4597473684210526,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002900998310946156,
+      "loss": 2.9804,
+      "step": 10919
+    },
+    {
+      "epoch": 0.4597894736842105,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00029006652336146,
+      "loss": 3.062,
+      "step": 10920
+    },
+    {
+      "epoch": 0.4598315789473684,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00029003321489833955,
+      "loss": 2.8508,
+      "step": 10921
+    },
+    {
+      "epoch": 0.4598736842105263,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002899999057058609,
+      "loss": 3.248,
+      "step": 10922
+    },
+    {
+      "epoch": 0.4599157894736842,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00028996659578463095,
+      "loss": 3.3538,
+      "step": 10923
+    },
+    {
+      "epoch": 0.4599578947368421,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002899332851352567,
+      "loss": 2.8977,
+      "step": 10924
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00028989997375834483,
+      "loss": 3.2788,
+      "step": 10925
+    },
+    {
+      "epoch": 0.4600421052631579,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00028986666165450235,
+      "loss": 2.8974,
+      "step": 10926
+    },
+    {
+      "epoch": 0.46008421052631576,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00028983334882433614,
+      "loss": 3.2715,
+      "step": 10927
+    },
+    {
+      "epoch": 0.46012631578947366,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00028980003526845303,
+      "loss": 3.5181,
+      "step": 10928
+    },
+    {
+      "epoch": 0.46016842105263156,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00028976672098746006,
+      "loss": 3.0002,
+      "step": 10929
+    },
+    {
+      "epoch": 0.46021052631578946,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002897334059819642,
+      "loss": 3.199,
+      "step": 10930
+    },
+    {
+      "epoch": 0.46025263157894736,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00028970009025257245,
+      "loss": 3.7068,
+      "step": 10931
+    },
+    {
+      "epoch": 0.46029473684210526,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00028966677379989157,
+      "loss": 3.3196,
+      "step": 10932
+    },
+    {
+      "epoch": 0.46033684210526316,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002896334566245288,
+      "loss": 2.9581,
+      "step": 10933
+    },
+    {
+      "epoch": 0.46037894736842105,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002896001387270909,
+      "loss": 3.3289,
+      "step": 10934
+    },
+    {
+      "epoch": 0.46042105263157895,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00028956682010818503,
+      "loss": 3.3524,
+      "step": 10935
+    },
+    {
+      "epoch": 0.46046315789473685,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00028953350076841824,
+      "loss": 3.3424,
+      "step": 10936
+    },
+    {
+      "epoch": 0.46050526315789475,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002895001807083974,
+      "loss": 3.4535,
+      "step": 10937
+    },
+    {
+      "epoch": 0.46054736842105265,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002894668599287298,
+      "loss": 2.9181,
+      "step": 10938
+    },
+    {
+      "epoch": 0.46058947368421055,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002894335384300223,
+      "loss": 3.1239,
+      "step": 10939
+    },
+    {
+      "epoch": 0.46063157894736845,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00028940021621288217,
+      "loss": 3.2357,
+      "step": 10940
+    },
+    {
+      "epoch": 0.4606736842105263,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002893668932779163,
+      "loss": 3.0255,
+      "step": 10941
+    },
+    {
+      "epoch": 0.4607157894736842,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002893335696257319,
+      "loss": 3.3788,
+      "step": 10942
+    },
+    {
+      "epoch": 0.4607578947368421,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002893002452569361,
+      "loss": 3.4051,
+      "step": 10943
+    },
+    {
+      "epoch": 0.4608,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.000289266920172136,
+      "loss": 2.7226,
+      "step": 10944
+    },
+    {
+      "epoch": 0.4608421052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002892335943719387,
+      "loss": 3.1065,
+      "step": 10945
+    },
+    {
+      "epoch": 0.4608842105263158,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00028920026785695155,
+      "loss": 3.1016,
+      "step": 10946
+    },
+    {
+      "epoch": 0.4609263157894737,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002891669406277815,
+      "loss": 3.5839,
+      "step": 10947
+    },
+    {
+      "epoch": 0.4609684210526316,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002891336126850358,
+      "loss": 3.0331,
+      "step": 10948
+    },
+    {
+      "epoch": 0.4610105263157895,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002891002840293216,
+      "loss": 3.5054,
+      "step": 10949
+    },
+    {
+      "epoch": 0.4610526315789474,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002890669546612462,
+      "loss": 3.1413,
+      "step": 10950
+    },
+    {
+      "epoch": 0.4610947368421053,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00028903362458141687,
+      "loss": 3.0297,
+      "step": 10951
+    },
+    {
+      "epoch": 0.4611368421052632,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002890002937904407,
+      "loss": 3.3813,
+      "step": 10952
+    },
+    {
+      "epoch": 0.4611789473684211,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.000288966962288925,
+      "loss": 3.4627,
+      "step": 10953
+    },
+    {
+      "epoch": 0.4612210526315789,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.000288933630077477,
+      "loss": 3.0975,
+      "step": 10954
+    },
+    {
+      "epoch": 0.4612631578947368,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00028890029715670406,
+      "loss": 2.6028,
+      "step": 10955
+    },
+    {
+      "epoch": 0.4613052631578947,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00028886696352721335,
+      "loss": 3.401,
+      "step": 10956
+    },
+    {
+      "epoch": 0.4613473684210526,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002888336291896123,
+      "loss": 3.2191,
+      "step": 10957
+    },
+    {
+      "epoch": 0.4613894736842105,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00028880029414450815,
+      "loss": 2.9675,
+      "step": 10958
+    },
+    {
+      "epoch": 0.4614315789473684,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00028876695839250824,
+      "loss": 3.4638,
+      "step": 10959
+    },
+    {
+      "epoch": 0.4614736842105263,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00028873362193421984,
+      "loss": 2.9631,
+      "step": 10960
+    },
+    {
+      "epoch": 0.4615157894736842,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002887002847702504,
+      "loss": 3.2601,
+      "step": 10961
+    },
+    {
+      "epoch": 0.4615578947368421,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00028866694690120716,
+      "loss": 3.5154,
+      "step": 10962
+    },
+    {
+      "epoch": 0.4616,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00028863360832769757,
+      "loss": 3.104,
+      "step": 10963
+    },
+    {
+      "epoch": 0.4616421052631579,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00028860026905032913,
+      "loss": 3.4063,
+      "step": 10964
+    },
+    {
+      "epoch": 0.4616842105263158,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00028856692906970916,
+      "loss": 3.114,
+      "step": 10965
+    },
+    {
+      "epoch": 0.4617263157894737,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.000288533588386445,
+      "loss": 2.7107,
+      "step": 10966
+    },
+    {
+      "epoch": 0.4617684210526316,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00028850024700114417,
+      "loss": 2.871,
+      "step": 10967
+    },
+    {
+      "epoch": 0.46181052631578945,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002884669049144141,
+      "loss": 3.3199,
+      "step": 10968
+    },
+    {
+      "epoch": 0.46185263157894735,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0002884335621268621,
+      "loss": 3.0826,
+      "step": 10969
+    },
+    {
+      "epoch": 0.46189473684210525,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00028840021863909583,
+      "loss": 3.0368,
+      "step": 10970
+    },
+    {
+      "epoch": 0.46193684210526315,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00028836687445172264,
+      "loss": 3.1435,
+      "step": 10971
+    },
+    {
+      "epoch": 0.46197894736842104,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002883335295653501,
+      "loss": 3.3272,
+      "step": 10972
+    },
+    {
+      "epoch": 0.46202105263157894,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00028830018398058577,
+      "loss": 3.5651,
+      "step": 10973
+    },
+    {
+      "epoch": 0.46206315789473684,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002882668376980371,
+      "loss": 3.0045,
+      "step": 10974
+    },
+    {
+      "epoch": 0.46210526315789474,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002882334907183115,
+      "loss": 3.0445,
+      "step": 10975
+    },
+    {
+      "epoch": 0.46214736842105264,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002882001430420166,
+      "loss": 3.1384,
+      "step": 10976
+    },
+    {
+      "epoch": 0.46218947368421054,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002881667946697601,
+      "loss": 3.1792,
+      "step": 10977
+    },
+    {
+      "epoch": 0.46223157894736844,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002881334456021494,
+      "loss": 3.0474,
+      "step": 10978
+    },
+    {
+      "epoch": 0.46227368421052634,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00028810009583979206,
+      "loss": 3.3346,
+      "step": 10979
+    },
+    {
+      "epoch": 0.46231578947368424,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002880667453832958,
+      "loss": 2.6771,
+      "step": 10980
+    },
+    {
+      "epoch": 0.4623578947368421,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002880333942332682,
+      "loss": 3.0706,
+      "step": 10981
+    },
+    {
+      "epoch": 0.4624,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00028800004239031685,
+      "loss": 3.3357,
+      "step": 10982
+    },
+    {
+      "epoch": 0.4624421052631579,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00028796668985504924,
+      "loss": 2.5734,
+      "step": 10983
+    },
+    {
+      "epoch": 0.4624842105263158,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002879333366280733,
+      "loss": 3.1539,
+      "step": 10984
+    },
+    {
+      "epoch": 0.4625263157894737,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00028789998270999643,
+      "loss": 3.2092,
+      "step": 10985
+    },
+    {
+      "epoch": 0.4625684210526316,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002878666281014265,
+      "loss": 3.2848,
+      "step": 10986
+    },
+    {
+      "epoch": 0.46261052631578947,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.000287833272802971,
+      "loss": 3.0889,
+      "step": 10987
+    },
+    {
+      "epoch": 0.46265263157894737,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002877999168152378,
+      "loss": 3.4342,
+      "step": 10988
+    },
+    {
+      "epoch": 0.46269473684210527,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00028776656013883447,
+      "loss": 3.1307,
+      "step": 10989
+    },
+    {
+      "epoch": 0.46273684210526317,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00028773320277436883,
+      "loss": 3.1539,
+      "step": 10990
+    },
+    {
+      "epoch": 0.46277894736842107,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002876998447224486,
+      "loss": 3.0921,
+      "step": 10991
+    },
+    {
+      "epoch": 0.46282105263157897,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00028766648598368147,
+      "loss": 3.2349,
+      "step": 10992
+    },
+    {
+      "epoch": 0.46286315789473687,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002876331265586752,
+      "loss": 3.487,
+      "step": 10993
+    },
+    {
+      "epoch": 0.46290526315789476,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002875997664480376,
+      "loss": 3.1138,
+      "step": 10994
+    },
+    {
+      "epoch": 0.4629473684210526,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00028756640565237633,
+      "loss": 3.7036,
+      "step": 10995
+    },
+    {
+      "epoch": 0.4629894736842105,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002875330441722994,
+      "loss": 3.1889,
+      "step": 10996
+    },
+    {
+      "epoch": 0.4630315789473684,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00028749968200841446,
+      "loss": 3.1696,
+      "step": 10997
+    },
+    {
+      "epoch": 0.4630736842105263,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002874663191613294,
+      "loss": 3.0466,
+      "step": 10998
+    },
+    {
+      "epoch": 0.4631157894736842,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00028743295563165205,
+      "loss": 2.8668,
+      "step": 10999
+    },
+    {
+      "epoch": 0.4631578947368421,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002873995914199902,
+      "loss": 3.1494,
+      "step": 11000
+    },
+    {
+      "epoch": 0.4632,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00028736622652695176,
+      "loss": 3.2486,
+      "step": 11001
+    },
+    {
+      "epoch": 0.4632421052631579,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002873328609531446,
+      "loss": 3.0619,
+      "step": 11002
+    },
+    {
+      "epoch": 0.4632842105263158,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002872994946991765,
+      "loss": 2.9887,
+      "step": 11003
+    },
+    {
+      "epoch": 0.4633263157894737,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002872661277656555,
+      "loss": 3.2071,
+      "step": 11004
+    },
+    {
+      "epoch": 0.4633684210526316,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002872327601531893,
+      "loss": 2.7406,
+      "step": 11005
+    },
+    {
+      "epoch": 0.4634105263157895,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002871993918623861,
+      "loss": 3.0966,
+      "step": 11006
+    },
+    {
+      "epoch": 0.4634526315789474,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00028716602289385364,
+      "loss": 3.284,
+      "step": 11007
+    },
+    {
+      "epoch": 0.46349473684210524,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00028713265324819995,
+      "loss": 3.2656,
+      "step": 11008
+    },
+    {
+      "epoch": 0.46353684210526314,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002870992829260329,
+      "loss": 3.0845,
+      "step": 11009
+    },
+    {
+      "epoch": 0.46357894736842103,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002870659119279605,
+      "loss": 3.3118,
+      "step": 11010
+    },
+    {
+      "epoch": 0.46362105263157893,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002870325402545908,
+      "loss": 2.9708,
+      "step": 11011
+    },
+    {
+      "epoch": 0.46366315789473683,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002869991679065317,
+      "loss": 3.1585,
+      "step": 11012
+    },
+    {
+      "epoch": 0.46370526315789473,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00028696579488439114,
+      "loss": 3.1989,
+      "step": 11013
+    },
+    {
+      "epoch": 0.46374736842105263,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002869324211887773,
+      "loss": 3.3521,
+      "step": 11014
+    },
+    {
+      "epoch": 0.46378947368421053,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002868990468202982,
+      "loss": 2.7115,
+      "step": 11015
+    },
+    {
+      "epoch": 0.46383157894736843,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002868656717795617,
+      "loss": 3.1267,
+      "step": 11016
+    },
+    {
+      "epoch": 0.4638736842105263,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002868322960671761,
+      "loss": 3.3949,
+      "step": 11017
+    },
+    {
+      "epoch": 0.4639157894736842,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002867989196837493,
+      "loss": 3.2507,
+      "step": 11018
+    },
+    {
+      "epoch": 0.4639578947368421,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002867655426298894,
+      "loss": 3.4806,
+      "step": 11019
+    },
+    {
+      "epoch": 0.464,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002867321649062045,
+      "loss": 3.3975,
+      "step": 11020
+    },
+    {
+      "epoch": 0.4640421052631579,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002866987865133027,
+      "loss": 3.2164,
+      "step": 11021
+    },
+    {
+      "epoch": 0.46408421052631577,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002866654074517922,
+      "loss": 3.0578,
+      "step": 11022
+    },
+    {
+      "epoch": 0.46412631578947366,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00028663202772228095,
+      "loss": 3.2089,
+      "step": 11023
+    },
+    {
+      "epoch": 0.46416842105263156,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00028659864732537727,
+      "loss": 3.5458,
+      "step": 11024
+    },
+    {
+      "epoch": 0.46421052631578946,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00028656526626168926,
+      "loss": 3.1491,
+      "step": 11025
+    },
+    {
+      "epoch": 0.46425263157894736,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00028653188453182503,
+      "loss": 3.1083,
+      "step": 11026
+    },
+    {
+      "epoch": 0.46429473684210526,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00028649850213639284,
+      "loss": 2.7537,
+      "step": 11027
+    },
+    {
+      "epoch": 0.46433684210526316,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00028646511907600076,
+      "loss": 3.3336,
+      "step": 11028
+    },
+    {
+      "epoch": 0.46437894736842106,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00028643173535125707,
+      "loss": 3.2345,
+      "step": 11029
+    },
+    {
+      "epoch": 0.46442105263157896,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002863983509627699,
+      "loss": 3.2096,
+      "step": 11030
+    },
+    {
+      "epoch": 0.46446315789473686,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00028636496591114756,
+      "loss": 3.0876,
+      "step": 11031
+    },
+    {
+      "epoch": 0.46450526315789475,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002863315801969984,
+      "loss": 3.0769,
+      "step": 11032
+    },
+    {
+      "epoch": 0.46454736842105265,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00028629819382093043,
+      "loss": 3.7223,
+      "step": 11033
+    },
+    {
+      "epoch": 0.46458947368421055,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00028626480678355207,
+      "loss": 3.6163,
+      "step": 11034
+    },
+    {
+      "epoch": 0.4646315789473684,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00028623141908547153,
+      "loss": 3.3951,
+      "step": 11035
+    },
+    {
+      "epoch": 0.4646736842105263,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00028619803072729705,
+      "loss": 2.9789,
+      "step": 11036
+    },
+    {
+      "epoch": 0.4647157894736842,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.000286164641709637,
+      "loss": 3.3617,
+      "step": 11037
+    },
+    {
+      "epoch": 0.4647578947368421,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00028613125203309965,
+      "loss": 4.0109,
+      "step": 11038
+    },
+    {
+      "epoch": 0.4648,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00028609786169829336,
+      "loss": 2.8828,
+      "step": 11039
+    },
+    {
+      "epoch": 0.4648421052631579,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002860644707058265,
+      "loss": 3.1537,
+      "step": 11040
+    },
+    {
+      "epoch": 0.4648842105263158,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002860310790563073,
+      "loss": 3.1394,
+      "step": 11041
+    },
+    {
+      "epoch": 0.4649263157894737,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002859976867503442,
+      "loss": 3.222,
+      "step": 11042
+    },
+    {
+      "epoch": 0.4649684210526316,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00028596429378854556,
+      "loss": 3.1944,
+      "step": 11043
+    },
+    {
+      "epoch": 0.4650105263157895,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002859309001715197,
+      "loss": 3.3569,
+      "step": 11044
+    },
+    {
+      "epoch": 0.4650526315789474,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00028589750589987507,
+      "loss": 3.0646,
+      "step": 11045
+    },
+    {
+      "epoch": 0.4650947368421053,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00028586411097422003,
+      "loss": 3.0007,
+      "step": 11046
+    },
+    {
+      "epoch": 0.4651368421052632,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002858307153951631,
+      "loss": 3.1132,
+      "step": 11047
+    },
+    {
+      "epoch": 0.465178947368421,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002857973191633126,
+      "loss": 2.786,
+      "step": 11048
+    },
+    {
+      "epoch": 0.4652210526315789,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.000285763922279277,
+      "loss": 2.9335,
+      "step": 11049
+    },
+    {
+      "epoch": 0.4652631578947368,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00028573052474366474,
+      "loss": 3.4274,
+      "step": 11050
+    },
+    {
+      "epoch": 0.4653052631578947,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002856971265570843,
+      "loss": 3.6876,
+      "step": 11051
+    },
+    {
+      "epoch": 0.4653473684210526,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002856637277201442,
+      "loss": 3.0635,
+      "step": 11052
+    },
+    {
+      "epoch": 0.4653894736842105,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002856303282334529,
+      "loss": 3.123,
+      "step": 11053
+    },
+    {
+      "epoch": 0.4654315789473684,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00028559692809761877,
+      "loss": 3.3445,
+      "step": 11054
+    },
+    {
+      "epoch": 0.4654736842105263,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00028556352731325054,
+      "loss": 3.2822,
+      "step": 11055
+    },
+    {
+      "epoch": 0.4655157894736842,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002855301258809566,
+      "loss": 3.2106,
+      "step": 11056
+    },
+    {
+      "epoch": 0.4655578947368421,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00028549672380134546,
+      "loss": 3.3361,
+      "step": 11057
+    },
+    {
+      "epoch": 0.4656,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00028546332107502574,
+      "loss": 3.286,
+      "step": 11058
+    },
+    {
+      "epoch": 0.4656421052631579,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002854299177026061,
+      "loss": 3.0591,
+      "step": 11059
+    },
+    {
+      "epoch": 0.4656842105263158,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002853965136846948,
+      "loss": 3.2394,
+      "step": 11060
+    },
+    {
+      "epoch": 0.4657263157894737,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002853631090219008,
+      "loss": 2.9695,
+      "step": 11061
+    },
+    {
+      "epoch": 0.46576842105263155,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002853297037148323,
+      "loss": 3.101,
+      "step": 11062
+    },
+    {
+      "epoch": 0.46581052631578945,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00028529629776409825,
+      "loss": 3.5741,
+      "step": 11063
+    },
+    {
+      "epoch": 0.46585263157894735,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002852628911703071,
+      "loss": 3.1201,
+      "step": 11064
+    },
+    {
+      "epoch": 0.46589473684210525,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002852294839340674,
+      "loss": 2.9888,
+      "step": 11065
+    },
+    {
+      "epoch": 0.46593684210526315,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002851960760559881,
+      "loss": 3.2819,
+      "step": 11066
+    },
+    {
+      "epoch": 0.46597894736842105,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002851626675366775,
+      "loss": 3.3095,
+      "step": 11067
+    },
+    {
+      "epoch": 0.46602105263157895,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002851292583767445,
+      "loss": 3.3439,
+      "step": 11068
+    },
+    {
+      "epoch": 0.46606315789473685,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002850958485767976,
+      "loss": 3.1526,
+      "step": 11069
+    },
+    {
+      "epoch": 0.46610526315789474,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002850624381374456,
+      "loss": 3.1498,
+      "step": 11070
+    },
+    {
+      "epoch": 0.46614736842105264,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00028502902705929725,
+      "loss": 3.1174,
+      "step": 11071
+    },
+    {
+      "epoch": 0.46618947368421054,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.000284995615342961,
+      "loss": 3.1359,
+      "step": 11072
+    },
+    {
+      "epoch": 0.46623157894736844,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002849622029890459,
+      "loss": 2.6997,
+      "step": 11073
+    },
+    {
+      "epoch": 0.46627368421052634,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002849287899981606,
+      "loss": 2.4903,
+      "step": 11074
+    },
+    {
+      "epoch": 0.4663157894736842,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00028489537637091367,
+      "loss": 3.2007,
+      "step": 11075
+    },
+    {
+      "epoch": 0.4663578947368421,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00028486196210791403,
+      "loss": 3.5014,
+      "step": 11076
+    },
+    {
+      "epoch": 0.4664,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002848285472097704,
+      "loss": 3.1761,
+      "step": 11077
+    },
+    {
+      "epoch": 0.4664421052631579,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00028479513167709155,
+      "loss": 3.4365,
+      "step": 11078
+    },
+    {
+      "epoch": 0.4664842105263158,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002847617155104862,
+      "loss": 3.5588,
+      "step": 11079
+    },
+    {
+      "epoch": 0.4665263157894737,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002847282987105633,
+      "loss": 3.3281,
+      "step": 11080
+    },
+    {
+      "epoch": 0.4665684210526316,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002846948812779317,
+      "loss": 2.904,
+      "step": 11081
+    },
+    {
+      "epoch": 0.4666105263157895,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002846614632132,
+      "loss": 3.5838,
+      "step": 11082
+    },
+    {
+      "epoch": 0.4666526315789474,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002846280445169772,
+      "loss": 2.8639,
+      "step": 11083
+    },
+    {
+      "epoch": 0.4666947368421053,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002845946251898722,
+      "loss": 3.0193,
+      "step": 11084
+    },
+    {
+      "epoch": 0.46673684210526317,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00028456120523249367,
+      "loss": 3.4019,
+      "step": 11085
+    },
+    {
+      "epoch": 0.46677894736842107,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00028452778464545063,
+      "loss": 2.9672,
+      "step": 11086
+    },
+    {
+      "epoch": 0.46682105263157897,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.000284494363429352,
+      "loss": 3.1525,
+      "step": 11087
+    },
+    {
+      "epoch": 0.46686315789473687,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002844609415848064,
+      "loss": 3.7454,
+      "step": 11088
+    },
+    {
+      "epoch": 0.4669052631578947,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002844275191124231,
+      "loss": 3.264,
+      "step": 11089
+    },
+    {
+      "epoch": 0.4669473684210526,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00028439409601281075,
+      "loss": 3.7794,
+      "step": 11090
+    },
+    {
+      "epoch": 0.4669894736842105,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00028436067228657845,
+      "loss": 3.3333,
+      "step": 11091
+    },
+    {
+      "epoch": 0.4670315789473684,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002843272479343351,
+      "loss": 3.0639,
+      "step": 11092
+    },
+    {
+      "epoch": 0.4670736842105263,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002842938229566895,
+      "loss": 3.3473,
+      "step": 11093
+    },
+    {
+      "epoch": 0.4671157894736842,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002842603973542509,
+      "loss": 3.4347,
+      "step": 11094
+    },
+    {
+      "epoch": 0.4671578947368421,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.000284226971127628,
+      "loss": 3.0985,
+      "step": 11095
+    },
+    {
+      "epoch": 0.4672,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00028419354427742995,
+      "loss": 3.1642,
+      "step": 11096
+    },
+    {
+      "epoch": 0.4672421052631579,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00028416011680426563,
+      "loss": 2.8935,
+      "step": 11097
+    },
+    {
+      "epoch": 0.4672842105263158,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002841266887087442,
+      "loss": 3.3013,
+      "step": 11098
+    },
+    {
+      "epoch": 0.4673263157894737,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00028409325999147455,
+      "loss": 2.9669,
+      "step": 11099
+    },
+    {
+      "epoch": 0.4673684210526316,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00028405983065306583,
+      "loss": 2.8304,
+      "step": 11100
+    },
+    {
+      "epoch": 0.4674105263157895,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.000284026400694127,
+      "loss": 3.3224,
+      "step": 11101
+    },
+    {
+      "epoch": 0.46745263157894734,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002839929701152671,
+      "loss": 3.4339,
+      "step": 11102
+    },
+    {
+      "epoch": 0.46749473684210524,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00028395953891709527,
+      "loss": 3.136,
+      "step": 11103
+    },
+    {
+      "epoch": 0.46753684210526314,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002839261071002204,
+      "loss": 3.067,
+      "step": 11104
+    },
+    {
+      "epoch": 0.46757894736842104,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002838926746652518,
+      "loss": 3.2678,
+      "step": 11105
+    },
+    {
+      "epoch": 0.46762105263157894,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00028385924161279853,
+      "loss": 3.1551,
+      "step": 11106
+    },
+    {
+      "epoch": 0.46766315789473684,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00028382580794346964,
+      "loss": 3.562,
+      "step": 11107
+    },
+    {
+      "epoch": 0.46770526315789474,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002837923736578742,
+      "loss": 3.3161,
+      "step": 11108
+    },
+    {
+      "epoch": 0.46774736842105263,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00028375893875662154,
+      "loss": 3.3308,
+      "step": 11109
+    },
+    {
+      "epoch": 0.46778947368421053,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002837255032403206,
+      "loss": 3.0422,
+      "step": 11110
+    },
+    {
+      "epoch": 0.46783157894736843,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00028369206710958064,
+      "loss": 3.3357,
+      "step": 11111
+    },
+    {
+      "epoch": 0.46787368421052633,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00028365863036501084,
+      "loss": 2.9283,
+      "step": 11112
+    },
+    {
+      "epoch": 0.46791578947368423,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002836251930072203,
+      "loss": 3.2735,
+      "step": 11113
+    },
+    {
+      "epoch": 0.46795789473684213,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002835917550368182,
+      "loss": 3.0573,
+      "step": 11114
+    },
+    {
+      "epoch": 0.468,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002835583164544139,
+      "loss": 3.2945,
+      "step": 11115
+    },
+    {
+      "epoch": 0.46804210526315787,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002835248772606165,
+      "loss": 3.3422,
+      "step": 11116
+    },
+    {
+      "epoch": 0.46808421052631577,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00028349143745603513,
+      "loss": 3.2905,
+      "step": 11117
+    },
+    {
+      "epoch": 0.46812631578947367,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00028345799704127916,
+      "loss": 3.8991,
+      "step": 11118
+    },
+    {
+      "epoch": 0.46816842105263157,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00028342455601695785,
+      "loss": 3.3379,
+      "step": 11119
+    },
+    {
+      "epoch": 0.46821052631578947,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00028339111438368034,
+      "loss": 2.9561,
+      "step": 11120
+    },
+    {
+      "epoch": 0.46825263157894736,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00028335767214205597,
+      "loss": 3.0417,
+      "step": 11121
+    },
+    {
+      "epoch": 0.46829473684210526,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.000283324229292694,
+      "loss": 3.5106,
+      "step": 11122
+    },
+    {
+      "epoch": 0.46833684210526316,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00028329078583620373,
+      "loss": 3.2043,
+      "step": 11123
+    },
+    {
+      "epoch": 0.46837894736842106,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00028325734177319446,
+      "loss": 3.1109,
+      "step": 11124
+    },
+    {
+      "epoch": 0.46842105263157896,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0002832238971042755,
+      "loss": 3.38,
+      "step": 11125
+    },
+    {
+      "epoch": 0.46846315789473686,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00028319045183005623,
+      "loss": 3.1454,
+      "step": 11126
+    },
+    {
+      "epoch": 0.46850526315789476,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00028315700595114585,
+      "loss": 3.8016,
+      "step": 11127
+    },
+    {
+      "epoch": 0.46854736842105266,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002831235594681538,
+      "loss": 3.1104,
+      "step": 11128
+    },
+    {
+      "epoch": 0.4685894736842105,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00028309011238168946,
+      "loss": 3.2327,
+      "step": 11129
+    },
+    {
+      "epoch": 0.4686315789473684,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00028305666469236203,
+      "loss": 2.708,
+      "step": 11130
+    },
+    {
+      "epoch": 0.4686736842105263,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00028302321640078106,
+      "loss": 3.0652,
+      "step": 11131
+    },
+    {
+      "epoch": 0.4687157894736842,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.000282989767507556,
+      "loss": 3.0473,
+      "step": 11132
+    },
+    {
+      "epoch": 0.4687578947368421,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00028295631801329604,
+      "loss": 3.4193,
+      "step": 11133
+    },
+    {
+      "epoch": 0.4688,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00028292286791861064,
+      "loss": 3.3724,
+      "step": 11134
+    },
+    {
+      "epoch": 0.4688421052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00028288941722410935,
+      "loss": 3.113,
+      "step": 11135
+    },
+    {
+      "epoch": 0.4688842105263158,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0002828559659304015,
+      "loss": 3.3249,
+      "step": 11136
+    },
+    {
+      "epoch": 0.4689263157894737,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00028282251403809655,
+      "loss": 3.0884,
+      "step": 11137
+    },
+    {
+      "epoch": 0.4689684210526316,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.000282789061547804,
+      "loss": 3.3311,
+      "step": 11138
+    },
+    {
+      "epoch": 0.4690105263157895,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00028275560846013323,
+      "loss": 3.0631,
+      "step": 11139
+    },
+    {
+      "epoch": 0.4690526315789474,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00028272215477569373,
+      "loss": 3.1261,
+      "step": 11140
+    },
+    {
+      "epoch": 0.4690947368421053,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002826887004950951,
+      "loss": 3.1957,
+      "step": 11141
+    },
+    {
+      "epoch": 0.4691368421052632,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002826552456189467,
+      "loss": 3.0753,
+      "step": 11142
+    },
+    {
+      "epoch": 0.46917894736842103,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002826217901478581,
+      "loss": 3.0499,
+      "step": 11143
+    },
+    {
+      "epoch": 0.4692210526315789,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00028258833408243875,
+      "loss": 3.5921,
+      "step": 11144
+    },
+    {
+      "epoch": 0.4692631578947368,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00028255487742329837,
+      "loss": 3.0914,
+      "step": 11145
+    },
+    {
+      "epoch": 0.4693052631578947,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002825214201710463,
+      "loss": 3.5112,
+      "step": 11146
+    },
+    {
+      "epoch": 0.4693473684210526,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00028248796232629216,
+      "loss": 3.1328,
+      "step": 11147
+    },
+    {
+      "epoch": 0.4693894736842105,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002824545038896456,
+      "loss": 3.3992,
+      "step": 11148
+    },
+    {
+      "epoch": 0.4694315789473684,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00028242104486171605,
+      "loss": 3.2938,
+      "step": 11149
+    },
+    {
+      "epoch": 0.4694736842105263,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00028238758524311313,
+      "loss": 3.1996,
+      "step": 11150
+    },
+    {
+      "epoch": 0.4695157894736842,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00028235412503444646,
+      "loss": 3.1004,
+      "step": 11151
+    },
+    {
+      "epoch": 0.4695578947368421,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002823206642363257,
+      "loss": 3.4519,
+      "step": 11152
+    },
+    {
+      "epoch": 0.4696,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002822872028493604,
+      "loss": 3.4161,
+      "step": 11153
+    },
+    {
+      "epoch": 0.4696421052631579,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002822537408741602,
+      "loss": 3.5683,
+      "step": 11154
+    },
+    {
+      "epoch": 0.4696842105263158,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00028222027831133466,
+      "loss": 3.3038,
+      "step": 11155
+    },
+    {
+      "epoch": 0.46972631578947366,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00028218681516149356,
+      "loss": 3.1519,
+      "step": 11156
+    },
+    {
+      "epoch": 0.46976842105263156,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002821533514252466,
+      "loss": 3.2105,
+      "step": 11157
+    },
+    {
+      "epoch": 0.46981052631578946,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002821198871032032,
+      "loss": 3.2832,
+      "step": 11158
+    },
+    {
+      "epoch": 0.46985263157894736,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00028208642219597325,
+      "loss": 3.3814,
+      "step": 11159
+    },
+    {
+      "epoch": 0.46989473684210525,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00028205295670416644,
+      "loss": 3.0818,
+      "step": 11160
+    },
+    {
+      "epoch": 0.46993684210526315,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002820194906283924,
+      "loss": 3.2789,
+      "step": 11161
+    },
+    {
+      "epoch": 0.46997894736842105,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002819860239692609,
+      "loss": 3.1111,
+      "step": 11162
+    },
+    {
+      "epoch": 0.47002105263157895,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002819525567273815,
+      "loss": 3.519,
+      "step": 11163
+    },
+    {
+      "epoch": 0.47006315789473685,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00028191908890336416,
+      "loss": 3.4117,
+      "step": 11164
+    },
+    {
+      "epoch": 0.47010526315789475,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002818856204978184,
+      "loss": 3.4008,
+      "step": 11165
+    },
+    {
+      "epoch": 0.47014736842105265,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002818521515113542,
+      "loss": 3.3098,
+      "step": 11166
+    },
+    {
+      "epoch": 0.47018947368421055,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002818186819445813,
+      "loss": 3.1341,
+      "step": 11167
+    },
+    {
+      "epoch": 0.47023157894736844,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00028178521179810927,
+      "loss": 3.1343,
+      "step": 11168
+    },
+    {
+      "epoch": 0.4702736842105263,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00028175174107254806,
+      "loss": 3.2277,
+      "step": 11169
+    },
+    {
+      "epoch": 0.4703157894736842,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00028171826976850753,
+      "loss": 3.4406,
+      "step": 11170
+    },
+    {
+      "epoch": 0.4703578947368421,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002816847978865972,
+      "loss": 3.0975,
+      "step": 11171
+    },
+    {
+      "epoch": 0.4704,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00028165132542742725,
+      "loss": 3.553,
+      "step": 11172
+    },
+    {
+      "epoch": 0.4704421052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00028161785239160723,
+      "loss": 3.2445,
+      "step": 11173
+    },
+    {
+      "epoch": 0.4704842105263158,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002815843787797472,
+      "loss": 3.0592,
+      "step": 11174
+    },
+    {
+      "epoch": 0.4705263157894737,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002815509045924568,
+      "loss": 3.1512,
+      "step": 11175
+    },
+    {
+      "epoch": 0.4705684210526316,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002815174298303461,
+      "loss": 3.6098,
+      "step": 11176
+    },
+    {
+      "epoch": 0.4706105263157895,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00028148395449402474,
+      "loss": 3.4264,
+      "step": 11177
+    },
+    {
+      "epoch": 0.4706526315789474,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002814504785841028,
+      "loss": 2.8674,
+      "step": 11178
+    },
+    {
+      "epoch": 0.4706947368421053,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00028141700210119005,
+      "loss": 3.0892,
+      "step": 11179
+    },
+    {
+      "epoch": 0.4707368421052632,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002813835250458964,
+      "loss": 3.4468,
+      "step": 11180
+    },
+    {
+      "epoch": 0.4707789473684211,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00028135004741883186,
+      "loss": 3.2925,
+      "step": 11181
+    },
+    {
+      "epoch": 0.470821052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002813165692206063,
+      "loss": 3.2443,
+      "step": 11182
+    },
+    {
+      "epoch": 0.4708631578947368,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002812830904518296,
+      "loss": 3.084,
+      "step": 11183
+    },
+    {
+      "epoch": 0.4709052631578947,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00028124961111311174,
+      "loss": 3.6488,
+      "step": 11184
+    },
+    {
+      "epoch": 0.4709473684210526,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00028121613120506275,
+      "loss": 3.1927,
+      "step": 11185
+    },
+    {
+      "epoch": 0.4709894736842105,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002811826507282925,
+      "loss": 3.3869,
+      "step": 11186
+    },
+    {
+      "epoch": 0.4710315789473684,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.000281149169683411,
+      "loss": 3.2963,
+      "step": 11187
+    },
+    {
+      "epoch": 0.4710736842105263,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0002811156880710282,
+      "loss": 3.0905,
+      "step": 11188
+    },
+    {
+      "epoch": 0.4711157894736842,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002810822058917541,
+      "loss": 3.5734,
+      "step": 11189
+    },
+    {
+      "epoch": 0.4711578947368421,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00028104872314619873,
+      "loss": 3.2389,
+      "step": 11190
+    },
+    {
+      "epoch": 0.4712,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002810152398349721,
+      "loss": 3.1316,
+      "step": 11191
+    },
+    {
+      "epoch": 0.4712421052631579,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002809817559586843,
+      "loss": 2.9237,
+      "step": 11192
+    },
+    {
+      "epoch": 0.4712842105263158,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002809482715179453,
+      "loss": 3.2665,
+      "step": 11193
+    },
+    {
+      "epoch": 0.4713263157894737,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002809147865133652,
+      "loss": 3.4429,
+      "step": 11194
+    },
+    {
+      "epoch": 0.4713684210526316,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00028088130094555393,
+      "loss": 3.2462,
+      "step": 11195
+    },
+    {
+      "epoch": 0.47141052631578945,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00028084781481512165,
+      "loss": 2.5726,
+      "step": 11196
+    },
+    {
+      "epoch": 0.47145263157894735,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00028081432812267844,
+      "loss": 3.3107,
+      "step": 11197
+    },
+    {
+      "epoch": 0.47149473684210524,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002807808408688343,
+      "loss": 3.199,
+      "step": 11198
+    },
+    {
+      "epoch": 0.47153684210526314,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002807473530541995,
+      "loss": 3.2169,
+      "step": 11199
+    },
+    {
+      "epoch": 0.47157894736842104,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.000280713864679384,
+      "loss": 3.3756,
+      "step": 11200
+    },
+    {
+      "epoch": 0.47162105263157894,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00028068037574499805,
+      "loss": 2.849,
+      "step": 11201
+    },
+    {
+      "epoch": 0.47166315789473684,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00028064688625165167,
+      "loss": 3.1861,
+      "step": 11202
+    },
+    {
+      "epoch": 0.47170526315789474,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002806133961999551,
+      "loss": 3.4251,
+      "step": 11203
+    },
+    {
+      "epoch": 0.47174736842105264,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0002805799055905183,
+      "loss": 3.0857,
+      "step": 11204
+    },
+    {
+      "epoch": 0.47178947368421054,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00028054641442395154,
+      "loss": 2.8668,
+      "step": 11205
+    },
+    {
+      "epoch": 0.47183157894736844,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00028051292270086505,
+      "loss": 3.4576,
+      "step": 11206
+    },
+    {
+      "epoch": 0.47187368421052633,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.000280479430421869,
+      "loss": 3.1642,
+      "step": 11207
+    },
+    {
+      "epoch": 0.47191578947368423,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00028044593758757354,
+      "loss": 2.6202,
+      "step": 11208
+    },
+    {
+      "epoch": 0.47195789473684213,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002804124441985888,
+      "loss": 3.0581,
+      "step": 11209
+    },
+    {
+      "epoch": 0.472,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002803789502555251,
+      "loss": 2.8977,
+      "step": 11210
+    },
+    {
+      "epoch": 0.4720421052631579,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00028034545575899254,
+      "loss": 3.1538,
+      "step": 11211
+    },
+    {
+      "epoch": 0.4720842105263158,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002803119607096016,
+      "loss": 3.2904,
+      "step": 11212
+    },
+    {
+      "epoch": 0.47212631578947367,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00028027846510796225,
+      "loss": 3.2363,
+      "step": 11213
+    },
+    {
+      "epoch": 0.47216842105263157,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00028024496895468474,
+      "loss": 3.1192,
+      "step": 11214
+    },
+    {
+      "epoch": 0.47221052631578947,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002802114722503796,
+      "loss": 2.9258,
+      "step": 11215
+    },
+    {
+      "epoch": 0.47225263157894737,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002801779749956569,
+      "loss": 3.3017,
+      "step": 11216
+    },
+    {
+      "epoch": 0.47229473684210527,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002801444771911269,
+      "loss": 3.1346,
+      "step": 11217
+    },
+    {
+      "epoch": 0.47233684210526317,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00028011097883739994,
+      "loss": 3.2025,
+      "step": 11218
+    },
+    {
+      "epoch": 0.47237894736842106,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00028007747993508635,
+      "loss": 3.4768,
+      "step": 11219
+    },
+    {
+      "epoch": 0.47242105263157896,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00028004398048479644,
+      "loss": 2.9053,
+      "step": 11220
+    },
+    {
+      "epoch": 0.47246315789473686,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002800104804871405,
+      "loss": 2.9437,
+      "step": 11221
+    },
+    {
+      "epoch": 0.47250526315789476,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00027997697994272883,
+      "loss": 3.2862,
+      "step": 11222
+    },
+    {
+      "epoch": 0.4725473684210526,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002799434788521718,
+      "loss": 3.3606,
+      "step": 11223
+    },
+    {
+      "epoch": 0.4725894736842105,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002799099772160798,
+      "loss": 3.2482,
+      "step": 11224
+    },
+    {
+      "epoch": 0.4726315789473684,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0002798764750350631,
+      "loss": 2.9727,
+      "step": 11225
+    },
+    {
+      "epoch": 0.4726736842105263,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002798429723097322,
+      "loss": 2.7849,
+      "step": 11226
+    },
+    {
+      "epoch": 0.4727157894736842,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00027980946904069745,
+      "loss": 3.3083,
+      "step": 11227
+    },
+    {
+      "epoch": 0.4727578947368421,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002797759652285691,
+      "loss": 2.9202,
+      "step": 11228
+    },
+    {
+      "epoch": 0.4728,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00027974246087395774,
+      "loss": 2.8535,
+      "step": 11229
+    },
+    {
+      "epoch": 0.4728421052631579,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002797089559774736,
+      "loss": 2.8304,
+      "step": 11230
+    },
+    {
+      "epoch": 0.4728842105263158,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002796754505397272,
+      "loss": 2.9279,
+      "step": 11231
+    },
+    {
+      "epoch": 0.4729263157894737,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00027964194456132896,
+      "loss": 3.1045,
+      "step": 11232
+    },
+    {
+      "epoch": 0.4729684210526316,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002796084380428894,
+      "loss": 2.9357,
+      "step": 11233
+    },
+    {
+      "epoch": 0.4730105263157895,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002795749309850188,
+      "loss": 3.2696,
+      "step": 11234
+    },
+    {
+      "epoch": 0.4730526315789474,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002795414233883278,
+      "loss": 3.3796,
+      "step": 11235
+    },
+    {
+      "epoch": 0.4730947368421053,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00027950791525342667,
+      "loss": 3.3979,
+      "step": 11236
+    },
+    {
+      "epoch": 0.47313684210526313,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00027947440658092603,
+      "loss": 3.1063,
+      "step": 11237
+    },
+    {
+      "epoch": 0.47317894736842103,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002794408973714363,
+      "loss": 3.7891,
+      "step": 11238
+    },
+    {
+      "epoch": 0.47322105263157893,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00027940738762556805,
+      "loss": 3.162,
+      "step": 11239
+    },
+    {
+      "epoch": 0.47326315789473683,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00027937387734393174,
+      "loss": 3.1884,
+      "step": 11240
+    },
+    {
+      "epoch": 0.47330526315789473,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002793403665271379,
+      "loss": 3.5732,
+      "step": 11241
+    },
+    {
+      "epoch": 0.47334736842105263,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002793068551757971,
+      "loss": 3.3219,
+      "step": 11242
+    },
+    {
+      "epoch": 0.4733894736842105,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0002792733432905198,
+      "loss": 2.9995,
+      "step": 11243
+    },
+    {
+      "epoch": 0.4734315789473684,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00027923983087191654,
+      "loss": 2.6831,
+      "step": 11244
+    },
+    {
+      "epoch": 0.4734736842105263,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.000279206317920598,
+      "loss": 2.9121,
+      "step": 11245
+    },
+    {
+      "epoch": 0.4735157894736842,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002791728044371746,
+      "loss": 2.9025,
+      "step": 11246
+    },
+    {
+      "epoch": 0.4735578947368421,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.000279139290422257,
+      "loss": 3.5811,
+      "step": 11247
+    },
+    {
+      "epoch": 0.4736,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00027910577587645567,
+      "loss": 3.4719,
+      "step": 11248
+    },
+    {
+      "epoch": 0.4736421052631579,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00027907226080038145,
+      "loss": 3.3567,
+      "step": 11249
+    },
+    {
+      "epoch": 0.47368421052631576,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002790387451946447,
+      "loss": 3.1155,
+      "step": 11250
+    },
+    {
+      "epoch": 0.47372631578947366,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00027900522905985624,
+      "loss": 2.2488,
+      "step": 11251
+    },
+    {
+      "epoch": 0.47376842105263156,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00027897171239662643,
+      "loss": 3.37,
+      "step": 11252
+    },
+    {
+      "epoch": 0.47381052631578946,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002789381952055662,
+      "loss": 3.3016,
+      "step": 11253
+    },
+    {
+      "epoch": 0.47385263157894736,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.000278904677487286,
+      "loss": 3.3227,
+      "step": 11254
+    },
+    {
+      "epoch": 0.47389473684210526,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002788711592423966,
+      "loss": 2.9876,
+      "step": 11255
+    },
+    {
+      "epoch": 0.47393684210526316,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00027883764047150844,
+      "loss": 2.6306,
+      "step": 11256
+    },
+    {
+      "epoch": 0.47397894736842106,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002788041211752325,
+      "loss": 2.464,
+      "step": 11257
+    },
+    {
+      "epoch": 0.47402105263157895,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002787706013541792,
+      "loss": 2.8837,
+      "step": 11258
+    },
+    {
+      "epoch": 0.47406315789473685,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00027873708100895945,
+      "loss": 3.2407,
+      "step": 11259
+    },
+    {
+      "epoch": 0.47410526315789475,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00027870356014018385,
+      "loss": 3.3901,
+      "step": 11260
+    },
+    {
+      "epoch": 0.47414736842105265,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002786700387484631,
+      "loss": 2.7232,
+      "step": 11261
+    },
+    {
+      "epoch": 0.47418947368421055,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00027863651683440787,
+      "loss": 2.9739,
+      "step": 11262
+    },
+    {
+      "epoch": 0.47423157894736845,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.000278602994398629,
+      "loss": 2.8542,
+      "step": 11263
+    },
+    {
+      "epoch": 0.4742736842105263,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0002785694714417371,
+      "loss": 2.9496,
+      "step": 11264
+    },
+    {
+      "epoch": 0.4743157894736842,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00027853594796434304,
+      "loss": 3.1237,
+      "step": 11265
+    },
+    {
+      "epoch": 0.4743578947368421,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002785024239670576,
+      "loss": 3.4626,
+      "step": 11266
+    },
+    {
+      "epoch": 0.4744,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00027846889945049144,
+      "loss": 3.5521,
+      "step": 11267
+    },
+    {
+      "epoch": 0.4744421052631579,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002784353744152554,
+      "loss": 3.1624,
+      "step": 11268
+    },
+    {
+      "epoch": 0.4744842105263158,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002784018488619602,
+      "loss": 3.1416,
+      "step": 11269
+    },
+    {
+      "epoch": 0.4745263157894737,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00027836832279121674,
+      "loss": 2.9493,
+      "step": 11270
+    },
+    {
+      "epoch": 0.4745684210526316,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002783347962036358,
+      "loss": 3.3596,
+      "step": 11271
+    },
+    {
+      "epoch": 0.4746105263157895,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00027830126909982804,
+      "loss": 2.7841,
+      "step": 11272
+    },
+    {
+      "epoch": 0.4746526315789474,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00027826774148040454,
+      "loss": 3.5661,
+      "step": 11273
+    },
+    {
+      "epoch": 0.4746947368421053,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00027823421334597585,
+      "loss": 3.0382,
+      "step": 11274
+    },
+    {
+      "epoch": 0.4747368421052632,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002782006846971531,
+      "loss": 2.6756,
+      "step": 11275
+    },
+    {
+      "epoch": 0.4747789473684211,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00027816715553454695,
+      "loss": 3.4888,
+      "step": 11276
+    },
+    {
+      "epoch": 0.4748210526315789,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0002781336258587684,
+      "loss": 2.9153,
+      "step": 11277
+    },
+    {
+      "epoch": 0.4748631578947368,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002781000956704282,
+      "loss": 3.0417,
+      "step": 11278
+    },
+    {
+      "epoch": 0.4749052631578947,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00027806656497013723,
+      "loss": 3.6996,
+      "step": 11279
+    },
+    {
+      "epoch": 0.4749473684210526,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0002780330337585065,
+      "loss": 3.3797,
+      "step": 11280
+    },
+    {
+      "epoch": 0.4749894736842105,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002779995020361468,
+      "loss": 3.0935,
+      "step": 11281
+    },
+    {
+      "epoch": 0.4750315789473684,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00027796596980366893,
+      "loss": 2.8673,
+      "step": 11282
+    },
+    {
+      "epoch": 0.4750736842105263,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00027793243706168415,
+      "loss": 3.3431,
+      "step": 11283
+    },
+    {
+      "epoch": 0.4751157894736842,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00027789890381080306,
+      "loss": 2.7922,
+      "step": 11284
+    },
+    {
+      "epoch": 0.4751578947368421,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00027786537005163676,
+      "loss": 3.5673,
+      "step": 11285
+    },
+    {
+      "epoch": 0.4752,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002778318357847962,
+      "loss": 2.8825,
+      "step": 11286
+    },
+    {
+      "epoch": 0.4752421052631579,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00027779830101089226,
+      "loss": 3.315,
+      "step": 11287
+    },
+    {
+      "epoch": 0.4752842105263158,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00027776476573053595,
+      "loss": 3.214,
+      "step": 11288
+    },
+    {
+      "epoch": 0.4753263157894737,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00027773122994433823,
+      "loss": 3.4841,
+      "step": 11289
+    },
+    {
+      "epoch": 0.47536842105263155,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00027769769365291005,
+      "loss": 3.425,
+      "step": 11290
+    },
+    {
+      "epoch": 0.47541052631578945,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00027766415685686246,
+      "loss": 3.4371,
+      "step": 11291
+    },
+    {
+      "epoch": 0.47545263157894735,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002776306195568064,
+      "loss": 3.7009,
+      "step": 11292
+    },
+    {
+      "epoch": 0.47549473684210525,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.000277597081753353,
+      "loss": 2.9753,
+      "step": 11293
+    },
+    {
+      "epoch": 0.47553684210526315,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002775635434471131,
+      "loss": 2.8093,
+      "step": 11294
+    },
+    {
+      "epoch": 0.47557894736842105,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002775300046386979,
+      "loss": 3.0758,
+      "step": 11295
+    },
+    {
+      "epoch": 0.47562105263157894,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00027749646532871833,
+      "loss": 3.5004,
+      "step": 11296
+    },
+    {
+      "epoch": 0.47566315789473684,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00027746292551778553,
+      "loss": 3.3972,
+      "step": 11297
+    },
+    {
+      "epoch": 0.47570526315789474,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002774293852065104,
+      "loss": 3.3036,
+      "step": 11298
+    },
+    {
+      "epoch": 0.47574736842105264,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002773958443955041,
+      "loss": 3.2322,
+      "step": 11299
+    },
+    {
+      "epoch": 0.47578947368421054,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002773623030853777,
+      "loss": 3.0336,
+      "step": 11300
+    },
+    {
+      "epoch": 0.47583157894736844,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00027732876127674235,
+      "loss": 3.303,
+      "step": 11301
+    },
+    {
+      "epoch": 0.47587368421052634,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002772952189702091,
+      "loss": 2.9711,
+      "step": 11302
+    },
+    {
+      "epoch": 0.47591578947368424,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.000277261676166389,
+      "loss": 2.7269,
+      "step": 11303
+    },
+    {
+      "epoch": 0.4759578947368421,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00027722813286589317,
+      "loss": 3.3304,
+      "step": 11304
+    },
+    {
+      "epoch": 0.476,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002771945890693328,
+      "loss": 3.4456,
+      "step": 11305
+    },
+    {
+      "epoch": 0.4760421052631579,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00027716104477731884,
+      "loss": 2.9679,
+      "step": 11306
+    },
+    {
+      "epoch": 0.4760842105263158,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002771274999904626,
+      "loss": 3.0881,
+      "step": 11307
+    },
+    {
+      "epoch": 0.4761263157894737,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00027709395470937526,
+      "loss": 2.6589,
+      "step": 11308
+    },
+    {
+      "epoch": 0.4761684210526316,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002770604089346679,
+      "loss": 3.2297,
+      "step": 11309
+    },
+    {
+      "epoch": 0.4762105263157895,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00027702686266695164,
+      "loss": 3.1835,
+      "step": 11310
+    },
+    {
+      "epoch": 0.47625263157894737,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00027699331590683775,
+      "loss": 3.2934,
+      "step": 11311
+    },
+    {
+      "epoch": 0.47629473684210527,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002769597686549373,
+      "loss": 3.6355,
+      "step": 11312
+    },
+    {
+      "epoch": 0.47633684210526317,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00027692622091186154,
+      "loss": 3.1638,
+      "step": 11313
+    },
+    {
+      "epoch": 0.47637894736842107,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002768926726782217,
+      "loss": 2.9164,
+      "step": 11314
+    },
+    {
+      "epoch": 0.47642105263157897,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002768591239546289,
+      "loss": 3.1783,
+      "step": 11315
+    },
+    {
+      "epoch": 0.47646315789473687,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002768255747416945,
+      "loss": 3.1852,
+      "step": 11316
+    },
+    {
+      "epoch": 0.4765052631578947,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00027679202504002966,
+      "loss": 3.161,
+      "step": 11317
+    },
+    {
+      "epoch": 0.4765473684210526,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002767584748502456,
+      "loss": 2.7965,
+      "step": 11318
+    },
+    {
+      "epoch": 0.4765894736842105,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002767249241729535,
+      "loss": 3.2595,
+      "step": 11319
+    },
+    {
+      "epoch": 0.4766315789473684,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002766913730087647,
+      "loss": 3.1578,
+      "step": 11320
+    },
+    {
+      "epoch": 0.4766736842105263,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002766578213582905,
+      "loss": 2.8727,
+      "step": 11321
+    },
+    {
+      "epoch": 0.4767157894736842,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002766242692221421,
+      "loss": 3.1883,
+      "step": 11322
+    },
+    {
+      "epoch": 0.4767578947368421,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00027659071660093073,
+      "loss": 3.4165,
+      "step": 11323
+    },
+    {
+      "epoch": 0.4768,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00027655716349526775,
+      "loss": 3.1915,
+      "step": 11324
+    },
+    {
+      "epoch": 0.4768421052631579,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00027652360990576455,
+      "loss": 3.0349,
+      "step": 11325
+    },
+    {
+      "epoch": 0.4768842105263158,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002764900558330322,
+      "loss": 3.0085,
+      "step": 11326
+    },
+    {
+      "epoch": 0.4769263157894737,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002764565012776823,
+      "loss": 3.1023,
+      "step": 11327
+    },
+    {
+      "epoch": 0.4769684210526316,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00027642294624032597,
+      "loss": 2.7274,
+      "step": 11328
+    },
+    {
+      "epoch": 0.4770105263157895,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00027638939072157466,
+      "loss": 3.3183,
+      "step": 11329
+    },
+    {
+      "epoch": 0.4770526315789474,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002763558347220396,
+      "loss": 3.506,
+      "step": 11330
+    },
+    {
+      "epoch": 0.47709473684210524,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00027632227824233213,
+      "loss": 3.4384,
+      "step": 11331
+    },
+    {
+      "epoch": 0.47713684210526314,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002762887212830638,
+      "loss": 3.1726,
+      "step": 11332
+    },
+    {
+      "epoch": 0.47717894736842104,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002762551638448458,
+      "loss": 3.1624,
+      "step": 11333
+    },
+    {
+      "epoch": 0.47722105263157893,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00027622160592828953,
+      "loss": 2.8228,
+      "step": 11334
+    },
+    {
+      "epoch": 0.47726315789473683,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002761880475340065,
+      "loss": 3.2728,
+      "step": 11335
+    },
+    {
+      "epoch": 0.47730526315789473,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00027615448866260796,
+      "loss": 3.145,
+      "step": 11336
+    },
+    {
+      "epoch": 0.47734736842105263,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002761209293147054,
+      "loss": 3.2462,
+      "step": 11337
+    },
+    {
+      "epoch": 0.47738947368421053,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002760873694909102,
+      "loss": 2.6945,
+      "step": 11338
+    },
+    {
+      "epoch": 0.47743157894736843,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002760538091918337,
+      "loss": 2.8207,
+      "step": 11339
+    },
+    {
+      "epoch": 0.47747368421052633,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0002760202484180875,
+      "loss": 2.6326,
+      "step": 11340
+    },
+    {
+      "epoch": 0.4775157894736842,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00027598668717028284,
+      "loss": 3.2766,
+      "step": 11341
+    },
+    {
+      "epoch": 0.4775578947368421,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002759531254490313,
+      "loss": 3.0902,
+      "step": 11342
+    },
+    {
+      "epoch": 0.4776,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002759195632549444,
+      "loss": 2.6482,
+      "step": 11343
+    },
+    {
+      "epoch": 0.47764210526315787,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00027588600058863345,
+      "loss": 3.321,
+      "step": 11344
+    },
+    {
+      "epoch": 0.47768421052631577,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00027585243745071003,
+      "loss": 3.1135,
+      "step": 11345
+    },
+    {
+      "epoch": 0.47772631578947367,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002758188738417855,
+      "loss": 2.8064,
+      "step": 11346
+    },
+    {
+      "epoch": 0.47776842105263156,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00027578530976247144,
+      "loss": 3.0362,
+      "step": 11347
+    },
+    {
+      "epoch": 0.47781052631578946,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002757517452133794,
+      "loss": 3.6248,
+      "step": 11348
+    },
+    {
+      "epoch": 0.47785263157894736,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00027571818019512075,
+      "loss": 2.9248,
+      "step": 11349
+    },
+    {
+      "epoch": 0.47789473684210526,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002756846147083071,
+      "loss": 3.3977,
+      "step": 11350
+    },
+    {
+      "epoch": 0.47793684210526316,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00027565104875354994,
+      "loss": 3.2024,
+      "step": 11351
+    },
+    {
+      "epoch": 0.47797894736842106,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002756174823314608,
+      "loss": 3.227,
+      "step": 11352
+    },
+    {
+      "epoch": 0.47802105263157896,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00027558391544265124,
+      "loss": 3.1085,
+      "step": 11353
+    },
+    {
+      "epoch": 0.47806315789473686,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00027555034808773285,
+      "loss": 2.9561,
+      "step": 11354
+    },
+    {
+      "epoch": 0.47810526315789476,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002755167802673171,
+      "loss": 2.9616,
+      "step": 11355
+    },
+    {
+      "epoch": 0.47814736842105265,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002754832119820156,
+      "loss": 3.0042,
+      "step": 11356
+    },
+    {
+      "epoch": 0.47818947368421055,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002754496432324399,
+      "loss": 3.0959,
+      "step": 11357
+    },
+    {
+      "epoch": 0.4782315789473684,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002754160740192016,
+      "loss": 2.9433,
+      "step": 11358
+    },
+    {
+      "epoch": 0.4782736842105263,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002753825043429123,
+      "loss": 3.1231,
+      "step": 11359
+    },
+    {
+      "epoch": 0.4783157894736842,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00027534893420418354,
+      "loss": 3.3139,
+      "step": 11360
+    },
+    {
+      "epoch": 0.4783578947368421,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00027531536360362707,
+      "loss": 3.1961,
+      "step": 11361
+    },
+    {
+      "epoch": 0.4784,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002752817925418544,
+      "loss": 3.0347,
+      "step": 11362
+    },
+    {
+      "epoch": 0.4784421052631579,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002752482210194772,
+      "loss": 2.9467,
+      "step": 11363
+    },
+    {
+      "epoch": 0.4784842105263158,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00027521464903710707,
+      "loss": 2.8194,
+      "step": 11364
+    },
+    {
+      "epoch": 0.4785263157894737,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00027518107659535565,
+      "loss": 3.4005,
+      "step": 11365
+    },
+    {
+      "epoch": 0.4785684210526316,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002751475036948345,
+      "loss": 3.2878,
+      "step": 11366
+    },
+    {
+      "epoch": 0.4786105263157895,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002751139303361555,
+      "loss": 3.4864,
+      "step": 11367
+    },
+    {
+      "epoch": 0.4786526315789474,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00027508035651993014,
+      "loss": 3.2432,
+      "step": 11368
+    },
+    {
+      "epoch": 0.4786947368421053,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002750467822467702,
+      "loss": 3.0233,
+      "step": 11369
+    },
+    {
+      "epoch": 0.4787368421052632,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002750132075172873,
+      "loss": 3.1092,
+      "step": 11370
+    },
+    {
+      "epoch": 0.478778947368421,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00027497963233209316,
+      "loss": 3.1232,
+      "step": 11371
+    },
+    {
+      "epoch": 0.4788210526315789,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00027494605669179945,
+      "loss": 2.6268,
+      "step": 11372
+    },
+    {
+      "epoch": 0.4788631578947368,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002749124805970179,
+      "loss": 2.7704,
+      "step": 11373
+    },
+    {
+      "epoch": 0.4789052631578947,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002748789040483602,
+      "loss": 3.0863,
+      "step": 11374
+    },
+    {
+      "epoch": 0.4789473684210526,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002748453270464381,
+      "loss": 3.7434,
+      "step": 11375
+    },
+    {
+      "epoch": 0.4789894736842105,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002748117495918634,
+      "loss": 3.4542,
+      "step": 11376
+    },
+    {
+      "epoch": 0.4790315789473684,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002747781716852477,
+      "loss": 3.398,
+      "step": 11377
+    },
+    {
+      "epoch": 0.4790736842105263,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00027474459332720285,
+      "loss": 3.355,
+      "step": 11378
+    },
+    {
+      "epoch": 0.4791157894736842,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00027471101451834065,
+      "loss": 3.0224,
+      "step": 11379
+    },
+    {
+      "epoch": 0.4791578947368421,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002746774352592727,
+      "loss": 2.6568,
+      "step": 11380
+    },
+    {
+      "epoch": 0.4792,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002746438555506109,
+      "loss": 2.9874,
+      "step": 11381
+    },
+    {
+      "epoch": 0.4792421052631579,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.000274610275392967,
+      "loss": 3.3681,
+      "step": 11382
+    },
+    {
+      "epoch": 0.4792842105263158,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002745766947869528,
+      "loss": 2.9196,
+      "step": 11383
+    },
+    {
+      "epoch": 0.4793263157894737,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00027454311373318013,
+      "loss": 3.179,
+      "step": 11384
+    },
+    {
+      "epoch": 0.47936842105263155,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002745095322322608,
+      "loss": 3.1272,
+      "step": 11385
+    },
+    {
+      "epoch": 0.47941052631578945,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002744759502848065,
+      "loss": 3.2385,
+      "step": 11386
+    },
+    {
+      "epoch": 0.47945263157894735,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002744423678914292,
+      "loss": 3.8109,
+      "step": 11387
+    },
+    {
+      "epoch": 0.47949473684210525,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00027440878505274066,
+      "loss": 2.9637,
+      "step": 11388
+    },
+    {
+      "epoch": 0.47953684210526315,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002743752017693528,
+      "loss": 3.0334,
+      "step": 11389
+    },
+    {
+      "epoch": 0.47957894736842105,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002743416180418772,
+      "loss": 2.7744,
+      "step": 11390
+    },
+    {
+      "epoch": 0.47962105263157895,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002743080338709261,
+      "loss": 2.9044,
+      "step": 11391
+    },
+    {
+      "epoch": 0.47966315789473685,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002742744492571112,
+      "loss": 3.008,
+      "step": 11392
+    },
+    {
+      "epoch": 0.47970526315789475,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002742408642010442,
+      "loss": 3.4472,
+      "step": 11393
+    },
+    {
+      "epoch": 0.47974736842105264,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00027420727870333726,
+      "loss": 3.4503,
+      "step": 11394
+    },
+    {
+      "epoch": 0.47978947368421054,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00027417369276460215,
+      "loss": 2.9477,
+      "step": 11395
+    },
+    {
+      "epoch": 0.47983157894736844,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002741401063854507,
+      "loss": 3.2224,
+      "step": 11396
+    },
+    {
+      "epoch": 0.47987368421052634,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00027410651956649495,
+      "loss": 2.95,
+      "step": 11397
+    },
+    {
+      "epoch": 0.4799157894736842,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00027407293230834667,
+      "loss": 2.6958,
+      "step": 11398
+    },
+    {
+      "epoch": 0.4799578947368421,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002740393446116179,
+      "loss": 3.5148,
+      "step": 11399
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00027400575647692046,
+      "loss": 3.1592,
+      "step": 11400
+    },
+    {
+      "epoch": 0.4800421052631579,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002739721679048663,
+      "loss": 3.054,
+      "step": 11401
+    },
+    {
+      "epoch": 0.4800842105263158,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002739385788960675,
+      "loss": 3.4561,
+      "step": 11402
+    },
+    {
+      "epoch": 0.4801263157894737,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00027390498945113593,
+      "loss": 3.1696,
+      "step": 11403
+    },
+    {
+      "epoch": 0.4801684210526316,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00027387139957068353,
+      "loss": 3.2177,
+      "step": 11404
+    },
+    {
+      "epoch": 0.4802105263157895,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00027383780925532223,
+      "loss": 2.7724,
+      "step": 11405
+    },
+    {
+      "epoch": 0.4802526315789474,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002738042185056641,
+      "loss": 3.1676,
+      "step": 11406
+    },
+    {
+      "epoch": 0.4802947368421053,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00027377062732232105,
+      "loss": 3.0837,
+      "step": 11407
+    },
+    {
+      "epoch": 0.4803368421052632,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.000273737035705905,
+      "loss": 3.0646,
+      "step": 11408
+    },
+    {
+      "epoch": 0.48037894736842107,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002737034436570282,
+      "loss": 2.853,
+      "step": 11409
+    },
+    {
+      "epoch": 0.48042105263157897,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002736698511763025,
+      "loss": 2.9993,
+      "step": 11410
+    },
+    {
+      "epoch": 0.48046315789473687,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00027363625826433993,
+      "loss": 3.0365,
+      "step": 11411
+    },
+    {
+      "epoch": 0.4805052631578947,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002736026649217524,
+      "loss": 2.3459,
+      "step": 11412
+    },
+    {
+      "epoch": 0.4805473684210526,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002735690711491521,
+      "loss": 2.957,
+      "step": 11413
+    },
+    {
+      "epoch": 0.4805894736842105,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00027353547694715103,
+      "loss": 3.1379,
+      "step": 11414
+    },
+    {
+      "epoch": 0.4806315789473684,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002735018823163612,
+      "loss": 3.2797,
+      "step": 11415
+    },
+    {
+      "epoch": 0.4806736842105263,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002734682872573947,
+      "loss": 3.0173,
+      "step": 11416
+    },
+    {
+      "epoch": 0.4807157894736842,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00027343469177086354,
+      "loss": 3.4544,
+      "step": 11417
+    },
+    {
+      "epoch": 0.4807578947368421,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002734010958573799,
+      "loss": 3.2111,
+      "step": 11418
+    },
+    {
+      "epoch": 0.4808,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002733674995175558,
+      "loss": 3.4395,
+      "step": 11419
+    },
+    {
+      "epoch": 0.4808421052631579,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002733339027520032,
+      "loss": 3.3871,
+      "step": 11420
+    },
+    {
+      "epoch": 0.4808842105263158,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00027330030556133446,
+      "loss": 3.1829,
+      "step": 11421
+    },
+    {
+      "epoch": 0.4809263157894737,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002732667079461614,
+      "loss": 2.9954,
+      "step": 11422
+    },
+    {
+      "epoch": 0.4809684210526316,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002732331099070964,
+      "loss": 2.6275,
+      "step": 11423
+    },
+    {
+      "epoch": 0.4810105263157895,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00027319951144475125,
+      "loss": 3.1068,
+      "step": 11424
+    },
+    {
+      "epoch": 0.48105263157894734,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002731659125597384,
+      "loss": 2.7786,
+      "step": 11425
+    },
+    {
+      "epoch": 0.48109473684210524,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002731323132526698,
+      "loss": 3.0806,
+      "step": 11426
+    },
+    {
+      "epoch": 0.48113684210526314,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002730987135241576,
+      "loss": 2.7643,
+      "step": 11427
+    },
+    {
+      "epoch": 0.48117894736842104,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.000273065113374814,
+      "loss": 3.2343,
+      "step": 11428
+    },
+    {
+      "epoch": 0.48122105263157894,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002730315128052512,
+      "loss": 3.299,
+      "step": 11429
+    },
+    {
+      "epoch": 0.48126315789473684,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002729979118160812,
+      "loss": 3.0913,
+      "step": 11430
+    },
+    {
+      "epoch": 0.48130526315789474,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002729643104079164,
+      "loss": 2.9683,
+      "step": 11431
+    },
+    {
+      "epoch": 0.48134736842105263,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00027293070858136866,
+      "loss": 2.9203,
+      "step": 11432
+    },
+    {
+      "epoch": 0.48138947368421053,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00027289710633705047,
+      "loss": 2.8536,
+      "step": 11433
+    },
+    {
+      "epoch": 0.48143157894736843,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00027286350367557383,
+      "loss": 3.4811,
+      "step": 11434
+    },
+    {
+      "epoch": 0.48147368421052633,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002728299005975511,
+      "loss": 3.1947,
+      "step": 11435
+    },
+    {
+      "epoch": 0.48151578947368423,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002727962971035944,
+      "loss": 2.8504,
+      "step": 11436
+    },
+    {
+      "epoch": 0.48155789473684213,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00027276269319431593,
+      "loss": 3.1503,
+      "step": 11437
+    },
+    {
+      "epoch": 0.4816,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002727290888703279,
+      "loss": 3.274,
+      "step": 11438
+    },
+    {
+      "epoch": 0.48164210526315787,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00027269548413224264,
+      "loss": 2.8836,
+      "step": 11439
+    },
+    {
+      "epoch": 0.48168421052631577,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002726618789806723,
+      "loss": 3.3781,
+      "step": 11440
+    },
+    {
+      "epoch": 0.48172631578947367,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00027262827341622913,
+      "loss": 2.7527,
+      "step": 11441
+    },
+    {
+      "epoch": 0.48176842105263157,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002725946674395254,
+      "loss": 3.0498,
+      "step": 11442
+    },
+    {
+      "epoch": 0.48181052631578947,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002725610610511734,
+      "loss": 3.2121,
+      "step": 11443
+    },
+    {
+      "epoch": 0.48185263157894737,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002725274542517854,
+      "loss": 3.2327,
+      "step": 11444
+    },
+    {
+      "epoch": 0.48189473684210526,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002724938470419736,
+      "loss": 3.3195,
+      "step": 11445
+    },
+    {
+      "epoch": 0.48193684210526316,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00027246023942235036,
+      "loss": 2.9329,
+      "step": 11446
+    },
+    {
+      "epoch": 0.48197894736842106,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.000272426631393528,
+      "loss": 3.2601,
+      "step": 11447
+    },
+    {
+      "epoch": 0.48202105263157896,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002723930229561187,
+      "loss": 3.3901,
+      "step": 11448
+    },
+    {
+      "epoch": 0.48206315789473686,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002723594141107348,
+      "loss": 3.2576,
+      "step": 11449
+    },
+    {
+      "epoch": 0.48210526315789476,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002723258048579887,
+      "loss": 3.0759,
+      "step": 11450
+    },
+    {
+      "epoch": 0.48214736842105266,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00027229219519849267,
+      "loss": 3.1091,
+      "step": 11451
+    },
+    {
+      "epoch": 0.4821894736842105,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.000272258585132859,
+      "loss": 3.2236,
+      "step": 11452
+    },
+    {
+      "epoch": 0.4822315789473684,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00027222497466170014,
+      "loss": 3.5546,
+      "step": 11453
+    },
+    {
+      "epoch": 0.4822736842105263,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00027219136378562827,
+      "loss": 3.2331,
+      "step": 11454
+    },
+    {
+      "epoch": 0.4823157894736842,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0002721577525052559,
+      "loss": 3.1089,
+      "step": 11455
+    },
+    {
+      "epoch": 0.4823578947368421,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00027212414082119525,
+      "loss": 3.2191,
+      "step": 11456
+    },
+    {
+      "epoch": 0.4824,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00027209052873405877,
+      "loss": 3.0693,
+      "step": 11457
+    },
+    {
+      "epoch": 0.4824421052631579,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002720569162444588,
+      "loss": 3.0904,
+      "step": 11458
+    },
+    {
+      "epoch": 0.4824842105263158,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00027202330335300773,
+      "loss": 3.1729,
+      "step": 11459
+    },
+    {
+      "epoch": 0.4825263157894737,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002719896900603179,
+      "loss": 3.3877,
+      "step": 11460
+    },
+    {
+      "epoch": 0.4825684210526316,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00027195607636700187,
+      "loss": 2.7472,
+      "step": 11461
+    },
+    {
+      "epoch": 0.4826105263157895,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002719224622736719,
+      "loss": 3.1258,
+      "step": 11462
+    },
+    {
+      "epoch": 0.4826526315789474,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002718888477809404,
+      "loss": 3.5352,
+      "step": 11463
+    },
+    {
+      "epoch": 0.4826947368421053,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002718552328894198,
+      "loss": 3.4957,
+      "step": 11464
+    },
+    {
+      "epoch": 0.48273684210526313,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00027182161759972256,
+      "loss": 3.4153,
+      "step": 11465
+    },
+    {
+      "epoch": 0.48277894736842103,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00027178800191246104,
+      "loss": 2.9373,
+      "step": 11466
+    },
+    {
+      "epoch": 0.48282105263157893,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00027175438582824774,
+      "loss": 3.1656,
+      "step": 11467
+    },
+    {
+      "epoch": 0.4828631578947368,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002717207693476951,
+      "loss": 3.4927,
+      "step": 11468
+    },
+    {
+      "epoch": 0.4829052631578947,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002716871524714156,
+      "loss": 3.4933,
+      "step": 11469
+    },
+    {
+      "epoch": 0.4829473684210526,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002716535352000217,
+      "loss": 3.2775,
+      "step": 11470
+    },
+    {
+      "epoch": 0.4829894736842105,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00027161991753412573,
+      "loss": 2.7777,
+      "step": 11471
+    },
+    {
+      "epoch": 0.4830315789473684,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00027158629947434033,
+      "loss": 3.3784,
+      "step": 11472
+    },
+    {
+      "epoch": 0.4830736842105263,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002715526810212779,
+      "loss": 3.0621,
+      "step": 11473
+    },
+    {
+      "epoch": 0.4831157894736842,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002715190621755509,
+      "loss": 2.7169,
+      "step": 11474
+    },
+    {
+      "epoch": 0.4831578947368421,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002714854429377718,
+      "loss": 3.3297,
+      "step": 11475
+    },
+    {
+      "epoch": 0.4832,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00027145182330855336,
+      "loss": 3.4379,
+      "step": 11476
+    },
+    {
+      "epoch": 0.4832421052631579,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002714182032885078,
+      "loss": 3.0992,
+      "step": 11477
+    },
+    {
+      "epoch": 0.4832842105263158,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002713845828782478,
+      "loss": 3.2114,
+      "step": 11478
+    },
+    {
+      "epoch": 0.48332631578947366,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00027135096207838577,
+      "loss": 2.8559,
+      "step": 11479
+    },
+    {
+      "epoch": 0.48336842105263156,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002713173408895343,
+      "loss": 3.0378,
+      "step": 11480
+    },
+    {
+      "epoch": 0.48341052631578946,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00027128371931230594,
+      "loss": 2.9799,
+      "step": 11481
+    },
+    {
+      "epoch": 0.48345263157894736,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00027125009734731327,
+      "loss": 2.6369,
+      "step": 11482
+    },
+    {
+      "epoch": 0.48349473684210525,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00027121647499516865,
+      "loss": 2.882,
+      "step": 11483
+    },
+    {
+      "epoch": 0.48353684210526315,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002711828522564849,
+      "loss": 3.2796,
+      "step": 11484
+    },
+    {
+      "epoch": 0.48357894736842105,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0002711492291318745,
+      "loss": 3.0068,
+      "step": 11485
+    },
+    {
+      "epoch": 0.48362105263157895,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00027111560562195,
+      "loss": 3.4852,
+      "step": 11486
+    },
+    {
+      "epoch": 0.48366315789473685,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00027108198172732384,
+      "loss": 2.7816,
+      "step": 11487
+    },
+    {
+      "epoch": 0.48370526315789475,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002710483574486088,
+      "loss": 3.4087,
+      "step": 11488
+    },
+    {
+      "epoch": 0.48374736842105265,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002710147327864175,
+      "loss": 2.9015,
+      "step": 11489
+    },
+    {
+      "epoch": 0.48378947368421055,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00027098110774136244,
+      "loss": 3.228,
+      "step": 11490
+    },
+    {
+      "epoch": 0.48383157894736845,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00027094748231405616,
+      "loss": 3.1674,
+      "step": 11491
+    },
+    {
+      "epoch": 0.4838736842105263,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002709138565051114,
+      "loss": 2.7901,
+      "step": 11492
+    },
+    {
+      "epoch": 0.4839157894736842,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002708802303151408,
+      "loss": 3.1453,
+      "step": 11493
+    },
+    {
+      "epoch": 0.4839578947368421,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002708466037447568,
+      "loss": 3.2396,
+      "step": 11494
+    },
+    {
+      "epoch": 0.484,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00027081297679457236,
+      "loss": 2.8111,
+      "step": 11495
+    },
+    {
+      "epoch": 0.4840421052631579,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00027077934946519985,
+      "loss": 3.3532,
+      "step": 11496
+    },
+    {
+      "epoch": 0.4840842105263158,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.000270745721757252,
+      "loss": 2.8147,
+      "step": 11497
+    },
+    {
+      "epoch": 0.4841263157894737,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002707120936713415,
+      "loss": 3.1329,
+      "step": 11498
+    },
+    {
+      "epoch": 0.4841684210526316,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.000270678465208081,
+      "loss": 3.5328,
+      "step": 11499
+    },
+    {
+      "epoch": 0.4842105263157895,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00027064483636808314,
+      "loss": 2.9755,
+      "step": 11500
+    },
+    {
+      "epoch": 0.4842526315789474,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00027061120715196054,
+      "loss": 3.1495,
+      "step": 11501
+    },
+    {
+      "epoch": 0.4842947368421053,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00027057757756032604,
+      "loss": 3.3247,
+      "step": 11502
+    },
+    {
+      "epoch": 0.4843368421052632,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00027054394759379227,
+      "loss": 2.9325,
+      "step": 11503
+    },
+    {
+      "epoch": 0.4843789473684211,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00027051031725297195,
+      "loss": 2.8036,
+      "step": 11504
+    },
+    {
+      "epoch": 0.484421052631579,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00027047668653847767,
+      "loss": 3.2401,
+      "step": 11505
+    },
+    {
+      "epoch": 0.4844631578947368,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002704430554509222,
+      "loss": 3.0875,
+      "step": 11506
+    },
+    {
+      "epoch": 0.4845052631578947,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002704094239909183,
+      "loss": 2.9815,
+      "step": 11507
+    },
+    {
+      "epoch": 0.4845473684210526,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00027037579215907873,
+      "loss": 3.2354,
+      "step": 11508
+    },
+    {
+      "epoch": 0.4845894736842105,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00027034215995601606,
+      "loss": 3.3104,
+      "step": 11509
+    },
+    {
+      "epoch": 0.4846315789473684,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00027030852738234323,
+      "loss": 2.9534,
+      "step": 11510
+    },
+    {
+      "epoch": 0.4846736842105263,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002702748944386729,
+      "loss": 3.0815,
+      "step": 11511
+    },
+    {
+      "epoch": 0.4847157894736842,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002702412611256178,
+      "loss": 3.0287,
+      "step": 11512
+    },
+    {
+      "epoch": 0.4847578947368421,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002702076274437907,
+      "loss": 3.0993,
+      "step": 11513
+    },
+    {
+      "epoch": 0.4848,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00027017399339380435,
+      "loss": 3.2814,
+      "step": 11514
+    },
+    {
+      "epoch": 0.4848421052631579,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00027014035897627155,
+      "loss": 3.2356,
+      "step": 11515
+    },
+    {
+      "epoch": 0.4848842105263158,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002701067241918051,
+      "loss": 3.0912,
+      "step": 11516
+    },
+    {
+      "epoch": 0.4849263157894737,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002700730890410177,
+      "loss": 3.2995,
+      "step": 11517
+    },
+    {
+      "epoch": 0.4849684210526316,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00027003945352452227,
+      "loss": 2.6678,
+      "step": 11518
+    },
+    {
+      "epoch": 0.48501052631578945,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002700058176429315,
+      "loss": 3.0785,
+      "step": 11519
+    },
+    {
+      "epoch": 0.48505263157894735,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00026997218139685825,
+      "loss": 2.8501,
+      "step": 11520
+    },
+    {
+      "epoch": 0.48509473684210525,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00026993854478691534,
+      "loss": 3.2432,
+      "step": 11521
+    },
+    {
+      "epoch": 0.48513684210526314,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00026990490781371554,
+      "loss": 3.3365,
+      "step": 11522
+    },
+    {
+      "epoch": 0.48517894736842104,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00026987127047787174,
+      "loss": 3.3956,
+      "step": 11523
+    },
+    {
+      "epoch": 0.48522105263157894,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00026983763277999673,
+      "loss": 2.6143,
+      "step": 11524
+    },
+    {
+      "epoch": 0.48526315789473684,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002698039947207033,
+      "loss": 2.7039,
+      "step": 11525
+    },
+    {
+      "epoch": 0.48530526315789474,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00026977035630060443,
+      "loss": 3.0832,
+      "step": 11526
+    },
+    {
+      "epoch": 0.48534736842105264,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002697367175203129,
+      "loss": 3.0168,
+      "step": 11527
+    },
+    {
+      "epoch": 0.48538947368421054,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002697030783804415,
+      "loss": 2.8895,
+      "step": 11528
+    },
+    {
+      "epoch": 0.48543157894736844,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00026966943888160315,
+      "loss": 3.2859,
+      "step": 11529
+    },
+    {
+      "epoch": 0.48547368421052634,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002696357990244108,
+      "loss": 2.8773,
+      "step": 11530
+    },
+    {
+      "epoch": 0.48551578947368423,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002696021588094773,
+      "loss": 3.3982,
+      "step": 11531
+    },
+    {
+      "epoch": 0.48555789473684213,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00026956851823741547,
+      "loss": 2.501,
+      "step": 11532
+    },
+    {
+      "epoch": 0.4856,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00026953487730883814,
+      "loss": 3.1027,
+      "step": 11533
+    },
+    {
+      "epoch": 0.4856421052631579,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002695012360243583,
+      "loss": 3.0517,
+      "step": 11534
+    },
+    {
+      "epoch": 0.4856842105263158,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.000269467594384589,
+      "loss": 3.277,
+      "step": 11535
+    },
+    {
+      "epoch": 0.4857263157894737,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002694339523901429,
+      "loss": 3.3383,
+      "step": 11536
+    },
+    {
+      "epoch": 0.48576842105263157,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002694003100416331,
+      "loss": 2.7619,
+      "step": 11537
+    },
+    {
+      "epoch": 0.48581052631578947,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00026936666733967234,
+      "loss": 3.0152,
+      "step": 11538
+    },
+    {
+      "epoch": 0.48585263157894737,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002693330242848737,
+      "loss": 2.8482,
+      "step": 11539
+    },
+    {
+      "epoch": 0.48589473684210527,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002692993808778501,
+      "loss": 3.1498,
+      "step": 11540
+    },
+    {
+      "epoch": 0.48593684210526317,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00026926573711921444,
+      "loss": 2.6681,
+      "step": 11541
+    },
+    {
+      "epoch": 0.48597894736842107,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00026923209300957964,
+      "loss": 2.9871,
+      "step": 11542
+    },
+    {
+      "epoch": 0.48602105263157896,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002691984485495587,
+      "loss": 2.7575,
+      "step": 11543
+    },
+    {
+      "epoch": 0.48606315789473686,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00026916480373976464,
+      "loss": 3.1203,
+      "step": 11544
+    },
+    {
+      "epoch": 0.48610526315789476,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002691311585808104,
+      "loss": 3.366,
+      "step": 11545
+    },
+    {
+      "epoch": 0.4861473684210526,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00026909751307330887,
+      "loss": 2.9958,
+      "step": 11546
+    },
+    {
+      "epoch": 0.4861894736842105,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002690638672178732,
+      "loss": 3.1085,
+      "step": 11547
+    },
+    {
+      "epoch": 0.4862315789473684,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002690302210151161,
+      "loss": 3.3156,
+      "step": 11548
+    },
+    {
+      "epoch": 0.4862736842105263,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002689965744656508,
+      "loss": 3.2489,
+      "step": 11549
+    },
+    {
+      "epoch": 0.4863157894736842,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002689629275700902,
+      "loss": 3.1426,
+      "step": 11550
+    },
+    {
+      "epoch": 0.4863578947368421,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002689292803290474,
+      "loss": 3.2146,
+      "step": 11551
+    },
+    {
+      "epoch": 0.4864,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0002688956327431354,
+      "loss": 3.7328,
+      "step": 11552
+    },
+    {
+      "epoch": 0.4864421052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002688619848129671,
+      "loss": 3.2254,
+      "step": 11553
+    },
+    {
+      "epoch": 0.4864842105263158,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00026882833653915563,
+      "loss": 3.2479,
+      "step": 11554
+    },
+    {
+      "epoch": 0.4865263157894737,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002687946879223139,
+      "loss": 3.381,
+      "step": 11555
+    },
+    {
+      "epoch": 0.4865684210526316,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002687610389630551,
+      "loss": 3.5601,
+      "step": 11556
+    },
+    {
+      "epoch": 0.4866105263157895,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002687273896619923,
+      "loss": 2.9476,
+      "step": 11557
+    },
+    {
+      "epoch": 0.4866526315789474,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002686937400197383,
+      "loss": 2.7184,
+      "step": 11558
+    },
+    {
+      "epoch": 0.48669473684210524,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00026866009003690647,
+      "loss": 2.8956,
+      "step": 11559
+    },
+    {
+      "epoch": 0.48673684210526313,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002686264397141096,
+      "loss": 2.8442,
+      "step": 11560
+    },
+    {
+      "epoch": 0.48677894736842103,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00026859278905196093,
+      "loss": 3.2941,
+      "step": 11561
+    },
+    {
+      "epoch": 0.48682105263157893,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00026855913805107353,
+      "loss": 3.2551,
+      "step": 11562
+    },
+    {
+      "epoch": 0.48686315789473683,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002685254867120604,
+      "loss": 3.1075,
+      "step": 11563
+    },
+    {
+      "epoch": 0.48690526315789473,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002684918350355347,
+      "loss": 3.1631,
+      "step": 11564
+    },
+    {
+      "epoch": 0.48694736842105263,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002684581830221095,
+      "loss": 3.2481,
+      "step": 11565
+    },
+    {
+      "epoch": 0.4869894736842105,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00026842453067239786,
+      "loss": 3.2536,
+      "step": 11566
+    },
+    {
+      "epoch": 0.4870315789473684,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00026839087798701287,
+      "loss": 3.1771,
+      "step": 11567
+    },
+    {
+      "epoch": 0.4870736842105263,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0002683572249665677,
+      "loss": 2.5556,
+      "step": 11568
+    },
+    {
+      "epoch": 0.4871157894736842,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002683235716116756,
+      "loss": 3.1457,
+      "step": 11569
+    },
+    {
+      "epoch": 0.4871578947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002682899179229494,
+      "loss": 3.3454,
+      "step": 11570
+    },
+    {
+      "epoch": 0.4872,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002682562639010025,
+      "loss": 3.3969,
+      "step": 11571
+    },
+    {
+      "epoch": 0.4872421052631579,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002682226095464479,
+      "loss": 3.3061,
+      "step": 11572
+    },
+    {
+      "epoch": 0.48728421052631576,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00026818895485989877,
+      "loss": 3.3871,
+      "step": 11573
+    },
+    {
+      "epoch": 0.48732631578947366,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00026815529984196816,
+      "loss": 2.8728,
+      "step": 11574
+    },
+    {
+      "epoch": 0.48736842105263156,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00026812164449326937,
+      "loss": 3.1885,
+      "step": 11575
+    },
+    {
+      "epoch": 0.48741052631578946,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00026808798881441546,
+      "loss": 3.3922,
+      "step": 11576
+    },
+    {
+      "epoch": 0.48745263157894736,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002680543328060197,
+      "loss": 3.0753,
+      "step": 11577
+    },
+    {
+      "epoch": 0.48749473684210526,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0002680206764686953,
+      "loss": 3.1138,
+      "step": 11578
+    },
+    {
+      "epoch": 0.48753684210526316,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002679870198030552,
+      "loss": 3.2394,
+      "step": 11579
+    },
+    {
+      "epoch": 0.48757894736842106,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00026795336280971286,
+      "loss": 3.2322,
+      "step": 11580
+    },
+    {
+      "epoch": 0.48762105263157896,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00026791970548928123,
+      "loss": 2.6476,
+      "step": 11581
+    },
+    {
+      "epoch": 0.48766315789473685,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002678860478423737,
+      "loss": 3.3496,
+      "step": 11582
+    },
+    {
+      "epoch": 0.48770526315789475,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00026785238986960335,
+      "loss": 3.3752,
+      "step": 11583
+    },
+    {
+      "epoch": 0.48774736842105265,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002678187315715834,
+      "loss": 2.8132,
+      "step": 11584
+    },
+    {
+      "epoch": 0.48778947368421055,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002677850729489271,
+      "loss": 3.052,
+      "step": 11585
+    },
+    {
+      "epoch": 0.4878315789473684,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002677514140022478,
+      "loss": 3.1179,
+      "step": 11586
+    },
+    {
+      "epoch": 0.4878736842105263,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002677177547321585,
+      "loss": 3.5733,
+      "step": 11587
+    },
+    {
+      "epoch": 0.4879157894736842,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00026768409513927246,
+      "loss": 2.9489,
+      "step": 11588
+    },
+    {
+      "epoch": 0.4879578947368421,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00026765043522420306,
+      "loss": 3.0931,
+      "step": 11589
+    },
+    {
+      "epoch": 0.488,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00026761677498756347,
+      "loss": 2.8685,
+      "step": 11590
+    },
+    {
+      "epoch": 0.4880421052631579,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00026758311442996697,
+      "loss": 2.9735,
+      "step": 11591
+    },
+    {
+      "epoch": 0.4880842105263158,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002675494535520267,
+      "loss": 3.2333,
+      "step": 11592
+    },
+    {
+      "epoch": 0.4881263157894737,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002675157923543561,
+      "loss": 2.9443,
+      "step": 11593
+    },
+    {
+      "epoch": 0.4881684210526316,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002674821308375683,
+      "loss": 2.9414,
+      "step": 11594
+    },
+    {
+      "epoch": 0.4882105263157895,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002674484690022765,
+      "loss": 2.993,
+      "step": 11595
+    },
+    {
+      "epoch": 0.4882526315789474,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00026741480684909426,
+      "loss": 3.1588,
+      "step": 11596
+    },
+    {
+      "epoch": 0.4882947368421053,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00026738114437863467,
+      "loss": 3.072,
+      "step": 11597
+    },
+    {
+      "epoch": 0.4883368421052632,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.000267347481591511,
+      "loss": 3.397,
+      "step": 11598
+    },
+    {
+      "epoch": 0.4883789473684211,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002673138184883367,
+      "loss": 3.0259,
+      "step": 11599
+    },
+    {
+      "epoch": 0.4884210526315789,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002672801550697248,
+      "loss": 3.0678,
+      "step": 11600
+    },
+    {
+      "epoch": 0.4884631578947368,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00026724649133628894,
+      "loss": 3.4206,
+      "step": 11601
+    },
+    {
+      "epoch": 0.4885052631578947,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002672128272886422,
+      "loss": 3.2407,
+      "step": 11602
+    },
+    {
+      "epoch": 0.4885473684210526,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002671791629273979,
+      "loss": 3.533,
+      "step": 11603
+    },
+    {
+      "epoch": 0.4885894736842105,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00026714549825316956,
+      "loss": 3.5265,
+      "step": 11604
+    },
+    {
+      "epoch": 0.4886315789473684,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002671118332665703,
+      "loss": 3.1144,
+      "step": 11605
+    },
+    {
+      "epoch": 0.4886736842105263,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002670781679682137,
+      "loss": 3.4031,
+      "step": 11606
+    },
+    {
+      "epoch": 0.4887157894736842,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002670445023587128,
+      "loss": 2.9668,
+      "step": 11607
+    },
+    {
+      "epoch": 0.4887578947368421,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00026701083643868105,
+      "loss": 2.5575,
+      "step": 11608
+    },
+    {
+      "epoch": 0.4888,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.000266977170208732,
+      "loss": 3.3395,
+      "step": 11609
+    },
+    {
+      "epoch": 0.4888421052631579,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002669435036694787,
+      "loss": 2.9966,
+      "step": 11610
+    },
+    {
+      "epoch": 0.4888842105263158,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00026690983682153477,
+      "loss": 3.0507,
+      "step": 11611
+    },
+    {
+      "epoch": 0.4889263157894737,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00026687616966551345,
+      "loss": 3.5903,
+      "step": 11612
+    },
+    {
+      "epoch": 0.48896842105263155,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00026684250220202816,
+      "loss": 3.5714,
+      "step": 11613
+    },
+    {
+      "epoch": 0.48901052631578945,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002668088344316923,
+      "loss": 3.2942,
+      "step": 11614
+    },
+    {
+      "epoch": 0.48905263157894735,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002667751663551191,
+      "loss": 2.7759,
+      "step": 11615
+    },
+    {
+      "epoch": 0.48909473684210525,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00026674149797292226,
+      "loss": 2.6355,
+      "step": 11616
+    },
+    {
+      "epoch": 0.48913684210526315,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002667078292857149,
+      "loss": 3.0265,
+      "step": 11617
+    },
+    {
+      "epoch": 0.48917894736842105,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00026667416029411045,
+      "loss": 2.7497,
+      "step": 11618
+    },
+    {
+      "epoch": 0.48922105263157895,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002666404909987225,
+      "loss": 3.5654,
+      "step": 11619
+    },
+    {
+      "epoch": 0.48926315789473684,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00026660682140016437,
+      "loss": 3.1773,
+      "step": 11620
+    },
+    {
+      "epoch": 0.48930526315789474,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002665731514990494,
+      "loss": 2.2933,
+      "step": 11621
+    },
+    {
+      "epoch": 0.48934736842105264,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002665394812959911,
+      "loss": 2.8146,
+      "step": 11622
+    },
+    {
+      "epoch": 0.48938947368421054,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00026650581079160285,
+      "loss": 3.4819,
+      "step": 11623
+    },
+    {
+      "epoch": 0.48943157894736844,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002664721399864982,
+      "loss": 3.2617,
+      "step": 11624
+    },
+    {
+      "epoch": 0.48947368421052634,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002664384688812905,
+      "loss": 2.7383,
+      "step": 11625
+    },
+    {
+      "epoch": 0.48951578947368424,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00026640479747659315,
+      "loss": 3.0708,
+      "step": 11626
+    },
+    {
+      "epoch": 0.4895578947368421,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002663711257730197,
+      "loss": 2.9493,
+      "step": 11627
+    },
+    {
+      "epoch": 0.4896,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00026633745377118356,
+      "loss": 3.0606,
+      "step": 11628
+    },
+    {
+      "epoch": 0.4896421052631579,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002663037814716982,
+      "loss": 2.9282,
+      "step": 11629
+    },
+    {
+      "epoch": 0.4896842105263158,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002662701088751771,
+      "loss": 3.1941,
+      "step": 11630
+    },
+    {
+      "epoch": 0.4897263157894737,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002662364359822338,
+      "loss": 3.7339,
+      "step": 11631
+    },
+    {
+      "epoch": 0.4897684210526316,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00026620276279348164,
+      "loss": 3.281,
+      "step": 11632
+    },
+    {
+      "epoch": 0.4898105263157895,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002661690893095343,
+      "loss": 2.9923,
+      "step": 11633
+    },
+    {
+      "epoch": 0.4898526315789474,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.000266135415531005,
+      "loss": 3.0403,
+      "step": 11634
+    },
+    {
+      "epoch": 0.48989473684210527,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00026610174145850736,
+      "loss": 3.0099,
+      "step": 11635
+    },
+    {
+      "epoch": 0.48993684210526317,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.000266068067092655,
+      "loss": 3.6336,
+      "step": 11636
+    },
+    {
+      "epoch": 0.48997894736842107,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002660343924340614,
+      "loss": 2.8436,
+      "step": 11637
+    },
+    {
+      "epoch": 0.49002105263157897,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00026600071748333984,
+      "loss": 2.9269,
+      "step": 11638
+    },
+    {
+      "epoch": 0.49006315789473687,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002659670422411042,
+      "loss": 2.776,
+      "step": 11639
+    },
+    {
+      "epoch": 0.4901052631578947,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002659333667079677,
+      "loss": 3.1351,
+      "step": 11640
+    },
+    {
+      "epoch": 0.4901473684210526,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.000265899690884544,
+      "loss": 3.032,
+      "step": 11641
+    },
+    {
+      "epoch": 0.4901894736842105,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00026586601477144656,
+      "loss": 3.2799,
+      "step": 11642
+    },
+    {
+      "epoch": 0.4902315789473684,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.000265832338369289,
+      "loss": 2.9265,
+      "step": 11643
+    },
+    {
+      "epoch": 0.4902736842105263,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00026579866167868486,
+      "loss": 3.6595,
+      "step": 11644
+    },
+    {
+      "epoch": 0.4903157894736842,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002657649847002477,
+      "loss": 3.282,
+      "step": 11645
+    },
+    {
+      "epoch": 0.4903578947368421,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00026573130743459103,
+      "loss": 3.1675,
+      "step": 11646
+    },
+    {
+      "epoch": 0.4904,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002656976298823284,
+      "loss": 3.0274,
+      "step": 11647
+    },
+    {
+      "epoch": 0.4904421052631579,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00026566395204407337,
+      "loss": 2.991,
+      "step": 11648
+    },
+    {
+      "epoch": 0.4904842105263158,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00026563027392043954,
+      "loss": 2.6073,
+      "step": 11649
+    },
+    {
+      "epoch": 0.4905263157894737,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002655965955120405,
+      "loss": 2.9554,
+      "step": 11650
+    },
+    {
+      "epoch": 0.4905684210526316,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002655629168194898,
+      "loss": 2.9446,
+      "step": 11651
+    },
+    {
+      "epoch": 0.4906105263157895,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0002655292378434011,
+      "loss": 2.9311,
+      "step": 11652
+    },
+    {
+      "epoch": 0.4906526315789474,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002654955585843879,
+      "loss": 3.2881,
+      "step": 11653
+    },
+    {
+      "epoch": 0.49069473684210524,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00026546187904306386,
+      "loss": 2.8006,
+      "step": 11654
+    },
+    {
+      "epoch": 0.49073684210526314,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002654281992200425,
+      "loss": 3.1704,
+      "step": 11655
+    },
+    {
+      "epoch": 0.49077894736842104,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002653945191159375,
+      "loss": 3.3464,
+      "step": 11656
+    },
+    {
+      "epoch": 0.49082105263157894,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00026536083873136243,
+      "loss": 3.0944,
+      "step": 11657
+    },
+    {
+      "epoch": 0.49086315789473683,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00026532715806693095,
+      "loss": 2.3493,
+      "step": 11658
+    },
+    {
+      "epoch": 0.49090526315789473,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00026529347712325664,
+      "loss": 3.2935,
+      "step": 11659
+    },
+    {
+      "epoch": 0.49094736842105263,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002652597959009531,
+      "loss": 3.1645,
+      "step": 11660
+    },
+    {
+      "epoch": 0.49098947368421053,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002652261144006341,
+      "loss": 3.3046,
+      "step": 11661
+    },
+    {
+      "epoch": 0.49103157894736843,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00026519243262291317,
+      "loss": 2.8383,
+      "step": 11662
+    },
+    {
+      "epoch": 0.49107368421052633,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00026515875056840396,
+      "loss": 3.2705,
+      "step": 11663
+    },
+    {
+      "epoch": 0.49111578947368423,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002651250682377201,
+      "loss": 3.4802,
+      "step": 11664
+    },
+    {
+      "epoch": 0.4911578947368421,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002650913856314753,
+      "loss": 2.9501,
+      "step": 11665
+    },
+    {
+      "epoch": 0.4912,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00026505770275028317,
+      "loss": 3.1806,
+      "step": 11666
+    },
+    {
+      "epoch": 0.49124210526315787,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002650240195947573,
+      "loss": 3.0393,
+      "step": 11667
+    },
+    {
+      "epoch": 0.49128421052631577,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00026499033616551155,
+      "loss": 3.4221,
+      "step": 11668
+    },
+    {
+      "epoch": 0.49132631578947367,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00026495665246315946,
+      "loss": 3.484,
+      "step": 11669
+    },
+    {
+      "epoch": 0.49136842105263157,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002649229684883148,
+      "loss": 3.3588,
+      "step": 11670
+    },
+    {
+      "epoch": 0.49141052631578946,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00026488928424159107,
+      "loss": 3.2207,
+      "step": 11671
+    },
+    {
+      "epoch": 0.49145263157894736,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002648555997236022,
+      "loss": 2.8685,
+      "step": 11672
+    },
+    {
+      "epoch": 0.49149473684210526,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00026482191493496166,
+      "loss": 3.5406,
+      "step": 11673
+    },
+    {
+      "epoch": 0.49153684210526316,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00026478822987628333,
+      "loss": 3.3571,
+      "step": 11674
+    },
+    {
+      "epoch": 0.49157894736842106,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002647545445481807,
+      "loss": 3.335,
+      "step": 11675
+    },
+    {
+      "epoch": 0.49162105263157896,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00026472085895126767,
+      "loss": 3.2036,
+      "step": 11676
+    },
+    {
+      "epoch": 0.49166315789473686,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002646871730861579,
+      "loss": 3.0012,
+      "step": 11677
+    },
+    {
+      "epoch": 0.49170526315789476,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002646534869534651,
+      "loss": 3.1464,
+      "step": 11678
+    },
+    {
+      "epoch": 0.49174736842105266,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00026461980055380294,
+      "loss": 3.1173,
+      "step": 11679
+    },
+    {
+      "epoch": 0.4917894736842105,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002645861138877853,
+      "loss": 2.8559,
+      "step": 11680
+    },
+    {
+      "epoch": 0.4918315789473684,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002645524269560257,
+      "loss": 3.3412,
+      "step": 11681
+    },
+    {
+      "epoch": 0.4918736842105263,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.000264518739759138,
+      "loss": 2.6561,
+      "step": 11682
+    },
+    {
+      "epoch": 0.4919157894736842,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002644850522977359,
+      "loss": 2.8107,
+      "step": 11683
+    },
+    {
+      "epoch": 0.4919578947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00026445136457243324,
+      "loss": 3.1022,
+      "step": 11684
+    },
+    {
+      "epoch": 0.492,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002644176765838436,
+      "loss": 3.2419,
+      "step": 11685
+    },
+    {
+      "epoch": 0.4920421052631579,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002643839883325809,
+      "loss": 3.4385,
+      "step": 11686
+    },
+    {
+      "epoch": 0.4920842105263158,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00026435029981925883,
+      "loss": 2.5535,
+      "step": 11687
+    },
+    {
+      "epoch": 0.4921263157894737,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00026431661104449116,
+      "loss": 3.0868,
+      "step": 11688
+    },
+    {
+      "epoch": 0.4921684210526316,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002642829220088916,
+      "loss": 3.3267,
+      "step": 11689
+    },
+    {
+      "epoch": 0.4922105263157895,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00026424923271307405,
+      "loss": 3.0876,
+      "step": 11690
+    },
+    {
+      "epoch": 0.4922526315789474,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002642155431576522,
+      "loss": 3.0237,
+      "step": 11691
+    },
+    {
+      "epoch": 0.4922947368421053,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002641818533432399,
+      "loss": 2.9536,
+      "step": 11692
+    },
+    {
+      "epoch": 0.4923368421052632,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00026414816327045084,
+      "loss": 3.0437,
+      "step": 11693
+    },
+    {
+      "epoch": 0.492378947368421,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00026411447293989886,
+      "loss": 3.1427,
+      "step": 11694
+    },
+    {
+      "epoch": 0.4924210526315789,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002640807823521978,
+      "loss": 3.5289,
+      "step": 11695
+    },
+    {
+      "epoch": 0.4924631578947368,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00026404709150796137,
+      "loss": 3.1923,
+      "step": 11696
+    },
+    {
+      "epoch": 0.4925052631578947,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002640134004078035,
+      "loss": 2.9184,
+      "step": 11697
+    },
+    {
+      "epoch": 0.4925473684210526,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002639797090523379,
+      "loss": 3.0645,
+      "step": 11698
+    },
+    {
+      "epoch": 0.4925894736842105,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002639460174421785,
+      "loss": 3.2219,
+      "step": 11699
+    },
+    {
+      "epoch": 0.4926315789473684,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.000263912325577939,
+      "loss": 3.1886,
+      "step": 11700
+    },
+    {
+      "epoch": 0.4926736842105263,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002638786334602332,
+      "loss": 2.7134,
+      "step": 11701
+    },
+    {
+      "epoch": 0.4927157894736842,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002638449410896751,
+      "loss": 2.821,
+      "step": 11702
+    },
+    {
+      "epoch": 0.4927578947368421,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002638112484668784,
+      "loss": 3.1285,
+      "step": 11703
+    },
+    {
+      "epoch": 0.4928,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00026377755559245706,
+      "loss": 2.9273,
+      "step": 11704
+    },
+    {
+      "epoch": 0.4928421052631579,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00026374386246702476,
+      "loss": 3.5778,
+      "step": 11705
+    },
+    {
+      "epoch": 0.4928842105263158,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002637101690911954,
+      "loss": 2.8352,
+      "step": 11706
+    },
+    {
+      "epoch": 0.49292631578947366,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.000263676475465583,
+      "loss": 3.4082,
+      "step": 11707
+    },
+    {
+      "epoch": 0.49296842105263156,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002636427815908012,
+      "loss": 2.9746,
+      "step": 11708
+    },
+    {
+      "epoch": 0.49301052631578945,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002636090874674639,
+      "loss": 3.482,
+      "step": 11709
+    },
+    {
+      "epoch": 0.49305263157894735,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00026357539309618505,
+      "loss": 2.7955,
+      "step": 11710
+    },
+    {
+      "epoch": 0.49309473684210525,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002635416984775785,
+      "loss": 2.7271,
+      "step": 11711
+    },
+    {
+      "epoch": 0.49313684210526315,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002635080036122582,
+      "loss": 3.7043,
+      "step": 11712
+    },
+    {
+      "epoch": 0.49317894736842105,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00026347430850083786,
+      "loss": 3.3319,
+      "step": 11713
+    },
+    {
+      "epoch": 0.49322105263157895,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002634406131439315,
+      "loss": 3.0794,
+      "step": 11714
+    },
+    {
+      "epoch": 0.49326315789473685,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.000263406917542153,
+      "loss": 3.3191,
+      "step": 11715
+    },
+    {
+      "epoch": 0.49330526315789475,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00026337322169611607,
+      "loss": 3.4763,
+      "step": 11716
+    },
+    {
+      "epoch": 0.49334736842105265,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002633395256064348,
+      "loss": 3.601,
+      "step": 11717
+    },
+    {
+      "epoch": 0.49338947368421054,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.000263305829273723,
+      "loss": 2.9146,
+      "step": 11718
+    },
+    {
+      "epoch": 0.49343157894736844,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00026327213269859473,
+      "loss": 2.6144,
+      "step": 11719
+    },
+    {
+      "epoch": 0.49347368421052634,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002632384358816638,
+      "loss": 3.3835,
+      "step": 11720
+    },
+    {
+      "epoch": 0.4935157894736842,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002632047388235441,
+      "loss": 3.0755,
+      "step": 11721
+    },
+    {
+      "epoch": 0.4935578947368421,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00026317104152484956,
+      "loss": 2.9746,
+      "step": 11722
+    },
+    {
+      "epoch": 0.4936,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002631373439861941,
+      "loss": 3.1699,
+      "step": 11723
+    },
+    {
+      "epoch": 0.4936421052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00026310364620819166,
+      "loss": 3.038,
+      "step": 11724
+    },
+    {
+      "epoch": 0.4936842105263158,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00026306994819145623,
+      "loss": 2.8827,
+      "step": 11725
+    },
+    {
+      "epoch": 0.4937263157894737,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002630362499366017,
+      "loss": 2.8776,
+      "step": 11726
+    },
+    {
+      "epoch": 0.4937684210526316,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00026300255144424197,
+      "loss": 3.3501,
+      "step": 11727
+    },
+    {
+      "epoch": 0.4938105263157895,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00026296885271499105,
+      "loss": 3.0118,
+      "step": 11728
+    },
+    {
+      "epoch": 0.4938526315789474,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002629351537494628,
+      "loss": 3.0965,
+      "step": 11729
+    },
+    {
+      "epoch": 0.4938947368421053,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002629014545482714,
+      "loss": 3.1056,
+      "step": 11730
+    },
+    {
+      "epoch": 0.4939368421052632,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00026286775511203054,
+      "loss": 3.1787,
+      "step": 11731
+    },
+    {
+      "epoch": 0.4939789473684211,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00026283405544135435,
+      "loss": 3.2477,
+      "step": 11732
+    },
+    {
+      "epoch": 0.49402105263157897,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002628003555368568,
+      "loss": 2.9984,
+      "step": 11733
+    },
+    {
+      "epoch": 0.4940631578947368,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002627666553991517,
+      "loss": 2.4976,
+      "step": 11734
+    },
+    {
+      "epoch": 0.4941052631578947,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00026273295502885315,
+      "loss": 2.9531,
+      "step": 11735
+    },
+    {
+      "epoch": 0.4941473684210526,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0002626992544265751,
+      "loss": 3.2128,
+      "step": 11736
+    },
+    {
+      "epoch": 0.4941894736842105,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002626655535929317,
+      "loss": 3.0804,
+      "step": 11737
+    },
+    {
+      "epoch": 0.4942315789473684,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002626318525285367,
+      "loss": 3.0013,
+      "step": 11738
+    },
+    {
+      "epoch": 0.4942736842105263,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002625981512340042,
+      "loss": 3.5655,
+      "step": 11739
+    },
+    {
+      "epoch": 0.4943157894736842,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00026256444970994817,
+      "loss": 2.5667,
+      "step": 11740
+    },
+    {
+      "epoch": 0.4943578947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00026253074795698267,
+      "loss": 3.4955,
+      "step": 11741
+    },
+    {
+      "epoch": 0.4944,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00026249704597572165,
+      "loss": 3.1781,
+      "step": 11742
+    },
+    {
+      "epoch": 0.4944421052631579,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00026246334376677904,
+      "loss": 2.9139,
+      "step": 11743
+    },
+    {
+      "epoch": 0.4944842105263158,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00026242964133076904,
+      "loss": 2.9561,
+      "step": 11744
+    },
+    {
+      "epoch": 0.4945263157894737,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002623959386683056,
+      "loss": 3.4379,
+      "step": 11745
+    },
+    {
+      "epoch": 0.4945684210526316,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00026236223578000265,
+      "loss": 3.5519,
+      "step": 11746
+    },
+    {
+      "epoch": 0.4946105263157895,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002623285326664744,
+      "loss": 3.1626,
+      "step": 11747
+    },
+    {
+      "epoch": 0.49465263157894734,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00026229482932833473,
+      "loss": 3.4296,
+      "step": 11748
+    },
+    {
+      "epoch": 0.49469473684210524,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00026226112576619766,
+      "loss": 3.5605,
+      "step": 11749
+    },
+    {
+      "epoch": 0.49473684210526314,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00026222742198067726,
+      "loss": 3.4966,
+      "step": 11750
+    },
+    {
+      "epoch": 0.49477894736842104,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002621937179723877,
+      "loss": 2.9615,
+      "step": 11751
+    },
+    {
+      "epoch": 0.49482105263157894,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002621600137419428,
+      "loss": 3.2592,
+      "step": 11752
+    },
+    {
+      "epoch": 0.49486315789473684,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002621263092899568,
+      "loss": 3.3818,
+      "step": 11753
+    },
+    {
+      "epoch": 0.49490526315789474,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002620926046170438,
+      "loss": 3.4676,
+      "step": 11754
+    },
+    {
+      "epoch": 0.49494736842105264,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00026205889972381766,
+      "loss": 3.3511,
+      "step": 11755
+    },
+    {
+      "epoch": 0.49498947368421053,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002620251946108925,
+      "loss": 2.9201,
+      "step": 11756
+    },
+    {
+      "epoch": 0.49503157894736843,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00026199148927888243,
+      "loss": 2.9879,
+      "step": 11757
+    },
+    {
+      "epoch": 0.49507368421052633,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00026195778372840157,
+      "loss": 3.1764,
+      "step": 11758
+    },
+    {
+      "epoch": 0.49511578947368423,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00026192407796006394,
+      "loss": 3.0332,
+      "step": 11759
+    },
+    {
+      "epoch": 0.49515789473684213,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002618903719744836,
+      "loss": 3.1114,
+      "step": 11760
+    },
+    {
+      "epoch": 0.4952,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00026185666577227464,
+      "loss": 3.2835,
+      "step": 11761
+    },
+    {
+      "epoch": 0.4952421052631579,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00026182295935405116,
+      "loss": 3.1296,
+      "step": 11762
+    },
+    {
+      "epoch": 0.49528421052631577,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00026178925272042717,
+      "loss": 2.964,
+      "step": 11763
+    },
+    {
+      "epoch": 0.49532631578947367,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.000261755545872017,
+      "loss": 2.9682,
+      "step": 11764
+    },
+    {
+      "epoch": 0.49536842105263157,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00026172183880943453,
+      "loss": 3.1118,
+      "step": 11765
+    },
+    {
+      "epoch": 0.49541052631578947,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00026168813153329396,
+      "loss": 3.347,
+      "step": 11766
+    },
+    {
+      "epoch": 0.49545263157894737,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00026165442404420936,
+      "loss": 3.2379,
+      "step": 11767
+    },
+    {
+      "epoch": 0.49549473684210527,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00026162071634279477,
+      "loss": 2.97,
+      "step": 11768
+    },
+    {
+      "epoch": 0.49553684210526316,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00026158700842966444,
+      "loss": 2.6273,
+      "step": 11769
+    },
+    {
+      "epoch": 0.49557894736842106,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002615533003054324,
+      "loss": 2.9531,
+      "step": 11770
+    },
+    {
+      "epoch": 0.49562105263157896,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00026151959197071283,
+      "loss": 2.8991,
+      "step": 11771
+    },
+    {
+      "epoch": 0.49566315789473686,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00026148588342611985,
+      "loss": 3.0404,
+      "step": 11772
+    },
+    {
+      "epoch": 0.49570526315789476,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00026145217467226755,
+      "loss": 3.4929,
+      "step": 11773
+    },
+    {
+      "epoch": 0.49574736842105266,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00026141846570977014,
+      "loss": 2.8661,
+      "step": 11774
+    },
+    {
+      "epoch": 0.4957894736842105,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00026138475653924166,
+      "loss": 3.3141,
+      "step": 11775
+    },
+    {
+      "epoch": 0.4958315789473684,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002613510471612962,
+      "loss": 3.2719,
+      "step": 11776
+    },
+    {
+      "epoch": 0.4958736842105263,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002613173375765481,
+      "loss": 2.9511,
+      "step": 11777
+    },
+    {
+      "epoch": 0.4959157894736842,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002612836277856113,
+      "loss": 3.0608,
+      "step": 11778
+    },
+    {
+      "epoch": 0.4959578947368421,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00026124991778910016,
+      "loss": 2.9175,
+      "step": 11779
+    },
+    {
+      "epoch": 0.496,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00026121620758762877,
+      "loss": 3.3419,
+      "step": 11780
+    },
+    {
+      "epoch": 0.4960421052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002611824971818112,
+      "loss": 3.0833,
+      "step": 11781
+    },
+    {
+      "epoch": 0.4960842105263158,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002611487865722617,
+      "loss": 3.0832,
+      "step": 11782
+    },
+    {
+      "epoch": 0.4961263157894737,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0002611150757595943,
+      "loss": 2.9122,
+      "step": 11783
+    },
+    {
+      "epoch": 0.4961684210526316,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002610813647444234,
+      "loss": 2.7323,
+      "step": 11784
+    },
+    {
+      "epoch": 0.4962105263157895,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.000261047653527363,
+      "loss": 3.015,
+      "step": 11785
+    },
+    {
+      "epoch": 0.4962526315789474,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00026101394210902734,
+      "loss": 3.2002,
+      "step": 11786
+    },
+    {
+      "epoch": 0.4962947368421053,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002609802304900306,
+      "loss": 3.1196,
+      "step": 11787
+    },
+    {
+      "epoch": 0.49633684210526313,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.000260946518670987,
+      "loss": 3.1333,
+      "step": 11788
+    },
+    {
+      "epoch": 0.49637894736842103,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00026091280665251066,
+      "loss": 3.5919,
+      "step": 11789
+    },
+    {
+      "epoch": 0.49642105263157893,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002608790944352157,
+      "loss": 2.9109,
+      "step": 11790
+    },
+    {
+      "epoch": 0.49646315789473683,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002608453820197166,
+      "loss": 3.1266,
+      "step": 11791
+    },
+    {
+      "epoch": 0.4965052631578947,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002608116694066273,
+      "loss": 3.3968,
+      "step": 11792
+    },
+    {
+      "epoch": 0.4965473684210526,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002607779565965621,
+      "loss": 3.098,
+      "step": 11793
+    },
+    {
+      "epoch": 0.4965894736842105,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00026074424359013513,
+      "loss": 3.2611,
+      "step": 11794
+    },
+    {
+      "epoch": 0.4966315789473684,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00026071053038796075,
+      "loss": 3.5618,
+      "step": 11795
+    },
+    {
+      "epoch": 0.4966736842105263,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002606768169906531,
+      "loss": 3.5454,
+      "step": 11796
+    },
+    {
+      "epoch": 0.4967157894736842,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002606431033988263,
+      "loss": 2.9509,
+      "step": 11797
+    },
+    {
+      "epoch": 0.4967578947368421,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00026060938961309474,
+      "loss": 3.616,
+      "step": 11798
+    },
+    {
+      "epoch": 0.4968,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00026057567563407254,
+      "loss": 2.7526,
+      "step": 11799
+    },
+    {
+      "epoch": 0.4968421052631579,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00026054196146237395,
+      "loss": 3.2414,
+      "step": 11800
+    },
+    {
+      "epoch": 0.4968842105263158,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002605082470986133,
+      "loss": 3.4821,
+      "step": 11801
+    },
+    {
+      "epoch": 0.49692631578947366,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00026047453254340457,
+      "loss": 3.0048,
+      "step": 11802
+    },
+    {
+      "epoch": 0.49696842105263156,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002604408177973622,
+      "loss": 3.2276,
+      "step": 11803
+    },
+    {
+      "epoch": 0.49701052631578946,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00026040710286110043,
+      "loss": 3.0849,
+      "step": 11804
+    },
+    {
+      "epoch": 0.49705263157894736,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002603733877352335,
+      "loss": 3.2406,
+      "step": 11805
+    },
+    {
+      "epoch": 0.49709473684210526,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002603396724203757,
+      "loss": 3.1194,
+      "step": 11806
+    },
+    {
+      "epoch": 0.49713684210526315,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002603059569171411,
+      "loss": 3.3074,
+      "step": 11807
+    },
+    {
+      "epoch": 0.49717894736842105,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002602722412261441,
+      "loss": 3.1868,
+      "step": 11808
+    },
+    {
+      "epoch": 0.49722105263157895,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.000260238525347999,
+      "loss": 3.376,
+      "step": 11809
+    },
+    {
+      "epoch": 0.49726315789473685,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002602048092833199,
+      "loss": 3.4394,
+      "step": 11810
+    },
+    {
+      "epoch": 0.49730526315789475,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00026017109303272113,
+      "loss": 3.0019,
+      "step": 11811
+    },
+    {
+      "epoch": 0.49734736842105265,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00026013737659681706,
+      "loss": 3.1913,
+      "step": 11812
+    },
+    {
+      "epoch": 0.49738947368421055,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002601036599762219,
+      "loss": 3.1737,
+      "step": 11813
+    },
+    {
+      "epoch": 0.49743157894736845,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00026006994317155,
+      "loss": 3.2834,
+      "step": 11814
+    },
+    {
+      "epoch": 0.4974736842105263,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00026003622618341553,
+      "loss": 3.6543,
+      "step": 11815
+    },
+    {
+      "epoch": 0.4975157894736842,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002600025090124328,
+      "loss": 3.3228,
+      "step": 11816
+    },
+    {
+      "epoch": 0.4975578947368421,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00025996879165921606,
+      "loss": 3.434,
+      "step": 11817
+    },
+    {
+      "epoch": 0.4976,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002599350741243797,
+      "loss": 3.4325,
+      "step": 11818
+    },
+    {
+      "epoch": 0.4976421052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002599013564085379,
+      "loss": 3.0194,
+      "step": 11819
+    },
+    {
+      "epoch": 0.4976842105263158,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00025986763851230504,
+      "loss": 2.9731,
+      "step": 11820
+    },
+    {
+      "epoch": 0.4977263157894737,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00025983392043629544,
+      "loss": 3.3173,
+      "step": 11821
+    },
+    {
+      "epoch": 0.4977684210526316,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002598002021811234,
+      "loss": 3.2137,
+      "step": 11822
+    },
+    {
+      "epoch": 0.4978105263157895,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002597664837474031,
+      "loss": 3.4264,
+      "step": 11823
+    },
+    {
+      "epoch": 0.4978526315789474,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00025973276513574896,
+      "loss": 3.1748,
+      "step": 11824
+    },
+    {
+      "epoch": 0.4978947368421053,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00025969904634677533,
+      "loss": 2.7568,
+      "step": 11825
+    },
+    {
+      "epoch": 0.4979368421052632,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002596653273810964,
+      "loss": 3.4514,
+      "step": 11826
+    },
+    {
+      "epoch": 0.4979789473684211,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002596316082393266,
+      "loss": 3.0647,
+      "step": 11827
+    },
+    {
+      "epoch": 0.4980210526315789,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00025959788892208023,
+      "loss": 3.1129,
+      "step": 11828
+    },
+    {
+      "epoch": 0.4980631578947368,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00025956416942997157,
+      "loss": 3.181,
+      "step": 11829
+    },
+    {
+      "epoch": 0.4981052631578947,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.000259530449763615,
+      "loss": 3.2907,
+      "step": 11830
+    },
+    {
+      "epoch": 0.4981473684210526,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002594967299236248,
+      "loss": 3.1404,
+      "step": 11831
+    },
+    {
+      "epoch": 0.4981894736842105,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002594630099106154,
+      "loss": 3.0203,
+      "step": 11832
+    },
+    {
+      "epoch": 0.4982315789473684,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00025942928972520106,
+      "loss": 3.284,
+      "step": 11833
+    },
+    {
+      "epoch": 0.4982736842105263,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00025939556936799616,
+      "loss": 2.9613,
+      "step": 11834
+    },
+    {
+      "epoch": 0.4983157894736842,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00025936184883961497,
+      "loss": 3.2107,
+      "step": 11835
+    },
+    {
+      "epoch": 0.4983578947368421,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002593281281406719,
+      "loss": 3.4872,
+      "step": 11836
+    },
+    {
+      "epoch": 0.4984,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00025929440727178127,
+      "loss": 2.4908,
+      "step": 11837
+    },
+    {
+      "epoch": 0.4984421052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00025926068623355754,
+      "loss": 3.2222,
+      "step": 11838
+    },
+    {
+      "epoch": 0.4984842105263158,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002592269650266149,
+      "loss": 3.5823,
+      "step": 11839
+    },
+    {
+      "epoch": 0.4985263157894737,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002591932436515679,
+      "loss": 3.4019,
+      "step": 11840
+    },
+    {
+      "epoch": 0.4985684210526316,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002591595221090308,
+      "loss": 3.2646,
+      "step": 11841
+    },
+    {
+      "epoch": 0.49861052631578945,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00025912580039961786,
+      "loss": 3.4698,
+      "step": 11842
+    },
+    {
+      "epoch": 0.49865263157894735,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00025909207852394363,
+      "loss": 3.3831,
+      "step": 11843
+    },
+    {
+      "epoch": 0.49869473684210525,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002590583564826224,
+      "loss": 2.6153,
+      "step": 11844
+    },
+    {
+      "epoch": 0.49873684210526315,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002590246342762685,
+      "loss": 2.8707,
+      "step": 11845
+    },
+    {
+      "epoch": 0.49877894736842104,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00025899091190549644,
+      "loss": 3.2202,
+      "step": 11846
+    },
+    {
+      "epoch": 0.49882105263157894,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002589571893709205,
+      "loss": 3.2959,
+      "step": 11847
+    },
+    {
+      "epoch": 0.49886315789473684,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00025892346667315513,
+      "loss": 3.4278,
+      "step": 11848
+    },
+    {
+      "epoch": 0.49890526315789474,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00025888974381281467,
+      "loss": 3.1796,
+      "step": 11849
+    },
+    {
+      "epoch": 0.49894736842105264,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00025885602079051353,
+      "loss": 2.8782,
+      "step": 11850
+    },
+    {
+      "epoch": 0.49898947368421054,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00025882229760686604,
+      "loss": 2.8731,
+      "step": 11851
+    },
+    {
+      "epoch": 0.49903157894736844,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002587885742624867,
+      "loss": 3.3547,
+      "step": 11852
+    },
+    {
+      "epoch": 0.49907368421052634,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002587548507579898,
+      "loss": 3.0811,
+      "step": 11853
+    },
+    {
+      "epoch": 0.49911578947368423,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00025872112709398985,
+      "loss": 3.2308,
+      "step": 11854
+    },
+    {
+      "epoch": 0.4991578947368421,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002586874032711013,
+      "loss": 3.591,
+      "step": 11855
+    },
+    {
+      "epoch": 0.4992,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002586536792899384,
+      "loss": 2.988,
+      "step": 11856
+    },
+    {
+      "epoch": 0.4992421052631579,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00025861995515111556,
+      "loss": 3.383,
+      "step": 11857
+    },
+    {
+      "epoch": 0.4992842105263158,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00025858623085524735,
+      "loss": 2.7262,
+      "step": 11858
+    },
+    {
+      "epoch": 0.4993263157894737,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00025855250640294806,
+      "loss": 2.8443,
+      "step": 11859
+    },
+    {
+      "epoch": 0.4993684210526316,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002585187817948322,
+      "loss": 3.2173,
+      "step": 11860
+    },
+    {
+      "epoch": 0.49941052631578947,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002584850570315141,
+      "loss": 3.369,
+      "step": 11861
+    },
+    {
+      "epoch": 0.49945263157894737,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00025845133211360824,
+      "loss": 3.1172,
+      "step": 11862
+    },
+    {
+      "epoch": 0.49949473684210527,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00025841760704172905,
+      "loss": 2.3914,
+      "step": 11863
+    },
+    {
+      "epoch": 0.49953684210526317,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002583838818164909,
+      "loss": 3.3364,
+      "step": 11864
+    },
+    {
+      "epoch": 0.49957894736842107,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002583501564385083,
+      "loss": 3.0707,
+      "step": 11865
+    },
+    {
+      "epoch": 0.49962105263157897,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00025831643090839574,
+      "loss": 3.2036,
+      "step": 11866
+    },
+    {
+      "epoch": 0.49966315789473686,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00025828270522676753,
+      "loss": 2.9932,
+      "step": 11867
+    },
+    {
+      "epoch": 0.49970526315789476,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00025824897939423817,
+      "loss": 3.3564,
+      "step": 11868
+    },
+    {
+      "epoch": 0.4997473684210526,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00025821525341142206,
+      "loss": 2.5034,
+      "step": 11869
+    },
+    {
+      "epoch": 0.4997894736842105,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002581815272789337,
+      "loss": 3.0827,
+      "step": 11870
+    },
+    {
+      "epoch": 0.4998315789473684,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00025814780099738747,
+      "loss": 3.2617,
+      "step": 11871
+    },
+    {
+      "epoch": 0.4998736842105263,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.000258114074567398,
+      "loss": 3.366,
+      "step": 11872
+    },
+    {
+      "epoch": 0.4999157894736842,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002580803479895796,
+      "loss": 2.9887,
+      "step": 11873
+    },
+    {
+      "epoch": 0.4999578947368421,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002580466212645467,
+      "loss": 3.4758,
+      "step": 11874
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00025801289439291385,
+      "loss": 2.5793,
+      "step": 11875
+    },
+    {
+      "epoch": 0.5000421052631578,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00025797916737529546,
+      "loss": 3.5292,
+      "step": 11876
+    },
+    {
+      "epoch": 0.5000842105263158,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.000257945440212306,
+      "loss": 3.0543,
+      "step": 11877
+    },
+    {
+      "epoch": 0.5001263157894736,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00025791171290455987,
+      "loss": 3.3211,
+      "step": 11878
+    },
+    {
+      "epoch": 0.5001684210526316,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002578779854526718,
+      "loss": 3.1241,
+      "step": 11879
+    },
+    {
+      "epoch": 0.5002105263157894,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.000257844257857256,
+      "loss": 3.6102,
+      "step": 11880
+    },
+    {
+      "epoch": 0.5002526315789474,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002578105301189271,
+      "loss": 3.3974,
+      "step": 11881
+    },
+    {
+      "epoch": 0.5002947368421052,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002577768022382995,
+      "loss": 3.3197,
+      "step": 11882
+    },
+    {
+      "epoch": 0.5003368421052632,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002577430742159877,
+      "loss": 2.8323,
+      "step": 11883
+    },
+    {
+      "epoch": 0.500378947368421,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002577093460526061,
+      "loss": 2.7986,
+      "step": 11884
+    },
+    {
+      "epoch": 0.500421052631579,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00025767561774876936,
+      "loss": 3.0178,
+      "step": 11885
+    },
+    {
+      "epoch": 0.5004631578947368,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002576418893050919,
+      "loss": 3.3002,
+      "step": 11886
+    },
+    {
+      "epoch": 0.5005052631578948,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00025760816072218804,
+      "loss": 3.0102,
+      "step": 11887
+    },
+    {
+      "epoch": 0.5005473684210526,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002575744320006726,
+      "loss": 3.3164,
+      "step": 11888
+    },
+    {
+      "epoch": 0.5005894736842105,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002575407031411599,
+      "loss": 2.9908,
+      "step": 11889
+    },
+    {
+      "epoch": 0.5006315789473684,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002575069741442644,
+      "loss": 3.0685,
+      "step": 11890
+    },
+    {
+      "epoch": 0.5006736842105263,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002574732450106006,
+      "loss": 2.9118,
+      "step": 11891
+    },
+    {
+      "epoch": 0.5007157894736842,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00025743951574078313,
+      "loss": 3.1362,
+      "step": 11892
+    },
+    {
+      "epoch": 0.5007578947368421,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00025740578633542645,
+      "loss": 2.7773,
+      "step": 11893
+    },
+    {
+      "epoch": 0.5008,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.000257372056795145,
+      "loss": 3.1506,
+      "step": 11894
+    },
+    {
+      "epoch": 0.5008421052631579,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00025733832712055327,
+      "loss": 2.9039,
+      "step": 11895
+    },
+    {
+      "epoch": 0.5008842105263158,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.000257304597312266,
+      "loss": 2.9184,
+      "step": 11896
+    },
+    {
+      "epoch": 0.5009263157894737,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002572708673708974,
+      "loss": 2.8301,
+      "step": 11897
+    },
+    {
+      "epoch": 0.5009684210526316,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002572371372970622,
+      "loss": 2.8448,
+      "step": 11898
+    },
+    {
+      "epoch": 0.5010105263157895,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00025720340709137483,
+      "loss": 2.9318,
+      "step": 11899
+    },
+    {
+      "epoch": 0.5010526315789474,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002571696767544499,
+      "loss": 3.0695,
+      "step": 11900
+    },
+    {
+      "epoch": 0.5010947368421053,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002571359462869019,
+      "loss": 3.5978,
+      "step": 11901
+    },
+    {
+      "epoch": 0.5011368421052631,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00025710221568934534,
+      "loss": 3.0096,
+      "step": 11902
+    },
+    {
+      "epoch": 0.5011789473684211,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002570684849623947,
+      "loss": 3.462,
+      "step": 11903
+    },
+    {
+      "epoch": 0.5012210526315789,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002570347541066645,
+      "loss": 3.1149,
+      "step": 11904
+    },
+    {
+      "epoch": 0.5012631578947369,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00025700102312276946,
+      "loss": 2.7461,
+      "step": 11905
+    },
+    {
+      "epoch": 0.5013052631578947,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00025696729201132396,
+      "loss": 3.3615,
+      "step": 11906
+    },
+    {
+      "epoch": 0.5013473684210527,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002569335607729426,
+      "loss": 3.2187,
+      "step": 11907
+    },
+    {
+      "epoch": 0.5013894736842105,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00025689982940823995,
+      "loss": 2.4131,
+      "step": 11908
+    },
+    {
+      "epoch": 0.5014315789473685,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002568660979178305,
+      "loss": 2.9688,
+      "step": 11909
+    },
+    {
+      "epoch": 0.5014736842105263,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002568323663023288,
+      "loss": 2.9243,
+      "step": 11910
+    },
+    {
+      "epoch": 0.5015157894736842,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00025679863456234937,
+      "loss": 3.2165,
+      "step": 11911
+    },
+    {
+      "epoch": 0.5015578947368421,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002567649026985068,
+      "loss": 3.2529,
+      "step": 11912
+    },
+    {
+      "epoch": 0.5016,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002567311707114157,
+      "loss": 3.3742,
+      "step": 11913
+    },
+    {
+      "epoch": 0.5016421052631579,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00025669743860169057,
+      "loss": 3.3712,
+      "step": 11914
+    },
+    {
+      "epoch": 0.5016842105263158,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.000256663706369946,
+      "loss": 2.9002,
+      "step": 11915
+    },
+    {
+      "epoch": 0.5017263157894737,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00025662997401679655,
+      "loss": 3.1564,
+      "step": 11916
+    },
+    {
+      "epoch": 0.5017684210526315,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00025659624154285673,
+      "loss": 3.456,
+      "step": 11917
+    },
+    {
+      "epoch": 0.5018105263157895,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00025656250894874106,
+      "loss": 3.2188,
+      "step": 11918
+    },
+    {
+      "epoch": 0.5018526315789473,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00025652877623506424,
+      "loss": 3.2334,
+      "step": 11919
+    },
+    {
+      "epoch": 0.5018947368421053,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00025649504340244085,
+      "loss": 2.9544,
+      "step": 11920
+    },
+    {
+      "epoch": 0.5019368421052631,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00025646131045148536,
+      "loss": 3.4514,
+      "step": 11921
+    },
+    {
+      "epoch": 0.5019789473684211,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00025642757738281234,
+      "loss": 2.8682,
+      "step": 11922
+    },
+    {
+      "epoch": 0.5020210526315789,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002563938441970365,
+      "loss": 3.3967,
+      "step": 11923
+    },
+    {
+      "epoch": 0.5020631578947369,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002563601108947723,
+      "loss": 2.9892,
+      "step": 11924
+    },
+    {
+      "epoch": 0.5021052631578947,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002563263774766343,
+      "loss": 3.3888,
+      "step": 11925
+    },
+    {
+      "epoch": 0.5021473684210527,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002562926439432372,
+      "loss": 2.9315,
+      "step": 11926
+    },
+    {
+      "epoch": 0.5021894736842105,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002562589102951955,
+      "loss": 3.1879,
+      "step": 11927
+    },
+    {
+      "epoch": 0.5022315789473685,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002562251765331238,
+      "loss": 3.6553,
+      "step": 11928
+    },
+    {
+      "epoch": 0.5022736842105263,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002561914426576367,
+      "loss": 3.1002,
+      "step": 11929
+    },
+    {
+      "epoch": 0.5023157894736842,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002561577086693488,
+      "loss": 3.1038,
+      "step": 11930
+    },
+    {
+      "epoch": 0.5023578947368421,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002561239745688746,
+      "loss": 2.7897,
+      "step": 11931
+    },
+    {
+      "epoch": 0.5024,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002560902403568289,
+      "loss": 3.1365,
+      "step": 11932
+    },
+    {
+      "epoch": 0.5024421052631579,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002560565060338261,
+      "loss": 3.1951,
+      "step": 11933
+    },
+    {
+      "epoch": 0.5024842105263158,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00025602277160048095,
+      "loss": 3.7368,
+      "step": 11934
+    },
+    {
+      "epoch": 0.5025263157894737,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00025598903705740796,
+      "loss": 2.8116,
+      "step": 11935
+    },
+    {
+      "epoch": 0.5025684210526316,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002559553024052216,
+      "loss": 3.518,
+      "step": 11936
+    },
+    {
+      "epoch": 0.5026105263157895,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00025592156764453684,
+      "loss": 3.515,
+      "step": 11937
+    },
+    {
+      "epoch": 0.5026526315789474,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002558878327759679,
+      "loss": 3.3112,
+      "step": 11938
+    },
+    {
+      "epoch": 0.5026947368421053,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00025585409780012967,
+      "loss": 3.3916,
+      "step": 11939
+    },
+    {
+      "epoch": 0.5027368421052631,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002558203627176367,
+      "loss": 2.9375,
+      "step": 11940
+    },
+    {
+      "epoch": 0.5027789473684211,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002557866275291035,
+      "loss": 2.9133,
+      "step": 11941
+    },
+    {
+      "epoch": 0.502821052631579,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00025575289223514475,
+      "loss": 2.917,
+      "step": 11942
+    },
+    {
+      "epoch": 0.5028631578947368,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002557191568363751,
+      "loss": 3.0016,
+      "step": 11943
+    },
+    {
+      "epoch": 0.5029052631578947,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.000255685421333409,
+      "loss": 3.3195,
+      "step": 11944
+    },
+    {
+      "epoch": 0.5029473684210526,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00025565168572686133,
+      "loss": 3.2903,
+      "step": 11945
+    },
+    {
+      "epoch": 0.5029894736842105,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002556179500173465,
+      "loss": 3.6244,
+      "step": 11946
+    },
+    {
+      "epoch": 0.5030315789473684,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002555842142054793,
+      "loss": 3.1248,
+      "step": 11947
+    },
+    {
+      "epoch": 0.5030736842105263,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002555504782918742,
+      "loss": 3.5486,
+      "step": 11948
+    },
+    {
+      "epoch": 0.5031157894736842,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.000255516742277146,
+      "loss": 2.926,
+      "step": 11949
+    },
+    {
+      "epoch": 0.5031578947368421,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00025548300616190923,
+      "loss": 2.6506,
+      "step": 11950
+    },
+    {
+      "epoch": 0.5032,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00025544926994677846,
+      "loss": 2.9996,
+      "step": 11951
+    },
+    {
+      "epoch": 0.5032421052631579,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00025541553363236843,
+      "loss": 2.9144,
+      "step": 11952
+    },
+    {
+      "epoch": 0.5032842105263158,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0002553817972192937,
+      "loss": 2.6542,
+      "step": 11953
+    },
+    {
+      "epoch": 0.5033263157894737,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.000255348060708169,
+      "loss": 3.3758,
+      "step": 11954
+    },
+    {
+      "epoch": 0.5033684210526316,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00025531432409960894,
+      "loss": 3.543,
+      "step": 11955
+    },
+    {
+      "epoch": 0.5034105263157894,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00025528058739422813,
+      "loss": 2.9508,
+      "step": 11956
+    },
+    {
+      "epoch": 0.5034526315789474,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002552468505926411,
+      "loss": 3.195,
+      "step": 11957
+    },
+    {
+      "epoch": 0.5034947368421052,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002552131136954627,
+      "loss": 3.1182,
+      "step": 11958
+    },
+    {
+      "epoch": 0.5035368421052632,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00025517937670330756,
+      "loss": 3.4034,
+      "step": 11959
+    },
+    {
+      "epoch": 0.503578947368421,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002551456396167902,
+      "loss": 2.8364,
+      "step": 11960
+    },
+    {
+      "epoch": 0.503621052631579,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0002551119024365254,
+      "loss": 3.4364,
+      "step": 11961
+    },
+    {
+      "epoch": 0.5036631578947368,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002550781651631276,
+      "loss": 3.1737,
+      "step": 11962
+    },
+    {
+      "epoch": 0.5037052631578948,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002550444277972117,
+      "loss": 2.8238,
+      "step": 11963
+    },
+    {
+      "epoch": 0.5037473684210526,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002550106903393922,
+      "loss": 3.046,
+      "step": 11964
+    },
+    {
+      "epoch": 0.5037894736842106,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002549769527902838,
+      "loss": 3.4918,
+      "step": 11965
+    },
+    {
+      "epoch": 0.5038315789473684,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002549432151505013,
+      "loss": 3.7213,
+      "step": 11966
+    },
+    {
+      "epoch": 0.5038736842105264,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002549094774206591,
+      "loss": 2.9697,
+      "step": 11967
+    },
+    {
+      "epoch": 0.5039157894736842,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00025487573960137206,
+      "loss": 3.066,
+      "step": 11968
+    },
+    {
+      "epoch": 0.5039578947368422,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00025484200169325477,
+      "loss": 3.2236,
+      "step": 11969
+    },
+    {
+      "epoch": 0.504,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002548082636969218,
+      "loss": 3.4485,
+      "step": 11970
+    },
+    {
+      "epoch": 0.5040421052631578,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.000254774525612988,
+      "loss": 3.5289,
+      "step": 11971
+    },
+    {
+      "epoch": 0.5040842105263158,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002547407874420679,
+      "loss": 3.4352,
+      "step": 11972
+    },
+    {
+      "epoch": 0.5041263157894736,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002547070491847763,
+      "loss": 3.6419,
+      "step": 11973
+    },
+    {
+      "epoch": 0.5041684210526316,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00025467331084172776,
+      "loss": 3.2991,
+      "step": 11974
+    },
+    {
+      "epoch": 0.5042105263157894,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00025463957241353695,
+      "loss": 2.7614,
+      "step": 11975
+    },
+    {
+      "epoch": 0.5042526315789474,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00025460583390081866,
+      "loss": 3.1424,
+      "step": 11976
+    },
+    {
+      "epoch": 0.5042947368421052,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00025457209530418745,
+      "loss": 3.0214,
+      "step": 11977
+    },
+    {
+      "epoch": 0.5043368421052632,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.000254538356624258,
+      "loss": 3.3662,
+      "step": 11978
+    },
+    {
+      "epoch": 0.504378947368421,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00025450461786164495,
+      "loss": 3.1607,
+      "step": 11979
+    },
+    {
+      "epoch": 0.504421052631579,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002544708790169632,
+      "loss": 3.3502,
+      "step": 11980
+    },
+    {
+      "epoch": 0.5044631578947368,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002544371400908273,
+      "loss": 2.8075,
+      "step": 11981
+    },
+    {
+      "epoch": 0.5045052631578948,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00025440340108385184,
+      "loss": 3.174,
+      "step": 11982
+    },
+    {
+      "epoch": 0.5045473684210526,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002543696619966516,
+      "loss": 2.7678,
+      "step": 11983
+    },
+    {
+      "epoch": 0.5045894736842105,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002543359228298413,
+      "loss": 3.1831,
+      "step": 11984
+    },
+    {
+      "epoch": 0.5046315789473684,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002543021835840355,
+      "loss": 2.8303,
+      "step": 11985
+    },
+    {
+      "epoch": 0.5046736842105263,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00025426844425984905,
+      "loss": 3.5595,
+      "step": 11986
+    },
+    {
+      "epoch": 0.5047157894736842,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00025423470485789643,
+      "loss": 3.2853,
+      "step": 11987
+    },
+    {
+      "epoch": 0.5047578947368421,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00025420096537879264,
+      "loss": 2.78,
+      "step": 11988
+    },
+    {
+      "epoch": 0.5048,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002541672258231521,
+      "loss": 2.9894,
+      "step": 11989
+    },
+    {
+      "epoch": 0.5048421052631579,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002541334861915897,
+      "loss": 3.319,
+      "step": 11990
+    },
+    {
+      "epoch": 0.5048842105263158,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002540997464847199,
+      "loss": 3.0665,
+      "step": 11991
+    },
+    {
+      "epoch": 0.5049263157894737,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00025406600670315756,
+      "loss": 3.1564,
+      "step": 11992
+    },
+    {
+      "epoch": 0.5049684210526316,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002540322668475174,
+      "loss": 3.6432,
+      "step": 11993
+    },
+    {
+      "epoch": 0.5050105263157895,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002539985269184141,
+      "loss": 3.6628,
+      "step": 11994
+    },
+    {
+      "epoch": 0.5050526315789474,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00025396478691646225,
+      "loss": 3.0249,
+      "step": 11995
+    },
+    {
+      "epoch": 0.5050947368421053,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002539310468422767,
+      "loss": 3.1347,
+      "step": 11996
+    },
+    {
+      "epoch": 0.5051368421052631,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002538973066964721,
+      "loss": 3.2627,
+      "step": 11997
+    },
+    {
+      "epoch": 0.5051789473684211,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00025386356647966315,
+      "loss": 3.1881,
+      "step": 11998
+    },
+    {
+      "epoch": 0.5052210526315789,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002538298261924645,
+      "loss": 3.4062,
+      "step": 11999
+    },
+    {
+      "epoch": 0.5052631578947369,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.000253796085835491,
+      "loss": 3.5201,
+      "step": 12000
+    },
+    {
+      "epoch": 0.5052631578947369,
+      "eval_loss": 3.1277520656585693,
+      "eval_runtime": 335.5448,
+      "eval_samples_per_second": 44.703,
+      "eval_steps_per_second": 5.588,
+      "step": 12000
+    },
+    {
+      "epoch": 0.5053052631578947,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00025376234540935724,
+      "loss": 3.1084,
+      "step": 12001
+    },
+    {
+      "epoch": 0.5053473684210527,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00025372860491467797,
+      "loss": 3.5781,
+      "step": 12002
+    },
+    {
+      "epoch": 0.5053894736842105,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00025369486435206784,
+      "loss": 3.2449,
+      "step": 12003
+    },
+    {
+      "epoch": 0.5054315789473685,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002536611237221417,
+      "loss": 3.1852,
+      "step": 12004
+    },
+    {
+      "epoch": 0.5054736842105263,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002536273830255141,
+      "loss": 3.1418,
+      "step": 12005
+    },
+    {
+      "epoch": 0.5055157894736843,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00025359364226279995,
+      "loss": 3.1013,
+      "step": 12006
+    },
+    {
+      "epoch": 0.5055578947368421,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00025355990143461385,
+      "loss": 3.0224,
+      "step": 12007
+    },
+    {
+      "epoch": 0.5056,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002535261605415705,
+      "loss": 2.8494,
+      "step": 12008
+    },
+    {
+      "epoch": 0.5056421052631579,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00025349241958428465,
+      "loss": 3.1999,
+      "step": 12009
+    },
+    {
+      "epoch": 0.5056842105263157,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.000253458678563371,
+      "loss": 2.7035,
+      "step": 12010
+    },
+    {
+      "epoch": 0.5057263157894737,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00025342493747944426,
+      "loss": 2.6765,
+      "step": 12011
+    },
+    {
+      "epoch": 0.5057684210526315,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00025339119633311925,
+      "loss": 3.5958,
+      "step": 12012
+    },
+    {
+      "epoch": 0.5058105263157895,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00025335745512501055,
+      "loss": 3.2025,
+      "step": 12013
+    },
+    {
+      "epoch": 0.5058526315789473,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00025332371385573306,
+      "loss": 3.2225,
+      "step": 12014
+    },
+    {
+      "epoch": 0.5058947368421053,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002532899725259014,
+      "loss": 3.1904,
+      "step": 12015
+    },
+    {
+      "epoch": 0.5059368421052631,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002532562311361303,
+      "loss": 3.1173,
+      "step": 12016
+    },
+    {
+      "epoch": 0.5059789473684211,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002532224896870344,
+      "loss": 3.4095,
+      "step": 12017
+    },
+    {
+      "epoch": 0.5060210526315789,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00025318874817922864,
+      "loss": 3.4642,
+      "step": 12018
+    },
+    {
+      "epoch": 0.5060631578947369,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002531550066133276,
+      "loss": 3.3437,
+      "step": 12019
+    },
+    {
+      "epoch": 0.5061052631578947,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00025312126498994603,
+      "loss": 2.8973,
+      "step": 12020
+    },
+    {
+      "epoch": 0.5061473684210527,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002530875233096986,
+      "loss": 3.1795,
+      "step": 12021
+    },
+    {
+      "epoch": 0.5061894736842105,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00025305378157320025,
+      "loss": 2.9435,
+      "step": 12022
+    },
+    {
+      "epoch": 0.5062315789473684,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002530200397810655,
+      "loss": 3.1416,
+      "step": 12023
+    },
+    {
+      "epoch": 0.5062736842105263,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002529862979339093,
+      "loss": 3.2411,
+      "step": 12024
+    },
+    {
+      "epoch": 0.5063157894736842,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00025295255603234617,
+      "loss": 2.8795,
+      "step": 12025
+    },
+    {
+      "epoch": 0.5063578947368421,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002529188140769909,
+      "loss": 3.1479,
+      "step": 12026
+    },
+    {
+      "epoch": 0.5064,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00025288507206845835,
+      "loss": 3.1442,
+      "step": 12027
+    },
+    {
+      "epoch": 0.5064421052631579,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002528513300073631,
+      "loss": 2.9922,
+      "step": 12028
+    },
+    {
+      "epoch": 0.5064842105263158,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00025281758789432,
+      "loss": 2.7362,
+      "step": 12029
+    },
+    {
+      "epoch": 0.5065263157894737,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002527838457299438,
+      "loss": 2.718,
+      "step": 12030
+    },
+    {
+      "epoch": 0.5065684210526316,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002527501035148492,
+      "loss": 3.397,
+      "step": 12031
+    },
+    {
+      "epoch": 0.5066105263157895,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002527163612496508,
+      "loss": 2.4836,
+      "step": 12032
+    },
+    {
+      "epoch": 0.5066526315789474,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002526826189349636,
+      "loss": 3.2031,
+      "step": 12033
+    },
+    {
+      "epoch": 0.5066947368421053,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002526488765714023,
+      "loss": 3.0282,
+      "step": 12034
+    },
+    {
+      "epoch": 0.5067368421052632,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002526151341595815,
+      "loss": 3.2466,
+      "step": 12035
+    },
+    {
+      "epoch": 0.5067789473684211,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00025258139170011607,
+      "loss": 3.0334,
+      "step": 12036
+    },
+    {
+      "epoch": 0.506821052631579,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002525476491936207,
+      "loss": 2.9378,
+      "step": 12037
+    },
+    {
+      "epoch": 0.5068631578947368,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002525139066407101,
+      "loss": 3.3028,
+      "step": 12038
+    },
+    {
+      "epoch": 0.5069052631578947,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002524801640419991,
+      "loss": 3.4621,
+      "step": 12039
+    },
+    {
+      "epoch": 0.5069473684210526,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002524464213981025,
+      "loss": 3.241,
+      "step": 12040
+    },
+    {
+      "epoch": 0.5069894736842105,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002524126787096349,
+      "loss": 3.3807,
+      "step": 12041
+    },
+    {
+      "epoch": 0.5070315789473684,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002523789359772112,
+      "loss": 3.2486,
+      "step": 12042
+    },
+    {
+      "epoch": 0.5070736842105263,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.000252345193201446,
+      "loss": 3.6243,
+      "step": 12043
+    },
+    {
+      "epoch": 0.5071157894736842,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00025231145038295415,
+      "loss": 3.0721,
+      "step": 12044
+    },
+    {
+      "epoch": 0.5071578947368421,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0002522777075223504,
+      "loss": 3.4299,
+      "step": 12045
+    },
+    {
+      "epoch": 0.5072,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00025224396462024943,
+      "loss": 3.1958,
+      "step": 12046
+    },
+    {
+      "epoch": 0.5072421052631579,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002522102216772662,
+      "loss": 3.0929,
+      "step": 12047
+    },
+    {
+      "epoch": 0.5072842105263158,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00025217647869401526,
+      "loss": 3.1264,
+      "step": 12048
+    },
+    {
+      "epoch": 0.5073263157894737,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00025214273567111144,
+      "loss": 3.2995,
+      "step": 12049
+    },
+    {
+      "epoch": 0.5073684210526316,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002521089926091695,
+      "loss": 2.853,
+      "step": 12050
+    },
+    {
+      "epoch": 0.5074105263157894,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00025207524950880423,
+      "loss": 3.5925,
+      "step": 12051
+    },
+    {
+      "epoch": 0.5074526315789474,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00025204150637063026,
+      "loss": 3.211,
+      "step": 12052
+    },
+    {
+      "epoch": 0.5074947368421052,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00025200776319526247,
+      "loss": 3.6927,
+      "step": 12053
+    },
+    {
+      "epoch": 0.5075368421052632,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00025197401998331564,
+      "loss": 2.9643,
+      "step": 12054
+    },
+    {
+      "epoch": 0.507578947368421,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0002519402767354044,
+      "loss": 3.0827,
+      "step": 12055
+    },
+    {
+      "epoch": 0.507621052631579,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0002519065334521437,
+      "loss": 3.1019,
+      "step": 12056
+    },
+    {
+      "epoch": 0.5076631578947368,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00025187279013414815,
+      "loss": 3.094,
+      "step": 12057
+    },
+    {
+      "epoch": 0.5077052631578948,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00025183904678203263,
+      "loss": 3.1969,
+      "step": 12058
+    },
+    {
+      "epoch": 0.5077473684210526,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00025180530339641175,
+      "loss": 3.5149,
+      "step": 12059
+    },
+    {
+      "epoch": 0.5077894736842106,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00025177155997790034,
+      "loss": 2.7354,
+      "step": 12060
+    },
+    {
+      "epoch": 0.5078315789473684,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002517378165271133,
+      "loss": 3.0681,
+      "step": 12061
+    },
+    {
+      "epoch": 0.5078736842105264,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00025170407304466526,
+      "loss": 3.3541,
+      "step": 12062
+    },
+    {
+      "epoch": 0.5079157894736842,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002516703295311709,
+      "loss": 2.9952,
+      "step": 12063
+    },
+    {
+      "epoch": 0.507957894736842,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002516365859872452,
+      "loss": 2.8894,
+      "step": 12064
+    },
+    {
+      "epoch": 0.508,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002516028424135028,
+      "loss": 3.0108,
+      "step": 12065
+    },
+    {
+      "epoch": 0.5080421052631579,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00025156909881055847,
+      "loss": 3.2818,
+      "step": 12066
+    },
+    {
+      "epoch": 0.5080842105263158,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00025153535517902703,
+      "loss": 2.9561,
+      "step": 12067
+    },
+    {
+      "epoch": 0.5081263157894736,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00025150161151952323,
+      "loss": 2.9998,
+      "step": 12068
+    },
+    {
+      "epoch": 0.5081684210526316,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00025146786783266186,
+      "loss": 3.1814,
+      "step": 12069
+    },
+    {
+      "epoch": 0.5082105263157894,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002514341241190576,
+      "loss": 3.215,
+      "step": 12070
+    },
+    {
+      "epoch": 0.5082526315789474,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002514003803793253,
+      "loss": 3.3154,
+      "step": 12071
+    },
+    {
+      "epoch": 0.5082947368421052,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00025136663661407975,
+      "loss": 3.0814,
+      "step": 12072
+    },
+    {
+      "epoch": 0.5083368421052632,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002513328928239356,
+      "loss": 3.1449,
+      "step": 12073
+    },
+    {
+      "epoch": 0.508378947368421,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00025129914900950777,
+      "loss": 3.1711,
+      "step": 12074
+    },
+    {
+      "epoch": 0.508421052631579,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.000251265405171411,
+      "loss": 3.1284,
+      "step": 12075
+    },
+    {
+      "epoch": 0.5084631578947368,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00025123166131026006,
+      "loss": 2.1731,
+      "step": 12076
+    },
+    {
+      "epoch": 0.5085052631578947,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00025119791742666967,
+      "loss": 3.0633,
+      "step": 12077
+    },
+    {
+      "epoch": 0.5085473684210526,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002511641735212546,
+      "loss": 3.2886,
+      "step": 12078
+    },
+    {
+      "epoch": 0.5085894736842105,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00025113042959462963,
+      "loss": 3.3694,
+      "step": 12079
+    },
+    {
+      "epoch": 0.5086315789473684,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002510966856474096,
+      "loss": 3.1678,
+      "step": 12080
+    },
+    {
+      "epoch": 0.5086736842105263,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002510629416802092,
+      "loss": 3.2288,
+      "step": 12081
+    },
+    {
+      "epoch": 0.5087157894736842,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002510291976936433,
+      "loss": 3.2909,
+      "step": 12082
+    },
+    {
+      "epoch": 0.5087578947368421,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00025099545368832666,
+      "loss": 3.3384,
+      "step": 12083
+    },
+    {
+      "epoch": 0.5088,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00025096170966487397,
+      "loss": 3.3906,
+      "step": 12084
+    },
+    {
+      "epoch": 0.5088421052631579,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002509279656239001,
+      "loss": 3.2106,
+      "step": 12085
+    },
+    {
+      "epoch": 0.5088842105263158,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002508942215660197,
+      "loss": 3.0348,
+      "step": 12086
+    },
+    {
+      "epoch": 0.5089263157894737,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002508604774918477,
+      "loss": 3.1138,
+      "step": 12087
+    },
+    {
+      "epoch": 0.5089684210526316,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002508267334019988,
+      "loss": 3.0862,
+      "step": 12088
+    },
+    {
+      "epoch": 0.5090105263157895,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002507929892970878,
+      "loss": 3.2202,
+      "step": 12089
+    },
+    {
+      "epoch": 0.5090526315789474,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002507592451777295,
+      "loss": 3.2161,
+      "step": 12090
+    },
+    {
+      "epoch": 0.5090947368421053,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0002507255010445386,
+      "loss": 3.1247,
+      "step": 12091
+    },
+    {
+      "epoch": 0.5091368421052631,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002506917568981299,
+      "loss": 2.7256,
+      "step": 12092
+    },
+    {
+      "epoch": 0.5091789473684211,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002506580127391182,
+      "loss": 3.1457,
+      "step": 12093
+    },
+    {
+      "epoch": 0.5092210526315789,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00025062426856811834,
+      "loss": 3.0477,
+      "step": 12094
+    },
+    {
+      "epoch": 0.5092631578947369,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00025059052438574503,
+      "loss": 3.3582,
+      "step": 12095
+    },
+    {
+      "epoch": 0.5093052631578947,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00025055678019261297,
+      "loss": 3.2627,
+      "step": 12096
+    },
+    {
+      "epoch": 0.5093473684210527,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002505230359893371,
+      "loss": 3.0709,
+      "step": 12097
+    },
+    {
+      "epoch": 0.5093894736842105,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002504892917765322,
+      "loss": 3.1097,
+      "step": 12098
+    },
+    {
+      "epoch": 0.5094315789473685,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002504555475548129,
+      "loss": 3.1494,
+      "step": 12099
+    },
+    {
+      "epoch": 0.5094736842105263,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.000250421803324794,
+      "loss": 3.3128,
+      "step": 12100
+    },
+    {
+      "epoch": 0.5095157894736843,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00025038805908709046,
+      "loss": 3.0236,
+      "step": 12101
+    },
+    {
+      "epoch": 0.5095578947368421,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00025035431484231687,
+      "loss": 2.6067,
+      "step": 12102
+    },
+    {
+      "epoch": 0.5096,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002503205705910881,
+      "loss": 3.1457,
+      "step": 12103
+    },
+    {
+      "epoch": 0.5096421052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002502868263340189,
+      "loss": 3.1816,
+      "step": 12104
+    },
+    {
+      "epoch": 0.5096842105263157,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.000250253082071724,
+      "loss": 3.1244,
+      "step": 12105
+    },
+    {
+      "epoch": 0.5097263157894737,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00025021933780481834,
+      "loss": 3.3044,
+      "step": 12106
+    },
+    {
+      "epoch": 0.5097684210526315,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00025018559353391656,
+      "loss": 3.0747,
+      "step": 12107
+    },
+    {
+      "epoch": 0.5098105263157895,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00025015184925963354,
+      "loss": 3.0653,
+      "step": 12108
+    },
+    {
+      "epoch": 0.5098526315789473,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00025011810498258395,
+      "loss": 3.2287,
+      "step": 12109
+    },
+    {
+      "epoch": 0.5098947368421053,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00025008436070338266,
+      "loss": 2.655,
+      "step": 12110
+    },
+    {
+      "epoch": 0.5099368421052631,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002500506164226444,
+      "loss": 3.2279,
+      "step": 12111
+    },
+    {
+      "epoch": 0.5099789473684211,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00025001687214098393,
+      "loss": 3.1841,
+      "step": 12112
+    },
+    {
+      "epoch": 0.5100210526315789,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00024998312785901613,
+      "loss": 3.4688,
+      "step": 12113
+    },
+    {
+      "epoch": 0.5100631578947369,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002499493835773557,
+      "loss": 3.502,
+      "step": 12114
+    },
+    {
+      "epoch": 0.5101052631578947,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00024991563929661735,
+      "loss": 3.0315,
+      "step": 12115
+    },
+    {
+      "epoch": 0.5101473684210527,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00024988189501741606,
+      "loss": 3.0181,
+      "step": 12116
+    },
+    {
+      "epoch": 0.5101894736842105,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00024984815074036653,
+      "loss": 3.5937,
+      "step": 12117
+    },
+    {
+      "epoch": 0.5102315789473684,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0002498144064660834,
+      "loss": 2.9864,
+      "step": 12118
+    },
+    {
+      "epoch": 0.5102736842105263,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00024978066219518173,
+      "loss": 3.341,
+      "step": 12119
+    },
+    {
+      "epoch": 0.5103157894736842,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.000249746917928276,
+      "loss": 3.2266,
+      "step": 12120
+    },
+    {
+      "epoch": 0.5103578947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00024971317366598116,
+      "loss": 3.032,
+      "step": 12121
+    },
+    {
+      "epoch": 0.5104,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002496794294089119,
+      "loss": 3.4523,
+      "step": 12122
+    },
+    {
+      "epoch": 0.5104421052631579,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002496456851576831,
+      "loss": 3.3017,
+      "step": 12123
+    },
+    {
+      "epoch": 0.5104842105263158,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002496119409129096,
+      "loss": 2.8967,
+      "step": 12124
+    },
+    {
+      "epoch": 0.5105263157894737,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00024957819667520593,
+      "loss": 3.4734,
+      "step": 12125
+    },
+    {
+      "epoch": 0.5105684210526316,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002495444524451872,
+      "loss": 2.9296,
+      "step": 12126
+    },
+    {
+      "epoch": 0.5106105263157895,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00024951070822346787,
+      "loss": 2.9928,
+      "step": 12127
+    },
+    {
+      "epoch": 0.5106526315789474,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002494769640106629,
+      "loss": 3.6764,
+      "step": 12128
+    },
+    {
+      "epoch": 0.5106947368421053,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00024944321980738704,
+      "loss": 3.5326,
+      "step": 12129
+    },
+    {
+      "epoch": 0.5107368421052632,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00024940947561425504,
+      "loss": 2.9893,
+      "step": 12130
+    },
+    {
+      "epoch": 0.510778947368421,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002493757314318817,
+      "loss": 3.2129,
+      "step": 12131
+    },
+    {
+      "epoch": 0.510821052631579,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002493419872608818,
+      "loss": 3.1903,
+      "step": 12132
+    },
+    {
+      "epoch": 0.5108631578947368,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002493082431018702,
+      "loss": 2.9323,
+      "step": 12133
+    },
+    {
+      "epoch": 0.5109052631578948,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002492744989554615,
+      "loss": 2.8906,
+      "step": 12134
+    },
+    {
+      "epoch": 0.5109473684210526,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002492407548222706,
+      "loss": 3.4325,
+      "step": 12135
+    },
+    {
+      "epoch": 0.5109894736842105,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002492070107029123,
+      "loss": 3.6045,
+      "step": 12136
+    },
+    {
+      "epoch": 0.5110315789473684,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002491732665980012,
+      "loss": 3.3395,
+      "step": 12137
+    },
+    {
+      "epoch": 0.5110736842105263,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002491395225081523,
+      "loss": 3.3997,
+      "step": 12138
+    },
+    {
+      "epoch": 0.5111157894736842,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00024910577843398026,
+      "loss": 2.8872,
+      "step": 12139
+    },
+    {
+      "epoch": 0.5111578947368421,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00024907203437610003,
+      "loss": 3.0892,
+      "step": 12140
+    },
+    {
+      "epoch": 0.5112,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00024903829033512604,
+      "loss": 3.5654,
+      "step": 12141
+    },
+    {
+      "epoch": 0.5112421052631579,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00024900454631167346,
+      "loss": 3.155,
+      "step": 12142
+    },
+    {
+      "epoch": 0.5112842105263158,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00024897080230635676,
+      "loss": 3.1129,
+      "step": 12143
+    },
+    {
+      "epoch": 0.5113263157894736,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002489370583197908,
+      "loss": 3.2347,
+      "step": 12144
+    },
+    {
+      "epoch": 0.5113684210526316,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0002489033143525905,
+      "loss": 2.69,
+      "step": 12145
+    },
+    {
+      "epoch": 0.5114105263157894,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00024886957040537043,
+      "loss": 3.3338,
+      "step": 12146
+    },
+    {
+      "epoch": 0.5114526315789474,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002488358264787455,
+      "loss": 2.9988,
+      "step": 12147
+    },
+    {
+      "epoch": 0.5114947368421052,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002488020825733304,
+      "loss": 2.9507,
+      "step": 12148
+    },
+    {
+      "epoch": 0.5115368421052632,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00024876833868974006,
+      "loss": 2.901,
+      "step": 12149
+    },
+    {
+      "epoch": 0.511578947368421,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.000248734594828589,
+      "loss": 3.1979,
+      "step": 12150
+    },
+    {
+      "epoch": 0.511621052631579,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0002487008509904922,
+      "loss": 3.811,
+      "step": 12151
+    },
+    {
+      "epoch": 0.5116631578947368,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002486671071760644,
+      "loss": 3.4091,
+      "step": 12152
+    },
+    {
+      "epoch": 0.5117052631578948,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002486333633859203,
+      "loss": 3.0215,
+      "step": 12153
+    },
+    {
+      "epoch": 0.5117473684210526,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002485996196206747,
+      "loss": 3.2772,
+      "step": 12154
+    },
+    {
+      "epoch": 0.5117894736842106,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002485658758809424,
+      "loss": 3.1454,
+      "step": 12155
+    },
+    {
+      "epoch": 0.5118315789473684,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00024853213216733826,
+      "loss": 2.9137,
+      "step": 12156
+    },
+    {
+      "epoch": 0.5118736842105264,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002484983884804768,
+      "loss": 3.0146,
+      "step": 12157
+    },
+    {
+      "epoch": 0.5119157894736842,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00024846464482097293,
+      "loss": 3.1763,
+      "step": 12158
+    },
+    {
+      "epoch": 0.5119578947368421,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002484309011894416,
+      "loss": 3.2043,
+      "step": 12159
+    },
+    {
+      "epoch": 0.512,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002483971575864972,
+      "loss": 3.3013,
+      "step": 12160
+    },
+    {
+      "epoch": 0.5120421052631579,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00024836341401275494,
+      "loss": 3.1218,
+      "step": 12161
+    },
+    {
+      "epoch": 0.5120842105263158,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00024832967046882915,
+      "loss": 3.0082,
+      "step": 12162
+    },
+    {
+      "epoch": 0.5121263157894737,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002482959269553348,
+      "loss": 3.3194,
+      "step": 12163
+    },
+    {
+      "epoch": 0.5121684210526316,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00024826218347288675,
+      "loss": 3.3184,
+      "step": 12164
+    },
+    {
+      "epoch": 0.5122105263157894,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0002482284400220996,
+      "loss": 2.9509,
+      "step": 12165
+    },
+    {
+      "epoch": 0.5122526315789474,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002481946966035883,
+      "loss": 3.5582,
+      "step": 12166
+    },
+    {
+      "epoch": 0.5122947368421052,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00024816095321796743,
+      "loss": 3.4749,
+      "step": 12167
+    },
+    {
+      "epoch": 0.5123368421052632,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002481272098658519,
+      "loss": 3.3094,
+      "step": 12168
+    },
+    {
+      "epoch": 0.512378947368421,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00024809346654785635,
+      "loss": 2.8079,
+      "step": 12169
+    },
+    {
+      "epoch": 0.512421052631579,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00024805972326459553,
+      "loss": 3.0434,
+      "step": 12170
+    },
+    {
+      "epoch": 0.5124631578947368,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00024802598001668437,
+      "loss": 2.9434,
+      "step": 12171
+    },
+    {
+      "epoch": 0.5125052631578947,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00024799223680473754,
+      "loss": 3.3398,
+      "step": 12172
+    },
+    {
+      "epoch": 0.5125473684210526,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002479584936293698,
+      "loss": 3.215,
+      "step": 12173
+    },
+    {
+      "epoch": 0.5125894736842105,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00024792475049119583,
+      "loss": 2.7238,
+      "step": 12174
+    },
+    {
+      "epoch": 0.5126315789473684,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00024789100739083054,
+      "loss": 3.1644,
+      "step": 12175
+    },
+    {
+      "epoch": 0.5126736842105263,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00024785726432888857,
+      "loss": 2.9208,
+      "step": 12176
+    },
+    {
+      "epoch": 0.5127157894736842,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00024782352130598475,
+      "loss": 3.4491,
+      "step": 12177
+    },
+    {
+      "epoch": 0.5127578947368421,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00024778977832273384,
+      "loss": 3.524,
+      "step": 12178
+    },
+    {
+      "epoch": 0.5128,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002477560353797505,
+      "loss": 3.1669,
+      "step": 12179
+    },
+    {
+      "epoch": 0.5128421052631579,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00024772229247764965,
+      "loss": 3.3603,
+      "step": 12180
+    },
+    {
+      "epoch": 0.5128842105263158,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00024768854961704586,
+      "loss": 3.387,
+      "step": 12181
+    },
+    {
+      "epoch": 0.5129263157894737,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002476548067985541,
+      "loss": 3.3329,
+      "step": 12182
+    },
+    {
+      "epoch": 0.5129684210526316,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00024762106402278887,
+      "loss": 3.0421,
+      "step": 12183
+    },
+    {
+      "epoch": 0.5130105263157895,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002475873212903651,
+      "loss": 3.0694,
+      "step": 12184
+    },
+    {
+      "epoch": 0.5130526315789473,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002475535786018976,
+      "loss": 3.4284,
+      "step": 12185
+    },
+    {
+      "epoch": 0.5130947368421053,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0002475198359580009,
+      "loss": 3.0238,
+      "step": 12186
+    },
+    {
+      "epoch": 0.5131368421052631,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00024748609335928997,
+      "loss": 2.8265,
+      "step": 12187
+    },
+    {
+      "epoch": 0.5131789473684211,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002474523508063794,
+      "loss": 2.8474,
+      "step": 12188
+    },
+    {
+      "epoch": 0.5132210526315789,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.000247418608299884,
+      "loss": 2.464,
+      "step": 12189
+    },
+    {
+      "epoch": 0.5132631578947369,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00024738486584041853,
+      "loss": 3.3961,
+      "step": 12190
+    },
+    {
+      "epoch": 0.5133052631578947,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002473511234285977,
+      "loss": 2.8654,
+      "step": 12191
+    },
+    {
+      "epoch": 0.5133473684210527,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002473173810650364,
+      "loss": 3.6859,
+      "step": 12192
+    },
+    {
+      "epoch": 0.5133894736842105,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00024728363875034916,
+      "loss": 3.2257,
+      "step": 12193
+    },
+    {
+      "epoch": 0.5134315789473685,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00024724989648515093,
+      "loss": 2.8332,
+      "step": 12194
+    },
+    {
+      "epoch": 0.5134736842105263,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002472161542700563,
+      "loss": 3.198,
+      "step": 12195
+    },
+    {
+      "epoch": 0.5135157894736843,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00024718241210568,
+      "loss": 3.609,
+      "step": 12196
+    },
+    {
+      "epoch": 0.5135578947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00024714866999263693,
+      "loss": 3.1794,
+      "step": 12197
+    },
+    {
+      "epoch": 0.5136,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00024711492793154166,
+      "loss": 3.2849,
+      "step": 12198
+    },
+    {
+      "epoch": 0.5136421052631579,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00024708118592300915,
+      "loss": 3.2719,
+      "step": 12199
+    },
+    {
+      "epoch": 0.5136842105263157,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0002470474439676539,
+      "loss": 2.8864,
+      "step": 12200
+    },
+    {
+      "epoch": 0.5137263157894737,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00024701370206609084,
+      "loss": 3.0759,
+      "step": 12201
+    },
+    {
+      "epoch": 0.5137684210526315,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002469799602189345,
+      "loss": 2.9301,
+      "step": 12202
+    },
+    {
+      "epoch": 0.5138105263157895,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00024694621842679987,
+      "loss": 2.9565,
+      "step": 12203
+    },
+    {
+      "epoch": 0.5138526315789473,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00024691247669030145,
+      "loss": 2.9639,
+      "step": 12204
+    },
+    {
+      "epoch": 0.5138947368421053,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00024687873501005403,
+      "loss": 3.2476,
+      "step": 12205
+    },
+    {
+      "epoch": 0.5139368421052631,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002468449933866725,
+      "loss": 3.1205,
+      "step": 12206
+    },
+    {
+      "epoch": 0.5139789473684211,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00024681125182077137,
+      "loss": 3.1007,
+      "step": 12207
+    },
+    {
+      "epoch": 0.5140210526315789,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00024677751031296566,
+      "loss": 2.6997,
+      "step": 12208
+    },
+    {
+      "epoch": 0.5140631578947369,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00024674376886386977,
+      "loss": 3.1469,
+      "step": 12209
+    },
+    {
+      "epoch": 0.5141052631578947,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002467100274740987,
+      "loss": 2.9534,
+      "step": 12210
+    },
+    {
+      "epoch": 0.5141473684210527,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.000246676286144267,
+      "loss": 3.0507,
+      "step": 12211
+    },
+    {
+      "epoch": 0.5141894736842105,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002466425448749894,
+      "loss": 3.4129,
+      "step": 12212
+    },
+    {
+      "epoch": 0.5142315789473684,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00024660880366688076,
+      "loss": 3.2609,
+      "step": 12213
+    },
+    {
+      "epoch": 0.5142736842105263,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00024657506252055575,
+      "loss": 3.2974,
+      "step": 12214
+    },
+    {
+      "epoch": 0.5143157894736842,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002465413214366291,
+      "loss": 3.147,
+      "step": 12215
+    },
+    {
+      "epoch": 0.5143578947368421,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002465075804157154,
+      "loss": 2.4456,
+      "step": 12216
+    },
+    {
+      "epoch": 0.5144,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002464738394584295,
+      "loss": 3.297,
+      "step": 12217
+    },
+    {
+      "epoch": 0.5144421052631579,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002464400985653862,
+      "loss": 3.6292,
+      "step": 12218
+    },
+    {
+      "epoch": 0.5144842105263158,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00024640635773720006,
+      "loss": 3.2386,
+      "step": 12219
+    },
+    {
+      "epoch": 0.5145263157894737,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002463726169744859,
+      "loss": 3.0935,
+      "step": 12220
+    },
+    {
+      "epoch": 0.5145684210526316,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00024633887627785835,
+      "loss": 2.9387,
+      "step": 12221
+    },
+    {
+      "epoch": 0.5146105263157895,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00024630513564793217,
+      "loss": 2.6977,
+      "step": 12222
+    },
+    {
+      "epoch": 0.5146526315789474,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00024627139508532204,
+      "loss": 3.055,
+      "step": 12223
+    },
+    {
+      "epoch": 0.5146947368421053,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00024623765459064277,
+      "loss": 3.5394,
+      "step": 12224
+    },
+    {
+      "epoch": 0.5147368421052632,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002462039141645091,
+      "loss": 3.0072,
+      "step": 12225
+    },
+    {
+      "epoch": 0.514778947368421,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00024617017380753546,
+      "loss": 2.8815,
+      "step": 12226
+    },
+    {
+      "epoch": 0.514821052631579,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00024613643352033697,
+      "loss": 2.8721,
+      "step": 12227
+    },
+    {
+      "epoch": 0.5148631578947368,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0002461026933035279,
+      "loss": 2.9788,
+      "step": 12228
+    },
+    {
+      "epoch": 0.5149052631578948,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002460689531577233,
+      "loss": 3.3951,
+      "step": 12229
+    },
+    {
+      "epoch": 0.5149473684210526,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0002460352130835378,
+      "loss": 2.9128,
+      "step": 12230
+    },
+    {
+      "epoch": 0.5149894736842106,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.000246001473081586,
+      "loss": 3.2425,
+      "step": 12231
+    },
+    {
+      "epoch": 0.5150315789473684,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002459677331524826,
+      "loss": 3.0483,
+      "step": 12232
+    },
+    {
+      "epoch": 0.5150736842105264,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002459339932968424,
+      "loss": 3.0331,
+      "step": 12233
+    },
+    {
+      "epoch": 0.5151157894736842,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002459002535152802,
+      "loss": 2.8052,
+      "step": 12234
+    },
+    {
+      "epoch": 0.5151578947368421,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002458665138084104,
+      "loss": 3.1245,
+      "step": 12235
+    },
+    {
+      "epoch": 0.5152,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.000245832774176848,
+      "loss": 3.4676,
+      "step": 12236
+    },
+    {
+      "epoch": 0.515242105263158,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002457990346212074,
+      "loss": 3.2111,
+      "step": 12237
+    },
+    {
+      "epoch": 0.5152842105263158,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002457652951421035,
+      "loss": 3.2141,
+      "step": 12238
+    },
+    {
+      "epoch": 0.5153263157894736,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.000245731555740151,
+      "loss": 2.7014,
+      "step": 12239
+    },
+    {
+      "epoch": 0.5153684210526316,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00024569781641596447,
+      "loss": 3.2407,
+      "step": 12240
+    },
+    {
+      "epoch": 0.5154105263157894,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002456640771701588,
+      "loss": 3.2367,
+      "step": 12241
+    },
+    {
+      "epoch": 0.5154526315789474,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002456303380033484,
+      "loss": 3.0961,
+      "step": 12242
+    },
+    {
+      "epoch": 0.5154947368421052,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00024559659891614823,
+      "loss": 3.3362,
+      "step": 12243
+    },
+    {
+      "epoch": 0.5155368421052632,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002455628599091728,
+      "loss": 3.2345,
+      "step": 12244
+    },
+    {
+      "epoch": 0.515578947368421,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00024552912098303677,
+      "loss": 3.051,
+      "step": 12245
+    },
+    {
+      "epoch": 0.515621052631579,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00024549538213835506,
+      "loss": 3.188,
+      "step": 12246
+    },
+    {
+      "epoch": 0.5156631578947368,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00024546164337574205,
+      "loss": 3.2395,
+      "step": 12247
+    },
+    {
+      "epoch": 0.5157052631578948,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002454279046958126,
+      "loss": 3.292,
+      "step": 12248
+    },
+    {
+      "epoch": 0.5157473684210526,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002453941660991814,
+      "loss": 3.4256,
+      "step": 12249
+    },
+    {
+      "epoch": 0.5157894736842106,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002453604275864631,
+      "loss": 3.2756,
+      "step": 12250
+    },
+    {
+      "epoch": 0.5158315789473684,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0002453266891582723,
+      "loss": 3.3899,
+      "step": 12251
+    },
+    {
+      "epoch": 0.5158736842105263,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002452929508152237,
+      "loss": 2.4669,
+      "step": 12252
+    },
+    {
+      "epoch": 0.5159157894736842,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00024525921255793214,
+      "loss": 3.1027,
+      "step": 12253
+    },
+    {
+      "epoch": 0.5159578947368421,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00024522547438701205,
+      "loss": 3.2144,
+      "step": 12254
+    },
+    {
+      "epoch": 0.516,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002451917363030782,
+      "loss": 3.2645,
+      "step": 12255
+    },
+    {
+      "epoch": 0.5160421052631579,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002451579983067453,
+      "loss": 3.2968,
+      "step": 12256
+    },
+    {
+      "epoch": 0.5160842105263158,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00024512426039862805,
+      "loss": 3.382,
+      "step": 12257
+    },
+    {
+      "epoch": 0.5161263157894737,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002450905225793409,
+      "loss": 3.0617,
+      "step": 12258
+    },
+    {
+      "epoch": 0.5161684210526316,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002450567848494987,
+      "loss": 3.0794,
+      "step": 12259
+    },
+    {
+      "epoch": 0.5162105263157895,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002450230472097162,
+      "loss": 3.5174,
+      "step": 12260
+    },
+    {
+      "epoch": 0.5162526315789474,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002449893096606078,
+      "loss": 3.0724,
+      "step": 12261
+    },
+    {
+      "epoch": 0.5162947368421053,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002449555722027884,
+      "loss": 3.0555,
+      "step": 12262
+    },
+    {
+      "epoch": 0.5163368421052632,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002449218348368724,
+      "loss": 3.5196,
+      "step": 12263
+    },
+    {
+      "epoch": 0.516378947368421,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002448880975634747,
+      "loss": 3.1938,
+      "step": 12264
+    },
+    {
+      "epoch": 0.516421052631579,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00024485436038320984,
+      "loss": 2.8785,
+      "step": 12265
+    },
+    {
+      "epoch": 0.5164631578947368,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00024482062329669245,
+      "loss": 3.3379,
+      "step": 12266
+    },
+    {
+      "epoch": 0.5165052631578947,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00024478688630453734,
+      "loss": 3.3396,
+      "step": 12267
+    },
+    {
+      "epoch": 0.5165473684210526,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002447531494073589,
+      "loss": 3.0865,
+      "step": 12268
+    },
+    {
+      "epoch": 0.5165894736842105,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00024471941260577204,
+      "loss": 2.5941,
+      "step": 12269
+    },
+    {
+      "epoch": 0.5166315789473684,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002446856759003911,
+      "loss": 2.9426,
+      "step": 12270
+    },
+    {
+      "epoch": 0.5166736842105263,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.000244651939291831,
+      "loss": 3.482,
+      "step": 12271
+    },
+    {
+      "epoch": 0.5167157894736842,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002446182027807063,
+      "loss": 3.0566,
+      "step": 12272
+    },
+    {
+      "epoch": 0.5167578947368421,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00024458446636763163,
+      "loss": 3.3214,
+      "step": 12273
+    },
+    {
+      "epoch": 0.5168,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002445507300532216,
+      "loss": 3.0885,
+      "step": 12274
+    },
+    {
+      "epoch": 0.5168421052631579,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00024451699383809083,
+      "loss": 2.9159,
+      "step": 12275
+    },
+    {
+      "epoch": 0.5168842105263158,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002444832577228541,
+      "loss": 3.3702,
+      "step": 12276
+    },
+    {
+      "epoch": 0.5169263157894737,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002444495217081258,
+      "loss": 3.1155,
+      "step": 12277
+    },
+    {
+      "epoch": 0.5169684210526316,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002444157857945207,
+      "loss": 2.7856,
+      "step": 12278
+    },
+    {
+      "epoch": 0.5170105263157895,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00024438204998265356,
+      "loss": 3.3925,
+      "step": 12279
+    },
+    {
+      "epoch": 0.5170526315789473,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00024434831427313873,
+      "loss": 3.1533,
+      "step": 12280
+    },
+    {
+      "epoch": 0.5170947368421053,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.000244314578666591,
+      "loss": 3.1405,
+      "step": 12281
+    },
+    {
+      "epoch": 0.5171368421052631,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00024428084316362495,
+      "loss": 3.144,
+      "step": 12282
+    },
+    {
+      "epoch": 0.5171789473684211,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002442471077648553,
+      "loss": 3.4128,
+      "step": 12283
+    },
+    {
+      "epoch": 0.5172210526315789,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00024421337247089653,
+      "loss": 2.9597,
+      "step": 12284
+    },
+    {
+      "epoch": 0.5172631578947369,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002441796372823633,
+      "loss": 3.3521,
+      "step": 12285
+    },
+    {
+      "epoch": 0.5173052631578947,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00024414590219987034,
+      "loss": 3.2852,
+      "step": 12286
+    },
+    {
+      "epoch": 0.5173473684210527,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00024411216722403206,
+      "loss": 2.9238,
+      "step": 12287
+    },
+    {
+      "epoch": 0.5173894736842105,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002440784323554632,
+      "loss": 3.099,
+      "step": 12288
+    },
+    {
+      "epoch": 0.5174315789473685,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00024404469759477837,
+      "loss": 3.1167,
+      "step": 12289
+    },
+    {
+      "epoch": 0.5174736842105263,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00024401096294259216,
+      "loss": 3.13,
+      "step": 12290
+    },
+    {
+      "epoch": 0.5175157894736843,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00024397722839951912,
+      "loss": 3.5161,
+      "step": 12291
+    },
+    {
+      "epoch": 0.5175578947368421,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00024394349396617386,
+      "loss": 3.1147,
+      "step": 12292
+    },
+    {
+      "epoch": 0.5176,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00024390975964317117,
+      "loss": 3.2513,
+      "step": 12293
+    },
+    {
+      "epoch": 0.5176421052631579,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00024387602543112538,
+      "loss": 2.9772,
+      "step": 12294
+    },
+    {
+      "epoch": 0.5176842105263157,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002438422913306513,
+      "loss": 3.2483,
+      "step": 12295
+    },
+    {
+      "epoch": 0.5177263157894737,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00024380855734236338,
+      "loss": 3.0495,
+      "step": 12296
+    },
+    {
+      "epoch": 0.5177684210526315,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002437748234668763,
+      "loss": 2.714,
+      "step": 12297
+    },
+    {
+      "epoch": 0.5178105263157895,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00024374108970480455,
+      "loss": 3.358,
+      "step": 12298
+    },
+    {
+      "epoch": 0.5178526315789473,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002437073560567628,
+      "loss": 3.141,
+      "step": 12299
+    },
+    {
+      "epoch": 0.5178947368421053,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00024367362252336575,
+      "loss": 2.8486,
+      "step": 12300
+    },
+    {
+      "epoch": 0.5179368421052631,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00024363988910522776,
+      "loss": 2.6724,
+      "step": 12301
+    },
+    {
+      "epoch": 0.5179789473684211,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00024360615580296357,
+      "loss": 3.2516,
+      "step": 12302
+    },
+    {
+      "epoch": 0.5180210526315789,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00024357242261718767,
+      "loss": 3.4128,
+      "step": 12303
+    },
+    {
+      "epoch": 0.5180631578947369,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002435386895485147,
+      "loss": 2.6205,
+      "step": 12304
+    },
+    {
+      "epoch": 0.5181052631578947,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0002435049565975592,
+      "loss": 3.1658,
+      "step": 12305
+    },
+    {
+      "epoch": 0.5181473684210526,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00024347122376493574,
+      "loss": 2.8273,
+      "step": 12306
+    },
+    {
+      "epoch": 0.5181894736842105,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00024343749105125895,
+      "loss": 2.8882,
+      "step": 12307
+    },
+    {
+      "epoch": 0.5182315789473684,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00024340375845714334,
+      "loss": 3.1004,
+      "step": 12308
+    },
+    {
+      "epoch": 0.5182736842105263,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00024337002598320357,
+      "loss": 2.9158,
+      "step": 12309
+    },
+    {
+      "epoch": 0.5183157894736842,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00024333629363005403,
+      "loss": 3.1073,
+      "step": 12310
+    },
+    {
+      "epoch": 0.5183578947368421,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002433025613983095,
+      "loss": 3.5573,
+      "step": 12311
+    },
+    {
+      "epoch": 0.5184,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00024326882928858435,
+      "loss": 3.2337,
+      "step": 12312
+    },
+    {
+      "epoch": 0.5184421052631579,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00024323509730149318,
+      "loss": 3.5741,
+      "step": 12313
+    },
+    {
+      "epoch": 0.5184842105263158,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00024320136543765067,
+      "loss": 3.2324,
+      "step": 12314
+    },
+    {
+      "epoch": 0.5185263157894737,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00024316763369767127,
+      "loss": 3.2759,
+      "step": 12315
+    },
+    {
+      "epoch": 0.5185684210526316,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002431339020821696,
+      "loss": 3.2133,
+      "step": 12316
+    },
+    {
+      "epoch": 0.5186105263157895,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002431001705917601,
+      "loss": 3.2245,
+      "step": 12317
+    },
+    {
+      "epoch": 0.5186526315789474,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00024306643922705737,
+      "loss": 2.9696,
+      "step": 12318
+    },
+    {
+      "epoch": 0.5186947368421052,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00024303270798867608,
+      "loss": 3.1864,
+      "step": 12319
+    },
+    {
+      "epoch": 0.5187368421052632,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00024299897687723055,
+      "loss": 3.1686,
+      "step": 12320
+    },
+    {
+      "epoch": 0.518778947368421,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00024296524589333555,
+      "loss": 3.2667,
+      "step": 12321
+    },
+    {
+      "epoch": 0.518821052631579,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00024293151503760541,
+      "loss": 3.3006,
+      "step": 12322
+    },
+    {
+      "epoch": 0.5188631578947368,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00024289778431065478,
+      "loss": 2.759,
+      "step": 12323
+    },
+    {
+      "epoch": 0.5189052631578948,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00024286405371309818,
+      "loss": 3.1951,
+      "step": 12324
+    },
+    {
+      "epoch": 0.5189473684210526,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00024283032324555007,
+      "loss": 3.4533,
+      "step": 12325
+    },
+    {
+      "epoch": 0.5189894736842106,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00024279659290862518,
+      "loss": 3.0924,
+      "step": 12326
+    },
+    {
+      "epoch": 0.5190315789473684,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00024276286270293783,
+      "loss": 2.7843,
+      "step": 12327
+    },
+    {
+      "epoch": 0.5190736842105264,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00024272913262910267,
+      "loss": 3.1106,
+      "step": 12328
+    },
+    {
+      "epoch": 0.5191157894736842,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002426954026877341,
+      "loss": 2.5767,
+      "step": 12329
+    },
+    {
+      "epoch": 0.5191578947368422,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00024266167287944674,
+      "loss": 2.9513,
+      "step": 12330
+    },
+    {
+      "epoch": 0.5192,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00024262794320485505,
+      "loss": 3.1142,
+      "step": 12331
+    },
+    {
+      "epoch": 0.519242105263158,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00024259421366457364,
+      "loss": 3.1643,
+      "step": 12332
+    },
+    {
+      "epoch": 0.5192842105263158,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002425604842592169,
+      "loss": 3.1648,
+      "step": 12333
+    },
+    {
+      "epoch": 0.5193263157894736,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00024252675498939938,
+      "loss": 3.6169,
+      "step": 12334
+    },
+    {
+      "epoch": 0.5193684210526316,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00024249302585573572,
+      "loss": 3.1232,
+      "step": 12335
+    },
+    {
+      "epoch": 0.5194105263157894,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00024245929685884015,
+      "loss": 3.4468,
+      "step": 12336
+    },
+    {
+      "epoch": 0.5194526315789474,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002424255679993275,
+      "loss": 3.1718,
+      "step": 12337
+    },
+    {
+      "epoch": 0.5194947368421052,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00024239183927781194,
+      "loss": 2.8976,
+      "step": 12338
+    },
+    {
+      "epoch": 0.5195368421052632,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00024235811069490818,
+      "loss": 2.8872,
+      "step": 12339
+    },
+    {
+      "epoch": 0.519578947368421,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00024232438225123067,
+      "loss": 3.3939,
+      "step": 12340
+    },
+    {
+      "epoch": 0.519621052631579,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002422906539473939,
+      "loss": 2.9512,
+      "step": 12341
+    },
+    {
+      "epoch": 0.5196631578947368,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002422569257840124,
+      "loss": 3.4896,
+      "step": 12342
+    },
+    {
+      "epoch": 0.5197052631578948,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00024222319776170053,
+      "loss": 3.3663,
+      "step": 12343
+    },
+    {
+      "epoch": 0.5197473684210526,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00024218946988107298,
+      "loss": 3.1689,
+      "step": 12344
+    },
+    {
+      "epoch": 0.5197894736842106,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00024215574214274402,
+      "loss": 2.9509,
+      "step": 12345
+    },
+    {
+      "epoch": 0.5198315789473684,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00024212201454732818,
+      "loss": 3.1362,
+      "step": 12346
+    },
+    {
+      "epoch": 0.5198736842105263,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00024208828709544012,
+      "loss": 2.6479,
+      "step": 12347
+    },
+    {
+      "epoch": 0.5199157894736842,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00024205455978769405,
+      "loss": 2.8719,
+      "step": 12348
+    },
+    {
+      "epoch": 0.5199578947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00024202083262470463,
+      "loss": 3.0595,
+      "step": 12349
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002419871056070862,
+      "loss": 3.6503,
+      "step": 12350
+    },
+    {
+      "epoch": 0.5200421052631579,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002419533787354534,
+      "loss": 2.7708,
+      "step": 12351
+    },
+    {
+      "epoch": 0.5200842105263158,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00024191965201042048,
+      "loss": 2.873,
+      "step": 12352
+    },
+    {
+      "epoch": 0.5201263157894737,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00024188592543260203,
+      "loss": 3.283,
+      "step": 12353
+    },
+    {
+      "epoch": 0.5201684210526316,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00024185219900261254,
+      "loss": 3.2003,
+      "step": 12354
+    },
+    {
+      "epoch": 0.5202105263157895,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00024181847272106632,
+      "loss": 3.0545,
+      "step": 12355
+    },
+    {
+      "epoch": 0.5202526315789474,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.000241784746588578,
+      "loss": 2.8742,
+      "step": 12356
+    },
+    {
+      "epoch": 0.5202947368421053,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00024175102060576187,
+      "loss": 3.1363,
+      "step": 12357
+    },
+    {
+      "epoch": 0.5203368421052632,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00024171729477323256,
+      "loss": 2.8911,
+      "step": 12358
+    },
+    {
+      "epoch": 0.520378947368421,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002416835690916043,
+      "loss": 3.1137,
+      "step": 12359
+    },
+    {
+      "epoch": 0.5204210526315789,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00024164984356149163,
+      "loss": 2.8964,
+      "step": 12360
+    },
+    {
+      "epoch": 0.5204631578947368,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00024161611818350912,
+      "loss": 2.9419,
+      "step": 12361
+    },
+    {
+      "epoch": 0.5205052631578947,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00024158239295827096,
+      "loss": 3.023,
+      "step": 12362
+    },
+    {
+      "epoch": 0.5205473684210526,
+      "grad_norm": 4.0,
+      "learning_rate": 0.00024154866788639185,
+      "loss": 3.3454,
+      "step": 12363
+    },
+    {
+      "epoch": 0.5205894736842105,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00024151494296848597,
+      "loss": 2.8844,
+      "step": 12364
+    },
+    {
+      "epoch": 0.5206315789473684,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00024148121820516782,
+      "loss": 3.4306,
+      "step": 12365
+    },
+    {
+      "epoch": 0.5206736842105263,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00024144749359705195,
+      "loss": 3.0017,
+      "step": 12366
+    },
+    {
+      "epoch": 0.5207157894736842,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00024141376914475266,
+      "loss": 3.5576,
+      "step": 12367
+    },
+    {
+      "epoch": 0.5207578947368421,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00024138004484888448,
+      "loss": 3.3512,
+      "step": 12368
+    },
+    {
+      "epoch": 0.5208,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00024134632071006165,
+      "loss": 2.8676,
+      "step": 12369
+    },
+    {
+      "epoch": 0.5208421052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00024131259672889883,
+      "loss": 2.8283,
+      "step": 12370
+    },
+    {
+      "epoch": 0.5208842105263158,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00024127887290601013,
+      "loss": 3.3267,
+      "step": 12371
+    },
+    {
+      "epoch": 0.5209263157894737,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00024124514924201015,
+      "loss": 3.4884,
+      "step": 12372
+    },
+    {
+      "epoch": 0.5209684210526315,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00024121142573751335,
+      "loss": 3.0178,
+      "step": 12373
+    },
+    {
+      "epoch": 0.5210105263157895,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.000241177702393134,
+      "loss": 3.0593,
+      "step": 12374
+    },
+    {
+      "epoch": 0.5210526315789473,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00024114397920948656,
+      "loss": 2.8292,
+      "step": 12375
+    },
+    {
+      "epoch": 0.5210947368421053,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00024111025618718534,
+      "loss": 3.2152,
+      "step": 12376
+    },
+    {
+      "epoch": 0.5211368421052631,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00024107653332684496,
+      "loss": 3.0594,
+      "step": 12377
+    },
+    {
+      "epoch": 0.5211789473684211,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00024104281062907953,
+      "loss": 3.4461,
+      "step": 12378
+    },
+    {
+      "epoch": 0.5212210526315789,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00024100908809450357,
+      "loss": 3.1054,
+      "step": 12379
+    },
+    {
+      "epoch": 0.5212631578947369,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00024097536572373156,
+      "loss": 3.2267,
+      "step": 12380
+    },
+    {
+      "epoch": 0.5213052631578947,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00024094164351737765,
+      "loss": 3.2725,
+      "step": 12381
+    },
+    {
+      "epoch": 0.5213473684210527,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00024090792147605646,
+      "loss": 2.9615,
+      "step": 12382
+    },
+    {
+      "epoch": 0.5213894736842105,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00024087419960038215,
+      "loss": 3.0452,
+      "step": 12383
+    },
+    {
+      "epoch": 0.5214315789473685,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00024084047789096937,
+      "loss": 2.9175,
+      "step": 12384
+    },
+    {
+      "epoch": 0.5214736842105263,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00024080675634843217,
+      "loss": 2.9063,
+      "step": 12385
+    },
+    {
+      "epoch": 0.5215157894736843,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00024077303497338506,
+      "loss": 3.0822,
+      "step": 12386
+    },
+    {
+      "epoch": 0.5215578947368421,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00024073931376644253,
+      "loss": 2.9695,
+      "step": 12387
+    },
+    {
+      "epoch": 0.5216,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00024070559272821874,
+      "loss": 3.2468,
+      "step": 12388
+    },
+    {
+      "epoch": 0.5216421052631579,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00024067187185932814,
+      "loss": 3.3241,
+      "step": 12389
+    },
+    {
+      "epoch": 0.5216842105263158,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00024063815116038512,
+      "loss": 3.1056,
+      "step": 12390
+    },
+    {
+      "epoch": 0.5217263157894737,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00024060443063200396,
+      "loss": 2.5764,
+      "step": 12391
+    },
+    {
+      "epoch": 0.5217684210526315,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.000240570710274799,
+      "loss": 3.4007,
+      "step": 12392
+    },
+    {
+      "epoch": 0.5218105263157895,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00024053699008938457,
+      "loss": 3.2498,
+      "step": 12393
+    },
+    {
+      "epoch": 0.5218526315789473,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002405032700763752,
+      "loss": 3.559,
+      "step": 12394
+    },
+    {
+      "epoch": 0.5218947368421053,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.000240469550236385,
+      "loss": 2.9602,
+      "step": 12395
+    },
+    {
+      "epoch": 0.5219368421052631,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002404358305700285,
+      "loss": 2.6052,
+      "step": 12396
+    },
+    {
+      "epoch": 0.5219789473684211,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002404021110779198,
+      "loss": 3.1265,
+      "step": 12397
+    },
+    {
+      "epoch": 0.5220210526315789,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00024036839176067343,
+      "loss": 3.3766,
+      "step": 12398
+    },
+    {
+      "epoch": 0.5220631578947369,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002403346726189036,
+      "loss": 2.6975,
+      "step": 12399
+    },
+    {
+      "epoch": 0.5221052631578947,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00024030095365322465,
+      "loss": 3.1492,
+      "step": 12400
+    },
+    {
+      "epoch": 0.5221473684210526,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00024026723486425105,
+      "loss": 3.1906,
+      "step": 12401
+    },
+    {
+      "epoch": 0.5221894736842105,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002402335162525969,
+      "loss": 3.0686,
+      "step": 12402
+    },
+    {
+      "epoch": 0.5222315789473684,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00024019979781887671,
+      "loss": 3.06,
+      "step": 12403
+    },
+    {
+      "epoch": 0.5222736842105263,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002401660795637046,
+      "loss": 3.5952,
+      "step": 12404
+    },
+    {
+      "epoch": 0.5223157894736842,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.000240132361487695,
+      "loss": 2.8188,
+      "step": 12405
+    },
+    {
+      "epoch": 0.5223578947368421,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00024009864359146216,
+      "loss": 2.7708,
+      "step": 12406
+    },
+    {
+      "epoch": 0.5224,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00024006492587562038,
+      "loss": 3.3041,
+      "step": 12407
+    },
+    {
+      "epoch": 0.5224421052631579,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.000240031208340784,
+      "loss": 3.1785,
+      "step": 12408
+    },
+    {
+      "epoch": 0.5224842105263158,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00023999749098756727,
+      "loss": 3.3944,
+      "step": 12409
+    },
+    {
+      "epoch": 0.5225263157894737,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002399637738165846,
+      "loss": 3.2215,
+      "step": 12410
+    },
+    {
+      "epoch": 0.5225684210526316,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00023993005682845003,
+      "loss": 3.2354,
+      "step": 12411
+    },
+    {
+      "epoch": 0.5226105263157895,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00023989634002377804,
+      "loss": 2.4986,
+      "step": 12412
+    },
+    {
+      "epoch": 0.5226526315789474,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00023986262340318297,
+      "loss": 2.9684,
+      "step": 12413
+    },
+    {
+      "epoch": 0.5226947368421052,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00023982890696727885,
+      "loss": 2.9437,
+      "step": 12414
+    },
+    {
+      "epoch": 0.5227368421052632,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002397951907166802,
+      "loss": 3.2296,
+      "step": 12415
+    },
+    {
+      "epoch": 0.522778947368421,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00023976147465200106,
+      "loss": 3.078,
+      "step": 12416
+    },
+    {
+      "epoch": 0.522821052631579,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.000239727758773856,
+      "loss": 3.1443,
+      "step": 12417
+    },
+    {
+      "epoch": 0.5228631578947368,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00023969404308285893,
+      "loss": 3.0736,
+      "step": 12418
+    },
+    {
+      "epoch": 0.5229052631578948,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00023966032757962433,
+      "loss": 3.089,
+      "step": 12419
+    },
+    {
+      "epoch": 0.5229473684210526,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00023962661226476653,
+      "loss": 2.7387,
+      "step": 12420
+    },
+    {
+      "epoch": 0.5229894736842106,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00023959289713889953,
+      "loss": 3.1779,
+      "step": 12421
+    },
+    {
+      "epoch": 0.5230315789473684,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00023955918220263783,
+      "loss": 3.1233,
+      "step": 12422
+    },
+    {
+      "epoch": 0.5230736842105264,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00023952546745659547,
+      "loss": 3.515,
+      "step": 12423
+    },
+    {
+      "epoch": 0.5231157894736842,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00023949175290138684,
+      "loss": 3.2101,
+      "step": 12424
+    },
+    {
+      "epoch": 0.5231578947368422,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00023945803853762606,
+      "loss": 2.875,
+      "step": 12425
+    },
+    {
+      "epoch": 0.5232,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00023942432436592742,
+      "loss": 3.1816,
+      "step": 12426
+    },
+    {
+      "epoch": 0.5232421052631578,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00023939061038690532,
+      "loss": 2.8526,
+      "step": 12427
+    },
+    {
+      "epoch": 0.5232842105263158,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00023935689660117368,
+      "loss": 2.7425,
+      "step": 12428
+    },
+    {
+      "epoch": 0.5233263157894736,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.000239323183009347,
+      "loss": 3.0416,
+      "step": 12429
+    },
+    {
+      "epoch": 0.5233684210526316,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00023928946961203926,
+      "loss": 3.2021,
+      "step": 12430
+    },
+    {
+      "epoch": 0.5234105263157894,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00023925575640986488,
+      "loss": 3.0565,
+      "step": 12431
+    },
+    {
+      "epoch": 0.5234526315789474,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00023922204340343792,
+      "loss": 3.5443,
+      "step": 12432
+    },
+    {
+      "epoch": 0.5234947368421052,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00023918833059337273,
+      "loss": 3.4299,
+      "step": 12433
+    },
+    {
+      "epoch": 0.5235368421052632,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00023915461798028345,
+      "loss": 3.169,
+      "step": 12434
+    },
+    {
+      "epoch": 0.523578947368421,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002391209055647842,
+      "loss": 2.9163,
+      "step": 12435
+    },
+    {
+      "epoch": 0.523621052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00023908719334748944,
+      "loss": 2.7838,
+      "step": 12436
+    },
+    {
+      "epoch": 0.5236631578947368,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00023905348132901305,
+      "loss": 2.5605,
+      "step": 12437
+    },
+    {
+      "epoch": 0.5237052631578948,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00023901976950996948,
+      "loss": 3.4295,
+      "step": 12438
+    },
+    {
+      "epoch": 0.5237473684210526,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00023898605789097273,
+      "loss": 2.8216,
+      "step": 12439
+    },
+    {
+      "epoch": 0.5237894736842105,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00023895234647263702,
+      "loss": 3.7811,
+      "step": 12440
+    },
+    {
+      "epoch": 0.5238315789473684,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00023891863525557665,
+      "loss": 2.6695,
+      "step": 12441
+    },
+    {
+      "epoch": 0.5238736842105263,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00023888492424040567,
+      "loss": 3.2999,
+      "step": 12442
+    },
+    {
+      "epoch": 0.5239157894736842,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00023885121342773842,
+      "loss": 3.2658,
+      "step": 12443
+    },
+    {
+      "epoch": 0.5239578947368421,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00023881750281818883,
+      "loss": 3.54,
+      "step": 12444
+    },
+    {
+      "epoch": 0.524,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00023878379241237135,
+      "loss": 3.2714,
+      "step": 12445
+    },
+    {
+      "epoch": 0.5240421052631579,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00023875008221089985,
+      "loss": 3.3491,
+      "step": 12446
+    },
+    {
+      "epoch": 0.5240842105263158,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00023871637221438866,
+      "loss": 2.7296,
+      "step": 12447
+    },
+    {
+      "epoch": 0.5241263157894737,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00023868266242345196,
+      "loss": 3.15,
+      "step": 12448
+    },
+    {
+      "epoch": 0.5241684210526316,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00023864895283870382,
+      "loss": 2.7114,
+      "step": 12449
+    },
+    {
+      "epoch": 0.5242105263157895,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00023861524346075846,
+      "loss": 3.1248,
+      "step": 12450
+    },
+    {
+      "epoch": 0.5242526315789474,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00023858153429022995,
+      "loss": 3.1649,
+      "step": 12451
+    },
+    {
+      "epoch": 0.5242947368421053,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00023854782532773254,
+      "loss": 3.5739,
+      "step": 12452
+    },
+    {
+      "epoch": 0.5243368421052632,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002385141165738802,
+      "loss": 2.9997,
+      "step": 12453
+    },
+    {
+      "epoch": 0.5243789473684211,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00023848040802928718,
+      "loss": 2.8612,
+      "step": 12454
+    },
+    {
+      "epoch": 0.5244210526315789,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00023844669969456765,
+      "loss": 2.985,
+      "step": 12455
+    },
+    {
+      "epoch": 0.5244631578947369,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002384129915703356,
+      "loss": 3.0751,
+      "step": 12456
+    },
+    {
+      "epoch": 0.5245052631578947,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002383792836572053,
+      "loss": 3.1573,
+      "step": 12457
+    },
+    {
+      "epoch": 0.5245473684210527,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002383455759557907,
+      "loss": 3.4282,
+      "step": 12458
+    },
+    {
+      "epoch": 0.5245894736842105,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00023831186846670616,
+      "loss": 2.9263,
+      "step": 12459
+    },
+    {
+      "epoch": 0.5246315789473684,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002382781611905655,
+      "loss": 3.4646,
+      "step": 12460
+    },
+    {
+      "epoch": 0.5246736842105263,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00023824445412798303,
+      "loss": 3.1308,
+      "step": 12461
+    },
+    {
+      "epoch": 0.5247157894736842,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00023821074727957284,
+      "loss": 3.3393,
+      "step": 12462
+    },
+    {
+      "epoch": 0.5247578947368421,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00023817704064594888,
+      "loss": 3.2434,
+      "step": 12463
+    },
+    {
+      "epoch": 0.5248,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00023814333422772545,
+      "loss": 3.036,
+      "step": 12464
+    },
+    {
+      "epoch": 0.5248421052631579,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002381096280255165,
+      "loss": 3.1238,
+      "step": 12465
+    },
+    {
+      "epoch": 0.5248842105263158,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002380759220399361,
+      "loss": 2.7108,
+      "step": 12466
+    },
+    {
+      "epoch": 0.5249263157894737,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00023804221627159847,
+      "loss": 3.4026,
+      "step": 12467
+    },
+    {
+      "epoch": 0.5249684210526315,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00023800851072111753,
+      "loss": 3.3602,
+      "step": 12468
+    },
+    {
+      "epoch": 0.5250105263157895,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00023797480538910757,
+      "loss": 2.8494,
+      "step": 12469
+    },
+    {
+      "epoch": 0.5250526315789473,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002379411002761824,
+      "loss": 3.1315,
+      "step": 12470
+    },
+    {
+      "epoch": 0.5250947368421053,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00023790739538295634,
+      "loss": 3.1601,
+      "step": 12471
+    },
+    {
+      "epoch": 0.5251368421052631,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002378736907100432,
+      "loss": 2.8565,
+      "step": 12472
+    },
+    {
+      "epoch": 0.5251789473684211,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00023783998625805715,
+      "loss": 3.1007,
+      "step": 12473
+    },
+    {
+      "epoch": 0.5252210526315789,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00023780628202761233,
+      "loss": 2.9499,
+      "step": 12474
+    },
+    {
+      "epoch": 0.5252631578947369,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002377725780193227,
+      "loss": 2.7723,
+      "step": 12475
+    },
+    {
+      "epoch": 0.5253052631578947,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002377388742338024,
+      "loss": 2.924,
+      "step": 12476
+    },
+    {
+      "epoch": 0.5253473684210527,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00023770517067166533,
+      "loss": 2.8851,
+      "step": 12477
+    },
+    {
+      "epoch": 0.5253894736842105,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002376714673335257,
+      "loss": 3.3263,
+      "step": 12478
+    },
+    {
+      "epoch": 0.5254315789473685,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00023763776421999736,
+      "loss": 3.231,
+      "step": 12479
+    },
+    {
+      "epoch": 0.5254736842105263,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002376040613316944,
+      "loss": 3.1015,
+      "step": 12480
+    },
+    {
+      "epoch": 0.5255157894736842,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00023757035866923102,
+      "loss": 3.1895,
+      "step": 12481
+    },
+    {
+      "epoch": 0.5255578947368421,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00023753665623322097,
+      "loss": 2.3863,
+      "step": 12482
+    },
+    {
+      "epoch": 0.5256,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00023750295402427847,
+      "loss": 2.8034,
+      "step": 12483
+    },
+    {
+      "epoch": 0.5256421052631579,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00023746925204301737,
+      "loss": 3.6204,
+      "step": 12484
+    },
+    {
+      "epoch": 0.5256842105263158,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00023743555029005192,
+      "loss": 3.2046,
+      "step": 12485
+    },
+    {
+      "epoch": 0.5257263157894737,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00023740184876599584,
+      "loss": 3.0912,
+      "step": 12486
+    },
+    {
+      "epoch": 0.5257684210526316,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00023736814747146333,
+      "loss": 2.7349,
+      "step": 12487
+    },
+    {
+      "epoch": 0.5258105263157895,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002373344464070684,
+      "loss": 3.0671,
+      "step": 12488
+    },
+    {
+      "epoch": 0.5258526315789473,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00023730074557342485,
+      "loss": 3.4956,
+      "step": 12489
+    },
+    {
+      "epoch": 0.5258947368421053,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00023726704497114686,
+      "loss": 3.4693,
+      "step": 12490
+    },
+    {
+      "epoch": 0.5259368421052631,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00023723334460084834,
+      "loss": 3.1823,
+      "step": 12491
+    },
+    {
+      "epoch": 0.5259789473684211,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00023719964446314334,
+      "loss": 2.8025,
+      "step": 12492
+    },
+    {
+      "epoch": 0.526021052631579,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002371659445586457,
+      "loss": 2.956,
+      "step": 12493
+    },
+    {
+      "epoch": 0.5260631578947368,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00023713224488796945,
+      "loss": 2.7892,
+      "step": 12494
+    },
+    {
+      "epoch": 0.5261052631578947,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002370985454517287,
+      "loss": 3.2175,
+      "step": 12495
+    },
+    {
+      "epoch": 0.5261473684210526,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00023706484625053714,
+      "loss": 3.4431,
+      "step": 12496
+    },
+    {
+      "epoch": 0.5261894736842105,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00023703114728500905,
+      "loss": 3.3788,
+      "step": 12497
+    },
+    {
+      "epoch": 0.5262315789473684,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00023699744855575807,
+      "loss": 3.214,
+      "step": 12498
+    },
+    {
+      "epoch": 0.5262736842105263,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002369637500633984,
+      "loss": 3.044,
+      "step": 12499
+    },
+    {
+      "epoch": 0.5263157894736842,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002369300518085438,
+      "loss": 3.1763,
+      "step": 12500
+    },
+    {
+      "epoch": 0.5263578947368421,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00023689635379180833,
+      "loss": 3.1968,
+      "step": 12501
+    },
+    {
+      "epoch": 0.5264,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00023686265601380598,
+      "loss": 3.2779,
+      "step": 12502
+    },
+    {
+      "epoch": 0.5264421052631579,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00023682895847515047,
+      "loss": 3.3576,
+      "step": 12503
+    },
+    {
+      "epoch": 0.5264842105263158,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.000236795261176456,
+      "loss": 3.3213,
+      "step": 12504
+    },
+    {
+      "epoch": 0.5265263157894737,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00023676156411833624,
+      "loss": 2.7958,
+      "step": 12505
+    },
+    {
+      "epoch": 0.5265684210526316,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002367278673014053,
+      "loss": 2.6401,
+      "step": 12506
+    },
+    {
+      "epoch": 0.5266105263157895,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.000236694170726277,
+      "loss": 3.4331,
+      "step": 12507
+    },
+    {
+      "epoch": 0.5266526315789474,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00023666047439356523,
+      "loss": 2.7475,
+      "step": 12508
+    },
+    {
+      "epoch": 0.5266947368421052,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00023662677830388402,
+      "loss": 3.0667,
+      "step": 12509
+    },
+    {
+      "epoch": 0.5267368421052632,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002365930824578471,
+      "loss": 3.4498,
+      "step": 12510
+    },
+    {
+      "epoch": 0.526778947368421,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002365593868560686,
+      "loss": 2.9331,
+      "step": 12511
+    },
+    {
+      "epoch": 0.526821052631579,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00023652569149916217,
+      "loss": 2.9081,
+      "step": 12512
+    },
+    {
+      "epoch": 0.5268631578947368,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002364919963877418,
+      "loss": 3.2116,
+      "step": 12513
+    },
+    {
+      "epoch": 0.5269052631578948,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00023645830152242147,
+      "loss": 3.2564,
+      "step": 12514
+    },
+    {
+      "epoch": 0.5269473684210526,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00023642460690381494,
+      "loss": 3.0385,
+      "step": 12515
+    },
+    {
+      "epoch": 0.5269894736842106,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0002363909125325361,
+      "loss": 2.4723,
+      "step": 12516
+    },
+    {
+      "epoch": 0.5270315789473684,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00023635721840919882,
+      "loss": 3.1462,
+      "step": 12517
+    },
+    {
+      "epoch": 0.5270736842105264,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002363235245344171,
+      "loss": 3.2589,
+      "step": 12518
+    },
+    {
+      "epoch": 0.5271157894736842,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002362898309088046,
+      "loss": 2.8462,
+      "step": 12519
+    },
+    {
+      "epoch": 0.5271578947368422,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00023625613753297523,
+      "loss": 2.7998,
+      "step": 12520
+    },
+    {
+      "epoch": 0.5272,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00023622244440754304,
+      "loss": 3.1943,
+      "step": 12521
+    },
+    {
+      "epoch": 0.5272421052631578,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002361887515331216,
+      "loss": 3.0822,
+      "step": 12522
+    },
+    {
+      "epoch": 0.5272842105263158,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00023615505891032496,
+      "loss": 3.4414,
+      "step": 12523
+    },
+    {
+      "epoch": 0.5273263157894736,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00023612136653976683,
+      "loss": 2.8385,
+      "step": 12524
+    },
+    {
+      "epoch": 0.5273684210526316,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002360876744220611,
+      "loss": 3.2219,
+      "step": 12525
+    },
+    {
+      "epoch": 0.5274105263157894,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00023605398255782155,
+      "loss": 3.537,
+      "step": 12526
+    },
+    {
+      "epoch": 0.5274526315789474,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00023602029094766205,
+      "loss": 3.1937,
+      "step": 12527
+    },
+    {
+      "epoch": 0.5274947368421052,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00023598659959219654,
+      "loss": 3.4697,
+      "step": 12528
+    },
+    {
+      "epoch": 0.5275368421052632,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00023595290849203862,
+      "loss": 3.3632,
+      "step": 12529
+    },
+    {
+      "epoch": 0.527578947368421,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002359192176478023,
+      "loss": 3.4115,
+      "step": 12530
+    },
+    {
+      "epoch": 0.527621052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00023588552706010118,
+      "loss": 3.1885,
+      "step": 12531
+    },
+    {
+      "epoch": 0.5276631578947368,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00023585183672954923,
+      "loss": 3.1806,
+      "step": 12532
+    },
+    {
+      "epoch": 0.5277052631578948,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00023581814665676017,
+      "loss": 3.2838,
+      "step": 12533
+    },
+    {
+      "epoch": 0.5277473684210526,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002357844568423478,
+      "loss": 3.2951,
+      "step": 12534
+    },
+    {
+      "epoch": 0.5277894736842105,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00023575076728692598,
+      "loss": 3.1688,
+      "step": 12535
+    },
+    {
+      "epoch": 0.5278315789473684,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00023571707799110837,
+      "loss": 3.0596,
+      "step": 12536
+    },
+    {
+      "epoch": 0.5278736842105263,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00023568338895550893,
+      "loss": 3.3758,
+      "step": 12537
+    },
+    {
+      "epoch": 0.5279157894736842,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002356497001807412,
+      "loss": 3.3384,
+      "step": 12538
+    },
+    {
+      "epoch": 0.5279578947368421,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002356160116674192,
+      "loss": 3.1275,
+      "step": 12539
+    },
+    {
+      "epoch": 0.528,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00023558232341615643,
+      "loss": 3.7641,
+      "step": 12540
+    },
+    {
+      "epoch": 0.5280421052631579,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002355486354275668,
+      "loss": 3.3112,
+      "step": 12541
+    },
+    {
+      "epoch": 0.5280842105263158,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002355149477022641,
+      "loss": 3.35,
+      "step": 12542
+    },
+    {
+      "epoch": 0.5281263157894737,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.000235481260240862,
+      "loss": 3.3637,
+      "step": 12543
+    },
+    {
+      "epoch": 0.5281684210526316,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00023544757304397437,
+      "loss": 3.4667,
+      "step": 12544
+    },
+    {
+      "epoch": 0.5282105263157895,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00023541388611221476,
+      "loss": 2.8021,
+      "step": 12545
+    },
+    {
+      "epoch": 0.5282526315789474,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002353801994461971,
+      "loss": 3.2508,
+      "step": 12546
+    },
+    {
+      "epoch": 0.5282947368421053,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00023534651304653496,
+      "loss": 2.4988,
+      "step": 12547
+    },
+    {
+      "epoch": 0.5283368421052631,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002353128269138421,
+      "loss": 3.3198,
+      "step": 12548
+    },
+    {
+      "epoch": 0.5283789473684211,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002352791410487323,
+      "loss": 3.0241,
+      "step": 12549
+    },
+    {
+      "epoch": 0.5284210526315789,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00023524545545181933,
+      "loss": 2.8186,
+      "step": 12550
+    },
+    {
+      "epoch": 0.5284631578947369,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002352117701237168,
+      "loss": 2.9601,
+      "step": 12551
+    },
+    {
+      "epoch": 0.5285052631578947,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00023517808506503835,
+      "loss": 3.2577,
+      "step": 12552
+    },
+    {
+      "epoch": 0.5285473684210527,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00023514440027639792,
+      "loss": 2.7615,
+      "step": 12553
+    },
+    {
+      "epoch": 0.5285894736842105,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00023511071575840894,
+      "loss": 3.4114,
+      "step": 12554
+    },
+    {
+      "epoch": 0.5286315789473685,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00023507703151168526,
+      "loss": 3.2768,
+      "step": 12555
+    },
+    {
+      "epoch": 0.5286736842105263,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00023504334753684058,
+      "loss": 2.928,
+      "step": 12556
+    },
+    {
+      "epoch": 0.5287157894736843,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00023500966383448846,
+      "loss": 3.1946,
+      "step": 12557
+    },
+    {
+      "epoch": 0.5287578947368421,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002349759804052427,
+      "loss": 3.11,
+      "step": 12558
+    },
+    {
+      "epoch": 0.5288,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0002349422972497169,
+      "loss": 3.0096,
+      "step": 12559
+    },
+    {
+      "epoch": 0.5288421052631579,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002349086143685247,
+      "loss": 3.206,
+      "step": 12560
+    },
+    {
+      "epoch": 0.5288842105263157,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00023487493176227995,
+      "loss": 3.1264,
+      "step": 12561
+    },
+    {
+      "epoch": 0.5289263157894737,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00023484124943159607,
+      "loss": 3.5353,
+      "step": 12562
+    },
+    {
+      "epoch": 0.5289684210526315,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00023480756737708693,
+      "loss": 3.3653,
+      "step": 12563
+    },
+    {
+      "epoch": 0.5290105263157895,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00023477388559936592,
+      "loss": 3.0333,
+      "step": 12564
+    },
+    {
+      "epoch": 0.5290526315789473,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002347402040990469,
+      "loss": 3.2067,
+      "step": 12565
+    },
+    {
+      "epoch": 0.5290947368421053,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00023470652287674342,
+      "loss": 3.1623,
+      "step": 12566
+    },
+    {
+      "epoch": 0.5291368421052631,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00023467284193306908,
+      "loss": 3.2997,
+      "step": 12567
+    },
+    {
+      "epoch": 0.5291789473684211,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0002346391612686376,
+      "loss": 2.574,
+      "step": 12568
+    },
+    {
+      "epoch": 0.5292210526315789,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00023460548088406253,
+      "loss": 3.3146,
+      "step": 12569
+    },
+    {
+      "epoch": 0.5292631578947369,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00023457180077995758,
+      "loss": 3.1554,
+      "step": 12570
+    },
+    {
+      "epoch": 0.5293052631578947,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00023453812095693618,
+      "loss": 2.8136,
+      "step": 12571
+    },
+    {
+      "epoch": 0.5293473684210527,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002345044414156122,
+      "loss": 2.8707,
+      "step": 12572
+    },
+    {
+      "epoch": 0.5293894736842105,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00023447076215659895,
+      "loss": 3.3357,
+      "step": 12573
+    },
+    {
+      "epoch": 0.5294315789473685,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00023443708318051016,
+      "loss": 3.4162,
+      "step": 12574
+    },
+    {
+      "epoch": 0.5294736842105263,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002344034044879595,
+      "loss": 2.9379,
+      "step": 12575
+    },
+    {
+      "epoch": 0.5295157894736842,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00023436972607956044,
+      "loss": 2.9,
+      "step": 12576
+    },
+    {
+      "epoch": 0.5295578947368421,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002343360479559267,
+      "loss": 2.9898,
+      "step": 12577
+    },
+    {
+      "epoch": 0.5296,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00023430237011767165,
+      "loss": 3.0563,
+      "step": 12578
+    },
+    {
+      "epoch": 0.5296421052631579,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002342686925654091,
+      "loss": 3.2241,
+      "step": 12579
+    },
+    {
+      "epoch": 0.5296842105263158,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00023423501529975236,
+      "loss": 3.4646,
+      "step": 12580
+    },
+    {
+      "epoch": 0.5297263157894737,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00023420133832131513,
+      "loss": 3.0999,
+      "step": 12581
+    },
+    {
+      "epoch": 0.5297684210526316,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00023416766163071104,
+      "loss": 3.2165,
+      "step": 12582
+    },
+    {
+      "epoch": 0.5298105263157895,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00023413398522855347,
+      "loss": 3.371,
+      "step": 12583
+    },
+    {
+      "epoch": 0.5298526315789474,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002341003091154561,
+      "loss": 2.5699,
+      "step": 12584
+    },
+    {
+      "epoch": 0.5298947368421053,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00023406663329203235,
+      "loss": 2.6565,
+      "step": 12585
+    },
+    {
+      "epoch": 0.5299368421052632,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00023403295775889594,
+      "loss": 3.0076,
+      "step": 12586
+    },
+    {
+      "epoch": 0.5299789473684211,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00023399928251666014,
+      "loss": 3.0702,
+      "step": 12587
+    },
+    {
+      "epoch": 0.530021052631579,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00023396560756593865,
+      "loss": 3.1357,
+      "step": 12588
+    },
+    {
+      "epoch": 0.5300631578947368,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00023393193290734504,
+      "loss": 3.0796,
+      "step": 12589
+    },
+    {
+      "epoch": 0.5301052631578947,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00023389825854149262,
+      "loss": 2.7889,
+      "step": 12590
+    },
+    {
+      "epoch": 0.5301473684210526,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002338645844689951,
+      "loss": 2.6947,
+      "step": 12591
+    },
+    {
+      "epoch": 0.5301894736842105,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002338309106904658,
+      "loss": 2.9049,
+      "step": 12592
+    },
+    {
+      "epoch": 0.5302315789473684,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00023379723720651845,
+      "loss": 2.9446,
+      "step": 12593
+    },
+    {
+      "epoch": 0.5302736842105263,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00023376356401776625,
+      "loss": 3.0439,
+      "step": 12594
+    },
+    {
+      "epoch": 0.5303157894736842,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00023372989112482287,
+      "loss": 3.2283,
+      "step": 12595
+    },
+    {
+      "epoch": 0.5303578947368421,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00023369621852830185,
+      "loss": 3.1895,
+      "step": 12596
+    },
+    {
+      "epoch": 0.5304,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00023366254622881648,
+      "loss": 2.9715,
+      "step": 12597
+    },
+    {
+      "epoch": 0.5304421052631579,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00023362887422698043,
+      "loss": 3.0714,
+      "step": 12598
+    },
+    {
+      "epoch": 0.5304842105263158,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00023359520252340691,
+      "loss": 3.1533,
+      "step": 12599
+    },
+    {
+      "epoch": 0.5305263157894737,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0002335615311187096,
+      "loss": 2.7955,
+      "step": 12600
+    },
+    {
+      "epoch": 0.5305684210526316,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00023352786001350185,
+      "loss": 2.9551,
+      "step": 12601
+    },
+    {
+      "epoch": 0.5306105263157894,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002334941892083971,
+      "loss": 2.6214,
+      "step": 12602
+    },
+    {
+      "epoch": 0.5306526315789474,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00023346051870400898,
+      "loss": 3.1395,
+      "step": 12603
+    },
+    {
+      "epoch": 0.5306947368421052,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00023342684850095064,
+      "loss": 3.1144,
+      "step": 12604
+    },
+    {
+      "epoch": 0.5307368421052632,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00023339317859983575,
+      "loss": 3.4439,
+      "step": 12605
+    },
+    {
+      "epoch": 0.530778947368421,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00023335950900127753,
+      "loss": 2.9315,
+      "step": 12606
+    },
+    {
+      "epoch": 0.530821052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00023332583970588956,
+      "loss": 3.504,
+      "step": 12607
+    },
+    {
+      "epoch": 0.5308631578947368,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00023329217071428514,
+      "loss": 2.6561,
+      "step": 12608
+    },
+    {
+      "epoch": 0.5309052631578948,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00023325850202707784,
+      "loss": 3.4882,
+      "step": 12609
+    },
+    {
+      "epoch": 0.5309473684210526,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00023322483364488088,
+      "loss": 3.1438,
+      "step": 12610
+    },
+    {
+      "epoch": 0.5309894736842106,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00023319116556830773,
+      "loss": 2.9365,
+      "step": 12611
+    },
+    {
+      "epoch": 0.5310315789473684,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00023315749779797193,
+      "loss": 3.1366,
+      "step": 12612
+    },
+    {
+      "epoch": 0.5310736842105264,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002331238303344866,
+      "loss": 3.1568,
+      "step": 12613
+    },
+    {
+      "epoch": 0.5311157894736842,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00023309016317846524,
+      "loss": 3.5282,
+      "step": 12614
+    },
+    {
+      "epoch": 0.531157894736842,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00023305649633052135,
+      "loss": 2.6739,
+      "step": 12615
+    },
+    {
+      "epoch": 0.5312,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002330228297912681,
+      "loss": 3.6034,
+      "step": 12616
+    },
+    {
+      "epoch": 0.5312421052631578,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00023298916356131896,
+      "loss": 3.169,
+      "step": 12617
+    },
+    {
+      "epoch": 0.5312842105263158,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00023295549764128725,
+      "loss": 3.2674,
+      "step": 12618
+    },
+    {
+      "epoch": 0.5313263157894736,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00023292183203178646,
+      "loss": 2.5441,
+      "step": 12619
+    },
+    {
+      "epoch": 0.5313684210526316,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00023288816673342969,
+      "loss": 2.7802,
+      "step": 12620
+    },
+    {
+      "epoch": 0.5314105263157894,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00023285450174683042,
+      "loss": 3.0206,
+      "step": 12621
+    },
+    {
+      "epoch": 0.5314526315789474,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002328208370726021,
+      "loss": 2.464,
+      "step": 12622
+    },
+    {
+      "epoch": 0.5314947368421052,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00023278717271135786,
+      "loss": 2.9625,
+      "step": 12623
+    },
+    {
+      "epoch": 0.5315368421052632,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00023275350866371113,
+      "loss": 3.0688,
+      "step": 12624
+    },
+    {
+      "epoch": 0.531578947368421,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002327198449302752,
+      "loss": 3.3764,
+      "step": 12625
+    },
+    {
+      "epoch": 0.531621052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00023268618151166343,
+      "loss": 3.4584,
+      "step": 12626
+    },
+    {
+      "epoch": 0.5316631578947368,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.000232652518408489,
+      "loss": 3.0768,
+      "step": 12627
+    },
+    {
+      "epoch": 0.5317052631578948,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00023261885562136535,
+      "loss": 3.1833,
+      "step": 12628
+    },
+    {
+      "epoch": 0.5317473684210526,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00023258519315090578,
+      "loss": 3.4424,
+      "step": 12629
+    },
+    {
+      "epoch": 0.5317894736842105,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00023255153099772344,
+      "loss": 3.1627,
+      "step": 12630
+    },
+    {
+      "epoch": 0.5318315789473684,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00023251786916243184,
+      "loss": 3.2133,
+      "step": 12631
+    },
+    {
+      "epoch": 0.5318736842105263,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00023248420764564398,
+      "loss": 3.3195,
+      "step": 12632
+    },
+    {
+      "epoch": 0.5319157894736842,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00023245054644797334,
+      "loss": 2.9517,
+      "step": 12633
+    },
+    {
+      "epoch": 0.5319578947368421,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00023241688557003307,
+      "loss": 3.0739,
+      "step": 12634
+    },
+    {
+      "epoch": 0.532,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00023238322501243649,
+      "loss": 3.0601,
+      "step": 12635
+    },
+    {
+      "epoch": 0.5320421052631579,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00023234956477579698,
+      "loss": 2.9509,
+      "step": 12636
+    },
+    {
+      "epoch": 0.5320842105263158,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002323159048607275,
+      "loss": 3.0394,
+      "step": 12637
+    },
+    {
+      "epoch": 0.5321263157894737,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00023228224526784162,
+      "loss": 2.9287,
+      "step": 12638
+    },
+    {
+      "epoch": 0.5321684210526316,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00023224858599775228,
+      "loss": 3.185,
+      "step": 12639
+    },
+    {
+      "epoch": 0.5322105263157895,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00023221492705107293,
+      "loss": 2.8927,
+      "step": 12640
+    },
+    {
+      "epoch": 0.5322526315789474,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00023218126842841663,
+      "loss": 3.2685,
+      "step": 12641
+    },
+    {
+      "epoch": 0.5322947368421053,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002321476101303967,
+      "loss": 3.0433,
+      "step": 12642
+    },
+    {
+      "epoch": 0.5323368421052631,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00023211395215762636,
+      "loss": 3.1253,
+      "step": 12643
+    },
+    {
+      "epoch": 0.5323789473684211,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00023208029451071878,
+      "loss": 3.5283,
+      "step": 12644
+    },
+    {
+      "epoch": 0.5324210526315789,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00023204663719028726,
+      "loss": 3.4994,
+      "step": 12645
+    },
+    {
+      "epoch": 0.5324631578947369,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002320129801969448,
+      "loss": 3.1769,
+      "step": 12646
+    },
+    {
+      "epoch": 0.5325052631578947,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00023197932353130485,
+      "loss": 2.718,
+      "step": 12647
+    },
+    {
+      "epoch": 0.5325473684210527,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002319456671939803,
+      "loss": 2.8629,
+      "step": 12648
+    },
+    {
+      "epoch": 0.5325894736842105,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00023191201118558452,
+      "loss": 2.8033,
+      "step": 12649
+    },
+    {
+      "epoch": 0.5326315789473685,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00023187835550673067,
+      "loss": 3.1805,
+      "step": 12650
+    },
+    {
+      "epoch": 0.5326736842105263,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00023184470015803185,
+      "loss": 3.4089,
+      "step": 12651
+    },
+    {
+      "epoch": 0.5327157894736843,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00023181104514010138,
+      "loss": 3.0306,
+      "step": 12652
+    },
+    {
+      "epoch": 0.5327578947368421,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00023177739045355215,
+      "loss": 3.2424,
+      "step": 12653
+    },
+    {
+      "epoch": 0.5328,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00023174373609899759,
+      "loss": 2.9711,
+      "step": 12654
+    },
+    {
+      "epoch": 0.5328421052631579,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0002317100820770506,
+      "loss": 3.0502,
+      "step": 12655
+    },
+    {
+      "epoch": 0.5328842105263157,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00023167642838832443,
+      "loss": 2.8089,
+      "step": 12656
+    },
+    {
+      "epoch": 0.5329263157894737,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00023164277503343232,
+      "loss": 3.4106,
+      "step": 12657
+    },
+    {
+      "epoch": 0.5329684210526315,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00023160912201298714,
+      "loss": 3.6643,
+      "step": 12658
+    },
+    {
+      "epoch": 0.5330105263157895,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002315754693276022,
+      "loss": 2.6841,
+      "step": 12659
+    },
+    {
+      "epoch": 0.5330526315789473,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00023154181697789054,
+      "loss": 3.1348,
+      "step": 12660
+    },
+    {
+      "epoch": 0.5330947368421053,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00023150816496446527,
+      "loss": 3.4332,
+      "step": 12661
+    },
+    {
+      "epoch": 0.5331368421052631,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00023147451328793962,
+      "loss": 3.7398,
+      "step": 12662
+    },
+    {
+      "epoch": 0.5331789473684211,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00023144086194892648,
+      "loss": 2.7761,
+      "step": 12663
+    },
+    {
+      "epoch": 0.5332210526315789,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00023140721094803914,
+      "loss": 2.8294,
+      "step": 12664
+    },
+    {
+      "epoch": 0.5332631578947369,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002313735602858904,
+      "loss": 2.7632,
+      "step": 12665
+    },
+    {
+      "epoch": 0.5333052631578947,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00023133990996309362,
+      "loss": 2.5546,
+      "step": 12666
+    },
+    {
+      "epoch": 0.5333473684210527,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00023130625998026174,
+      "loss": 3.4901,
+      "step": 12667
+    },
+    {
+      "epoch": 0.5333894736842105,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00023127261033800779,
+      "loss": 3.3508,
+      "step": 12668
+    },
+    {
+      "epoch": 0.5334315789473684,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00023123896103694491,
+      "loss": 2.7997,
+      "step": 12669
+    },
+    {
+      "epoch": 0.5334736842105263,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00023120531207768607,
+      "loss": 2.8876,
+      "step": 12670
+    },
+    {
+      "epoch": 0.5335157894736842,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002311716634608445,
+      "loss": 3.0089,
+      "step": 12671
+    },
+    {
+      "epoch": 0.5335578947368421,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00023113801518703295,
+      "loss": 3.1059,
+      "step": 12672
+    },
+    {
+      "epoch": 0.5336,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00023110436725686475,
+      "loss": 3.026,
+      "step": 12673
+    },
+    {
+      "epoch": 0.5336421052631579,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00023107071967095263,
+      "loss": 3.6347,
+      "step": 12674
+    },
+    {
+      "epoch": 0.5336842105263158,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00023103707242990977,
+      "loss": 2.9436,
+      "step": 12675
+    },
+    {
+      "epoch": 0.5337263157894737,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002310034255343492,
+      "loss": 3.2199,
+      "step": 12676
+    },
+    {
+      "epoch": 0.5337684210526316,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002309697789848839,
+      "loss": 2.837,
+      "step": 12677
+    },
+    {
+      "epoch": 0.5338105263157895,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00023093613278212693,
+      "loss": 3.2399,
+      "step": 12678
+    },
+    {
+      "epoch": 0.5338526315789474,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00023090248692669114,
+      "loss": 2.9832,
+      "step": 12679
+    },
+    {
+      "epoch": 0.5338947368421053,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00023086884141918968,
+      "loss": 2.8075,
+      "step": 12680
+    },
+    {
+      "epoch": 0.5339368421052632,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00023083519626023537,
+      "loss": 2.996,
+      "step": 12681
+    },
+    {
+      "epoch": 0.533978947368421,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00023080155145044126,
+      "loss": 3.3807,
+      "step": 12682
+    },
+    {
+      "epoch": 0.534021052631579,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00023076790699042045,
+      "loss": 3.3397,
+      "step": 12683
+    },
+    {
+      "epoch": 0.5340631578947368,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00023073426288078562,
+      "loss": 2.9738,
+      "step": 12684
+    },
+    {
+      "epoch": 0.5341052631578947,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00023070061912214995,
+      "loss": 3.1122,
+      "step": 12685
+    },
+    {
+      "epoch": 0.5341473684210526,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002306669757151263,
+      "loss": 3.2446,
+      "step": 12686
+    },
+    {
+      "epoch": 0.5341894736842105,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00023063333266032775,
+      "loss": 3.2109,
+      "step": 12687
+    },
+    {
+      "epoch": 0.5342315789473684,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00023059968995836698,
+      "loss": 3.0674,
+      "step": 12688
+    },
+    {
+      "epoch": 0.5342736842105263,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00023056604760985708,
+      "loss": 2.9957,
+      "step": 12689
+    },
+    {
+      "epoch": 0.5343157894736842,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00023053240561541106,
+      "loss": 3.1843,
+      "step": 12690
+    },
+    {
+      "epoch": 0.5343578947368421,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00023049876397564166,
+      "loss": 3.1221,
+      "step": 12691
+    },
+    {
+      "epoch": 0.5344,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00023046512269116187,
+      "loss": 3.128,
+      "step": 12692
+    },
+    {
+      "epoch": 0.5344421052631579,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00023043148176258457,
+      "loss": 3.1448,
+      "step": 12693
+    },
+    {
+      "epoch": 0.5344842105263158,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00023039784119052278,
+      "loss": 2.991,
+      "step": 12694
+    },
+    {
+      "epoch": 0.5345263157894737,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002303642009755892,
+      "loss": 3.4573,
+      "step": 12695
+    },
+    {
+      "epoch": 0.5345684210526316,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00023033056111839678,
+      "loss": 3.2706,
+      "step": 12696
+    },
+    {
+      "epoch": 0.5346105263157894,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00023029692161955856,
+      "loss": 2.9666,
+      "step": 12697
+    },
+    {
+      "epoch": 0.5346526315789474,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00023026328247968716,
+      "loss": 3.0784,
+      "step": 12698
+    },
+    {
+      "epoch": 0.5346947368421052,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00023022964369939566,
+      "loss": 3.332,
+      "step": 12699
+    },
+    {
+      "epoch": 0.5347368421052632,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00023019600527929675,
+      "loss": 3.3824,
+      "step": 12700
+    },
+    {
+      "epoch": 0.534778947368421,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00023016236722000338,
+      "loss": 3.0983,
+      "step": 12701
+    },
+    {
+      "epoch": 0.534821052631579,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00023012872952212832,
+      "loss": 2.9692,
+      "step": 12702
+    },
+    {
+      "epoch": 0.5348631578947368,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00023009509218628444,
+      "loss": 3.6116,
+      "step": 12703
+    },
+    {
+      "epoch": 0.5349052631578948,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00023006145521308475,
+      "loss": 3.3333,
+      "step": 12704
+    },
+    {
+      "epoch": 0.5349473684210526,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00023002781860314176,
+      "loss": 2.5794,
+      "step": 12705
+    },
+    {
+      "epoch": 0.5349894736842106,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002299941823570686,
+      "loss": 3.3274,
+      "step": 12706
+    },
+    {
+      "epoch": 0.5350315789473684,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002299605464754778,
+      "loss": 3.0704,
+      "step": 12707
+    },
+    {
+      "epoch": 0.5350736842105264,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00022992691095898235,
+      "loss": 2.479,
+      "step": 12708
+    },
+    {
+      "epoch": 0.5351157894736842,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00022989327580819494,
+      "loss": 3.4943,
+      "step": 12709
+    },
+    {
+      "epoch": 0.535157894736842,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002298596410237285,
+      "loss": 3.3364,
+      "step": 12710
+    },
+    {
+      "epoch": 0.5352,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0002298260066061957,
+      "loss": 2.7743,
+      "step": 12711
+    },
+    {
+      "epoch": 0.5352421052631579,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00022979237255620934,
+      "loss": 3.2902,
+      "step": 12712
+    },
+    {
+      "epoch": 0.5352842105263158,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0002297587388743823,
+      "loss": 3.3277,
+      "step": 12713
+    },
+    {
+      "epoch": 0.5353263157894737,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00022972510556132714,
+      "loss": 3.544,
+      "step": 12714
+    },
+    {
+      "epoch": 0.5353684210526316,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00022969147261765676,
+      "loss": 3.2444,
+      "step": 12715
+    },
+    {
+      "epoch": 0.5354105263157894,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00022965784004398397,
+      "loss": 3.186,
+      "step": 12716
+    },
+    {
+      "epoch": 0.5354526315789474,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002296242078409213,
+      "loss": 3.4802,
+      "step": 12717
+    },
+    {
+      "epoch": 0.5354947368421052,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00022959057600908174,
+      "loss": 3.1814,
+      "step": 12718
+    },
+    {
+      "epoch": 0.5355368421052632,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002295569445490778,
+      "loss": 2.8028,
+      "step": 12719
+    },
+    {
+      "epoch": 0.535578947368421,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00022952331346152245,
+      "loss": 2.7376,
+      "step": 12720
+    },
+    {
+      "epoch": 0.535621052631579,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00022948968274702814,
+      "loss": 2.7118,
+      "step": 12721
+    },
+    {
+      "epoch": 0.5356631578947368,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00022945605240620774,
+      "loss": 2.7902,
+      "step": 12722
+    },
+    {
+      "epoch": 0.5357052631578947,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.000229422422439674,
+      "loss": 3.1385,
+      "step": 12723
+    },
+    {
+      "epoch": 0.5357473684210526,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00022938879284803947,
+      "loss": 3.6393,
+      "step": 12724
+    },
+    {
+      "epoch": 0.5357894736842105,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00022935516363191695,
+      "loss": 3.3796,
+      "step": 12725
+    },
+    {
+      "epoch": 0.5358315789473684,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00022932153479191908,
+      "loss": 3.0666,
+      "step": 12726
+    },
+    {
+      "epoch": 0.5358736842105263,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00022928790632865856,
+      "loss": 2.6989,
+      "step": 12727
+    },
+    {
+      "epoch": 0.5359157894736842,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00022925427824274804,
+      "loss": 3.2232,
+      "step": 12728
+    },
+    {
+      "epoch": 0.5359578947368421,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00022922065053480013,
+      "loss": 3.3304,
+      "step": 12729
+    },
+    {
+      "epoch": 0.536,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002291870232054277,
+      "loss": 3.2385,
+      "step": 12730
+    },
+    {
+      "epoch": 0.5360421052631579,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00022915339625524312,
+      "loss": 3.0788,
+      "step": 12731
+    },
+    {
+      "epoch": 0.5360842105263158,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002291197696848593,
+      "loss": 3.4092,
+      "step": 12732
+    },
+    {
+      "epoch": 0.5361263157894737,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00022908614349488861,
+      "loss": 3.0481,
+      "step": 12733
+    },
+    {
+      "epoch": 0.5361684210526316,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002290525176859439,
+      "loss": 2.987,
+      "step": 12734
+    },
+    {
+      "epoch": 0.5362105263157895,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00022901889225863763,
+      "loss": 3.137,
+      "step": 12735
+    },
+    {
+      "epoch": 0.5362526315789473,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00022898526721358247,
+      "loss": 2.9314,
+      "step": 12736
+    },
+    {
+      "epoch": 0.5362947368421053,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002289516425513912,
+      "loss": 3.1312,
+      "step": 12737
+    },
+    {
+      "epoch": 0.5363368421052631,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00022891801827267614,
+      "loss": 2.8295,
+      "step": 12738
+    },
+    {
+      "epoch": 0.5363789473684211,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00022888439437805015,
+      "loss": 2.8614,
+      "step": 12739
+    },
+    {
+      "epoch": 0.5364210526315789,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00022885077086812552,
+      "loss": 3.3308,
+      "step": 12740
+    },
+    {
+      "epoch": 0.5364631578947369,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00022881714774351512,
+      "loss": 3.3051,
+      "step": 12741
+    },
+    {
+      "epoch": 0.5365052631578947,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00022878352500483134,
+      "loss": 2.389,
+      "step": 12742
+    },
+    {
+      "epoch": 0.5365473684210527,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0002287499026526868,
+      "loss": 3.3163,
+      "step": 12743
+    },
+    {
+      "epoch": 0.5365894736842105,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002287162806876941,
+      "loss": 3.2768,
+      "step": 12744
+    },
+    {
+      "epoch": 0.5366315789473685,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00022868265911046569,
+      "loss": 3.2537,
+      "step": 12745
+    },
+    {
+      "epoch": 0.5366736842105263,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002286490379216143,
+      "loss": 3.0788,
+      "step": 12746
+    },
+    {
+      "epoch": 0.5367157894736843,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00022861541712175228,
+      "loss": 2.7111,
+      "step": 12747
+    },
+    {
+      "epoch": 0.5367578947368421,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002285817967114923,
+      "loss": 3.0657,
+      "step": 12748
+    },
+    {
+      "epoch": 0.5368,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002285481766914467,
+      "loss": 2.9346,
+      "step": 12749
+    },
+    {
+      "epoch": 0.5368421052631579,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00022851455706222816,
+      "loss": 3.2789,
+      "step": 12750
+    },
+    {
+      "epoch": 0.5368842105263157,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00022848093782444916,
+      "loss": 3.2632,
+      "step": 12751
+    },
+    {
+      "epoch": 0.5369263157894737,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00022844731897872216,
+      "loss": 3.3419,
+      "step": 12752
+    },
+    {
+      "epoch": 0.5369684210526315,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00022841370052565976,
+      "loss": 3.1362,
+      "step": 12753
+    },
+    {
+      "epoch": 0.5370105263157895,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002283800824658743,
+      "loss": 3.3425,
+      "step": 12754
+    },
+    {
+      "epoch": 0.5370526315789473,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00022834646479997844,
+      "loss": 3.169,
+      "step": 12755
+    },
+    {
+      "epoch": 0.5370947368421053,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00022831284752858444,
+      "loss": 2.8892,
+      "step": 12756
+    },
+    {
+      "epoch": 0.5371368421052631,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002282792306523049,
+      "loss": 3.0912,
+      "step": 12757
+    },
+    {
+      "epoch": 0.5371789473684211,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002282456141717523,
+      "loss": 3.2636,
+      "step": 12758
+    },
+    {
+      "epoch": 0.5372210526315789,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00022821199808753897,
+      "loss": 3.4331,
+      "step": 12759
+    },
+    {
+      "epoch": 0.5372631578947369,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00022817838240027754,
+      "loss": 3.2414,
+      "step": 12760
+    },
+    {
+      "epoch": 0.5373052631578947,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00022814476711058022,
+      "loss": 2.6413,
+      "step": 12761
+    },
+    {
+      "epoch": 0.5373473684210527,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002281111522190596,
+      "loss": 2.685,
+      "step": 12762
+    },
+    {
+      "epoch": 0.5373894736842105,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00022807753772632818,
+      "loss": 3.3057,
+      "step": 12763
+    },
+    {
+      "epoch": 0.5374315789473684,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0002280439236329981,
+      "loss": 3.4439,
+      "step": 12764
+    },
+    {
+      "epoch": 0.5374736842105263,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002280103099396821,
+      "loss": 2.6983,
+      "step": 12765
+    },
+    {
+      "epoch": 0.5375157894736842,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002279766966469923,
+      "loss": 3.5582,
+      "step": 12766
+    },
+    {
+      "epoch": 0.5375578947368421,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00022794308375554128,
+      "loss": 3.3419,
+      "step": 12767
+    },
+    {
+      "epoch": 0.5376,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00022790947126594124,
+      "loss": 2.8263,
+      "step": 12768
+    },
+    {
+      "epoch": 0.5376421052631579,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00022787585917880479,
+      "loss": 3.1683,
+      "step": 12769
+    },
+    {
+      "epoch": 0.5376842105263158,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002278422474947442,
+      "loss": 2.8922,
+      "step": 12770
+    },
+    {
+      "epoch": 0.5377263157894737,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00022780863621437174,
+      "loss": 2.9278,
+      "step": 12771
+    },
+    {
+      "epoch": 0.5377684210526316,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00022777502533829997,
+      "loss": 3.5117,
+      "step": 12772
+    },
+    {
+      "epoch": 0.5378105263157895,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.000227741414867141,
+      "loss": 3.3157,
+      "step": 12773
+    },
+    {
+      "epoch": 0.5378526315789474,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00022770780480150743,
+      "loss": 2.7496,
+      "step": 12774
+    },
+    {
+      "epoch": 0.5378947368421053,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00022767419514201136,
+      "loss": 2.9335,
+      "step": 12775
+    },
+    {
+      "epoch": 0.5379368421052632,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002276405858892652,
+      "loss": 2.9917,
+      "step": 12776
+    },
+    {
+      "epoch": 0.537978947368421,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00022760697704388136,
+      "loss": 3.0267,
+      "step": 12777
+    },
+    {
+      "epoch": 0.538021052631579,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00022757336860647204,
+      "loss": 3.0119,
+      "step": 12778
+    },
+    {
+      "epoch": 0.5380631578947368,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00022753976057764968,
+      "loss": 3.0938,
+      "step": 12779
+    },
+    {
+      "epoch": 0.5381052631578948,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00022750615295802644,
+      "loss": 3.0561,
+      "step": 12780
+    },
+    {
+      "epoch": 0.5381473684210526,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002274725457482147,
+      "loss": 3.4779,
+      "step": 12781
+    },
+    {
+      "epoch": 0.5381894736842106,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00022743893894882665,
+      "loss": 2.5003,
+      "step": 12782
+    },
+    {
+      "epoch": 0.5382315789473684,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002274053325604746,
+      "loss": 3.4628,
+      "step": 12783
+    },
+    {
+      "epoch": 0.5382736842105263,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00022737172658377093,
+      "loss": 2.8963,
+      "step": 12784
+    },
+    {
+      "epoch": 0.5383157894736842,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00022733812101932778,
+      "loss": 2.941,
+      "step": 12785
+    },
+    {
+      "epoch": 0.5383578947368421,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00022730451586775745,
+      "loss": 2.7317,
+      "step": 12786
+    },
+    {
+      "epoch": 0.5384,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002272709111296721,
+      "loss": 2.7255,
+      "step": 12787
+    },
+    {
+      "epoch": 0.5384421052631579,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00022723730680568416,
+      "loss": 3.1659,
+      "step": 12788
+    },
+    {
+      "epoch": 0.5384842105263158,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00022720370289640564,
+      "loss": 3.1836,
+      "step": 12789
+    },
+    {
+      "epoch": 0.5385263157894736,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002271700994024489,
+      "loss": 3.4736,
+      "step": 12790
+    },
+    {
+      "epoch": 0.5385684210526316,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002271364963244262,
+      "loss": 3.1953,
+      "step": 12791
+    },
+    {
+      "epoch": 0.5386105263157894,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0002271028936629496,
+      "loss": 2.4013,
+      "step": 12792
+    },
+    {
+      "epoch": 0.5386526315789474,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002270692914186314,
+      "loss": 2.7199,
+      "step": 12793
+    },
+    {
+      "epoch": 0.5386947368421052,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002270356895920837,
+      "loss": 3.281,
+      "step": 12794
+    },
+    {
+      "epoch": 0.5387368421052632,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00022700208818391887,
+      "loss": 3.1711,
+      "step": 12795
+    },
+    {
+      "epoch": 0.538778947368421,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00022696848719474887,
+      "loss": 3.2292,
+      "step": 12796
+    },
+    {
+      "epoch": 0.538821052631579,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00022693488662518595,
+      "loss": 3.1128,
+      "step": 12797
+    },
+    {
+      "epoch": 0.5388631578947368,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00022690128647584245,
+      "loss": 3.1284,
+      "step": 12798
+    },
+    {
+      "epoch": 0.5389052631578948,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00022686768674733024,
+      "loss": 3.068,
+      "step": 12799
+    },
+    {
+      "epoch": 0.5389473684210526,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00022683408744026171,
+      "loss": 2.8425,
+      "step": 12800
+    },
+    {
+      "epoch": 0.5389894736842106,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00022680048855524878,
+      "loss": 3.1166,
+      "step": 12801
+    },
+    {
+      "epoch": 0.5390315789473684,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00022676689009290375,
+      "loss": 3.3533,
+      "step": 12802
+    },
+    {
+      "epoch": 0.5390736842105263,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00022673329205383862,
+      "loss": 2.9523,
+      "step": 12803
+    },
+    {
+      "epoch": 0.5391157894736842,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00022669969443866555,
+      "loss": 3.4934,
+      "step": 12804
+    },
+    {
+      "epoch": 0.5391578947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002266660972479968,
+      "loss": 3.3707,
+      "step": 12805
+    },
+    {
+      "epoch": 0.5392,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00022663250048244425,
+      "loss": 2.6656,
+      "step": 12806
+    },
+    {
+      "epoch": 0.5392421052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00022659890414262018,
+      "loss": 3.4155,
+      "step": 12807
+    },
+    {
+      "epoch": 0.5392842105263158,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00022656530822913647,
+      "loss": 3.1296,
+      "step": 12808
+    },
+    {
+      "epoch": 0.5393263157894737,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002265317127426053,
+      "loss": 2.7463,
+      "step": 12809
+    },
+    {
+      "epoch": 0.5393684210526316,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00022649811768363882,
+      "loss": 2.7752,
+      "step": 12810
+    },
+    {
+      "epoch": 0.5394105263157895,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00022646452305284898,
+      "loss": 3.2148,
+      "step": 12811
+    },
+    {
+      "epoch": 0.5394526315789474,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00022643092885084795,
+      "loss": 3.0524,
+      "step": 12812
+    },
+    {
+      "epoch": 0.5394947368421052,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002263973350782476,
+      "loss": 2.7869,
+      "step": 12813
+    },
+    {
+      "epoch": 0.5395368421052632,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002263637417356602,
+      "loss": 2.6607,
+      "step": 12814
+    },
+    {
+      "epoch": 0.539578947368421,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00022633014882369757,
+      "loss": 3.1635,
+      "step": 12815
+    },
+    {
+      "epoch": 0.539621052631579,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0002262965563429718,
+      "loss": 2.4586,
+      "step": 12816
+    },
+    {
+      "epoch": 0.5396631578947368,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00022626296429409498,
+      "loss": 3.3586,
+      "step": 12817
+    },
+    {
+      "epoch": 0.5397052631578947,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00022622937267767902,
+      "loss": 3.0057,
+      "step": 12818
+    },
+    {
+      "epoch": 0.5397473684210526,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00022619578149433599,
+      "loss": 3.1835,
+      "step": 12819
+    },
+    {
+      "epoch": 0.5397894736842105,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00022616219074467778,
+      "loss": 2.8604,
+      "step": 12820
+    },
+    {
+      "epoch": 0.5398315789473684,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00022612860042931656,
+      "loss": 3.3622,
+      "step": 12821
+    },
+    {
+      "epoch": 0.5398736842105263,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002260950105488641,
+      "loss": 3.2495,
+      "step": 12822
+    },
+    {
+      "epoch": 0.5399157894736842,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00022606142110393246,
+      "loss": 3.1095,
+      "step": 12823
+    },
+    {
+      "epoch": 0.5399578947368421,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002260278320951337,
+      "loss": 2.8223,
+      "step": 12824
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00022599424352307955,
+      "loss": 3.3127,
+      "step": 12825
+    },
+    {
+      "epoch": 0.5400421052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00022596065538838217,
+      "loss": 3.2161,
+      "step": 12826
+    },
+    {
+      "epoch": 0.5400842105263158,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002259270676916534,
+      "loss": 3.2006,
+      "step": 12827
+    },
+    {
+      "epoch": 0.5401263157894737,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00022589348043350514,
+      "loss": 3.3813,
+      "step": 12828
+    },
+    {
+      "epoch": 0.5401684210526316,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002258598936145493,
+      "loss": 3.3803,
+      "step": 12829
+    },
+    {
+      "epoch": 0.5402105263157895,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002258263072353978,
+      "loss": 3.2231,
+      "step": 12830
+    },
+    {
+      "epoch": 0.5402526315789473,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00022579272129666275,
+      "loss": 2.9005,
+      "step": 12831
+    },
+    {
+      "epoch": 0.5402947368421053,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00022575913579895576,
+      "loss": 3.5052,
+      "step": 12832
+    },
+    {
+      "epoch": 0.5403368421052631,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00022572555074288893,
+      "loss": 3.1204,
+      "step": 12833
+    },
+    {
+      "epoch": 0.5403789473684211,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002256919661290739,
+      "loss": 3.284,
+      "step": 12834
+    },
+    {
+      "epoch": 0.5404210526315789,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002256583819581228,
+      "loss": 3.4327,
+      "step": 12835
+    },
+    {
+      "epoch": 0.5404631578947369,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002256247982306473,
+      "loss": 3.0934,
+      "step": 12836
+    },
+    {
+      "epoch": 0.5405052631578947,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00022559121494725932,
+      "loss": 3.1555,
+      "step": 12837
+    },
+    {
+      "epoch": 0.5405473684210527,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00022555763210857087,
+      "loss": 3.2322,
+      "step": 12838
+    },
+    {
+      "epoch": 0.5405894736842105,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002255240497151935,
+      "loss": 3.1816,
+      "step": 12839
+    },
+    {
+      "epoch": 0.5406315789473685,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00022549046776773932,
+      "loss": 3.1325,
+      "step": 12840
+    },
+    {
+      "epoch": 0.5406736842105263,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002254568862668199,
+      "loss": 3.2035,
+      "step": 12841
+    },
+    {
+      "epoch": 0.5407157894736843,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002254233052130472,
+      "loss": 2.8648,
+      "step": 12842
+    },
+    {
+      "epoch": 0.5407578947368421,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00022538972460703306,
+      "loss": 3.1046,
+      "step": 12843
+    },
+    {
+      "epoch": 0.5408,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00022535614444938912,
+      "loss": 2.9757,
+      "step": 12844
+    },
+    {
+      "epoch": 0.5408421052631579,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00022532256474072735,
+      "loss": 2.9846,
+      "step": 12845
+    },
+    {
+      "epoch": 0.5408842105263157,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002252889854816594,
+      "loss": 2.9782,
+      "step": 12846
+    },
+    {
+      "epoch": 0.5409263157894737,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00022525540667279721,
+      "loss": 3.0351,
+      "step": 12847
+    },
+    {
+      "epoch": 0.5409684210526315,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00022522182831475232,
+      "loss": 2.6855,
+      "step": 12848
+    },
+    {
+      "epoch": 0.5410105263157895,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00022518825040813672,
+      "loss": 3.3166,
+      "step": 12849
+    },
+    {
+      "epoch": 0.5410526315789473,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00022515467295356194,
+      "loss": 3.3633,
+      "step": 12850
+    },
+    {
+      "epoch": 0.5410947368421053,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002251210959516398,
+      "loss": 3.0377,
+      "step": 12851
+    },
+    {
+      "epoch": 0.5411368421052631,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00022508751940298217,
+      "loss": 3.2266,
+      "step": 12852
+    },
+    {
+      "epoch": 0.5411789473684211,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00022505394330820056,
+      "loss": 2.6111,
+      "step": 12853
+    },
+    {
+      "epoch": 0.5412210526315789,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00022502036766790693,
+      "loss": 2.9188,
+      "step": 12854
+    },
+    {
+      "epoch": 0.5412631578947369,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00022498679248271272,
+      "loss": 3.301,
+      "step": 12855
+    },
+    {
+      "epoch": 0.5413052631578947,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00022495321775322988,
+      "loss": 2.6922,
+      "step": 12856
+    },
+    {
+      "epoch": 0.5413473684210526,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00022491964348006987,
+      "loss": 2.8302,
+      "step": 12857
+    },
+    {
+      "epoch": 0.5413894736842105,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002248860696638445,
+      "loss": 3.3484,
+      "step": 12858
+    },
+    {
+      "epoch": 0.5414315789473684,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00022485249630516555,
+      "loss": 2.8995,
+      "step": 12859
+    },
+    {
+      "epoch": 0.5414736842105263,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00022481892340464444,
+      "loss": 2.9962,
+      "step": 12860
+    },
+    {
+      "epoch": 0.5415157894736842,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00022478535096289302,
+      "loss": 3.1667,
+      "step": 12861
+    },
+    {
+      "epoch": 0.5415578947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00022475177898052283,
+      "loss": 3.3313,
+      "step": 12862
+    },
+    {
+      "epoch": 0.5416,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00022471820745814555,
+      "loss": 3.1258,
+      "step": 12863
+    },
+    {
+      "epoch": 0.5416421052631579,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00022468463639637294,
+      "loss": 2.9057,
+      "step": 12864
+    },
+    {
+      "epoch": 0.5416842105263158,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002246510657958164,
+      "loss": 2.7129,
+      "step": 12865
+    },
+    {
+      "epoch": 0.5417263157894737,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00022461749565708778,
+      "loss": 3.1935,
+      "step": 12866
+    },
+    {
+      "epoch": 0.5417684210526316,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00022458392598079843,
+      "loss": 3.1722,
+      "step": 12867
+    },
+    {
+      "epoch": 0.5418105263157895,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00022455035676756017,
+      "loss": 3.047,
+      "step": 12868
+    },
+    {
+      "epoch": 0.5418526315789474,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00022451678801798446,
+      "loss": 3.0507,
+      "step": 12869
+    },
+    {
+      "epoch": 0.5418947368421053,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00022448321973268293,
+      "loss": 3.0461,
+      "step": 12870
+    },
+    {
+      "epoch": 0.5419368421052632,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00022444965191226722,
+      "loss": 3.2732,
+      "step": 12871
+    },
+    {
+      "epoch": 0.541978947368421,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00022441608455734874,
+      "loss": 3.188,
+      "step": 12872
+    },
+    {
+      "epoch": 0.542021052631579,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00022438251766853927,
+      "loss": 3.0988,
+      "step": 12873
+    },
+    {
+      "epoch": 0.5420631578947368,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00022434895124645007,
+      "loss": 3.1495,
+      "step": 12874
+    },
+    {
+      "epoch": 0.5421052631578948,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.000224315385291693,
+      "loss": 2.893,
+      "step": 12875
+    },
+    {
+      "epoch": 0.5421473684210526,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002242818198048793,
+      "loss": 3.1744,
+      "step": 12876
+    },
+    {
+      "epoch": 0.5421894736842106,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00022424825478662065,
+      "loss": 2.8689,
+      "step": 12877
+    },
+    {
+      "epoch": 0.5422315789473684,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00022421469023752857,
+      "loss": 3.0869,
+      "step": 12878
+    },
+    {
+      "epoch": 0.5422736842105264,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002241811261582145,
+      "loss": 3.2633,
+      "step": 12879
+    },
+    {
+      "epoch": 0.5423157894736842,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00022414756254929006,
+      "loss": 3.3491,
+      "step": 12880
+    },
+    {
+      "epoch": 0.5423578947368421,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00022411399941136656,
+      "loss": 3.4902,
+      "step": 12881
+    },
+    {
+      "epoch": 0.5424,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00022408043674505568,
+      "loss": 3.3735,
+      "step": 12882
+    },
+    {
+      "epoch": 0.542442105263158,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002240468745509687,
+      "loss": 2.8697,
+      "step": 12883
+    },
+    {
+      "epoch": 0.5424842105263158,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00022401331282971717,
+      "loss": 3.3974,
+      "step": 12884
+    },
+    {
+      "epoch": 0.5425263157894736,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00022397975158191258,
+      "loss": 3.6307,
+      "step": 12885
+    },
+    {
+      "epoch": 0.5425684210526316,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00022394619080816637,
+      "loss": 2.809,
+      "step": 12886
+    },
+    {
+      "epoch": 0.5426105263157894,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00022391263050908992,
+      "loss": 2.9324,
+      "step": 12887
+    },
+    {
+      "epoch": 0.5426526315789474,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00022387907068529466,
+      "loss": 2.7167,
+      "step": 12888
+    },
+    {
+      "epoch": 0.5426947368421052,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00022384551133739213,
+      "loss": 3.0064,
+      "step": 12889
+    },
+    {
+      "epoch": 0.5427368421052632,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00022381195246599356,
+      "loss": 3.0278,
+      "step": 12890
+    },
+    {
+      "epoch": 0.542778947368421,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00022377839407171043,
+      "loss": 2.9579,
+      "step": 12891
+    },
+    {
+      "epoch": 0.542821052631579,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002237448361551543,
+      "loss": 2.8737,
+      "step": 12892
+    },
+    {
+      "epoch": 0.5428631578947368,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00022371127871693625,
+      "loss": 3.0868,
+      "step": 12893
+    },
+    {
+      "epoch": 0.5429052631578948,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00022367772175766788,
+      "loss": 3.3868,
+      "step": 12894
+    },
+    {
+      "epoch": 0.5429473684210526,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00022364416527796046,
+      "loss": 2.7747,
+      "step": 12895
+    },
+    {
+      "epoch": 0.5429894736842106,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00022361060927842546,
+      "loss": 3.3441,
+      "step": 12896
+    },
+    {
+      "epoch": 0.5430315789473684,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00022357705375967407,
+      "loss": 3.1039,
+      "step": 12897
+    },
+    {
+      "epoch": 0.5430736842105263,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00022354349872231769,
+      "loss": 3.1572,
+      "step": 12898
+    },
+    {
+      "epoch": 0.5431157894736842,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00022350994416696783,
+      "loss": 4.0993,
+      "step": 12899
+    },
+    {
+      "epoch": 0.5431578947368421,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00022347639009423552,
+      "loss": 3.196,
+      "step": 12900
+    },
+    {
+      "epoch": 0.5432,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00022344283650473226,
+      "loss": 3.2129,
+      "step": 12901
+    },
+    {
+      "epoch": 0.5432421052631579,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00022340928339906936,
+      "loss": 2.7864,
+      "step": 12902
+    },
+    {
+      "epoch": 0.5432842105263158,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00022337573077785804,
+      "loss": 2.785,
+      "step": 12903
+    },
+    {
+      "epoch": 0.5433263157894737,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00022334217864170954,
+      "loss": 3.0181,
+      "step": 12904
+    },
+    {
+      "epoch": 0.5433684210526316,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00022330862699123527,
+      "loss": 3.2711,
+      "step": 12905
+    },
+    {
+      "epoch": 0.5434105263157895,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00022327507582704657,
+      "loss": 3.0973,
+      "step": 12906
+    },
+    {
+      "epoch": 0.5434526315789474,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00022324152514975447,
+      "loss": 3.5134,
+      "step": 12907
+    },
+    {
+      "epoch": 0.5434947368421053,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00022320797495997046,
+      "loss": 3.1497,
+      "step": 12908
+    },
+    {
+      "epoch": 0.5435368421052632,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002231744252583055,
+      "loss": 3.5638,
+      "step": 12909
+    },
+    {
+      "epoch": 0.543578947368421,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00022314087604537102,
+      "loss": 3.4367,
+      "step": 12910
+    },
+    {
+      "epoch": 0.5436210526315789,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00022310732732177833,
+      "loss": 3.1139,
+      "step": 12911
+    },
+    {
+      "epoch": 0.5436631578947368,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00022307377908813841,
+      "loss": 2.6976,
+      "step": 12912
+    },
+    {
+      "epoch": 0.5437052631578947,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00022304023134506277,
+      "loss": 3.1863,
+      "step": 12913
+    },
+    {
+      "epoch": 0.5437473684210526,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002230066840931623,
+      "loss": 2.6206,
+      "step": 12914
+    },
+    {
+      "epoch": 0.5437894736842105,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00022297313733304845,
+      "loss": 3.1941,
+      "step": 12915
+    },
+    {
+      "epoch": 0.5438315789473684,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00022293959106533216,
+      "loss": 2.7666,
+      "step": 12916
+    },
+    {
+      "epoch": 0.5438736842105263,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00022290604529062472,
+      "loss": 2.9412,
+      "step": 12917
+    },
+    {
+      "epoch": 0.5439157894736842,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002228725000095374,
+      "loss": 3.2147,
+      "step": 12918
+    },
+    {
+      "epoch": 0.5439578947368421,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00022283895522268117,
+      "loss": 2.8983,
+      "step": 12919
+    },
+    {
+      "epoch": 0.544,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002228054109306673,
+      "loss": 2.9777,
+      "step": 12920
+    },
+    {
+      "epoch": 0.5440421052631579,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00022277186713410684,
+      "loss": 3.1171,
+      "step": 12921
+    },
+    {
+      "epoch": 0.5440842105263158,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00022273832383361113,
+      "loss": 3.5142,
+      "step": 12922
+    },
+    {
+      "epoch": 0.5441263157894737,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00022270478102979097,
+      "loss": 3.3884,
+      "step": 12923
+    },
+    {
+      "epoch": 0.5441684210526315,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00022267123872325763,
+      "loss": 3.2416,
+      "step": 12924
+    },
+    {
+      "epoch": 0.5442105263157895,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00022263769691462231,
+      "loss": 3.2345,
+      "step": 12925
+    },
+    {
+      "epoch": 0.5442526315789473,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00022260415560449592,
+      "loss": 2.8929,
+      "step": 12926
+    },
+    {
+      "epoch": 0.5442947368421053,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00022257061479348967,
+      "loss": 3.526,
+      "step": 12927
+    },
+    {
+      "epoch": 0.5443368421052631,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00022253707448221456,
+      "loss": 3.3578,
+      "step": 12928
+    },
+    {
+      "epoch": 0.5443789473684211,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00022250353467128176,
+      "loss": 3.1546,
+      "step": 12929
+    },
+    {
+      "epoch": 0.5444210526315789,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00022246999536130213,
+      "loss": 3.1805,
+      "step": 12930
+    },
+    {
+      "epoch": 0.5444631578947369,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00022243645655288687,
+      "loss": 2.9529,
+      "step": 12931
+    },
+    {
+      "epoch": 0.5445052631578947,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00022240291824664708,
+      "loss": 3.0559,
+      "step": 12932
+    },
+    {
+      "epoch": 0.5445473684210527,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00022236938044319362,
+      "loss": 3.3177,
+      "step": 12933
+    },
+    {
+      "epoch": 0.5445894736842105,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00022233584314313766,
+      "loss": 3.3355,
+      "step": 12934
+    },
+    {
+      "epoch": 0.5446315789473685,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00022230230634709002,
+      "loss": 2.9472,
+      "step": 12935
+    },
+    {
+      "epoch": 0.5446736842105263,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00022226877005566186,
+      "loss": 3.1843,
+      "step": 12936
+    },
+    {
+      "epoch": 0.5447157894736843,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002222352342694641,
+      "loss": 2.5554,
+      "step": 12937
+    },
+    {
+      "epoch": 0.5447578947368421,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00022220169898910772,
+      "loss": 3.4046,
+      "step": 12938
+    },
+    {
+      "epoch": 0.5448,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00022216816421520387,
+      "loss": 3.0626,
+      "step": 12939
+    },
+    {
+      "epoch": 0.5448421052631579,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00022213462994836322,
+      "loss": 3.0481,
+      "step": 12940
+    },
+    {
+      "epoch": 0.5448842105263157,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.000222101096189197,
+      "loss": 3.3538,
+      "step": 12941
+    },
+    {
+      "epoch": 0.5449263157894737,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00022206756293831592,
+      "loss": 3.3664,
+      "step": 12942
+    },
+    {
+      "epoch": 0.5449684210526315,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00022203403019633108,
+      "loss": 3.3199,
+      "step": 12943
+    },
+    {
+      "epoch": 0.5450105263157895,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002220004979638533,
+      "loss": 2.788,
+      "step": 12944
+    },
+    {
+      "epoch": 0.5450526315789473,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002219669662414936,
+      "loss": 3.4296,
+      "step": 12945
+    },
+    {
+      "epoch": 0.5450947368421053,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00022193343502986283,
+      "loss": 2.6046,
+      "step": 12946
+    },
+    {
+      "epoch": 0.5451368421052631,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00022189990432957185,
+      "loss": 3.1761,
+      "step": 12947
+    },
+    {
+      "epoch": 0.5451789473684211,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002218663741412317,
+      "loss": 2.9329,
+      "step": 12948
+    },
+    {
+      "epoch": 0.5452210526315789,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00022183284446545304,
+      "loss": 3.317,
+      "step": 12949
+    },
+    {
+      "epoch": 0.5452631578947369,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.000221799315302847,
+      "loss": 2.9712,
+      "step": 12950
+    },
+    {
+      "epoch": 0.5453052631578947,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00022176578665402416,
+      "loss": 2.8621,
+      "step": 12951
+    },
+    {
+      "epoch": 0.5453473684210526,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00022173225851959553,
+      "loss": 3.1188,
+      "step": 12952
+    },
+    {
+      "epoch": 0.5453894736842105,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00022169873090017197,
+      "loss": 3.3152,
+      "step": 12953
+    },
+    {
+      "epoch": 0.5454315789473684,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00022166520379636424,
+      "loss": 2.4804,
+      "step": 12954
+    },
+    {
+      "epoch": 0.5454736842105263,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00022163167720878332,
+      "loss": 3.3844,
+      "step": 12955
+    },
+    {
+      "epoch": 0.5455157894736842,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002215981511380398,
+      "loss": 2.8548,
+      "step": 12956
+    },
+    {
+      "epoch": 0.5455578947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00022156462558474456,
+      "loss": 2.9641,
+      "step": 12957
+    },
+    {
+      "epoch": 0.5456,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002215311005495086,
+      "loss": 2.8783,
+      "step": 12958
+    },
+    {
+      "epoch": 0.5456421052631579,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002214975760329424,
+      "loss": 2.9841,
+      "step": 12959
+    },
+    {
+      "epoch": 0.5456842105263158,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00022146405203565697,
+      "loss": 3.2366,
+      "step": 12960
+    },
+    {
+      "epoch": 0.5457263157894737,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002214305285582629,
+      "loss": 3.3482,
+      "step": 12961
+    },
+    {
+      "epoch": 0.5457684210526316,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002213970056013711,
+      "loss": 3.2304,
+      "step": 12962
+    },
+    {
+      "epoch": 0.5458105263157895,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00022136348316559215,
+      "loss": 3.3084,
+      "step": 12963
+    },
+    {
+      "epoch": 0.5458526315789474,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00022132996125153693,
+      "loss": 3.0206,
+      "step": 12964
+    },
+    {
+      "epoch": 0.5458947368421052,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00022129643985981622,
+      "loss": 3.0426,
+      "step": 12965
+    },
+    {
+      "epoch": 0.5459368421052632,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00022126291899104053,
+      "loss": 3.0565,
+      "step": 12966
+    },
+    {
+      "epoch": 0.545978947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00022122939864582081,
+      "loss": 3.2017,
+      "step": 12967
+    },
+    {
+      "epoch": 0.546021052631579,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00022119587882476756,
+      "loss": 3.1762,
+      "step": 12968
+    },
+    {
+      "epoch": 0.5460631578947368,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00022116235952849157,
+      "loss": 2.8913,
+      "step": 12969
+    },
+    {
+      "epoch": 0.5461052631578948,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00022112884075760347,
+      "loss": 3.1523,
+      "step": 12970
+    },
+    {
+      "epoch": 0.5461473684210526,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00022109532251271397,
+      "loss": 2.9942,
+      "step": 12971
+    },
+    {
+      "epoch": 0.5461894736842106,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00022106180479443383,
+      "loss": 3.0992,
+      "step": 12972
+    },
+    {
+      "epoch": 0.5462315789473684,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0002210282876033735,
+      "loss": 3.1492,
+      "step": 12973
+    },
+    {
+      "epoch": 0.5462736842105264,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00022099477094014386,
+      "loss": 3.43,
+      "step": 12974
+    },
+    {
+      "epoch": 0.5463157894736842,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00022096125480535527,
+      "loss": 2.8924,
+      "step": 12975
+    },
+    {
+      "epoch": 0.5463578947368422,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00022092773919961864,
+      "loss": 3.3664,
+      "step": 12976
+    },
+    {
+      "epoch": 0.5464,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002208942241235443,
+      "loss": 3.2815,
+      "step": 12977
+    },
+    {
+      "epoch": 0.5464421052631578,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00022086070957774305,
+      "loss": 3.0154,
+      "step": 12978
+    },
+    {
+      "epoch": 0.5464842105263158,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00022082719556282547,
+      "loss": 2.691,
+      "step": 12979
+    },
+    {
+      "epoch": 0.5465263157894736,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00022079368207940204,
+      "loss": 2.9776,
+      "step": 12980
+    },
+    {
+      "epoch": 0.5465684210526316,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002207601691280835,
+      "loss": 3.3093,
+      "step": 12981
+    },
+    {
+      "epoch": 0.5466105263157894,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00022072665670948025,
+      "loss": 3.3581,
+      "step": 12982
+    },
+    {
+      "epoch": 0.5466526315789474,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.000220693144824203,
+      "loss": 3.0329,
+      "step": 12983
+    },
+    {
+      "epoch": 0.5466947368421052,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00022065963347286211,
+      "loss": 3.4624,
+      "step": 12984
+    },
+    {
+      "epoch": 0.5467368421052632,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00022062612265606824,
+      "loss": 3.4156,
+      "step": 12985
+    },
+    {
+      "epoch": 0.546778947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00022059261237443196,
+      "loss": 3.272,
+      "step": 12986
+    },
+    {
+      "epoch": 0.546821052631579,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00022055910262856372,
+      "loss": 2.9962,
+      "step": 12987
+    },
+    {
+      "epoch": 0.5468631578947368,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00022052559341907406,
+      "loss": 3.574,
+      "step": 12988
+    },
+    {
+      "epoch": 0.5469052631578948,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00022049208474657337,
+      "loss": 2.7729,
+      "step": 12989
+    },
+    {
+      "epoch": 0.5469473684210526,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00022045857661167235,
+      "loss": 3.0983,
+      "step": 12990
+    },
+    {
+      "epoch": 0.5469894736842106,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00022042506901498123,
+      "loss": 3.4261,
+      "step": 12991
+    },
+    {
+      "epoch": 0.5470315789473684,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00022039156195711063,
+      "loss": 3.0474,
+      "step": 12992
+    },
+    {
+      "epoch": 0.5470736842105263,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00022035805543867108,
+      "loss": 3.1227,
+      "step": 12993
+    },
+    {
+      "epoch": 0.5471157894736842,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002203245494602728,
+      "loss": 3.6214,
+      "step": 12994
+    },
+    {
+      "epoch": 0.5471578947368421,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00022029104402252644,
+      "loss": 3.3922,
+      "step": 12995
+    },
+    {
+      "epoch": 0.5472,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0002202575391260423,
+      "loss": 2.772,
+      "step": 12996
+    },
+    {
+      "epoch": 0.5472421052631579,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00022022403477143097,
+      "loss": 2.9054,
+      "step": 12997
+    },
+    {
+      "epoch": 0.5472842105263158,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002201905309593026,
+      "loss": 2.9836,
+      "step": 12998
+    },
+    {
+      "epoch": 0.5473263157894737,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00022015702769026774,
+      "loss": 3.1555,
+      "step": 12999
+    },
+    {
+      "epoch": 0.5473684210526316,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002201235249649369,
+      "loss": 3.1737,
+      "step": 13000
+    },
+    {
+      "epoch": 0.5474105263157895,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00022009002278392022,
+      "loss": 3.1369,
+      "step": 13001
+    },
+    {
+      "epoch": 0.5474526315789474,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002200565211478282,
+      "loss": 3.1964,
+      "step": 13002
+    },
+    {
+      "epoch": 0.5474947368421053,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00022002302005727124,
+      "loss": 2.7732,
+      "step": 13003
+    },
+    {
+      "epoch": 0.5475368421052632,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002199895195128596,
+      "loss": 3.3481,
+      "step": 13004
+    },
+    {
+      "epoch": 0.547578947368421,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00021995601951520357,
+      "loss": 3.5979,
+      "step": 13005
+    },
+    {
+      "epoch": 0.5476210526315789,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002199225200649136,
+      "loss": 3.057,
+      "step": 13006
+    },
+    {
+      "epoch": 0.5476631578947369,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002198890211626001,
+      "loss": 2.7924,
+      "step": 13007
+    },
+    {
+      "epoch": 0.5477052631578947,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002198555228088731,
+      "loss": 3.0095,
+      "step": 13008
+    },
+    {
+      "epoch": 0.5477473684210526,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00021982202500434324,
+      "loss": 3.2217,
+      "step": 13009
+    },
+    {
+      "epoch": 0.5477894736842105,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00021978852774962043,
+      "loss": 3.042,
+      "step": 13010
+    },
+    {
+      "epoch": 0.5478315789473684,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002197550310453152,
+      "loss": 3.1071,
+      "step": 13011
+    },
+    {
+      "epoch": 0.5478736842105263,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002197215348920378,
+      "loss": 3.0826,
+      "step": 13012
+    },
+    {
+      "epoch": 0.5479157894736842,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00021968803929039844,
+      "loss": 3.2932,
+      "step": 13013
+    },
+    {
+      "epoch": 0.5479578947368421,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00021965454424100744,
+      "loss": 3.1672,
+      "step": 13014
+    },
+    {
+      "epoch": 0.548,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00021962104974447493,
+      "loss": 3.3085,
+      "step": 13015
+    },
+    {
+      "epoch": 0.5480421052631579,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00021958755580141128,
+      "loss": 2.853,
+      "step": 13016
+    },
+    {
+      "epoch": 0.5480842105263158,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00021955406241242652,
+      "loss": 3.1304,
+      "step": 13017
+    },
+    {
+      "epoch": 0.5481263157894737,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00021952056957813098,
+      "loss": 3.2672,
+      "step": 13018
+    },
+    {
+      "epoch": 0.5481684210526315,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00021948707729913498,
+      "loss": 3.1367,
+      "step": 13019
+    },
+    {
+      "epoch": 0.5482105263157895,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00021945358557604844,
+      "loss": 3.4987,
+      "step": 13020
+    },
+    {
+      "epoch": 0.5482526315789473,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00021942009440948176,
+      "loss": 3.2441,
+      "step": 13021
+    },
+    {
+      "epoch": 0.5482947368421053,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00021938660380004497,
+      "loss": 2.8075,
+      "step": 13022
+    },
+    {
+      "epoch": 0.5483368421052631,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00021935311374834837,
+      "loss": 3.238,
+      "step": 13023
+    },
+    {
+      "epoch": 0.5483789473684211,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00021931962425500196,
+      "loss": 2.8012,
+      "step": 13024
+    },
+    {
+      "epoch": 0.5484210526315789,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00021928613532061592,
+      "loss": 3.6176,
+      "step": 13025
+    },
+    {
+      "epoch": 0.5484631578947369,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00021925264694580055,
+      "loss": 2.7136,
+      "step": 13026
+    },
+    {
+      "epoch": 0.5485052631578947,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002192191591311657,
+      "loss": 2.8669,
+      "step": 13027
+    },
+    {
+      "epoch": 0.5485473684210527,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00021918567187732162,
+      "loss": 2.7375,
+      "step": 13028
+    },
+    {
+      "epoch": 0.5485894736842105,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00021915218518487838,
+      "loss": 3.2815,
+      "step": 13029
+    },
+    {
+      "epoch": 0.5486315789473685,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00021911869905444616,
+      "loss": 3.1672,
+      "step": 13030
+    },
+    {
+      "epoch": 0.5486736842105263,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002190852134866349,
+      "loss": 3.0757,
+      "step": 13031
+    },
+    {
+      "epoch": 0.5487157894736842,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00021905172848205468,
+      "loss": 2.9415,
+      "step": 13032
+    },
+    {
+      "epoch": 0.5487578947368421,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00021901824404131574,
+      "loss": 2.8345,
+      "step": 13033
+    },
+    {
+      "epoch": 0.5488,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00021898476016502788,
+      "loss": 3.1218,
+      "step": 13034
+    },
+    {
+      "epoch": 0.5488421052631579,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00021895127685380133,
+      "loss": 3.0361,
+      "step": 13035
+    },
+    {
+      "epoch": 0.5488842105263158,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00021891779410824596,
+      "loss": 3.1933,
+      "step": 13036
+    },
+    {
+      "epoch": 0.5489263157894737,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002188843119289719,
+      "loss": 3.0485,
+      "step": 13037
+    },
+    {
+      "epoch": 0.5489684210526315,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002188508303165891,
+      "loss": 3.4201,
+      "step": 13038
+    },
+    {
+      "epoch": 0.5490105263157895,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002188173492717075,
+      "loss": 2.993,
+      "step": 13039
+    },
+    {
+      "epoch": 0.5490526315789473,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00021878386879493731,
+      "loss": 2.7195,
+      "step": 13040
+    },
+    {
+      "epoch": 0.5490947368421053,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00021875038888688825,
+      "loss": 3.0586,
+      "step": 13041
+    },
+    {
+      "epoch": 0.5491368421052631,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00021871690954817046,
+      "loss": 3.5585,
+      "step": 13042
+    },
+    {
+      "epoch": 0.5491789473684211,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00021868343077939377,
+      "loss": 2.8942,
+      "step": 13043
+    },
+    {
+      "epoch": 0.5492210526315789,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002186499525811682,
+      "loss": 3.2324,
+      "step": 13044
+    },
+    {
+      "epoch": 0.5492631578947369,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002186164749541036,
+      "loss": 3.1854,
+      "step": 13045
+    },
+    {
+      "epoch": 0.5493052631578947,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00021858299789881001,
+      "loss": 3.2194,
+      "step": 13046
+    },
+    {
+      "epoch": 0.5493473684210526,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00021854952141589729,
+      "loss": 3.2336,
+      "step": 13047
+    },
+    {
+      "epoch": 0.5493894736842105,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00021851604550597527,
+      "loss": 2.9783,
+      "step": 13048
+    },
+    {
+      "epoch": 0.5494315789473684,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00021848257016965404,
+      "loss": 3.5081,
+      "step": 13049
+    },
+    {
+      "epoch": 0.5494736842105263,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002184490954075432,
+      "loss": 3.31,
+      "step": 13050
+    },
+    {
+      "epoch": 0.5495157894736842,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00021841562122025293,
+      "loss": 2.9811,
+      "step": 13051
+    },
+    {
+      "epoch": 0.5495578947368421,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00021838214760839278,
+      "loss": 3.1103,
+      "step": 13052
+    },
+    {
+      "epoch": 0.5496,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00021834867457257279,
+      "loss": 3.4132,
+      "step": 13053
+    },
+    {
+      "epoch": 0.5496421052631579,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00021831520211340277,
+      "loss": 3.1326,
+      "step": 13054
+    },
+    {
+      "epoch": 0.5496842105263158,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00021828173023149254,
+      "loss": 2.724,
+      "step": 13055
+    },
+    {
+      "epoch": 0.5497263157894737,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00021824825892745198,
+      "loss": 3.5347,
+      "step": 13056
+    },
+    {
+      "epoch": 0.5497684210526316,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00021821478820189074,
+      "loss": 3.0379,
+      "step": 13057
+    },
+    {
+      "epoch": 0.5498105263157895,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00021818131805541874,
+      "loss": 3.6,
+      "step": 13058
+    },
+    {
+      "epoch": 0.5498526315789474,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00021814784848864582,
+      "loss": 3.241,
+      "step": 13059
+    },
+    {
+      "epoch": 0.5498947368421052,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00021811437950218154,
+      "loss": 2.9015,
+      "step": 13060
+    },
+    {
+      "epoch": 0.5499368421052632,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002180809110966359,
+      "loss": 3.4148,
+      "step": 13061
+    },
+    {
+      "epoch": 0.549978947368421,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00021804744327261856,
+      "loss": 2.6951,
+      "step": 13062
+    },
+    {
+      "epoch": 0.550021052631579,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00021801397603073925,
+      "loss": 3.2257,
+      "step": 13063
+    },
+    {
+      "epoch": 0.5500631578947368,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00021798050937160766,
+      "loss": 3.0777,
+      "step": 13064
+    },
+    {
+      "epoch": 0.5501052631578948,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00021794704329583354,
+      "loss": 3.4933,
+      "step": 13065
+    },
+    {
+      "epoch": 0.5501473684210526,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00021791357780402676,
+      "loss": 3.3695,
+      "step": 13066
+    },
+    {
+      "epoch": 0.5501894736842106,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002178801128967968,
+      "loss": 2.7132,
+      "step": 13067
+    },
+    {
+      "epoch": 0.5502315789473684,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00021784664857475353,
+      "loss": 3.1979,
+      "step": 13068
+    },
+    {
+      "epoch": 0.5502736842105264,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00021781318483850645,
+      "loss": 3.114,
+      "step": 13069
+    },
+    {
+      "epoch": 0.5503157894736842,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0002177797216886654,
+      "loss": 3.1233,
+      "step": 13070
+    },
+    {
+      "epoch": 0.5503578947368422,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00021774625912583984,
+      "loss": 3.4415,
+      "step": 13071
+    },
+    {
+      "epoch": 0.5504,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002177127971506396,
+      "loss": 2.7947,
+      "step": 13072
+    },
+    {
+      "epoch": 0.5504421052631578,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00021767933576367435,
+      "loss": 3.2409,
+      "step": 13073
+    },
+    {
+      "epoch": 0.5504842105263158,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00021764587496555352,
+      "loss": 3.2877,
+      "step": 13074
+    },
+    {
+      "epoch": 0.5505263157894736,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00021761241475688696,
+      "loss": 3.4383,
+      "step": 13075
+    },
+    {
+      "epoch": 0.5505684210526316,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.000217578955138284,
+      "loss": 3.3968,
+      "step": 13076
+    },
+    {
+      "epoch": 0.5506105263157894,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002175454961103545,
+      "loss": 2.9002,
+      "step": 13077
+    },
+    {
+      "epoch": 0.5506526315789474,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00021751203767370788,
+      "loss": 2.8556,
+      "step": 13078
+    },
+    {
+      "epoch": 0.5506947368421052,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00021747857982895368,
+      "loss": 3.1114,
+      "step": 13079
+    },
+    {
+      "epoch": 0.5507368421052632,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00021744512257670167,
+      "loss": 3.2508,
+      "step": 13080
+    },
+    {
+      "epoch": 0.550778947368421,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00021741166591756117,
+      "loss": 3.3409,
+      "step": 13081
+    },
+    {
+      "epoch": 0.550821052631579,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00021737820985214196,
+      "loss": 3.1382,
+      "step": 13082
+    },
+    {
+      "epoch": 0.5508631578947368,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00021734475438105332,
+      "loss": 2.769,
+      "step": 13083
+    },
+    {
+      "epoch": 0.5509052631578948,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00021731129950490501,
+      "loss": 3.1677,
+      "step": 13084
+    },
+    {
+      "epoch": 0.5509473684210526,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002172778452243063,
+      "loss": 3.457,
+      "step": 13085
+    },
+    {
+      "epoch": 0.5509894736842105,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002172443915398668,
+      "loss": 3.3314,
+      "step": 13086
+    },
+    {
+      "epoch": 0.5510315789473684,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00021721093845219604,
+      "loss": 3.4098,
+      "step": 13087
+    },
+    {
+      "epoch": 0.5510736842105263,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00021717748596190343,
+      "loss": 3.3239,
+      "step": 13088
+    },
+    {
+      "epoch": 0.5511157894736842,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00021714403406959855,
+      "loss": 2.972,
+      "step": 13089
+    },
+    {
+      "epoch": 0.5511578947368421,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00021711058277589066,
+      "loss": 2.827,
+      "step": 13090
+    },
+    {
+      "epoch": 0.5512,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002170771320813894,
+      "loss": 3.5012,
+      "step": 13091
+    },
+    {
+      "epoch": 0.5512421052631579,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00021704368198670405,
+      "loss": 2.9355,
+      "step": 13092
+    },
+    {
+      "epoch": 0.5512842105263158,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00021701023249244406,
+      "loss": 2.9697,
+      "step": 13093
+    },
+    {
+      "epoch": 0.5513263157894737,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00021697678359921895,
+      "loss": 2.8756,
+      "step": 13094
+    },
+    {
+      "epoch": 0.5513684210526316,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00021694333530763798,
+      "loss": 3.2732,
+      "step": 13095
+    },
+    {
+      "epoch": 0.5514105263157895,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00021690988761831066,
+      "loss": 3.3246,
+      "step": 13096
+    },
+    {
+      "epoch": 0.5514526315789474,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00021687644053184623,
+      "loss": 3.0394,
+      "step": 13097
+    },
+    {
+      "epoch": 0.5514947368421053,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00021684299404885427,
+      "loss": 2.6014,
+      "step": 13098
+    },
+    {
+      "epoch": 0.5515368421052631,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00021680954816994386,
+      "loss": 2.7838,
+      "step": 13099
+    },
+    {
+      "epoch": 0.5515789473684211,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002167761028957245,
+      "loss": 3.5711,
+      "step": 13100
+    },
+    {
+      "epoch": 0.5516210526315789,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002167426582268056,
+      "loss": 2.8942,
+      "step": 13101
+    },
+    {
+      "epoch": 0.5516631578947369,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0002167092141637963,
+      "loss": 3.3247,
+      "step": 13102
+    },
+    {
+      "epoch": 0.5517052631578947,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00021667577070730605,
+      "loss": 3.5109,
+      "step": 13103
+    },
+    {
+      "epoch": 0.5517473684210527,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00021664232785794404,
+      "loss": 3.1419,
+      "step": 13104
+    },
+    {
+      "epoch": 0.5517894736842105,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00021660888561631975,
+      "loss": 2.8525,
+      "step": 13105
+    },
+    {
+      "epoch": 0.5518315789473685,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00021657544398304224,
+      "loss": 2.774,
+      "step": 13106
+    },
+    {
+      "epoch": 0.5518736842105263,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00021654200295872085,
+      "loss": 2.8581,
+      "step": 13107
+    },
+    {
+      "epoch": 0.5519157894736842,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00021650856254396494,
+      "loss": 3.122,
+      "step": 13108
+    },
+    {
+      "epoch": 0.5519578947368421,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002164751227393836,
+      "loss": 3.2237,
+      "step": 13109
+    },
+    {
+      "epoch": 0.552,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002164416835455862,
+      "loss": 3.129,
+      "step": 13110
+    },
+    {
+      "epoch": 0.5520421052631579,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0002164082449631818,
+      "loss": 3.2775,
+      "step": 13111
+    },
+    {
+      "epoch": 0.5520842105263158,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00021637480699277973,
+      "loss": 3.2526,
+      "step": 13112
+    },
+    {
+      "epoch": 0.5521263157894737,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00021634136963498922,
+      "loss": 3.0274,
+      "step": 13113
+    },
+    {
+      "epoch": 0.5521684210526315,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00021630793289041935,
+      "loss": 3.6067,
+      "step": 13114
+    },
+    {
+      "epoch": 0.5522105263157895,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00021627449675967945,
+      "loss": 2.595,
+      "step": 13115
+    },
+    {
+      "epoch": 0.5522526315789473,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002162410612433785,
+      "loss": 3.5471,
+      "step": 13116
+    },
+    {
+      "epoch": 0.5522947368421053,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00021620762634212584,
+      "loss": 3.3747,
+      "step": 13117
+    },
+    {
+      "epoch": 0.5523368421052631,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00021617419205653045,
+      "loss": 3.3234,
+      "step": 13118
+    },
+    {
+      "epoch": 0.5523789473684211,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00021614075838720148,
+      "loss": 3.3878,
+      "step": 13119
+    },
+    {
+      "epoch": 0.5524210526315789,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00021610732533474826,
+      "loss": 3.4799,
+      "step": 13120
+    },
+    {
+      "epoch": 0.5524631578947369,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00021607389289977964,
+      "loss": 2.7874,
+      "step": 13121
+    },
+    {
+      "epoch": 0.5525052631578947,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00021604046108290487,
+      "loss": 3.0119,
+      "step": 13122
+    },
+    {
+      "epoch": 0.5525473684210527,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00021600702988473295,
+      "loss": 3.2949,
+      "step": 13123
+    },
+    {
+      "epoch": 0.5525894736842105,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00021597359930587313,
+      "loss": 3.185,
+      "step": 13124
+    },
+    {
+      "epoch": 0.5526315789473685,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00021594016934693423,
+      "loss": 3.4764,
+      "step": 13125
+    },
+    {
+      "epoch": 0.5526736842105263,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00021590674000852538,
+      "loss": 3.2228,
+      "step": 13126
+    },
+    {
+      "epoch": 0.5527157894736842,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00021587331129125585,
+      "loss": 2.9108,
+      "step": 13127
+    },
+    {
+      "epoch": 0.5527578947368421,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00021583988319573435,
+      "loss": 3.4249,
+      "step": 13128
+    },
+    {
+      "epoch": 0.5528,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002158064557225701,
+      "loss": 2.8893,
+      "step": 13129
+    },
+    {
+      "epoch": 0.5528421052631579,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.000215773028872372,
+      "loss": 3.0339,
+      "step": 13130
+    },
+    {
+      "epoch": 0.5528842105263158,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00021573960264574922,
+      "loss": 3.1813,
+      "step": 13131
+    },
+    {
+      "epoch": 0.5529263157894737,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00021570617704331046,
+      "loss": 3.1033,
+      "step": 13132
+    },
+    {
+      "epoch": 0.5529684210526316,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00021567275206566493,
+      "loss": 3.1489,
+      "step": 13133
+    },
+    {
+      "epoch": 0.5530105263157895,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00021563932771342159,
+      "loss": 3.0849,
+      "step": 13134
+    },
+    {
+      "epoch": 0.5530526315789474,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00021560590398718923,
+      "loss": 2.319,
+      "step": 13135
+    },
+    {
+      "epoch": 0.5530947368421053,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00021557248088757702,
+      "loss": 2.8203,
+      "step": 13136
+    },
+    {
+      "epoch": 0.5531368421052631,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002155390584151936,
+      "loss": 3.0241,
+      "step": 13137
+    },
+    {
+      "epoch": 0.5531789473684211,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00021550563657064814,
+      "loss": 3.1614,
+      "step": 13138
+    },
+    {
+      "epoch": 0.553221052631579,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00021547221535454938,
+      "loss": 3.0889,
+      "step": 13139
+    },
+    {
+      "epoch": 0.5532631578947368,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002154387947675063,
+      "loss": 3.2377,
+      "step": 13140
+    },
+    {
+      "epoch": 0.5533052631578947,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002154053748101279,
+      "loss": 2.959,
+      "step": 13141
+    },
+    {
+      "epoch": 0.5533473684210526,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00021537195548302276,
+      "loss": 3.2823,
+      "step": 13142
+    },
+    {
+      "epoch": 0.5533894736842105,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00021533853678680004,
+      "loss": 3.052,
+      "step": 13143
+    },
+    {
+      "epoch": 0.5534315789473684,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00021530511872206836,
+      "loss": 3.2303,
+      "step": 13144
+    },
+    {
+      "epoch": 0.5534736842105263,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002152717012894367,
+      "loss": 3.2891,
+      "step": 13145
+    },
+    {
+      "epoch": 0.5535157894736842,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00021523828448951376,
+      "loss": 2.7642,
+      "step": 13146
+    },
+    {
+      "epoch": 0.5535578947368421,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00021520486832290854,
+      "loss": 3.2216,
+      "step": 13147
+    },
+    {
+      "epoch": 0.5536,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002151714527902297,
+      "loss": 2.9512,
+      "step": 13148
+    },
+    {
+      "epoch": 0.5536421052631579,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00021513803789208603,
+      "loss": 3.574,
+      "step": 13149
+    },
+    {
+      "epoch": 0.5536842105263158,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00021510462362908642,
+      "loss": 3.1369,
+      "step": 13150
+    },
+    {
+      "epoch": 0.5537263157894737,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002150712100018395,
+      "loss": 3.0302,
+      "step": 13151
+    },
+    {
+      "epoch": 0.5537684210526316,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00021503779701095415,
+      "loss": 3.2992,
+      "step": 13152
+    },
+    {
+      "epoch": 0.5538105263157894,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.000215004384657039,
+      "loss": 3.0181,
+      "step": 13153
+    },
+    {
+      "epoch": 0.5538526315789474,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00021497097294070284,
+      "loss": 2.7274,
+      "step": 13154
+    },
+    {
+      "epoch": 0.5538947368421052,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00021493756186255445,
+      "loss": 3.0258,
+      "step": 13155
+    },
+    {
+      "epoch": 0.5539368421052632,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002149041514232024,
+      "loss": 3.3478,
+      "step": 13156
+    },
+    {
+      "epoch": 0.553978947368421,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00021487074162325563,
+      "loss": 3.3587,
+      "step": 13157
+    },
+    {
+      "epoch": 0.554021052631579,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00021483733246332255,
+      "loss": 3.0506,
+      "step": 13158
+    },
+    {
+      "epoch": 0.5540631578947368,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00021480392394401194,
+      "loss": 2.8549,
+      "step": 13159
+    },
+    {
+      "epoch": 0.5541052631578948,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002147705160659326,
+      "loss": 2.9636,
+      "step": 13160
+    },
+    {
+      "epoch": 0.5541473684210526,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00021473710882969295,
+      "loss": 3.3841,
+      "step": 13161
+    },
+    {
+      "epoch": 0.5541894736842106,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002147037022359018,
+      "loss": 3.3301,
+      "step": 13162
+    },
+    {
+      "epoch": 0.5542315789473684,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002146702962851677,
+      "loss": 3.0933,
+      "step": 13163
+    },
+    {
+      "epoch": 0.5542736842105264,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00021463689097809933,
+      "loss": 2.9962,
+      "step": 13164
+    },
+    {
+      "epoch": 0.5543157894736842,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002146034863153052,
+      "loss": 3.4274,
+      "step": 13165
+    },
+    {
+      "epoch": 0.5543578947368422,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00021457008229739394,
+      "loss": 3.4441,
+      "step": 13166
+    },
+    {
+      "epoch": 0.5544,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00021453667892497427,
+      "loss": 2.9326,
+      "step": 13167
+    },
+    {
+      "epoch": 0.5544421052631578,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00021450327619865453,
+      "loss": 3.173,
+      "step": 13168
+    },
+    {
+      "epoch": 0.5544842105263158,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00021446987411904351,
+      "loss": 3.3385,
+      "step": 13169
+    },
+    {
+      "epoch": 0.5545263157894736,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00021443647268674953,
+      "loss": 3.0514,
+      "step": 13170
+    },
+    {
+      "epoch": 0.5545684210526316,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002144030719023813,
+      "loss": 3.0018,
+      "step": 13171
+    },
+    {
+      "epoch": 0.5546105263157894,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00021436967176654718,
+      "loss": 3.1869,
+      "step": 13172
+    },
+    {
+      "epoch": 0.5546526315789474,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002143362722798558,
+      "loss": 3.1425,
+      "step": 13173
+    },
+    {
+      "epoch": 0.5546947368421052,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00021430287344291574,
+      "loss": 2.9003,
+      "step": 13174
+    },
+    {
+      "epoch": 0.5547368421052632,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00021426947525633527,
+      "loss": 2.9766,
+      "step": 13175
+    },
+    {
+      "epoch": 0.554778947368421,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0002142360777207231,
+      "loss": 3.4352,
+      "step": 13176
+    },
+    {
+      "epoch": 0.554821052631579,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00021420268083668746,
+      "loss": 3.0748,
+      "step": 13177
+    },
+    {
+      "epoch": 0.5548631578947368,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.000214169284604837,
+      "loss": 3.0347,
+      "step": 13178
+    },
+    {
+      "epoch": 0.5549052631578948,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00021413588902578,
+      "loss": 3.0456,
+      "step": 13179
+    },
+    {
+      "epoch": 0.5549473684210526,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00021410249410012494,
+      "loss": 3.3114,
+      "step": 13180
+    },
+    {
+      "epoch": 0.5549894736842105,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00021406909982848033,
+      "loss": 3.1245,
+      "step": 13181
+    },
+    {
+      "epoch": 0.5550315789473684,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00021403570621145447,
+      "loss": 3.5848,
+      "step": 13182
+    },
+    {
+      "epoch": 0.5550736842105263,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00021400231324965584,
+      "loss": 3.3441,
+      "step": 13183
+    },
+    {
+      "epoch": 0.5551157894736842,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002139689209436927,
+      "loss": 3.2104,
+      "step": 13184
+    },
+    {
+      "epoch": 0.5551578947368421,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002139355292941736,
+      "loss": 2.8626,
+      "step": 13185
+    },
+    {
+      "epoch": 0.5552,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00021390213830170665,
+      "loss": 2.9954,
+      "step": 13186
+    },
+    {
+      "epoch": 0.5552421052631579,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00021386874796690034,
+      "loss": 3.1758,
+      "step": 13187
+    },
+    {
+      "epoch": 0.5552842105263158,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00021383535829036304,
+      "loss": 2.4312,
+      "step": 13188
+    },
+    {
+      "epoch": 0.5553263157894737,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00021380196927270296,
+      "loss": 2.9523,
+      "step": 13189
+    },
+    {
+      "epoch": 0.5553684210526316,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0002137685809145286,
+      "loss": 2.8671,
+      "step": 13190
+    },
+    {
+      "epoch": 0.5554105263157895,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00021373519321644797,
+      "loss": 2.7751,
+      "step": 13191
+    },
+    {
+      "epoch": 0.5554526315789474,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00021370180617906963,
+      "loss": 3.5387,
+      "step": 13192
+    },
+    {
+      "epoch": 0.5554947368421053,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00021366841980300166,
+      "loss": 3.1991,
+      "step": 13193
+    },
+    {
+      "epoch": 0.5555368421052631,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00021363503408885237,
+      "loss": 2.947,
+      "step": 13194
+    },
+    {
+      "epoch": 0.5555789473684211,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00021360164903723013,
+      "loss": 3.5365,
+      "step": 13195
+    },
+    {
+      "epoch": 0.5556210526315789,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00021356826464874302,
+      "loss": 3.1408,
+      "step": 13196
+    },
+    {
+      "epoch": 0.5556631578947369,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00021353488092399933,
+      "loss": 3.1382,
+      "step": 13197
+    },
+    {
+      "epoch": 0.5557052631578947,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00021350149786360722,
+      "loss": 2.6871,
+      "step": 13198
+    },
+    {
+      "epoch": 0.5557473684210527,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00021346811546817506,
+      "loss": 3.373,
+      "step": 13199
+    },
+    {
+      "epoch": 0.5557894736842105,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00021343473373831078,
+      "loss": 3.1158,
+      "step": 13200
+    },
+    {
+      "epoch": 0.5558315789473685,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00021340135267462269,
+      "loss": 2.678,
+      "step": 13201
+    },
+    {
+      "epoch": 0.5558736842105263,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002133679722777191,
+      "loss": 3.304,
+      "step": 13202
+    },
+    {
+      "epoch": 0.5559157894736843,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00021333459254820785,
+      "loss": 3.0153,
+      "step": 13203
+    },
+    {
+      "epoch": 0.5559578947368421,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00021330121348669734,
+      "loss": 3.1213,
+      "step": 13204
+    },
+    {
+      "epoch": 0.556,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00021326783509379552,
+      "loss": 2.9011,
+      "step": 13205
+    },
+    {
+      "epoch": 0.5560421052631579,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00021323445737011067,
+      "loss": 3.4657,
+      "step": 13206
+    },
+    {
+      "epoch": 0.5560842105263157,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00021320108031625077,
+      "loss": 3.085,
+      "step": 13207
+    },
+    {
+      "epoch": 0.5561263157894737,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00021316770393282392,
+      "loss": 3.4868,
+      "step": 13208
+    },
+    {
+      "epoch": 0.5561684210526315,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00021313432822043832,
+      "loss": 3.2705,
+      "step": 13209
+    },
+    {
+      "epoch": 0.5562105263157895,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00021310095317970184,
+      "loss": 3.2855,
+      "step": 13210
+    },
+    {
+      "epoch": 0.5562526315789473,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00021306757881122277,
+      "loss": 2.8415,
+      "step": 13211
+    },
+    {
+      "epoch": 0.5562947368421053,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002130342051156089,
+      "loss": 2.9086,
+      "step": 13212
+    },
+    {
+      "epoch": 0.5563368421052631,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00021300083209346837,
+      "loss": 3.3894,
+      "step": 13213
+    },
+    {
+      "epoch": 0.5563789473684211,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00021296745974540927,
+      "loss": 2.801,
+      "step": 13214
+    },
+    {
+      "epoch": 0.5564210526315789,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00021293408807203948,
+      "loss": 3.3947,
+      "step": 13215
+    },
+    {
+      "epoch": 0.5564631578947369,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00021290071707396718,
+      "loss": 2.9878,
+      "step": 13216
+    },
+    {
+      "epoch": 0.5565052631578947,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002128673467518001,
+      "loss": 3.0717,
+      "step": 13217
+    },
+    {
+      "epoch": 0.5565473684210527,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00021283397710614643,
+      "loss": 3.0413,
+      "step": 13218
+    },
+    {
+      "epoch": 0.5565894736842105,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00021280060813761397,
+      "loss": 3.104,
+      "step": 13219
+    },
+    {
+      "epoch": 0.5566315789473684,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00021276723984681067,
+      "loss": 3.3639,
+      "step": 13220
+    },
+    {
+      "epoch": 0.5566736842105263,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002127338722343446,
+      "loss": 3.316,
+      "step": 13221
+    },
+    {
+      "epoch": 0.5567157894736842,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00021270050530082358,
+      "loss": 3.1257,
+      "step": 13222
+    },
+    {
+      "epoch": 0.5567578947368421,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00021266713904685553,
+      "loss": 3.0225,
+      "step": 13223
+    },
+    {
+      "epoch": 0.5568,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00021263377347304827,
+      "loss": 3.2181,
+      "step": 13224
+    },
+    {
+      "epoch": 0.5568421052631579,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00021260040858000988,
+      "loss": 3.7013,
+      "step": 13225
+    },
+    {
+      "epoch": 0.5568842105263158,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.000212567044368348,
+      "loss": 3.1312,
+      "step": 13226
+    },
+    {
+      "epoch": 0.5569263157894737,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002125336808386706,
+      "loss": 2.6004,
+      "step": 13227
+    },
+    {
+      "epoch": 0.5569684210526316,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00021250031799158558,
+      "loss": 3.2772,
+      "step": 13228
+    },
+    {
+      "epoch": 0.5570105263157895,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00021246695582770062,
+      "loss": 3.3376,
+      "step": 13229
+    },
+    {
+      "epoch": 0.5570526315789474,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00021243359434762368,
+      "loss": 2.8049,
+      "step": 13230
+    },
+    {
+      "epoch": 0.5570947368421053,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00021240023355196248,
+      "loss": 2.6925,
+      "step": 13231
+    },
+    {
+      "epoch": 0.5571368421052632,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0002123668734413249,
+      "loss": 3.4016,
+      "step": 13232
+    },
+    {
+      "epoch": 0.5571789473684211,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00021233351401631862,
+      "loss": 2.9465,
+      "step": 13233
+    },
+    {
+      "epoch": 0.557221052631579,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00021230015527755143,
+      "loss": 2.9792,
+      "step": 13234
+    },
+    {
+      "epoch": 0.5572631578947368,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00021226679722563123,
+      "loss": 2.9459,
+      "step": 13235
+    },
+    {
+      "epoch": 0.5573052631578947,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00021223343986116552,
+      "loss": 2.9992,
+      "step": 13236
+    },
+    {
+      "epoch": 0.5573473684210526,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00021220008318476225,
+      "loss": 2.8997,
+      "step": 13237
+    },
+    {
+      "epoch": 0.5573894736842105,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00021216672719702904,
+      "loss": 2.4982,
+      "step": 13238
+    },
+    {
+      "epoch": 0.5574315789473684,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002121333718985736,
+      "loss": 3.0885,
+      "step": 13239
+    },
+    {
+      "epoch": 0.5574736842105263,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00021210001729000355,
+      "loss": 3.1118,
+      "step": 13240
+    },
+    {
+      "epoch": 0.5575157894736842,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002120666633719267,
+      "loss": 3.0469,
+      "step": 13241
+    },
+    {
+      "epoch": 0.5575578947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00021203331014495077,
+      "loss": 2.8936,
+      "step": 13242
+    },
+    {
+      "epoch": 0.5576,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00021199995760968321,
+      "loss": 2.9686,
+      "step": 13243
+    },
+    {
+      "epoch": 0.5576421052631579,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00021196660576673187,
+      "loss": 3.6513,
+      "step": 13244
+    },
+    {
+      "epoch": 0.5576842105263158,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00021193325461670422,
+      "loss": 3.2334,
+      "step": 13245
+    },
+    {
+      "epoch": 0.5577263157894737,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00021189990416020795,
+      "loss": 3.0281,
+      "step": 13246
+    },
+    {
+      "epoch": 0.5577684210526316,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00021186655439785063,
+      "loss": 3.295,
+      "step": 13247
+    },
+    {
+      "epoch": 0.5578105263157894,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00021183320533023987,
+      "loss": 3.3307,
+      "step": 13248
+    },
+    {
+      "epoch": 0.5578526315789474,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0002117998569579834,
+      "loss": 3.3982,
+      "step": 13249
+    },
+    {
+      "epoch": 0.5578947368421052,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0002117665092816885,
+      "loss": 2.8229,
+      "step": 13250
+    },
+    {
+      "epoch": 0.5579368421052632,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00021173316230196305,
+      "loss": 3.106,
+      "step": 13251
+    },
+    {
+      "epoch": 0.557978947368421,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00021169981601941427,
+      "loss": 2.9214,
+      "step": 13252
+    },
+    {
+      "epoch": 0.558021052631579,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00021166647043464992,
+      "loss": 3.4026,
+      "step": 13253
+    },
+    {
+      "epoch": 0.5580631578947368,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002116331255482774,
+      "loss": 2.9294,
+      "step": 13254
+    },
+    {
+      "epoch": 0.5581052631578948,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002115997813609042,
+      "loss": 3.4222,
+      "step": 13255
+    },
+    {
+      "epoch": 0.5581473684210526,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00021156643787313795,
+      "loss": 2.8854,
+      "step": 13256
+    },
+    {
+      "epoch": 0.5581894736842106,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.000211533095085586,
+      "loss": 2.9743,
+      "step": 13257
+    },
+    {
+      "epoch": 0.5582315789473684,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00021149975299885592,
+      "loss": 3.3284,
+      "step": 13258
+    },
+    {
+      "epoch": 0.5582736842105264,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.000211466411613555,
+      "loss": 2.5828,
+      "step": 13259
+    },
+    {
+      "epoch": 0.5583157894736842,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00021143307093029085,
+      "loss": 3.3491,
+      "step": 13260
+    },
+    {
+      "epoch": 0.558357894736842,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00021139973094967088,
+      "loss": 3.2432,
+      "step": 13261
+    },
+    {
+      "epoch": 0.5584,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00021136639167230236,
+      "loss": 3.2038,
+      "step": 13262
+    },
+    {
+      "epoch": 0.5584421052631579,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00021133305309879285,
+      "loss": 2.6996,
+      "step": 13263
+    },
+    {
+      "epoch": 0.5584842105263158,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00021129971522974966,
+      "loss": 3.6287,
+      "step": 13264
+    },
+    {
+      "epoch": 0.5585263157894736,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00021126637806578025,
+      "loss": 3.115,
+      "step": 13265
+    },
+    {
+      "epoch": 0.5585684210526316,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00021123304160749185,
+      "loss": 3.419,
+      "step": 13266
+    },
+    {
+      "epoch": 0.5586105263157894,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00021119970585549186,
+      "loss": 3.1656,
+      "step": 13267
+    },
+    {
+      "epoch": 0.5586526315789474,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00021116637081038776,
+      "loss": 2.9235,
+      "step": 13268
+    },
+    {
+      "epoch": 0.5586947368421052,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00021113303647278664,
+      "loss": 3.0173,
+      "step": 13269
+    },
+    {
+      "epoch": 0.5587368421052632,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00021109970284329606,
+      "loss": 3.3627,
+      "step": 13270
+    },
+    {
+      "epoch": 0.558778947368421,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00021106636992252305,
+      "loss": 3.0347,
+      "step": 13271
+    },
+    {
+      "epoch": 0.558821052631579,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00021103303771107508,
+      "loss": 3.2351,
+      "step": 13272
+    },
+    {
+      "epoch": 0.5588631578947368,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00021099970620955933,
+      "loss": 3.0244,
+      "step": 13273
+    },
+    {
+      "epoch": 0.5589052631578947,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00021096637541858314,
+      "loss": 3.1168,
+      "step": 13274
+    },
+    {
+      "epoch": 0.5589473684210526,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002109330453387538,
+      "loss": 2.8197,
+      "step": 13275
+    },
+    {
+      "epoch": 0.5589894736842105,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0002108997159706784,
+      "loss": 3.0742,
+      "step": 13276
+    },
+    {
+      "epoch": 0.5590315789473684,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00021086638731496432,
+      "loss": 3.2655,
+      "step": 13277
+    },
+    {
+      "epoch": 0.5590736842105263,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00021083305937221857,
+      "loss": 2.867,
+      "step": 13278
+    },
+    {
+      "epoch": 0.5591157894736842,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00021079973214304854,
+      "loss": 3.1748,
+      "step": 13279
+    },
+    {
+      "epoch": 0.5591578947368421,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00021076640562806129,
+      "loss": 3.2145,
+      "step": 13280
+    },
+    {
+      "epoch": 0.5592,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.000210733079827864,
+      "loss": 3.2894,
+      "step": 13281
+    },
+    {
+      "epoch": 0.5592421052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002106997547430639,
+      "loss": 3.2616,
+      "step": 13282
+    },
+    {
+      "epoch": 0.5592842105263158,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00021066643037426808,
+      "loss": 3.1266,
+      "step": 13283
+    },
+    {
+      "epoch": 0.5593263157894737,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00021063310672208375,
+      "loss": 3.0293,
+      "step": 13284
+    },
+    {
+      "epoch": 0.5593684210526316,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002105997837871179,
+      "loss": 3.1444,
+      "step": 13285
+    },
+    {
+      "epoch": 0.5594105263157895,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00021056646156997774,
+      "loss": 3.3858,
+      "step": 13286
+    },
+    {
+      "epoch": 0.5594526315789474,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00021053314007127025,
+      "loss": 3.1317,
+      "step": 13287
+    },
+    {
+      "epoch": 0.5594947368421053,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00021049981929160256,
+      "loss": 2.911,
+      "step": 13288
+    },
+    {
+      "epoch": 0.5595368421052631,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002104664992315818,
+      "loss": 3.145,
+      "step": 13289
+    },
+    {
+      "epoch": 0.5595789473684211,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00021043317989181495,
+      "loss": 3.3867,
+      "step": 13290
+    },
+    {
+      "epoch": 0.5596210526315789,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00021039986127290917,
+      "loss": 2.7112,
+      "step": 13291
+    },
+    {
+      "epoch": 0.5596631578947369,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002103665433754713,
+      "loss": 2.9776,
+      "step": 13292
+    },
+    {
+      "epoch": 0.5597052631578947,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002103332262001085,
+      "loss": 3.252,
+      "step": 13293
+    },
+    {
+      "epoch": 0.5597473684210527,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00021029990974742762,
+      "loss": 3.6021,
+      "step": 13294
+    },
+    {
+      "epoch": 0.5597894736842105,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00021026659401803576,
+      "loss": 2.9986,
+      "step": 13295
+    },
+    {
+      "epoch": 0.5598315789473685,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00021023327901253998,
+      "loss": 3.4741,
+      "step": 13296
+    },
+    {
+      "epoch": 0.5598736842105263,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.000210199964731547,
+      "loss": 3.3788,
+      "step": 13297
+    },
+    {
+      "epoch": 0.5599157894736843,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00021016665117566398,
+      "loss": 3.2551,
+      "step": 13298
+    },
+    {
+      "epoch": 0.5599578947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00021013333834549769,
+      "loss": 3.3287,
+      "step": 13299
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00021010002624165526,
+      "loss": 3.3162,
+      "step": 13300
+    },
+    {
+      "epoch": 0.5600421052631579,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00021006671486474337,
+      "loss": 3.3367,
+      "step": 13301
+    },
+    {
+      "epoch": 0.5600842105263157,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00021003340421536898,
+      "loss": 2.652,
+      "step": 13302
+    },
+    {
+      "epoch": 0.5601263157894737,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00021000009429413916,
+      "loss": 3.1295,
+      "step": 13303
+    },
+    {
+      "epoch": 0.5601684210526315,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002099667851016605,
+      "loss": 3.0544,
+      "step": 13304
+    },
+    {
+      "epoch": 0.5602105263157895,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00020993347663854004,
+      "loss": 3.033,
+      "step": 13305
+    },
+    {
+      "epoch": 0.5602526315789473,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002099001689053845,
+      "loss": 3.1356,
+      "step": 13306
+    },
+    {
+      "epoch": 0.5602947368421053,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00020986686190280074,
+      "loss": 3.076,
+      "step": 13307
+    },
+    {
+      "epoch": 0.5603368421052631,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00020983355563139572,
+      "loss": 2.8803,
+      "step": 13308
+    },
+    {
+      "epoch": 0.5603789473684211,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00020980025009177607,
+      "loss": 3.1783,
+      "step": 13309
+    },
+    {
+      "epoch": 0.5604210526315789,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00020976694528454867,
+      "loss": 3.2449,
+      "step": 13310
+    },
+    {
+      "epoch": 0.5604631578947369,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00020973364121032017,
+      "loss": 3.416,
+      "step": 13311
+    },
+    {
+      "epoch": 0.5605052631578947,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00020970033786969753,
+      "loss": 3.5466,
+      "step": 13312
+    },
+    {
+      "epoch": 0.5605473684210527,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002096670352632873,
+      "loss": 3.234,
+      "step": 13313
+    },
+    {
+      "epoch": 0.5605894736842105,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00020963373339169623,
+      "loss": 2.6963,
+      "step": 13314
+    },
+    {
+      "epoch": 0.5606315789473684,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00020960043225553123,
+      "loss": 3.085,
+      "step": 13315
+    },
+    {
+      "epoch": 0.5606736842105263,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00020956713185539877,
+      "loss": 3.2071,
+      "step": 13316
+    },
+    {
+      "epoch": 0.5607157894736842,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00020953383219190583,
+      "loss": 3.2566,
+      "step": 13317
+    },
+    {
+      "epoch": 0.5607578947368421,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0002095005332656588,
+      "loss": 3.4362,
+      "step": 13318
+    },
+    {
+      "epoch": 0.5608,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00020946723507726456,
+      "loss": 3.3992,
+      "step": 13319
+    },
+    {
+      "epoch": 0.5608421052631579,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00020943393762732956,
+      "loss": 2.8092,
+      "step": 13320
+    },
+    {
+      "epoch": 0.5608842105263158,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00020940064091646055,
+      "loss": 2.942,
+      "step": 13321
+    },
+    {
+      "epoch": 0.5609263157894737,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00020936734494526424,
+      "loss": 2.9566,
+      "step": 13322
+    },
+    {
+      "epoch": 0.5609684210526316,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00020933404971434715,
+      "loss": 2.6166,
+      "step": 13323
+    },
+    {
+      "epoch": 0.5610105263157895,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002093007552243159,
+      "loss": 3.0619,
+      "step": 13324
+    },
+    {
+      "epoch": 0.5610526315789474,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.000209267461475777,
+      "loss": 3.3991,
+      "step": 13325
+    },
+    {
+      "epoch": 0.5610947368421053,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00020923416846933723,
+      "loss": 2.6836,
+      "step": 13326
+    },
+    {
+      "epoch": 0.5611368421052632,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00020920087620560288,
+      "loss": 3.4332,
+      "step": 13327
+    },
+    {
+      "epoch": 0.561178947368421,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00020916758468518065,
+      "loss": 2.9093,
+      "step": 13328
+    },
+    {
+      "epoch": 0.561221052631579,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00020913429390867713,
+      "loss": 2.8755,
+      "step": 13329
+    },
+    {
+      "epoch": 0.5612631578947368,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00020910100387669866,
+      "loss": 3.0418,
+      "step": 13330
+    },
+    {
+      "epoch": 0.5613052631578948,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002090677145898519,
+      "loss": 3.116,
+      "step": 13331
+    },
+    {
+      "epoch": 0.5613473684210526,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00020903442604874325,
+      "loss": 3.3644,
+      "step": 13332
+    },
+    {
+      "epoch": 0.5613894736842105,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002090011382539793,
+      "loss": 3.4315,
+      "step": 13333
+    },
+    {
+      "epoch": 0.5614315789473684,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00020896785120616638,
+      "loss": 2.8542,
+      "step": 13334
+    },
+    {
+      "epoch": 0.5614736842105263,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00020893456490591098,
+      "loss": 3.2463,
+      "step": 13335
+    },
+    {
+      "epoch": 0.5615157894736842,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00020890127935381964,
+      "loss": 3.2973,
+      "step": 13336
+    },
+    {
+      "epoch": 0.5615578947368421,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00020886799455049863,
+      "loss": 2.9064,
+      "step": 13337
+    },
+    {
+      "epoch": 0.5616,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00020883471049655444,
+      "loss": 3.1614,
+      "step": 13338
+    },
+    {
+      "epoch": 0.5616421052631579,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00020880142719259352,
+      "loss": 3.0627,
+      "step": 13339
+    },
+    {
+      "epoch": 0.5616842105263158,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00020876814463922214,
+      "loss": 2.9105,
+      "step": 13340
+    },
+    {
+      "epoch": 0.5617263157894736,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00020873486283704668,
+      "loss": 3.061,
+      "step": 13341
+    },
+    {
+      "epoch": 0.5617684210526316,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00020870158178667354,
+      "loss": 3.2333,
+      "step": 13342
+    },
+    {
+      "epoch": 0.5618105263157894,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00020866830148870915,
+      "loss": 3.0262,
+      "step": 13343
+    },
+    {
+      "epoch": 0.5618526315789474,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00020863502194375967,
+      "loss": 2.8341,
+      "step": 13344
+    },
+    {
+      "epoch": 0.5618947368421052,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00020860174315243154,
+      "loss": 3.0954,
+      "step": 13345
+    },
+    {
+      "epoch": 0.5619368421052632,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00020856846511533095,
+      "loss": 2.6248,
+      "step": 13346
+    },
+    {
+      "epoch": 0.561978947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00020853518783306427,
+      "loss": 3.5041,
+      "step": 13347
+    },
+    {
+      "epoch": 0.562021052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0002085019113062377,
+      "loss": 2.7965,
+      "step": 13348
+    },
+    {
+      "epoch": 0.5620631578947368,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00020846863553545755,
+      "loss": 2.8155,
+      "step": 13349
+    },
+    {
+      "epoch": 0.5621052631578948,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002084353605213302,
+      "loss": 3.212,
+      "step": 13350
+    },
+    {
+      "epoch": 0.5621473684210526,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002084020862644616,
+      "loss": 3.1989,
+      "step": 13351
+    },
+    {
+      "epoch": 0.5621894736842106,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00020836881276545822,
+      "loss": 3.0888,
+      "step": 13352
+    },
+    {
+      "epoch": 0.5622315789473684,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00020833554002492607,
+      "loss": 3.1422,
+      "step": 13353
+    },
+    {
+      "epoch": 0.5622736842105264,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00020830226804347142,
+      "loss": 3.2071,
+      "step": 13354
+    },
+    {
+      "epoch": 0.5623157894736842,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00020826899682170055,
+      "loss": 3.0148,
+      "step": 13355
+    },
+    {
+      "epoch": 0.5623578947368421,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00020823572636021942,
+      "loss": 3.335,
+      "step": 13356
+    },
+    {
+      "epoch": 0.5624,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00020820245665963437,
+      "loss": 3.038,
+      "step": 13357
+    },
+    {
+      "epoch": 0.5624421052631579,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002081691877205514,
+      "loss": 3.4129,
+      "step": 13358
+    },
+    {
+      "epoch": 0.5624842105263158,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00020813591954357675,
+      "loss": 3.0938,
+      "step": 13359
+    },
+    {
+      "epoch": 0.5625263157894737,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00020810265212931633,
+      "loss": 3.2146,
+      "step": 13360
+    },
+    {
+      "epoch": 0.5625684210526316,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002080693854783764,
+      "loss": 3.0276,
+      "step": 13361
+    },
+    {
+      "epoch": 0.5626105263157894,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00020803611959136308,
+      "loss": 2.8753,
+      "step": 13362
+    },
+    {
+      "epoch": 0.5626526315789474,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00020800285446888228,
+      "loss": 3.0602,
+      "step": 13363
+    },
+    {
+      "epoch": 0.5626947368421052,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00020796959011154019,
+      "loss": 3.2112,
+      "step": 13364
+    },
+    {
+      "epoch": 0.5627368421052632,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00020793632651994269,
+      "loss": 3.4673,
+      "step": 13365
+    },
+    {
+      "epoch": 0.562778947368421,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00020790306369469603,
+      "loss": 3.3115,
+      "step": 13366
+    },
+    {
+      "epoch": 0.562821052631579,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00020786980163640595,
+      "loss": 2.5923,
+      "step": 13367
+    },
+    {
+      "epoch": 0.5628631578947368,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002078365403456786,
+      "loss": 3.4835,
+      "step": 13368
+    },
+    {
+      "epoch": 0.5629052631578947,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00020780327982312008,
+      "loss": 3.0649,
+      "step": 13369
+    },
+    {
+      "epoch": 0.5629473684210526,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00020777002006933608,
+      "loss": 3.021,
+      "step": 13370
+    },
+    {
+      "epoch": 0.5629894736842105,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00020773676108493285,
+      "loss": 2.9052,
+      "step": 13371
+    },
+    {
+      "epoch": 0.5630315789473684,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00020770350287051603,
+      "loss": 3.2196,
+      "step": 13372
+    },
+    {
+      "epoch": 0.5630736842105263,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00020767024542669174,
+      "loss": 3.1854,
+      "step": 13373
+    },
+    {
+      "epoch": 0.5631157894736842,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00020763698875406582,
+      "loss": 2.9665,
+      "step": 13374
+    },
+    {
+      "epoch": 0.5631578947368421,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00020760373285324415,
+      "loss": 2.7489,
+      "step": 13375
+    },
+    {
+      "epoch": 0.5632,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00020757047772483278,
+      "loss": 3.5204,
+      "step": 13376
+    },
+    {
+      "epoch": 0.5632421052631579,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00020753722336943736,
+      "loss": 2.9893,
+      "step": 13377
+    },
+    {
+      "epoch": 0.5632842105263158,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00020750396978766396,
+      "loss": 2.6112,
+      "step": 13378
+    },
+    {
+      "epoch": 0.5633263157894737,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00020747071698011817,
+      "loss": 3.3848,
+      "step": 13379
+    },
+    {
+      "epoch": 0.5633684210526316,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00020743746494740597,
+      "loss": 2.846,
+      "step": 13380
+    },
+    {
+      "epoch": 0.5634105263157895,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00020740421369013312,
+      "loss": 3.4135,
+      "step": 13381
+    },
+    {
+      "epoch": 0.5634526315789473,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002073709632089055,
+      "loss": 3.1989,
+      "step": 13382
+    },
+    {
+      "epoch": 0.5634947368421053,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00020733771350432885,
+      "loss": 3.1562,
+      "step": 13383
+    },
+    {
+      "epoch": 0.5635368421052631,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0002073044645770089,
+      "loss": 2.7636,
+      "step": 13384
+    },
+    {
+      "epoch": 0.5635789473684211,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002072712164275515,
+      "loss": 2.8281,
+      "step": 13385
+    },
+    {
+      "epoch": 0.5636210526315789,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00020723796905656224,
+      "loss": 2.9482,
+      "step": 13386
+    },
+    {
+      "epoch": 0.5636631578947369,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00020720472246464708,
+      "loss": 3.2882,
+      "step": 13387
+    },
+    {
+      "epoch": 0.5637052631578947,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002071714766524115,
+      "loss": 3.0299,
+      "step": 13388
+    },
+    {
+      "epoch": 0.5637473684210527,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00020713823162046122,
+      "loss": 2.6648,
+      "step": 13389
+    },
+    {
+      "epoch": 0.5637894736842105,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0002071049873694021,
+      "loss": 3.2732,
+      "step": 13390
+    },
+    {
+      "epoch": 0.5638315789473685,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00020707174389983965,
+      "loss": 2.9669,
+      "step": 13391
+    },
+    {
+      "epoch": 0.5638736842105263,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002070385012123797,
+      "loss": 3.3212,
+      "step": 13392
+    },
+    {
+      "epoch": 0.5639157894736843,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00020700525930762766,
+      "loss": 3.5706,
+      "step": 13393
+    },
+    {
+      "epoch": 0.5639578947368421,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002069720181861894,
+      "loss": 3.2916,
+      "step": 13394
+    },
+    {
+      "epoch": 0.564,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00020693877784867028,
+      "loss": 3.0355,
+      "step": 13395
+    },
+    {
+      "epoch": 0.5640421052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00020690553829567606,
+      "loss": 3.3719,
+      "step": 13396
+    },
+    {
+      "epoch": 0.5640842105263157,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00020687229952781236,
+      "loss": 3.3088,
+      "step": 13397
+    },
+    {
+      "epoch": 0.5641263157894737,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00020683906154568465,
+      "loss": 3.159,
+      "step": 13398
+    },
+    {
+      "epoch": 0.5641684210526315,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00020680582434989857,
+      "loss": 3.0954,
+      "step": 13399
+    },
+    {
+      "epoch": 0.5642105263157895,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00020677258794105956,
+      "loss": 3.2769,
+      "step": 13400
+    },
+    {
+      "epoch": 0.5642526315789473,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002067393523197733,
+      "loss": 3.5096,
+      "step": 13401
+    },
+    {
+      "epoch": 0.5642947368421053,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00020670611748664508,
+      "loss": 3.0812,
+      "step": 13402
+    },
+    {
+      "epoch": 0.5643368421052631,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00020667288344228057,
+      "loss": 2.9125,
+      "step": 13403
+    },
+    {
+      "epoch": 0.5643789473684211,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0002066396501872853,
+      "loss": 3.1512,
+      "step": 13404
+    },
+    {
+      "epoch": 0.5644210526315789,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00020660641772226457,
+      "loss": 2.9029,
+      "step": 13405
+    },
+    {
+      "epoch": 0.5644631578947369,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00020657318604782397,
+      "loss": 3.4337,
+      "step": 13406
+    },
+    {
+      "epoch": 0.5645052631578947,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00020653995516456883,
+      "loss": 2.7975,
+      "step": 13407
+    },
+    {
+      "epoch": 0.5645473684210527,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00020650672507310463,
+      "loss": 2.9489,
+      "step": 13408
+    },
+    {
+      "epoch": 0.5645894736842105,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00020647349577403696,
+      "loss": 2.9926,
+      "step": 13409
+    },
+    {
+      "epoch": 0.5646315789473684,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00020644026726797092,
+      "loss": 2.9842,
+      "step": 13410
+    },
+    {
+      "epoch": 0.5646736842105263,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00020640703955551214,
+      "loss": 3.0026,
+      "step": 13411
+    },
+    {
+      "epoch": 0.5647157894736842,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00020637381263726573,
+      "loss": 3.5585,
+      "step": 13412
+    },
+    {
+      "epoch": 0.5647578947368421,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00020634058651383737,
+      "loss": 3.1831,
+      "step": 13413
+    },
+    {
+      "epoch": 0.5648,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00020630736118583206,
+      "loss": 3.178,
+      "step": 13414
+    },
+    {
+      "epoch": 0.5648421052631579,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00020627413665385533,
+      "loss": 3.0248,
+      "step": 13415
+    },
+    {
+      "epoch": 0.5648842105263158,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0002062409129185125,
+      "loss": 3.0945,
+      "step": 13416
+    },
+    {
+      "epoch": 0.5649263157894737,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00020620768998040874,
+      "loss": 3.2937,
+      "step": 13417
+    },
+    {
+      "epoch": 0.5649684210526316,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00020617446784014956,
+      "loss": 3.4015,
+      "step": 13418
+    },
+    {
+      "epoch": 0.5650105263157895,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00020614124649833996,
+      "loss": 2.9013,
+      "step": 13419
+    },
+    {
+      "epoch": 0.5650526315789474,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00020610802595558543,
+      "loss": 2.8324,
+      "step": 13420
+    },
+    {
+      "epoch": 0.5650947368421053,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00020607480621249097,
+      "loss": 3.1197,
+      "step": 13421
+    },
+    {
+      "epoch": 0.5651368421052632,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00020604158726966195,
+      "loss": 3.2631,
+      "step": 13422
+    },
+    {
+      "epoch": 0.565178947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00020600836912770362,
+      "loss": 2.8708,
+      "step": 13423
+    },
+    {
+      "epoch": 0.565221052631579,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00020597515178722106,
+      "loss": 2.5407,
+      "step": 13424
+    },
+    {
+      "epoch": 0.5652631578947368,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0002059419352488196,
+      "loss": 2.6628,
+      "step": 13425
+    },
+    {
+      "epoch": 0.5653052631578948,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00020590871951310422,
+      "loss": 3.3524,
+      "step": 13426
+    },
+    {
+      "epoch": 0.5653473684210526,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00020587550458068027,
+      "loss": 3.0766,
+      "step": 13427
+    },
+    {
+      "epoch": 0.5653894736842106,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00020584229045215267,
+      "loss": 3.2858,
+      "step": 13428
+    },
+    {
+      "epoch": 0.5654315789473684,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00020580907712812666,
+      "loss": 3.3141,
+      "step": 13429
+    },
+    {
+      "epoch": 0.5654736842105264,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00020577586460920745,
+      "loss": 3.1747,
+      "step": 13430
+    },
+    {
+      "epoch": 0.5655157894736842,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00020574265289599993,
+      "loss": 3.3121,
+      "step": 13431
+    },
+    {
+      "epoch": 0.5655578947368421,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00020570944198910928,
+      "loss": 2.8888,
+      "step": 13432
+    },
+    {
+      "epoch": 0.5656,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00020567623188914055,
+      "loss": 3.4697,
+      "step": 13433
+    },
+    {
+      "epoch": 0.565642105263158,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002056430225966989,
+      "loss": 3.2755,
+      "step": 13434
+    },
+    {
+      "epoch": 0.5656842105263158,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00020560981411238915,
+      "loss": 2.9606,
+      "step": 13435
+    },
+    {
+      "epoch": 0.5657263157894736,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0002055766064368164,
+      "loss": 2.9596,
+      "step": 13436
+    },
+    {
+      "epoch": 0.5657684210526316,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00020554339957058583,
+      "loss": 2.8059,
+      "step": 13437
+    },
+    {
+      "epoch": 0.5658105263157894,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00020551019351430216,
+      "loss": 2.9254,
+      "step": 13438
+    },
+    {
+      "epoch": 0.5658526315789474,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00020547698826857056,
+      "loss": 3.0345,
+      "step": 13439
+    },
+    {
+      "epoch": 0.5658947368421052,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00020544378383399592,
+      "loss": 3.2259,
+      "step": 13440
+    },
+    {
+      "epoch": 0.5659368421052632,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00020541058021118316,
+      "loss": 3.4507,
+      "step": 13441
+    },
+    {
+      "epoch": 0.565978947368421,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002053773774007372,
+      "loss": 3.0356,
+      "step": 13442
+    },
+    {
+      "epoch": 0.566021052631579,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.000205344175403263,
+      "loss": 3.0072,
+      "step": 13443
+    },
+    {
+      "epoch": 0.5660631578947368,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00020531097421936558,
+      "loss": 3.2416,
+      "step": 13444
+    },
+    {
+      "epoch": 0.5661052631578948,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00020527777384964958,
+      "loss": 3.0714,
+      "step": 13445
+    },
+    {
+      "epoch": 0.5661473684210526,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0002052445742947201,
+      "loss": 2.8334,
+      "step": 13446
+    },
+    {
+      "epoch": 0.5661894736842106,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00020521137555518177,
+      "loss": 3.6101,
+      "step": 13447
+    },
+    {
+      "epoch": 0.5662315789473684,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002051781776316396,
+      "loss": 2.8802,
+      "step": 13448
+    },
+    {
+      "epoch": 0.5662736842105263,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00020514498052469834,
+      "loss": 3.4606,
+      "step": 13449
+    },
+    {
+      "epoch": 0.5663157894736842,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0002051117842349628,
+      "loss": 2.4606,
+      "step": 13450
+    },
+    {
+      "epoch": 0.5663578947368421,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00020507858876303797,
+      "loss": 2.9818,
+      "step": 13451
+    },
+    {
+      "epoch": 0.5664,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00020504539410952832,
+      "loss": 3.2285,
+      "step": 13452
+    },
+    {
+      "epoch": 0.5664421052631579,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00020501220027503893,
+      "loss": 3.2174,
+      "step": 13453
+    },
+    {
+      "epoch": 0.5664842105263158,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00020497900726017425,
+      "loss": 3.2938,
+      "step": 13454
+    },
+    {
+      "epoch": 0.5665263157894737,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00020494581506553915,
+      "loss": 2.8193,
+      "step": 13455
+    },
+    {
+      "epoch": 0.5665684210526316,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00020491262369173847,
+      "loss": 2.9598,
+      "step": 13456
+    },
+    {
+      "epoch": 0.5666105263157895,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00020487943313937673,
+      "loss": 2.7091,
+      "step": 13457
+    },
+    {
+      "epoch": 0.5666526315789474,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002048462434090587,
+      "loss": 2.9085,
+      "step": 13458
+    },
+    {
+      "epoch": 0.5666947368421053,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00020481305450138908,
+      "loss": 3.0619,
+      "step": 13459
+    },
+    {
+      "epoch": 0.5667368421052632,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002047798664169726,
+      "loss": 3.0536,
+      "step": 13460
+    },
+    {
+      "epoch": 0.566778947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002047466791564137,
+      "loss": 2.9131,
+      "step": 13461
+    },
+    {
+      "epoch": 0.5668210526315789,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00020471349272031713,
+      "loss": 2.8271,
+      "step": 13462
+    },
+    {
+      "epoch": 0.5668631578947368,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00020468030710928765,
+      "loss": 2.9499,
+      "step": 13463
+    },
+    {
+      "epoch": 0.5669052631578947,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002046471223239296,
+      "loss": 3.1213,
+      "step": 13464
+    },
+    {
+      "epoch": 0.5669473684210526,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00020461393836484777,
+      "loss": 3.49,
+      "step": 13465
+    },
+    {
+      "epoch": 0.5669894736842105,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00020458075523264655,
+      "loss": 3.5118,
+      "step": 13466
+    },
+    {
+      "epoch": 0.5670315789473684,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00020454757292793076,
+      "loss": 3.5503,
+      "step": 13467
+    },
+    {
+      "epoch": 0.5670736842105263,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00020451439145130465,
+      "loss": 3.2772,
+      "step": 13468
+    },
+    {
+      "epoch": 0.5671157894736842,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002044812108033729,
+      "loss": 2.7237,
+      "step": 13469
+    },
+    {
+      "epoch": 0.5671578947368421,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00020444803098474011,
+      "loss": 2.7521,
+      "step": 13470
+    },
+    {
+      "epoch": 0.5672,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00020441485199601057,
+      "loss": 2.8323,
+      "step": 13471
+    },
+    {
+      "epoch": 0.5672421052631579,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00020438167383778896,
+      "loss": 3.7848,
+      "step": 13472
+    },
+    {
+      "epoch": 0.5672842105263158,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00020434849651067958,
+      "loss": 3.0707,
+      "step": 13473
+    },
+    {
+      "epoch": 0.5673263157894737,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00020431532001528697,
+      "loss": 2.8751,
+      "step": 13474
+    },
+    {
+      "epoch": 0.5673684210526316,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0002042821443522155,
+      "loss": 3.2152,
+      "step": 13475
+    },
+    {
+      "epoch": 0.5674105263157895,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00020424896952206967,
+      "loss": 2.8275,
+      "step": 13476
+    },
+    {
+      "epoch": 0.5674526315789473,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00020421579552545397,
+      "loss": 3.0393,
+      "step": 13477
+    },
+    {
+      "epoch": 0.5674947368421053,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00020418262236297256,
+      "loss": 3.7078,
+      "step": 13478
+    },
+    {
+      "epoch": 0.5675368421052631,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00020414945003523004,
+      "loss": 3.1412,
+      "step": 13479
+    },
+    {
+      "epoch": 0.5675789473684211,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00020411627854283057,
+      "loss": 3.2693,
+      "step": 13480
+    },
+    {
+      "epoch": 0.5676210526315789,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00020408310788637863,
+      "loss": 3.086,
+      "step": 13481
+    },
+    {
+      "epoch": 0.5676631578947369,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00020404993806647848,
+      "loss": 2.898,
+      "step": 13482
+    },
+    {
+      "epoch": 0.5677052631578947,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002040167690837345,
+      "loss": 3.0452,
+      "step": 13483
+    },
+    {
+      "epoch": 0.5677473684210527,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.000203983600938751,
+      "loss": 3.1846,
+      "step": 13484
+    },
+    {
+      "epoch": 0.5677894736842105,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00020395043363213215,
+      "loss": 3.0197,
+      "step": 13485
+    },
+    {
+      "epoch": 0.5678315789473685,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002039172671644824,
+      "loss": 3.1228,
+      "step": 13486
+    },
+    {
+      "epoch": 0.5678736842105263,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00020388410153640576,
+      "loss": 3.2838,
+      "step": 13487
+    },
+    {
+      "epoch": 0.5679157894736843,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00020385093674850675,
+      "loss": 3.1748,
+      "step": 13488
+    },
+    {
+      "epoch": 0.5679578947368421,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00020381777280138933,
+      "loss": 3.039,
+      "step": 13489
+    },
+    {
+      "epoch": 0.568,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00020378460969565782,
+      "loss": 2.3598,
+      "step": 13490
+    },
+    {
+      "epoch": 0.5680421052631579,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00020375144743191648,
+      "loss": 3.4648,
+      "step": 13491
+    },
+    {
+      "epoch": 0.5680842105263157,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00020371828601076938,
+      "loss": 3.5287,
+      "step": 13492
+    },
+    {
+      "epoch": 0.5681263157894737,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00020368512543282087,
+      "loss": 3.3216,
+      "step": 13493
+    },
+    {
+      "epoch": 0.5681684210526315,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00020365196569867479,
+      "loss": 3.1009,
+      "step": 13494
+    },
+    {
+      "epoch": 0.5682105263157895,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00020361880680893558,
+      "loss": 3.4478,
+      "step": 13495
+    },
+    {
+      "epoch": 0.5682526315789473,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002035856487642071,
+      "loss": 3.0307,
+      "step": 13496
+    },
+    {
+      "epoch": 0.5682947368421053,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00020355249156509355,
+      "loss": 2.9178,
+      "step": 13497
+    },
+    {
+      "epoch": 0.5683368421052631,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00020351933521219913,
+      "loss": 3.0776,
+      "step": 13498
+    },
+    {
+      "epoch": 0.5683789473684211,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00020348617970612776,
+      "loss": 3.2554,
+      "step": 13499
+    },
+    {
+      "epoch": 0.5684210526315789,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00020345302504748357,
+      "loss": 2.7334,
+      "step": 13500
+    },
+    {
+      "epoch": 0.5684631578947369,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002034198712368705,
+      "loss": 3.3914,
+      "step": 13501
+    },
+    {
+      "epoch": 0.5685052631578947,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00020338671827489275,
+      "loss": 2.8601,
+      "step": 13502
+    },
+    {
+      "epoch": 0.5685473684210526,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00020335356616215414,
+      "loss": 3.218,
+      "step": 13503
+    },
+    {
+      "epoch": 0.5685894736842105,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00020332041489925872,
+      "loss": 3.3275,
+      "step": 13504
+    },
+    {
+      "epoch": 0.5686315789473684,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00020328726448681062,
+      "loss": 3.2204,
+      "step": 13505
+    },
+    {
+      "epoch": 0.5686736842105263,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00020325411492541358,
+      "loss": 2.5485,
+      "step": 13506
+    },
+    {
+      "epoch": 0.5687157894736842,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00020322096621567168,
+      "loss": 2.8415,
+      "step": 13507
+    },
+    {
+      "epoch": 0.5687578947368421,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00020318781835818877,
+      "loss": 2.9219,
+      "step": 13508
+    },
+    {
+      "epoch": 0.5688,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002031546713535688,
+      "loss": 3.1259,
+      "step": 13509
+    },
+    {
+      "epoch": 0.5688421052631579,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00020312152520241577,
+      "loss": 2.7639,
+      "step": 13510
+    },
+    {
+      "epoch": 0.5688842105263158,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00020308837990533338,
+      "loss": 3.5421,
+      "step": 13511
+    },
+    {
+      "epoch": 0.5689263157894737,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00020305523546292571,
+      "loss": 2.8726,
+      "step": 13512
+    },
+    {
+      "epoch": 0.5689684210526316,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00020302209187579638,
+      "loss": 2.8697,
+      "step": 13513
+    },
+    {
+      "epoch": 0.5690105263157895,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002029889491445494,
+      "loss": 3.183,
+      "step": 13514
+    },
+    {
+      "epoch": 0.5690526315789474,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00020295580726978848,
+      "loss": 3.3963,
+      "step": 13515
+    },
+    {
+      "epoch": 0.5690947368421052,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00020292266625211746,
+      "loss": 2.8119,
+      "step": 13516
+    },
+    {
+      "epoch": 0.5691368421052632,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002028895260921402,
+      "loss": 2.8115,
+      "step": 13517
+    },
+    {
+      "epoch": 0.569178947368421,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00020285638679046038,
+      "loss": 3.203,
+      "step": 13518
+    },
+    {
+      "epoch": 0.569221052631579,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00020282324834768186,
+      "loss": 3.0444,
+      "step": 13519
+    },
+    {
+      "epoch": 0.5692631578947368,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00020279011076440825,
+      "loss": 3.3087,
+      "step": 13520
+    },
+    {
+      "epoch": 0.5693052631578948,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0002027569740412435,
+      "loss": 3.3042,
+      "step": 13521
+    },
+    {
+      "epoch": 0.5693473684210526,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.000202723838178791,
+      "loss": 2.9935,
+      "step": 13522
+    },
+    {
+      "epoch": 0.5693894736842106,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00020269070317765464,
+      "loss": 3.1422,
+      "step": 13523
+    },
+    {
+      "epoch": 0.5694315789473684,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0002026575690384382,
+      "loss": 3.0019,
+      "step": 13524
+    },
+    {
+      "epoch": 0.5694736842105264,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0002026244357617451,
+      "loss": 2.9676,
+      "step": 13525
+    },
+    {
+      "epoch": 0.5695157894736842,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00020259130334817922,
+      "loss": 3.4595,
+      "step": 13526
+    },
+    {
+      "epoch": 0.5695578947368422,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.000202558171798344,
+      "loss": 3.3711,
+      "step": 13527
+    },
+    {
+      "epoch": 0.5696,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00020252504111284328,
+      "loss": 3.2391,
+      "step": 13528
+    },
+    {
+      "epoch": 0.569642105263158,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0002024919112922804,
+      "loss": 3.2737,
+      "step": 13529
+    },
+    {
+      "epoch": 0.5696842105263158,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00020245878233725908,
+      "loss": 3.2462,
+      "step": 13530
+    },
+    {
+      "epoch": 0.5697263157894736,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00020242565424838298,
+      "loss": 3.1948,
+      "step": 13531
+    },
+    {
+      "epoch": 0.5697684210526316,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00020239252702625548,
+      "loss": 2.7891,
+      "step": 13532
+    },
+    {
+      "epoch": 0.5698105263157894,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00020235940067148025,
+      "loss": 3.0025,
+      "step": 13533
+    },
+    {
+      "epoch": 0.5698526315789474,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00020232627518466072,
+      "loss": 2.8918,
+      "step": 13534
+    },
+    {
+      "epoch": 0.5698947368421052,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00020229315056640053,
+      "loss": 2.9107,
+      "step": 13535
+    },
+    {
+      "epoch": 0.5699368421052632,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00020226002681730297,
+      "loss": 3.2809,
+      "step": 13536
+    },
+    {
+      "epoch": 0.569978947368421,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00020222690393797163,
+      "loss": 2.9792,
+      "step": 13537
+    },
+    {
+      "epoch": 0.570021052631579,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0002021937819290101,
+      "loss": 3.0968,
+      "step": 13538
+    },
+    {
+      "epoch": 0.5700631578947368,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00020216066079102157,
+      "loss": 3.0045,
+      "step": 13539
+    },
+    {
+      "epoch": 0.5701052631578948,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00020212754052460963,
+      "loss": 3.0371,
+      "step": 13540
+    },
+    {
+      "epoch": 0.5701473684210526,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00020209442113037763,
+      "loss": 2.9851,
+      "step": 13541
+    },
+    {
+      "epoch": 0.5701894736842106,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002020613026089291,
+      "loss": 2.9925,
+      "step": 13542
+    },
+    {
+      "epoch": 0.5702315789473684,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00020202818496086716,
+      "loss": 3.1685,
+      "step": 13543
+    },
+    {
+      "epoch": 0.5702736842105263,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00020199506818679536,
+      "loss": 3.0505,
+      "step": 13544
+    },
+    {
+      "epoch": 0.5703157894736842,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00020196195228731713,
+      "loss": 2.7883,
+      "step": 13545
+    },
+    {
+      "epoch": 0.5703578947368421,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00020192883726303558,
+      "loss": 2.734,
+      "step": 13546
+    },
+    {
+      "epoch": 0.5704,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00020189572311455421,
+      "loss": 3.1089,
+      "step": 13547
+    },
+    {
+      "epoch": 0.5704421052631579,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00020186260984247616,
+      "loss": 3.1768,
+      "step": 13548
+    },
+    {
+      "epoch": 0.5704842105263158,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00020182949744740482,
+      "loss": 3.0338,
+      "step": 13549
+    },
+    {
+      "epoch": 0.5705263157894737,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00020179638592994343,
+      "loss": 2.866,
+      "step": 13550
+    },
+    {
+      "epoch": 0.5705684210526316,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0002017632752906952,
+      "loss": 3.2152,
+      "step": 13551
+    },
+    {
+      "epoch": 0.5706105263157895,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00020173016553026358,
+      "loss": 3.4708,
+      "step": 13552
+    },
+    {
+      "epoch": 0.5706526315789474,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002016970566492515,
+      "loss": 3.3534,
+      "step": 13553
+    },
+    {
+      "epoch": 0.5706947368421053,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00020166394864826244,
+      "loss": 3.2921,
+      "step": 13554
+    },
+    {
+      "epoch": 0.5707368421052632,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00020163084152789926,
+      "loss": 3.3696,
+      "step": 13555
+    },
+    {
+      "epoch": 0.570778947368421,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00020159773528876534,
+      "loss": 3.0982,
+      "step": 13556
+    },
+    {
+      "epoch": 0.5708210526315789,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0002015646299314639,
+      "loss": 3.3727,
+      "step": 13557
+    },
+    {
+      "epoch": 0.5708631578947369,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00020153152545659798,
+      "loss": 2.8047,
+      "step": 13558
+    },
+    {
+      "epoch": 0.5709052631578947,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00020149842186477072,
+      "loss": 3.143,
+      "step": 13559
+    },
+    {
+      "epoch": 0.5709473684210526,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00020146531915658518,
+      "loss": 3.2787,
+      "step": 13560
+    },
+    {
+      "epoch": 0.5709894736842105,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00020143221733264456,
+      "loss": 3.1574,
+      "step": 13561
+    },
+    {
+      "epoch": 0.5710315789473684,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0002013991163935518,
+      "loss": 3.353,
+      "step": 13562
+    },
+    {
+      "epoch": 0.5710736842105263,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00020136601633991003,
+      "loss": 3.1434,
+      "step": 13563
+    },
+    {
+      "epoch": 0.5711157894736842,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00020133291717232244,
+      "loss": 2.7114,
+      "step": 13564
+    },
+    {
+      "epoch": 0.5711578947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00020129981889139176,
+      "loss": 3.0325,
+      "step": 13565
+    },
+    {
+      "epoch": 0.5712,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00020126672149772123,
+      "loss": 3.1738,
+      "step": 13566
+    },
+    {
+      "epoch": 0.5712421052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00020123362499191373,
+      "loss": 3.1487,
+      "step": 13567
+    },
+    {
+      "epoch": 0.5712842105263158,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0002012005293745724,
+      "loss": 2.931,
+      "step": 13568
+    },
+    {
+      "epoch": 0.5713263157894737,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00020116743464629997,
+      "loss": 3.248,
+      "step": 13569
+    },
+    {
+      "epoch": 0.5713684210526315,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0002011343408076995,
+      "loss": 3.3934,
+      "step": 13570
+    },
+    {
+      "epoch": 0.5714105263157895,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00020110124785937403,
+      "loss": 3.0199,
+      "step": 13571
+    },
+    {
+      "epoch": 0.5714526315789473,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00020106815580192629,
+      "loss": 3.3724,
+      "step": 13572
+    },
+    {
+      "epoch": 0.5714947368421053,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00020103506463595933,
+      "loss": 2.919,
+      "step": 13573
+    },
+    {
+      "epoch": 0.5715368421052631,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00020100197436207588,
+      "loss": 3.3387,
+      "step": 13574
+    },
+    {
+      "epoch": 0.5715789473684211,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00020096888498087893,
+      "loss": 2.9242,
+      "step": 13575
+    },
+    {
+      "epoch": 0.5716210526315789,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00020093579649297123,
+      "loss": 3.3454,
+      "step": 13576
+    },
+    {
+      "epoch": 0.5716631578947369,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00020090270889895569,
+      "loss": 2.5785,
+      "step": 13577
+    },
+    {
+      "epoch": 0.5717052631578947,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00020086962219943515,
+      "loss": 3.1434,
+      "step": 13578
+    },
+    {
+      "epoch": 0.5717473684210527,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0002008365363950123,
+      "loss": 3.1326,
+      "step": 13579
+    },
+    {
+      "epoch": 0.5717894736842105,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00020080345148629013,
+      "loss": 3.085,
+      "step": 13580
+    },
+    {
+      "epoch": 0.5718315789473685,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00020077036747387115,
+      "loss": 3.3156,
+      "step": 13581
+    },
+    {
+      "epoch": 0.5718736842105263,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00020073728435835827,
+      "loss": 3.3149,
+      "step": 13582
+    },
+    {
+      "epoch": 0.5719157894736842,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00020070420214035416,
+      "loss": 2.9619,
+      "step": 13583
+    },
+    {
+      "epoch": 0.5719578947368421,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00020067112082046152,
+      "loss": 2.7903,
+      "step": 13584
+    },
+    {
+      "epoch": 0.572,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00020063804039928323,
+      "loss": 3.2269,
+      "step": 13585
+    },
+    {
+      "epoch": 0.5720421052631579,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00020060496087742174,
+      "loss": 2.8271,
+      "step": 13586
+    },
+    {
+      "epoch": 0.5720842105263158,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00020057188225547996,
+      "loss": 3.1412,
+      "step": 13587
+    },
+    {
+      "epoch": 0.5721263157894737,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00020053880453406027,
+      "loss": 3.0548,
+      "step": 13588
+    },
+    {
+      "epoch": 0.5721684210526315,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00020050572771376557,
+      "loss": 3.0698,
+      "step": 13589
+    },
+    {
+      "epoch": 0.5722105263157895,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00020047265179519826,
+      "loss": 2.918,
+      "step": 13590
+    },
+    {
+      "epoch": 0.5722526315789473,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.000200439576778961,
+      "loss": 2.994,
+      "step": 13591
+    },
+    {
+      "epoch": 0.5722947368421053,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0002004065026656565,
+      "loss": 3.0818,
+      "step": 13592
+    },
+    {
+      "epoch": 0.5723368421052631,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002003734294558872,
+      "loss": 2.8728,
+      "step": 13593
+    },
+    {
+      "epoch": 0.5723789473684211,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00020034035715025582,
+      "loss": 3.0596,
+      "step": 13594
+    },
+    {
+      "epoch": 0.5724210526315789,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00020030728574936467,
+      "loss": 2.934,
+      "step": 13595
+    },
+    {
+      "epoch": 0.5724631578947369,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00020027421525381647,
+      "loss": 3.0167,
+      "step": 13596
+    },
+    {
+      "epoch": 0.5725052631578947,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00020024114566421356,
+      "loss": 3.0947,
+      "step": 13597
+    },
+    {
+      "epoch": 0.5725473684210526,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0002002080769811585,
+      "loss": 3.2543,
+      "step": 13598
+    },
+    {
+      "epoch": 0.5725894736842105,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00020017500920525383,
+      "loss": 3.1688,
+      "step": 13599
+    },
+    {
+      "epoch": 0.5726315789473684,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0002001419423371019,
+      "loss": 3.0509,
+      "step": 13600
+    },
+    {
+      "epoch": 0.5726736842105263,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0002001088763773053,
+      "loss": 3.1073,
+      "step": 13601
+    },
+    {
+      "epoch": 0.5727157894736842,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0002000758113264663,
+      "loss": 3.0342,
+      "step": 13602
+    },
+    {
+      "epoch": 0.5727578947368421,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0002000427471851873,
+      "loss": 2.999,
+      "step": 13603
+    },
+    {
+      "epoch": 0.5728,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00020000968395407089,
+      "loss": 3.1953,
+      "step": 13604
+    },
+    {
+      "epoch": 0.5728421052631579,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00019997662163371915,
+      "loss": 3.1256,
+      "step": 13605
+    },
+    {
+      "epoch": 0.5728842105263158,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00019994356022473472,
+      "loss": 3.0893,
+      "step": 13606
+    },
+    {
+      "epoch": 0.5729263157894737,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001999104997277197,
+      "loss": 3.2867,
+      "step": 13607
+    },
+    {
+      "epoch": 0.5729684210526316,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0001998774401432766,
+      "loss": 2.9404,
+      "step": 13608
+    },
+    {
+      "epoch": 0.5730105263157895,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00019984438147200757,
+      "loss": 2.9398,
+      "step": 13609
+    },
+    {
+      "epoch": 0.5730526315789474,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00019981132371451498,
+      "loss": 3.4143,
+      "step": 13610
+    },
+    {
+      "epoch": 0.5730947368421052,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001997782668714012,
+      "loss": 3.0714,
+      "step": 13611
+    },
+    {
+      "epoch": 0.5731368421052632,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001997452109432683,
+      "loss": 3.3885,
+      "step": 13612
+    },
+    {
+      "epoch": 0.573178947368421,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00019971215593071872,
+      "loss": 3.4412,
+      "step": 13613
+    },
+    {
+      "epoch": 0.573221052631579,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00019967910183435448,
+      "loss": 2.8344,
+      "step": 13614
+    },
+    {
+      "epoch": 0.5732631578947368,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001996460486547779,
+      "loss": 3.0669,
+      "step": 13615
+    },
+    {
+      "epoch": 0.5733052631578948,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00019961299639259118,
+      "loss": 3.5075,
+      "step": 13616
+    },
+    {
+      "epoch": 0.5733473684210526,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00019957994504839639,
+      "loss": 3.4828,
+      "step": 13617
+    },
+    {
+      "epoch": 0.5733894736842106,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00019954689462279578,
+      "loss": 2.9475,
+      "step": 13618
+    },
+    {
+      "epoch": 0.5734315789473684,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00019951384511639149,
+      "loss": 3.3577,
+      "step": 13619
+    },
+    {
+      "epoch": 0.5734736842105264,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019948079652978569,
+      "loss": 3.1781,
+      "step": 13620
+    },
+    {
+      "epoch": 0.5735157894736842,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001994477488635803,
+      "loss": 3.4171,
+      "step": 13621
+    },
+    {
+      "epoch": 0.5735578947368422,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00019941470211837768,
+      "loss": 3.2132,
+      "step": 13622
+    },
+    {
+      "epoch": 0.5736,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001993816562947796,
+      "loss": 2.7517,
+      "step": 13623
+    },
+    {
+      "epoch": 0.5736421052631578,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00019934861139338827,
+      "loss": 3.0417,
+      "step": 13624
+    },
+    {
+      "epoch": 0.5736842105263158,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001993155674148058,
+      "loss": 3.4557,
+      "step": 13625
+    },
+    {
+      "epoch": 0.5737263157894736,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001992825243596341,
+      "loss": 2.835,
+      "step": 13626
+    },
+    {
+      "epoch": 0.5737684210526316,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0001992494822284753,
+      "loss": 3.0413,
+      "step": 13627
+    },
+    {
+      "epoch": 0.5738105263157894,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0001992164410219312,
+      "loss": 3.423,
+      "step": 13628
+    },
+    {
+      "epoch": 0.5738526315789474,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00019918340074060397,
+      "loss": 3.0479,
+      "step": 13629
+    },
+    {
+      "epoch": 0.5738947368421052,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0001991503613850954,
+      "loss": 3.2468,
+      "step": 13630
+    },
+    {
+      "epoch": 0.5739368421052632,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0001991173229560075,
+      "loss": 3.0716,
+      "step": 13631
+    },
+    {
+      "epoch": 0.573978947368421,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00019908428545394226,
+      "loss": 3.3137,
+      "step": 13632
+    },
+    {
+      "epoch": 0.574021052631579,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00019905124887950143,
+      "loss": 2.8699,
+      "step": 13633
+    },
+    {
+      "epoch": 0.5740631578947368,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00019901821323328704,
+      "loss": 3.4341,
+      "step": 13634
+    },
+    {
+      "epoch": 0.5741052631578948,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00019898517851590084,
+      "loss": 2.8827,
+      "step": 13635
+    },
+    {
+      "epoch": 0.5741473684210526,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00019895214472794484,
+      "loss": 3.211,
+      "step": 13636
+    },
+    {
+      "epoch": 0.5741894736842105,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00019891911187002072,
+      "loss": 3.2151,
+      "step": 13637
+    },
+    {
+      "epoch": 0.5742315789473684,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00019888607994273034,
+      "loss": 3.1404,
+      "step": 13638
+    },
+    {
+      "epoch": 0.5742736842105263,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0001988530489466756,
+      "loss": 2.8683,
+      "step": 13639
+    },
+    {
+      "epoch": 0.5743157894736842,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00019882001888245815,
+      "loss": 2.3499,
+      "step": 13640
+    },
+    {
+      "epoch": 0.5743578947368421,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00019878698975067988,
+      "loss": 3.0614,
+      "step": 13641
+    },
+    {
+      "epoch": 0.5744,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00019875396155194242,
+      "loss": 2.8017,
+      "step": 13642
+    },
+    {
+      "epoch": 0.5744421052631579,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00019872093428684767,
+      "loss": 3.169,
+      "step": 13643
+    },
+    {
+      "epoch": 0.5744842105263158,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00019868790795599713,
+      "loss": 3.2016,
+      "step": 13644
+    },
+    {
+      "epoch": 0.5745263157894737,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00019865488255999266,
+      "loss": 2.6418,
+      "step": 13645
+    },
+    {
+      "epoch": 0.5745684210526316,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00019862185809943594,
+      "loss": 3.6396,
+      "step": 13646
+    },
+    {
+      "epoch": 0.5746105263157895,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00019858883457492854,
+      "loss": 2.9667,
+      "step": 13647
+    },
+    {
+      "epoch": 0.5746526315789474,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00019855581198707225,
+      "loss": 2.6119,
+      "step": 13648
+    },
+    {
+      "epoch": 0.5746947368421053,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00019852279033646852,
+      "loss": 3.2095,
+      "step": 13649
+    },
+    {
+      "epoch": 0.5747368421052632,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00019848976962371916,
+      "loss": 3.0505,
+      "step": 13650
+    },
+    {
+      "epoch": 0.5747789473684211,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00019845674984942557,
+      "loss": 3.4954,
+      "step": 13651
+    },
+    {
+      "epoch": 0.5748210526315789,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00019842373101418948,
+      "loss": 3.6447,
+      "step": 13652
+    },
+    {
+      "epoch": 0.5748631578947369,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0001983907131186125,
+      "loss": 3.1914,
+      "step": 13653
+    },
+    {
+      "epoch": 0.5749052631578947,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00019835769616329595,
+      "loss": 2.9919,
+      "step": 13654
+    },
+    {
+      "epoch": 0.5749473684210527,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00019832468014884165,
+      "loss": 3.0797,
+      "step": 13655
+    },
+    {
+      "epoch": 0.5749894736842105,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00019829166507585083,
+      "loss": 3.087,
+      "step": 13656
+    },
+    {
+      "epoch": 0.5750315789473684,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00019825865094492512,
+      "loss": 2.8431,
+      "step": 13657
+    },
+    {
+      "epoch": 0.5750736842105263,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0001982256377566661,
+      "loss": 3.5012,
+      "step": 13658
+    },
+    {
+      "epoch": 0.5751157894736842,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00019819262551167504,
+      "loss": 3.4376,
+      "step": 13659
+    },
+    {
+      "epoch": 0.5751578947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00019815961421055354,
+      "loss": 3.0873,
+      "step": 13660
+    },
+    {
+      "epoch": 0.5752,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00019812660385390288,
+      "loss": 2.9918,
+      "step": 13661
+    },
+    {
+      "epoch": 0.5752421052631579,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00019809359444232467,
+      "loss": 2.7084,
+      "step": 13662
+    },
+    {
+      "epoch": 0.5752842105263158,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00019806058597642004,
+      "loss": 3.4384,
+      "step": 13663
+    },
+    {
+      "epoch": 0.5753263157894737,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00019802757845679053,
+      "loss": 3.2723,
+      "step": 13664
+    },
+    {
+      "epoch": 0.5753684210526315,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00019799457188403758,
+      "loss": 3.1454,
+      "step": 13665
+    },
+    {
+      "epoch": 0.5754105263157895,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00019796156625876232,
+      "loss": 2.7084,
+      "step": 13666
+    },
+    {
+      "epoch": 0.5754526315789473,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00019792856158156624,
+      "loss": 3.069,
+      "step": 13667
+    },
+    {
+      "epoch": 0.5754947368421053,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00019789555785305053,
+      "loss": 3.2069,
+      "step": 13668
+    },
+    {
+      "epoch": 0.5755368421052631,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00019786255507381665,
+      "loss": 3.3254,
+      "step": 13669
+    },
+    {
+      "epoch": 0.5755789473684211,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00019782955324446563,
+      "loss": 2.7682,
+      "step": 13670
+    },
+    {
+      "epoch": 0.5756210526315789,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00019779655236559885,
+      "loss": 3.0039,
+      "step": 13671
+    },
+    {
+      "epoch": 0.5756631578947369,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00019776355243781768,
+      "loss": 3.0551,
+      "step": 13672
+    },
+    {
+      "epoch": 0.5757052631578947,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00019773055346172309,
+      "loss": 3.0055,
+      "step": 13673
+    },
+    {
+      "epoch": 0.5757473684210527,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00019769755543791646,
+      "loss": 3.3821,
+      "step": 13674
+    },
+    {
+      "epoch": 0.5757894736842105,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00019766455836699891,
+      "loss": 2.7475,
+      "step": 13675
+    },
+    {
+      "epoch": 0.5758315789473685,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00019763156224957164,
+      "loss": 3.2862,
+      "step": 13676
+    },
+    {
+      "epoch": 0.5758736842105263,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0001975985670862357,
+      "loss": 2.9635,
+      "step": 13677
+    },
+    {
+      "epoch": 0.5759157894736842,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001975655728775923,
+      "loss": 2.8314,
+      "step": 13678
+    },
+    {
+      "epoch": 0.5759578947368421,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00019753257962424264,
+      "loss": 3.2541,
+      "step": 13679
+    },
+    {
+      "epoch": 0.576,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00019749958732678767,
+      "loss": 3.2719,
+      "step": 13680
+    },
+    {
+      "epoch": 0.5760421052631579,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00019746659598582862,
+      "loss": 3.1393,
+      "step": 13681
+    },
+    {
+      "epoch": 0.5760842105263158,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00019743360560196634,
+      "loss": 2.8229,
+      "step": 13682
+    },
+    {
+      "epoch": 0.5761263157894737,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00019740061617580206,
+      "loss": 2.6648,
+      "step": 13683
+    },
+    {
+      "epoch": 0.5761684210526316,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00019736762770793673,
+      "loss": 3.2121,
+      "step": 13684
+    },
+    {
+      "epoch": 0.5762105263157895,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00019733464019897136,
+      "loss": 2.9435,
+      "step": 13685
+    },
+    {
+      "epoch": 0.5762526315789473,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0001973016536495071,
+      "loss": 3.3146,
+      "step": 13686
+    },
+    {
+      "epoch": 0.5762947368421053,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019726866806014465,
+      "loss": 2.9521,
+      "step": 13687
+    },
+    {
+      "epoch": 0.5763368421052631,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00019723568343148524,
+      "loss": 3.2617,
+      "step": 13688
+    },
+    {
+      "epoch": 0.5763789473684211,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00019720269976412956,
+      "loss": 3.2851,
+      "step": 13689
+    },
+    {
+      "epoch": 0.576421052631579,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0001971697170586787,
+      "loss": 3.0564,
+      "step": 13690
+    },
+    {
+      "epoch": 0.5764631578947368,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00019713673531573353,
+      "loss": 3.2995,
+      "step": 13691
+    },
+    {
+      "epoch": 0.5765052631578947,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0001971037545358949,
+      "loss": 3.3089,
+      "step": 13692
+    },
+    {
+      "epoch": 0.5765473684210526,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00019707077471976375,
+      "loss": 3.106,
+      "step": 13693
+    },
+    {
+      "epoch": 0.5765894736842105,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00019703779586794086,
+      "loss": 2.9998,
+      "step": 13694
+    },
+    {
+      "epoch": 0.5766315789473684,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00019700481798102718,
+      "loss": 2.9926,
+      "step": 13695
+    },
+    {
+      "epoch": 0.5766736842105263,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00019697184105962334,
+      "loss": 3.4573,
+      "step": 13696
+    },
+    {
+      "epoch": 0.5767157894736842,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00019693886510433037,
+      "loss": 2.6345,
+      "step": 13697
+    },
+    {
+      "epoch": 0.5767578947368421,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00019690589011574884,
+      "loss": 3.0796,
+      "step": 13698
+    },
+    {
+      "epoch": 0.5768,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001968729160944796,
+      "loss": 3.4196,
+      "step": 13699
+    },
+    {
+      "epoch": 0.5768421052631579,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00019683994304112347,
+      "loss": 3.1593,
+      "step": 13700
+    },
+    {
+      "epoch": 0.5768842105263158,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00019680697095628103,
+      "loss": 2.9102,
+      "step": 13701
+    },
+    {
+      "epoch": 0.5769263157894737,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00019677399984055323,
+      "loss": 3.1848,
+      "step": 13702
+    },
+    {
+      "epoch": 0.5769684210526316,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00019674102969454048,
+      "loss": 2.7659,
+      "step": 13703
+    },
+    {
+      "epoch": 0.5770105263157894,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00019670806051884357,
+      "loss": 2.865,
+      "step": 13704
+    },
+    {
+      "epoch": 0.5770526315789474,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00019667509231406332,
+      "loss": 3.0852,
+      "step": 13705
+    },
+    {
+      "epoch": 0.5770947368421052,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00019664212508080012,
+      "loss": 2.9547,
+      "step": 13706
+    },
+    {
+      "epoch": 0.5771368421052632,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00019660915881965482,
+      "loss": 3.28,
+      "step": 13707
+    },
+    {
+      "epoch": 0.577178947368421,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001965761935312278,
+      "loss": 2.5273,
+      "step": 13708
+    },
+    {
+      "epoch": 0.577221052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00019654322921611983,
+      "loss": 3.3062,
+      "step": 13709
+    },
+    {
+      "epoch": 0.5772631578947368,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00019651026587493137,
+      "loss": 2.8975,
+      "step": 13710
+    },
+    {
+      "epoch": 0.5773052631578948,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00019647730350826298,
+      "loss": 3.2323,
+      "step": 13711
+    },
+    {
+      "epoch": 0.5773473684210526,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00019644434211671537,
+      "loss": 2.9513,
+      "step": 13712
+    },
+    {
+      "epoch": 0.5773894736842106,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00019641138170088882,
+      "loss": 3.2068,
+      "step": 13713
+    },
+    {
+      "epoch": 0.5774315789473684,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00019637842226138402,
+      "loss": 3.1636,
+      "step": 13714
+    },
+    {
+      "epoch": 0.5774736842105264,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00019634546379880126,
+      "loss": 3.102,
+      "step": 13715
+    },
+    {
+      "epoch": 0.5775157894736842,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00019631250631374118,
+      "loss": 3.246,
+      "step": 13716
+    },
+    {
+      "epoch": 0.5775578947368422,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00019627954980680406,
+      "loss": 3.2551,
+      "step": 13717
+    },
+    {
+      "epoch": 0.5776,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00019624659427859052,
+      "loss": 3.3713,
+      "step": 13718
+    },
+    {
+      "epoch": 0.5776421052631578,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00019621363972970087,
+      "loss": 3.1488,
+      "step": 13719
+    },
+    {
+      "epoch": 0.5776842105263158,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00019618068616073542,
+      "loss": 3.0424,
+      "step": 13720
+    },
+    {
+      "epoch": 0.5777263157894736,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00019614773357229476,
+      "loss": 2.7422,
+      "step": 13721
+    },
+    {
+      "epoch": 0.5777684210526316,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00019611478196497902,
+      "loss": 2.8786,
+      "step": 13722
+    },
+    {
+      "epoch": 0.5778105263157894,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00019608183133938874,
+      "loss": 2.9028,
+      "step": 13723
+    },
+    {
+      "epoch": 0.5778526315789474,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00019604888169612403,
+      "loss": 3.0107,
+      "step": 13724
+    },
+    {
+      "epoch": 0.5778947368421052,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00019601593303578534,
+      "loss": 3.3451,
+      "step": 13725
+    },
+    {
+      "epoch": 0.5779368421052632,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00019598298535897292,
+      "loss": 3.1659,
+      "step": 13726
+    },
+    {
+      "epoch": 0.577978947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00019595003866628702,
+      "loss": 3.0619,
+      "step": 13727
+    },
+    {
+      "epoch": 0.578021052631579,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00019591709295832798,
+      "loss": 3.0275,
+      "step": 13728
+    },
+    {
+      "epoch": 0.5780631578947368,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00019588414823569592,
+      "loss": 3.1347,
+      "step": 13729
+    },
+    {
+      "epoch": 0.5781052631578948,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00019585120449899115,
+      "loss": 3.3909,
+      "step": 13730
+    },
+    {
+      "epoch": 0.5781473684210526,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00019581826174881373,
+      "loss": 3.0728,
+      "step": 13731
+    },
+    {
+      "epoch": 0.5781894736842105,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001957853199857639,
+      "loss": 3.0892,
+      "step": 13732
+    },
+    {
+      "epoch": 0.5782315789473684,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00019575237921044198,
+      "loss": 3.2125,
+      "step": 13733
+    },
+    {
+      "epoch": 0.5782736842105263,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00019571943942344784,
+      "loss": 3.2148,
+      "step": 13734
+    },
+    {
+      "epoch": 0.5783157894736842,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0001956865006253818,
+      "loss": 3.2875,
+      "step": 13735
+    },
+    {
+      "epoch": 0.5783578947368421,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019565356281684384,
+      "loss": 2.9895,
+      "step": 13736
+    },
+    {
+      "epoch": 0.5784,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00019562062599843422,
+      "loss": 2.9685,
+      "step": 13737
+    },
+    {
+      "epoch": 0.5784421052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0001955876901707528,
+      "loss": 3.3921,
+      "step": 13738
+    },
+    {
+      "epoch": 0.5784842105263158,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001955547553343997,
+      "loss": 3.2356,
+      "step": 13739
+    },
+    {
+      "epoch": 0.5785263157894737,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00019552182148997513,
+      "loss": 3.1637,
+      "step": 13740
+    },
+    {
+      "epoch": 0.5785684210526316,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00019548888863807885,
+      "loss": 3.0512,
+      "step": 13741
+    },
+    {
+      "epoch": 0.5786105263157895,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00019545595677931103,
+      "loss": 3.1227,
+      "step": 13742
+    },
+    {
+      "epoch": 0.5786526315789474,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00019542302591427153,
+      "loss": 2.885,
+      "step": 13743
+    },
+    {
+      "epoch": 0.5786947368421053,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00019539009604356049,
+      "loss": 3.1261,
+      "step": 13744
+    },
+    {
+      "epoch": 0.5787368421052631,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00019535716716777762,
+      "loss": 2.8873,
+      "step": 13745
+    },
+    {
+      "epoch": 0.5787789473684211,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00019532423928752296,
+      "loss": 3.0553,
+      "step": 13746
+    },
+    {
+      "epoch": 0.5788210526315789,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00019529131240339653,
+      "loss": 3.3112,
+      "step": 13747
+    },
+    {
+      "epoch": 0.5788631578947369,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00019525838651599797,
+      "loss": 3.3617,
+      "step": 13748
+    },
+    {
+      "epoch": 0.5789052631578947,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00019522546162592744,
+      "loss": 3.3629,
+      "step": 13749
+    },
+    {
+      "epoch": 0.5789473684210527,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00019519253773378454,
+      "loss": 3.076,
+      "step": 13750
+    },
+    {
+      "epoch": 0.5789894736842105,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00019515961484016918,
+      "loss": 2.8628,
+      "step": 13751
+    },
+    {
+      "epoch": 0.5790315789473685,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00019512669294568125,
+      "loss": 3.2886,
+      "step": 13752
+    },
+    {
+      "epoch": 0.5790736842105263,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00019509377205092045,
+      "loss": 3.3293,
+      "step": 13753
+    },
+    {
+      "epoch": 0.5791157894736843,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00019506085215648672,
+      "loss": 2.9399,
+      "step": 13754
+    },
+    {
+      "epoch": 0.5791578947368421,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00019502793326297963,
+      "loss": 2.8272,
+      "step": 13755
+    },
+    {
+      "epoch": 0.5792,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001949950153709991,
+      "loss": 3.0928,
+      "step": 13756
+    },
+    {
+      "epoch": 0.5792421052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00019496209848114465,
+      "loss": 3.1687,
+      "step": 13757
+    },
+    {
+      "epoch": 0.5792842105263157,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0001949291825940161,
+      "loss": 3.147,
+      "step": 13758
+    },
+    {
+      "epoch": 0.5793263157894737,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001948962677102132,
+      "loss": 3.055,
+      "step": 13759
+    },
+    {
+      "epoch": 0.5793684210526315,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00019486335383033548,
+      "loss": 2.8733,
+      "step": 13760
+    },
+    {
+      "epoch": 0.5794105263157895,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00019483044095498282,
+      "loss": 2.8951,
+      "step": 13761
+    },
+    {
+      "epoch": 0.5794526315789473,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00019479752908475455,
+      "loss": 3.1999,
+      "step": 13762
+    },
+    {
+      "epoch": 0.5794947368421053,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00019476461822025057,
+      "loss": 3.1258,
+      "step": 13763
+    },
+    {
+      "epoch": 0.5795368421052631,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00019473170836207024,
+      "loss": 3.0004,
+      "step": 13764
+    },
+    {
+      "epoch": 0.5795789473684211,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00019469879951081322,
+      "loss": 3.4338,
+      "step": 13765
+    },
+    {
+      "epoch": 0.5796210526315789,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019466589166707924,
+      "loss": 3.0753,
+      "step": 13766
+    },
+    {
+      "epoch": 0.5796631578947369,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00019463298483146757,
+      "loss": 2.815,
+      "step": 13767
+    },
+    {
+      "epoch": 0.5797052631578947,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00019460007900457793,
+      "loss": 3.2149,
+      "step": 13768
+    },
+    {
+      "epoch": 0.5797473684210527,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0001945671741870097,
+      "loss": 3.1791,
+      "step": 13769
+    },
+    {
+      "epoch": 0.5797894736842105,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00019453427037936257,
+      "loss": 3.16,
+      "step": 13770
+    },
+    {
+      "epoch": 0.5798315789473685,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0001945013675822357,
+      "loss": 3.202,
+      "step": 13771
+    },
+    {
+      "epoch": 0.5798736842105263,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0001944684657962287,
+      "loss": 2.7253,
+      "step": 13772
+    },
+    {
+      "epoch": 0.5799157894736842,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00019443556502194114,
+      "loss": 2.9746,
+      "step": 13773
+    },
+    {
+      "epoch": 0.5799578947368421,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00019440266525997218,
+      "loss": 2.7486,
+      "step": 13774
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00019436976651092142,
+      "loss": 3.0587,
+      "step": 13775
+    },
+    {
+      "epoch": 0.5800421052631579,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00019433686877538814,
+      "loss": 3.3073,
+      "step": 13776
+    },
+    {
+      "epoch": 0.5800842105263158,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00019430397205397168,
+      "loss": 3.2184,
+      "step": 13777
+    },
+    {
+      "epoch": 0.5801263157894737,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00019427107634727137,
+      "loss": 3.1135,
+      "step": 13778
+    },
+    {
+      "epoch": 0.5801684210526316,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00019423818165588657,
+      "loss": 3.0284,
+      "step": 13779
+    },
+    {
+      "epoch": 0.5802105263157895,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001942052879804167,
+      "loss": 3.1657,
+      "step": 13780
+    },
+    {
+      "epoch": 0.5802526315789474,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00019417239532146084,
+      "loss": 2.7498,
+      "step": 13781
+    },
+    {
+      "epoch": 0.5802947368421053,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001941395036796184,
+      "loss": 3.1181,
+      "step": 13782
+    },
+    {
+      "epoch": 0.5803368421052632,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001941066130554885,
+      "loss": 3.1329,
+      "step": 13783
+    },
+    {
+      "epoch": 0.5803789473684211,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001940737234496705,
+      "loss": 2.8607,
+      "step": 13784
+    },
+    {
+      "epoch": 0.580421052631579,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00019404083486276345,
+      "loss": 2.8194,
+      "step": 13785
+    },
+    {
+      "epoch": 0.5804631578947368,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00019400794729536666,
+      "loss": 2.9065,
+      "step": 13786
+    },
+    {
+      "epoch": 0.5805052631578947,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00019397506074807938,
+      "loss": 2.9228,
+      "step": 13787
+    },
+    {
+      "epoch": 0.5805473684210526,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0001939421752215006,
+      "loss": 3.3125,
+      "step": 13788
+    },
+    {
+      "epoch": 0.5805894736842105,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0001939092907162296,
+      "loss": 3.2819,
+      "step": 13789
+    },
+    {
+      "epoch": 0.5806315789473684,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00019387640723286532,
+      "loss": 3.1888,
+      "step": 13790
+    },
+    {
+      "epoch": 0.5806736842105263,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00019384352477200698,
+      "loss": 3.503,
+      "step": 13791
+    },
+    {
+      "epoch": 0.5807157894736842,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00019381064333425367,
+      "loss": 3.3177,
+      "step": 13792
+    },
+    {
+      "epoch": 0.5807578947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00019377776292020435,
+      "loss": 2.8954,
+      "step": 13793
+    },
+    {
+      "epoch": 0.5808,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0001937448835304582,
+      "loss": 3.0454,
+      "step": 13794
+    },
+    {
+      "epoch": 0.5808421052631579,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0001937120051656141,
+      "loss": 3.3597,
+      "step": 13795
+    },
+    {
+      "epoch": 0.5808842105263158,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00019367912782627126,
+      "loss": 2.838,
+      "step": 13796
+    },
+    {
+      "epoch": 0.5809263157894737,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00019364625151302843,
+      "loss": 3.1315,
+      "step": 13797
+    },
+    {
+      "epoch": 0.5809684210526316,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00019361337622648479,
+      "loss": 2.9543,
+      "step": 13798
+    },
+    {
+      "epoch": 0.5810105263157894,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00019358050196723904,
+      "loss": 2.5844,
+      "step": 13799
+    },
+    {
+      "epoch": 0.5810526315789474,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001935476287358903,
+      "loss": 3.595,
+      "step": 13800
+    },
+    {
+      "epoch": 0.5810947368421052,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00019351475653303746,
+      "loss": 3.1196,
+      "step": 13801
+    },
+    {
+      "epoch": 0.5811368421052632,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00019348188535927937,
+      "loss": 3.3868,
+      "step": 13802
+    },
+    {
+      "epoch": 0.581178947368421,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.000193449015215215,
+      "loss": 3.2094,
+      "step": 13803
+    },
+    {
+      "epoch": 0.581221052631579,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.000193416146101443,
+      "loss": 2.9218,
+      "step": 13804
+    },
+    {
+      "epoch": 0.5812631578947368,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00019338327801856237,
+      "loss": 2.6842,
+      "step": 13805
+    },
+    {
+      "epoch": 0.5813052631578948,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00019335041096717198,
+      "loss": 3.0885,
+      "step": 13806
+    },
+    {
+      "epoch": 0.5813473684210526,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00019331754494787047,
+      "loss": 2.9682,
+      "step": 13807
+    },
+    {
+      "epoch": 0.5813894736842106,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019328467996125676,
+      "loss": 3.0242,
+      "step": 13808
+    },
+    {
+      "epoch": 0.5814315789473684,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00019325181600792946,
+      "loss": 2.9842,
+      "step": 13809
+    },
+    {
+      "epoch": 0.5814736842105264,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00019321895308848744,
+      "loss": 2.784,
+      "step": 13810
+    },
+    {
+      "epoch": 0.5815157894736842,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00019318609120352932,
+      "loss": 3.1663,
+      "step": 13811
+    },
+    {
+      "epoch": 0.581557894736842,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00019315323035365388,
+      "loss": 3.3585,
+      "step": 13812
+    },
+    {
+      "epoch": 0.5816,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0001931203705394599,
+      "loss": 2.9644,
+      "step": 13813
+    },
+    {
+      "epoch": 0.5816421052631578,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00019308751176154578,
+      "loss": 2.9935,
+      "step": 13814
+    },
+    {
+      "epoch": 0.5816842105263158,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00019305465402051048,
+      "loss": 3.4496,
+      "step": 13815
+    },
+    {
+      "epoch": 0.5817263157894736,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00019302179731695238,
+      "loss": 3.1577,
+      "step": 13816
+    },
+    {
+      "epoch": 0.5817684210526316,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0001929889416514702,
+      "loss": 2.6438,
+      "step": 13817
+    },
+    {
+      "epoch": 0.5818105263157894,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001929560870246625,
+      "loss": 3.1945,
+      "step": 13818
+    },
+    {
+      "epoch": 0.5818526315789474,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00019292323343712793,
+      "loss": 3.381,
+      "step": 13819
+    },
+    {
+      "epoch": 0.5818947368421052,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00019289038088946494,
+      "loss": 3.0547,
+      "step": 13820
+    },
+    {
+      "epoch": 0.5819368421052632,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00019285752938227207,
+      "loss": 3.0595,
+      "step": 13821
+    },
+    {
+      "epoch": 0.581978947368421,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00019282467891614801,
+      "loss": 3.0542,
+      "step": 13822
+    },
+    {
+      "epoch": 0.582021052631579,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019279182949169098,
+      "loss": 2.9763,
+      "step": 13823
+    },
+    {
+      "epoch": 0.5820631578947368,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00019275898110949975,
+      "loss": 3.2207,
+      "step": 13824
+    },
+    {
+      "epoch": 0.5821052631578948,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0001927261337701725,
+      "loss": 2.6718,
+      "step": 13825
+    },
+    {
+      "epoch": 0.5821473684210526,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00019269328747430778,
+      "loss": 3.2863,
+      "step": 13826
+    },
+    {
+      "epoch": 0.5821894736842105,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00019266044222250413,
+      "loss": 3.1867,
+      "step": 13827
+    },
+    {
+      "epoch": 0.5822315789473684,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00019262759801535974,
+      "loss": 2.7265,
+      "step": 13828
+    },
+    {
+      "epoch": 0.5822736842105263,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00019259475485347325,
+      "loss": 3.2155,
+      "step": 13829
+    },
+    {
+      "epoch": 0.5823157894736842,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00019256191273744276,
+      "loss": 3.4406,
+      "step": 13830
+    },
+    {
+      "epoch": 0.5823578947368421,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00019252907166786683,
+      "loss": 2.801,
+      "step": 13831
+    },
+    {
+      "epoch": 0.5824,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00019249623164534357,
+      "loss": 3.2797,
+      "step": 13832
+    },
+    {
+      "epoch": 0.5824421052631579,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00019246339267047145,
+      "loss": 3.0109,
+      "step": 13833
+    },
+    {
+      "epoch": 0.5824842105263158,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00019243055474384873,
+      "loss": 3.1749,
+      "step": 13834
+    },
+    {
+      "epoch": 0.5825263157894737,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00019239771786607368,
+      "loss": 3.1003,
+      "step": 13835
+    },
+    {
+      "epoch": 0.5825684210526316,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001923648820377445,
+      "loss": 3.1046,
+      "step": 13836
+    },
+    {
+      "epoch": 0.5826105263157895,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00019233204725945945,
+      "loss": 2.919,
+      "step": 13837
+    },
+    {
+      "epoch": 0.5826526315789474,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00019229921353181682,
+      "loss": 2.955,
+      "step": 13838
+    },
+    {
+      "epoch": 0.5826947368421053,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00019226638085541458,
+      "loss": 3.0983,
+      "step": 13839
+    },
+    {
+      "epoch": 0.5827368421052631,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001922335492308511,
+      "loss": 3.0569,
+      "step": 13840
+    },
+    {
+      "epoch": 0.5827789473684211,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00019220071865872458,
+      "loss": 3.2637,
+      "step": 13841
+    },
+    {
+      "epoch": 0.5828210526315789,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00019216788913963294,
+      "loss": 3.421,
+      "step": 13842
+    },
+    {
+      "epoch": 0.5828631578947369,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00019213506067417447,
+      "loss": 2.7806,
+      "step": 13843
+    },
+    {
+      "epoch": 0.5829052631578947,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00019210223326294714,
+      "loss": 3.4063,
+      "step": 13844
+    },
+    {
+      "epoch": 0.5829473684210527,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001920694069065492,
+      "loss": 3.3458,
+      "step": 13845
+    },
+    {
+      "epoch": 0.5829894736842105,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00019203658160557848,
+      "loss": 3.3923,
+      "step": 13846
+    },
+    {
+      "epoch": 0.5830315789473685,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00019200375736063316,
+      "loss": 3.1868,
+      "step": 13847
+    },
+    {
+      "epoch": 0.5830736842105263,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00019197093417231136,
+      "loss": 3.0917,
+      "step": 13848
+    },
+    {
+      "epoch": 0.5831157894736843,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001919381120412108,
+      "loss": 3.1577,
+      "step": 13849
+    },
+    {
+      "epoch": 0.5831578947368421,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0001919052909679297,
+      "loss": 2.9812,
+      "step": 13850
+    },
+    {
+      "epoch": 0.5832,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00019187247095306598,
+      "loss": 3.6194,
+      "step": 13851
+    },
+    {
+      "epoch": 0.5832421052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00019183965199721746,
+      "loss": 2.9838,
+      "step": 13852
+    },
+    {
+      "epoch": 0.5832842105263157,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00019180683410098217,
+      "loss": 2.6956,
+      "step": 13853
+    },
+    {
+      "epoch": 0.5833263157894737,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00019177401726495796,
+      "loss": 2.8155,
+      "step": 13854
+    },
+    {
+      "epoch": 0.5833684210526315,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00019174120148974282,
+      "loss": 2.8291,
+      "step": 13855
+    },
+    {
+      "epoch": 0.5834105263157895,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00019170838677593448,
+      "loss": 3.076,
+      "step": 13856
+    },
+    {
+      "epoch": 0.5834526315789473,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00019167557312413093,
+      "loss": 3.3579,
+      "step": 13857
+    },
+    {
+      "epoch": 0.5834947368421053,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0001916427605349298,
+      "loss": 2.6468,
+      "step": 13858
+    },
+    {
+      "epoch": 0.5835368421052631,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.000191609949008929,
+      "loss": 3.21,
+      "step": 13859
+    },
+    {
+      "epoch": 0.5835789473684211,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001915771385467264,
+      "loss": 3.4829,
+      "step": 13860
+    },
+    {
+      "epoch": 0.5836210526315789,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001915443291489196,
+      "loss": 2.8703,
+      "step": 13861
+    },
+    {
+      "epoch": 0.5836631578947369,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00019151152081610655,
+      "loss": 3.3593,
+      "step": 13862
+    },
+    {
+      "epoch": 0.5837052631578947,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001914787135488848,
+      "loss": 3.2309,
+      "step": 13863
+    },
+    {
+      "epoch": 0.5837473684210527,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00019144590734785223,
+      "loss": 3.3431,
+      "step": 13864
+    },
+    {
+      "epoch": 0.5837894736842105,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00019141310221360632,
+      "loss": 3.0165,
+      "step": 13865
+    },
+    {
+      "epoch": 0.5838315789473684,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00019138029814674481,
+      "loss": 3.5003,
+      "step": 13866
+    },
+    {
+      "epoch": 0.5838736842105263,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00019134749514786555,
+      "loss": 3.181,
+      "step": 13867
+    },
+    {
+      "epoch": 0.5839157894736842,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0001913146932175659,
+      "loss": 3.1319,
+      "step": 13868
+    },
+    {
+      "epoch": 0.5839578947368421,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001912818923564436,
+      "loss": 2.715,
+      "step": 13869
+    },
+    {
+      "epoch": 0.584,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0001912490925650962,
+      "loss": 3.1688,
+      "step": 13870
+    },
+    {
+      "epoch": 0.5840421052631579,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00019121629384412142,
+      "loss": 3.3224,
+      "step": 13871
+    },
+    {
+      "epoch": 0.5840842105263158,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001911834961941166,
+      "loss": 2.9578,
+      "step": 13872
+    },
+    {
+      "epoch": 0.5841263157894737,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001911506996156794,
+      "loss": 3.3652,
+      "step": 13873
+    },
+    {
+      "epoch": 0.5841684210526316,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00019111790410940733,
+      "loss": 3.2564,
+      "step": 13874
+    },
+    {
+      "epoch": 0.5842105263157895,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00019108510967589783,
+      "loss": 3.3644,
+      "step": 13875
+    },
+    {
+      "epoch": 0.5842526315789474,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00019105231631574843,
+      "loss": 3.3257,
+      "step": 13876
+    },
+    {
+      "epoch": 0.5842947368421053,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00019101952402955652,
+      "loss": 2.8879,
+      "step": 13877
+    },
+    {
+      "epoch": 0.5843368421052632,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00019098673281791969,
+      "loss": 3.1453,
+      "step": 13878
+    },
+    {
+      "epoch": 0.584378947368421,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0001909539426814351,
+      "loss": 3.2128,
+      "step": 13879
+    },
+    {
+      "epoch": 0.584421052631579,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00019092115362070037,
+      "loss": 3.2869,
+      "step": 13880
+    },
+    {
+      "epoch": 0.5844631578947368,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00019088836563631287,
+      "loss": 3.3193,
+      "step": 13881
+    },
+    {
+      "epoch": 0.5845052631578947,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0001908555787288698,
+      "loss": 2.8254,
+      "step": 13882
+    },
+    {
+      "epoch": 0.5845473684210526,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0001908227928989687,
+      "loss": 3.1346,
+      "step": 13883
+    },
+    {
+      "epoch": 0.5845894736842105,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00019079000814720669,
+      "loss": 2.9287,
+      "step": 13884
+    },
+    {
+      "epoch": 0.5846315789473684,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001907572244741812,
+      "loss": 3.2555,
+      "step": 13885
+    },
+    {
+      "epoch": 0.5846736842105263,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00019072444188048942,
+      "loss": 3.1583,
+      "step": 13886
+    },
+    {
+      "epoch": 0.5847157894736842,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00019069166036672866,
+      "loss": 3.4248,
+      "step": 13887
+    },
+    {
+      "epoch": 0.5847578947368421,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00019065887993349628,
+      "loss": 3.4637,
+      "step": 13888
+    },
+    {
+      "epoch": 0.5848,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00019062610058138927,
+      "loss": 2.6711,
+      "step": 13889
+    },
+    {
+      "epoch": 0.5848421052631579,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00019059332231100506,
+      "loss": 3.3968,
+      "step": 13890
+    },
+    {
+      "epoch": 0.5848842105263158,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0001905605451229406,
+      "loss": 3.2266,
+      "step": 13891
+    },
+    {
+      "epoch": 0.5849263157894737,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00019052776901779323,
+      "loss": 3.2517,
+      "step": 13892
+    },
+    {
+      "epoch": 0.5849684210526316,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00019049499399616,
+      "loss": 2.68,
+      "step": 13893
+    },
+    {
+      "epoch": 0.5850105263157894,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00019046222005863808,
+      "loss": 3.0779,
+      "step": 13894
+    },
+    {
+      "epoch": 0.5850526315789474,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00019042944720582455,
+      "loss": 2.8453,
+      "step": 13895
+    },
+    {
+      "epoch": 0.5850947368421052,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00019039667543831648,
+      "loss": 3.3318,
+      "step": 13896
+    },
+    {
+      "epoch": 0.5851368421052632,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00019036390475671107,
+      "loss": 3.3269,
+      "step": 13897
+    },
+    {
+      "epoch": 0.585178947368421,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0001903311351616051,
+      "loss": 3.3125,
+      "step": 13898
+    },
+    {
+      "epoch": 0.585221052631579,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00019029836665359587,
+      "loss": 3.172,
+      "step": 13899
+    },
+    {
+      "epoch": 0.5852631578947368,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00019026559923328014,
+      "loss": 2.8946,
+      "step": 13900
+    },
+    {
+      "epoch": 0.5853052631578948,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00019023283290125497,
+      "loss": 3.0234,
+      "step": 13901
+    },
+    {
+      "epoch": 0.5853473684210526,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0001902000676581175,
+      "loss": 2.5206,
+      "step": 13902
+    },
+    {
+      "epoch": 0.5853894736842106,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001901673035044644,
+      "loss": 2.8556,
+      "step": 13903
+    },
+    {
+      "epoch": 0.5854315789473684,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019013454044089286,
+      "loss": 3.3084,
+      "step": 13904
+    },
+    {
+      "epoch": 0.5854736842105264,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0001901017784679996,
+      "loss": 3.2,
+      "step": 13905
+    },
+    {
+      "epoch": 0.5855157894736842,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019006901758638147,
+      "loss": 3.023,
+      "step": 13906
+    },
+    {
+      "epoch": 0.585557894736842,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00019003625779663558,
+      "loss": 3.0901,
+      "step": 13907
+    },
+    {
+      "epoch": 0.5856,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00019000349909935853,
+      "loss": 3.022,
+      "step": 13908
+    },
+    {
+      "epoch": 0.5856421052631579,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00018997074149514732,
+      "loss": 2.5702,
+      "step": 13909
+    },
+    {
+      "epoch": 0.5856842105263158,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00018993798498459854,
+      "loss": 3.3927,
+      "step": 13910
+    },
+    {
+      "epoch": 0.5857263157894737,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0001899052295683092,
+      "loss": 3.46,
+      "step": 13911
+    },
+    {
+      "epoch": 0.5857684210526316,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00018987247524687587,
+      "loss": 3.0356,
+      "step": 13912
+    },
+    {
+      "epoch": 0.5858105263157894,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001898397220208954,
+      "loss": 3.3943,
+      "step": 13913
+    },
+    {
+      "epoch": 0.5858526315789474,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00018980696989096463,
+      "loss": 3.3657,
+      "step": 13914
+    },
+    {
+      "epoch": 0.5858947368421052,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00018977421885768003,
+      "loss": 3.5954,
+      "step": 13915
+    },
+    {
+      "epoch": 0.5859368421052632,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00018974146892163855,
+      "loss": 3.2644,
+      "step": 13916
+    },
+    {
+      "epoch": 0.585978947368421,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00018970872008343653,
+      "loss": 3.1405,
+      "step": 13917
+    },
+    {
+      "epoch": 0.586021052631579,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0001896759723436709,
+      "loss": 3.1546,
+      "step": 13918
+    },
+    {
+      "epoch": 0.5860631578947368,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0001896432257029381,
+      "loss": 3.055,
+      "step": 13919
+    },
+    {
+      "epoch": 0.5861052631578947,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0001896104801618348,
+      "loss": 3.0267,
+      "step": 13920
+    },
+    {
+      "epoch": 0.5861473684210526,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00018957773572095773,
+      "loss": 3.4306,
+      "step": 13921
+    },
+    {
+      "epoch": 0.5861894736842105,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001895449923809032,
+      "loss": 3.4683,
+      "step": 13922
+    },
+    {
+      "epoch": 0.5862315789473684,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00018951225014226798,
+      "loss": 3.274,
+      "step": 13923
+    },
+    {
+      "epoch": 0.5862736842105263,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0001894795090056484,
+      "loss": 2.6348,
+      "step": 13924
+    },
+    {
+      "epoch": 0.5863157894736842,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00018944676897164115,
+      "loss": 3.2133,
+      "step": 13925
+    },
+    {
+      "epoch": 0.5863578947368421,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0001894140300408425,
+      "loss": 3.1667,
+      "step": 13926
+    },
+    {
+      "epoch": 0.5864,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0001893812922138491,
+      "loss": 2.7082,
+      "step": 13927
+    },
+    {
+      "epoch": 0.5864421052631579,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0001893485554912573,
+      "loss": 2.8815,
+      "step": 13928
+    },
+    {
+      "epoch": 0.5864842105263158,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00018931581987366355,
+      "loss": 3.5537,
+      "step": 13929
+    },
+    {
+      "epoch": 0.5865263157894737,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00018928308536166438,
+      "loss": 3.0808,
+      "step": 13930
+    },
+    {
+      "epoch": 0.5865684210526316,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00018925035195585594,
+      "loss": 3.0665,
+      "step": 13931
+    },
+    {
+      "epoch": 0.5866105263157895,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00018921761965683478,
+      "loss": 2.887,
+      "step": 13932
+    },
+    {
+      "epoch": 0.5866526315789473,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0001891848884651971,
+      "loss": 2.6216,
+      "step": 13933
+    },
+    {
+      "epoch": 0.5866947368421053,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00018915215838153932,
+      "loss": 2.9539,
+      "step": 13934
+    },
+    {
+      "epoch": 0.5867368421052631,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00018911942940645773,
+      "loss": 2.6452,
+      "step": 13935
+    },
+    {
+      "epoch": 0.5867789473684211,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00018908670154054862,
+      "loss": 3.0001,
+      "step": 13936
+    },
+    {
+      "epoch": 0.5868210526315789,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00018905397478440823,
+      "loss": 2.9021,
+      "step": 13937
+    },
+    {
+      "epoch": 0.5868631578947369,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00018902124913863278,
+      "loss": 3.0226,
+      "step": 13938
+    },
+    {
+      "epoch": 0.5869052631578947,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001889885246038186,
+      "loss": 2.8644,
+      "step": 13939
+    },
+    {
+      "epoch": 0.5869473684210527,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00018895580118056174,
+      "loss": 2.5821,
+      "step": 13940
+    },
+    {
+      "epoch": 0.5869894736842105,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00018892307886945846,
+      "loss": 2.953,
+      "step": 13941
+    },
+    {
+      "epoch": 0.5870315789473685,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.000188890357671105,
+      "loss": 3.3639,
+      "step": 13942
+    },
+    {
+      "epoch": 0.5870736842105263,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00018885763758609732,
+      "loss": 3.2949,
+      "step": 13943
+    },
+    {
+      "epoch": 0.5871157894736843,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0001888249186150317,
+      "loss": 2.6449,
+      "step": 13944
+    },
+    {
+      "epoch": 0.5871578947368421,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00018879220075850416,
+      "loss": 2.9494,
+      "step": 13945
+    },
+    {
+      "epoch": 0.5872,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00018875948401711086,
+      "loss": 3.1981,
+      "step": 13946
+    },
+    {
+      "epoch": 0.5872421052631579,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00018872676839144773,
+      "loss": 3.3015,
+      "step": 13947
+    },
+    {
+      "epoch": 0.5872842105263157,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0001886940538821109,
+      "loss": 3.6655,
+      "step": 13948
+    },
+    {
+      "epoch": 0.5873263157894737,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00018866134048969646,
+      "loss": 3.2806,
+      "step": 13949
+    },
+    {
+      "epoch": 0.5873684210526315,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00018862862821480023,
+      "loss": 3.1046,
+      "step": 13950
+    },
+    {
+      "epoch": 0.5874105263157895,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00018859591705801832,
+      "loss": 3.1942,
+      "step": 13951
+    },
+    {
+      "epoch": 0.5874526315789473,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00018856320701994667,
+      "loss": 3.3292,
+      "step": 13952
+    },
+    {
+      "epoch": 0.5874947368421053,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00018853049810118116,
+      "loss": 3.3935,
+      "step": 13953
+    },
+    {
+      "epoch": 0.5875368421052631,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0001884977903023178,
+      "loss": 2.9382,
+      "step": 13954
+    },
+    {
+      "epoch": 0.5875789473684211,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00018846508362395237,
+      "loss": 3.0085,
+      "step": 13955
+    },
+    {
+      "epoch": 0.5876210526315789,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00018843237806668095,
+      "loss": 2.9271,
+      "step": 13956
+    },
+    {
+      "epoch": 0.5876631578947369,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00018839967363109913,
+      "loss": 3.158,
+      "step": 13957
+    },
+    {
+      "epoch": 0.5877052631578947,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.000188366970317803,
+      "loss": 2.9561,
+      "step": 13958
+    },
+    {
+      "epoch": 0.5877473684210527,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00018833426812738812,
+      "loss": 2.8339,
+      "step": 13959
+    },
+    {
+      "epoch": 0.5877894736842105,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00018830156706045045,
+      "loss": 2.9765,
+      "step": 13960
+    },
+    {
+      "epoch": 0.5878315789473684,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00018826886711758578,
+      "loss": 2.9063,
+      "step": 13961
+    },
+    {
+      "epoch": 0.5878736842105263,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00018823616829938977,
+      "loss": 2.951,
+      "step": 13962
+    },
+    {
+      "epoch": 0.5879157894736842,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0001882034706064583,
+      "loss": 2.8705,
+      "step": 13963
+    },
+    {
+      "epoch": 0.5879578947368421,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00018817077403938693,
+      "loss": 3.7721,
+      "step": 13964
+    },
+    {
+      "epoch": 0.588,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00018813807859877147,
+      "loss": 3.269,
+      "step": 13965
+    },
+    {
+      "epoch": 0.5880421052631579,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00018810538428520743,
+      "loss": 3.1864,
+      "step": 13966
+    },
+    {
+      "epoch": 0.5880842105263158,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00018807269109929057,
+      "loss": 3.3164,
+      "step": 13967
+    },
+    {
+      "epoch": 0.5881263157894737,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001880399990416166,
+      "loss": 3.0805,
+      "step": 13968
+    },
+    {
+      "epoch": 0.5881684210526316,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.000188007308112781,
+      "loss": 2.9917,
+      "step": 13969
+    },
+    {
+      "epoch": 0.5882105263157895,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00018797461831337943,
+      "loss": 2.8359,
+      "step": 13970
+    },
+    {
+      "epoch": 0.5882526315789474,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00018794192964400735,
+      "loss": 3.1658,
+      "step": 13971
+    },
+    {
+      "epoch": 0.5882947368421053,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00018790924210526056,
+      "loss": 3.0243,
+      "step": 13972
+    },
+    {
+      "epoch": 0.5883368421052632,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00018787655569773426,
+      "loss": 3.083,
+      "step": 13973
+    },
+    {
+      "epoch": 0.588378947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00018784387042202415,
+      "loss": 3.137,
+      "step": 13974
+    },
+    {
+      "epoch": 0.588421052631579,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001878111862787258,
+      "loss": 2.7689,
+      "step": 13975
+    },
+    {
+      "epoch": 0.5884631578947368,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00018777850326843448,
+      "loss": 2.8201,
+      "step": 13976
+    },
+    {
+      "epoch": 0.5885052631578948,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00018774582139174576,
+      "loss": 3.4679,
+      "step": 13977
+    },
+    {
+      "epoch": 0.5885473684210526,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.000187713140649255,
+      "loss": 3.1368,
+      "step": 13978
+    },
+    {
+      "epoch": 0.5885894736842106,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00018768046104155768,
+      "loss": 3.3511,
+      "step": 13979
+    },
+    {
+      "epoch": 0.5886315789473684,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00018764778256924908,
+      "loss": 2.6194,
+      "step": 13980
+    },
+    {
+      "epoch": 0.5886736842105263,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00018761510523292459,
+      "loss": 3.1724,
+      "step": 13981
+    },
+    {
+      "epoch": 0.5887157894736842,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00018758242903317972,
+      "loss": 3.2743,
+      "step": 13982
+    },
+    {
+      "epoch": 0.5887578947368421,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00018754975397060951,
+      "loss": 3.3447,
+      "step": 13983
+    },
+    {
+      "epoch": 0.5888,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00018751708004580955,
+      "loss": 3.3171,
+      "step": 13984
+    },
+    {
+      "epoch": 0.5888421052631579,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00018748440725937485,
+      "loss": 3.805,
+      "step": 13985
+    },
+    {
+      "epoch": 0.5888842105263158,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00018745173561190087,
+      "loss": 3.2026,
+      "step": 13986
+    },
+    {
+      "epoch": 0.5889263157894736,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0001874190651039827,
+      "loss": 3.4327,
+      "step": 13987
+    },
+    {
+      "epoch": 0.5889684210526316,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00018738639573621563,
+      "loss": 2.8129,
+      "step": 13988
+    },
+    {
+      "epoch": 0.5890105263157894,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.000187353727509195,
+      "loss": 3.4215,
+      "step": 13989
+    },
+    {
+      "epoch": 0.5890526315789474,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00018732106042351572,
+      "loss": 2.9689,
+      "step": 13990
+    },
+    {
+      "epoch": 0.5890947368421052,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001872883944797732,
+      "loss": 3.0947,
+      "step": 13991
+    },
+    {
+      "epoch": 0.5891368421052632,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0001872557296785623,
+      "loss": 2.9188,
+      "step": 13992
+    },
+    {
+      "epoch": 0.589178947368421,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001872230660204784,
+      "loss": 3.483,
+      "step": 13993
+    },
+    {
+      "epoch": 0.589221052631579,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00018719040350611638,
+      "loss": 2.8715,
+      "step": 13994
+    },
+    {
+      "epoch": 0.5892631578947368,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00018715774213607145,
+      "loss": 2.9801,
+      "step": 13995
+    },
+    {
+      "epoch": 0.5893052631578948,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00018712508191093865,
+      "loss": 3.3026,
+      "step": 13996
+    },
+    {
+      "epoch": 0.5893473684210526,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00018709242283131294,
+      "loss": 3.444,
+      "step": 13997
+    },
+    {
+      "epoch": 0.5893894736842106,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00018705976489778947,
+      "loss": 3.209,
+      "step": 13998
+    },
+    {
+      "epoch": 0.5894315789473684,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.000187027108110963,
+      "loss": 3.3482,
+      "step": 13999
+    },
+    {
+      "epoch": 0.5894736842105263,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00018699445247142866,
+      "loss": 3.333,
+      "step": 14000
+    },
+    {
+      "epoch": 0.5895157894736842,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00018696179797978146,
+      "loss": 3.0912,
+      "step": 14001
+    },
+    {
+      "epoch": 0.5895578947368421,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00018692914463661615,
+      "loss": 3.3171,
+      "step": 14002
+    },
+    {
+      "epoch": 0.5896,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00018689649244252774,
+      "loss": 3.2943,
+      "step": 14003
+    },
+    {
+      "epoch": 0.5896421052631579,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00018686384139811105,
+      "loss": 2.6557,
+      "step": 14004
+    },
+    {
+      "epoch": 0.5896842105263158,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0001868311915039611,
+      "loss": 3.171,
+      "step": 14005
+    },
+    {
+      "epoch": 0.5897263157894737,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00018679854276067252,
+      "loss": 3.1507,
+      "step": 14006
+    },
+    {
+      "epoch": 0.5897684210526316,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0001867658951688402,
+      "loss": 2.9782,
+      "step": 14007
+    },
+    {
+      "epoch": 0.5898105263157895,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00018673324872905912,
+      "loss": 2.9733,
+      "step": 14008
+    },
+    {
+      "epoch": 0.5898526315789474,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0001867006034419238,
+      "loss": 3.087,
+      "step": 14009
+    },
+    {
+      "epoch": 0.5898947368421052,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00018666795930802915,
+      "loss": 3.3937,
+      "step": 14010
+    },
+    {
+      "epoch": 0.5899368421052632,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00018663531632796989,
+      "loss": 2.9664,
+      "step": 14011
+    },
+    {
+      "epoch": 0.589978947368421,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00018660267450234066,
+      "loss": 3.0335,
+      "step": 14012
+    },
+    {
+      "epoch": 0.590021052631579,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0001865700338317362,
+      "loss": 2.774,
+      "step": 14013
+    },
+    {
+      "epoch": 0.5900631578947368,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001865373943167512,
+      "loss": 3.3908,
+      "step": 14014
+    },
+    {
+      "epoch": 0.5901052631578947,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00018650475595798042,
+      "loss": 2.8827,
+      "step": 14015
+    },
+    {
+      "epoch": 0.5901473684210526,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00018647211875601828,
+      "loss": 2.8737,
+      "step": 14016
+    },
+    {
+      "epoch": 0.5901894736842105,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00018643948271145958,
+      "loss": 2.3997,
+      "step": 14017
+    },
+    {
+      "epoch": 0.5902315789473684,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0001864068478248987,
+      "loss": 3.5256,
+      "step": 14018
+    },
+    {
+      "epoch": 0.5902736842105263,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00018637421409693045,
+      "loss": 2.7102,
+      "step": 14019
+    },
+    {
+      "epoch": 0.5903157894736842,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00018634158152814917,
+      "loss": 3.1982,
+      "step": 14020
+    },
+    {
+      "epoch": 0.5903578947368421,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00018630895011914947,
+      "loss": 2.8601,
+      "step": 14021
+    },
+    {
+      "epoch": 0.5904,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00018627631987052598,
+      "loss": 3.4766,
+      "step": 14022
+    },
+    {
+      "epoch": 0.5904421052631579,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00018624369078287296,
+      "loss": 2.8024,
+      "step": 14023
+    },
+    {
+      "epoch": 0.5904842105263158,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0001862110628567851,
+      "loss": 2.7287,
+      "step": 14024
+    },
+    {
+      "epoch": 0.5905263157894737,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00018617843609285663,
+      "loss": 3.0883,
+      "step": 14025
+    },
+    {
+      "epoch": 0.5905684210526316,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00018614581049168218,
+      "loss": 3.0341,
+      "step": 14026
+    },
+    {
+      "epoch": 0.5906105263157895,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00018611318605385585,
+      "loss": 3.2233,
+      "step": 14027
+    },
+    {
+      "epoch": 0.5906526315789473,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0001860805627799723,
+      "loss": 3.2316,
+      "step": 14028
+    },
+    {
+      "epoch": 0.5906947368421053,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00018604794067062581,
+      "loss": 3.2186,
+      "step": 14029
+    },
+    {
+      "epoch": 0.5907368421052631,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00018601531972641064,
+      "loss": 3.1209,
+      "step": 14030
+    },
+    {
+      "epoch": 0.5907789473684211,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00018598269994792127,
+      "loss": 3.4186,
+      "step": 14031
+    },
+    {
+      "epoch": 0.5908210526315789,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00018595008133575182,
+      "loss": 2.9376,
+      "step": 14032
+    },
+    {
+      "epoch": 0.5908631578947369,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0001859174638904967,
+      "loss": 3.4776,
+      "step": 14033
+    },
+    {
+      "epoch": 0.5909052631578947,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00018588484761275,
+      "loss": 2.7281,
+      "step": 14034
+    },
+    {
+      "epoch": 0.5909473684210527,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00018585223250310606,
+      "loss": 3.1854,
+      "step": 14035
+    },
+    {
+      "epoch": 0.5909894736842105,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001858196185621591,
+      "loss": 2.9249,
+      "step": 14036
+    },
+    {
+      "epoch": 0.5910315789473685,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00018578700579050322,
+      "loss": 3.1483,
+      "step": 14037
+    },
+    {
+      "epoch": 0.5910736842105263,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0001857543941887328,
+      "loss": 3.0011,
+      "step": 14038
+    },
+    {
+      "epoch": 0.5911157894736843,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0001857217837574417,
+      "loss": 3.1335,
+      "step": 14039
+    },
+    {
+      "epoch": 0.5911578947368421,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0001856891744972243,
+      "loss": 3.5151,
+      "step": 14040
+    },
+    {
+      "epoch": 0.5912,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00018565656640867447,
+      "loss": 3.0257,
+      "step": 14041
+    },
+    {
+      "epoch": 0.5912421052631579,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00018562395949238636,
+      "loss": 3.2532,
+      "step": 14042
+    },
+    {
+      "epoch": 0.5912842105263157,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001855913537489542,
+      "loss": 3.2271,
+      "step": 14043
+    },
+    {
+      "epoch": 0.5913263157894737,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001855587491789718,
+      "loss": 3.1459,
+      "step": 14044
+    },
+    {
+      "epoch": 0.5913684210526315,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00018552614578303328,
+      "loss": 2.7192,
+      "step": 14045
+    },
+    {
+      "epoch": 0.5914105263157895,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00018549354356173266,
+      "loss": 2.8455,
+      "step": 14046
+    },
+    {
+      "epoch": 0.5914526315789473,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00018546094251566393,
+      "loss": 2.9592,
+      "step": 14047
+    },
+    {
+      "epoch": 0.5914947368421053,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00018542834264542092,
+      "loss": 3.1317,
+      "step": 14048
+    },
+    {
+      "epoch": 0.5915368421052631,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0001853957439515976,
+      "loss": 3.6409,
+      "step": 14049
+    },
+    {
+      "epoch": 0.5915789473684211,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001853631464347881,
+      "loss": 2.8056,
+      "step": 14050
+    },
+    {
+      "epoch": 0.5916210526315789,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001853305500955859,
+      "loss": 3.4607,
+      "step": 14051
+    },
+    {
+      "epoch": 0.5916631578947369,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00018529795493458522,
+      "loss": 3.1425,
+      "step": 14052
+    },
+    {
+      "epoch": 0.5917052631578947,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00018526536095237974,
+      "loss": 3.2086,
+      "step": 14053
+    },
+    {
+      "epoch": 0.5917473684210526,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00018523276814956337,
+      "loss": 3.8333,
+      "step": 14054
+    },
+    {
+      "epoch": 0.5917894736842105,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00018520017652672984,
+      "loss": 3.4118,
+      "step": 14055
+    },
+    {
+      "epoch": 0.5918315789473684,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00018516758608447292,
+      "loss": 2.9002,
+      "step": 14056
+    },
+    {
+      "epoch": 0.5918736842105263,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00018513499682338652,
+      "loss": 2.9023,
+      "step": 14057
+    },
+    {
+      "epoch": 0.5919157894736842,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00018510240874406415,
+      "loss": 3.3603,
+      "step": 14058
+    },
+    {
+      "epoch": 0.5919578947368421,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00018506982184709976,
+      "loss": 3.4403,
+      "step": 14059
+    },
+    {
+      "epoch": 0.592,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001850372361330868,
+      "loss": 3.006,
+      "step": 14060
+    },
+    {
+      "epoch": 0.5920421052631579,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001850046516026191,
+      "loss": 3.1424,
+      "step": 14061
+    },
+    {
+      "epoch": 0.5920842105263158,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0001849720682562903,
+      "loss": 2.6964,
+      "step": 14062
+    },
+    {
+      "epoch": 0.5921263157894737,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00018493948609469397,
+      "loss": 3.0775,
+      "step": 14063
+    },
+    {
+      "epoch": 0.5921684210526316,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00018490690511842388,
+      "loss": 2.7162,
+      "step": 14064
+    },
+    {
+      "epoch": 0.5922105263157895,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00018487432532807336,
+      "loss": 3.132,
+      "step": 14065
+    },
+    {
+      "epoch": 0.5922526315789474,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00018484174672423625,
+      "loss": 2.975,
+      "step": 14066
+    },
+    {
+      "epoch": 0.5922947368421053,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00018480916930750587,
+      "loss": 3.3118,
+      "step": 14067
+    },
+    {
+      "epoch": 0.5923368421052632,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00018477659307847578,
+      "loss": 3.5919,
+      "step": 14068
+    },
+    {
+      "epoch": 0.592378947368421,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00018474401803773966,
+      "loss": 3.3821,
+      "step": 14069
+    },
+    {
+      "epoch": 0.592421052631579,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00018471144418589073,
+      "loss": 3.0293,
+      "step": 14070
+    },
+    {
+      "epoch": 0.5924631578947368,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00018467887152352264,
+      "loss": 3.1185,
+      "step": 14071
+    },
+    {
+      "epoch": 0.5925052631578948,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00018464630005122874,
+      "loss": 3.027,
+      "step": 14072
+    },
+    {
+      "epoch": 0.5925473684210526,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00018461372976960252,
+      "loss": 3.3056,
+      "step": 14073
+    },
+    {
+      "epoch": 0.5925894736842106,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00018458116067923724,
+      "loss": 2.9311,
+      "step": 14074
+    },
+    {
+      "epoch": 0.5926315789473684,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00018454859278072637,
+      "loss": 3.242,
+      "step": 14075
+    },
+    {
+      "epoch": 0.5926736842105264,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0001845160260746633,
+      "loss": 3.3032,
+      "step": 14076
+    },
+    {
+      "epoch": 0.5927157894736842,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00018448346056164122,
+      "loss": 3.4215,
+      "step": 14077
+    },
+    {
+      "epoch": 0.5927578947368422,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00018445089624225352,
+      "loss": 2.9424,
+      "step": 14078
+    },
+    {
+      "epoch": 0.5928,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001844183331170935,
+      "loss": 2.9477,
+      "step": 14079
+    },
+    {
+      "epoch": 0.592842105263158,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00018438577118675442,
+      "loss": 3.2466,
+      "step": 14080
+    },
+    {
+      "epoch": 0.5928842105263158,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00018435321045182944,
+      "loss": 2.7994,
+      "step": 14081
+    },
+    {
+      "epoch": 0.5929263157894736,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0001843206509129118,
+      "loss": 2.945,
+      "step": 14082
+    },
+    {
+      "epoch": 0.5929684210526316,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00018428809257059486,
+      "loss": 2.9767,
+      "step": 14083
+    },
+    {
+      "epoch": 0.5930105263157894,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00018425553542547154,
+      "loss": 3.0225,
+      "step": 14084
+    },
+    {
+      "epoch": 0.5930526315789474,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00018422297947813526,
+      "loss": 3.1694,
+      "step": 14085
+    },
+    {
+      "epoch": 0.5930947368421052,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00018419042472917884,
+      "loss": 2.7224,
+      "step": 14086
+    },
+    {
+      "epoch": 0.5931368421052632,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00018415787117919568,
+      "loss": 2.8289,
+      "step": 14087
+    },
+    {
+      "epoch": 0.593178947368421,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00018412531882877863,
+      "loss": 3.5368,
+      "step": 14088
+    },
+    {
+      "epoch": 0.593221052631579,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00018409276767852094,
+      "loss": 3.0918,
+      "step": 14089
+    },
+    {
+      "epoch": 0.5932631578947368,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00018406021772901567,
+      "loss": 2.9173,
+      "step": 14090
+    },
+    {
+      "epoch": 0.5933052631578948,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0001840276689808556,
+      "loss": 3.6685,
+      "step": 14091
+    },
+    {
+      "epoch": 0.5933473684210526,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00018399512143463403,
+      "loss": 3.4018,
+      "step": 14092
+    },
+    {
+      "epoch": 0.5933894736842106,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00018396257509094372,
+      "loss": 3.2447,
+      "step": 14093
+    },
+    {
+      "epoch": 0.5934315789473684,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00018393002995037772,
+      "loss": 3.2238,
+      "step": 14094
+    },
+    {
+      "epoch": 0.5934736842105263,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0001838974860135289,
+      "loss": 2.8675,
+      "step": 14095
+    },
+    {
+      "epoch": 0.5935157894736842,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00018386494328099024,
+      "loss": 2.8125,
+      "step": 14096
+    },
+    {
+      "epoch": 0.5935578947368421,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00018383240175335464,
+      "loss": 2.7587,
+      "step": 14097
+    },
+    {
+      "epoch": 0.5936,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0001837998614312149,
+      "loss": 2.9503,
+      "step": 14098
+    },
+    {
+      "epoch": 0.5936421052631579,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00018376732231516397,
+      "loss": 2.7711,
+      "step": 14099
+    },
+    {
+      "epoch": 0.5936842105263158,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0001837347844057945,
+      "loss": 2.7605,
+      "step": 14100
+    },
+    {
+      "epoch": 0.5937263157894737,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00018370224770369943,
+      "loss": 3.0096,
+      "step": 14101
+    },
+    {
+      "epoch": 0.5937684210526316,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0001836697122094716,
+      "loss": 3.3535,
+      "step": 14102
+    },
+    {
+      "epoch": 0.5938105263157895,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00018363717792370358,
+      "loss": 3.5111,
+      "step": 14103
+    },
+    {
+      "epoch": 0.5938526315789474,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00018360464484698825,
+      "loss": 3.2068,
+      "step": 14104
+    },
+    {
+      "epoch": 0.5938947368421053,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0001835721129799182,
+      "loss": 3.2665,
+      "step": 14105
+    },
+    {
+      "epoch": 0.5939368421052632,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00018353958232308638,
+      "loss": 3.4318,
+      "step": 14106
+    },
+    {
+      "epoch": 0.593978947368421,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00018350705287708513,
+      "loss": 3.0891,
+      "step": 14107
+    },
+    {
+      "epoch": 0.5940210526315789,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00018347452464250724,
+      "loss": 3.3439,
+      "step": 14108
+    },
+    {
+      "epoch": 0.5940631578947368,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00018344199761994546,
+      "loss": 2.7869,
+      "step": 14109
+    },
+    {
+      "epoch": 0.5941052631578947,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0001834094718099922,
+      "loss": 3.3625,
+      "step": 14110
+    },
+    {
+      "epoch": 0.5941473684210526,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00018337694721324018,
+      "loss": 3.3111,
+      "step": 14111
+    },
+    {
+      "epoch": 0.5941894736842105,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00018334442383028185,
+      "loss": 3.2014,
+      "step": 14112
+    },
+    {
+      "epoch": 0.5942315789473684,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00018331190166170983,
+      "loss": 2.9869,
+      "step": 14113
+    },
+    {
+      "epoch": 0.5942736842105263,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00018327938070811657,
+      "loss": 3.1566,
+      "step": 14114
+    },
+    {
+      "epoch": 0.5943157894736842,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00018324686097009456,
+      "loss": 3.0855,
+      "step": 14115
+    },
+    {
+      "epoch": 0.5943578947368421,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00018321434244823645,
+      "loss": 2.6995,
+      "step": 14116
+    },
+    {
+      "epoch": 0.5944,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018318182514313444,
+      "loss": 3.3796,
+      "step": 14117
+    },
+    {
+      "epoch": 0.5944421052631579,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00018314930905538119,
+      "loss": 2.9373,
+      "step": 14118
+    },
+    {
+      "epoch": 0.5944842105263158,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00018311679418556883,
+      "loss": 2.9837,
+      "step": 14119
+    },
+    {
+      "epoch": 0.5945263157894737,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00018308428053429,
+      "loss": 2.9227,
+      "step": 14120
+    },
+    {
+      "epoch": 0.5945684210526315,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00018305176810213687,
+      "loss": 2.7834,
+      "step": 14121
+    },
+    {
+      "epoch": 0.5946105263157895,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00018301925688970187,
+      "loss": 2.7403,
+      "step": 14122
+    },
+    {
+      "epoch": 0.5946526315789473,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00018298674689757745,
+      "loss": 2.8032,
+      "step": 14123
+    },
+    {
+      "epoch": 0.5946947368421053,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00018295423812635559,
+      "loss": 2.873,
+      "step": 14124
+    },
+    {
+      "epoch": 0.5947368421052631,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0001829217305766289,
+      "loss": 3.2694,
+      "step": 14125
+    },
+    {
+      "epoch": 0.5947789473684211,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00018288922424898932,
+      "loss": 3.4606,
+      "step": 14126
+    },
+    {
+      "epoch": 0.5948210526315789,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0001828567191440293,
+      "loss": 2.9249,
+      "step": 14127
+    },
+    {
+      "epoch": 0.5948631578947369,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00018282421526234093,
+      "loss": 2.5941,
+      "step": 14128
+    },
+    {
+      "epoch": 0.5949052631578947,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00018279171260451643,
+      "loss": 3.0786,
+      "step": 14129
+    },
+    {
+      "epoch": 0.5949473684210527,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.000182759211171148,
+      "loss": 3.0047,
+      "step": 14130
+    },
+    {
+      "epoch": 0.5949894736842105,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00018272671096282767,
+      "loss": 3.0047,
+      "step": 14131
+    },
+    {
+      "epoch": 0.5950315789473685,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00018269421198014775,
+      "loss": 2.9246,
+      "step": 14132
+    },
+    {
+      "epoch": 0.5950736842105263,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0001826617142237001,
+      "loss": 3.2797,
+      "step": 14133
+    },
+    {
+      "epoch": 0.5951157894736843,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00018262921769407697,
+      "loss": 3.1566,
+      "step": 14134
+    },
+    {
+      "epoch": 0.5951578947368421,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00018259672239187024,
+      "loss": 2.2889,
+      "step": 14135
+    },
+    {
+      "epoch": 0.5952,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00018256422831767206,
+      "loss": 3.2808,
+      "step": 14136
+    },
+    {
+      "epoch": 0.5952421052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0001825317354720744,
+      "loss": 2.9301,
+      "step": 14137
+    },
+    {
+      "epoch": 0.5952842105263157,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00018249924385566928,
+      "loss": 3.2682,
+      "step": 14138
+    },
+    {
+      "epoch": 0.5953263157894737,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00018246675346904868,
+      "loss": 3.2212,
+      "step": 14139
+    },
+    {
+      "epoch": 0.5953684210526315,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00018243426431280437,
+      "loss": 3.192,
+      "step": 14140
+    },
+    {
+      "epoch": 0.5954105263157895,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00018240177638752852,
+      "loss": 3.4714,
+      "step": 14141
+    },
+    {
+      "epoch": 0.5954526315789473,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00018236928969381274,
+      "loss": 3.3307,
+      "step": 14142
+    },
+    {
+      "epoch": 0.5954947368421053,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00018233680423224906,
+      "loss": 3.0748,
+      "step": 14143
+    },
+    {
+      "epoch": 0.5955368421052631,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001823043200034294,
+      "loss": 3.4481,
+      "step": 14144
+    },
+    {
+      "epoch": 0.5955789473684211,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001822718370079454,
+      "loss": 2.7483,
+      "step": 14145
+    },
+    {
+      "epoch": 0.5956210526315789,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00018223935524638897,
+      "loss": 2.9262,
+      "step": 14146
+    },
+    {
+      "epoch": 0.5956631578947369,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00018220687471935185,
+      "loss": 2.971,
+      "step": 14147
+    },
+    {
+      "epoch": 0.5957052631578947,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0001821743954274258,
+      "loss": 3.0555,
+      "step": 14148
+    },
+    {
+      "epoch": 0.5957473684210526,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00018214191737120271,
+      "loss": 2.9495,
+      "step": 14149
+    },
+    {
+      "epoch": 0.5957894736842105,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00018210944055127405,
+      "loss": 2.8069,
+      "step": 14150
+    },
+    {
+      "epoch": 0.5958315789473684,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001820769649682317,
+      "loss": 3.209,
+      "step": 14151
+    },
+    {
+      "epoch": 0.5958736842105263,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001820444906226671,
+      "loss": 3.172,
+      "step": 14152
+    },
+    {
+      "epoch": 0.5959157894736842,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00018201201751517215,
+      "loss": 3.3699,
+      "step": 14153
+    },
+    {
+      "epoch": 0.5959578947368421,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00018197954564633828,
+      "loss": 2.7971,
+      "step": 14154
+    },
+    {
+      "epoch": 0.596,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00018194707501675722,
+      "loss": 3.0257,
+      "step": 14155
+    },
+    {
+      "epoch": 0.5960421052631579,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001819146056270205,
+      "loss": 3.318,
+      "step": 14156
+    },
+    {
+      "epoch": 0.5960842105263158,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00018188213747771963,
+      "loss": 2.9438,
+      "step": 14157
+    },
+    {
+      "epoch": 0.5961263157894737,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00018184967056944623,
+      "loss": 3.0333,
+      "step": 14158
+    },
+    {
+      "epoch": 0.5961684210526316,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00018181720490279173,
+      "loss": 3.274,
+      "step": 14159
+    },
+    {
+      "epoch": 0.5962105263157895,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0001817847404783477,
+      "loss": 3.0457,
+      "step": 14160
+    },
+    {
+      "epoch": 0.5962526315789474,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00018175227729670546,
+      "loss": 2.7163,
+      "step": 14161
+    },
+    {
+      "epoch": 0.5962947368421052,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00018171981535845656,
+      "loss": 3.4618,
+      "step": 14162
+    },
+    {
+      "epoch": 0.5963368421052632,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0001816873546641924,
+      "loss": 3.4997,
+      "step": 14163
+    },
+    {
+      "epoch": 0.596378947368421,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00018165489521450434,
+      "loss": 3.3044,
+      "step": 14164
+    },
+    {
+      "epoch": 0.596421052631579,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001816224370099839,
+      "loss": 3.0795,
+      "step": 14165
+    },
+    {
+      "epoch": 0.5964631578947368,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00018158998005122224,
+      "loss": 3.0373,
+      "step": 14166
+    },
+    {
+      "epoch": 0.5965052631578948,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00018155752433881084,
+      "loss": 3.0506,
+      "step": 14167
+    },
+    {
+      "epoch": 0.5965473684210526,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00018152506987334082,
+      "loss": 3.4617,
+      "step": 14168
+    },
+    {
+      "epoch": 0.5965894736842106,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00018149261665540356,
+      "loss": 2.9665,
+      "step": 14169
+    },
+    {
+      "epoch": 0.5966315789473684,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0001814601646855904,
+      "loss": 3.1314,
+      "step": 14170
+    },
+    {
+      "epoch": 0.5966736842105264,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00018142771396449254,
+      "loss": 3.0709,
+      "step": 14171
+    },
+    {
+      "epoch": 0.5967157894736842,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00018139526449270112,
+      "loss": 3.292,
+      "step": 14172
+    },
+    {
+      "epoch": 0.5967578947368422,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00018136281627080734,
+      "loss": 3.0269,
+      "step": 14173
+    },
+    {
+      "epoch": 0.5968,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0001813303692994025,
+      "loss": 3.3575,
+      "step": 14174
+    },
+    {
+      "epoch": 0.5968421052631578,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00018129792357907755,
+      "loss": 3.191,
+      "step": 14175
+    },
+    {
+      "epoch": 0.5968842105263158,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0001812654791104237,
+      "loss": 3.1151,
+      "step": 14176
+    },
+    {
+      "epoch": 0.5969263157894736,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00018123303589403218,
+      "loss": 2.8759,
+      "step": 14177
+    },
+    {
+      "epoch": 0.5969684210526316,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00018120059393049383,
+      "loss": 2.8783,
+      "step": 14178
+    },
+    {
+      "epoch": 0.5970105263157894,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001811681532203999,
+      "loss": 3.4004,
+      "step": 14179
+    },
+    {
+      "epoch": 0.5970526315789474,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00018113571376434127,
+      "loss": 3.1573,
+      "step": 14180
+    },
+    {
+      "epoch": 0.5970947368421052,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0001811032755629091,
+      "loss": 2.8983,
+      "step": 14181
+    },
+    {
+      "epoch": 0.5971368421052632,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00018107083861669426,
+      "loss": 3.4293,
+      "step": 14182
+    },
+    {
+      "epoch": 0.597178947368421,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0001810384029262877,
+      "loss": 3.0085,
+      "step": 14183
+    },
+    {
+      "epoch": 0.597221052631579,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00018100596849228055,
+      "loss": 3.0895,
+      "step": 14184
+    },
+    {
+      "epoch": 0.5972631578947368,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00018097353531526346,
+      "loss": 3.0738,
+      "step": 14185
+    },
+    {
+      "epoch": 0.5973052631578948,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00018094110339582757,
+      "loss": 3.2546,
+      "step": 14186
+    },
+    {
+      "epoch": 0.5973473684210526,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00018090867273456353,
+      "loss": 3.2452,
+      "step": 14187
+    },
+    {
+      "epoch": 0.5973894736842106,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018087624333206233,
+      "loss": 3.3662,
+      "step": 14188
+    },
+    {
+      "epoch": 0.5974315789473684,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00018084381518891475,
+      "loss": 2.9624,
+      "step": 14189
+    },
+    {
+      "epoch": 0.5974736842105263,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00018081138830571152,
+      "loss": 3.3758,
+      "step": 14190
+    },
+    {
+      "epoch": 0.5975157894736842,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0001807789626830437,
+      "loss": 3.2225,
+      "step": 14191
+    },
+    {
+      "epoch": 0.5975578947368421,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00018074653832150168,
+      "loss": 3.0898,
+      "step": 14192
+    },
+    {
+      "epoch": 0.5976,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00018071411522167652,
+      "loss": 3.403,
+      "step": 14193
+    },
+    {
+      "epoch": 0.5976421052631579,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00018068169338415863,
+      "loss": 2.8114,
+      "step": 14194
+    },
+    {
+      "epoch": 0.5976842105263158,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00018064927280953891,
+      "loss": 3.4682,
+      "step": 14195
+    },
+    {
+      "epoch": 0.5977263157894737,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001806168534984079,
+      "loss": 3.3691,
+      "step": 14196
+    },
+    {
+      "epoch": 0.5977684210526316,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00018058443545135628,
+      "loss": 2.8175,
+      "step": 14197
+    },
+    {
+      "epoch": 0.5978105263157895,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00018055201866897482,
+      "loss": 3.1892,
+      "step": 14198
+    },
+    {
+      "epoch": 0.5978526315789474,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00018051960315185384,
+      "loss": 3.3063,
+      "step": 14199
+    },
+    {
+      "epoch": 0.5978947368421053,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00018048718890058418,
+      "loss": 3.1537,
+      "step": 14200
+    },
+    {
+      "epoch": 0.5979368421052632,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00018045477591575616,
+      "loss": 3.1764,
+      "step": 14201
+    },
+    {
+      "epoch": 0.597978947368421,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001804223641979604,
+      "loss": 3.09,
+      "step": 14202
+    },
+    {
+      "epoch": 0.5980210526315789,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00018038995374778754,
+      "loss": 3.3642,
+      "step": 14203
+    },
+    {
+      "epoch": 0.5980631578947369,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0001803575445658278,
+      "loss": 3.1946,
+      "step": 14204
+    },
+    {
+      "epoch": 0.5981052631578947,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00018032513665267185,
+      "loss": 2.9327,
+      "step": 14205
+    },
+    {
+      "epoch": 0.5981473684210526,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00018029273000891,
+      "loss": 3.1497,
+      "step": 14206
+    },
+    {
+      "epoch": 0.5981894736842105,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00018026032463513278,
+      "loss": 2.7971,
+      "step": 14207
+    },
+    {
+      "epoch": 0.5982315789473684,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00018022792053193042,
+      "loss": 2.958,
+      "step": 14208
+    },
+    {
+      "epoch": 0.5982736842105263,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00018019551769989336,
+      "loss": 3.3272,
+      "step": 14209
+    },
+    {
+      "epoch": 0.5983157894736842,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00018016311613961206,
+      "loss": 2.4949,
+      "step": 14210
+    },
+    {
+      "epoch": 0.5983578947368421,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00018013071585167663,
+      "loss": 2.9405,
+      "step": 14211
+    },
+    {
+      "epoch": 0.5984,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0001800983168366775,
+      "loss": 3.173,
+      "step": 14212
+    },
+    {
+      "epoch": 0.5984421052631579,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00018006591909520487,
+      "loss": 3.0187,
+      "step": 14213
+    },
+    {
+      "epoch": 0.5984842105263158,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00018003352262784914,
+      "loss": 3.0295,
+      "step": 14214
+    },
+    {
+      "epoch": 0.5985263157894737,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0001800011274352003,
+      "loss": 3.4337,
+      "step": 14215
+    },
+    {
+      "epoch": 0.5985684210526315,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00017996873351784864,
+      "loss": 2.989,
+      "step": 14216
+    },
+    {
+      "epoch": 0.5986105263157895,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00017993634087638455,
+      "loss": 3.1541,
+      "step": 14217
+    },
+    {
+      "epoch": 0.5986526315789473,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00017990394951139784,
+      "loss": 3.0405,
+      "step": 14218
+    },
+    {
+      "epoch": 0.5986947368421053,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00017987155942347894,
+      "loss": 3.3537,
+      "step": 14219
+    },
+    {
+      "epoch": 0.5987368421052631,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00017983917061321775,
+      "loss": 3.0125,
+      "step": 14220
+    },
+    {
+      "epoch": 0.5987789473684211,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00017980678308120447,
+      "loss": 3.3385,
+      "step": 14221
+    },
+    {
+      "epoch": 0.5988210526315789,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00017977439682802905,
+      "loss": 3.2128,
+      "step": 14222
+    },
+    {
+      "epoch": 0.5988631578947369,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00017974201185428162,
+      "loss": 3.344,
+      "step": 14223
+    },
+    {
+      "epoch": 0.5989052631578947,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001797096281605523,
+      "loss": 3.4571,
+      "step": 14224
+    },
+    {
+      "epoch": 0.5989473684210527,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00017967724574743084,
+      "loss": 3.192,
+      "step": 14225
+    },
+    {
+      "epoch": 0.5989894736842105,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00017964486461550746,
+      "loss": 3.0205,
+      "step": 14226
+    },
+    {
+      "epoch": 0.5990315789473685,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0001796124847653719,
+      "loss": 3.0975,
+      "step": 14227
+    },
+    {
+      "epoch": 0.5990736842105263,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00017958010619761416,
+      "loss": 2.7796,
+      "step": 14228
+    },
+    {
+      "epoch": 0.5991157894736842,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0001795477289128242,
+      "loss": 2.9655,
+      "step": 14229
+    },
+    {
+      "epoch": 0.5991578947368421,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00017951535291159176,
+      "loss": 2.9468,
+      "step": 14230
+    },
+    {
+      "epoch": 0.5992,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00017948297819450685,
+      "loss": 2.9989,
+      "step": 14231
+    },
+    {
+      "epoch": 0.5992421052631579,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00017945060476215913,
+      "loss": 3.1823,
+      "step": 14232
+    },
+    {
+      "epoch": 0.5992842105263158,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00017941823261513864,
+      "loss": 3.3522,
+      "step": 14233
+    },
+    {
+      "epoch": 0.5993263157894737,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00017938586175403492,
+      "loss": 3.2526,
+      "step": 14234
+    },
+    {
+      "epoch": 0.5993684210526316,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00017935349217943793,
+      "loss": 3.1417,
+      "step": 14235
+    },
+    {
+      "epoch": 0.5994105263157895,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00017932112389193716,
+      "loss": 2.8684,
+      "step": 14236
+    },
+    {
+      "epoch": 0.5994526315789473,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00017928875689212254,
+      "loss": 3.1602,
+      "step": 14237
+    },
+    {
+      "epoch": 0.5994947368421053,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00017925639118058367,
+      "loss": 3.1476,
+      "step": 14238
+    },
+    {
+      "epoch": 0.5995368421052631,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00017922402675791022,
+      "loss": 2.9712,
+      "step": 14239
+    },
+    {
+      "epoch": 0.5995789473684211,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00017919166362469194,
+      "loss": 3.0147,
+      "step": 14240
+    },
+    {
+      "epoch": 0.5996210526315789,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00017915930178151823,
+      "loss": 3.1166,
+      "step": 14241
+    },
+    {
+      "epoch": 0.5996631578947368,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00017912694122897894,
+      "loss": 2.8975,
+      "step": 14242
+    },
+    {
+      "epoch": 0.5997052631578947,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001790945819676634,
+      "loss": 3.0567,
+      "step": 14243
+    },
+    {
+      "epoch": 0.5997473684210526,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00017906222399816123,
+      "loss": 3.4858,
+      "step": 14244
+    },
+    {
+      "epoch": 0.5997894736842105,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00017902986732106218,
+      "loss": 2.8606,
+      "step": 14245
+    },
+    {
+      "epoch": 0.5998315789473684,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0001789975119369554,
+      "loss": 3.4899,
+      "step": 14246
+    },
+    {
+      "epoch": 0.5998736842105263,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0001789651578464306,
+      "loss": 3.1858,
+      "step": 14247
+    },
+    {
+      "epoch": 0.5999157894736842,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00017893280505007706,
+      "loss": 2.8699,
+      "step": 14248
+    },
+    {
+      "epoch": 0.5999578947368421,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00017890045354848434,
+      "loss": 3.1221,
+      "step": 14249
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0001788681033422419,
+      "loss": 2.7105,
+      "step": 14250
+    },
+    {
+      "epoch": 0.6000421052631579,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00017883575443193898,
+      "loss": 3.1565,
+      "step": 14251
+    },
+    {
+      "epoch": 0.6000842105263158,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0001788034068181651,
+      "loss": 3.221,
+      "step": 14252
+    },
+    {
+      "epoch": 0.6001263157894737,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0001787710605015094,
+      "loss": 3.2445,
+      "step": 14253
+    },
+    {
+      "epoch": 0.6001684210526316,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00017873871548256125,
+      "loss": 3.1422,
+      "step": 14254
+    },
+    {
+      "epoch": 0.6002105263157895,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00017870637176191,
+      "loss": 3.4716,
+      "step": 14255
+    },
+    {
+      "epoch": 0.6002526315789474,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00017867402934014493,
+      "loss": 2.8835,
+      "step": 14256
+    },
+    {
+      "epoch": 0.6002947368421052,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00017864168821785526,
+      "loss": 3.517,
+      "step": 14257
+    },
+    {
+      "epoch": 0.6003368421052632,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0001786093483956301,
+      "loss": 2.9048,
+      "step": 14258
+    },
+    {
+      "epoch": 0.600378947368421,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00017857700987405887,
+      "loss": 2.8359,
+      "step": 14259
+    },
+    {
+      "epoch": 0.600421052631579,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00017854467265373048,
+      "loss": 3.2564,
+      "step": 14260
+    },
+    {
+      "epoch": 0.6004631578947368,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0001785123367352343,
+      "loss": 3.2064,
+      "step": 14261
+    },
+    {
+      "epoch": 0.6005052631578948,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00017848000211915924,
+      "loss": 2.9936,
+      "step": 14262
+    },
+    {
+      "epoch": 0.6005473684210526,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00017844766880609453,
+      "loss": 3.3684,
+      "step": 14263
+    },
+    {
+      "epoch": 0.6005894736842106,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00017841533679662923,
+      "loss": 3.0844,
+      "step": 14264
+    },
+    {
+      "epoch": 0.6006315789473684,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001783830060913524,
+      "loss": 3.1581,
+      "step": 14265
+    },
+    {
+      "epoch": 0.6006736842105264,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0001783506766908531,
+      "loss": 2.9269,
+      "step": 14266
+    },
+    {
+      "epoch": 0.6007157894736842,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00017831834859572022,
+      "loss": 3.4592,
+      "step": 14267
+    },
+    {
+      "epoch": 0.6007578947368422,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001782860218065429,
+      "loss": 3.322,
+      "step": 14268
+    },
+    {
+      "epoch": 0.6008,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00017825369632390987,
+      "loss": 2.8601,
+      "step": 14269
+    },
+    {
+      "epoch": 0.6008421052631578,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00017822137214841023,
+      "loss": 2.9781,
+      "step": 14270
+    },
+    {
+      "epoch": 0.6008842105263158,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0001781890492806329,
+      "loss": 2.7142,
+      "step": 14271
+    },
+    {
+      "epoch": 0.6009263157894736,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001781567277211667,
+      "loss": 3.2353,
+      "step": 14272
+    },
+    {
+      "epoch": 0.6009684210526316,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00017812440747060052,
+      "loss": 2.9058,
+      "step": 14273
+    },
+    {
+      "epoch": 0.6010105263157894,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00017809208852952312,
+      "loss": 3.1842,
+      "step": 14274
+    },
+    {
+      "epoch": 0.6010526315789474,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00017805977089852348,
+      "loss": 3.3471,
+      "step": 14275
+    },
+    {
+      "epoch": 0.6010947368421052,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00017802745457819023,
+      "loss": 3.102,
+      "step": 14276
+    },
+    {
+      "epoch": 0.6011368421052632,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00017799513956911215,
+      "loss": 2.7908,
+      "step": 14277
+    },
+    {
+      "epoch": 0.601178947368421,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00017796282587187814,
+      "loss": 3.1129,
+      "step": 14278
+    },
+    {
+      "epoch": 0.601221052631579,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0001779305134870767,
+      "loss": 3.0738,
+      "step": 14279
+    },
+    {
+      "epoch": 0.6012631578947368,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0001778982024152967,
+      "loss": 2.9089,
+      "step": 14280
+    },
+    {
+      "epoch": 0.6013052631578948,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00017786589265712666,
+      "loss": 3.0549,
+      "step": 14281
+    },
+    {
+      "epoch": 0.6013473684210526,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00017783358421315538,
+      "loss": 2.6364,
+      "step": 14282
+    },
+    {
+      "epoch": 0.6013894736842105,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0001778012770839713,
+      "loss": 2.8628,
+      "step": 14283
+    },
+    {
+      "epoch": 0.6014315789473684,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00017776897127016317,
+      "loss": 3.1314,
+      "step": 14284
+    },
+    {
+      "epoch": 0.6014736842105263,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00017773666677231963,
+      "loss": 3.1663,
+      "step": 14285
+    },
+    {
+      "epoch": 0.6015157894736842,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00017770436359102897,
+      "loss": 3.3835,
+      "step": 14286
+    },
+    {
+      "epoch": 0.6015578947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0001776720617268799,
+      "loss": 3.1776,
+      "step": 14287
+    },
+    {
+      "epoch": 0.6016,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00017763976118046094,
+      "loss": 3.2507,
+      "step": 14288
+    },
+    {
+      "epoch": 0.6016421052631579,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00017760746195236043,
+      "loss": 3.4731,
+      "step": 14289
+    },
+    {
+      "epoch": 0.6016842105263158,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00017757516404316693,
+      "loss": 2.7301,
+      "step": 14290
+    },
+    {
+      "epoch": 0.6017263157894737,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00017754286745346883,
+      "loss": 3.5825,
+      "step": 14291
+    },
+    {
+      "epoch": 0.6017684210526316,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00017751057218385463,
+      "loss": 3.1869,
+      "step": 14292
+    },
+    {
+      "epoch": 0.6018105263157895,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00017747827823491252,
+      "loss": 2.8144,
+      "step": 14293
+    },
+    {
+      "epoch": 0.6018526315789474,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00017744598560723114,
+      "loss": 3.1766,
+      "step": 14294
+    },
+    {
+      "epoch": 0.6018947368421053,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00017741369430139848,
+      "loss": 3.2164,
+      "step": 14295
+    },
+    {
+      "epoch": 0.6019368421052631,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0001773814043180031,
+      "loss": 3.6473,
+      "step": 14296
+    },
+    {
+      "epoch": 0.6019789473684211,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00017734911565763317,
+      "loss": 3.3819,
+      "step": 14297
+    },
+    {
+      "epoch": 0.6020210526315789,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00017731682832087698,
+      "loss": 3.141,
+      "step": 14298
+    },
+    {
+      "epoch": 0.6020631578947369,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00017728454230832286,
+      "loss": 2.921,
+      "step": 14299
+    },
+    {
+      "epoch": 0.6021052631578947,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001772522576205589,
+      "loss": 2.9058,
+      "step": 14300
+    },
+    {
+      "epoch": 0.6021473684210527,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00017721997425817338,
+      "loss": 3.1806,
+      "step": 14301
+    },
+    {
+      "epoch": 0.6021894736842105,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00017718769222175435,
+      "loss": 3.222,
+      "step": 14302
+    },
+    {
+      "epoch": 0.6022315789473685,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00017715541151188996,
+      "loss": 3.1514,
+      "step": 14303
+    },
+    {
+      "epoch": 0.6022736842105263,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00017712313212916853,
+      "loss": 2.8569,
+      "step": 14304
+    },
+    {
+      "epoch": 0.6023157894736842,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00017709085407417792,
+      "loss": 3.4402,
+      "step": 14305
+    },
+    {
+      "epoch": 0.6023578947368421,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0001770585773475063,
+      "loss": 3.3504,
+      "step": 14306
+    },
+    {
+      "epoch": 0.6024,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00017702630194974168,
+      "loss": 3.1977,
+      "step": 14307
+    },
+    {
+      "epoch": 0.6024421052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00017699402788147218,
+      "loss": 3.2827,
+      "step": 14308
+    },
+    {
+      "epoch": 0.6024842105263158,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0001769617551432856,
+      "loss": 3.2181,
+      "step": 14309
+    },
+    {
+      "epoch": 0.6025263157894737,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00017692948373577004,
+      "loss": 3.0767,
+      "step": 14310
+    },
+    {
+      "epoch": 0.6025684210526315,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001768972136595135,
+      "loss": 2.9875,
+      "step": 14311
+    },
+    {
+      "epoch": 0.6026105263157895,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0001768649449151038,
+      "loss": 3.1384,
+      "step": 14312
+    },
+    {
+      "epoch": 0.6026526315789473,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00017683267750312888,
+      "loss": 3.0977,
+      "step": 14313
+    },
+    {
+      "epoch": 0.6026947368421053,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00017680041142417656,
+      "loss": 3.4675,
+      "step": 14314
+    },
+    {
+      "epoch": 0.6027368421052631,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00017676814667883485,
+      "loss": 3.0921,
+      "step": 14315
+    },
+    {
+      "epoch": 0.6027789473684211,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00017673588326769135,
+      "loss": 3.4372,
+      "step": 14316
+    },
+    {
+      "epoch": 0.6028210526315789,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00017670362119133397,
+      "loss": 3.1493,
+      "step": 14317
+    },
+    {
+      "epoch": 0.6028631578947369,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0001766713604503506,
+      "loss": 3.4734,
+      "step": 14318
+    },
+    {
+      "epoch": 0.6029052631578947,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0001766391010453288,
+      "loss": 3.4723,
+      "step": 14319
+    },
+    {
+      "epoch": 0.6029473684210527,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00017660684297685647,
+      "loss": 3.2245,
+      "step": 14320
+    },
+    {
+      "epoch": 0.6029894736842105,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00017657458624552113,
+      "loss": 3.166,
+      "step": 14321
+    },
+    {
+      "epoch": 0.6030315789473685,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001765423308519106,
+      "loss": 3.1929,
+      "step": 14322
+    },
+    {
+      "epoch": 0.6030736842105263,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00017651007679661247,
+      "loss": 3.1554,
+      "step": 14323
+    },
+    {
+      "epoch": 0.6031157894736842,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00017647782408021436,
+      "loss": 3.4395,
+      "step": 14324
+    },
+    {
+      "epoch": 0.6031578947368421,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00017644557270330403,
+      "loss": 3.1518,
+      "step": 14325
+    },
+    {
+      "epoch": 0.6032,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0001764133226664688,
+      "loss": 3.0421,
+      "step": 14326
+    },
+    {
+      "epoch": 0.6032421052631579,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00017638107397029655,
+      "loss": 3.3119,
+      "step": 14327
+    },
+    {
+      "epoch": 0.6032842105263158,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00017634882661537448,
+      "loss": 3.4664,
+      "step": 14328
+    },
+    {
+      "epoch": 0.6033263157894737,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0001763165806022903,
+      "loss": 3.1714,
+      "step": 14329
+    },
+    {
+      "epoch": 0.6033684210526316,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001762843359316314,
+      "loss": 3.3592,
+      "step": 14330
+    },
+    {
+      "epoch": 0.6034105263157895,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00017625209260398536,
+      "loss": 2.9916,
+      "step": 14331
+    },
+    {
+      "epoch": 0.6034526315789474,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00017621985061993953,
+      "loss": 3.6419,
+      "step": 14332
+    },
+    {
+      "epoch": 0.6034947368421053,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00017618760998008127,
+      "loss": 3.3885,
+      "step": 14333
+    },
+    {
+      "epoch": 0.6035368421052631,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00017615537068499816,
+      "loss": 3.0778,
+      "step": 14334
+    },
+    {
+      "epoch": 0.6035789473684211,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0001761231327352773,
+      "loss": 2.5775,
+      "step": 14335
+    },
+    {
+      "epoch": 0.603621052631579,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00017609089613150627,
+      "loss": 3.2107,
+      "step": 14336
+    },
+    {
+      "epoch": 0.6036631578947368,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001760586608742722,
+      "loss": 3.0254,
+      "step": 14337
+    },
+    {
+      "epoch": 0.6037052631578947,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00017602642696416243,
+      "loss": 3.3475,
+      "step": 14338
+    },
+    {
+      "epoch": 0.6037473684210526,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00017599419440176428,
+      "loss": 3.2766,
+      "step": 14339
+    },
+    {
+      "epoch": 0.6037894736842105,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001759619631876649,
+      "loss": 3.3514,
+      "step": 14340
+    },
+    {
+      "epoch": 0.6038315789473684,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00017592973332245172,
+      "loss": 3.0353,
+      "step": 14341
+    },
+    {
+      "epoch": 0.6038736842105263,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0001758975048067116,
+      "loss": 3.0156,
+      "step": 14342
+    },
+    {
+      "epoch": 0.6039157894736842,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.000175865277641032,
+      "loss": 2.9943,
+      "step": 14343
+    },
+    {
+      "epoch": 0.6039578947368421,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00017583305182599984,
+      "loss": 3.2597,
+      "step": 14344
+    },
+    {
+      "epoch": 0.604,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00017580082736220234,
+      "loss": 3.1668,
+      "step": 14345
+    },
+    {
+      "epoch": 0.6040421052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00017576860425022662,
+      "loss": 2.7546,
+      "step": 14346
+    },
+    {
+      "epoch": 0.6040842105263158,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00017573638249065974,
+      "loss": 3.0973,
+      "step": 14347
+    },
+    {
+      "epoch": 0.6041263157894737,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001757041620840887,
+      "loss": 2.8762,
+      "step": 14348
+    },
+    {
+      "epoch": 0.6041684210526316,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00017567194303110047,
+      "loss": 3.2078,
+      "step": 14349
+    },
+    {
+      "epoch": 0.6042105263157894,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00017563972533228211,
+      "loss": 3.1475,
+      "step": 14350
+    },
+    {
+      "epoch": 0.6042526315789474,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0001756075089882207,
+      "loss": 3.1275,
+      "step": 14351
+    },
+    {
+      "epoch": 0.6042947368421052,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00017557529399950294,
+      "loss": 3.2045,
+      "step": 14352
+    },
+    {
+      "epoch": 0.6043368421052632,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.000175543080366716,
+      "loss": 3.0218,
+      "step": 14353
+    },
+    {
+      "epoch": 0.604378947368421,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00017551086809044656,
+      "loss": 2.8475,
+      "step": 14354
+    },
+    {
+      "epoch": 0.604421052631579,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0001754786571712816,
+      "loss": 3.1377,
+      "step": 14355
+    },
+    {
+      "epoch": 0.6044631578947368,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00017544644760980794,
+      "loss": 3.4219,
+      "step": 14356
+    },
+    {
+      "epoch": 0.6045052631578948,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00017541423940661238,
+      "loss": 3.1462,
+      "step": 14357
+    },
+    {
+      "epoch": 0.6045473684210526,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00017538203256228186,
+      "loss": 3.2937,
+      "step": 14358
+    },
+    {
+      "epoch": 0.6045894736842106,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0001753498270774029,
+      "loss": 2.7347,
+      "step": 14359
+    },
+    {
+      "epoch": 0.6046315789473684,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00017531762295256253,
+      "loss": 2.7794,
+      "step": 14360
+    },
+    {
+      "epoch": 0.6046736842105264,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001752854201883472,
+      "loss": 3.3034,
+      "step": 14361
+    },
+    {
+      "epoch": 0.6047157894736842,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00017525321878534382,
+      "loss": 2.7878,
+      "step": 14362
+    },
+    {
+      "epoch": 0.604757894736842,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00017522101874413887,
+      "loss": 3.1952,
+      "step": 14363
+    },
+    {
+      "epoch": 0.6048,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00017518882006531912,
+      "loss": 3.2335,
+      "step": 14364
+    },
+    {
+      "epoch": 0.6048421052631578,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00017515662274947117,
+      "loss": 2.9054,
+      "step": 14365
+    },
+    {
+      "epoch": 0.6048842105263158,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0001751244267971816,
+      "loss": 2.7199,
+      "step": 14366
+    },
+    {
+      "epoch": 0.6049263157894736,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00017509223220903708,
+      "loss": 2.8467,
+      "step": 14367
+    },
+    {
+      "epoch": 0.6049684210526316,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.000175060038985624,
+      "loss": 3.2566,
+      "step": 14368
+    },
+    {
+      "epoch": 0.6050105263157894,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00017502784712752906,
+      "loss": 3.4888,
+      "step": 14369
+    },
+    {
+      "epoch": 0.6050526315789474,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00017499565663533852,
+      "loss": 2.6347,
+      "step": 14370
+    },
+    {
+      "epoch": 0.6050947368421052,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00017496346750963903,
+      "loss": 3.0466,
+      "step": 14371
+    },
+    {
+      "epoch": 0.6051368421052632,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00017493127975101703,
+      "loss": 3.6157,
+      "step": 14372
+    },
+    {
+      "epoch": 0.605178947368421,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00017489909336005887,
+      "loss": 3.4147,
+      "step": 14373
+    },
+    {
+      "epoch": 0.605221052631579,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00017486690833735108,
+      "loss": 3.1085,
+      "step": 14374
+    },
+    {
+      "epoch": 0.6052631578947368,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00017483472468347982,
+      "loss": 2.8484,
+      "step": 14375
+    },
+    {
+      "epoch": 0.6053052631578948,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0001748025423990317,
+      "loss": 3.1112,
+      "step": 14376
+    },
+    {
+      "epoch": 0.6053473684210526,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0001747703614845928,
+      "loss": 3.2114,
+      "step": 14377
+    },
+    {
+      "epoch": 0.6053894736842105,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00017473818194074951,
+      "loss": 2.9145,
+      "step": 14378
+    },
+    {
+      "epoch": 0.6054315789473684,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001747060037680882,
+      "loss": 3.224,
+      "step": 14379
+    },
+    {
+      "epoch": 0.6054736842105263,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00017467382696719497,
+      "loss": 3.0835,
+      "step": 14380
+    },
+    {
+      "epoch": 0.6055157894736842,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00017464165153865613,
+      "loss": 3.1289,
+      "step": 14381
+    },
+    {
+      "epoch": 0.6055578947368421,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00017460947748305783,
+      "loss": 3.0134,
+      "step": 14382
+    },
+    {
+      "epoch": 0.6056,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00017457730480098638,
+      "loss": 3.1964,
+      "step": 14383
+    },
+    {
+      "epoch": 0.6056421052631579,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00017454513349302766,
+      "loss": 3.1444,
+      "step": 14384
+    },
+    {
+      "epoch": 0.6056842105263158,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.000174512963559768,
+      "loss": 3.1645,
+      "step": 14385
+    },
+    {
+      "epoch": 0.6057263157894737,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00017448079500179353,
+      "loss": 2.9484,
+      "step": 14386
+    },
+    {
+      "epoch": 0.6057684210526316,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00017444862781969018,
+      "loss": 2.868,
+      "step": 14387
+    },
+    {
+      "epoch": 0.6058105263157895,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00017441646201404407,
+      "loss": 3.1535,
+      "step": 14388
+    },
+    {
+      "epoch": 0.6058526315789474,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0001743842975854412,
+      "loss": 2.8054,
+      "step": 14389
+    },
+    {
+      "epoch": 0.6058947368421053,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00017435213453446765,
+      "loss": 3.1087,
+      "step": 14390
+    },
+    {
+      "epoch": 0.6059368421052631,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00017431997286170922,
+      "loss": 3.1549,
+      "step": 14391
+    },
+    {
+      "epoch": 0.6059789473684211,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.000174287812567752,
+      "loss": 3.1221,
+      "step": 14392
+    },
+    {
+      "epoch": 0.6060210526315789,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00017425565365318196,
+      "loss": 2.8602,
+      "step": 14393
+    },
+    {
+      "epoch": 0.6060631578947369,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0001742234961185848,
+      "loss": 2.8558,
+      "step": 14394
+    },
+    {
+      "epoch": 0.6061052631578947,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001741913399645466,
+      "loss": 2.8347,
+      "step": 14395
+    },
+    {
+      "epoch": 0.6061473684210527,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0001741591851916531,
+      "loss": 3.4093,
+      "step": 14396
+    },
+    {
+      "epoch": 0.6061894736842105,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00017412703180049002,
+      "loss": 3.1927,
+      "step": 14397
+    },
+    {
+      "epoch": 0.6062315789473685,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0001740948797916434,
+      "loss": 3.0297,
+      "step": 14398
+    },
+    {
+      "epoch": 0.6062736842105263,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001740627291656988,
+      "loss": 3.3566,
+      "step": 14399
+    },
+    {
+      "epoch": 0.6063157894736843,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00017403057992324217,
+      "loss": 3.1013,
+      "step": 14400
+    },
+    {
+      "epoch": 0.6063578947368421,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00017399843206485898,
+      "loss": 3.3346,
+      "step": 14401
+    },
+    {
+      "epoch": 0.6064,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0001739662855911352,
+      "loss": 3.3243,
+      "step": 14402
+    },
+    {
+      "epoch": 0.6064421052631579,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00017393414050265627,
+      "loss": 3.4896,
+      "step": 14403
+    },
+    {
+      "epoch": 0.6064842105263157,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00017390199680000787,
+      "loss": 3.4679,
+      "step": 14404
+    },
+    {
+      "epoch": 0.6065263157894737,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0001738698544837758,
+      "loss": 2.815,
+      "step": 14405
+    },
+    {
+      "epoch": 0.6065684210526315,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00017383771355454548,
+      "loss": 3.3122,
+      "step": 14406
+    },
+    {
+      "epoch": 0.6066105263157895,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00017380557401290252,
+      "loss": 2.6552,
+      "step": 14407
+    },
+    {
+      "epoch": 0.6066526315789473,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00017377343585943245,
+      "loss": 3.1834,
+      "step": 14408
+    },
+    {
+      "epoch": 0.6066947368421053,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00017374129909472093,
+      "loss": 3.0601,
+      "step": 14409
+    },
+    {
+      "epoch": 0.6067368421052631,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00017370916371935323,
+      "loss": 2.8227,
+      "step": 14410
+    },
+    {
+      "epoch": 0.6067789473684211,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001736770297339149,
+      "loss": 3.3021,
+      "step": 14411
+    },
+    {
+      "epoch": 0.6068210526315789,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00017364489713899157,
+      "loss": 3.0529,
+      "step": 14412
+    },
+    {
+      "epoch": 0.6068631578947369,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00017361276593516833,
+      "loss": 2.9552,
+      "step": 14413
+    },
+    {
+      "epoch": 0.6069052631578947,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00017358063612303085,
+      "loss": 2.8452,
+      "step": 14414
+    },
+    {
+      "epoch": 0.6069473684210527,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00017354850770316433,
+      "loss": 3.105,
+      "step": 14415
+    },
+    {
+      "epoch": 0.6069894736842105,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00017351638067615428,
+      "loss": 3.0665,
+      "step": 14416
+    },
+    {
+      "epoch": 0.6070315789473684,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00017348425504258577,
+      "loss": 2.6447,
+      "step": 14417
+    },
+    {
+      "epoch": 0.6070736842105263,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00017345213080304425,
+      "loss": 3.4866,
+      "step": 14418
+    },
+    {
+      "epoch": 0.6071157894736842,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00017342000795811507,
+      "loss": 3.4621,
+      "step": 14419
+    },
+    {
+      "epoch": 0.6071578947368421,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00017338788650838328,
+      "loss": 3.1926,
+      "step": 14420
+    },
+    {
+      "epoch": 0.6072,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001733557664544343,
+      "loss": 2.6116,
+      "step": 14421
+    },
+    {
+      "epoch": 0.6072421052631579,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00017332364779685305,
+      "loss": 2.8597,
+      "step": 14422
+    },
+    {
+      "epoch": 0.6072842105263158,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00017329153053622493,
+      "loss": 2.5189,
+      "step": 14423
+    },
+    {
+      "epoch": 0.6073263157894737,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00017325941467313493,
+      "loss": 3.4193,
+      "step": 14424
+    },
+    {
+      "epoch": 0.6073684210526316,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00017322730020816825,
+      "loss": 2.8177,
+      "step": 14425
+    },
+    {
+      "epoch": 0.6074105263157895,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00017319518714191003,
+      "loss": 3.1547,
+      "step": 14426
+    },
+    {
+      "epoch": 0.6074526315789474,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00017316307547494514,
+      "loss": 2.8267,
+      "step": 14427
+    },
+    {
+      "epoch": 0.6074947368421053,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00017313096520785886,
+      "loss": 2.7881,
+      "step": 14428
+    },
+    {
+      "epoch": 0.6075368421052632,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00017309885634123602,
+      "loss": 3.4356,
+      "step": 14429
+    },
+    {
+      "epoch": 0.6075789473684211,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00017306674887566166,
+      "loss": 3.4469,
+      "step": 14430
+    },
+    {
+      "epoch": 0.607621052631579,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001730346428117207,
+      "loss": 3.0541,
+      "step": 14431
+    },
+    {
+      "epoch": 0.6076631578947368,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00017300253814999818,
+      "loss": 2.9103,
+      "step": 14432
+    },
+    {
+      "epoch": 0.6077052631578947,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00017297043489107895,
+      "loss": 3.3339,
+      "step": 14433
+    },
+    {
+      "epoch": 0.6077473684210526,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0001729383330355478,
+      "loss": 3.1711,
+      "step": 14434
+    },
+    {
+      "epoch": 0.6077894736842105,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001729062325839898,
+      "loss": 3.0314,
+      "step": 14435
+    },
+    {
+      "epoch": 0.6078315789473684,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00017287413353698957,
+      "loss": 3.287,
+      "step": 14436
+    },
+    {
+      "epoch": 0.6078736842105263,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00017284203589513212,
+      "loss": 3.3059,
+      "step": 14437
+    },
+    {
+      "epoch": 0.6079157894736842,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00017280993965900198,
+      "loss": 2.8258,
+      "step": 14438
+    },
+    {
+      "epoch": 0.6079578947368421,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00017277784482918408,
+      "loss": 3.0593,
+      "step": 14439
+    },
+    {
+      "epoch": 0.608,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00017274575140626317,
+      "loss": 2.933,
+      "step": 14440
+    },
+    {
+      "epoch": 0.6080421052631579,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00017271365939082383,
+      "loss": 3.0344,
+      "step": 14441
+    },
+    {
+      "epoch": 0.6080842105263158,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00017268156878345092,
+      "loss": 2.9966,
+      "step": 14442
+    },
+    {
+      "epoch": 0.6081263157894737,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00017264947958472887,
+      "loss": 3.2886,
+      "step": 14443
+    },
+    {
+      "epoch": 0.6081684210526316,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00017261739179524253,
+      "loss": 3.0775,
+      "step": 14444
+    },
+    {
+      "epoch": 0.6082105263157894,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00017258530541557627,
+      "loss": 3.5416,
+      "step": 14445
+    },
+    {
+      "epoch": 0.6082526315789474,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0001725532204463148,
+      "loss": 2.4764,
+      "step": 14446
+    },
+    {
+      "epoch": 0.6082947368421052,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00017252113688804273,
+      "loss": 2.9827,
+      "step": 14447
+    },
+    {
+      "epoch": 0.6083368421052632,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00017248905474134447,
+      "loss": 3.0291,
+      "step": 14448
+    },
+    {
+      "epoch": 0.608378947368421,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00017245697400680458,
+      "loss": 3.4183,
+      "step": 14449
+    },
+    {
+      "epoch": 0.608421052631579,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00017242489468500746,
+      "loss": 3.1488,
+      "step": 14450
+    },
+    {
+      "epoch": 0.6084631578947368,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00017239281677653762,
+      "loss": 2.9957,
+      "step": 14451
+    },
+    {
+      "epoch": 0.6085052631578948,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00017236074028197958,
+      "loss": 3.5276,
+      "step": 14452
+    },
+    {
+      "epoch": 0.6085473684210526,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0001723286652019175,
+      "loss": 3.3975,
+      "step": 14453
+    },
+    {
+      "epoch": 0.6085894736842106,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.000172296591536936,
+      "loss": 3.3564,
+      "step": 14454
+    },
+    {
+      "epoch": 0.6086315789473684,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00017226451928761921,
+      "loss": 3.3337,
+      "step": 14455
+    },
+    {
+      "epoch": 0.6086736842105264,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001722324484545516,
+      "loss": 3.3109,
+      "step": 14456
+    },
+    {
+      "epoch": 0.6087157894736842,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00017220037903831732,
+      "loss": 3.1215,
+      "step": 14457
+    },
+    {
+      "epoch": 0.608757894736842,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00017216831103950076,
+      "loss": 2.4311,
+      "step": 14458
+    },
+    {
+      "epoch": 0.6088,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0001721362444586862,
+      "loss": 3.2418,
+      "step": 14459
+    },
+    {
+      "epoch": 0.6088421052631579,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001721041792964577,
+      "loss": 3.4157,
+      "step": 14460
+    },
+    {
+      "epoch": 0.6088842105263158,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00017207211555339964,
+      "loss": 2.9562,
+      "step": 14461
+    },
+    {
+      "epoch": 0.6089263157894736,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0001720400532300959,
+      "loss": 2.8664,
+      "step": 14462
+    },
+    {
+      "epoch": 0.6089684210526316,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00017200799232713093,
+      "loss": 3.0115,
+      "step": 14463
+    },
+    {
+      "epoch": 0.6090105263157894,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00017197593284508866,
+      "loss": 3.2859,
+      "step": 14464
+    },
+    {
+      "epoch": 0.6090526315789474,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00017194387478455318,
+      "loss": 2.8892,
+      "step": 14465
+    },
+    {
+      "epoch": 0.6090947368421052,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00017191181814610863,
+      "loss": 3.1776,
+      "step": 14466
+    },
+    {
+      "epoch": 0.6091368421052632,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00017187976293033897,
+      "loss": 2.6483,
+      "step": 14467
+    },
+    {
+      "epoch": 0.609178947368421,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00017184770913782833,
+      "loss": 3.0819,
+      "step": 14468
+    },
+    {
+      "epoch": 0.609221052631579,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001718156567691605,
+      "loss": 3.0569,
+      "step": 14469
+    },
+    {
+      "epoch": 0.6092631578947368,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00017178360582491964,
+      "loss": 2.9479,
+      "step": 14470
+    },
+    {
+      "epoch": 0.6093052631578947,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0001717515563056895,
+      "loss": 3.2937,
+      "step": 14471
+    },
+    {
+      "epoch": 0.6093473684210526,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00017171950821205404,
+      "loss": 3.1747,
+      "step": 14472
+    },
+    {
+      "epoch": 0.6093894736842105,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00017168746154459726,
+      "loss": 3.1403,
+      "step": 14473
+    },
+    {
+      "epoch": 0.6094315789473684,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0001716554163039028,
+      "loss": 2.6385,
+      "step": 14474
+    },
+    {
+      "epoch": 0.6094736842105263,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00017162337249055476,
+      "loss": 3.3172,
+      "step": 14475
+    },
+    {
+      "epoch": 0.6095157894736842,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00017159133010513666,
+      "loss": 3.3779,
+      "step": 14476
+    },
+    {
+      "epoch": 0.6095578947368421,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00017155928914823249,
+      "loss": 3.1768,
+      "step": 14477
+    },
+    {
+      "epoch": 0.6096,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00017152724962042582,
+      "loss": 2.9544,
+      "step": 14478
+    },
+    {
+      "epoch": 0.6096421052631579,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00017149521152230046,
+      "loss": 3.2507,
+      "step": 14479
+    },
+    {
+      "epoch": 0.6096842105263158,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0001714631748544402,
+      "loss": 3.1608,
+      "step": 14480
+    },
+    {
+      "epoch": 0.6097263157894737,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00017143113961742853,
+      "loss": 2.7929,
+      "step": 14481
+    },
+    {
+      "epoch": 0.6097684210526316,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00017139910581184922,
+      "loss": 2.8111,
+      "step": 14482
+    },
+    {
+      "epoch": 0.6098105263157895,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00017136707343828582,
+      "loss": 3.2001,
+      "step": 14483
+    },
+    {
+      "epoch": 0.6098526315789474,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00017133504249732204,
+      "loss": 2.8702,
+      "step": 14484
+    },
+    {
+      "epoch": 0.6098947368421053,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00017130301298954126,
+      "loss": 2.9646,
+      "step": 14485
+    },
+    {
+      "epoch": 0.6099368421052631,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00017127098491552712,
+      "loss": 3.202,
+      "step": 14486
+    },
+    {
+      "epoch": 0.6099789473684211,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00017123895827586322,
+      "loss": 2.5163,
+      "step": 14487
+    },
+    {
+      "epoch": 0.6100210526315789,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00017120693307113286,
+      "loss": 2.7098,
+      "step": 14488
+    },
+    {
+      "epoch": 0.6100631578947369,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00017117490930191963,
+      "loss": 2.8221,
+      "step": 14489
+    },
+    {
+      "epoch": 0.6101052631578947,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0001711428869688069,
+      "loss": 3.1792,
+      "step": 14490
+    },
+    {
+      "epoch": 0.6101473684210527,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00017111086607237826,
+      "loss": 2.894,
+      "step": 14491
+    },
+    {
+      "epoch": 0.6101894736842105,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00017107884661321683,
+      "loss": 3.3188,
+      "step": 14492
+    },
+    {
+      "epoch": 0.6102315789473685,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00017104682859190605,
+      "loss": 2.8844,
+      "step": 14493
+    },
+    {
+      "epoch": 0.6102736842105263,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00017101481200902945,
+      "loss": 3.2214,
+      "step": 14494
+    },
+    {
+      "epoch": 0.6103157894736843,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00017098279686517,
+      "loss": 2.9975,
+      "step": 14495
+    },
+    {
+      "epoch": 0.6103578947368421,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00017095078316091132,
+      "loss": 2.7352,
+      "step": 14496
+    },
+    {
+      "epoch": 0.6104,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00017091877089683632,
+      "loss": 3.2231,
+      "step": 14497
+    },
+    {
+      "epoch": 0.6104421052631579,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00017088676007352843,
+      "loss": 3.0825,
+      "step": 14498
+    },
+    {
+      "epoch": 0.6104842105263157,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00017085475069157088,
+      "loss": 3.0315,
+      "step": 14499
+    },
+    {
+      "epoch": 0.6105263157894737,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00017082274275154674,
+      "loss": 3.0517,
+      "step": 14500
+    },
+    {
+      "epoch": 0.6105684210526315,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001707907362540393,
+      "loss": 3.0769,
+      "step": 14501
+    },
+    {
+      "epoch": 0.6106105263157895,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00017075873119963144,
+      "loss": 2.7907,
+      "step": 14502
+    },
+    {
+      "epoch": 0.6106526315789473,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001707267275889065,
+      "loss": 3.003,
+      "step": 14503
+    },
+    {
+      "epoch": 0.6106947368421053,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00017069472542244736,
+      "loss": 3.1957,
+      "step": 14504
+    },
+    {
+      "epoch": 0.6107368421052631,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00017066272470083713,
+      "loss": 3.1265,
+      "step": 14505
+    },
+    {
+      "epoch": 0.6107789473684211,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0001706307254246589,
+      "loss": 3.106,
+      "step": 14506
+    },
+    {
+      "epoch": 0.6108210526315789,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00017059872759449558,
+      "loss": 3.1876,
+      "step": 14507
+    },
+    {
+      "epoch": 0.6108631578947369,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00017056673121093016,
+      "loss": 3.278,
+      "step": 14508
+    },
+    {
+      "epoch": 0.6109052631578947,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00017053473627454553,
+      "loss": 3.056,
+      "step": 14509
+    },
+    {
+      "epoch": 0.6109473684210527,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001705027427859247,
+      "loss": 3.3199,
+      "step": 14510
+    },
+    {
+      "epoch": 0.6109894736842105,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00017047075074565045,
+      "loss": 3.2345,
+      "step": 14511
+    },
+    {
+      "epoch": 0.6110315789473684,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00017043876015430563,
+      "loss": 2.9218,
+      "step": 14512
+    },
+    {
+      "epoch": 0.6110736842105263,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00017040677101247322,
+      "loss": 2.8957,
+      "step": 14513
+    },
+    {
+      "epoch": 0.6111157894736842,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00017037478332073585,
+      "loss": 3.4303,
+      "step": 14514
+    },
+    {
+      "epoch": 0.6111578947368421,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0001703427970796764,
+      "loss": 3.5862,
+      "step": 14515
+    },
+    {
+      "epoch": 0.6112,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00017031081228987755,
+      "loss": 2.6488,
+      "step": 14516
+    },
+    {
+      "epoch": 0.6112421052631579,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00017027882895192219,
+      "loss": 2.9731,
+      "step": 14517
+    },
+    {
+      "epoch": 0.6112842105263158,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00017024684706639276,
+      "loss": 3.0858,
+      "step": 14518
+    },
+    {
+      "epoch": 0.6113263157894737,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00017021486663387208,
+      "loss": 3.0784,
+      "step": 14519
+    },
+    {
+      "epoch": 0.6113684210526316,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00017018288765494294,
+      "loss": 2.9581,
+      "step": 14520
+    },
+    {
+      "epoch": 0.6114105263157895,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00017015091013018762,
+      "loss": 2.9755,
+      "step": 14521
+    },
+    {
+      "epoch": 0.6114526315789474,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00017011893406018909,
+      "loss": 3.0269,
+      "step": 14522
+    },
+    {
+      "epoch": 0.6114947368421053,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00017008695944552955,
+      "loss": 2.6009,
+      "step": 14523
+    },
+    {
+      "epoch": 0.6115368421052632,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0001700549862867918,
+      "loss": 2.8842,
+      "step": 14524
+    },
+    {
+      "epoch": 0.611578947368421,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00017002301458455822,
+      "loss": 3.007,
+      "step": 14525
+    },
+    {
+      "epoch": 0.611621052631579,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00016999104433941133,
+      "loss": 3.1766,
+      "step": 14526
+    },
+    {
+      "epoch": 0.6116631578947368,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00016995907555193368,
+      "loss": 3.0891,
+      "step": 14527
+    },
+    {
+      "epoch": 0.6117052631578948,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00016992710822270757,
+      "loss": 3.0675,
+      "step": 14528
+    },
+    {
+      "epoch": 0.6117473684210526,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00016989514235231558,
+      "loss": 3.0202,
+      "step": 14529
+    },
+    {
+      "epoch": 0.6117894736842105,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00016986317794133985,
+      "loss": 2.9388,
+      "step": 14530
+    },
+    {
+      "epoch": 0.6118315789473684,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00016983121499036292,
+      "loss": 2.9369,
+      "step": 14531
+    },
+    {
+      "epoch": 0.6118736842105263,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.000169799253499967,
+      "loss": 2.9923,
+      "step": 14532
+    },
+    {
+      "epoch": 0.6119157894736842,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00016976729347073445,
+      "loss": 3.2774,
+      "step": 14533
+    },
+    {
+      "epoch": 0.6119578947368421,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00016973533490324764,
+      "loss": 3.3321,
+      "step": 14534
+    },
+    {
+      "epoch": 0.612,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00016970337779808863,
+      "loss": 3.2423,
+      "step": 14535
+    },
+    {
+      "epoch": 0.6120421052631579,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0001696714221558398,
+      "loss": 2.9124,
+      "step": 14536
+    },
+    {
+      "epoch": 0.6120842105263158,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00016963946797708324,
+      "loss": 3.2309,
+      "step": 14537
+    },
+    {
+      "epoch": 0.6121263157894736,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001696075152624012,
+      "loss": 2.9732,
+      "step": 14538
+    },
+    {
+      "epoch": 0.6121684210526316,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0001695755640123757,
+      "loss": 2.654,
+      "step": 14539
+    },
+    {
+      "epoch": 0.6122105263157894,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001695436142275889,
+      "loss": 2.905,
+      "step": 14540
+    },
+    {
+      "epoch": 0.6122526315789474,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00016951166590862299,
+      "loss": 2.971,
+      "step": 14541
+    },
+    {
+      "epoch": 0.6122947368421052,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00016947971905605991,
+      "loss": 2.767,
+      "step": 14542
+    },
+    {
+      "epoch": 0.6123368421052632,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00016944777367048185,
+      "loss": 3.1406,
+      "step": 14543
+    },
+    {
+      "epoch": 0.612378947368421,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0001694158297524706,
+      "loss": 3.0883,
+      "step": 14544
+    },
+    {
+      "epoch": 0.612421052631579,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00016938388730260823,
+      "loss": 3.1209,
+      "step": 14545
+    },
+    {
+      "epoch": 0.6124631578947368,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001693519463214768,
+      "loss": 3.3674,
+      "step": 14546
+    },
+    {
+      "epoch": 0.6125052631578948,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00016932000680965813,
+      "loss": 3.3928,
+      "step": 14547
+    },
+    {
+      "epoch": 0.6125473684210526,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00016928806876773412,
+      "loss": 2.9182,
+      "step": 14548
+    },
+    {
+      "epoch": 0.6125894736842106,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0001692561321962867,
+      "loss": 3.1807,
+      "step": 14549
+    },
+    {
+      "epoch": 0.6126315789473684,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00016922419709589765,
+      "loss": 3.2928,
+      "step": 14550
+    },
+    {
+      "epoch": 0.6126736842105264,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0001691922634671488,
+      "loss": 3.2731,
+      "step": 14551
+    },
+    {
+      "epoch": 0.6127157894736842,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00016916033131062194,
+      "loss": 2.9972,
+      "step": 14552
+    },
+    {
+      "epoch": 0.6127578947368421,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.000169128400626899,
+      "loss": 2.6307,
+      "step": 14553
+    },
+    {
+      "epoch": 0.6128,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00016909647141656142,
+      "loss": 3.1296,
+      "step": 14554
+    },
+    {
+      "epoch": 0.6128421052631579,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0001690645436801912,
+      "loss": 2.9623,
+      "step": 14555
+    },
+    {
+      "epoch": 0.6128842105263158,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00016903261741836983,
+      "loss": 3.3081,
+      "step": 14556
+    },
+    {
+      "epoch": 0.6129263157894737,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00016900069263167905,
+      "loss": 2.6553,
+      "step": 14557
+    },
+    {
+      "epoch": 0.6129684210526316,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00016896876932070043,
+      "loss": 3.2837,
+      "step": 14558
+    },
+    {
+      "epoch": 0.6130105263157895,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00016893684748601565,
+      "loss": 3.2106,
+      "step": 14559
+    },
+    {
+      "epoch": 0.6130526315789474,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00016890492712820634,
+      "loss": 2.9639,
+      "step": 14560
+    },
+    {
+      "epoch": 0.6130947368421052,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0001688730082478539,
+      "loss": 3.3748,
+      "step": 14561
+    },
+    {
+      "epoch": 0.6131368421052632,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00016884109084554,
+      "loss": 2.7156,
+      "step": 14562
+    },
+    {
+      "epoch": 0.613178947368421,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00016880917492184599,
+      "loss": 2.9551,
+      "step": 14563
+    },
+    {
+      "epoch": 0.613221052631579,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00016877726047735347,
+      "loss": 2.8629,
+      "step": 14564
+    },
+    {
+      "epoch": 0.6132631578947368,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0001687453475126438,
+      "loss": 3.2514,
+      "step": 14565
+    },
+    {
+      "epoch": 0.6133052631578947,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00016871343602829842,
+      "loss": 3.2552,
+      "step": 14566
+    },
+    {
+      "epoch": 0.6133473684210526,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00016868152602489875,
+      "loss": 3.5586,
+      "step": 14567
+    },
+    {
+      "epoch": 0.6133894736842105,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0001686496175030261,
+      "loss": 3.4255,
+      "step": 14568
+    },
+    {
+      "epoch": 0.6134315789473684,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001686177104632619,
+      "loss": 2.9049,
+      "step": 14569
+    },
+    {
+      "epoch": 0.6134736842105263,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00016858580490618734,
+      "loss": 3.1637,
+      "step": 14570
+    },
+    {
+      "epoch": 0.6135157894736842,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001685539008323838,
+      "loss": 2.8351,
+      "step": 14571
+    },
+    {
+      "epoch": 0.6135578947368421,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00016852199824243246,
+      "loss": 3.4535,
+      "step": 14572
+    },
+    {
+      "epoch": 0.6136,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00016849009713691454,
+      "loss": 3.362,
+      "step": 14573
+    },
+    {
+      "epoch": 0.6136421052631579,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00016845819751641134,
+      "loss": 3.1529,
+      "step": 14574
+    },
+    {
+      "epoch": 0.6136842105263158,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0001684262993815039,
+      "loss": 3.1031,
+      "step": 14575
+    },
+    {
+      "epoch": 0.6137263157894737,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00016839440273277358,
+      "loss": 3.0688,
+      "step": 14576
+    },
+    {
+      "epoch": 0.6137684210526316,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00016836250757080122,
+      "loss": 2.9009,
+      "step": 14577
+    },
+    {
+      "epoch": 0.6138105263157895,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00016833061389616818,
+      "loss": 3.3902,
+      "step": 14578
+    },
+    {
+      "epoch": 0.6138526315789473,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00016829872170945532,
+      "loss": 3.1091,
+      "step": 14579
+    },
+    {
+      "epoch": 0.6138947368421053,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00016826683101124368,
+      "loss": 2.526,
+      "step": 14580
+    },
+    {
+      "epoch": 0.6139368421052631,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00016823494180211453,
+      "loss": 2.7999,
+      "step": 14581
+    },
+    {
+      "epoch": 0.6139789473684211,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00016820305408264853,
+      "loss": 2.6407,
+      "step": 14582
+    },
+    {
+      "epoch": 0.6140210526315789,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00016817116785342682,
+      "loss": 3.1273,
+      "step": 14583
+    },
+    {
+      "epoch": 0.6140631578947369,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00016813928311503024,
+      "loss": 3.4547,
+      "step": 14584
+    },
+    {
+      "epoch": 0.6141052631578947,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00016810739986803985,
+      "loss": 2.978,
+      "step": 14585
+    },
+    {
+      "epoch": 0.6141473684210527,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00016807551811303629,
+      "loss": 2.7872,
+      "step": 14586
+    },
+    {
+      "epoch": 0.6141894736842105,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00016804363785060056,
+      "loss": 3.059,
+      "step": 14587
+    },
+    {
+      "epoch": 0.6142315789473685,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00016801175908131354,
+      "loss": 3.4983,
+      "step": 14588
+    },
+    {
+      "epoch": 0.6142736842105263,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00016797988180575584,
+      "loss": 2.5859,
+      "step": 14589
+    },
+    {
+      "epoch": 0.6143157894736843,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00016794800602450838,
+      "loss": 3.3268,
+      "step": 14590
+    },
+    {
+      "epoch": 0.6143578947368421,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001679161317381518,
+      "loss": 3.0141,
+      "step": 14591
+    },
+    {
+      "epoch": 0.6144,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00016788425894726696,
+      "loss": 3.1844,
+      "step": 14592
+    },
+    {
+      "epoch": 0.6144421052631579,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00016785238765243431,
+      "loss": 3.1795,
+      "step": 14593
+    },
+    {
+      "epoch": 0.6144842105263157,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00016782051785423462,
+      "loss": 2.6298,
+      "step": 14594
+    },
+    {
+      "epoch": 0.6145263157894737,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00016778864955324869,
+      "loss": 3.3081,
+      "step": 14595
+    },
+    {
+      "epoch": 0.6145684210526315,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00016775678275005685,
+      "loss": 3.1323,
+      "step": 14596
+    },
+    {
+      "epoch": 0.6146105263157895,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0001677249174452399,
+      "loss": 2.9106,
+      "step": 14597
+    },
+    {
+      "epoch": 0.6146526315789473,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0001676930536393782,
+      "loss": 3.4599,
+      "step": 14598
+    },
+    {
+      "epoch": 0.6146947368421053,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0001676611913330523,
+      "loss": 2.8165,
+      "step": 14599
+    },
+    {
+      "epoch": 0.6147368421052631,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00016762933052684288,
+      "loss": 3.2782,
+      "step": 14600
+    },
+    {
+      "epoch": 0.6147789473684211,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00016759747122133017,
+      "loss": 3.232,
+      "step": 14601
+    },
+    {
+      "epoch": 0.6148210526315789,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00016756561341709484,
+      "loss": 3.3507,
+      "step": 14602
+    },
+    {
+      "epoch": 0.6148631578947369,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00016753375711471707,
+      "loss": 3.1859,
+      "step": 14603
+    },
+    {
+      "epoch": 0.6149052631578947,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00016750190231477744,
+      "loss": 3.6782,
+      "step": 14604
+    },
+    {
+      "epoch": 0.6149473684210527,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00016747004901785612,
+      "loss": 3.0056,
+      "step": 14605
+    },
+    {
+      "epoch": 0.6149894736842105,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00016743819722453356,
+      "loss": 3.2818,
+      "step": 14606
+    },
+    {
+      "epoch": 0.6150315789473684,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00016740634693539005,
+      "loss": 3.1684,
+      "step": 14607
+    },
+    {
+      "epoch": 0.6150736842105263,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001673744981510059,
+      "loss": 2.8754,
+      "step": 14608
+    },
+    {
+      "epoch": 0.6151157894736842,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0001673426508719613,
+      "loss": 2.8765,
+      "step": 14609
+    },
+    {
+      "epoch": 0.6151578947368421,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0001673108050988364,
+      "loss": 2.8824,
+      "step": 14610
+    },
+    {
+      "epoch": 0.6152,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00016727896083221163,
+      "loss": 3.6416,
+      "step": 14611
+    },
+    {
+      "epoch": 0.6152421052631579,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00016724711807266686,
+      "loss": 3.4658,
+      "step": 14612
+    },
+    {
+      "epoch": 0.6152842105263158,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00016721527682078235,
+      "loss": 2.8821,
+      "step": 14613
+    },
+    {
+      "epoch": 0.6153263157894737,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0001671834370771384,
+      "loss": 2.9586,
+      "step": 14614
+    },
+    {
+      "epoch": 0.6153684210526316,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00016715159884231475,
+      "loss": 2.7713,
+      "step": 14615
+    },
+    {
+      "epoch": 0.6154105263157895,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00016711976211689173,
+      "loss": 2.9707,
+      "step": 14616
+    },
+    {
+      "epoch": 0.6154526315789474,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00016708792690144918,
+      "loss": 2.7867,
+      "step": 14617
+    },
+    {
+      "epoch": 0.6154947368421053,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00016705609319656728,
+      "loss": 2.6267,
+      "step": 14618
+    },
+    {
+      "epoch": 0.6155368421052632,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00016702426100282584,
+      "loss": 2.7877,
+      "step": 14619
+    },
+    {
+      "epoch": 0.615578947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00016699243032080484,
+      "loss": 2.967,
+      "step": 14620
+    },
+    {
+      "epoch": 0.615621052631579,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00016696060115108436,
+      "loss": 2.7963,
+      "step": 14621
+    },
+    {
+      "epoch": 0.6156631578947368,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00016692877349424402,
+      "loss": 3.3531,
+      "step": 14622
+    },
+    {
+      "epoch": 0.6157052631578948,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00016689694735086392,
+      "loss": 3.166,
+      "step": 14623
+    },
+    {
+      "epoch": 0.6157473684210526,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00016686512272152376,
+      "loss": 3.0746,
+      "step": 14624
+    },
+    {
+      "epoch": 0.6157894736842106,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0001668332996068034,
+      "loss": 3.3734,
+      "step": 14625
+    },
+    {
+      "epoch": 0.6158315789473684,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00016680147800728255,
+      "loss": 3.1566,
+      "step": 14626
+    },
+    {
+      "epoch": 0.6158736842105264,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00016676965792354102,
+      "loss": 2.7957,
+      "step": 14627
+    },
+    {
+      "epoch": 0.6159157894736842,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0001667378393561586,
+      "loss": 3.2673,
+      "step": 14628
+    },
+    {
+      "epoch": 0.6159578947368421,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00016670602230571486,
+      "loss": 2.9194,
+      "step": 14629
+    },
+    {
+      "epoch": 0.616,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001666742067727896,
+      "loss": 3.4422,
+      "step": 14630
+    },
+    {
+      "epoch": 0.616042105263158,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00016664239275796232,
+      "loss": 2.9945,
+      "step": 14631
+    },
+    {
+      "epoch": 0.6160842105263158,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00016661058026181275,
+      "loss": 3.1402,
+      "step": 14632
+    },
+    {
+      "epoch": 0.6161263157894736,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00016657876928492034,
+      "loss": 3.1204,
+      "step": 14633
+    },
+    {
+      "epoch": 0.6161684210526316,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00016654695982786479,
+      "loss": 3.5275,
+      "step": 14634
+    },
+    {
+      "epoch": 0.6162105263157894,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00016651515189122562,
+      "loss": 3.0454,
+      "step": 14635
+    },
+    {
+      "epoch": 0.6162526315789474,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00016648334547558224,
+      "loss": 3.2069,
+      "step": 14636
+    },
+    {
+      "epoch": 0.6162947368421052,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00016645154058151423,
+      "loss": 3.0012,
+      "step": 14637
+    },
+    {
+      "epoch": 0.6163368421052632,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00016641973720960094,
+      "loss": 3.1157,
+      "step": 14638
+    },
+    {
+      "epoch": 0.616378947368421,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00016638793536042184,
+      "loss": 3.1975,
+      "step": 14639
+    },
+    {
+      "epoch": 0.616421052631579,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00016635613503455636,
+      "loss": 3.2474,
+      "step": 14640
+    },
+    {
+      "epoch": 0.6164631578947368,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00016632433623258374,
+      "loss": 3.3385,
+      "step": 14641
+    },
+    {
+      "epoch": 0.6165052631578948,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00016629253895508346,
+      "loss": 2.9157,
+      "step": 14642
+    },
+    {
+      "epoch": 0.6165473684210526,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00016626074320263474,
+      "loss": 3.3105,
+      "step": 14643
+    },
+    {
+      "epoch": 0.6165894736842106,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00016622894897581694,
+      "loss": 3.0823,
+      "step": 14644
+    },
+    {
+      "epoch": 0.6166315789473684,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001661971562752092,
+      "loss": 2.8648,
+      "step": 14645
+    },
+    {
+      "epoch": 0.6166736842105263,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001661653651013908,
+      "loss": 3.2991,
+      "step": 14646
+    },
+    {
+      "epoch": 0.6167157894736842,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00016613357545494103,
+      "loss": 3.2676,
+      "step": 14647
+    },
+    {
+      "epoch": 0.6167578947368421,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00016610178733643887,
+      "loss": 3.0841,
+      "step": 14648
+    },
+    {
+      "epoch": 0.6168,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00016607000074646367,
+      "loss": 3.2915,
+      "step": 14649
+    },
+    {
+      "epoch": 0.6168421052631579,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00016603821568559436,
+      "loss": 3.2486,
+      "step": 14650
+    },
+    {
+      "epoch": 0.6168842105263158,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001660064321544102,
+      "loss": 3.4859,
+      "step": 14651
+    },
+    {
+      "epoch": 0.6169263157894737,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00016597465015349007,
+      "loss": 3.4322,
+      "step": 14652
+    },
+    {
+      "epoch": 0.6169684210526316,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001659428696834131,
+      "loss": 2.9866,
+      "step": 14653
+    },
+    {
+      "epoch": 0.6170105263157895,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00016591109074475836,
+      "loss": 3.317,
+      "step": 14654
+    },
+    {
+      "epoch": 0.6170526315789474,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00016587931333810464,
+      "loss": 3.0822,
+      "step": 14655
+    },
+    {
+      "epoch": 0.6170947368421053,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00016584753746403113,
+      "loss": 2.8402,
+      "step": 14656
+    },
+    {
+      "epoch": 0.6171368421052632,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00016581576312311647,
+      "loss": 2.8914,
+      "step": 14657
+    },
+    {
+      "epoch": 0.617178947368421,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00016578399031593976,
+      "loss": 2.9862,
+      "step": 14658
+    },
+    {
+      "epoch": 0.6172210526315789,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00016575221904307976,
+      "loss": 3.3821,
+      "step": 14659
+    },
+    {
+      "epoch": 0.6172631578947368,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001657204493051153,
+      "loss": 2.8875,
+      "step": 14660
+    },
+    {
+      "epoch": 0.6173052631578947,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00016568868110262538,
+      "loss": 3.1469,
+      "step": 14661
+    },
+    {
+      "epoch": 0.6173473684210526,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00016565691443618853,
+      "loss": 3.2223,
+      "step": 14662
+    },
+    {
+      "epoch": 0.6173894736842105,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00016562514930638366,
+      "loss": 3.313,
+      "step": 14663
+    },
+    {
+      "epoch": 0.6174315789473684,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00016559338571378934,
+      "loss": 3.0802,
+      "step": 14664
+    },
+    {
+      "epoch": 0.6174736842105263,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00016556162365898445,
+      "loss": 3.2864,
+      "step": 14665
+    },
+    {
+      "epoch": 0.6175157894736842,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001655298631425475,
+      "loss": 3.5397,
+      "step": 14666
+    },
+    {
+      "epoch": 0.6175578947368421,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0001654981041650572,
+      "loss": 2.9549,
+      "step": 14667
+    },
+    {
+      "epoch": 0.6176,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001654663467270922,
+      "loss": 3.1089,
+      "step": 14668
+    },
+    {
+      "epoch": 0.6176421052631579,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00016543459082923095,
+      "loss": 3.3306,
+      "step": 14669
+    },
+    {
+      "epoch": 0.6176842105263158,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00016540283647205223,
+      "loss": 2.5111,
+      "step": 14670
+    },
+    {
+      "epoch": 0.6177263157894737,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00016537108365613436,
+      "loss": 3.2455,
+      "step": 14671
+    },
+    {
+      "epoch": 0.6177684210526316,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00016533933238205596,
+      "loss": 3.1339,
+      "step": 14672
+    },
+    {
+      "epoch": 0.6178105263157895,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00016530758265039536,
+      "loss": 3.1826,
+      "step": 14673
+    },
+    {
+      "epoch": 0.6178526315789473,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00016527583446173112,
+      "loss": 3.2293,
+      "step": 14674
+    },
+    {
+      "epoch": 0.6178947368421053,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00016524408781664162,
+      "loss": 2.9576,
+      "step": 14675
+    },
+    {
+      "epoch": 0.6179368421052631,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0001652123427157053,
+      "loss": 3.1977,
+      "step": 14676
+    },
+    {
+      "epoch": 0.6179789473684211,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0001651805991595005,
+      "loss": 2.9556,
+      "step": 14677
+    },
+    {
+      "epoch": 0.6180210526315789,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00016514885714860546,
+      "loss": 2.7214,
+      "step": 14678
+    },
+    {
+      "epoch": 0.6180631578947369,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00016511711668359863,
+      "loss": 3.0968,
+      "step": 14679
+    },
+    {
+      "epoch": 0.6181052631578947,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00016508537776505812,
+      "loss": 3.2297,
+      "step": 14680
+    },
+    {
+      "epoch": 0.6181473684210527,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00016505364039356223,
+      "loss": 3.832,
+      "step": 14681
+    },
+    {
+      "epoch": 0.6181894736842105,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00016502190456968929,
+      "loss": 3.6666,
+      "step": 14682
+    },
+    {
+      "epoch": 0.6182315789473685,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00016499017029401734,
+      "loss": 3.0906,
+      "step": 14683
+    },
+    {
+      "epoch": 0.6182736842105263,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00016495843756712464,
+      "loss": 3.1767,
+      "step": 14684
+    },
+    {
+      "epoch": 0.6183157894736843,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00016492670638958924,
+      "loss": 2.901,
+      "step": 14685
+    },
+    {
+      "epoch": 0.6183578947368421,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0001648949767619894,
+      "loss": 2.8862,
+      "step": 14686
+    },
+    {
+      "epoch": 0.6184,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00016486324868490298,
+      "loss": 3.1939,
+      "step": 14687
+    },
+    {
+      "epoch": 0.6184421052631579,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00016483152215890812,
+      "loss": 3.2774,
+      "step": 14688
+    },
+    {
+      "epoch": 0.6184842105263157,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.000164799797184583,
+      "loss": 2.9213,
+      "step": 14689
+    },
+    {
+      "epoch": 0.6185263157894737,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00016476807376250536,
+      "loss": 3.2491,
+      "step": 14690
+    },
+    {
+      "epoch": 0.6185684210526315,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00016473635189325332,
+      "loss": 2.8205,
+      "step": 14691
+    },
+    {
+      "epoch": 0.6186105263157895,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00016470463157740466,
+      "loss": 3.2021,
+      "step": 14692
+    },
+    {
+      "epoch": 0.6186526315789473,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00016467291281553755,
+      "loss": 3.0599,
+      "step": 14693
+    },
+    {
+      "epoch": 0.6186947368421053,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001646411956082296,
+      "loss": 3.1228,
+      "step": 14694
+    },
+    {
+      "epoch": 0.6187368421052631,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0001646094799560588,
+      "loss": 2.8678,
+      "step": 14695
+    },
+    {
+      "epoch": 0.6187789473684211,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.000164577765859603,
+      "loss": 3.1835,
+      "step": 14696
+    },
+    {
+      "epoch": 0.6188210526315789,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00016454605331943988,
+      "loss": 3.0552,
+      "step": 14697
+    },
+    {
+      "epoch": 0.6188631578947369,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00016451434233614738,
+      "loss": 3.1266,
+      "step": 14698
+    },
+    {
+      "epoch": 0.6189052631578947,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00016448263291030297,
+      "loss": 3.0081,
+      "step": 14699
+    },
+    {
+      "epoch": 0.6189473684210526,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00016445092504248452,
+      "loss": 3.1335,
+      "step": 14700
+    },
+    {
+      "epoch": 0.6189894736842105,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00016441921873326976,
+      "loss": 3.0132,
+      "step": 14701
+    },
+    {
+      "epoch": 0.6190315789473684,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0001643875139832362,
+      "loss": 3.0719,
+      "step": 14702
+    },
+    {
+      "epoch": 0.6190736842105263,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0001643558107929617,
+      "loss": 3.2791,
+      "step": 14703
+    },
+    {
+      "epoch": 0.6191157894736842,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00016432410916302357,
+      "loss": 2.4616,
+      "step": 14704
+    },
+    {
+      "epoch": 0.6191578947368421,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001642924090939996,
+      "loss": 3.0128,
+      "step": 14705
+    },
+    {
+      "epoch": 0.6192,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00016426071058646717,
+      "loss": 2.8629,
+      "step": 14706
+    },
+    {
+      "epoch": 0.6192421052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00016422901364100383,
+      "loss": 3.2047,
+      "step": 14707
+    },
+    {
+      "epoch": 0.6192842105263158,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00016419731825818712,
+      "loss": 2.9342,
+      "step": 14708
+    },
+    {
+      "epoch": 0.6193263157894737,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00016416562443859444,
+      "loss": 2.9637,
+      "step": 14709
+    },
+    {
+      "epoch": 0.6193684210526316,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001641339321828033,
+      "loss": 3.4904,
+      "step": 14710
+    },
+    {
+      "epoch": 0.6194105263157895,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0001641022414913909,
+      "loss": 3.3665,
+      "step": 14711
+    },
+    {
+      "epoch": 0.6194526315789474,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0001640705523649349,
+      "loss": 2.6715,
+      "step": 14712
+    },
+    {
+      "epoch": 0.6194947368421052,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00016403886480401234,
+      "loss": 3.5207,
+      "step": 14713
+    },
+    {
+      "epoch": 0.6195368421052632,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00016400717880920069,
+      "loss": 2.9443,
+      "step": 14714
+    },
+    {
+      "epoch": 0.619578947368421,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00016397549438107727,
+      "loss": 3.294,
+      "step": 14715
+    },
+    {
+      "epoch": 0.619621052631579,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00016394381152021916,
+      "loss": 2.8237,
+      "step": 14716
+    },
+    {
+      "epoch": 0.6196631578947368,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00016391213022720378,
+      "loss": 2.4853,
+      "step": 14717
+    },
+    {
+      "epoch": 0.6197052631578948,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00016388045050260816,
+      "loss": 3.3355,
+      "step": 14718
+    },
+    {
+      "epoch": 0.6197473684210526,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00016384877234700967,
+      "loss": 3.7743,
+      "step": 14719
+    },
+    {
+      "epoch": 0.6197894736842106,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00016381709576098518,
+      "loss": 2.9684,
+      "step": 14720
+    },
+    {
+      "epoch": 0.6198315789473684,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.000163785420745112,
+      "loss": 3.1094,
+      "step": 14721
+    },
+    {
+      "epoch": 0.6198736842105264,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00016375374729996723,
+      "loss": 3.2501,
+      "step": 14722
+    },
+    {
+      "epoch": 0.6199157894736842,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00016372207542612776,
+      "loss": 3.146,
+      "step": 14723
+    },
+    {
+      "epoch": 0.6199578947368422,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00016369040512417073,
+      "loss": 3.0607,
+      "step": 14724
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00016365873639467314,
+      "loss": 2.838,
+      "step": 14725
+    },
+    {
+      "epoch": 0.620042105263158,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001636270692382119,
+      "loss": 3.2375,
+      "step": 14726
+    },
+    {
+      "epoch": 0.6200842105263158,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00016359540365536393,
+      "loss": 3.2377,
+      "step": 14727
+    },
+    {
+      "epoch": 0.6201263157894736,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00016356373964670617,
+      "loss": 3.3782,
+      "step": 14728
+    },
+    {
+      "epoch": 0.6201684210526316,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00016353207721281566,
+      "loss": 3.1545,
+      "step": 14729
+    },
+    {
+      "epoch": 0.6202105263157894,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00016350041635426898,
+      "loss": 3.169,
+      "step": 14730
+    },
+    {
+      "epoch": 0.6202526315789474,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00016346875707164317,
+      "loss": 3.2718,
+      "step": 14731
+    },
+    {
+      "epoch": 0.6202947368421052,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00016343709936551484,
+      "loss": 3.0835,
+      "step": 14732
+    },
+    {
+      "epoch": 0.6203368421052632,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00016340544323646092,
+      "loss": 2.9436,
+      "step": 14733
+    },
+    {
+      "epoch": 0.620378947368421,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00016337378868505804,
+      "loss": 2.7633,
+      "step": 14734
+    },
+    {
+      "epoch": 0.620421052631579,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001633421357118829,
+      "loss": 3.0181,
+      "step": 14735
+    },
+    {
+      "epoch": 0.6204631578947368,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001633104843175124,
+      "loss": 3.3796,
+      "step": 14736
+    },
+    {
+      "epoch": 0.6205052631578948,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00016327883450252285,
+      "loss": 2.9884,
+      "step": 14737
+    },
+    {
+      "epoch": 0.6205473684210526,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00016324718626749117,
+      "loss": 3.1744,
+      "step": 14738
+    },
+    {
+      "epoch": 0.6205894736842106,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00016321553961299375,
+      "loss": 2.9381,
+      "step": 14739
+    },
+    {
+      "epoch": 0.6206315789473684,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00016318389453960722,
+      "loss": 2.8592,
+      "step": 14740
+    },
+    {
+      "epoch": 0.6206736842105263,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00016315225104790814,
+      "loss": 3.2543,
+      "step": 14741
+    },
+    {
+      "epoch": 0.6207157894736842,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00016312060913847298,
+      "loss": 2.9172,
+      "step": 14742
+    },
+    {
+      "epoch": 0.6207578947368421,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0001630889688118783,
+      "loss": 3.3776,
+      "step": 14743
+    },
+    {
+      "epoch": 0.6208,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00016305733006870043,
+      "loss": 3.0296,
+      "step": 14744
+    },
+    {
+      "epoch": 0.6208421052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00016302569290951597,
+      "loss": 2.8502,
+      "step": 14745
+    },
+    {
+      "epoch": 0.6208842105263158,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0001629940573349011,
+      "loss": 3.4774,
+      "step": 14746
+    },
+    {
+      "epoch": 0.6209263157894737,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001629624233454322,
+      "loss": 3.2138,
+      "step": 14747
+    },
+    {
+      "epoch": 0.6209684210526316,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00016293079094168586,
+      "loss": 3.4904,
+      "step": 14748
+    },
+    {
+      "epoch": 0.6210105263157895,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00016289916012423807,
+      "loss": 2.8742,
+      "step": 14749
+    },
+    {
+      "epoch": 0.6210526315789474,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00016286753089366532,
+      "loss": 3.2267,
+      "step": 14750
+    },
+    {
+      "epoch": 0.6210947368421053,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001628359032505437,
+      "loss": 3.1515,
+      "step": 14751
+    },
+    {
+      "epoch": 0.6211368421052632,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00016280427719544965,
+      "loss": 3.3967,
+      "step": 14752
+    },
+    {
+      "epoch": 0.621178947368421,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001627726527289591,
+      "loss": 3.3758,
+      "step": 14753
+    },
+    {
+      "epoch": 0.6212210526315789,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00016274102985164833,
+      "loss": 2.9003,
+      "step": 14754
+    },
+    {
+      "epoch": 0.6212631578947369,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00016270940856409356,
+      "loss": 2.6935,
+      "step": 14755
+    },
+    {
+      "epoch": 0.6213052631578947,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0001626777888668707,
+      "loss": 3.0158,
+      "step": 14756
+    },
+    {
+      "epoch": 0.6213473684210526,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00016264617076055605,
+      "loss": 3.2252,
+      "step": 14757
+    },
+    {
+      "epoch": 0.6213894736842105,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0001626145542457254,
+      "loss": 3.0112,
+      "step": 14758
+    },
+    {
+      "epoch": 0.6214315789473684,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00016258293932295499,
+      "loss": 3.0777,
+      "step": 14759
+    },
+    {
+      "epoch": 0.6214736842105263,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00016255132599282064,
+      "loss": 3.1352,
+      "step": 14760
+    },
+    {
+      "epoch": 0.6215157894736842,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0001625197142558984,
+      "loss": 3.009,
+      "step": 14761
+    },
+    {
+      "epoch": 0.6215578947368421,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00016248810411276427,
+      "loss": 3.3425,
+      "step": 14762
+    },
+    {
+      "epoch": 0.6216,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00016245649556399395,
+      "loss": 3.132,
+      "step": 14763
+    },
+    {
+      "epoch": 0.6216421052631579,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00016242488861016352,
+      "loss": 2.775,
+      "step": 14764
+    },
+    {
+      "epoch": 0.6216842105263158,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00016239328325184864,
+      "loss": 3.3143,
+      "step": 14765
+    },
+    {
+      "epoch": 0.6217263157894737,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00016236167948962524,
+      "loss": 2.9878,
+      "step": 14766
+    },
+    {
+      "epoch": 0.6217684210526315,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00016233007732406902,
+      "loss": 2.799,
+      "step": 14767
+    },
+    {
+      "epoch": 0.6218105263157895,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00016229847675575584,
+      "loss": 3.5372,
+      "step": 14768
+    },
+    {
+      "epoch": 0.6218526315789473,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00016226687778526136,
+      "loss": 2.6357,
+      "step": 14769
+    },
+    {
+      "epoch": 0.6218947368421053,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00016223528041316127,
+      "loss": 3.2334,
+      "step": 14770
+    },
+    {
+      "epoch": 0.6219368421052631,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00016220368464003133,
+      "loss": 2.6858,
+      "step": 14771
+    },
+    {
+      "epoch": 0.6219789473684211,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00016217209046644697,
+      "loss": 3.527,
+      "step": 14772
+    },
+    {
+      "epoch": 0.6220210526315789,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00016214049789298405,
+      "loss": 3.4784,
+      "step": 14773
+    },
+    {
+      "epoch": 0.6220631578947369,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00016210890692021793,
+      "loss": 3.4383,
+      "step": 14774
+    },
+    {
+      "epoch": 0.6221052631578947,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00016207731754872427,
+      "loss": 2.8816,
+      "step": 14775
+    },
+    {
+      "epoch": 0.6221473684210527,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00016204572977907862,
+      "loss": 3.0155,
+      "step": 14776
+    },
+    {
+      "epoch": 0.6221894736842105,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00016201414361185636,
+      "loss": 3.5335,
+      "step": 14777
+    },
+    {
+      "epoch": 0.6222315789473685,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00016198255904763316,
+      "loss": 3.0139,
+      "step": 14778
+    },
+    {
+      "epoch": 0.6222736842105263,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00016195097608698417,
+      "loss": 3.1913,
+      "step": 14779
+    },
+    {
+      "epoch": 0.6223157894736842,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0001619193947304851,
+      "loss": 3.5826,
+      "step": 14780
+    },
+    {
+      "epoch": 0.6223578947368421,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00016188781497871103,
+      "loss": 2.7906,
+      "step": 14781
+    },
+    {
+      "epoch": 0.6224,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00016185623683223743,
+      "loss": 3.1118,
+      "step": 14782
+    },
+    {
+      "epoch": 0.6224421052631579,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00016182466029163973,
+      "loss": 2.6,
+      "step": 14783
+    },
+    {
+      "epoch": 0.6224842105263158,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001617930853574931,
+      "loss": 3.299,
+      "step": 14784
+    },
+    {
+      "epoch": 0.6225263157894737,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00016176151203037278,
+      "loss": 2.865,
+      "step": 14785
+    },
+    {
+      "epoch": 0.6225684210526315,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00016172994031085402,
+      "loss": 3.1171,
+      "step": 14786
+    },
+    {
+      "epoch": 0.6226105263157895,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00016169837019951213,
+      "loss": 2.9711,
+      "step": 14787
+    },
+    {
+      "epoch": 0.6226526315789473,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00016166680169692207,
+      "loss": 3.1872,
+      "step": 14788
+    },
+    {
+      "epoch": 0.6226947368421053,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0001616352348036591,
+      "loss": 3.0893,
+      "step": 14789
+    },
+    {
+      "epoch": 0.6227368421052631,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00016160366952029846,
+      "loss": 3.5962,
+      "step": 14790
+    },
+    {
+      "epoch": 0.6227789473684211,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00016157210584741497,
+      "loss": 3.4068,
+      "step": 14791
+    },
+    {
+      "epoch": 0.6228210526315789,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00016154054378558387,
+      "loss": 2.9685,
+      "step": 14792
+    },
+    {
+      "epoch": 0.6228631578947369,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00016150898333538007,
+      "loss": 2.9176,
+      "step": 14793
+    },
+    {
+      "epoch": 0.6229052631578947,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001614774244973786,
+      "loss": 2.8883,
+      "step": 14794
+    },
+    {
+      "epoch": 0.6229473684210526,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0001614458672721546,
+      "loss": 3.1156,
+      "step": 14795
+    },
+    {
+      "epoch": 0.6229894736842105,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00016141431166028273,
+      "loss": 2.5544,
+      "step": 14796
+    },
+    {
+      "epoch": 0.6230315789473684,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00016138275766233812,
+      "loss": 3.2856,
+      "step": 14797
+    },
+    {
+      "epoch": 0.6230736842105263,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0001613512052788954,
+      "loss": 3.0189,
+      "step": 14798
+    },
+    {
+      "epoch": 0.6231157894736842,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00016131965451052967,
+      "loss": 2.8155,
+      "step": 14799
+    },
+    {
+      "epoch": 0.6231578947368421,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00016128810535781562,
+      "loss": 3.1113,
+      "step": 14800
+    },
+    {
+      "epoch": 0.6232,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.000161256557821328,
+      "loss": 2.7182,
+      "step": 14801
+    },
+    {
+      "epoch": 0.6232421052631579,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00016122501190164168,
+      "loss": 2.7206,
+      "step": 14802
+    },
+    {
+      "epoch": 0.6232842105263158,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00016119346759933128,
+      "loss": 3.4374,
+      "step": 14803
+    },
+    {
+      "epoch": 0.6233263157894737,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00016116192491497168,
+      "loss": 3.5363,
+      "step": 14804
+    },
+    {
+      "epoch": 0.6233684210526316,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00016113038384913731,
+      "loss": 3.5053,
+      "step": 14805
+    },
+    {
+      "epoch": 0.6234105263157895,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.00016109884440240302,
+      "loss": 3.1919,
+      "step": 14806
+    },
+    {
+      "epoch": 0.6234526315789474,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00016106730657534324,
+      "loss": 2.9201,
+      "step": 14807
+    },
+    {
+      "epoch": 0.6234947368421052,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00016103577036853268,
+      "loss": 2.9444,
+      "step": 14808
+    },
+    {
+      "epoch": 0.6235368421052632,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00016100423578254585,
+      "loss": 2.926,
+      "step": 14809
+    },
+    {
+      "epoch": 0.623578947368421,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00016097270281795724,
+      "loss": 2.9178,
+      "step": 14810
+    },
+    {
+      "epoch": 0.623621052631579,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0001609411714753415,
+      "loss": 2.6337,
+      "step": 14811
+    },
+    {
+      "epoch": 0.6236631578947368,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00016090964175527288,
+      "loss": 3.3811,
+      "step": 14812
+    },
+    {
+      "epoch": 0.6237052631578948,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.000160878113658326,
+      "loss": 2.9459,
+      "step": 14813
+    },
+    {
+      "epoch": 0.6237473684210526,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00016084658718507506,
+      "loss": 3.1945,
+      "step": 14814
+    },
+    {
+      "epoch": 0.6237894736842106,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00016081506233609456,
+      "loss": 3.469,
+      "step": 14815
+    },
+    {
+      "epoch": 0.6238315789473684,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00016078353911195891,
+      "loss": 2.9836,
+      "step": 14816
+    },
+    {
+      "epoch": 0.6238736842105264,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0001607520175132423,
+      "loss": 2.9011,
+      "step": 14817
+    },
+    {
+      "epoch": 0.6239157894736842,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00016072049754051909,
+      "loss": 3.0529,
+      "step": 14818
+    },
+    {
+      "epoch": 0.6239578947368422,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00016068897919436348,
+      "loss": 2.9474,
+      "step": 14819
+    },
+    {
+      "epoch": 0.624,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00016065746247534985,
+      "loss": 3.2194,
+      "step": 14820
+    },
+    {
+      "epoch": 0.6240421052631578,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00016062594738405216,
+      "loss": 3.2669,
+      "step": 14821
+    },
+    {
+      "epoch": 0.6240842105263158,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0001605944339210447,
+      "loss": 3.2098,
+      "step": 14822
+    },
+    {
+      "epoch": 0.6241263157894736,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001605629220869017,
+      "loss": 2.7981,
+      "step": 14823
+    },
+    {
+      "epoch": 0.6241684210526316,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00016053141188219708,
+      "loss": 2.8417,
+      "step": 14824
+    },
+    {
+      "epoch": 0.6242105263157894,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00016049990330750509,
+      "loss": 2.8781,
+      "step": 14825
+    },
+    {
+      "epoch": 0.6242526315789474,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001604683963633996,
+      "loss": 3.4556,
+      "step": 14826
+    },
+    {
+      "epoch": 0.6242947368421052,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00016043689105045488,
+      "loss": 2.9334,
+      "step": 14827
+    },
+    {
+      "epoch": 0.6243368421052632,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00016040538736924464,
+      "loss": 3.1528,
+      "step": 14828
+    },
+    {
+      "epoch": 0.624378947368421,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00016037388532034297,
+      "loss": 2.6867,
+      "step": 14829
+    },
+    {
+      "epoch": 0.624421052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00016034238490432394,
+      "loss": 2.971,
+      "step": 14830
+    },
+    {
+      "epoch": 0.6244631578947368,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00016031088612176116,
+      "loss": 3.0505,
+      "step": 14831
+    },
+    {
+      "epoch": 0.6245052631578948,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0001602793889732288,
+      "loss": 2.9823,
+      "step": 14832
+    },
+    {
+      "epoch": 0.6245473684210526,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00016024789345930047,
+      "loss": 3.2585,
+      "step": 14833
+    },
+    {
+      "epoch": 0.6245894736842105,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00016021639958055007,
+      "loss": 3.009,
+      "step": 14834
+    },
+    {
+      "epoch": 0.6246315789473684,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00016018490733755138,
+      "loss": 2.7402,
+      "step": 14835
+    },
+    {
+      "epoch": 0.6246736842105263,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001601534167308781,
+      "loss": 3.5637,
+      "step": 14836
+    },
+    {
+      "epoch": 0.6247157894736842,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00016012192776110413,
+      "loss": 3.4937,
+      "step": 14837
+    },
+    {
+      "epoch": 0.6247578947368421,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00016009044042880288,
+      "loss": 2.8601,
+      "step": 14838
+    },
+    {
+      "epoch": 0.6248,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00016005895473454834,
+      "loss": 3.1615,
+      "step": 14839
+    },
+    {
+      "epoch": 0.6248421052631579,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00016002747067891382,
+      "loss": 2.6641,
+      "step": 14840
+    },
+    {
+      "epoch": 0.6248842105263158,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00015999598826247315,
+      "loss": 3.3915,
+      "step": 14841
+    },
+    {
+      "epoch": 0.6249263157894737,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00015996450748579982,
+      "loss": 3.0779,
+      "step": 14842
+    },
+    {
+      "epoch": 0.6249684210526316,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00015993302834946732,
+      "loss": 2.717,
+      "step": 14843
+    },
+    {
+      "epoch": 0.6250105263157895,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00015990155085404925,
+      "loss": 3.4828,
+      "step": 14844
+    },
+    {
+      "epoch": 0.6250526315789474,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00015987007500011902,
+      "loss": 3.0511,
+      "step": 14845
+    },
+    {
+      "epoch": 0.6250947368421053,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00015983860078825024,
+      "loss": 3.191,
+      "step": 14846
+    },
+    {
+      "epoch": 0.6251368421052632,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00015980712821901612,
+      "loss": 2.8682,
+      "step": 14847
+    },
+    {
+      "epoch": 0.6251789473684211,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00015977565729299016,
+      "loss": 3.1973,
+      "step": 14848
+    },
+    {
+      "epoch": 0.6252210526315789,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001597441880107458,
+      "loss": 3.1877,
+      "step": 14849
+    },
+    {
+      "epoch": 0.6252631578947369,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00015971272037285618,
+      "loss": 2.9822,
+      "step": 14850
+    },
+    {
+      "epoch": 0.6253052631578947,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00015968125437989478,
+      "loss": 2.9245,
+      "step": 14851
+    },
+    {
+      "epoch": 0.6253473684210527,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00015964979003243475,
+      "loss": 3.0644,
+      "step": 14852
+    },
+    {
+      "epoch": 0.6253894736842105,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00015961832733104946,
+      "loss": 3.1771,
+      "step": 14853
+    },
+    {
+      "epoch": 0.6254315789473684,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00015958686627631197,
+      "loss": 3.1407,
+      "step": 14854
+    },
+    {
+      "epoch": 0.6254736842105263,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00015955540686879554,
+      "loss": 3.1409,
+      "step": 14855
+    },
+    {
+      "epoch": 0.6255157894736842,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00015952394910907346,
+      "loss": 3.355,
+      "step": 14856
+    },
+    {
+      "epoch": 0.6255578947368421,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00015949249299771861,
+      "loss": 3.1925,
+      "step": 14857
+    },
+    {
+      "epoch": 0.6256,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00015946103853530427,
+      "loss": 3.0282,
+      "step": 14858
+    },
+    {
+      "epoch": 0.6256421052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00015942958572240335,
+      "loss": 3.0676,
+      "step": 14859
+    },
+    {
+      "epoch": 0.6256842105263158,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00015939813455958902,
+      "loss": 3.397,
+      "step": 14860
+    },
+    {
+      "epoch": 0.6257263157894737,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00015936668504743415,
+      "loss": 3.5096,
+      "step": 14861
+    },
+    {
+      "epoch": 0.6257684210526315,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00015933523718651184,
+      "loss": 3.1072,
+      "step": 14862
+    },
+    {
+      "epoch": 0.6258105263157895,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.000159303790977395,
+      "loss": 2.6977,
+      "step": 14863
+    },
+    {
+      "epoch": 0.6258526315789473,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00015927234642065647,
+      "loss": 2.9145,
+      "step": 14864
+    },
+    {
+      "epoch": 0.6258947368421053,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0001592409035168693,
+      "loss": 3.4849,
+      "step": 14865
+    },
+    {
+      "epoch": 0.6259368421052631,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0001592094622666061,
+      "loss": 3.4892,
+      "step": 14866
+    },
+    {
+      "epoch": 0.6259789473684211,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0001591780226704399,
+      "loss": 2.7073,
+      "step": 14867
+    },
+    {
+      "epoch": 0.6260210526315789,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00015914658472894336,
+      "loss": 2.9459,
+      "step": 14868
+    },
+    {
+      "epoch": 0.6260631578947369,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00015911514844268936,
+      "loss": 3.397,
+      "step": 14869
+    },
+    {
+      "epoch": 0.6261052631578947,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00015908371381225054,
+      "loss": 3.1791,
+      "step": 14870
+    },
+    {
+      "epoch": 0.6261473684210527,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00015905228083819961,
+      "loss": 3.104,
+      "step": 14871
+    },
+    {
+      "epoch": 0.6261894736842105,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00015902084952110933,
+      "loss": 3.1175,
+      "step": 14872
+    },
+    {
+      "epoch": 0.6262315789473685,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0001589894198615522,
+      "loss": 2.6542,
+      "step": 14873
+    },
+    {
+      "epoch": 0.6262736842105263,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.000158957991860101,
+      "loss": 3.1141,
+      "step": 14874
+    },
+    {
+      "epoch": 0.6263157894736842,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00015892656551732813,
+      "loss": 3.1094,
+      "step": 14875
+    },
+    {
+      "epoch": 0.6263578947368421,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00015889514083380622,
+      "loss": 3.0647,
+      "step": 14876
+    },
+    {
+      "epoch": 0.6264,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00015886371781010782,
+      "loss": 2.6812,
+      "step": 14877
+    },
+    {
+      "epoch": 0.6264421052631579,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00015883229644680542,
+      "loss": 3.1418,
+      "step": 14878
+    },
+    {
+      "epoch": 0.6264842105263158,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00015880087674447152,
+      "loss": 3.1786,
+      "step": 14879
+    },
+    {
+      "epoch": 0.6265263157894737,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00015876945870367838,
+      "loss": 3.2833,
+      "step": 14880
+    },
+    {
+      "epoch": 0.6265684210526316,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0001587380423249986,
+      "loss": 3.2237,
+      "step": 14881
+    },
+    {
+      "epoch": 0.6266105263157895,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001587066276090044,
+      "loss": 3.1676,
+      "step": 14882
+    },
+    {
+      "epoch": 0.6266526315789474,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00015867521455626815,
+      "loss": 2.8551,
+      "step": 14883
+    },
+    {
+      "epoch": 0.6266947368421053,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00015864380316736227,
+      "loss": 3.1448,
+      "step": 14884
+    },
+    {
+      "epoch": 0.6267368421052631,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00015861239344285895,
+      "loss": 2.9195,
+      "step": 14885
+    },
+    {
+      "epoch": 0.6267789473684211,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00015858098538333042,
+      "loss": 2.9354,
+      "step": 14886
+    },
+    {
+      "epoch": 0.626821052631579,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0001585495789893489,
+      "loss": 3.7157,
+      "step": 14887
+    },
+    {
+      "epoch": 0.6268631578947368,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00015851817426148669,
+      "loss": 3.2879,
+      "step": 14888
+    },
+    {
+      "epoch": 0.6269052631578947,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00015848677120031577,
+      "loss": 3.3159,
+      "step": 14889
+    },
+    {
+      "epoch": 0.6269473684210526,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00015845536980640834,
+      "loss": 2.9835,
+      "step": 14890
+    },
+    {
+      "epoch": 0.6269894736842105,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0001584239700803366,
+      "loss": 3.1589,
+      "step": 14891
+    },
+    {
+      "epoch": 0.6270315789473684,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00015839257202267246,
+      "loss": 2.8193,
+      "step": 14892
+    },
+    {
+      "epoch": 0.6270736842105263,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00015836117563398805,
+      "loss": 3.2201,
+      "step": 14893
+    },
+    {
+      "epoch": 0.6271157894736842,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00015832978091485532,
+      "loss": 3.595,
+      "step": 14894
+    },
+    {
+      "epoch": 0.6271578947368421,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00015829838786584623,
+      "loss": 3.4672,
+      "step": 14895
+    },
+    {
+      "epoch": 0.6272,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001582669964875329,
+      "loss": 3.2728,
+      "step": 14896
+    },
+    {
+      "epoch": 0.6272421052631579,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.000158235606780487,
+      "loss": 2.7824,
+      "step": 14897
+    },
+    {
+      "epoch": 0.6272842105263158,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00015820421874528066,
+      "loss": 3.1381,
+      "step": 14898
+    },
+    {
+      "epoch": 0.6273263157894737,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00015817283238248546,
+      "loss": 3.3476,
+      "step": 14899
+    },
+    {
+      "epoch": 0.6273684210526316,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00015814144769267345,
+      "loss": 3.5594,
+      "step": 14900
+    },
+    {
+      "epoch": 0.6274105263157894,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0001581100646764163,
+      "loss": 3.0274,
+      "step": 14901
+    },
+    {
+      "epoch": 0.6274526315789474,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00015807868333428578,
+      "loss": 3.3255,
+      "step": 14902
+    },
+    {
+      "epoch": 0.6274947368421052,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00015804730366685365,
+      "loss": 3.161,
+      "step": 14903
+    },
+    {
+      "epoch": 0.6275368421052632,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00015801592567469159,
+      "loss": 3.0962,
+      "step": 14904
+    },
+    {
+      "epoch": 0.627578947368421,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00015798454935837139,
+      "loss": 3.0788,
+      "step": 14905
+    },
+    {
+      "epoch": 0.627621052631579,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00015795317471846448,
+      "loss": 2.9881,
+      "step": 14906
+    },
+    {
+      "epoch": 0.6276631578947368,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00015792180175554267,
+      "loss": 3.3221,
+      "step": 14907
+    },
+    {
+      "epoch": 0.6277052631578948,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00015789043047017737,
+      "loss": 3.2497,
+      "step": 14908
+    },
+    {
+      "epoch": 0.6277473684210526,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00015785906086294017,
+      "loss": 3.2467,
+      "step": 14909
+    },
+    {
+      "epoch": 0.6277894736842106,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00015782769293440268,
+      "loss": 2.998,
+      "step": 14910
+    },
+    {
+      "epoch": 0.6278315789473684,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0001577963266851363,
+      "loss": 3.075,
+      "step": 14911
+    },
+    {
+      "epoch": 0.6278736842105264,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00015776496211571258,
+      "loss": 2.7698,
+      "step": 14912
+    },
+    {
+      "epoch": 0.6279157894736842,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001577335992267028,
+      "loss": 3.0398,
+      "step": 14913
+    },
+    {
+      "epoch": 0.6279578947368422,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00015770223801867854,
+      "loss": 3.1419,
+      "step": 14914
+    },
+    {
+      "epoch": 0.628,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00015767087849221097,
+      "loss": 2.5886,
+      "step": 14915
+    },
+    {
+      "epoch": 0.6280421052631578,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00015763952064787147,
+      "loss": 3.1272,
+      "step": 14916
+    },
+    {
+      "epoch": 0.6280842105263158,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0001576081644862315,
+      "loss": 3.2196,
+      "step": 14917
+    },
+    {
+      "epoch": 0.6281263157894736,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00015757681000786212,
+      "loss": 3.3713,
+      "step": 14918
+    },
+    {
+      "epoch": 0.6281684210526316,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001575454572133347,
+      "loss": 3.2238,
+      "step": 14919
+    },
+    {
+      "epoch": 0.6282105263157894,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0001575141061032204,
+      "loss": 3.0077,
+      "step": 14920
+    },
+    {
+      "epoch": 0.6282526315789474,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00015748275667809051,
+      "loss": 2.8974,
+      "step": 14921
+    },
+    {
+      "epoch": 0.6282947368421052,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00015745140893851596,
+      "loss": 3.0965,
+      "step": 14922
+    },
+    {
+      "epoch": 0.6283368421052632,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00015742006288506805,
+      "loss": 2.7788,
+      "step": 14923
+    },
+    {
+      "epoch": 0.628378947368421,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00015738871851831787,
+      "loss": 2.8099,
+      "step": 14924
+    },
+    {
+      "epoch": 0.628421052631579,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00015735737583883635,
+      "loss": 3.4885,
+      "step": 14925
+    },
+    {
+      "epoch": 0.6284631578947368,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00015732603484719466,
+      "loss": 2.9779,
+      "step": 14926
+    },
+    {
+      "epoch": 0.6285052631578948,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00015729469554396364,
+      "loss": 3.539,
+      "step": 14927
+    },
+    {
+      "epoch": 0.6285473684210526,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00015726335792971447,
+      "loss": 3.0421,
+      "step": 14928
+    },
+    {
+      "epoch": 0.6285894736842105,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00015723202200501786,
+      "loss": 2.9923,
+      "step": 14929
+    },
+    {
+      "epoch": 0.6286315789473684,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00015720068777044476,
+      "loss": 3.1699,
+      "step": 14930
+    },
+    {
+      "epoch": 0.6286736842105263,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001571693552265663,
+      "loss": 3.3098,
+      "step": 14931
+    },
+    {
+      "epoch": 0.6287157894736842,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00015713802437395291,
+      "loss": 3.3099,
+      "step": 14932
+    },
+    {
+      "epoch": 0.6287578947368421,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00015710669521317577,
+      "loss": 2.8568,
+      "step": 14933
+    },
+    {
+      "epoch": 0.6288,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00015707536774480535,
+      "loss": 3.2563,
+      "step": 14934
+    },
+    {
+      "epoch": 0.6288421052631579,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00015704404196941263,
+      "loss": 3.1794,
+      "step": 14935
+    },
+    {
+      "epoch": 0.6288842105263158,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001570127178875682,
+      "loss": 3.225,
+      "step": 14936
+    },
+    {
+      "epoch": 0.6289263157894737,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00015698139549984278,
+      "loss": 3.476,
+      "step": 14937
+    },
+    {
+      "epoch": 0.6289684210526316,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00015695007480680714,
+      "loss": 2.9088,
+      "step": 14938
+    },
+    {
+      "epoch": 0.6290105263157895,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001569187558090317,
+      "loss": 3.3564,
+      "step": 14939
+    },
+    {
+      "epoch": 0.6290526315789474,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0001568874385070873,
+      "loss": 3.0855,
+      "step": 14940
+    },
+    {
+      "epoch": 0.6290947368421053,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00015685612290154425,
+      "loss": 3.1391,
+      "step": 14941
+    },
+    {
+      "epoch": 0.6291368421052631,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00015682480899297319,
+      "loss": 3.0318,
+      "step": 14942
+    },
+    {
+      "epoch": 0.6291789473684211,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00015679349678194467,
+      "loss": 3.0833,
+      "step": 14943
+    },
+    {
+      "epoch": 0.6292210526315789,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00015676218626902916,
+      "loss": 2.6311,
+      "step": 14944
+    },
+    {
+      "epoch": 0.6292631578947369,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00015673087745479705,
+      "loss": 3.2621,
+      "step": 14945
+    },
+    {
+      "epoch": 0.6293052631578947,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00015669957033981875,
+      "loss": 3.1662,
+      "step": 14946
+    },
+    {
+      "epoch": 0.6293473684210527,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001566682649246647,
+      "loss": 3.2021,
+      "step": 14947
+    },
+    {
+      "epoch": 0.6293894736842105,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001566369612099051,
+      "loss": 3.1501,
+      "step": 14948
+    },
+    {
+      "epoch": 0.6294315789473685,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0001566056591961104,
+      "loss": 3.3739,
+      "step": 14949
+    },
+    {
+      "epoch": 0.6294736842105263,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00015657435888385098,
+      "loss": 3.1056,
+      "step": 14950
+    },
+    {
+      "epoch": 0.6295157894736843,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00015654306027369682,
+      "loss": 2.8816,
+      "step": 14951
+    },
+    {
+      "epoch": 0.6295578947368421,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00015651176336621837,
+      "loss": 2.7512,
+      "step": 14952
+    },
+    {
+      "epoch": 0.6296,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00015648046816198568,
+      "loss": 3.5269,
+      "step": 14953
+    },
+    {
+      "epoch": 0.6296421052631579,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0001564491746615691,
+      "loss": 3.2092,
+      "step": 14954
+    },
+    {
+      "epoch": 0.6296842105263157,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00015641788286553855,
+      "loss": 2.9902,
+      "step": 14955
+    },
+    {
+      "epoch": 0.6297263157894737,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0001563865927744642,
+      "loss": 2.9846,
+      "step": 14956
+    },
+    {
+      "epoch": 0.6297684210526315,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00015635530438891623,
+      "loss": 3.8157,
+      "step": 14957
+    },
+    {
+      "epoch": 0.6298105263157895,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001563240177094645,
+      "loss": 2.7765,
+      "step": 14958
+    },
+    {
+      "epoch": 0.6298526315789473,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00015629273273667915,
+      "loss": 3.4377,
+      "step": 14959
+    },
+    {
+      "epoch": 0.6298947368421053,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00015626144947113012,
+      "loss": 3.4104,
+      "step": 14960
+    },
+    {
+      "epoch": 0.6299368421052631,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0001562301679133873,
+      "loss": 3.1417,
+      "step": 14961
+    },
+    {
+      "epoch": 0.6299789473684211,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00015619888806402065,
+      "loss": 2.893,
+      "step": 14962
+    },
+    {
+      "epoch": 0.6300210526315789,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001561676099236,
+      "loss": 3.422,
+      "step": 14963
+    },
+    {
+      "epoch": 0.6300631578947369,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0001561363334926954,
+      "loss": 3.1399,
+      "step": 14964
+    },
+    {
+      "epoch": 0.6301052631578947,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00015610505877187636,
+      "loss": 3.207,
+      "step": 14965
+    },
+    {
+      "epoch": 0.6301473684210527,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.000156073785761713,
+      "loss": 3.2552,
+      "step": 14966
+    },
+    {
+      "epoch": 0.6301894736842105,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00015604251446277474,
+      "loss": 2.9491,
+      "step": 14967
+    },
+    {
+      "epoch": 0.6302315789473685,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00015601124487563153,
+      "loss": 3.4595,
+      "step": 14968
+    },
+    {
+      "epoch": 0.6302736842105263,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.000155979977000853,
+      "loss": 3.6816,
+      "step": 14969
+    },
+    {
+      "epoch": 0.6303157894736842,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00015594871083900874,
+      "loss": 3.2167,
+      "step": 14970
+    },
+    {
+      "epoch": 0.6303578947368421,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00015591744639066862,
+      "loss": 2.7873,
+      "step": 14971
+    },
+    {
+      "epoch": 0.6304,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.000155886183656402,
+      "loss": 2.6726,
+      "step": 14972
+    },
+    {
+      "epoch": 0.6304421052631579,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00015585492263677858,
+      "loss": 3.301,
+      "step": 14973
+    },
+    {
+      "epoch": 0.6304842105263158,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0001558236633323678,
+      "loss": 3.1195,
+      "step": 14974
+    },
+    {
+      "epoch": 0.6305263157894737,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001557924057437393,
+      "loss": 3.2124,
+      "step": 14975
+    },
+    {
+      "epoch": 0.6305684210526316,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00015576114987146238,
+      "loss": 2.5183,
+      "step": 14976
+    },
+    {
+      "epoch": 0.6306105263157895,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0001557298957161066,
+      "loss": 2.826,
+      "step": 14977
+    },
+    {
+      "epoch": 0.6306526315789474,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0001556986432782414,
+      "loss": 2.817,
+      "step": 14978
+    },
+    {
+      "epoch": 0.6306947368421053,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00015566739255843607,
+      "loss": 3.1216,
+      "step": 14979
+    },
+    {
+      "epoch": 0.6307368421052632,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00015563614355726009,
+      "loss": 3.059,
+      "step": 14980
+    },
+    {
+      "epoch": 0.6307789473684211,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00015560489627528264,
+      "loss": 3.1275,
+      "step": 14981
+    },
+    {
+      "epoch": 0.630821052631579,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00015557365071307313,
+      "loss": 3.0089,
+      "step": 14982
+    },
+    {
+      "epoch": 0.6308631578947368,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00015554240687120068,
+      "loss": 2.8831,
+      "step": 14983
+    },
+    {
+      "epoch": 0.6309052631578947,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00015551116475023458,
+      "loss": 2.8487,
+      "step": 14984
+    },
+    {
+      "epoch": 0.6309473684210526,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0001554799243507441,
+      "loss": 3.6548,
+      "step": 14985
+    },
+    {
+      "epoch": 0.6309894736842105,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00015544868567329828,
+      "loss": 2.9701,
+      "step": 14986
+    },
+    {
+      "epoch": 0.6310315789473684,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00015541744871846642,
+      "loss": 2.5822,
+      "step": 14987
+    },
+    {
+      "epoch": 0.6310736842105263,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00015538621348681742,
+      "loss": 3.1166,
+      "step": 14988
+    },
+    {
+      "epoch": 0.6311157894736842,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00015535497997892057,
+      "loss": 3.0705,
+      "step": 14989
+    },
+    {
+      "epoch": 0.6311578947368421,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00015532374819534462,
+      "loss": 3.0401,
+      "step": 14990
+    },
+    {
+      "epoch": 0.6312,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00015529251813665878,
+      "loss": 3.2649,
+      "step": 14991
+    },
+    {
+      "epoch": 0.6312421052631579,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00015526128980343208,
+      "loss": 3.2711,
+      "step": 14992
+    },
+    {
+      "epoch": 0.6312842105263158,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00015523006319623326,
+      "loss": 3.0064,
+      "step": 14993
+    },
+    {
+      "epoch": 0.6313263157894737,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00015519883831563135,
+      "loss": 2.876,
+      "step": 14994
+    },
+    {
+      "epoch": 0.6313684210526316,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00015516761516219524,
+      "loss": 3.1223,
+      "step": 14995
+    },
+    {
+      "epoch": 0.6314105263157894,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00015513639373649374,
+      "loss": 3.1781,
+      "step": 14996
+    },
+    {
+      "epoch": 0.6314526315789474,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00015510517403909574,
+      "loss": 3.3544,
+      "step": 14997
+    },
+    {
+      "epoch": 0.6314947368421052,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0001550739560705699,
+      "loss": 3.7107,
+      "step": 14998
+    },
+    {
+      "epoch": 0.6315368421052632,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00015504273983148518,
+      "loss": 3.4328,
+      "step": 14999
+    },
+    {
+      "epoch": 0.631578947368421,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00015501152532241004,
+      "loss": 3.466,
+      "step": 15000
+    },
+    {
+      "epoch": 0.631578947368421,
+      "eval_loss": 3.0952181816101074,
+      "eval_runtime": 335.5119,
+      "eval_samples_per_second": 44.708,
+      "eval_steps_per_second": 5.588,
+      "step": 15000
+    },
+    {
+      "epoch": 0.631621052631579,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00015498031254391334,
+      "loss": 3.3284,
+      "step": 15001
+    },
+    {
+      "epoch": 0.6316631578947368,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00015494910149656371,
+      "loss": 2.9961,
+      "step": 15002
+    },
+    {
+      "epoch": 0.6317052631578948,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00015491789218092971,
+      "loss": 2.9405,
+      "step": 15003
+    },
+    {
+      "epoch": 0.6317473684210526,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001548866845975801,
+      "loss": 3.5555,
+      "step": 15004
+    },
+    {
+      "epoch": 0.6317894736842106,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00015485547874708323,
+      "loss": 3.0339,
+      "step": 15005
+    },
+    {
+      "epoch": 0.6318315789473684,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00015482427463000786,
+      "loss": 2.9503,
+      "step": 15006
+    },
+    {
+      "epoch": 0.6318736842105264,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0001547930722469223,
+      "loss": 3.3502,
+      "step": 15007
+    },
+    {
+      "epoch": 0.6319157894736842,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00015476187159839521,
+      "loss": 3.1998,
+      "step": 15008
+    },
+    {
+      "epoch": 0.631957894736842,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0001547306726849948,
+      "loss": 3.1636,
+      "step": 15009
+    },
+    {
+      "epoch": 0.632,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00015469947550728958,
+      "loss": 2.9483,
+      "step": 15010
+    },
+    {
+      "epoch": 0.6320421052631578,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.000154668280065848,
+      "loss": 3.0869,
+      "step": 15011
+    },
+    {
+      "epoch": 0.6320842105263158,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00015463708636123831,
+      "loss": 3.1895,
+      "step": 15012
+    },
+    {
+      "epoch": 0.6321263157894736,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00015460589439402896,
+      "loss": 2.8422,
+      "step": 15013
+    },
+    {
+      "epoch": 0.6321684210526316,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00015457470416478802,
+      "loss": 3.1311,
+      "step": 15014
+    },
+    {
+      "epoch": 0.6322105263157894,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.000154543515674084,
+      "loss": 3.2495,
+      "step": 15015
+    },
+    {
+      "epoch": 0.6322526315789474,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0001545123289224848,
+      "loss": 3.2504,
+      "step": 15016
+    },
+    {
+      "epoch": 0.6322947368421052,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00015448114391055878,
+      "loss": 3.506,
+      "step": 15017
+    },
+    {
+      "epoch": 0.6323368421052632,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00015444996063887422,
+      "loss": 2.7659,
+      "step": 15018
+    },
+    {
+      "epoch": 0.632378947368421,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.000154418779107999,
+      "loss": 3.3293,
+      "step": 15019
+    },
+    {
+      "epoch": 0.632421052631579,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00015438759931850137,
+      "loss": 3.3957,
+      "step": 15020
+    },
+    {
+      "epoch": 0.6324631578947368,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00015435642127094927,
+      "loss": 2.8634,
+      "step": 15021
+    },
+    {
+      "epoch": 0.6325052631578947,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00015432524496591094,
+      "loss": 2.6752,
+      "step": 15022
+    },
+    {
+      "epoch": 0.6325473684210526,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0001542940704039541,
+      "loss": 3.2996,
+      "step": 15023
+    },
+    {
+      "epoch": 0.6325894736842105,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00015426289758564688,
+      "loss": 3.3541,
+      "step": 15024
+    },
+    {
+      "epoch": 0.6326315789473684,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00015423172651155725,
+      "loss": 3.0191,
+      "step": 15025
+    },
+    {
+      "epoch": 0.6326736842105263,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00015420055718225295,
+      "loss": 3.0934,
+      "step": 15026
+    },
+    {
+      "epoch": 0.6327157894736842,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00015416938959830198,
+      "loss": 3.5385,
+      "step": 15027
+    },
+    {
+      "epoch": 0.6327578947368421,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0001541382237602721,
+      "loss": 2.8548,
+      "step": 15028
+    },
+    {
+      "epoch": 0.6328,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00015410705966873123,
+      "loss": 3.2134,
+      "step": 15029
+    },
+    {
+      "epoch": 0.6328421052631579,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.000154075897324247,
+      "loss": 3.0927,
+      "step": 15030
+    },
+    {
+      "epoch": 0.6328842105263158,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00015404473672738718,
+      "loss": 3.1764,
+      "step": 15031
+    },
+    {
+      "epoch": 0.6329263157894737,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0001540135778787196,
+      "loss": 3.2262,
+      "step": 15032
+    },
+    {
+      "epoch": 0.6329684210526316,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0001539824207788118,
+      "loss": 3.4183,
+      "step": 15033
+    },
+    {
+      "epoch": 0.6330105263157895,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00015395126542823155,
+      "loss": 3.7791,
+      "step": 15034
+    },
+    {
+      "epoch": 0.6330526315789474,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0001539201118275463,
+      "loss": 2.9635,
+      "step": 15035
+    },
+    {
+      "epoch": 0.6330947368421053,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00015388895997732374,
+      "loss": 2.9243,
+      "step": 15036
+    },
+    {
+      "epoch": 0.6331368421052631,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00015385780987813138,
+      "loss": 2.8828,
+      "step": 15037
+    },
+    {
+      "epoch": 0.6331789473684211,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00015382666153053676,
+      "loss": 3.8644,
+      "step": 15038
+    },
+    {
+      "epoch": 0.6332210526315789,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00015379551493510747,
+      "loss": 3.1287,
+      "step": 15039
+    },
+    {
+      "epoch": 0.6332631578947369,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001537643700924108,
+      "loss": 2.9927,
+      "step": 15040
+    },
+    {
+      "epoch": 0.6333052631578947,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00015373322700301429,
+      "loss": 2.4085,
+      "step": 15041
+    },
+    {
+      "epoch": 0.6333473684210527,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00015370208566748518,
+      "loss": 2.7481,
+      "step": 15042
+    },
+    {
+      "epoch": 0.6333894736842105,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0001536709460863909,
+      "loss": 2.9193,
+      "step": 15043
+    },
+    {
+      "epoch": 0.6334315789473685,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001536398082602989,
+      "loss": 3.3618,
+      "step": 15044
+    },
+    {
+      "epoch": 0.6334736842105263,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00015360867218977634,
+      "loss": 2.6084,
+      "step": 15045
+    },
+    {
+      "epoch": 0.6335157894736843,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00015357753787539053,
+      "loss": 3.1218,
+      "step": 15046
+    },
+    {
+      "epoch": 0.6335578947368421,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00015354640531770863,
+      "loss": 2.6928,
+      "step": 15047
+    },
+    {
+      "epoch": 0.6336,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.000153515274517298,
+      "loss": 3.1263,
+      "step": 15048
+    },
+    {
+      "epoch": 0.6336421052631579,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00015348414547472562,
+      "loss": 2.7719,
+      "step": 15049
+    },
+    {
+      "epoch": 0.6336842105263157,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001534530181905587,
+      "loss": 3.3164,
+      "step": 15050
+    },
+    {
+      "epoch": 0.6337263157894737,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00015342189266536444,
+      "loss": 3.4272,
+      "step": 15051
+    },
+    {
+      "epoch": 0.6337684210526315,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001533907688997097,
+      "loss": 3.1096,
+      "step": 15052
+    },
+    {
+      "epoch": 0.6338105263157895,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00015335964689416174,
+      "loss": 3.0531,
+      "step": 15053
+    },
+    {
+      "epoch": 0.6338526315789473,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00015332852664928738,
+      "loss": 3.2097,
+      "step": 15054
+    },
+    {
+      "epoch": 0.6338947368421053,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0001532974081656538,
+      "loss": 3.1711,
+      "step": 15055
+    },
+    {
+      "epoch": 0.6339368421052631,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001532662914438277,
+      "loss": 2.9236,
+      "step": 15056
+    },
+    {
+      "epoch": 0.6339789473684211,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00015323517648437612,
+      "loss": 3.537,
+      "step": 15057
+    },
+    {
+      "epoch": 0.6340210526315789,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00015320406328786607,
+      "loss": 2.6572,
+      "step": 15058
+    },
+    {
+      "epoch": 0.6340631578947369,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0001531729518548641,
+      "loss": 3.1046,
+      "step": 15059
+    },
+    {
+      "epoch": 0.6341052631578947,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00015314184218593723,
+      "loss": 2.8916,
+      "step": 15060
+    },
+    {
+      "epoch": 0.6341473684210527,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00015311073428165222,
+      "loss": 3.1758,
+      "step": 15061
+    },
+    {
+      "epoch": 0.6341894736842105,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0001530796281425758,
+      "loss": 3.246,
+      "step": 15062
+    },
+    {
+      "epoch": 0.6342315789473684,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00015304852376927464,
+      "loss": 2.6281,
+      "step": 15063
+    },
+    {
+      "epoch": 0.6342736842105263,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00015301742116231542,
+      "loss": 2.6975,
+      "step": 15064
+    },
+    {
+      "epoch": 0.6343157894736842,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00015298632032226494,
+      "loss": 3.1861,
+      "step": 15065
+    },
+    {
+      "epoch": 0.6343578947368421,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00015295522124968963,
+      "loss": 2.699,
+      "step": 15066
+    },
+    {
+      "epoch": 0.6344,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001529241239451563,
+      "loss": 2.9479,
+      "step": 15067
+    },
+    {
+      "epoch": 0.6344421052631579,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00015289302840923124,
+      "loss": 3.2398,
+      "step": 15068
+    },
+    {
+      "epoch": 0.6344842105263158,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00015286193464248115,
+      "loss": 3.2375,
+      "step": 15069
+    },
+    {
+      "epoch": 0.6345263157894737,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00015283084264547243,
+      "loss": 3.3047,
+      "step": 15070
+    },
+    {
+      "epoch": 0.6345684210526316,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0001527997524187716,
+      "loss": 3.2854,
+      "step": 15071
+    },
+    {
+      "epoch": 0.6346105263157895,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00015276866396294518,
+      "loss": 3.3843,
+      "step": 15072
+    },
+    {
+      "epoch": 0.6346526315789474,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00015273757727855933,
+      "loss": 3.0405,
+      "step": 15073
+    },
+    {
+      "epoch": 0.6346947368421053,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00015270649236618066,
+      "loss": 3.012,
+      "step": 15074
+    },
+    {
+      "epoch": 0.6347368421052632,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0001526754092263753,
+      "loss": 3.3896,
+      "step": 15075
+    },
+    {
+      "epoch": 0.634778947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00015264432785970966,
+      "loss": 3.2394,
+      "step": 15076
+    },
+    {
+      "epoch": 0.634821052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00015261324826675,
+      "loss": 2.9098,
+      "step": 15077
+    },
+    {
+      "epoch": 0.6348631578947368,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00015258217044806244,
+      "loss": 2.7584,
+      "step": 15078
+    },
+    {
+      "epoch": 0.6349052631578948,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00015255109440421338,
+      "loss": 2.8926,
+      "step": 15079
+    },
+    {
+      "epoch": 0.6349473684210526,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0001525200201357688,
+      "loss": 2.9187,
+      "step": 15080
+    },
+    {
+      "epoch": 0.6349894736842105,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00015248894764329502,
+      "loss": 3.4695,
+      "step": 15081
+    },
+    {
+      "epoch": 0.6350315789473684,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00015245787692735795,
+      "loss": 3.0178,
+      "step": 15082
+    },
+    {
+      "epoch": 0.6350736842105263,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00015242680798852385,
+      "loss": 3.2409,
+      "step": 15083
+    },
+    {
+      "epoch": 0.6351157894736842,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00015239574082735857,
+      "loss": 3.1375,
+      "step": 15084
+    },
+    {
+      "epoch": 0.6351578947368421,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0001523646754444282,
+      "loss": 2.7571,
+      "step": 15085
+    },
+    {
+      "epoch": 0.6352,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0001523336118402988,
+      "loss": 3.3967,
+      "step": 15086
+    },
+    {
+      "epoch": 0.6352421052631579,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00015230255001553617,
+      "loss": 3.0641,
+      "step": 15087
+    },
+    {
+      "epoch": 0.6352842105263158,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0001522714899707064,
+      "loss": 3.3358,
+      "step": 15088
+    },
+    {
+      "epoch": 0.6353263157894737,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00015224043170637515,
+      "loss": 3.2893,
+      "step": 15089
+    },
+    {
+      "epoch": 0.6353684210526316,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00015220937522310846,
+      "loss": 3.352,
+      "step": 15090
+    },
+    {
+      "epoch": 0.6354105263157894,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00015217832052147196,
+      "loss": 3.5444,
+      "step": 15091
+    },
+    {
+      "epoch": 0.6354526315789474,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001521472676020315,
+      "loss": 3.227,
+      "step": 15092
+    },
+    {
+      "epoch": 0.6354947368421052,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.000152116216465353,
+      "loss": 2.5507,
+      "step": 15093
+    },
+    {
+      "epoch": 0.6355368421052632,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00015208516711200194,
+      "loss": 3.1603,
+      "step": 15094
+    },
+    {
+      "epoch": 0.635578947368421,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001520541195425441,
+      "loss": 2.8496,
+      "step": 15095
+    },
+    {
+      "epoch": 0.635621052631579,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00015202307375754504,
+      "loss": 3.0032,
+      "step": 15096
+    },
+    {
+      "epoch": 0.6356631578947368,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0001519920297575705,
+      "loss": 3.1384,
+      "step": 15097
+    },
+    {
+      "epoch": 0.6357052631578948,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00015196098754318612,
+      "loss": 2.8778,
+      "step": 15098
+    },
+    {
+      "epoch": 0.6357473684210526,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001519299471149573,
+      "loss": 3.2905,
+      "step": 15099
+    },
+    {
+      "epoch": 0.6357894736842106,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00015189890847344965,
+      "loss": 2.9913,
+      "step": 15100
+    },
+    {
+      "epoch": 0.6358315789473684,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00015186787161922854,
+      "loss": 2.9145,
+      "step": 15101
+    },
+    {
+      "epoch": 0.6358736842105264,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00015183683655285956,
+      "loss": 2.7434,
+      "step": 15102
+    },
+    {
+      "epoch": 0.6359157894736842,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00015180580327490804,
+      "loss": 2.5147,
+      "step": 15103
+    },
+    {
+      "epoch": 0.635957894736842,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00015177477178593947,
+      "loss": 3.2603,
+      "step": 15104
+    },
+    {
+      "epoch": 0.636,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0001517437420865191,
+      "loss": 2.9065,
+      "step": 15105
+    },
+    {
+      "epoch": 0.6360421052631579,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0001517127141772123,
+      "loss": 2.8908,
+      "step": 15106
+    },
+    {
+      "epoch": 0.6360842105263158,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00015168168805858445,
+      "loss": 3.0568,
+      "step": 15107
+    },
+    {
+      "epoch": 0.6361263157894737,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00015165066373120062,
+      "loss": 2.9242,
+      "step": 15108
+    },
+    {
+      "epoch": 0.6361684210526316,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00015161964119562627,
+      "loss": 3.2599,
+      "step": 15109
+    },
+    {
+      "epoch": 0.6362105263157894,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00015158862045242633,
+      "loss": 2.648,
+      "step": 15110
+    },
+    {
+      "epoch": 0.6362526315789474,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001515576015021661,
+      "loss": 3.2012,
+      "step": 15111
+    },
+    {
+      "epoch": 0.6362947368421052,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0001515265843454108,
+      "loss": 3.125,
+      "step": 15112
+    },
+    {
+      "epoch": 0.6363368421052632,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00015149556898272533,
+      "loss": 3.186,
+      "step": 15113
+    },
+    {
+      "epoch": 0.636378947368421,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00015146455541467497,
+      "loss": 3.1077,
+      "step": 15114
+    },
+    {
+      "epoch": 0.636421052631579,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00015143354364182454,
+      "loss": 3.2735,
+      "step": 15115
+    },
+    {
+      "epoch": 0.6364631578947368,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00015140253366473922,
+      "loss": 3.1799,
+      "step": 15116
+    },
+    {
+      "epoch": 0.6365052631578947,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0001513715254839838,
+      "loss": 3.1229,
+      "step": 15117
+    },
+    {
+      "epoch": 0.6365473684210526,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00015134051910012332,
+      "loss": 3.0195,
+      "step": 15118
+    },
+    {
+      "epoch": 0.6365894736842105,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001513095145137227,
+      "loss": 2.8513,
+      "step": 15119
+    },
+    {
+      "epoch": 0.6366315789473684,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00015127851172534675,
+      "loss": 3.0837,
+      "step": 15120
+    },
+    {
+      "epoch": 0.6366736842105263,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001512475107355603,
+      "loss": 3.4265,
+      "step": 15121
+    },
+    {
+      "epoch": 0.6367157894736842,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00015121651154492819,
+      "loss": 3.086,
+      "step": 15122
+    },
+    {
+      "epoch": 0.6367578947368421,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00015118551415401526,
+      "loss": 3.1515,
+      "step": 15123
+    },
+    {
+      "epoch": 0.6368,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00015115451856338602,
+      "loss": 3.2297,
+      "step": 15124
+    },
+    {
+      "epoch": 0.6368421052631579,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00015112352477360537,
+      "loss": 2.9911,
+      "step": 15125
+    },
+    {
+      "epoch": 0.6368842105263158,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.000151092532785238,
+      "loss": 3.0715,
+      "step": 15126
+    },
+    {
+      "epoch": 0.6369263157894737,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00015106154259884836,
+      "loss": 3.424,
+      "step": 15127
+    },
+    {
+      "epoch": 0.6369684210526316,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001510305542150012,
+      "loss": 3.0849,
+      "step": 15128
+    },
+    {
+      "epoch": 0.6370105263157895,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00015099956763426104,
+      "loss": 3.1155,
+      "step": 15129
+    },
+    {
+      "epoch": 0.6370526315789473,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0001509685828571925,
+      "loss": 3.1721,
+      "step": 15130
+    },
+    {
+      "epoch": 0.6370947368421053,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00015093759988435995,
+      "loss": 3.3245,
+      "step": 15131
+    },
+    {
+      "epoch": 0.6371368421052631,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00015090661871632795,
+      "loss": 3.0829,
+      "step": 15132
+    },
+    {
+      "epoch": 0.6371789473684211,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.000150875639353661,
+      "loss": 3.1264,
+      "step": 15133
+    },
+    {
+      "epoch": 0.6372210526315789,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00015084466179692335,
+      "loss": 3.521,
+      "step": 15134
+    },
+    {
+      "epoch": 0.6372631578947369,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00015081368604667956,
+      "loss": 2.6497,
+      "step": 15135
+    },
+    {
+      "epoch": 0.6373052631578947,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00015078271210349376,
+      "loss": 3.2965,
+      "step": 15136
+    },
+    {
+      "epoch": 0.6373473684210527,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00015075173996793043,
+      "loss": 3.0335,
+      "step": 15137
+    },
+    {
+      "epoch": 0.6373894736842105,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00015072076964055373,
+      "loss": 3.0598,
+      "step": 15138
+    },
+    {
+      "epoch": 0.6374315789473685,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00015068980112192796,
+      "loss": 2.8235,
+      "step": 15139
+    },
+    {
+      "epoch": 0.6374736842105263,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00015065883441261744,
+      "loss": 3.0912,
+      "step": 15140
+    },
+    {
+      "epoch": 0.6375157894736843,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00015062786951318614,
+      "loss": 3.219,
+      "step": 15141
+    },
+    {
+      "epoch": 0.6375578947368421,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0001505969064241984,
+      "loss": 3.3337,
+      "step": 15142
+    },
+    {
+      "epoch": 0.6376,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00015056594514621806,
+      "loss": 3.2448,
+      "step": 15143
+    },
+    {
+      "epoch": 0.6376421052631579,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00015053498567980944,
+      "loss": 3.1603,
+      "step": 15144
+    },
+    {
+      "epoch": 0.6376842105263157,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00015050402802553657,
+      "loss": 3.5756,
+      "step": 15145
+    },
+    {
+      "epoch": 0.6377263157894737,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00015047307218396332,
+      "loss": 3.5932,
+      "step": 15146
+    },
+    {
+      "epoch": 0.6377684210526315,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0001504421181556539,
+      "loss": 3.2706,
+      "step": 15147
+    },
+    {
+      "epoch": 0.6378105263157895,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00015041116594117197,
+      "loss": 2.8318,
+      "step": 15148
+    },
+    {
+      "epoch": 0.6378526315789473,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0001503802155410817,
+      "loss": 3.4171,
+      "step": 15149
+    },
+    {
+      "epoch": 0.6378947368421053,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00015034926695594675,
+      "loss": 3.0776,
+      "step": 15150
+    },
+    {
+      "epoch": 0.6379368421052631,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00015031832018633107,
+      "loss": 2.8747,
+      "step": 15151
+    },
+    {
+      "epoch": 0.6379789473684211,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00015028737523279856,
+      "loss": 3.0641,
+      "step": 15152
+    },
+    {
+      "epoch": 0.6380210526315789,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00015025643209591285,
+      "loss": 3.1129,
+      "step": 15153
+    },
+    {
+      "epoch": 0.6380631578947369,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0001502254907762378,
+      "loss": 3.2763,
+      "step": 15154
+    },
+    {
+      "epoch": 0.6381052631578947,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.000150194551274337,
+      "loss": 3.1304,
+      "step": 15155
+    },
+    {
+      "epoch": 0.6381473684210527,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00015016361359077436,
+      "loss": 3.127,
+      "step": 15156
+    },
+    {
+      "epoch": 0.6381894736842105,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00015013267772611322,
+      "loss": 2.8736,
+      "step": 15157
+    },
+    {
+      "epoch": 0.6382315789473684,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001501017436809174,
+      "loss": 3.3996,
+      "step": 15158
+    },
+    {
+      "epoch": 0.6382736842105263,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00015007081145575047,
+      "loss": 3.4592,
+      "step": 15159
+    },
+    {
+      "epoch": 0.6383157894736842,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0001500398810511759,
+      "loss": 2.9642,
+      "step": 15160
+    },
+    {
+      "epoch": 0.6383578947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00015000895246775728,
+      "loss": 3.1632,
+      "step": 15161
+    },
+    {
+      "epoch": 0.6384,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00014997802570605807,
+      "loss": 2.8832,
+      "step": 15162
+    },
+    {
+      "epoch": 0.6384421052631579,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001499471007666417,
+      "loss": 3.3844,
+      "step": 15163
+    },
+    {
+      "epoch": 0.6384842105263158,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00014991617765007154,
+      "loss": 2.8697,
+      "step": 15164
+    },
+    {
+      "epoch": 0.6385263157894737,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00014988525635691106,
+      "loss": 2.8038,
+      "step": 15165
+    },
+    {
+      "epoch": 0.6385684210526316,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00014985433688772366,
+      "loss": 3.1418,
+      "step": 15166
+    },
+    {
+      "epoch": 0.6386105263157895,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00014982341924307246,
+      "loss": 3.5013,
+      "step": 15167
+    },
+    {
+      "epoch": 0.6386526315789474,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00014979250342352096,
+      "loss": 3.1737,
+      "step": 15168
+    },
+    {
+      "epoch": 0.6386947368421053,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00014976158942963224,
+      "loss": 3.0077,
+      "step": 15169
+    },
+    {
+      "epoch": 0.6387368421052632,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00014973067726196961,
+      "loss": 3.2396,
+      "step": 15170
+    },
+    {
+      "epoch": 0.638778947368421,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00014969976692109623,
+      "loss": 2.9719,
+      "step": 15171
+    },
+    {
+      "epoch": 0.638821052631579,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00014966885840757516,
+      "loss": 3.2424,
+      "step": 15172
+    },
+    {
+      "epoch": 0.6388631578947368,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00014963795172196975,
+      "loss": 3.2607,
+      "step": 15173
+    },
+    {
+      "epoch": 0.6389052631578948,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00014960704686484282,
+      "loss": 2.8867,
+      "step": 15174
+    },
+    {
+      "epoch": 0.6389473684210526,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001495761438367577,
+      "loss": 3.5469,
+      "step": 15175
+    },
+    {
+      "epoch": 0.6389894736842106,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0001495452426382771,
+      "loss": 3.2575,
+      "step": 15176
+    },
+    {
+      "epoch": 0.6390315789473684,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00014951434326996422,
+      "loss": 3.4603,
+      "step": 15177
+    },
+    {
+      "epoch": 0.6390736842105263,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00014948344573238192,
+      "loss": 3.3069,
+      "step": 15178
+    },
+    {
+      "epoch": 0.6391157894736842,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0001494525500260931,
+      "loss": 2.7507,
+      "step": 15179
+    },
+    {
+      "epoch": 0.6391578947368421,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0001494216561516607,
+      "loss": 3.0808,
+      "step": 15180
+    },
+    {
+      "epoch": 0.6392,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00014939076410964752,
+      "loss": 3.2244,
+      "step": 15181
+    },
+    {
+      "epoch": 0.639242105263158,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0001493598739006165,
+      "loss": 3.05,
+      "step": 15182
+    },
+    {
+      "epoch": 0.6392842105263158,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00014932898552513026,
+      "loss": 3.4507,
+      "step": 15183
+    },
+    {
+      "epoch": 0.6393263157894736,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0001492980989837517,
+      "loss": 3.3,
+      "step": 15184
+    },
+    {
+      "epoch": 0.6393684210526316,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00014926721427704334,
+      "loss": 2.7179,
+      "step": 15185
+    },
+    {
+      "epoch": 0.6394105263157894,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.000149236331405568,
+      "loss": 2.6776,
+      "step": 15186
+    },
+    {
+      "epoch": 0.6394526315789474,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00014920545036988837,
+      "loss": 2.9292,
+      "step": 15187
+    },
+    {
+      "epoch": 0.6394947368421052,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00014917457117056694,
+      "loss": 3.1369,
+      "step": 15188
+    },
+    {
+      "epoch": 0.6395368421052632,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00014914369380816645,
+      "loss": 3.387,
+      "step": 15189
+    },
+    {
+      "epoch": 0.639578947368421,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00014911281828324926,
+      "loss": 3.0294,
+      "step": 15190
+    },
+    {
+      "epoch": 0.639621052631579,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.000149081944596378,
+      "loss": 3.108,
+      "step": 15191
+    },
+    {
+      "epoch": 0.6396631578947368,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0001490510727481152,
+      "loss": 2.8377,
+      "step": 15192
+    },
+    {
+      "epoch": 0.6397052631578948,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0001490202027390232,
+      "loss": 3.5374,
+      "step": 15193
+    },
+    {
+      "epoch": 0.6397473684210526,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00014898933456966451,
+      "loss": 3.141,
+      "step": 15194
+    },
+    {
+      "epoch": 0.6397894736842106,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00014895846824060138,
+      "loss": 2.8468,
+      "step": 15195
+    },
+    {
+      "epoch": 0.6398315789473684,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001489276037523963,
+      "loss": 3.2337,
+      "step": 15196
+    },
+    {
+      "epoch": 0.6398736842105263,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00014889674110561146,
+      "loss": 3.1992,
+      "step": 15197
+    },
+    {
+      "epoch": 0.6399157894736842,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00014886588030080922,
+      "loss": 3.1638,
+      "step": 15198
+    },
+    {
+      "epoch": 0.6399578947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00014883502133855193,
+      "loss": 2.7004,
+      "step": 15199
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00014880416421940155,
+      "loss": 3.3718,
+      "step": 15200
+    },
+    {
+      "epoch": 0.6400421052631579,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001487733089439205,
+      "loss": 3.0877,
+      "step": 15201
+    },
+    {
+      "epoch": 0.6400842105263158,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00014874245551267073,
+      "loss": 3.2789,
+      "step": 15202
+    },
+    {
+      "epoch": 0.6401263157894737,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00014871160392621453,
+      "loss": 3.0484,
+      "step": 15203
+    },
+    {
+      "epoch": 0.6401684210526316,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0001486807541851138,
+      "loss": 3.4268,
+      "step": 15204
+    },
+    {
+      "epoch": 0.6402105263157895,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00014864990628993074,
+      "loss": 3.0317,
+      "step": 15205
+    },
+    {
+      "epoch": 0.6402526315789474,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00014861906024122735,
+      "loss": 3.4472,
+      "step": 15206
+    },
+    {
+      "epoch": 0.6402947368421052,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00014858821603956545,
+      "loss": 3.11,
+      "step": 15207
+    },
+    {
+      "epoch": 0.6403368421052632,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00014855737368550727,
+      "loss": 3.1971,
+      "step": 15208
+    },
+    {
+      "epoch": 0.640378947368421,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0001485265331796144,
+      "loss": 3.0998,
+      "step": 15209
+    },
+    {
+      "epoch": 0.640421052631579,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.000148495694522449,
+      "loss": 3.3118,
+      "step": 15210
+    },
+    {
+      "epoch": 0.6404631578947368,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00014846485771457263,
+      "loss": 2.8374,
+      "step": 15211
+    },
+    {
+      "epoch": 0.6405052631578947,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0001484340227565473,
+      "loss": 3.1539,
+      "step": 15212
+    },
+    {
+      "epoch": 0.6405473684210526,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0001484031896489348,
+      "loss": 3.1276,
+      "step": 15213
+    },
+    {
+      "epoch": 0.6405894736842105,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00014837235839229674,
+      "loss": 3.1338,
+      "step": 15214
+    },
+    {
+      "epoch": 0.6406315789473684,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.000148341528987195,
+      "loss": 2.9809,
+      "step": 15215
+    },
+    {
+      "epoch": 0.6406736842105263,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00014831070143419105,
+      "loss": 3.1864,
+      "step": 15216
+    },
+    {
+      "epoch": 0.6407157894736842,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00014827987573384676,
+      "loss": 3.3738,
+      "step": 15217
+    },
+    {
+      "epoch": 0.6407578947368421,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0001482490518867235,
+      "loss": 3.0045,
+      "step": 15218
+    },
+    {
+      "epoch": 0.6408,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00014821822989338296,
+      "loss": 2.6944,
+      "step": 15219
+    },
+    {
+      "epoch": 0.6408421052631579,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00014818740975438676,
+      "loss": 3.13,
+      "step": 15220
+    },
+    {
+      "epoch": 0.6408842105263158,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0001481565914702963,
+      "loss": 3.1615,
+      "step": 15221
+    },
+    {
+      "epoch": 0.6409263157894737,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00014812577504167313,
+      "loss": 3.1618,
+      "step": 15222
+    },
+    {
+      "epoch": 0.6409684210526316,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00014809496046907855,
+      "loss": 2.9308,
+      "step": 15223
+    },
+    {
+      "epoch": 0.6410105263157895,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00014806414775307418,
+      "loss": 2.7885,
+      "step": 15224
+    },
+    {
+      "epoch": 0.6410526315789473,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00014803333689422115,
+      "loss": 3.4629,
+      "step": 15225
+    },
+    {
+      "epoch": 0.6410947368421053,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00014800252789308092,
+      "loss": 3.203,
+      "step": 15226
+    },
+    {
+      "epoch": 0.6411368421052631,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00014797172075021488,
+      "loss": 2.954,
+      "step": 15227
+    },
+    {
+      "epoch": 0.6411789473684211,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00014794091546618414,
+      "loss": 3.2884,
+      "step": 15228
+    },
+    {
+      "epoch": 0.6412210526315789,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00014791011204155,
+      "loss": 3.1616,
+      "step": 15229
+    },
+    {
+      "epoch": 0.6412631578947369,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0001478793104768737,
+      "loss": 3.2868,
+      "step": 15230
+    },
+    {
+      "epoch": 0.6413052631578947,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0001478485107727164,
+      "loss": 3.31,
+      "step": 15231
+    },
+    {
+      "epoch": 0.6413473684210527,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00014781771292963914,
+      "loss": 3.2528,
+      "step": 15232
+    },
+    {
+      "epoch": 0.6413894736842105,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00014778691694820311,
+      "loss": 3.1324,
+      "step": 15233
+    },
+    {
+      "epoch": 0.6414315789473685,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00014775612282896943,
+      "loss": 3.1566,
+      "step": 15234
+    },
+    {
+      "epoch": 0.6414736842105263,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00014772533057249898,
+      "loss": 3.5487,
+      "step": 15235
+    },
+    {
+      "epoch": 0.6415157894736843,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00014769454017935285,
+      "loss": 3.2191,
+      "step": 15236
+    },
+    {
+      "epoch": 0.6415578947368421,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00014766375165009205,
+      "loss": 2.8905,
+      "step": 15237
+    },
+    {
+      "epoch": 0.6416,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00014763296498527742,
+      "loss": 3.0981,
+      "step": 15238
+    },
+    {
+      "epoch": 0.6416421052631579,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00014760218018546983,
+      "loss": 3.1496,
+      "step": 15239
+    },
+    {
+      "epoch": 0.6416842105263157,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00014757139725123024,
+      "loss": 3.0404,
+      "step": 15240
+    },
+    {
+      "epoch": 0.6417263157894737,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00014754061618311954,
+      "loss": 3.2482,
+      "step": 15241
+    },
+    {
+      "epoch": 0.6417684210526315,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00014750983698169832,
+      "loss": 2.8224,
+      "step": 15242
+    },
+    {
+      "epoch": 0.6418105263157895,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00014747905964752756,
+      "loss": 3.2683,
+      "step": 15243
+    },
+    {
+      "epoch": 0.6418526315789473,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00014744828418116775,
+      "loss": 2.7334,
+      "step": 15244
+    },
+    {
+      "epoch": 0.6418947368421053,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00014741751058317971,
+      "loss": 3.0247,
+      "step": 15245
+    },
+    {
+      "epoch": 0.6419368421052631,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00014738673885412415,
+      "loss": 3.3357,
+      "step": 15246
+    },
+    {
+      "epoch": 0.6419789473684211,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00014735596899456162,
+      "loss": 3.3715,
+      "step": 15247
+    },
+    {
+      "epoch": 0.6420210526315789,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00014732520100505282,
+      "loss": 2.9668,
+      "step": 15248
+    },
+    {
+      "epoch": 0.6420631578947369,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001472944348861581,
+      "loss": 3.1655,
+      "step": 15249
+    },
+    {
+      "epoch": 0.6421052631578947,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00014726367063843822,
+      "loss": 3.1824,
+      "step": 15250
+    },
+    {
+      "epoch": 0.6421473684210526,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00014723290826245344,
+      "loss": 3.167,
+      "step": 15251
+    },
+    {
+      "epoch": 0.6421894736842105,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00014720214775876435,
+      "loss": 2.9136,
+      "step": 15252
+    },
+    {
+      "epoch": 0.6422315789473684,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001471713891279314,
+      "loss": 3.1543,
+      "step": 15253
+    },
+    {
+      "epoch": 0.6422736842105263,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00014714063237051484,
+      "loss": 3.1397,
+      "step": 15254
+    },
+    {
+      "epoch": 0.6423157894736842,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00014710987748707517,
+      "loss": 2.9942,
+      "step": 15255
+    },
+    {
+      "epoch": 0.6423578947368421,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00014707912447817258,
+      "loss": 3.2469,
+      "step": 15256
+    },
+    {
+      "epoch": 0.6424,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00014704837334436754,
+      "loss": 3.1904,
+      "step": 15257
+    },
+    {
+      "epoch": 0.6424421052631579,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00014701762408622006,
+      "loss": 2.6426,
+      "step": 15258
+    },
+    {
+      "epoch": 0.6424842105263158,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00014698687670429045,
+      "loss": 2.8527,
+      "step": 15259
+    },
+    {
+      "epoch": 0.6425263157894737,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00014695613119913904,
+      "loss": 3.5967,
+      "step": 15260
+    },
+    {
+      "epoch": 0.6425684210526316,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0001469253875713257,
+      "loss": 3.0684,
+      "step": 15261
+    },
+    {
+      "epoch": 0.6426105263157895,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00014689464582141076,
+      "loss": 3.228,
+      "step": 15262
+    },
+    {
+      "epoch": 0.6426526315789474,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00014686390594995422,
+      "loss": 2.8606,
+      "step": 15263
+    },
+    {
+      "epoch": 0.6426947368421053,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0001468331679575162,
+      "loss": 3.2451,
+      "step": 15264
+    },
+    {
+      "epoch": 0.6427368421052632,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00014680243184465655,
+      "loss": 3.2359,
+      "step": 15265
+    },
+    {
+      "epoch": 0.642778947368421,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0001467716976119353,
+      "loss": 3.17,
+      "step": 15266
+    },
+    {
+      "epoch": 0.642821052631579,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001467409652599126,
+      "loss": 3.0633,
+      "step": 15267
+    },
+    {
+      "epoch": 0.6428631578947368,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00014671023478914807,
+      "loss": 3.0597,
+      "step": 15268
+    },
+    {
+      "epoch": 0.6429052631578948,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0001466795062002018,
+      "loss": 3.3747,
+      "step": 15269
+    },
+    {
+      "epoch": 0.6429473684210526,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00014664877949363341,
+      "loss": 2.9471,
+      "step": 15270
+    },
+    {
+      "epoch": 0.6429894736842106,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0001466180546700029,
+      "loss": 3.156,
+      "step": 15271
+    },
+    {
+      "epoch": 0.6430315789473684,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0001465873317298699,
+      "loss": 3.1558,
+      "step": 15272
+    },
+    {
+      "epoch": 0.6430736842105264,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001465566106737942,
+      "loss": 3.3786,
+      "step": 15273
+    },
+    {
+      "epoch": 0.6431157894736842,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0001465258915023356,
+      "loss": 2.8449,
+      "step": 15274
+    },
+    {
+      "epoch": 0.6431578947368422,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00014649517421605362,
+      "loss": 2.5784,
+      "step": 15275
+    },
+    {
+      "epoch": 0.6432,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00014646445881550803,
+      "loss": 3.403,
+      "step": 15276
+    },
+    {
+      "epoch": 0.643242105263158,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00014643374530125825,
+      "loss": 3.2902,
+      "step": 15277
+    },
+    {
+      "epoch": 0.6432842105263158,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00014640303367386405,
+      "loss": 2.8078,
+      "step": 15278
+    },
+    {
+      "epoch": 0.6433263157894736,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0001463723239338847,
+      "loss": 2.9995,
+      "step": 15279
+    },
+    {
+      "epoch": 0.6433684210526316,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00014634161608187998,
+      "loss": 3.4704,
+      "step": 15280
+    },
+    {
+      "epoch": 0.6434105263157894,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00014631091011840918,
+      "loss": 3.1849,
+      "step": 15281
+    },
+    {
+      "epoch": 0.6434526315789474,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00014628020604403176,
+      "loss": 2.7016,
+      "step": 15282
+    },
+    {
+      "epoch": 0.6434947368421052,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00014624950385930718,
+      "loss": 2.9955,
+      "step": 15283
+    },
+    {
+      "epoch": 0.6435368421052632,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00014621880356479466,
+      "loss": 3.0802,
+      "step": 15284
+    },
+    {
+      "epoch": 0.643578947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00014618810516105372,
+      "loss": 3.0921,
+      "step": 15285
+    },
+    {
+      "epoch": 0.643621052631579,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001461574086486434,
+      "loss": 3.0386,
+      "step": 15286
+    },
+    {
+      "epoch": 0.6436631578947368,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00014612671402812312,
+      "loss": 3.3815,
+      "step": 15287
+    },
+    {
+      "epoch": 0.6437052631578948,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00014609602130005207,
+      "loss": 2.9637,
+      "step": 15288
+    },
+    {
+      "epoch": 0.6437473684210526,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00014606533046498938,
+      "loss": 2.6867,
+      "step": 15289
+    },
+    {
+      "epoch": 0.6437894736842106,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001460346415234944,
+      "loss": 2.9471,
+      "step": 15290
+    },
+    {
+      "epoch": 0.6438315789473684,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00014600395447612595,
+      "loss": 3.2734,
+      "step": 15291
+    },
+    {
+      "epoch": 0.6438736842105263,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00014597326932344325,
+      "loss": 3.2952,
+      "step": 15292
+    },
+    {
+      "epoch": 0.6439157894736842,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00014594258606600547,
+      "loss": 3.4915,
+      "step": 15293
+    },
+    {
+      "epoch": 0.6439578947368421,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00014591190470437137,
+      "loss": 3.253,
+      "step": 15294
+    },
+    {
+      "epoch": 0.644,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00014588122523910031,
+      "loss": 2.4031,
+      "step": 15295
+    },
+    {
+      "epoch": 0.6440421052631579,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00014585054767075073,
+      "loss": 2.8718,
+      "step": 15296
+    },
+    {
+      "epoch": 0.6440842105263158,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.000145819871999882,
+      "loss": 2.9707,
+      "step": 15297
+    },
+    {
+      "epoch": 0.6441263157894737,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00014578919822705262,
+      "loss": 3.4652,
+      "step": 15298
+    },
+    {
+      "epoch": 0.6441684210526316,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00014575852635282172,
+      "loss": 3.3701,
+      "step": 15299
+    },
+    {
+      "epoch": 0.6442105263157895,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00014572785637774794,
+      "loss": 3.4632,
+      "step": 15300
+    },
+    {
+      "epoch": 0.6442526315789474,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00014569718830239014,
+      "loss": 3.4888,
+      "step": 15301
+    },
+    {
+      "epoch": 0.6442947368421053,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00014566652212730702,
+      "loss": 2.5655,
+      "step": 15302
+    },
+    {
+      "epoch": 0.6443368421052632,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00014563585785305717,
+      "loss": 3.2854,
+      "step": 15303
+    },
+    {
+      "epoch": 0.644378947368421,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00014560519548019956,
+      "loss": 2.5697,
+      "step": 15304
+    },
+    {
+      "epoch": 0.6444210526315789,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00014557453500929247,
+      "loss": 3.1477,
+      "step": 15305
+    },
+    {
+      "epoch": 0.6444631578947368,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0001455438764408947,
+      "loss": 2.4591,
+      "step": 15306
+    },
+    {
+      "epoch": 0.6445052631578947,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001455132197755648,
+      "loss": 2.9453,
+      "step": 15307
+    },
+    {
+      "epoch": 0.6445473684210526,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00014548256501386125,
+      "loss": 3.0365,
+      "step": 15308
+    },
+    {
+      "epoch": 0.6445894736842105,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0001454519121563426,
+      "loss": 2.9882,
+      "step": 15309
+    },
+    {
+      "epoch": 0.6446315789473684,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00014542126120356715,
+      "loss": 2.8996,
+      "step": 15310
+    },
+    {
+      "epoch": 0.6446736842105263,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00014539061215609365,
+      "loss": 2.7953,
+      "step": 15311
+    },
+    {
+      "epoch": 0.6447157894736842,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00014535996501448007,
+      "loss": 3.2627,
+      "step": 15312
+    },
+    {
+      "epoch": 0.6447578947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0001453293197792851,
+      "loss": 3.2114,
+      "step": 15313
+    },
+    {
+      "epoch": 0.6448,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00014529867645106693,
+      "loss": 2.8209,
+      "step": 15314
+    },
+    {
+      "epoch": 0.6448421052631579,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00014526803503038383,
+      "loss": 3.4961,
+      "step": 15315
+    },
+    {
+      "epoch": 0.6448842105263158,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00014523739551779413,
+      "loss": 2.7784,
+      "step": 15316
+    },
+    {
+      "epoch": 0.6449263157894737,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00014520675791385596,
+      "loss": 2.8801,
+      "step": 15317
+    },
+    {
+      "epoch": 0.6449684210526315,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00014517612221912754,
+      "loss": 3.4941,
+      "step": 15318
+    },
+    {
+      "epoch": 0.6450105263157895,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00014514548843416693,
+      "loss": 3.218,
+      "step": 15319
+    },
+    {
+      "epoch": 0.6450526315789473,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00014511485655953238,
+      "loss": 3.5963,
+      "step": 15320
+    },
+    {
+      "epoch": 0.6450947368421053,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00014508422659578193,
+      "loss": 3.0655,
+      "step": 15321
+    },
+    {
+      "epoch": 0.6451368421052631,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0001450535985434736,
+      "loss": 2.7694,
+      "step": 15322
+    },
+    {
+      "epoch": 0.6451789473684211,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00014502297240316542,
+      "loss": 2.6437,
+      "step": 15323
+    },
+    {
+      "epoch": 0.6452210526315789,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0001449923481754153,
+      "loss": 3.6341,
+      "step": 15324
+    },
+    {
+      "epoch": 0.6452631578947369,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00014496172586078124,
+      "loss": 2.7572,
+      "step": 15325
+    },
+    {
+      "epoch": 0.6453052631578947,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00014493110545982102,
+      "loss": 3.0701,
+      "step": 15326
+    },
+    {
+      "epoch": 0.6453473684210527,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0001449004869730927,
+      "loss": 3.0578,
+      "step": 15327
+    },
+    {
+      "epoch": 0.6453894736842105,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.000144869870401154,
+      "loss": 2.7235,
+      "step": 15328
+    },
+    {
+      "epoch": 0.6454315789473685,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0001448392557445628,
+      "loss": 3.0111,
+      "step": 15329
+    },
+    {
+      "epoch": 0.6454736842105263,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00014480864300387675,
+      "loss": 3.0969,
+      "step": 15330
+    },
+    {
+      "epoch": 0.6455157894736843,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00014477803217965364,
+      "loss": 3.1876,
+      "step": 15331
+    },
+    {
+      "epoch": 0.6455578947368421,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00014474742327245116,
+      "loss": 3.6041,
+      "step": 15332
+    },
+    {
+      "epoch": 0.6456,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00014471681628282695,
+      "loss": 2.6802,
+      "step": 15333
+    },
+    {
+      "epoch": 0.6456421052631579,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00014468621121133856,
+      "loss": 3.4278,
+      "step": 15334
+    },
+    {
+      "epoch": 0.6456842105263157,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00014465560805854377,
+      "loss": 2.8994,
+      "step": 15335
+    },
+    {
+      "epoch": 0.6457263157894737,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00014462500682500002,
+      "loss": 3.6074,
+      "step": 15336
+    },
+    {
+      "epoch": 0.6457684210526315,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001445944075112648,
+      "loss": 3.3545,
+      "step": 15337
+    },
+    {
+      "epoch": 0.6458105263157895,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00014456381011789565,
+      "loss": 2.6238,
+      "step": 15338
+    },
+    {
+      "epoch": 0.6458526315789473,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00014453321464544993,
+      "loss": 3.2071,
+      "step": 15339
+    },
+    {
+      "epoch": 0.6458947368421053,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001445026210944853,
+      "loss": 3.1965,
+      "step": 15340
+    },
+    {
+      "epoch": 0.6459368421052631,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00014447202946555877,
+      "loss": 3.0988,
+      "step": 15341
+    },
+    {
+      "epoch": 0.6459789473684211,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0001444414397592279,
+      "loss": 2.9545,
+      "step": 15342
+    },
+    {
+      "epoch": 0.6460210526315789,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00014441085197605005,
+      "loss": 3.3785,
+      "step": 15343
+    },
+    {
+      "epoch": 0.6460631578947369,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00014438026611658238,
+      "loss": 3.1818,
+      "step": 15344
+    },
+    {
+      "epoch": 0.6461052631578947,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0001443496821813822,
+      "loss": 3.1631,
+      "step": 15345
+    },
+    {
+      "epoch": 0.6461473684210526,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00014431910017100654,
+      "loss": 2.6291,
+      "step": 15346
+    },
+    {
+      "epoch": 0.6461894736842105,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00014428852008601292,
+      "loss": 3.2387,
+      "step": 15347
+    },
+    {
+      "epoch": 0.6462315789473684,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00014425794192695807,
+      "loss": 3.0379,
+      "step": 15348
+    },
+    {
+      "epoch": 0.6462736842105263,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00014422736569439946,
+      "loss": 2.9165,
+      "step": 15349
+    },
+    {
+      "epoch": 0.6463157894736842,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00014419679138889376,
+      "loss": 3.3941,
+      "step": 15350
+    },
+    {
+      "epoch": 0.6463578947368421,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00014416621901099832,
+      "loss": 2.898,
+      "step": 15351
+    },
+    {
+      "epoch": 0.6464,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00014413564856127003,
+      "loss": 3.0999,
+      "step": 15352
+    },
+    {
+      "epoch": 0.6464421052631579,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00014410508004026573,
+      "loss": 2.6489,
+      "step": 15353
+    },
+    {
+      "epoch": 0.6464842105263158,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00014407451344854264,
+      "loss": 3.4549,
+      "step": 15354
+    },
+    {
+      "epoch": 0.6465263157894737,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00014404394878665726,
+      "loss": 3.2674,
+      "step": 15355
+    },
+    {
+      "epoch": 0.6465684210526316,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00014401338605516683,
+      "loss": 3.0625,
+      "step": 15356
+    },
+    {
+      "epoch": 0.6466105263157895,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0001439828252546278,
+      "loss": 3.2687,
+      "step": 15357
+    },
+    {
+      "epoch": 0.6466526315789474,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001439522663855972,
+      "loss": 2.9323,
+      "step": 15358
+    },
+    {
+      "epoch": 0.6466947368421052,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00014392170944863177,
+      "loss": 3.2047,
+      "step": 15359
+    },
+    {
+      "epoch": 0.6467368421052632,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0001438911544442881,
+      "loss": 3.2014,
+      "step": 15360
+    },
+    {
+      "epoch": 0.646778947368421,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00014386060137312295,
+      "loss": 3.1416,
+      "step": 15361
+    },
+    {
+      "epoch": 0.646821052631579,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001438300502356928,
+      "loss": 3.29,
+      "step": 15362
+    },
+    {
+      "epoch": 0.6468631578947368,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00014379950103255462,
+      "loss": 3.3051,
+      "step": 15363
+    },
+    {
+      "epoch": 0.6469052631578948,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00014376895376426454,
+      "loss": 3.3226,
+      "step": 15364
+    },
+    {
+      "epoch": 0.6469473684210526,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00014373840843137954,
+      "loss": 3.1592,
+      "step": 15365
+    },
+    {
+      "epoch": 0.6469894736842106,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00014370786503445563,
+      "loss": 3.341,
+      "step": 15366
+    },
+    {
+      "epoch": 0.6470315789473684,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00014367732357404968,
+      "loss": 3.427,
+      "step": 15367
+    },
+    {
+      "epoch": 0.6470736842105264,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0001436467840507179,
+      "loss": 2.9153,
+      "step": 15368
+    },
+    {
+      "epoch": 0.6471157894736842,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001436162464650167,
+      "loss": 3.094,
+      "step": 15369
+    },
+    {
+      "epoch": 0.6471578947368422,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001435857108175027,
+      "loss": 3.1258,
+      "step": 15370
+    },
+    {
+      "epoch": 0.6472,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00014355517710873183,
+      "loss": 3.1314,
+      "step": 15371
+    },
+    {
+      "epoch": 0.6472421052631578,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00014352464533926075,
+      "loss": 3.3219,
+      "step": 15372
+    },
+    {
+      "epoch": 0.6472842105263158,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0001434941155096453,
+      "loss": 3.1367,
+      "step": 15373
+    },
+    {
+      "epoch": 0.6473263157894736,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00014346358762044206,
+      "loss": 3.3943,
+      "step": 15374
+    },
+    {
+      "epoch": 0.6473684210526316,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00014343306167220708,
+      "loss": 2.918,
+      "step": 15375
+    },
+    {
+      "epoch": 0.6474105263157894,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00014340253766549648,
+      "loss": 2.3796,
+      "step": 15376
+    },
+    {
+      "epoch": 0.6474526315789474,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00014337201560086645,
+      "loss": 2.9856,
+      "step": 15377
+    },
+    {
+      "epoch": 0.6474947368421052,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00014334149547887284,
+      "loss": 3.1874,
+      "step": 15378
+    },
+    {
+      "epoch": 0.6475368421052632,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0001433109773000721,
+      "loss": 3.1252,
+      "step": 15379
+    },
+    {
+      "epoch": 0.647578947368421,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0001432804610650198,
+      "loss": 3.4591,
+      "step": 15380
+    },
+    {
+      "epoch": 0.647621052631579,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00014324994677427223,
+      "loss": 3.1386,
+      "step": 15381
+    },
+    {
+      "epoch": 0.6476631578947368,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00014321943442838515,
+      "loss": 2.9095,
+      "step": 15382
+    },
+    {
+      "epoch": 0.6477052631578948,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00014318892402791455,
+      "loss": 2.8722,
+      "step": 15383
+    },
+    {
+      "epoch": 0.6477473684210526,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00014315841557341626,
+      "loss": 2.9733,
+      "step": 15384
+    },
+    {
+      "epoch": 0.6477894736842106,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.000143127909065446,
+      "loss": 3.1064,
+      "step": 15385
+    },
+    {
+      "epoch": 0.6478315789473684,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00014309740450455983,
+      "loss": 2.7143,
+      "step": 15386
+    },
+    {
+      "epoch": 0.6478736842105263,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001430669018913132,
+      "loss": 3.0695,
+      "step": 15387
+    },
+    {
+      "epoch": 0.6479157894736842,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00014303640122626204,
+      "loss": 3.6446,
+      "step": 15388
+    },
+    {
+      "epoch": 0.6479578947368421,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00014300590250996198,
+      "loss": 3.2076,
+      "step": 15389
+    },
+    {
+      "epoch": 0.648,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0001429754057429687,
+      "loss": 2.8745,
+      "step": 15390
+    },
+    {
+      "epoch": 0.6480421052631579,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00014294491092583774,
+      "loss": 3.1567,
+      "step": 15391
+    },
+    {
+      "epoch": 0.6480842105263158,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0001429144180591247,
+      "loss": 3.006,
+      "step": 15392
+    },
+    {
+      "epoch": 0.6481263157894737,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00014288392714338508,
+      "loss": 3.1309,
+      "step": 15393
+    },
+    {
+      "epoch": 0.6481684210526316,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001428534381791745,
+      "loss": 2.9069,
+      "step": 15394
+    },
+    {
+      "epoch": 0.6482105263157895,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00014282295116704847,
+      "loss": 3.1705,
+      "step": 15395
+    },
+    {
+      "epoch": 0.6482526315789474,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00014279246610756228,
+      "loss": 3.3319,
+      "step": 15396
+    },
+    {
+      "epoch": 0.6482947368421053,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001427619830012714,
+      "loss": 3.3487,
+      "step": 15397
+    },
+    {
+      "epoch": 0.6483368421052632,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00014273150184873122,
+      "loss": 3.3209,
+      "step": 15398
+    },
+    {
+      "epoch": 0.648378947368421,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.000142701022650497,
+      "loss": 3.0337,
+      "step": 15399
+    },
+    {
+      "epoch": 0.6484210526315789,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.000142670545407124,
+      "loss": 3.2667,
+      "step": 15400
+    },
+    {
+      "epoch": 0.6484631578947369,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00014264007011916766,
+      "loss": 3.547,
+      "step": 15401
+    },
+    {
+      "epoch": 0.6485052631578947,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0001426095967871831,
+      "loss": 2.7986,
+      "step": 15402
+    },
+    {
+      "epoch": 0.6485473684210527,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00014257912541172548,
+      "loss": 3.2803,
+      "step": 15403
+    },
+    {
+      "epoch": 0.6485894736842105,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00014254865599335002,
+      "loss": 3.2019,
+      "step": 15404
+    },
+    {
+      "epoch": 0.6486315789473684,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00014251818853261178,
+      "loss": 2.8432,
+      "step": 15405
+    },
+    {
+      "epoch": 0.6486736842105263,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00014248772303006587,
+      "loss": 3.1761,
+      "step": 15406
+    },
+    {
+      "epoch": 0.6487157894736842,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001424572594862672,
+      "loss": 3.0022,
+      "step": 15407
+    },
+    {
+      "epoch": 0.6487578947368421,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0001424267979017711,
+      "loss": 3.055,
+      "step": 15408
+    },
+    {
+      "epoch": 0.6488,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00014239633827713217,
+      "loss": 2.9692,
+      "step": 15409
+    },
+    {
+      "epoch": 0.6488421052631579,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00014236588061290563,
+      "loss": 3.5153,
+      "step": 15410
+    },
+    {
+      "epoch": 0.6488842105263158,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00014233542490964626,
+      "loss": 3.0271,
+      "step": 15411
+    },
+    {
+      "epoch": 0.6489263157894737,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00014230497116790898,
+      "loss": 3.244,
+      "step": 15412
+    },
+    {
+      "epoch": 0.6489684210526315,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00014227451938824857,
+      "loss": 3.2601,
+      "step": 15413
+    },
+    {
+      "epoch": 0.6490105263157895,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00014224406957121972,
+      "loss": 3.1056,
+      "step": 15414
+    },
+    {
+      "epoch": 0.6490526315789473,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00014221362171737756,
+      "loss": 2.8352,
+      "step": 15415
+    },
+    {
+      "epoch": 0.6490947368421053,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00014218317582727637,
+      "loss": 2.794,
+      "step": 15416
+    },
+    {
+      "epoch": 0.6491368421052631,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001421527319014711,
+      "loss": 2.7106,
+      "step": 15417
+    },
+    {
+      "epoch": 0.6491789473684211,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00014212228994051634,
+      "loss": 3.1676,
+      "step": 15418
+    },
+    {
+      "epoch": 0.6492210526315789,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001420918499449667,
+      "loss": 3.0585,
+      "step": 15419
+    },
+    {
+      "epoch": 0.6492631578947369,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0001420614119153768,
+      "loss": 3.1459,
+      "step": 15420
+    },
+    {
+      "epoch": 0.6493052631578947,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00014203097585230103,
+      "loss": 2.9756,
+      "step": 15421
+    },
+    {
+      "epoch": 0.6493473684210527,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00014200054175629424,
+      "loss": 3.3647,
+      "step": 15422
+    },
+    {
+      "epoch": 0.6493894736842105,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00014197010962791045,
+      "loss": 2.9855,
+      "step": 15423
+    },
+    {
+      "epoch": 0.6494315789473685,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0001419396794677046,
+      "loss": 3.2659,
+      "step": 15424
+    },
+    {
+      "epoch": 0.6494736842105263,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00014190925127623055,
+      "loss": 3.0454,
+      "step": 15425
+    },
+    {
+      "epoch": 0.6495157894736842,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00014187882505404308,
+      "loss": 3.428,
+      "step": 15426
+    },
+    {
+      "epoch": 0.6495578947368421,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00014184840080169638,
+      "loss": 2.9105,
+      "step": 15427
+    },
+    {
+      "epoch": 0.6496,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00014181797851974464,
+      "loss": 3.5866,
+      "step": 15428
+    },
+    {
+      "epoch": 0.6496421052631579,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00014178755820874243,
+      "loss": 3.1565,
+      "step": 15429
+    },
+    {
+      "epoch": 0.6496842105263158,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00014175713986924355,
+      "loss": 3.1828,
+      "step": 15430
+    },
+    {
+      "epoch": 0.6497263157894737,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001417267235018026,
+      "loss": 3.2198,
+      "step": 15431
+    },
+    {
+      "epoch": 0.6497684210526316,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00014169630910697335,
+      "loss": 3.3999,
+      "step": 15432
+    },
+    {
+      "epoch": 0.6498105263157895,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0001416658966853102,
+      "loss": 3.5442,
+      "step": 15433
+    },
+    {
+      "epoch": 0.6498526315789473,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001416354862373671,
+      "loss": 3.0459,
+      "step": 15434
+    },
+    {
+      "epoch": 0.6498947368421053,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00014160507776369812,
+      "loss": 3.2393,
+      "step": 15435
+    },
+    {
+      "epoch": 0.6499368421052631,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00014157467126485728,
+      "loss": 2.8662,
+      "step": 15436
+    },
+    {
+      "epoch": 0.6499789473684211,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00014154426674139843,
+      "loss": 3.4425,
+      "step": 15437
+    },
+    {
+      "epoch": 0.6500210526315789,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0001415138641938758,
+      "loss": 3.1926,
+      "step": 15438
+    },
+    {
+      "epoch": 0.6500631578947368,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00014148346362284292,
+      "loss": 3.2168,
+      "step": 15439
+    },
+    {
+      "epoch": 0.6501052631578947,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0001414530650288539,
+      "loss": 2.6032,
+      "step": 15440
+    },
+    {
+      "epoch": 0.6501473684210526,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00014142266841246255,
+      "loss": 3.1884,
+      "step": 15441
+    },
+    {
+      "epoch": 0.6501894736842105,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00014139227377422253,
+      "loss": 3.2056,
+      "step": 15442
+    },
+    {
+      "epoch": 0.6502315789473684,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00014136188111468773,
+      "loss": 3.1645,
+      "step": 15443
+    },
+    {
+      "epoch": 0.6502736842105263,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0001413314904344117,
+      "loss": 3.0849,
+      "step": 15444
+    },
+    {
+      "epoch": 0.6503157894736842,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00014130110173394845,
+      "loss": 2.6386,
+      "step": 15445
+    },
+    {
+      "epoch": 0.6503578947368421,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0001412707150138512,
+      "loss": 2.9644,
+      "step": 15446
+    },
+    {
+      "epoch": 0.6504,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00014124033027467384,
+      "loss": 3.0721,
+      "step": 15447
+    },
+    {
+      "epoch": 0.6504421052631579,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00014120994751696988,
+      "loss": 3.2342,
+      "step": 15448
+    },
+    {
+      "epoch": 0.6504842105263158,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001411795667412929,
+      "loss": 3.6151,
+      "step": 15449
+    },
+    {
+      "epoch": 0.6505263157894737,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0001411491879481963,
+      "loss": 3.0623,
+      "step": 15450
+    },
+    {
+      "epoch": 0.6505684210526316,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00014111881113823365,
+      "loss": 3.0185,
+      "step": 15451
+    },
+    {
+      "epoch": 0.6506105263157895,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0001410884363119583,
+      "loss": 3.2916,
+      "step": 15452
+    },
+    {
+      "epoch": 0.6506526315789474,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00014105806346992358,
+      "loss": 3.3551,
+      "step": 15453
+    },
+    {
+      "epoch": 0.6506947368421052,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00014102769261268306,
+      "loss": 3.0469,
+      "step": 15454
+    },
+    {
+      "epoch": 0.6507368421052632,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00014099732374078992,
+      "loss": 2.9465,
+      "step": 15455
+    },
+    {
+      "epoch": 0.650778947368421,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00014096695685479748,
+      "loss": 3.3318,
+      "step": 15456
+    },
+    {
+      "epoch": 0.650821052631579,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.000140936591955259,
+      "loss": 3.4713,
+      "step": 15457
+    },
+    {
+      "epoch": 0.6508631578947368,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00014090622904272762,
+      "loss": 3.5066,
+      "step": 15458
+    },
+    {
+      "epoch": 0.6509052631578948,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0001408758681177566,
+      "loss": 3.0179,
+      "step": 15459
+    },
+    {
+      "epoch": 0.6509473684210526,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00014084550918089895,
+      "loss": 3.1468,
+      "step": 15460
+    },
+    {
+      "epoch": 0.6509894736842106,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00014081515223270794,
+      "loss": 3.4201,
+      "step": 15461
+    },
+    {
+      "epoch": 0.6510315789473684,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0001407847972737366,
+      "loss": 3.1901,
+      "step": 15462
+    },
+    {
+      "epoch": 0.6510736842105264,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00014075444430453792,
+      "loss": 3.0143,
+      "step": 15463
+    },
+    {
+      "epoch": 0.6511157894736842,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00014072409332566488,
+      "loss": 3.5901,
+      "step": 15464
+    },
+    {
+      "epoch": 0.6511578947368422,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00014069374433767052,
+      "loss": 3.4214,
+      "step": 15465
+    },
+    {
+      "epoch": 0.6512,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00014066339734110768,
+      "loss": 2.8183,
+      "step": 15466
+    },
+    {
+      "epoch": 0.6512421052631578,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001406330523365293,
+      "loss": 3.2492,
+      "step": 15467
+    },
+    {
+      "epoch": 0.6512842105263158,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00014060270932448804,
+      "loss": 3.0104,
+      "step": 15468
+    },
+    {
+      "epoch": 0.6513263157894736,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00014057236830553705,
+      "loss": 3.0741,
+      "step": 15469
+    },
+    {
+      "epoch": 0.6513684210526316,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00014054202928022891,
+      "loss": 3.2965,
+      "step": 15470
+    },
+    {
+      "epoch": 0.6514105263157894,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00014051169224911636,
+      "loss": 3.0974,
+      "step": 15471
+    },
+    {
+      "epoch": 0.6514526315789474,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001404813572127522,
+      "loss": 2.4895,
+      "step": 15472
+    },
+    {
+      "epoch": 0.6514947368421052,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00014045102417168897,
+      "loss": 2.606,
+      "step": 15473
+    },
+    {
+      "epoch": 0.6515368421052632,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0001404206931264794,
+      "loss": 2.9481,
+      "step": 15474
+    },
+    {
+      "epoch": 0.651578947368421,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00014039036407767594,
+      "loss": 3.2244,
+      "step": 15475
+    },
+    {
+      "epoch": 0.651621052631579,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00014036003702583137,
+      "loss": 2.9741,
+      "step": 15476
+    },
+    {
+      "epoch": 0.6516631578947368,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00014032971197149808,
+      "loss": 3.1124,
+      "step": 15477
+    },
+    {
+      "epoch": 0.6517052631578948,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0001402993889152286,
+      "loss": 3.477,
+      "step": 15478
+    },
+    {
+      "epoch": 0.6517473684210526,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00014026906785757537,
+      "loss": 2.8771,
+      "step": 15479
+    },
+    {
+      "epoch": 0.6517894736842105,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00014023874879909076,
+      "loss": 3.2579,
+      "step": 15480
+    },
+    {
+      "epoch": 0.6518315789473684,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0001402084317403272,
+      "loss": 3.4636,
+      "step": 15481
+    },
+    {
+      "epoch": 0.6518736842105263,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00014017811668183693,
+      "loss": 3.0716,
+      "step": 15482
+    },
+    {
+      "epoch": 0.6519157894736842,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00014014780362417252,
+      "loss": 2.5203,
+      "step": 15483
+    },
+    {
+      "epoch": 0.6519578947368421,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00014011749256788585,
+      "loss": 3.2773,
+      "step": 15484
+    },
+    {
+      "epoch": 0.652,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00014008718351352952,
+      "loss": 3.2776,
+      "step": 15485
+    },
+    {
+      "epoch": 0.6520421052631579,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00014005687646165548,
+      "loss": 3.2386,
+      "step": 15486
+    },
+    {
+      "epoch": 0.6520842105263158,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00014002657141281603,
+      "loss": 2.7903,
+      "step": 15487
+    },
+    {
+      "epoch": 0.6521263157894737,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0001399962683675632,
+      "loss": 2.938,
+      "step": 15488
+    },
+    {
+      "epoch": 0.6521684210526316,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00013996596732644904,
+      "loss": 3.3932,
+      "step": 15489
+    },
+    {
+      "epoch": 0.6522105263157895,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00013993566829002585,
+      "loss": 3.4126,
+      "step": 15490
+    },
+    {
+      "epoch": 0.6522526315789474,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0001399053712588453,
+      "loss": 3.0689,
+      "step": 15491
+    },
+    {
+      "epoch": 0.6522947368421053,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00013987507623345957,
+      "loss": 2.8674,
+      "step": 15492
+    },
+    {
+      "epoch": 0.6523368421052631,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001398447832144206,
+      "loss": 3.3193,
+      "step": 15493
+    },
+    {
+      "epoch": 0.6523789473684211,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00013981449220228025,
+      "loss": 2.9399,
+      "step": 15494
+    },
+    {
+      "epoch": 0.6524210526315789,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00013978420319759037,
+      "loss": 3.4297,
+      "step": 15495
+    },
+    {
+      "epoch": 0.6524631578947369,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00013975391620090272,
+      "loss": 3.218,
+      "step": 15496
+    },
+    {
+      "epoch": 0.6525052631578947,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00013972363121276938,
+      "loss": 3.1009,
+      "step": 15497
+    },
+    {
+      "epoch": 0.6525473684210527,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0001396933482337417,
+      "loss": 3.1822,
+      "step": 15498
+    },
+    {
+      "epoch": 0.6525894736842105,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00013966306726437183,
+      "loss": 3.3537,
+      "step": 15499
+    },
+    {
+      "epoch": 0.6526315789473685,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00013963278830521103,
+      "loss": 3.0388,
+      "step": 15500
+    },
+    {
+      "epoch": 0.6526736842105263,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00013960251135681126,
+      "loss": 3.3139,
+      "step": 15501
+    },
+    {
+      "epoch": 0.6527157894736842,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.000139572236419724,
+      "loss": 3.41,
+      "step": 15502
+    },
+    {
+      "epoch": 0.6527578947368421,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00013954196349450076,
+      "loss": 3.4537,
+      "step": 15503
+    },
+    {
+      "epoch": 0.6528,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00013951169258169337,
+      "loss": 3.1323,
+      "step": 15504
+    },
+    {
+      "epoch": 0.6528421052631579,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001394814236818529,
+      "loss": 3.36,
+      "step": 15505
+    },
+    {
+      "epoch": 0.6528842105263158,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00013945115679553127,
+      "loss": 3.1089,
+      "step": 15506
+    },
+    {
+      "epoch": 0.6529263157894737,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00013942089192327945,
+      "loss": 3.1394,
+      "step": 15507
+    },
+    {
+      "epoch": 0.6529684210526315,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00013939062906564913,
+      "loss": 3.2209,
+      "step": 15508
+    },
+    {
+      "epoch": 0.6530105263157895,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00013936036822319165,
+      "loss": 3.4028,
+      "step": 15509
+    },
+    {
+      "epoch": 0.6530526315789473,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0001393301093964582,
+      "loss": 3.1186,
+      "step": 15510
+    },
+    {
+      "epoch": 0.6530947368421053,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00013929985258600018,
+      "loss": 3.4186,
+      "step": 15511
+    },
+    {
+      "epoch": 0.6531368421052631,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00013926959779236864,
+      "loss": 2.9905,
+      "step": 15512
+    },
+    {
+      "epoch": 0.6531789473684211,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00013923934501611514,
+      "loss": 3.0216,
+      "step": 15513
+    },
+    {
+      "epoch": 0.6532210526315789,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00013920909425779043,
+      "loss": 3.045,
+      "step": 15514
+    },
+    {
+      "epoch": 0.6532631578947369,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00013917884551794597,
+      "loss": 3.109,
+      "step": 15515
+    },
+    {
+      "epoch": 0.6533052631578947,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00013914859879713272,
+      "loss": 3.2266,
+      "step": 15516
+    },
+    {
+      "epoch": 0.6533473684210527,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00013911835409590172,
+      "loss": 3.1346,
+      "step": 15517
+    },
+    {
+      "epoch": 0.6533894736842105,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00013908811141480408,
+      "loss": 3.0094,
+      "step": 15518
+    },
+    {
+      "epoch": 0.6534315789473685,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00013905787075439057,
+      "loss": 3.1286,
+      "step": 15519
+    },
+    {
+      "epoch": 0.6534736842105263,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00013902763211521259,
+      "loss": 3.4562,
+      "step": 15520
+    },
+    {
+      "epoch": 0.6535157894736842,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0001389973954978205,
+      "loss": 2.8064,
+      "step": 15521
+    },
+    {
+      "epoch": 0.6535578947368421,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00013896716090276557,
+      "loss": 2.9919,
+      "step": 15522
+    },
+    {
+      "epoch": 0.6536,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00013893692833059852,
+      "loss": 3.1654,
+      "step": 15523
+    },
+    {
+      "epoch": 0.6536421052631579,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0001389066977818701,
+      "loss": 2.7809,
+      "step": 15524
+    },
+    {
+      "epoch": 0.6536842105263158,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00013887646925713116,
+      "loss": 3.1881,
+      "step": 15525
+    },
+    {
+      "epoch": 0.6537263157894737,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00013884624275693237,
+      "loss": 2.8581,
+      "step": 15526
+    },
+    {
+      "epoch": 0.6537684210526316,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00013881601828182445,
+      "loss": 3.0233,
+      "step": 15527
+    },
+    {
+      "epoch": 0.6538105263157895,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0001387857958323579,
+      "loss": 3.559,
+      "step": 15528
+    },
+    {
+      "epoch": 0.6538526315789474,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001387555754090836,
+      "loss": 3.0345,
+      "step": 15529
+    },
+    {
+      "epoch": 0.6538947368421053,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.000138725357012552,
+      "loss": 3.4478,
+      "step": 15530
+    },
+    {
+      "epoch": 0.6539368421052631,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00013869514064331369,
+      "loss": 3.0467,
+      "step": 15531
+    },
+    {
+      "epoch": 0.6539789473684211,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00013866492630191913,
+      "loss": 3.0422,
+      "step": 15532
+    },
+    {
+      "epoch": 0.654021052631579,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00013863471398891874,
+      "loss": 3.1498,
+      "step": 15533
+    },
+    {
+      "epoch": 0.6540631578947368,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00013860450370486307,
+      "loss": 2.9495,
+      "step": 15534
+    },
+    {
+      "epoch": 0.6541052631578947,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00013857429545030244,
+      "loss": 3.149,
+      "step": 15535
+    },
+    {
+      "epoch": 0.6541473684210526,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00013854408922578715,
+      "loss": 3.1901,
+      "step": 15536
+    },
+    {
+      "epoch": 0.6541894736842105,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0001385138850318677,
+      "loss": 3.2531,
+      "step": 15537
+    },
+    {
+      "epoch": 0.6542315789473684,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00013848368286909423,
+      "loss": 3.0485,
+      "step": 15538
+    },
+    {
+      "epoch": 0.6542736842105263,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00013845348273801707,
+      "loss": 2.9535,
+      "step": 15539
+    },
+    {
+      "epoch": 0.6543157894736842,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0001384232846391864,
+      "loss": 3.2329,
+      "step": 15540
+    },
+    {
+      "epoch": 0.6543578947368421,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00013839308857315226,
+      "loss": 3.3624,
+      "step": 15541
+    },
+    {
+      "epoch": 0.6544,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001383628945404651,
+      "loss": 2.7173,
+      "step": 15542
+    },
+    {
+      "epoch": 0.6544421052631579,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00013833270254167465,
+      "loss": 2.9578,
+      "step": 15543
+    },
+    {
+      "epoch": 0.6544842105263158,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00013830251257733128,
+      "loss": 3.4635,
+      "step": 15544
+    },
+    {
+      "epoch": 0.6545263157894737,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00013827232464798487,
+      "loss": 3.5202,
+      "step": 15545
+    },
+    {
+      "epoch": 0.6545684210526316,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00013824213875418542,
+      "loss": 2.6278,
+      "step": 15546
+    },
+    {
+      "epoch": 0.6546105263157894,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0001382119548964829,
+      "loss": 3.1953,
+      "step": 15547
+    },
+    {
+      "epoch": 0.6546526315789474,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00013818177307542713,
+      "loss": 3.0636,
+      "step": 15548
+    },
+    {
+      "epoch": 0.6546947368421052,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00013815159329156826,
+      "loss": 2.9344,
+      "step": 15549
+    },
+    {
+      "epoch": 0.6547368421052632,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00013812141554545577,
+      "loss": 3.1729,
+      "step": 15550
+    },
+    {
+      "epoch": 0.654778947368421,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0001380912398376398,
+      "loss": 3.0613,
+      "step": 15551
+    },
+    {
+      "epoch": 0.654821052631579,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00013806106616866975,
+      "loss": 3.012,
+      "step": 15552
+    },
+    {
+      "epoch": 0.6548631578947368,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001380308945390957,
+      "loss": 3.5513,
+      "step": 15553
+    },
+    {
+      "epoch": 0.6549052631578948,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00013800072494946715,
+      "loss": 3.428,
+      "step": 15554
+    },
+    {
+      "epoch": 0.6549473684210526,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00013797055740033372,
+      "loss": 2.5745,
+      "step": 15555
+    },
+    {
+      "epoch": 0.6549894736842106,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00013794039189224527,
+      "loss": 2.5773,
+      "step": 15556
+    },
+    {
+      "epoch": 0.6550315789473684,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00013791022842575104,
+      "loss": 3.2473,
+      "step": 15557
+    },
+    {
+      "epoch": 0.6550736842105264,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00013788006700140096,
+      "loss": 2.9793,
+      "step": 15558
+    },
+    {
+      "epoch": 0.6551157894736842,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001378499076197441,
+      "loss": 2.6976,
+      "step": 15559
+    },
+    {
+      "epoch": 0.655157894736842,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00013781975028133027,
+      "loss": 3.2225,
+      "step": 15560
+    },
+    {
+      "epoch": 0.6552,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00013778959498670878,
+      "loss": 3.2579,
+      "step": 15561
+    },
+    {
+      "epoch": 0.6552421052631578,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.000137759441736429,
+      "loss": 2.7472,
+      "step": 15562
+    },
+    {
+      "epoch": 0.6552842105263158,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00013772929053104037,
+      "loss": 2.9966,
+      "step": 15563
+    },
+    {
+      "epoch": 0.6553263157894736,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00013769914137109202,
+      "loss": 3.472,
+      "step": 15564
+    },
+    {
+      "epoch": 0.6553684210526316,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00013766899425713354,
+      "loss": 3.0146,
+      "step": 15565
+    },
+    {
+      "epoch": 0.6554105263157894,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00013763884918971387,
+      "loss": 3.2299,
+      "step": 15566
+    },
+    {
+      "epoch": 0.6554526315789474,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001376087061693825,
+      "loss": 3.017,
+      "step": 15567
+    },
+    {
+      "epoch": 0.6554947368421052,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00013757856519668827,
+      "loss": 3.5853,
+      "step": 15568
+    },
+    {
+      "epoch": 0.6555368421052632,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00013754842627218056,
+      "loss": 3.4511,
+      "step": 15569
+    },
+    {
+      "epoch": 0.655578947368421,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00013751828939640841,
+      "loss": 3.4127,
+      "step": 15570
+    },
+    {
+      "epoch": 0.655621052631579,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00013748815456992078,
+      "loss": 3.407,
+      "step": 15571
+    },
+    {
+      "epoch": 0.6556631578947368,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.000137458021793267,
+      "loss": 2.7952,
+      "step": 15572
+    },
+    {
+      "epoch": 0.6557052631578948,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0001374278910669956,
+      "loss": 3.0883,
+      "step": 15573
+    },
+    {
+      "epoch": 0.6557473684210526,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00013739776239165598,
+      "loss": 2.6413,
+      "step": 15574
+    },
+    {
+      "epoch": 0.6557894736842105,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0001373676357677966,
+      "loss": 2.8272,
+      "step": 15575
+    },
+    {
+      "epoch": 0.6558315789473684,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00013733751119596665,
+      "loss": 3.0791,
+      "step": 15576
+    },
+    {
+      "epoch": 0.6558736842105263,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0001373073886767149,
+      "loss": 3.1639,
+      "step": 15577
+    },
+    {
+      "epoch": 0.6559157894736842,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00013727726821059012,
+      "loss": 3.3191,
+      "step": 15578
+    },
+    {
+      "epoch": 0.6559578947368421,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00013724714979814105,
+      "loss": 3.3552,
+      "step": 15579
+    },
+    {
+      "epoch": 0.656,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0001372170334399163,
+      "loss": 2.9483,
+      "step": 15580
+    },
+    {
+      "epoch": 0.6560421052631579,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00013718691913646496,
+      "loss": 3.4547,
+      "step": 15581
+    },
+    {
+      "epoch": 0.6560842105263158,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00013715680688833519,
+      "loss": 3.3725,
+      "step": 15582
+    },
+    {
+      "epoch": 0.6561263157894737,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00013712669669607586,
+      "loss": 2.7827,
+      "step": 15583
+    },
+    {
+      "epoch": 0.6561684210526316,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00013709658856023556,
+      "loss": 3.3958,
+      "step": 15584
+    },
+    {
+      "epoch": 0.6562105263157895,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00013706648248136273,
+      "loss": 3.3138,
+      "step": 15585
+    },
+    {
+      "epoch": 0.6562526315789474,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0001370363784600059,
+      "loss": 3.574,
+      "step": 15586
+    },
+    {
+      "epoch": 0.6562947368421053,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00013700627649671345,
+      "loss": 3.1679,
+      "step": 15587
+    },
+    {
+      "epoch": 0.6563368421052631,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00013697617659203394,
+      "loss": 2.8231,
+      "step": 15588
+    },
+    {
+      "epoch": 0.6563789473684211,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00013694607874651572,
+      "loss": 3.2442,
+      "step": 15589
+    },
+    {
+      "epoch": 0.6564210526315789,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00013691598296070706,
+      "loss": 3.329,
+      "step": 15590
+    },
+    {
+      "epoch": 0.6564631578947369,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00013688588923515638,
+      "loss": 3.119,
+      "step": 15591
+    },
+    {
+      "epoch": 0.6565052631578947,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00013685579757041189,
+      "loss": 3.4377,
+      "step": 15592
+    },
+    {
+      "epoch": 0.6565473684210527,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00013682570796702182,
+      "loss": 2.9281,
+      "step": 15593
+    },
+    {
+      "epoch": 0.6565894736842105,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0001367956204255344,
+      "loss": 3.4774,
+      "step": 15594
+    },
+    {
+      "epoch": 0.6566315789473685,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00013676553494649762,
+      "loss": 3.0702,
+      "step": 15595
+    },
+    {
+      "epoch": 0.6566736842105263,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00013673545153045986,
+      "loss": 3.053,
+      "step": 15596
+    },
+    {
+      "epoch": 0.6567157894736843,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001367053701779691,
+      "loss": 3.0937,
+      "step": 15597
+    },
+    {
+      "epoch": 0.6567578947368421,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0001366752908895734,
+      "loss": 2.9983,
+      "step": 15598
+    },
+    {
+      "epoch": 0.6568,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00013664521366582073,
+      "loss": 3.1999,
+      "step": 15599
+    },
+    {
+      "epoch": 0.6568421052631579,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0001366151385072591,
+      "loss": 2.7007,
+      "step": 15600
+    },
+    {
+      "epoch": 0.6568842105263157,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0001365850654144364,
+      "loss": 3.0551,
+      "step": 15601
+    },
+    {
+      "epoch": 0.6569263157894737,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00013655499438790044,
+      "loss": 3.0709,
+      "step": 15602
+    },
+    {
+      "epoch": 0.6569684210526315,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00013652492542819933,
+      "loss": 3.3113,
+      "step": 15603
+    },
+    {
+      "epoch": 0.6570105263157895,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00013649485853588073,
+      "loss": 3.3274,
+      "step": 15604
+    },
+    {
+      "epoch": 0.6570526315789473,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00013646479371149246,
+      "loss": 2.5574,
+      "step": 15605
+    },
+    {
+      "epoch": 0.6570947368421053,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0001364347309555822,
+      "loss": 3.3983,
+      "step": 15606
+    },
+    {
+      "epoch": 0.6571368421052631,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00013640467026869774,
+      "loss": 2.9053,
+      "step": 15607
+    },
+    {
+      "epoch": 0.6571789473684211,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00013637461165138674,
+      "loss": 3.0288,
+      "step": 15608
+    },
+    {
+      "epoch": 0.6572210526315789,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00013634455510419668,
+      "loss": 2.7372,
+      "step": 15609
+    },
+    {
+      "epoch": 0.6572631578947369,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0001363145006276755,
+      "loss": 3.243,
+      "step": 15610
+    },
+    {
+      "epoch": 0.6573052631578947,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00013628444822237035,
+      "loss": 2.7422,
+      "step": 15611
+    },
+    {
+      "epoch": 0.6573473684210527,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00013625439788882905,
+      "loss": 3.0412,
+      "step": 15612
+    },
+    {
+      "epoch": 0.6573894736842105,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00013622434962759896,
+      "loss": 2.8283,
+      "step": 15613
+    },
+    {
+      "epoch": 0.6574315789473684,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0001361943034392275,
+      "loss": 3.1585,
+      "step": 15614
+    },
+    {
+      "epoch": 0.6574736842105263,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00013616425932426218,
+      "loss": 3.245,
+      "step": 15615
+    },
+    {
+      "epoch": 0.6575157894736842,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00013613421728325017,
+      "loss": 3.2405,
+      "step": 15616
+    },
+    {
+      "epoch": 0.6575578947368421,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00013610417731673913,
+      "loss": 2.8941,
+      "step": 15617
+    },
+    {
+      "epoch": 0.6576,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00013607413942527596,
+      "loss": 2.7251,
+      "step": 15618
+    },
+    {
+      "epoch": 0.6576421052631579,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00013604410360940826,
+      "loss": 3.286,
+      "step": 15619
+    },
+    {
+      "epoch": 0.6576842105263158,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00013601406986968307,
+      "loss": 3.3008,
+      "step": 15620
+    },
+    {
+      "epoch": 0.6577263157894737,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00013598403820664755,
+      "loss": 3.2681,
+      "step": 15621
+    },
+    {
+      "epoch": 0.6577684210526316,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00013595400862084894,
+      "loss": 2.8217,
+      "step": 15622
+    },
+    {
+      "epoch": 0.6578105263157895,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001359239811128342,
+      "loss": 3.0123,
+      "step": 15623
+    },
+    {
+      "epoch": 0.6578526315789474,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00013589395568315067,
+      "loss": 3.5839,
+      "step": 15624
+    },
+    {
+      "epoch": 0.6578947368421053,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00013586393233234502,
+      "loss": 3.2106,
+      "step": 15625
+    },
+    {
+      "epoch": 0.6579368421052632,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0001358339110609646,
+      "loss": 3.6317,
+      "step": 15626
+    },
+    {
+      "epoch": 0.6579789473684211,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.000135803891869556,
+      "loss": 2.8747,
+      "step": 15627
+    },
+    {
+      "epoch": 0.658021052631579,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00013577387475866638,
+      "loss": 2.9033,
+      "step": 15628
+    },
+    {
+      "epoch": 0.6580631578947368,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00013574385972884258,
+      "loss": 2.8642,
+      "step": 15629
+    },
+    {
+      "epoch": 0.6581052631578947,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00013571384678063127,
+      "loss": 3.1112,
+      "step": 15630
+    },
+    {
+      "epoch": 0.6581473684210526,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00013568383591457963,
+      "loss": 3.1546,
+      "step": 15631
+    },
+    {
+      "epoch": 0.6581894736842105,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00013565382713123399,
+      "loss": 3.1903,
+      "step": 15632
+    },
+    {
+      "epoch": 0.6582315789473684,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00013562382043114145,
+      "loss": 3.0894,
+      "step": 15633
+    },
+    {
+      "epoch": 0.6582736842105263,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001355938158148483,
+      "loss": 3.1193,
+      "step": 15634
+    },
+    {
+      "epoch": 0.6583157894736842,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00013556381328290153,
+      "loss": 3.3353,
+      "step": 15635
+    },
+    {
+      "epoch": 0.6583578947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00013553381283584764,
+      "loss": 2.6582,
+      "step": 15636
+    },
+    {
+      "epoch": 0.6584,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00013550381447423315,
+      "loss": 2.8553,
+      "step": 15637
+    },
+    {
+      "epoch": 0.6584421052631579,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00013547381819860465,
+      "loss": 3.4918,
+      "step": 15638
+    },
+    {
+      "epoch": 0.6584842105263158,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00013544382400950853,
+      "loss": 3.2636,
+      "step": 15639
+    },
+    {
+      "epoch": 0.6585263157894737,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00013541383190749152,
+      "loss": 2.7619,
+      "step": 15640
+    },
+    {
+      "epoch": 0.6585684210526316,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00013538384189309966,
+      "loss": 3.2107,
+      "step": 15641
+    },
+    {
+      "epoch": 0.6586105263157894,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00013535385396687966,
+      "loss": 2.4485,
+      "step": 15642
+    },
+    {
+      "epoch": 0.6586526315789474,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0001353238681293777,
+      "loss": 3.1182,
+      "step": 15643
+    },
+    {
+      "epoch": 0.6586947368421052,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00013529388438114016,
+      "loss": 3.0032,
+      "step": 15644
+    },
+    {
+      "epoch": 0.6587368421052632,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00013526390272271325,
+      "loss": 2.8326,
+      "step": 15645
+    },
+    {
+      "epoch": 0.658778947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0001352339231546431,
+      "loss": 3.1958,
+      "step": 15646
+    },
+    {
+      "epoch": 0.658821052631579,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00013520394567747627,
+      "loss": 2.7169,
+      "step": 15647
+    },
+    {
+      "epoch": 0.6588631578947368,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00013517397029175844,
+      "loss": 2.9366,
+      "step": 15648
+    },
+    {
+      "epoch": 0.6589052631578948,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00013514399699803608,
+      "loss": 2.972,
+      "step": 15649
+    },
+    {
+      "epoch": 0.6589473684210526,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0001351140257968551,
+      "loss": 3.2461,
+      "step": 15650
+    },
+    {
+      "epoch": 0.6589894736842106,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00013508405668876165,
+      "loss": 3.3538,
+      "step": 15651
+    },
+    {
+      "epoch": 0.6590315789473684,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001350540896743016,
+      "loss": 3.5014,
+      "step": 15652
+    },
+    {
+      "epoch": 0.6590736842105264,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00013502412475402105,
+      "loss": 3.1084,
+      "step": 15653
+    },
+    {
+      "epoch": 0.6591157894736842,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001349941619284658,
+      "loss": 3.0358,
+      "step": 15654
+    },
+    {
+      "epoch": 0.659157894736842,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0001349642011981817,
+      "loss": 2.8567,
+      "step": 15655
+    },
+    {
+      "epoch": 0.6592,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00013493424256371485,
+      "loss": 2.9034,
+      "step": 15656
+    },
+    {
+      "epoch": 0.6592421052631579,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00013490428602561084,
+      "loss": 3.0365,
+      "step": 15657
+    },
+    {
+      "epoch": 0.6592842105263158,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00013487433158441552,
+      "loss": 3.2563,
+      "step": 15658
+    },
+    {
+      "epoch": 0.6593263157894736,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0001348443792406746,
+      "loss": 3.2114,
+      "step": 15659
+    },
+    {
+      "epoch": 0.6593684210526316,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00013481442899493382,
+      "loss": 3.5359,
+      "step": 15660
+    },
+    {
+      "epoch": 0.6594105263157894,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0001347844808477388,
+      "loss": 3.1023,
+      "step": 15661
+    },
+    {
+      "epoch": 0.6594526315789474,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00013475453479963506,
+      "loss": 2.9235,
+      "step": 15662
+    },
+    {
+      "epoch": 0.6594947368421052,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00013472459085116839,
+      "loss": 2.7418,
+      "step": 15663
+    },
+    {
+      "epoch": 0.6595368421052632,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00013469464900288418,
+      "loss": 3.1674,
+      "step": 15664
+    },
+    {
+      "epoch": 0.659578947368421,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00013466470925532808,
+      "loss": 3.6334,
+      "step": 15665
+    },
+    {
+      "epoch": 0.659621052631579,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00013463477160904541,
+      "loss": 2.5271,
+      "step": 15666
+    },
+    {
+      "epoch": 0.6596631578947368,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00013460483606458166,
+      "loss": 2.8849,
+      "step": 15667
+    },
+    {
+      "epoch": 0.6597052631578947,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001345749026224822,
+      "loss": 3.0873,
+      "step": 15668
+    },
+    {
+      "epoch": 0.6597473684210526,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001345449712832924,
+      "loss": 3.1987,
+      "step": 15669
+    },
+    {
+      "epoch": 0.6597894736842105,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0001345150420475575,
+      "loss": 2.7454,
+      "step": 15670
+    },
+    {
+      "epoch": 0.6598315789473684,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00013448511491582295,
+      "loss": 3.2422,
+      "step": 15671
+    },
+    {
+      "epoch": 0.6598736842105263,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0001344551898886338,
+      "loss": 3.2176,
+      "step": 15672
+    },
+    {
+      "epoch": 0.6599157894736842,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0001344252669665354,
+      "loss": 3.2383,
+      "step": 15673
+    },
+    {
+      "epoch": 0.6599578947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00013439534615007284,
+      "loss": 3.0154,
+      "step": 15674
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00013436542743979125,
+      "loss": 3.4572,
+      "step": 15675
+    },
+    {
+      "epoch": 0.6600421052631579,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0001343355108362357,
+      "loss": 3.5481,
+      "step": 15676
+    },
+    {
+      "epoch": 0.6600842105263158,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001343055963399511,
+      "loss": 3.204,
+      "step": 15677
+    },
+    {
+      "epoch": 0.6601263157894737,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00013427568395148271,
+      "loss": 3.3478,
+      "step": 15678
+    },
+    {
+      "epoch": 0.6601684210526316,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0001342457736713754,
+      "loss": 3.2411,
+      "step": 15679
+    },
+    {
+      "epoch": 0.6602105263157895,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001342158655001741,
+      "loss": 2.8372,
+      "step": 15680
+    },
+    {
+      "epoch": 0.6602526315789473,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00013418595943842372,
+      "loss": 3.1275,
+      "step": 15681
+    },
+    {
+      "epoch": 0.6602947368421053,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00013415605548666905,
+      "loss": 3.1359,
+      "step": 15682
+    },
+    {
+      "epoch": 0.6603368421052631,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001341261536454549,
+      "loss": 3.167,
+      "step": 15683
+    },
+    {
+      "epoch": 0.6603789473684211,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00013409625391532605,
+      "loss": 3.4055,
+      "step": 15684
+    },
+    {
+      "epoch": 0.6604210526315789,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00013406635629682742,
+      "loss": 3.3793,
+      "step": 15685
+    },
+    {
+      "epoch": 0.6604631578947369,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001340364607905034,
+      "loss": 3.3549,
+      "step": 15686
+    },
+    {
+      "epoch": 0.6605052631578947,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00013400656739689888,
+      "loss": 3.3546,
+      "step": 15687
+    },
+    {
+      "epoch": 0.6605473684210527,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00013397667611655844,
+      "loss": 3.0565,
+      "step": 15688
+    },
+    {
+      "epoch": 0.6605894736842105,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00013394678695002654,
+      "loss": 3.0792,
+      "step": 15689
+    },
+    {
+      "epoch": 0.6606315789473685,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.000133916899897848,
+      "loss": 2.9204,
+      "step": 15690
+    },
+    {
+      "epoch": 0.6606736842105263,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00013388701496056698,
+      "loss": 3.0613,
+      "step": 15691
+    },
+    {
+      "epoch": 0.6607157894736843,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0001338571321387283,
+      "loss": 3.2818,
+      "step": 15692
+    },
+    {
+      "epoch": 0.6607578947368421,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00013382725143287603,
+      "loss": 3.4337,
+      "step": 15693
+    },
+    {
+      "epoch": 0.6608,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001337973728435548,
+      "loss": 2.7607,
+      "step": 15694
+    },
+    {
+      "epoch": 0.6608421052631579,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00013376749637130896,
+      "loss": 2.9089,
+      "step": 15695
+    },
+    {
+      "epoch": 0.6608842105263157,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00013373762201668275,
+      "loss": 3.4544,
+      "step": 15696
+    },
+    {
+      "epoch": 0.6609263157894737,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00013370774978022047,
+      "loss": 3.1535,
+      "step": 15697
+    },
+    {
+      "epoch": 0.6609684210526315,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00013367787966246623,
+      "loss": 2.8811,
+      "step": 15698
+    },
+    {
+      "epoch": 0.6610105263157895,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00013364801166396457,
+      "loss": 3.256,
+      "step": 15699
+    },
+    {
+      "epoch": 0.6610526315789473,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00013361814578525922,
+      "loss": 3.2203,
+      "step": 15700
+    },
+    {
+      "epoch": 0.6610947368421053,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0001335882820268947,
+      "loss": 2.9717,
+      "step": 15701
+    },
+    {
+      "epoch": 0.6611368421052631,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001335584203894147,
+      "loss": 3.589,
+      "step": 15702
+    },
+    {
+      "epoch": 0.6611789473684211,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00013352856087336357,
+      "loss": 3.4689,
+      "step": 15703
+    },
+    {
+      "epoch": 0.6612210526315789,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00013349870347928523,
+      "loss": 2.3425,
+      "step": 15704
+    },
+    {
+      "epoch": 0.6612631578947369,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00013346884820772354,
+      "loss": 3.1679,
+      "step": 15705
+    },
+    {
+      "epoch": 0.6613052631578947,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001334389950592227,
+      "loss": 3.2857,
+      "step": 15706
+    },
+    {
+      "epoch": 0.6613473684210527,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00013340914403432618,
+      "loss": 3.3305,
+      "step": 15707
+    },
+    {
+      "epoch": 0.6613894736842105,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00013337929513357831,
+      "loss": 3.0762,
+      "step": 15708
+    },
+    {
+      "epoch": 0.6614315789473684,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00013334944835752247,
+      "loss": 3.2639,
+      "step": 15709
+    },
+    {
+      "epoch": 0.6614736842105263,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001333196037067027,
+      "loss": 3.0184,
+      "step": 15710
+    },
+    {
+      "epoch": 0.6615157894736842,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00013328976118166263,
+      "loss": 3.3818,
+      "step": 15711
+    },
+    {
+      "epoch": 0.6615578947368421,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00013325992078294603,
+      "loss": 2.8022,
+      "step": 15712
+    },
+    {
+      "epoch": 0.6616,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0001332300825110965,
+      "loss": 3.0591,
+      "step": 15713
+    },
+    {
+      "epoch": 0.6616421052631579,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00013320024636665756,
+      "loss": 2.5197,
+      "step": 15714
+    },
+    {
+      "epoch": 0.6616842105263158,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001331704123501731,
+      "loss": 3.4446,
+      "step": 15715
+    },
+    {
+      "epoch": 0.6617263157894737,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00013314058046218625,
+      "loss": 2.8752,
+      "step": 15716
+    },
+    {
+      "epoch": 0.6617684210526316,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00013311075070324084,
+      "loss": 3.1053,
+      "step": 15717
+    },
+    {
+      "epoch": 0.6618105263157895,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0001330809230738802,
+      "loss": 2.795,
+      "step": 15718
+    },
+    {
+      "epoch": 0.6618526315789474,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00013305109757464774,
+      "loss": 3.0697,
+      "step": 15719
+    },
+    {
+      "epoch": 0.6618947368421053,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00013302127420608689,
+      "loss": 3.2741,
+      "step": 15720
+    },
+    {
+      "epoch": 0.6619368421052632,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0001329914529687409,
+      "loss": 3.6601,
+      "step": 15721
+    },
+    {
+      "epoch": 0.661978947368421,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00013296163386315336,
+      "loss": 2.5136,
+      "step": 15722
+    },
+    {
+      "epoch": 0.662021052631579,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00013293181688986712,
+      "loss": 3.1676,
+      "step": 15723
+    },
+    {
+      "epoch": 0.6620631578947368,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00013290200204942572,
+      "loss": 3.1506,
+      "step": 15724
+    },
+    {
+      "epoch": 0.6621052631578948,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00013287218934237226,
+      "loss": 2.6463,
+      "step": 15725
+    },
+    {
+      "epoch": 0.6621473684210526,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0001328423787692499,
+      "loss": 2.7638,
+      "step": 15726
+    },
+    {
+      "epoch": 0.6621894736842105,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00013281257033060173,
+      "loss": 3.1715,
+      "step": 15727
+    },
+    {
+      "epoch": 0.6622315789473684,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00013278276402697086,
+      "loss": 3.212,
+      "step": 15728
+    },
+    {
+      "epoch": 0.6622736842105263,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0001327529598589003,
+      "loss": 3.1292,
+      "step": 15729
+    },
+    {
+      "epoch": 0.6623157894736842,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00013272315782693296,
+      "loss": 3.3163,
+      "step": 15730
+    },
+    {
+      "epoch": 0.6623578947368421,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00013269335793161197,
+      "loss": 2.9168,
+      "step": 15731
+    },
+    {
+      "epoch": 0.6624,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00013266356017348015,
+      "loss": 3.2731,
+      "step": 15732
+    },
+    {
+      "epoch": 0.6624421052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00013263376455308043,
+      "loss": 3.0833,
+      "step": 15733
+    },
+    {
+      "epoch": 0.6624842105263158,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00013260397107095563,
+      "loss": 2.8879,
+      "step": 15734
+    },
+    {
+      "epoch": 0.6625263157894736,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0001325741797276485,
+      "loss": 3.2165,
+      "step": 15735
+    },
+    {
+      "epoch": 0.6625684210526316,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0001325443905237018,
+      "loss": 2.933,
+      "step": 15736
+    },
+    {
+      "epoch": 0.6626105263157894,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00013251460345965838,
+      "loss": 2.9521,
+      "step": 15737
+    },
+    {
+      "epoch": 0.6626526315789474,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001324848185360608,
+      "loss": 2.6726,
+      "step": 15738
+    },
+    {
+      "epoch": 0.6626947368421052,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001324550357534518,
+      "loss": 2.8963,
+      "step": 15739
+    },
+    {
+      "epoch": 0.6627368421052632,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00013242525511237394,
+      "loss": 2.882,
+      "step": 15740
+    },
+    {
+      "epoch": 0.662778947368421,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00013239547661336975,
+      "loss": 2.807,
+      "step": 15741
+    },
+    {
+      "epoch": 0.662821052631579,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00013236570025698186,
+      "loss": 3.15,
+      "step": 15742
+    },
+    {
+      "epoch": 0.6628631578947368,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00013233592604375256,
+      "loss": 3.0825,
+      "step": 15743
+    },
+    {
+      "epoch": 0.6629052631578948,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00013230615397422462,
+      "loss": 3.0947,
+      "step": 15744
+    },
+    {
+      "epoch": 0.6629473684210526,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0001322763840489401,
+      "loss": 3.1431,
+      "step": 15745
+    },
+    {
+      "epoch": 0.6629894736842106,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00013224661626844164,
+      "loss": 3.3134,
+      "step": 15746
+    },
+    {
+      "epoch": 0.6630315789473684,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00013221685063327144,
+      "loss": 3.2757,
+      "step": 15747
+    },
+    {
+      "epoch": 0.6630736842105264,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00013218708714397186,
+      "loss": 3.5356,
+      "step": 15748
+    },
+    {
+      "epoch": 0.6631157894736842,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00013215732580108513,
+      "loss": 3.2489,
+      "step": 15749
+    },
+    {
+      "epoch": 0.6631578947368421,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00013212756660515337,
+      "loss": 2.8478,
+      "step": 15750
+    },
+    {
+      "epoch": 0.6632,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00013209780955671904,
+      "loss": 3.173,
+      "step": 15751
+    },
+    {
+      "epoch": 0.6632421052631579,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00013206805465632387,
+      "loss": 2.9065,
+      "step": 15752
+    },
+    {
+      "epoch": 0.6632842105263158,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00013203830190451033,
+      "loss": 3.2499,
+      "step": 15753
+    },
+    {
+      "epoch": 0.6633263157894737,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00013200855130182033,
+      "loss": 2.926,
+      "step": 15754
+    },
+    {
+      "epoch": 0.6633684210526316,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00013197880284879582,
+      "loss": 3.4132,
+      "step": 15755
+    },
+    {
+      "epoch": 0.6634105263157895,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0001319490565459789,
+      "loss": 2.9822,
+      "step": 15756
+    },
+    {
+      "epoch": 0.6634526315789474,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00013191931239391137,
+      "loss": 2.9443,
+      "step": 15757
+    },
+    {
+      "epoch": 0.6634947368421052,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00013188957039313542,
+      "loss": 3.2769,
+      "step": 15758
+    },
+    {
+      "epoch": 0.6635368421052632,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00013185983054419254,
+      "loss": 3.15,
+      "step": 15759
+    },
+    {
+      "epoch": 0.663578947368421,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00013183009284762492,
+      "loss": 3.4536,
+      "step": 15760
+    },
+    {
+      "epoch": 0.663621052631579,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00013180035730397396,
+      "loss": 3.0111,
+      "step": 15761
+    },
+    {
+      "epoch": 0.6636631578947368,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00013177062391378172,
+      "loss": 3.0216,
+      "step": 15762
+    },
+    {
+      "epoch": 0.6637052631578947,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00013174089267758982,
+      "loss": 3.2921,
+      "step": 15763
+    },
+    {
+      "epoch": 0.6637473684210526,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00013171116359593976,
+      "loss": 2.9425,
+      "step": 15764
+    },
+    {
+      "epoch": 0.6637894736842105,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00013168143666937354,
+      "loss": 3.3366,
+      "step": 15765
+    },
+    {
+      "epoch": 0.6638315789473684,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00013165171189843234,
+      "loss": 3.2112,
+      "step": 15766
+    },
+    {
+      "epoch": 0.6638736842105263,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0001316219892836581,
+      "loss": 3.1818,
+      "step": 15767
+    },
+    {
+      "epoch": 0.6639157894736842,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0001315922688255919,
+      "loss": 2.6691,
+      "step": 15768
+    },
+    {
+      "epoch": 0.6639578947368421,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00013156255052477553,
+      "loss": 3.0826,
+      "step": 15769
+    },
+    {
+      "epoch": 0.664,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00013153283438175034,
+      "loss": 2.8736,
+      "step": 15770
+    },
+    {
+      "epoch": 0.6640421052631579,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00013150312039705774,
+      "loss": 2.9973,
+      "step": 15771
+    },
+    {
+      "epoch": 0.6640842105263158,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00013147340857123902,
+      "loss": 2.7762,
+      "step": 15772
+    },
+    {
+      "epoch": 0.6641263157894737,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00013144369890483548,
+      "loss": 2.8798,
+      "step": 15773
+    },
+    {
+      "epoch": 0.6641684210526316,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0001314139913983886,
+      "loss": 2.8144,
+      "step": 15774
+    },
+    {
+      "epoch": 0.6642105263157895,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0001313842860524393,
+      "loss": 2.5495,
+      "step": 15775
+    },
+    {
+      "epoch": 0.6642526315789473,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0001313545828675291,
+      "loss": 2.9559,
+      "step": 15776
+    },
+    {
+      "epoch": 0.6642947368421053,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00013132488184419884,
+      "loss": 2.9774,
+      "step": 15777
+    },
+    {
+      "epoch": 0.6643368421052631,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00013129518298298987,
+      "loss": 2.7229,
+      "step": 15778
+    },
+    {
+      "epoch": 0.6643789473684211,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00013126548628444323,
+      "loss": 2.816,
+      "step": 15779
+    },
+    {
+      "epoch": 0.6644210526315789,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00013123579174909977,
+      "loss": 3.2036,
+      "step": 15780
+    },
+    {
+      "epoch": 0.6644631578947369,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001312060993775009,
+      "loss": 2.5084,
+      "step": 15781
+    },
+    {
+      "epoch": 0.6645052631578947,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001311764091701871,
+      "loss": 3.0876,
+      "step": 15782
+    },
+    {
+      "epoch": 0.6645473684210527,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00013114672112769966,
+      "loss": 2.8164,
+      "step": 15783
+    },
+    {
+      "epoch": 0.6645894736842105,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00013111703525057916,
+      "loss": 2.9774,
+      "step": 15784
+    },
+    {
+      "epoch": 0.6646315789473685,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00013108735153936668,
+      "loss": 2.8797,
+      "step": 15785
+    },
+    {
+      "epoch": 0.6646736842105263,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00013105766999460297,
+      "loss": 2.8749,
+      "step": 15786
+    },
+    {
+      "epoch": 0.6647157894736843,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00013102799061682875,
+      "loss": 2.8532,
+      "step": 15787
+    },
+    {
+      "epoch": 0.6647578947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00013099831340658474,
+      "loss": 2.6231,
+      "step": 15788
+    },
+    {
+      "epoch": 0.6648,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00013096863836441154,
+      "loss": 2.8444,
+      "step": 15789
+    },
+    {
+      "epoch": 0.6648421052631579,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00013093896549085,
+      "loss": 3.0132,
+      "step": 15790
+    },
+    {
+      "epoch": 0.6648842105263157,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0001309092947864406,
+      "loss": 2.4935,
+      "step": 15791
+    },
+    {
+      "epoch": 0.6649263157894737,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00013087962625172396,
+      "loss": 3.209,
+      "step": 15792
+    },
+    {
+      "epoch": 0.6649684210526315,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00013084995988724054,
+      "loss": 3.0388,
+      "step": 15793
+    },
+    {
+      "epoch": 0.6650105263157895,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00013082029569353088,
+      "loss": 2.7978,
+      "step": 15794
+    },
+    {
+      "epoch": 0.6650526315789473,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00013079063367113535,
+      "loss": 2.6761,
+      "step": 15795
+    },
+    {
+      "epoch": 0.6650947368421053,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00013076097382059437,
+      "loss": 3.4351,
+      "step": 15796
+    },
+    {
+      "epoch": 0.6651368421052631,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00013073131614244843,
+      "loss": 3.068,
+      "step": 15797
+    },
+    {
+      "epoch": 0.6651789473684211,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00013070166063723775,
+      "loss": 3.0334,
+      "step": 15798
+    },
+    {
+      "epoch": 0.6652210526315789,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00013067200730550266,
+      "loss": 2.8792,
+      "step": 15799
+    },
+    {
+      "epoch": 0.6652631578947369,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00013064235614778342,
+      "loss": 3.2272,
+      "step": 15800
+    },
+    {
+      "epoch": 0.6653052631578947,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00013061270716462016,
+      "loss": 3.505,
+      "step": 15801
+    },
+    {
+      "epoch": 0.6653473684210526,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00013058306035655312,
+      "loss": 2.9735,
+      "step": 15802
+    },
+    {
+      "epoch": 0.6653894736842105,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00013055341572412244,
+      "loss": 2.9673,
+      "step": 15803
+    },
+    {
+      "epoch": 0.6654315789473684,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00013052377326786804,
+      "loss": 3.0999,
+      "step": 15804
+    },
+    {
+      "epoch": 0.6654736842105263,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00013049413298833024,
+      "loss": 3.18,
+      "step": 15805
+    },
+    {
+      "epoch": 0.6655157894736842,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0001304644948860489,
+      "loss": 2.9492,
+      "step": 15806
+    },
+    {
+      "epoch": 0.6655578947368421,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00013043485896156403,
+      "loss": 3.547,
+      "step": 15807
+    },
+    {
+      "epoch": 0.6656,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00013040522521541554,
+      "loss": 3.3056,
+      "step": 15808
+    },
+    {
+      "epoch": 0.6656421052631579,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00013037559364814332,
+      "loss": 2.8112,
+      "step": 15809
+    },
+    {
+      "epoch": 0.6656842105263158,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00013034596426028729,
+      "loss": 3.0902,
+      "step": 15810
+    },
+    {
+      "epoch": 0.6657263157894737,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00013031633705238704,
+      "loss": 3.2839,
+      "step": 15811
+    },
+    {
+      "epoch": 0.6657684210526316,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0001302867120249826,
+      "loss": 2.9039,
+      "step": 15812
+    },
+    {
+      "epoch": 0.6658105263157895,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00013025708917861365,
+      "loss": 3.1066,
+      "step": 15813
+    },
+    {
+      "epoch": 0.6658526315789474,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00013022746851381985,
+      "loss": 3.1412,
+      "step": 15814
+    },
+    {
+      "epoch": 0.6658947368421053,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0001301978500311408,
+      "loss": 2.9414,
+      "step": 15815
+    },
+    {
+      "epoch": 0.6659368421052632,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001301682337311162,
+      "loss": 3.0336,
+      "step": 15816
+    },
+    {
+      "epoch": 0.665978947368421,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00013013861961428555,
+      "loss": 3.1247,
+      "step": 15817
+    },
+    {
+      "epoch": 0.666021052631579,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00013010900768118834,
+      "loss": 3.245,
+      "step": 15818
+    },
+    {
+      "epoch": 0.6660631578947368,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00013007939793236435,
+      "loss": 3.4649,
+      "step": 15819
+    },
+    {
+      "epoch": 0.6661052631578948,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00013004979036835264,
+      "loss": 2.8322,
+      "step": 15820
+    },
+    {
+      "epoch": 0.6661473684210526,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00013002018498969287,
+      "loss": 3.3361,
+      "step": 15821
+    },
+    {
+      "epoch": 0.6661894736842106,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00012999058179692438,
+      "loss": 2.9639,
+      "step": 15822
+    },
+    {
+      "epoch": 0.6662315789473684,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00012996098079058648,
+      "loss": 3.3286,
+      "step": 15823
+    },
+    {
+      "epoch": 0.6662736842105264,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00012993138197121848,
+      "loss": 3.2119,
+      "step": 15824
+    },
+    {
+      "epoch": 0.6663157894736842,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00012990178533935954,
+      "loss": 2.9074,
+      "step": 15825
+    },
+    {
+      "epoch": 0.6663578947368421,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00012987219089554913,
+      "loss": 3.6743,
+      "step": 15826
+    },
+    {
+      "epoch": 0.6664,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00012984259864032608,
+      "loss": 3.5049,
+      "step": 15827
+    },
+    {
+      "epoch": 0.666442105263158,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001298130085742299,
+      "loss": 2.8127,
+      "step": 15828
+    },
+    {
+      "epoch": 0.6664842105263158,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00012978342069779928,
+      "loss": 3.1235,
+      "step": 15829
+    },
+    {
+      "epoch": 0.6665263157894736,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0001297538350115736,
+      "loss": 2.9444,
+      "step": 15830
+    },
+    {
+      "epoch": 0.6665684210526316,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00012972425151609173,
+      "loss": 2.8065,
+      "step": 15831
+    },
+    {
+      "epoch": 0.6666105263157894,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00012969467021189263,
+      "loss": 3.1415,
+      "step": 15832
+    },
+    {
+      "epoch": 0.6666526315789474,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00012966509109951546,
+      "loss": 2.776,
+      "step": 15833
+    },
+    {
+      "epoch": 0.6666947368421052,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00012963551417949874,
+      "loss": 3.0222,
+      "step": 15834
+    },
+    {
+      "epoch": 0.6667368421052632,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00012960593945238172,
+      "loss": 3.1074,
+      "step": 15835
+    },
+    {
+      "epoch": 0.666778947368421,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00012957636691870284,
+      "loss": 3.2742,
+      "step": 15836
+    },
+    {
+      "epoch": 0.666821052631579,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00012954679657900113,
+      "loss": 3.2158,
+      "step": 15837
+    },
+    {
+      "epoch": 0.6668631578947368,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0001295172284338153,
+      "loss": 3.1544,
+      "step": 15838
+    },
+    {
+      "epoch": 0.6669052631578948,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00012948766248368392,
+      "loss": 2.5699,
+      "step": 15839
+    },
+    {
+      "epoch": 0.6669473684210526,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00012945809872914592,
+      "loss": 3.2143,
+      "step": 15840
+    },
+    {
+      "epoch": 0.6669894736842106,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00012942853717073948,
+      "loss": 3.1279,
+      "step": 15841
+    },
+    {
+      "epoch": 0.6670315789473684,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001293989778090037,
+      "loss": 2.9067,
+      "step": 15842
+    },
+    {
+      "epoch": 0.6670736842105263,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001293694206444766,
+      "loss": 3.4485,
+      "step": 15843
+    },
+    {
+      "epoch": 0.6671157894736842,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00012933986567769705,
+      "loss": 2.917,
+      "step": 15844
+    },
+    {
+      "epoch": 0.6671578947368421,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00012931031290920337,
+      "loss": 2.8975,
+      "step": 15845
+    },
+    {
+      "epoch": 0.6672,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00012928076233953396,
+      "loss": 3.2469,
+      "step": 15846
+    },
+    {
+      "epoch": 0.6672421052631579,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00012925121396922722,
+      "loss": 3.114,
+      "step": 15847
+    },
+    {
+      "epoch": 0.6672842105263158,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00012922166779882142,
+      "loss": 3.2593,
+      "step": 15848
+    },
+    {
+      "epoch": 0.6673263157894737,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00012919212382885508,
+      "loss": 3.1166,
+      "step": 15849
+    },
+    {
+      "epoch": 0.6673684210526316,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00012916258205986615,
+      "loss": 3.1908,
+      "step": 15850
+    },
+    {
+      "epoch": 0.6674105263157895,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00012913304249239304,
+      "loss": 3.0478,
+      "step": 15851
+    },
+    {
+      "epoch": 0.6674526315789474,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00012910350512697395,
+      "loss": 2.9303,
+      "step": 15852
+    },
+    {
+      "epoch": 0.6674947368421053,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00012907396996414693,
+      "loss": 3.3043,
+      "step": 15853
+    },
+    {
+      "epoch": 0.6675368421052632,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00012904443700445008,
+      "loss": 2.8537,
+      "step": 15854
+    },
+    {
+      "epoch": 0.667578947368421,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00012901490624842143,
+      "loss": 2.9666,
+      "step": 15855
+    },
+    {
+      "epoch": 0.6676210526315789,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0001289853776965991,
+      "loss": 3.0076,
+      "step": 15856
+    },
+    {
+      "epoch": 0.6676631578947368,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00012895585134952087,
+      "loss": 2.7696,
+      "step": 15857
+    },
+    {
+      "epoch": 0.6677052631578947,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00012892632720772494,
+      "loss": 3.2988,
+      "step": 15858
+    },
+    {
+      "epoch": 0.6677473684210526,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00012889680527174903,
+      "loss": 2.9033,
+      "step": 15859
+    },
+    {
+      "epoch": 0.6677894736842105,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00012886728554213108,
+      "loss": 3.1483,
+      "step": 15860
+    },
+    {
+      "epoch": 0.6678315789473684,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00012883776801940884,
+      "loss": 3.1064,
+      "step": 15861
+    },
+    {
+      "epoch": 0.6678736842105263,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00012880825270412008,
+      "loss": 2.9161,
+      "step": 15862
+    },
+    {
+      "epoch": 0.6679157894736842,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00012877873959680258,
+      "loss": 2.645,
+      "step": 15863
+    },
+    {
+      "epoch": 0.6679578947368421,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00012874922869799394,
+      "loss": 2.843,
+      "step": 15864
+    },
+    {
+      "epoch": 0.668,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00012871972000823195,
+      "loss": 3.0943,
+      "step": 15865
+    },
+    {
+      "epoch": 0.6680421052631579,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00012869021352805422,
+      "loss": 3.4077,
+      "step": 15866
+    },
+    {
+      "epoch": 0.6680842105263158,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00012866070925799823,
+      "loss": 2.5298,
+      "step": 15867
+    },
+    {
+      "epoch": 0.6681263157894737,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00012863120719860155,
+      "loss": 3.514,
+      "step": 15868
+    },
+    {
+      "epoch": 0.6681684210526316,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00012860170735040167,
+      "loss": 2.8745,
+      "step": 15869
+    },
+    {
+      "epoch": 0.6682105263157895,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00012857220971393605,
+      "loss": 2.6445,
+      "step": 15870
+    },
+    {
+      "epoch": 0.6682526315789473,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001285427142897421,
+      "loss": 2.6437,
+      "step": 15871
+    },
+    {
+      "epoch": 0.6682947368421053,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00012851322107835709,
+      "loss": 2.8648,
+      "step": 15872
+    },
+    {
+      "epoch": 0.6683368421052631,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00012848373008031856,
+      "loss": 3.3423,
+      "step": 15873
+    },
+    {
+      "epoch": 0.6683789473684211,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00012845424129616365,
+      "loss": 3.0569,
+      "step": 15874
+    },
+    {
+      "epoch": 0.6684210526315789,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00012842475472642968,
+      "loss": 3.3816,
+      "step": 15875
+    },
+    {
+      "epoch": 0.6684631578947369,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00012839527037165384,
+      "loss": 3.1974,
+      "step": 15876
+    },
+    {
+      "epoch": 0.6685052631578947,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00012836578823237327,
+      "loss": 3.7311,
+      "step": 15877
+    },
+    {
+      "epoch": 0.6685473684210527,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00012833630830912516,
+      "loss": 3.2166,
+      "step": 15878
+    },
+    {
+      "epoch": 0.6685894736842105,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00012830683060244646,
+      "loss": 3.0744,
+      "step": 15879
+    },
+    {
+      "epoch": 0.6686315789473685,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00012827735511287443,
+      "loss": 3.2041,
+      "step": 15880
+    },
+    {
+      "epoch": 0.6686736842105263,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00012824788184094592,
+      "loss": 2.9794,
+      "step": 15881
+    },
+    {
+      "epoch": 0.6687157894736843,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00012821841078719797,
+      "loss": 3.0495,
+      "step": 15882
+    },
+    {
+      "epoch": 0.6687578947368421,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00012818894195216752,
+      "loss": 3.6279,
+      "step": 15883
+    },
+    {
+      "epoch": 0.6688,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00012815947533639137,
+      "loss": 2.7101,
+      "step": 15884
+    },
+    {
+      "epoch": 0.6688421052631579,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0001281300109404065,
+      "loss": 3.3523,
+      "step": 15885
+    },
+    {
+      "epoch": 0.6688842105263157,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00012810054876474946,
+      "loss": 3.1733,
+      "step": 15886
+    },
+    {
+      "epoch": 0.6689263157894737,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00012807108880995744,
+      "loss": 2.7345,
+      "step": 15887
+    },
+    {
+      "epoch": 0.6689684210526315,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00012804163107656675,
+      "loss": 3.5655,
+      "step": 15888
+    },
+    {
+      "epoch": 0.6690105263157895,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0001280121755651143,
+      "loss": 2.7114,
+      "step": 15889
+    },
+    {
+      "epoch": 0.6690526315789473,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00012798272227613667,
+      "loss": 3.3194,
+      "step": 15890
+    },
+    {
+      "epoch": 0.6690947368421053,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0001279532712101704,
+      "loss": 3.0179,
+      "step": 15891
+    },
+    {
+      "epoch": 0.6691368421052631,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00012792382236775235,
+      "loss": 3.1671,
+      "step": 15892
+    },
+    {
+      "epoch": 0.6691789473684211,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0001278943757494186,
+      "loss": 3.1426,
+      "step": 15893
+    },
+    {
+      "epoch": 0.6692210526315789,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00012786493135570604,
+      "loss": 3.3884,
+      "step": 15894
+    },
+    {
+      "epoch": 0.6692631578947369,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00012783548918715077,
+      "loss": 2.9857,
+      "step": 15895
+    },
+    {
+      "epoch": 0.6693052631578947,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0001278060492442894,
+      "loss": 2.8803,
+      "step": 15896
+    },
+    {
+      "epoch": 0.6693473684210526,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00012777661152765826,
+      "loss": 2.961,
+      "step": 15897
+    },
+    {
+      "epoch": 0.6693894736842105,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00012774717603779369,
+      "loss": 3.1851,
+      "step": 15898
+    },
+    {
+      "epoch": 0.6694315789473684,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00012771774277523188,
+      "loss": 2.7351,
+      "step": 15899
+    },
+    {
+      "epoch": 0.6694736842105263,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00012768831174050905,
+      "loss": 2.7822,
+      "step": 15900
+    },
+    {
+      "epoch": 0.6695157894736842,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00012765888293416163,
+      "loss": 3.2094,
+      "step": 15901
+    },
+    {
+      "epoch": 0.6695578947368421,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00012762945635672545,
+      "loss": 2.9743,
+      "step": 15902
+    },
+    {
+      "epoch": 0.6696,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00012760003200873698,
+      "loss": 3.468,
+      "step": 15903
+    },
+    {
+      "epoch": 0.6696421052631579,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00012757060989073194,
+      "loss": 3.1297,
+      "step": 15904
+    },
+    {
+      "epoch": 0.6696842105263158,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00012754119000324662,
+      "loss": 2.5427,
+      "step": 15905
+    },
+    {
+      "epoch": 0.6697263157894737,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00012751177234681695,
+      "loss": 3.3635,
+      "step": 15906
+    },
+    {
+      "epoch": 0.6697684210526316,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00012748235692197873,
+      "loss": 2.9842,
+      "step": 15907
+    },
+    {
+      "epoch": 0.6698105263157895,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00012745294372926825,
+      "loss": 3.0108,
+      "step": 15908
+    },
+    {
+      "epoch": 0.6698526315789474,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00012742353276922098,
+      "loss": 3.1352,
+      "step": 15909
+    },
+    {
+      "epoch": 0.6698947368421052,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00012739412404237305,
+      "loss": 3.3357,
+      "step": 15910
+    },
+    {
+      "epoch": 0.6699368421052632,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00012736471754925995,
+      "loss": 2.9537,
+      "step": 15911
+    },
+    {
+      "epoch": 0.669978947368421,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00012733531329041774,
+      "loss": 3.0622,
+      "step": 15912
+    },
+    {
+      "epoch": 0.670021052631579,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00012730591126638196,
+      "loss": 2.9481,
+      "step": 15913
+    },
+    {
+      "epoch": 0.6700631578947368,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0001272765114776883,
+      "loss": 3.0313,
+      "step": 15914
+    },
+    {
+      "epoch": 0.6701052631578948,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00012724711392487245,
+      "loss": 3.1777,
+      "step": 15915
+    },
+    {
+      "epoch": 0.6701473684210526,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00012721771860846987,
+      "loss": 3.0189,
+      "step": 15916
+    },
+    {
+      "epoch": 0.6701894736842106,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00012718832552901632,
+      "loss": 3.4344,
+      "step": 15917
+    },
+    {
+      "epoch": 0.6702315789473684,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00012715893468704704,
+      "loss": 2.7198,
+      "step": 15918
+    },
+    {
+      "epoch": 0.6702736842105264,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00012712954608309774,
+      "loss": 2.9259,
+      "step": 15919
+    },
+    {
+      "epoch": 0.6703157894736842,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00012710015971770372,
+      "loss": 3.0983,
+      "step": 15920
+    },
+    {
+      "epoch": 0.6703578947368422,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001270707755914004,
+      "loss": 2.9564,
+      "step": 15921
+    },
+    {
+      "epoch": 0.6704,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00012704139370472308,
+      "loss": 2.9131,
+      "step": 15922
+    },
+    {
+      "epoch": 0.6704421052631578,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00012701201405820704,
+      "loss": 3.1411,
+      "step": 15923
+    },
+    {
+      "epoch": 0.6704842105263158,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001269826366523878,
+      "loss": 2.3204,
+      "step": 15924
+    },
+    {
+      "epoch": 0.6705263157894736,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00012695326148780012,
+      "loss": 3.0725,
+      "step": 15925
+    },
+    {
+      "epoch": 0.6705684210526316,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001269238885649796,
+      "loss": 3.1156,
+      "step": 15926
+    },
+    {
+      "epoch": 0.6706105263157894,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001268945178844612,
+      "loss": 3.3699,
+      "step": 15927
+    },
+    {
+      "epoch": 0.6706526315789474,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00012686514944678008,
+      "loss": 3.2315,
+      "step": 15928
+    },
+    {
+      "epoch": 0.6706947368421052,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00012683578325247118,
+      "loss": 2.8811,
+      "step": 15929
+    },
+    {
+      "epoch": 0.6707368421052632,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00012680641930206966,
+      "loss": 3.1503,
+      "step": 15930
+    },
+    {
+      "epoch": 0.670778947368421,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0001267770575961104,
+      "loss": 3.0572,
+      "step": 15931
+    },
+    {
+      "epoch": 0.670821052631579,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00012674769813512832,
+      "loss": 3.3249,
+      "step": 15932
+    },
+    {
+      "epoch": 0.6708631578947368,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00012671834091965842,
+      "loss": 2.9281,
+      "step": 15933
+    },
+    {
+      "epoch": 0.6709052631578948,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00012668898595023553,
+      "loss": 2.9572,
+      "step": 15934
+    },
+    {
+      "epoch": 0.6709473684210526,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00012665963322739437,
+      "loss": 2.9123,
+      "step": 15935
+    },
+    {
+      "epoch": 0.6709894736842106,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00012663028275166981,
+      "loss": 3.1905,
+      "step": 15936
+    },
+    {
+      "epoch": 0.6710315789473684,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0001266009345235966,
+      "loss": 3.0376,
+      "step": 15937
+    },
+    {
+      "epoch": 0.6710736842105263,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0001265715885437092,
+      "loss": 2.8841,
+      "step": 15938
+    },
+    {
+      "epoch": 0.6711157894736842,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00012654224481254258,
+      "loss": 2.5057,
+      "step": 15939
+    },
+    {
+      "epoch": 0.6711578947368421,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00012651290333063117,
+      "loss": 2.7287,
+      "step": 15940
+    },
+    {
+      "epoch": 0.6712,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001264835640985096,
+      "loss": 3.2412,
+      "step": 15941
+    },
+    {
+      "epoch": 0.6712421052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00012645422711671234,
+      "loss": 3.3325,
+      "step": 15942
+    },
+    {
+      "epoch": 0.6712842105263158,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0001264248923857739,
+      "loss": 2.9342,
+      "step": 15943
+    },
+    {
+      "epoch": 0.6713263157894737,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00012639555990622876,
+      "loss": 3.1991,
+      "step": 15944
+    },
+    {
+      "epoch": 0.6713684210526316,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00012636622967861113,
+      "loss": 2.9292,
+      "step": 15945
+    },
+    {
+      "epoch": 0.6714105263157895,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0001263369017034558,
+      "loss": 3.3595,
+      "step": 15946
+    },
+    {
+      "epoch": 0.6714526315789474,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00012630757598129655,
+      "loss": 2.779,
+      "step": 15947
+    },
+    {
+      "epoch": 0.6714947368421053,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00012627825251266812,
+      "loss": 2.7281,
+      "step": 15948
+    },
+    {
+      "epoch": 0.6715368421052632,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0001262489312981045,
+      "loss": 3.003,
+      "step": 15949
+    },
+    {
+      "epoch": 0.671578947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00012621961233813994,
+      "loss": 3.3618,
+      "step": 15950
+    },
+    {
+      "epoch": 0.6716210526315789,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00012619029563330864,
+      "loss": 3.1853,
+      "step": 15951
+    },
+    {
+      "epoch": 0.6716631578947369,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001261609811841446,
+      "loss": 3.0302,
+      "step": 15952
+    },
+    {
+      "epoch": 0.6717052631578947,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00012613166899118212,
+      "loss": 3.0751,
+      "step": 15953
+    },
+    {
+      "epoch": 0.6717473684210526,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00012610235905495494,
+      "loss": 2.8066,
+      "step": 15954
+    },
+    {
+      "epoch": 0.6717894736842105,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00012607305137599727,
+      "loss": 3.1997,
+      "step": 15955
+    },
+    {
+      "epoch": 0.6718315789473684,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00012604374595484303,
+      "loss": 3.1338,
+      "step": 15956
+    },
+    {
+      "epoch": 0.6718736842105263,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00012601444279202608,
+      "loss": 3.2804,
+      "step": 15957
+    },
+    {
+      "epoch": 0.6719157894736842,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001259851418880803,
+      "loss": 2.7821,
+      "step": 15958
+    },
+    {
+      "epoch": 0.6719578947368421,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00012595584324353942,
+      "loss": 2.957,
+      "step": 15959
+    },
+    {
+      "epoch": 0.672,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00012592654685893756,
+      "loss": 3.1528,
+      "step": 15960
+    },
+    {
+      "epoch": 0.6720421052631579,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00012589725273480802,
+      "loss": 3.1736,
+      "step": 15961
+    },
+    {
+      "epoch": 0.6720842105263158,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001258679608716849,
+      "loss": 3.5325,
+      "step": 15962
+    },
+    {
+      "epoch": 0.6721263157894737,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0001258386712701015,
+      "loss": 2.9699,
+      "step": 15963
+    },
+    {
+      "epoch": 0.6721684210526315,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00012580938393059175,
+      "loss": 3.2067,
+      "step": 15964
+    },
+    {
+      "epoch": 0.6722105263157895,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00012578009885368907,
+      "loss": 2.9865,
+      "step": 15965
+    },
+    {
+      "epoch": 0.6722526315789473,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00012575081603992695,
+      "loss": 3.4989,
+      "step": 15966
+    },
+    {
+      "epoch": 0.6722947368421053,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00012572153548983915,
+      "loss": 3.1228,
+      "step": 15967
+    },
+    {
+      "epoch": 0.6723368421052631,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00012569225720395882,
+      "loss": 2.8226,
+      "step": 15968
+    },
+    {
+      "epoch": 0.6723789473684211,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00012566298118281966,
+      "loss": 2.6992,
+      "step": 15969
+    },
+    {
+      "epoch": 0.6724210526315789,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00012563370742695468,
+      "loss": 3.7462,
+      "step": 15970
+    },
+    {
+      "epoch": 0.6724631578947369,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00012560443593689755,
+      "loss": 3.1766,
+      "step": 15971
+    },
+    {
+      "epoch": 0.6725052631578947,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00012557516671318143,
+      "loss": 2.9605,
+      "step": 15972
+    },
+    {
+      "epoch": 0.6725473684210527,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00012554589975633958,
+      "loss": 3.0569,
+      "step": 15973
+    },
+    {
+      "epoch": 0.6725894736842105,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00012551663506690523,
+      "loss": 3.1149,
+      "step": 15974
+    },
+    {
+      "epoch": 0.6726315789473685,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00012548737264541138,
+      "loss": 2.9173,
+      "step": 15975
+    },
+    {
+      "epoch": 0.6726736842105263,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00012545811249239152,
+      "loss": 2.5799,
+      "step": 15976
+    },
+    {
+      "epoch": 0.6727157894736842,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00012542885460837836,
+      "loss": 3.2402,
+      "step": 15977
+    },
+    {
+      "epoch": 0.6727578947368421,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00012539959899390525,
+      "loss": 3.3866,
+      "step": 15978
+    },
+    {
+      "epoch": 0.6728,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0001253703456495049,
+      "loss": 2.9762,
+      "step": 15979
+    },
+    {
+      "epoch": 0.6728421052631579,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00012534109457571047,
+      "loss": 2.9285,
+      "step": 15980
+    },
+    {
+      "epoch": 0.6728842105263158,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00012531184577305488,
+      "loss": 3.2687,
+      "step": 15981
+    },
+    {
+      "epoch": 0.6729263157894737,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0001252825992420708,
+      "loss": 3.0216,
+      "step": 15982
+    },
+    {
+      "epoch": 0.6729684210526315,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00012525335498329146,
+      "loss": 3.246,
+      "step": 15983
+    },
+    {
+      "epoch": 0.6730105263157895,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00012522411299724923,
+      "loss": 2.8129,
+      "step": 15984
+    },
+    {
+      "epoch": 0.6730526315789473,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00012519487328447712,
+      "loss": 3.116,
+      "step": 15985
+    },
+    {
+      "epoch": 0.6730947368421053,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0001251656358455078,
+      "loss": 3.4185,
+      "step": 15986
+    },
+    {
+      "epoch": 0.6731368421052631,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00012513640068087393,
+      "loss": 3.5245,
+      "step": 15987
+    },
+    {
+      "epoch": 0.6731789473684211,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0001251071677911081,
+      "loss": 3.0252,
+      "step": 15988
+    },
+    {
+      "epoch": 0.6732210526315789,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00012507793717674294,
+      "loss": 3.0259,
+      "step": 15989
+    },
+    {
+      "epoch": 0.6732631578947369,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00012504870883831102,
+      "loss": 3.4577,
+      "step": 15990
+    },
+    {
+      "epoch": 0.6733052631578947,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00012501948277634468,
+      "loss": 3.2727,
+      "step": 15991
+    },
+    {
+      "epoch": 0.6733473684210526,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00012499025899137664,
+      "loss": 3.0052,
+      "step": 15992
+    },
+    {
+      "epoch": 0.6733894736842105,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0001249610374839392,
+      "loss": 3.4769,
+      "step": 15993
+    },
+    {
+      "epoch": 0.6734315789473684,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00012493181825456472,
+      "loss": 3.4875,
+      "step": 15994
+    },
+    {
+      "epoch": 0.6734736842105263,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00012490260130378555,
+      "loss": 3.2303,
+      "step": 15995
+    },
+    {
+      "epoch": 0.6735157894736842,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00012487338663213403,
+      "loss": 2.9853,
+      "step": 15996
+    },
+    {
+      "epoch": 0.6735578947368421,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00012484417424014237,
+      "loss": 3.518,
+      "step": 15997
+    },
+    {
+      "epoch": 0.6736,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00012481496412834273,
+      "loss": 3.184,
+      "step": 15998
+    },
+    {
+      "epoch": 0.6736421052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00012478575629726744,
+      "loss": 3.5022,
+      "step": 15999
+    },
+    {
+      "epoch": 0.6736842105263158,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00012475655074744855,
+      "loss": 3.3783,
+      "step": 16000
+    },
+    {
+      "epoch": 0.6737263157894737,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00012472734747941816,
+      "loss": 3.4482,
+      "step": 16001
+    },
+    {
+      "epoch": 0.6737684210526316,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0001246981464937083,
+      "loss": 2.9048,
+      "step": 16002
+    },
+    {
+      "epoch": 0.6738105263157895,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00012466894779085096,
+      "loss": 2.8888,
+      "step": 16003
+    },
+    {
+      "epoch": 0.6738526315789474,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00012463975137137819,
+      "loss": 2.5122,
+      "step": 16004
+    },
+    {
+      "epoch": 0.6738947368421052,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00012461055723582182,
+      "loss": 3.2735,
+      "step": 16005
+    },
+    {
+      "epoch": 0.6739368421052632,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001245813653847137,
+      "loss": 3.1916,
+      "step": 16006
+    },
+    {
+      "epoch": 0.673978947368421,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001245521758185858,
+      "loss": 3.0942,
+      "step": 16007
+    },
+    {
+      "epoch": 0.674021052631579,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0001245229885379699,
+      "loss": 3.4631,
+      "step": 16008
+    },
+    {
+      "epoch": 0.6740631578947368,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00012449380354339772,
+      "loss": 2.7936,
+      "step": 16009
+    },
+    {
+      "epoch": 0.6741052631578948,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00012446462083540093,
+      "loss": 2.7816,
+      "step": 16010
+    },
+    {
+      "epoch": 0.6741473684210526,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00012443544041451128,
+      "loss": 3.2789,
+      "step": 16011
+    },
+    {
+      "epoch": 0.6741894736842106,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00012440626228126038,
+      "loss": 3.3192,
+      "step": 16012
+    },
+    {
+      "epoch": 0.6742315789473684,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0001243770864361797,
+      "loss": 3.0779,
+      "step": 16013
+    },
+    {
+      "epoch": 0.6742736842105264,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00012434791287980102,
+      "loss": 3.1375,
+      "step": 16014
+    },
+    {
+      "epoch": 0.6743157894736842,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00012431874161265572,
+      "loss": 3.0256,
+      "step": 16015
+    },
+    {
+      "epoch": 0.6743578947368422,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00012428957263527527,
+      "loss": 3.4919,
+      "step": 16016
+    },
+    {
+      "epoch": 0.6744,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00012426040594819114,
+      "loss": 3.1525,
+      "step": 16017
+    },
+    {
+      "epoch": 0.6744421052631578,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00012423124155193463,
+      "loss": 2.9154,
+      "step": 16018
+    },
+    {
+      "epoch": 0.6744842105263158,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00012420207944703715,
+      "loss": 2.8667,
+      "step": 16019
+    },
+    {
+      "epoch": 0.6745263157894736,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00012417291963402987,
+      "loss": 3.2322,
+      "step": 16020
+    },
+    {
+      "epoch": 0.6745684210526316,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00012414376211344435,
+      "loss": 3.0335,
+      "step": 16021
+    },
+    {
+      "epoch": 0.6746105263157894,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00012411460688581142,
+      "loss": 2.7226,
+      "step": 16022
+    },
+    {
+      "epoch": 0.6746526315789474,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00012408545395166254,
+      "loss": 2.7735,
+      "step": 16023
+    },
+    {
+      "epoch": 0.6746947368421052,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00012405630331152875,
+      "loss": 3.1164,
+      "step": 16024
+    },
+    {
+      "epoch": 0.6747368421052632,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00012402715496594114,
+      "loss": 3.0255,
+      "step": 16025
+    },
+    {
+      "epoch": 0.674778947368421,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00012399800891543073,
+      "loss": 2.4151,
+      "step": 16026
+    },
+    {
+      "epoch": 0.674821052631579,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00012396886516052847,
+      "loss": 3.5808,
+      "step": 16027
+    },
+    {
+      "epoch": 0.6748631578947368,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00012393972370176558,
+      "loss": 2.8057,
+      "step": 16028
+    },
+    {
+      "epoch": 0.6749052631578948,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00012391058453967266,
+      "loss": 2.943,
+      "step": 16029
+    },
+    {
+      "epoch": 0.6749473684210526,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00012388144767478082,
+      "loss": 3.0527,
+      "step": 16030
+    },
+    {
+      "epoch": 0.6749894736842105,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00012385231310762078,
+      "loss": 3.149,
+      "step": 16031
+    },
+    {
+      "epoch": 0.6750315789473684,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00012382318083872343,
+      "loss": 3.2126,
+      "step": 16032
+    },
+    {
+      "epoch": 0.6750736842105263,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00012379405086861947,
+      "loss": 3.2796,
+      "step": 16033
+    },
+    {
+      "epoch": 0.6751157894736842,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0001237649231978395,
+      "loss": 3.1493,
+      "step": 16034
+    },
+    {
+      "epoch": 0.6751578947368421,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00012373579782691449,
+      "loss": 2.94,
+      "step": 16035
+    },
+    {
+      "epoch": 0.6752,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00012370667475637473,
+      "loss": 3.397,
+      "step": 16036
+    },
+    {
+      "epoch": 0.6752421052631579,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00012367755398675116,
+      "loss": 3.0456,
+      "step": 16037
+    },
+    {
+      "epoch": 0.6752842105263158,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00012364843551857392,
+      "loss": 3.1292,
+      "step": 16038
+    },
+    {
+      "epoch": 0.6753263157894737,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0001236193193523738,
+      "loss": 2.8186,
+      "step": 16039
+    },
+    {
+      "epoch": 0.6753684210526316,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00012359020548868125,
+      "loss": 2.7601,
+      "step": 16040
+    },
+    {
+      "epoch": 0.6754105263157895,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00012356109392802647,
+      "loss": 2.9628,
+      "step": 16041
+    },
+    {
+      "epoch": 0.6754526315789474,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00012353198467094024,
+      "loss": 2.6467,
+      "step": 16042
+    },
+    {
+      "epoch": 0.6754947368421053,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00012350287771795246,
+      "loss": 3.7813,
+      "step": 16043
+    },
+    {
+      "epoch": 0.6755368421052632,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00012347377306959377,
+      "loss": 3.2144,
+      "step": 16044
+    },
+    {
+      "epoch": 0.6755789473684211,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001234446707263941,
+      "loss": 3.1624,
+      "step": 16045
+    },
+    {
+      "epoch": 0.6756210526315789,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00012341557068888396,
+      "loss": 2.9355,
+      "step": 16046
+    },
+    {
+      "epoch": 0.6756631578947369,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0001233864729575934,
+      "loss": 3.156,
+      "step": 16047
+    },
+    {
+      "epoch": 0.6757052631578947,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00012335737753305247,
+      "loss": 3.7806,
+      "step": 16048
+    },
+    {
+      "epoch": 0.6757473684210527,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001233282844157914,
+      "loss": 3.3783,
+      "step": 16049
+    },
+    {
+      "epoch": 0.6757894736842105,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00012329919360634,
+      "loss": 2.9129,
+      "step": 16050
+    },
+    {
+      "epoch": 0.6758315789473684,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00012327010510522862,
+      "loss": 2.5533,
+      "step": 16051
+    },
+    {
+      "epoch": 0.6758736842105263,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00012324101891298685,
+      "loss": 2.9665,
+      "step": 16052
+    },
+    {
+      "epoch": 0.6759157894736842,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001232119350301449,
+      "loss": 2.595,
+      "step": 16053
+    },
+    {
+      "epoch": 0.6759578947368421,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00012318285345723249,
+      "loss": 2.9708,
+      "step": 16054
+    },
+    {
+      "epoch": 0.676,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0001231537741947795,
+      "loss": 3.4914,
+      "step": 16055
+    },
+    {
+      "epoch": 0.6760421052631579,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001231246972433157,
+      "loss": 3.0139,
+      "step": 16056
+    },
+    {
+      "epoch": 0.6760842105263158,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00012309562260337073,
+      "loss": 3.2564,
+      "step": 16057
+    },
+    {
+      "epoch": 0.6761263157894737,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00012306655027547465,
+      "loss": 3.0608,
+      "step": 16058
+    },
+    {
+      "epoch": 0.6761684210526315,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00012303748026015664,
+      "loss": 2.9952,
+      "step": 16059
+    },
+    {
+      "epoch": 0.6762105263157895,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00012300841255794668,
+      "loss": 3.4193,
+      "step": 16060
+    },
+    {
+      "epoch": 0.6762526315789473,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00012297934716937422,
+      "loss": 3.327,
+      "step": 16061
+    },
+    {
+      "epoch": 0.6762947368421053,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00012295028409496886,
+      "loss": 2.9947,
+      "step": 16062
+    },
+    {
+      "epoch": 0.6763368421052631,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00012292122333526002,
+      "loss": 3.1509,
+      "step": 16063
+    },
+    {
+      "epoch": 0.6763789473684211,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00012289216489077715,
+      "loss": 3.1413,
+      "step": 16064
+    },
+    {
+      "epoch": 0.6764210526315789,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00012286310876204973,
+      "loss": 3.0548,
+      "step": 16065
+    },
+    {
+      "epoch": 0.6764631578947369,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00012283405494960698,
+      "loss": 3.6163,
+      "step": 16066
+    },
+    {
+      "epoch": 0.6765052631578947,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0001228050034539784,
+      "loss": 2.5047,
+      "step": 16067
+    },
+    {
+      "epoch": 0.6765473684210527,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00012277595427569323,
+      "loss": 2.901,
+      "step": 16068
+    },
+    {
+      "epoch": 0.6765894736842105,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001227469074152807,
+      "loss": 3.1595,
+      "step": 16069
+    },
+    {
+      "epoch": 0.6766315789473685,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00012271786287327,
+      "loss": 3.2306,
+      "step": 16070
+    },
+    {
+      "epoch": 0.6766736842105263,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0001226888206501903,
+      "loss": 3.2328,
+      "step": 16071
+    },
+    {
+      "epoch": 0.6767157894736842,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0001226597807465707,
+      "loss": 3.0643,
+      "step": 16072
+    },
+    {
+      "epoch": 0.6767578947368421,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00012263074316294016,
+      "loss": 3.0075,
+      "step": 16073
+    },
+    {
+      "epoch": 0.6768,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00012260170789982794,
+      "loss": 2.8612,
+      "step": 16074
+    },
+    {
+      "epoch": 0.6768421052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0001225726749577629,
+      "loss": 3.2311,
+      "step": 16075
+    },
+    {
+      "epoch": 0.6768842105263158,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00012254364433727402,
+      "loss": 2.9804,
+      "step": 16076
+    },
+    {
+      "epoch": 0.6769263157894737,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00012251461603889017,
+      "loss": 3.4231,
+      "step": 16077
+    },
+    {
+      "epoch": 0.6769684210526316,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00012248559006314023,
+      "loss": 2.9016,
+      "step": 16078
+    },
+    {
+      "epoch": 0.6770105263157895,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00012245656641055302,
+      "loss": 3.2588,
+      "step": 16079
+    },
+    {
+      "epoch": 0.6770526315789474,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00012242754508165729,
+      "loss": 3.2213,
+      "step": 16080
+    },
+    {
+      "epoch": 0.6770947368421053,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00012239852607698174,
+      "loss": 3.1591,
+      "step": 16081
+    },
+    {
+      "epoch": 0.6771368421052631,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00012236950939705518,
+      "loss": 2.8133,
+      "step": 16082
+    },
+    {
+      "epoch": 0.6771789473684211,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0001223404950424062,
+      "loss": 3.2476,
+      "step": 16083
+    },
+    {
+      "epoch": 0.677221052631579,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00012231148301356345,
+      "loss": 3.0044,
+      "step": 16084
+    },
+    {
+      "epoch": 0.6772631578947368,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00012228247331105542,
+      "loss": 3.0845,
+      "step": 16085
+    },
+    {
+      "epoch": 0.6773052631578947,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00012225346593541057,
+      "loss": 3.3282,
+      "step": 16086
+    },
+    {
+      "epoch": 0.6773473684210526,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00012222446088715766,
+      "loss": 3.0728,
+      "step": 16087
+    },
+    {
+      "epoch": 0.6773894736842105,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00012219545816682476,
+      "loss": 2.9561,
+      "step": 16088
+    },
+    {
+      "epoch": 0.6774315789473684,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00012216645777494056,
+      "loss": 3.4179,
+      "step": 16089
+    },
+    {
+      "epoch": 0.6774736842105263,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00012213745971203328,
+      "loss": 2.75,
+      "step": 16090
+    },
+    {
+      "epoch": 0.6775157894736842,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00012210846397863126,
+      "loss": 3.3597,
+      "step": 16091
+    },
+    {
+      "epoch": 0.6775578947368421,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00012207947057526277,
+      "loss": 2.6293,
+      "step": 16092
+    },
+    {
+      "epoch": 0.6776,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001220504795024559,
+      "loss": 3.2396,
+      "step": 16093
+    },
+    {
+      "epoch": 0.6776421052631579,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00012202149076073915,
+      "loss": 2.7772,
+      "step": 16094
+    },
+    {
+      "epoch": 0.6776842105263158,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00012199250435064029,
+      "loss": 2.8994,
+      "step": 16095
+    },
+    {
+      "epoch": 0.6777263157894737,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00012196352027268782,
+      "loss": 3.221,
+      "step": 16096
+    },
+    {
+      "epoch": 0.6777684210526316,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00012193453852740935,
+      "loss": 3.1589,
+      "step": 16097
+    },
+    {
+      "epoch": 0.6778105263157894,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00012190555911533322,
+      "loss": 2.5522,
+      "step": 16098
+    },
+    {
+      "epoch": 0.6778526315789474,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00012187658203698729,
+      "loss": 3.1229,
+      "step": 16099
+    },
+    {
+      "epoch": 0.6778947368421052,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0001218476072928994,
+      "loss": 3.0717,
+      "step": 16100
+    },
+    {
+      "epoch": 0.6779368421052632,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00012181863488359774,
+      "loss": 3.14,
+      "step": 16101
+    },
+    {
+      "epoch": 0.677978947368421,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00012178966480960973,
+      "loss": 3.1837,
+      "step": 16102
+    },
+    {
+      "epoch": 0.678021052631579,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00012176069707146362,
+      "loss": 3.235,
+      "step": 16103
+    },
+    {
+      "epoch": 0.6780631578947368,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0001217317316696867,
+      "loss": 3.1908,
+      "step": 16104
+    },
+    {
+      "epoch": 0.6781052631578948,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00012170276860480703,
+      "loss": 3.2871,
+      "step": 16105
+    },
+    {
+      "epoch": 0.6781473684210526,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00012167380787735219,
+      "loss": 3.1537,
+      "step": 16106
+    },
+    {
+      "epoch": 0.6781894736842106,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00012164484948784976,
+      "loss": 3.4178,
+      "step": 16107
+    },
+    {
+      "epoch": 0.6782315789473684,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00012161589343682739,
+      "loss": 3.1416,
+      "step": 16108
+    },
+    {
+      "epoch": 0.6782736842105264,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00012158693972481249,
+      "loss": 2.8099,
+      "step": 16109
+    },
+    {
+      "epoch": 0.6783157894736842,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00012155798835233287,
+      "loss": 3.1204,
+      "step": 16110
+    },
+    {
+      "epoch": 0.6783578947368422,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00012152903931991558,
+      "loss": 3.3064,
+      "step": 16111
+    },
+    {
+      "epoch": 0.6784,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00012150009262808845,
+      "loss": 3.0845,
+      "step": 16112
+    },
+    {
+      "epoch": 0.6784421052631578,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00012147114827737846,
+      "loss": 3.2118,
+      "step": 16113
+    },
+    {
+      "epoch": 0.6784842105263158,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00012144220626831323,
+      "loss": 3.3953,
+      "step": 16114
+    },
+    {
+      "epoch": 0.6785263157894736,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00012141326660141993,
+      "loss": 2.2515,
+      "step": 16115
+    },
+    {
+      "epoch": 0.6785684210526316,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00012138432927722575,
+      "loss": 3.3975,
+      "step": 16116
+    },
+    {
+      "epoch": 0.6786105263157894,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0001213553942962582,
+      "loss": 2.9525,
+      "step": 16117
+    },
+    {
+      "epoch": 0.6786526315789474,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00012132646165904396,
+      "loss": 3.1446,
+      "step": 16118
+    },
+    {
+      "epoch": 0.6786947368421052,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00012129753136611063,
+      "loss": 2.9322,
+      "step": 16119
+    },
+    {
+      "epoch": 0.6787368421052632,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00012126860341798487,
+      "loss": 3.4651,
+      "step": 16120
+    },
+    {
+      "epoch": 0.678778947368421,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00012123967781519398,
+      "loss": 3.1704,
+      "step": 16121
+    },
+    {
+      "epoch": 0.678821052631579,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00012121075455826488,
+      "loss": 3.4154,
+      "step": 16122
+    },
+    {
+      "epoch": 0.6788631578947368,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00012118183364772453,
+      "loss": 2.9112,
+      "step": 16123
+    },
+    {
+      "epoch": 0.6789052631578948,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00012115291508409978,
+      "loss": 3.4438,
+      "step": 16124
+    },
+    {
+      "epoch": 0.6789473684210526,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00012112399886791747,
+      "loss": 3.5528,
+      "step": 16125
+    },
+    {
+      "epoch": 0.6789894736842105,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00012109508499970464,
+      "loss": 3.0166,
+      "step": 16126
+    },
+    {
+      "epoch": 0.6790315789473684,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00012106617347998775,
+      "loss": 2.7317,
+      "step": 16127
+    },
+    {
+      "epoch": 0.6790736842105263,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00012103726430929374,
+      "loss": 3.0632,
+      "step": 16128
+    },
+    {
+      "epoch": 0.6791157894736842,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00012100835748814929,
+      "loss": 2.9947,
+      "step": 16129
+    },
+    {
+      "epoch": 0.6791578947368421,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00012097945301708097,
+      "loss": 3.4151,
+      "step": 16130
+    },
+    {
+      "epoch": 0.6792,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00012095055089661545,
+      "loss": 2.9139,
+      "step": 16131
+    },
+    {
+      "epoch": 0.6792421052631579,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00012092165112727916,
+      "loss": 2.727,
+      "step": 16132
+    },
+    {
+      "epoch": 0.6792842105263158,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00012089275370959882,
+      "loss": 3.3195,
+      "step": 16133
+    },
+    {
+      "epoch": 0.6793263157894737,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00012086385864410079,
+      "loss": 3.408,
+      "step": 16134
+    },
+    {
+      "epoch": 0.6793684210526316,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00012083496593131154,
+      "loss": 3.1718,
+      "step": 16135
+    },
+    {
+      "epoch": 0.6794105263157895,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00012080607557175749,
+      "loss": 3.0446,
+      "step": 16136
+    },
+    {
+      "epoch": 0.6794526315789474,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0001207771875659649,
+      "loss": 3.4796,
+      "step": 16137
+    },
+    {
+      "epoch": 0.6794947368421053,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00012074830191446012,
+      "loss": 3.3173,
+      "step": 16138
+    },
+    {
+      "epoch": 0.6795368421052631,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00012071941861776944,
+      "loss": 2.4586,
+      "step": 16139
+    },
+    {
+      "epoch": 0.6795789473684211,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00012069053767641897,
+      "loss": 3.4476,
+      "step": 16140
+    },
+    {
+      "epoch": 0.6796210526315789,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00012066165909093504,
+      "loss": 3.1829,
+      "step": 16141
+    },
+    {
+      "epoch": 0.6796631578947369,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001206327828618437,
+      "loss": 2.7153,
+      "step": 16142
+    },
+    {
+      "epoch": 0.6797052631578947,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00012060390898967107,
+      "loss": 3.0918,
+      "step": 16143
+    },
+    {
+      "epoch": 0.6797473684210527,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00012057503747494322,
+      "loss": 2.9281,
+      "step": 16144
+    },
+    {
+      "epoch": 0.6797894736842105,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00012054616831818607,
+      "loss": 3.0911,
+      "step": 16145
+    },
+    {
+      "epoch": 0.6798315789473685,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00012051730151992566,
+      "loss": 3.4514,
+      "step": 16146
+    },
+    {
+      "epoch": 0.6798736842105263,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001204884370806878,
+      "loss": 3.1015,
+      "step": 16147
+    },
+    {
+      "epoch": 0.6799157894736843,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0001204595750009986,
+      "loss": 2.8403,
+      "step": 16148
+    },
+    {
+      "epoch": 0.6799578947368421,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00012043071528138355,
+      "loss": 3.4872,
+      "step": 16149
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00012040185792236874,
+      "loss": 2.7525,
+      "step": 16150
+    },
+    {
+      "epoch": 0.6800421052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00012037300292447981,
+      "loss": 3.2922,
+      "step": 16151
+    },
+    {
+      "epoch": 0.6800842105263157,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00012034415028824247,
+      "loss": 3.3392,
+      "step": 16152
+    },
+    {
+      "epoch": 0.6801263157894737,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00012031530001418234,
+      "loss": 3.3113,
+      "step": 16153
+    },
+    {
+      "epoch": 0.6801684210526315,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00012028645210282502,
+      "loss": 3.2078,
+      "step": 16154
+    },
+    {
+      "epoch": 0.6802105263157895,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00012025760655469628,
+      "loss": 2.7872,
+      "step": 16155
+    },
+    {
+      "epoch": 0.6802526315789473,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00012022876337032134,
+      "loss": 2.7714,
+      "step": 16156
+    },
+    {
+      "epoch": 0.6802947368421053,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00012019992255022597,
+      "loss": 3.0251,
+      "step": 16157
+    },
+    {
+      "epoch": 0.6803368421052631,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00012017108409493547,
+      "loss": 3.585,
+      "step": 16158
+    },
+    {
+      "epoch": 0.6803789473684211,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00012014224800497528,
+      "loss": 3.2309,
+      "step": 16159
+    },
+    {
+      "epoch": 0.6804210526315789,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00012011341428087078,
+      "loss": 2.8453,
+      "step": 16160
+    },
+    {
+      "epoch": 0.6804631578947369,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00012008458292314714,
+      "loss": 2.9876,
+      "step": 16161
+    },
+    {
+      "epoch": 0.6805052631578947,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00012005575393232995,
+      "loss": 2.9682,
+      "step": 16162
+    },
+    {
+      "epoch": 0.6805473684210527,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00012002692730894405,
+      "loss": 3.472,
+      "step": 16163
+    },
+    {
+      "epoch": 0.6805894736842105,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00011999810305351502,
+      "loss": 3.3375,
+      "step": 16164
+    },
+    {
+      "epoch": 0.6806315789473685,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00011996928116656761,
+      "loss": 3.1445,
+      "step": 16165
+    },
+    {
+      "epoch": 0.6806736842105263,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00011994046164862724,
+      "loss": 3.2009,
+      "step": 16166
+    },
+    {
+      "epoch": 0.6807157894736842,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001199116445002188,
+      "loss": 3.0228,
+      "step": 16167
+    },
+    {
+      "epoch": 0.6807578947368421,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00011988282972186729,
+      "loss": 3.1926,
+      "step": 16168
+    },
+    {
+      "epoch": 0.6808,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00011985401731409792,
+      "loss": 3.2791,
+      "step": 16169
+    },
+    {
+      "epoch": 0.6808421052631579,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00011982520727743523,
+      "loss": 3.0767,
+      "step": 16170
+    },
+    {
+      "epoch": 0.6808842105263158,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00011979639961240454,
+      "loss": 3.163,
+      "step": 16171
+    },
+    {
+      "epoch": 0.6809263157894737,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00011976759431953027,
+      "loss": 2.5058,
+      "step": 16172
+    },
+    {
+      "epoch": 0.6809684210526316,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001197387913993375,
+      "loss": 2.7328,
+      "step": 16173
+    },
+    {
+      "epoch": 0.6810105263157895,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00011970999085235088,
+      "loss": 2.9342,
+      "step": 16174
+    },
+    {
+      "epoch": 0.6810526315789474,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00011968119267909517,
+      "loss": 2.9526,
+      "step": 16175
+    },
+    {
+      "epoch": 0.6810947368421053,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00011965239688009499,
+      "loss": 3.4862,
+      "step": 16176
+    },
+    {
+      "epoch": 0.6811368421052632,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00011962360345587491,
+      "loss": 2.9894,
+      "step": 16177
+    },
+    {
+      "epoch": 0.6811789473684211,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00011959481240695977,
+      "loss": 3.0826,
+      "step": 16178
+    },
+    {
+      "epoch": 0.681221052631579,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00011956602373387377,
+      "loss": 2.5881,
+      "step": 16179
+    },
+    {
+      "epoch": 0.6812631578947368,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0001195372374371417,
+      "loss": 3.0944,
+      "step": 16180
+    },
+    {
+      "epoch": 0.6813052631578947,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00011950845351728772,
+      "loss": 3.0159,
+      "step": 16181
+    },
+    {
+      "epoch": 0.6813473684210526,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00011947967197483648,
+      "loss": 2.3523,
+      "step": 16182
+    },
+    {
+      "epoch": 0.6813894736842105,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00011945089281031227,
+      "loss": 3.0046,
+      "step": 16183
+    },
+    {
+      "epoch": 0.6814315789473684,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0001194221160242393,
+      "loss": 3.0739,
+      "step": 16184
+    },
+    {
+      "epoch": 0.6814736842105263,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00011939334161714215,
+      "loss": 3.2549,
+      "step": 16185
+    },
+    {
+      "epoch": 0.6815157894736842,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00011936456958954467,
+      "loss": 3.1204,
+      "step": 16186
+    },
+    {
+      "epoch": 0.6815578947368421,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00011933579994197133,
+      "loss": 3.1981,
+      "step": 16187
+    },
+    {
+      "epoch": 0.6816,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00011930703267494619,
+      "loss": 3.4245,
+      "step": 16188
+    },
+    {
+      "epoch": 0.6816421052631579,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00011927826778899336,
+      "loss": 3.1462,
+      "step": 16189
+    },
+    {
+      "epoch": 0.6816842105263158,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00011924950528463691,
+      "loss": 2.4574,
+      "step": 16190
+    },
+    {
+      "epoch": 0.6817263157894737,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00011922074516240083,
+      "loss": 2.5084,
+      "step": 16191
+    },
+    {
+      "epoch": 0.6817684210526316,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00011919198742280912,
+      "loss": 2.8044,
+      "step": 16192
+    },
+    {
+      "epoch": 0.6818105263157894,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0001191632320663856,
+      "loss": 3.0638,
+      "step": 16193
+    },
+    {
+      "epoch": 0.6818526315789474,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00011913447909365438,
+      "loss": 3.1425,
+      "step": 16194
+    },
+    {
+      "epoch": 0.6818947368421052,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00011910572850513918,
+      "loss": 2.9236,
+      "step": 16195
+    },
+    {
+      "epoch": 0.6819368421052632,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00011907698030136383,
+      "loss": 2.8871,
+      "step": 16196
+    },
+    {
+      "epoch": 0.681978947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00011904823448285204,
+      "loss": 2.871,
+      "step": 16197
+    },
+    {
+      "epoch": 0.682021052631579,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00011901949105012758,
+      "loss": 2.983,
+      "step": 16198
+    },
+    {
+      "epoch": 0.6820631578947368,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00011899075000371409,
+      "loss": 3.4097,
+      "step": 16199
+    },
+    {
+      "epoch": 0.6821052631578948,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00011896201134413512,
+      "loss": 3.0942,
+      "step": 16200
+    },
+    {
+      "epoch": 0.6821473684210526,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0001189332750719144,
+      "loss": 2.9146,
+      "step": 16201
+    },
+    {
+      "epoch": 0.6821894736842106,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00011890454118757544,
+      "loss": 3.0275,
+      "step": 16202
+    },
+    {
+      "epoch": 0.6822315789473684,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00011887580969164171,
+      "loss": 3.6323,
+      "step": 16203
+    },
+    {
+      "epoch": 0.6822736842105264,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00011884708058463667,
+      "loss": 3.4811,
+      "step": 16204
+    },
+    {
+      "epoch": 0.6823157894736842,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0001188183538670837,
+      "loss": 2.984,
+      "step": 16205
+    },
+    {
+      "epoch": 0.682357894736842,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0001187896295395062,
+      "loss": 3.2407,
+      "step": 16206
+    },
+    {
+      "epoch": 0.6824,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00011876090760242747,
+      "loss": 2.773,
+      "step": 16207
+    },
+    {
+      "epoch": 0.6824421052631578,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00011873218805637073,
+      "loss": 3.0291,
+      "step": 16208
+    },
+    {
+      "epoch": 0.6824842105263158,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00011870347090185937,
+      "loss": 3.1989,
+      "step": 16209
+    },
+    {
+      "epoch": 0.6825263157894736,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00011867475613941652,
+      "loss": 3.469,
+      "step": 16210
+    },
+    {
+      "epoch": 0.6825684210526316,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00011864604376956531,
+      "loss": 3.3692,
+      "step": 16211
+    },
+    {
+      "epoch": 0.6826105263157894,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00011861733379282885,
+      "loss": 3.0437,
+      "step": 16212
+    },
+    {
+      "epoch": 0.6826526315789474,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00011858862620973019,
+      "loss": 3.2612,
+      "step": 16213
+    },
+    {
+      "epoch": 0.6826947368421052,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00011855992102079235,
+      "loss": 3.2095,
+      "step": 16214
+    },
+    {
+      "epoch": 0.6827368421052632,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00011853121822653823,
+      "loss": 3.2219,
+      "step": 16215
+    },
+    {
+      "epoch": 0.682778947368421,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00011850251782749094,
+      "loss": 2.9198,
+      "step": 16216
+    },
+    {
+      "epoch": 0.682821052631579,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00011847381982417327,
+      "loss": 3.3388,
+      "step": 16217
+    },
+    {
+      "epoch": 0.6828631578947368,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00011844512421710807,
+      "loss": 2.7277,
+      "step": 16218
+    },
+    {
+      "epoch": 0.6829052631578947,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00011841643100681813,
+      "loss": 3.1026,
+      "step": 16219
+    },
+    {
+      "epoch": 0.6829473684210526,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00011838774019382623,
+      "loss": 2.8913,
+      "step": 16220
+    },
+    {
+      "epoch": 0.6829894736842105,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00011835905177865505,
+      "loss": 3.2164,
+      "step": 16221
+    },
+    {
+      "epoch": 0.6830315789473684,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00011833036576182718,
+      "loss": 2.4817,
+      "step": 16222
+    },
+    {
+      "epoch": 0.6830736842105263,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00011830168214386553,
+      "loss": 2.9741,
+      "step": 16223
+    },
+    {
+      "epoch": 0.6831157894736842,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00011827300092529231,
+      "loss": 3.0859,
+      "step": 16224
+    },
+    {
+      "epoch": 0.6831578947368421,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0001182443221066303,
+      "loss": 3.2365,
+      "step": 16225
+    },
+    {
+      "epoch": 0.6832,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00011821564568840199,
+      "loss": 3.3139,
+      "step": 16226
+    },
+    {
+      "epoch": 0.6832421052631579,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00011818697167112974,
+      "loss": 2.9414,
+      "step": 16227
+    },
+    {
+      "epoch": 0.6832842105263158,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00011815830005533598,
+      "loss": 3.3312,
+      "step": 16228
+    },
+    {
+      "epoch": 0.6833263157894737,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00011812963084154302,
+      "loss": 2.5778,
+      "step": 16229
+    },
+    {
+      "epoch": 0.6833684210526316,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00011810096403027343,
+      "loss": 2.9523,
+      "step": 16230
+    },
+    {
+      "epoch": 0.6834105263157895,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0001180722996220491,
+      "loss": 3.2114,
+      "step": 16231
+    },
+    {
+      "epoch": 0.6834526315789474,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00011804363761739257,
+      "loss": 2.6708,
+      "step": 16232
+    },
+    {
+      "epoch": 0.6834947368421053,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00011801497801682593,
+      "loss": 3.287,
+      "step": 16233
+    },
+    {
+      "epoch": 0.6835368421052631,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001179863208208713,
+      "loss": 2.8955,
+      "step": 16234
+    },
+    {
+      "epoch": 0.6835789473684211,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00011795766603005078,
+      "loss": 3.1961,
+      "step": 16235
+    },
+    {
+      "epoch": 0.6836210526315789,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00011792901364488637,
+      "loss": 3.46,
+      "step": 16236
+    },
+    {
+      "epoch": 0.6836631578947369,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00011790036366590032,
+      "loss": 3.2679,
+      "step": 16237
+    },
+    {
+      "epoch": 0.6837052631578947,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00011787171609361427,
+      "loss": 3.0178,
+      "step": 16238
+    },
+    {
+      "epoch": 0.6837473684210527,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00011784307092855048,
+      "loss": 2.7105,
+      "step": 16239
+    },
+    {
+      "epoch": 0.6837894736842105,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00011781442817123048,
+      "loss": 3.0033,
+      "step": 16240
+    },
+    {
+      "epoch": 0.6838315789473685,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00011778578782217636,
+      "loss": 2.8953,
+      "step": 16241
+    },
+    {
+      "epoch": 0.6838736842105263,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00011775714988190985,
+      "loss": 3.3307,
+      "step": 16242
+    },
+    {
+      "epoch": 0.6839157894736843,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00011772851435095261,
+      "loss": 2.8446,
+      "step": 16243
+    },
+    {
+      "epoch": 0.6839578947368421,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00011769988122982662,
+      "loss": 3.152,
+      "step": 16244
+    },
+    {
+      "epoch": 0.684,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00011767125051905314,
+      "loss": 2.8681,
+      "step": 16245
+    },
+    {
+      "epoch": 0.6840421052631579,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0001176426222191542,
+      "loss": 2.762,
+      "step": 16246
+    },
+    {
+      "epoch": 0.6840842105263157,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00011761399633065098,
+      "loss": 3.3489,
+      "step": 16247
+    },
+    {
+      "epoch": 0.6841263157894737,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00011758537285406528,
+      "loss": 2.9021,
+      "step": 16248
+    },
+    {
+      "epoch": 0.6841684210526315,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00011755675178991851,
+      "loss": 3.4785,
+      "step": 16249
+    },
+    {
+      "epoch": 0.6842105263157895,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00011752813313873211,
+      "loss": 2.7558,
+      "step": 16250
+    },
+    {
+      "epoch": 0.6842526315789473,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00011749951690102748,
+      "loss": 3.3634,
+      "step": 16251
+    },
+    {
+      "epoch": 0.6842947368421053,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00011747090307732589,
+      "loss": 3.4248,
+      "step": 16252
+    },
+    {
+      "epoch": 0.6843368421052631,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00011744229166814888,
+      "loss": 3.1195,
+      "step": 16253
+    },
+    {
+      "epoch": 0.6843789473684211,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00011741368267401742,
+      "loss": 3.1863,
+      "step": 16254
+    },
+    {
+      "epoch": 0.6844210526315789,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00011738507609545293,
+      "loss": 3.2127,
+      "step": 16255
+    },
+    {
+      "epoch": 0.6844631578947369,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00011735647193297658,
+      "loss": 3.1887,
+      "step": 16256
+    },
+    {
+      "epoch": 0.6845052631578947,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00011732787018710945,
+      "loss": 2.7824,
+      "step": 16257
+    },
+    {
+      "epoch": 0.6845473684210527,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00011729927085837264,
+      "loss": 3.0537,
+      "step": 16258
+    },
+    {
+      "epoch": 0.6845894736842105,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00011727067394728708,
+      "loss": 3.2699,
+      "step": 16259
+    },
+    {
+      "epoch": 0.6846315789473684,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0001172420794543741,
+      "loss": 2.9023,
+      "step": 16260
+    },
+    {
+      "epoch": 0.6846736842105263,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00011721348738015425,
+      "loss": 2.8951,
+      "step": 16261
+    },
+    {
+      "epoch": 0.6847157894736842,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00011718489772514876,
+      "loss": 2.9293,
+      "step": 16262
+    },
+    {
+      "epoch": 0.6847578947368421,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00011715631048987834,
+      "loss": 3.0959,
+      "step": 16263
+    },
+    {
+      "epoch": 0.6848,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001171277256748639,
+      "loss": 3.0854,
+      "step": 16264
+    },
+    {
+      "epoch": 0.6848421052631579,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00011709914328062616,
+      "loss": 3.3589,
+      "step": 16265
+    },
+    {
+      "epoch": 0.6848842105263158,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001170705633076859,
+      "loss": 3.5302,
+      "step": 16266
+    },
+    {
+      "epoch": 0.6849263157894737,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00011704198575656374,
+      "loss": 3.1648,
+      "step": 16267
+    },
+    {
+      "epoch": 0.6849684210526316,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00011701341062778034,
+      "loss": 3.3053,
+      "step": 16268
+    },
+    {
+      "epoch": 0.6850105263157895,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001169848379218564,
+      "loss": 2.9369,
+      "step": 16269
+    },
+    {
+      "epoch": 0.6850526315789474,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00011695626763931247,
+      "loss": 3.2693,
+      "step": 16270
+    },
+    {
+      "epoch": 0.6850947368421053,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00011692769978066897,
+      "loss": 3.2811,
+      "step": 16271
+    },
+    {
+      "epoch": 0.6851368421052632,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00011689913434644648,
+      "loss": 2.9411,
+      "step": 16272
+    },
+    {
+      "epoch": 0.685178947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00011687057133716533,
+      "loss": 2.8029,
+      "step": 16273
+    },
+    {
+      "epoch": 0.685221052631579,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00011684201075334597,
+      "loss": 2.9066,
+      "step": 16274
+    },
+    {
+      "epoch": 0.6852631578947368,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00011681345259550863,
+      "loss": 3.1444,
+      "step": 16275
+    },
+    {
+      "epoch": 0.6853052631578948,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00011678489686417379,
+      "loss": 3.1446,
+      "step": 16276
+    },
+    {
+      "epoch": 0.6853473684210526,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0001167563435598616,
+      "loss": 3.2857,
+      "step": 16277
+    },
+    {
+      "epoch": 0.6853894736842105,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00011672779268309227,
+      "loss": 3.0498,
+      "step": 16278
+    },
+    {
+      "epoch": 0.6854315789473684,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00011669924423438599,
+      "loss": 2.7376,
+      "step": 16279
+    },
+    {
+      "epoch": 0.6854736842105263,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00011667069821426285,
+      "loss": 3.0223,
+      "step": 16280
+    },
+    {
+      "epoch": 0.6855157894736842,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00011664215462324293,
+      "loss": 3.3344,
+      "step": 16281
+    },
+    {
+      "epoch": 0.6855578947368421,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00011661361346184627,
+      "loss": 2.9819,
+      "step": 16282
+    },
+    {
+      "epoch": 0.6856,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00011658507473059274,
+      "loss": 3.4173,
+      "step": 16283
+    },
+    {
+      "epoch": 0.6856421052631579,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0001165565384300025,
+      "loss": 2.9902,
+      "step": 16284
+    },
+    {
+      "epoch": 0.6856842105263158,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00011652800456059532,
+      "loss": 3.3064,
+      "step": 16285
+    },
+    {
+      "epoch": 0.6857263157894737,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00011649947312289108,
+      "loss": 3.1919,
+      "step": 16286
+    },
+    {
+      "epoch": 0.6857684210526316,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0001164709441174096,
+      "loss": 2.81,
+      "step": 16287
+    },
+    {
+      "epoch": 0.6858105263157894,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00011644241754467053,
+      "loss": 2.9744,
+      "step": 16288
+    },
+    {
+      "epoch": 0.6858526315789474,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00011641389340519387,
+      "loss": 3.1061,
+      "step": 16289
+    },
+    {
+      "epoch": 0.6858947368421052,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00011638537169949892,
+      "loss": 2.8245,
+      "step": 16290
+    },
+    {
+      "epoch": 0.6859368421052632,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00011635685242810562,
+      "loss": 3.1369,
+      "step": 16291
+    },
+    {
+      "epoch": 0.685978947368421,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00011632833559153345,
+      "loss": 3.5735,
+      "step": 16292
+    },
+    {
+      "epoch": 0.686021052631579,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00011629982119030191,
+      "loss": 3.2884,
+      "step": 16293
+    },
+    {
+      "epoch": 0.6860631578947368,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00011627130922493057,
+      "loss": 3.2307,
+      "step": 16294
+    },
+    {
+      "epoch": 0.6861052631578948,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00011624279969593873,
+      "loss": 3.2151,
+      "step": 16295
+    },
+    {
+      "epoch": 0.6861473684210526,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00011621429260384614,
+      "loss": 3.3047,
+      "step": 16296
+    },
+    {
+      "epoch": 0.6861894736842106,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00011618578794917176,
+      "loss": 2.8561,
+      "step": 16297
+    },
+    {
+      "epoch": 0.6862315789473684,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00011615728573243525,
+      "loss": 3.1308,
+      "step": 16298
+    },
+    {
+      "epoch": 0.6862736842105264,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00011612878595415557,
+      "loss": 3.2669,
+      "step": 16299
+    },
+    {
+      "epoch": 0.6863157894736842,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00011610028861485222,
+      "loss": 2.9908,
+      "step": 16300
+    },
+    {
+      "epoch": 0.6863578947368421,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00011607179371504429,
+      "loss": 2.6822,
+      "step": 16301
+    },
+    {
+      "epoch": 0.6864,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00011604330125525078,
+      "loss": 2.8336,
+      "step": 16302
+    },
+    {
+      "epoch": 0.6864421052631579,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00011601481123599117,
+      "loss": 3.1341,
+      "step": 16303
+    },
+    {
+      "epoch": 0.6864842105263158,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00011598632365778406,
+      "loss": 3.1834,
+      "step": 16304
+    },
+    {
+      "epoch": 0.6865263157894737,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00011595783852114885,
+      "loss": 2.9666,
+      "step": 16305
+    },
+    {
+      "epoch": 0.6865684210526316,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00011592935582660416,
+      "loss": 2.9744,
+      "step": 16306
+    },
+    {
+      "epoch": 0.6866105263157894,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00011590087557466916,
+      "loss": 3.0851,
+      "step": 16307
+    },
+    {
+      "epoch": 0.6866526315789474,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00011587239776586266,
+      "loss": 2.933,
+      "step": 16308
+    },
+    {
+      "epoch": 0.6866947368421052,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00011584392240070349,
+      "loss": 3.1021,
+      "step": 16309
+    },
+    {
+      "epoch": 0.6867368421052632,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0001158154494797104,
+      "loss": 3.2615,
+      "step": 16310
+    },
+    {
+      "epoch": 0.686778947368421,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00011578697900340207,
+      "loss": 3.1459,
+      "step": 16311
+    },
+    {
+      "epoch": 0.686821052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0001157585109722975,
+      "loss": 2.5403,
+      "step": 16312
+    },
+    {
+      "epoch": 0.6868631578947368,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00011573004538691492,
+      "loss": 2.9488,
+      "step": 16313
+    },
+    {
+      "epoch": 0.6869052631578947,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00011570158224777336,
+      "loss": 2.6024,
+      "step": 16314
+    },
+    {
+      "epoch": 0.6869473684210526,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00011567312155539097,
+      "loss": 3.0249,
+      "step": 16315
+    },
+    {
+      "epoch": 0.6869894736842105,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00011564466331028659,
+      "loss": 2.8846,
+      "step": 16316
+    },
+    {
+      "epoch": 0.6870315789473684,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00011561620751297857,
+      "loss": 2.745,
+      "step": 16317
+    },
+    {
+      "epoch": 0.6870736842105263,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00011558775416398529,
+      "loss": 3.01,
+      "step": 16318
+    },
+    {
+      "epoch": 0.6871157894736842,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00011555930326382536,
+      "loss": 3.5908,
+      "step": 16319
+    },
+    {
+      "epoch": 0.6871578947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00011553085481301678,
+      "loss": 3.2262,
+      "step": 16320
+    },
+    {
+      "epoch": 0.6872,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00011550240881207822,
+      "loss": 2.8055,
+      "step": 16321
+    },
+    {
+      "epoch": 0.6872421052631579,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00011547396526152754,
+      "loss": 2.879,
+      "step": 16322
+    },
+    {
+      "epoch": 0.6872842105263158,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00011544552416188325,
+      "loss": 3.464,
+      "step": 16323
+    },
+    {
+      "epoch": 0.6873263157894737,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00011541708551366345,
+      "loss": 3.1478,
+      "step": 16324
+    },
+    {
+      "epoch": 0.6873684210526316,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00011538864931738618,
+      "loss": 3.1797,
+      "step": 16325
+    },
+    {
+      "epoch": 0.6874105263157895,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00011536021557356957,
+      "loss": 3.2848,
+      "step": 16326
+    },
+    {
+      "epoch": 0.6874526315789473,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00011533178428273156,
+      "loss": 2.6118,
+      "step": 16327
+    },
+    {
+      "epoch": 0.6874947368421053,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0001153033554453904,
+      "loss": 3.0494,
+      "step": 16328
+    },
+    {
+      "epoch": 0.6875368421052631,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00011527492906206363,
+      "loss": 3.0882,
+      "step": 16329
+    },
+    {
+      "epoch": 0.6875789473684211,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00011524650513326946,
+      "loss": 2.642,
+      "step": 16330
+    },
+    {
+      "epoch": 0.6876210526315789,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00011521808365952566,
+      "loss": 2.6467,
+      "step": 16331
+    },
+    {
+      "epoch": 0.6876631578947369,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00011518966464134997,
+      "loss": 3.1021,
+      "step": 16332
+    },
+    {
+      "epoch": 0.6877052631578947,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00011516124807926024,
+      "loss": 3.1584,
+      "step": 16333
+    },
+    {
+      "epoch": 0.6877473684210527,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00011513283397377403,
+      "loss": 2.9561,
+      "step": 16334
+    },
+    {
+      "epoch": 0.6877894736842105,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00011510442232540918,
+      "loss": 2.9454,
+      "step": 16335
+    },
+    {
+      "epoch": 0.6878315789473685,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0001150760131346833,
+      "loss": 2.8391,
+      "step": 16336
+    },
+    {
+      "epoch": 0.6878736842105263,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00011504760640211392,
+      "loss": 2.9557,
+      "step": 16337
+    },
+    {
+      "epoch": 0.6879157894736843,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00011501920212821859,
+      "loss": 3.1737,
+      "step": 16338
+    },
+    {
+      "epoch": 0.6879578947368421,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00011499080031351484,
+      "loss": 2.9135,
+      "step": 16339
+    },
+    {
+      "epoch": 0.688,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00011496240095852001,
+      "loss": 3.2937,
+      "step": 16340
+    },
+    {
+      "epoch": 0.6880421052631579,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00011493400406375162,
+      "loss": 3.0662,
+      "step": 16341
+    },
+    {
+      "epoch": 0.6880842105263157,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00011490560962972688,
+      "loss": 3.2675,
+      "step": 16342
+    },
+    {
+      "epoch": 0.6881263157894737,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00011487721765696327,
+      "loss": 3.1199,
+      "step": 16343
+    },
+    {
+      "epoch": 0.6881684210526315,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00011484882814597802,
+      "loss": 2.7314,
+      "step": 16344
+    },
+    {
+      "epoch": 0.6882105263157895,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00011482044109728831,
+      "loss": 2.9474,
+      "step": 16345
+    },
+    {
+      "epoch": 0.6882526315789473,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00011479205651141134,
+      "loss": 3.3588,
+      "step": 16346
+    },
+    {
+      "epoch": 0.6882947368421053,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00011476367438886424,
+      "loss": 3.021,
+      "step": 16347
+    },
+    {
+      "epoch": 0.6883368421052631,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00011473529473016408,
+      "loss": 3.2984,
+      "step": 16348
+    },
+    {
+      "epoch": 0.6883789473684211,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00011470691753582784,
+      "loss": 3.074,
+      "step": 16349
+    },
+    {
+      "epoch": 0.6884210526315789,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00011467854280637269,
+      "loss": 3.3534,
+      "step": 16350
+    },
+    {
+      "epoch": 0.6884631578947369,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0001146501705423155,
+      "loss": 2.8463,
+      "step": 16351
+    },
+    {
+      "epoch": 0.6885052631578947,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00011462180074417317,
+      "loss": 2.8634,
+      "step": 16352
+    },
+    {
+      "epoch": 0.6885473684210527,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00011459343341246255,
+      "loss": 3.2214,
+      "step": 16353
+    },
+    {
+      "epoch": 0.6885894736842105,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00011456506854770051,
+      "loss": 3.0087,
+      "step": 16354
+    },
+    {
+      "epoch": 0.6886315789473684,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00011453670615040376,
+      "loss": 2.8773,
+      "step": 16355
+    },
+    {
+      "epoch": 0.6886736842105263,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.000114508346221089,
+      "loss": 3.0765,
+      "step": 16356
+    },
+    {
+      "epoch": 0.6887157894736842,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00011447998876027316,
+      "loss": 3.6035,
+      "step": 16357
+    },
+    {
+      "epoch": 0.6887578947368421,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00011445163376847248,
+      "loss": 3.2539,
+      "step": 16358
+    },
+    {
+      "epoch": 0.6888,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00011442328124620388,
+      "loss": 3.0006,
+      "step": 16359
+    },
+    {
+      "epoch": 0.6888421052631579,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00011439493119398381,
+      "loss": 2.8496,
+      "step": 16360
+    },
+    {
+      "epoch": 0.6888842105263158,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00011436658361232873,
+      "loss": 3.1002,
+      "step": 16361
+    },
+    {
+      "epoch": 0.6889263157894737,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00011433823850175514,
+      "loss": 3.2487,
+      "step": 16362
+    },
+    {
+      "epoch": 0.6889684210526316,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0001143098958627794,
+      "loss": 2.8381,
+      "step": 16363
+    },
+    {
+      "epoch": 0.6890105263157895,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00011428155569591809,
+      "loss": 2.7803,
+      "step": 16364
+    },
+    {
+      "epoch": 0.6890526315789474,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00011425321800168717,
+      "loss": 3.1177,
+      "step": 16365
+    },
+    {
+      "epoch": 0.6890947368421053,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00011422488278060322,
+      "loss": 2.5929,
+      "step": 16366
+    },
+    {
+      "epoch": 0.6891368421052632,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00011419655003318239,
+      "loss": 3.3445,
+      "step": 16367
+    },
+    {
+      "epoch": 0.689178947368421,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00011416821975994084,
+      "loss": 3.3862,
+      "step": 16368
+    },
+    {
+      "epoch": 0.689221052631579,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00011413989196139473,
+      "loss": 3.4973,
+      "step": 16369
+    },
+    {
+      "epoch": 0.6892631578947368,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0001141115666380601,
+      "loss": 2.8978,
+      "step": 16370
+    },
+    {
+      "epoch": 0.6893052631578948,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00011408324379045321,
+      "loss": 3.2052,
+      "step": 16371
+    },
+    {
+      "epoch": 0.6893473684210526,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00011405492341908977,
+      "loss": 3.3215,
+      "step": 16372
+    },
+    {
+      "epoch": 0.6893894736842106,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00011402660552448604,
+      "loss": 3.1445,
+      "step": 16373
+    },
+    {
+      "epoch": 0.6894315789473684,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00011399829010715765,
+      "loss": 3.2772,
+      "step": 16374
+    },
+    {
+      "epoch": 0.6894736842105263,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001139699771676207,
+      "loss": 3.0099,
+      "step": 16375
+    },
+    {
+      "epoch": 0.6895157894736842,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00011394166670639091,
+      "loss": 2.8397,
+      "step": 16376
+    },
+    {
+      "epoch": 0.6895578947368421,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00011391335872398404,
+      "loss": 2.8979,
+      "step": 16377
+    },
+    {
+      "epoch": 0.6896,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00011388505322091605,
+      "loss": 2.9595,
+      "step": 16378
+    },
+    {
+      "epoch": 0.689642105263158,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00011385675019770225,
+      "loss": 2.9557,
+      "step": 16379
+    },
+    {
+      "epoch": 0.6896842105263158,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00011382844965485873,
+      "loss": 3.1993,
+      "step": 16380
+    },
+    {
+      "epoch": 0.6897263157894736,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00011380015159290064,
+      "loss": 3.0748,
+      "step": 16381
+    },
+    {
+      "epoch": 0.6897684210526316,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00011377185601234386,
+      "loss": 3.0308,
+      "step": 16382
+    },
+    {
+      "epoch": 0.6898105263157894,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00011374356291370383,
+      "loss": 3.2683,
+      "step": 16383
+    },
+    {
+      "epoch": 0.6898526315789474,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00011371527229749596,
+      "loss": 2.7047,
+      "step": 16384
+    },
+    {
+      "epoch": 0.6898947368421052,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0001136869841642357,
+      "loss": 3.1632,
+      "step": 16385
+    },
+    {
+      "epoch": 0.6899368421052632,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00011365869851443834,
+      "loss": 3.2978,
+      "step": 16386
+    },
+    {
+      "epoch": 0.689978947368421,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00011363041534861948,
+      "loss": 3.0313,
+      "step": 16387
+    },
+    {
+      "epoch": 0.690021052631579,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00011360213466729405,
+      "loss": 3.3926,
+      "step": 16388
+    },
+    {
+      "epoch": 0.6900631578947368,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00011357385647097756,
+      "loss": 3.0301,
+      "step": 16389
+    },
+    {
+      "epoch": 0.6901052631578948,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00011354558076018508,
+      "loss": 3.2553,
+      "step": 16390
+    },
+    {
+      "epoch": 0.6901473684210526,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00011351730753543183,
+      "loss": 2.6601,
+      "step": 16391
+    },
+    {
+      "epoch": 0.6901894736842106,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00011348903679723289,
+      "loss": 3.2735,
+      "step": 16392
+    },
+    {
+      "epoch": 0.6902315789473684,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00011346076854610316,
+      "loss": 3.595,
+      "step": 16393
+    },
+    {
+      "epoch": 0.6902736842105263,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00011343250278255801,
+      "loss": 2.9079,
+      "step": 16394
+    },
+    {
+      "epoch": 0.6903157894736842,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.000113404239507112,
+      "loss": 3.2109,
+      "step": 16395
+    },
+    {
+      "epoch": 0.6903578947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00011337597872028038,
+      "loss": 2.8577,
+      "step": 16396
+    },
+    {
+      "epoch": 0.6904,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00011334772042257787,
+      "loss": 3.4765,
+      "step": 16397
+    },
+    {
+      "epoch": 0.6904421052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00011331946461451936,
+      "loss": 3.1288,
+      "step": 16398
+    },
+    {
+      "epoch": 0.6904842105263158,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00011329121129661959,
+      "loss": 3.2431,
+      "step": 16399
+    },
+    {
+      "epoch": 0.6905263157894737,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00011326296046939332,
+      "loss": 2.7721,
+      "step": 16400
+    },
+    {
+      "epoch": 0.6905684210526316,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00011323471213335526,
+      "loss": 3.0284,
+      "step": 16401
+    },
+    {
+      "epoch": 0.6906105263157895,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00011320646628901998,
+      "loss": 3.1735,
+      "step": 16402
+    },
+    {
+      "epoch": 0.6906526315789474,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00011317822293690222,
+      "loss": 3.4179,
+      "step": 16403
+    },
+    {
+      "epoch": 0.6906947368421053,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001131499820775165,
+      "loss": 3.2273,
+      "step": 16404
+    },
+    {
+      "epoch": 0.6907368421052632,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00011312174371137729,
+      "loss": 2.8546,
+      "step": 16405
+    },
+    {
+      "epoch": 0.690778947368421,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00011309350783899911,
+      "loss": 3.0019,
+      "step": 16406
+    },
+    {
+      "epoch": 0.690821052631579,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00011306527446089634,
+      "loss": 3.4377,
+      "step": 16407
+    },
+    {
+      "epoch": 0.6908631578947368,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0001130370435775834,
+      "loss": 3.2937,
+      "step": 16408
+    },
+    {
+      "epoch": 0.6909052631578947,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00011300881518957448,
+      "loss": 2.9478,
+      "step": 16409
+    },
+    {
+      "epoch": 0.6909473684210526,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00011298058929738412,
+      "loss": 3.4034,
+      "step": 16410
+    },
+    {
+      "epoch": 0.6909894736842105,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00011295236590152639,
+      "loss": 3.264,
+      "step": 16411
+    },
+    {
+      "epoch": 0.6910315789473684,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00011292414500251555,
+      "loss": 2.7518,
+      "step": 16412
+    },
+    {
+      "epoch": 0.6910736842105263,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00011289592660086573,
+      "loss": 2.9315,
+      "step": 16413
+    },
+    {
+      "epoch": 0.6911157894736842,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00011286771069709104,
+      "loss": 3.0569,
+      "step": 16414
+    },
+    {
+      "epoch": 0.6911578947368421,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00011283949729170553,
+      "loss": 3.3152,
+      "step": 16415
+    },
+    {
+      "epoch": 0.6912,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00011281128638522322,
+      "loss": 2.7624,
+      "step": 16416
+    },
+    {
+      "epoch": 0.6912421052631579,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00011278307797815798,
+      "loss": 3.1863,
+      "step": 16417
+    },
+    {
+      "epoch": 0.6912842105263158,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00011275487207102394,
+      "loss": 2.9469,
+      "step": 16418
+    },
+    {
+      "epoch": 0.6913263157894737,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0001127266686643349,
+      "loss": 3.2813,
+      "step": 16419
+    },
+    {
+      "epoch": 0.6913684210526316,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00011269846775860465,
+      "loss": 2.6729,
+      "step": 16420
+    },
+    {
+      "epoch": 0.6914105263157895,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00011267026935434699,
+      "loss": 3.4617,
+      "step": 16421
+    },
+    {
+      "epoch": 0.6914526315789473,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00011264207345207564,
+      "loss": 3.06,
+      "step": 16422
+    },
+    {
+      "epoch": 0.6914947368421053,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00011261388005230436,
+      "loss": 3.1881,
+      "step": 16423
+    },
+    {
+      "epoch": 0.6915368421052631,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00011258568915554666,
+      "loss": 2.7078,
+      "step": 16424
+    },
+    {
+      "epoch": 0.6915789473684211,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00011255750076231632,
+      "loss": 2.9213,
+      "step": 16425
+    },
+    {
+      "epoch": 0.6916210526315789,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00011252931487312687,
+      "loss": 2.9451,
+      "step": 16426
+    },
+    {
+      "epoch": 0.6916631578947369,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00011250113148849175,
+      "loss": 3.046,
+      "step": 16427
+    },
+    {
+      "epoch": 0.6917052631578947,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00011247295060892448,
+      "loss": 2.7479,
+      "step": 16428
+    },
+    {
+      "epoch": 0.6917473684210527,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00011244477223493845,
+      "loss": 2.9175,
+      "step": 16429
+    },
+    {
+      "epoch": 0.6917894736842105,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00011241659636704704,
+      "loss": 3.3704,
+      "step": 16430
+    },
+    {
+      "epoch": 0.6918315789473685,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00011238842300576352,
+      "loss": 2.9725,
+      "step": 16431
+    },
+    {
+      "epoch": 0.6918736842105263,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00011236025215160143,
+      "loss": 3.3457,
+      "step": 16432
+    },
+    {
+      "epoch": 0.6919157894736843,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00011233208380507362,
+      "loss": 3.2574,
+      "step": 16433
+    },
+    {
+      "epoch": 0.6919578947368421,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0001123039179666936,
+      "loss": 3.3999,
+      "step": 16434
+    },
+    {
+      "epoch": 0.692,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0001122757546369744,
+      "loss": 3.1016,
+      "step": 16435
+    },
+    {
+      "epoch": 0.6920421052631579,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00011224759381642902,
+      "loss": 3.3405,
+      "step": 16436
+    },
+    {
+      "epoch": 0.6920842105263157,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00011221943550557085,
+      "loss": 2.5575,
+      "step": 16437
+    },
+    {
+      "epoch": 0.6921263157894737,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00011219127970491247,
+      "loss": 3.1215,
+      "step": 16438
+    },
+    {
+      "epoch": 0.6921684210526315,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00011216312641496726,
+      "loss": 3.3596,
+      "step": 16439
+    },
+    {
+      "epoch": 0.6922105263157895,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00011213497563624772,
+      "loss": 3.0254,
+      "step": 16440
+    },
+    {
+      "epoch": 0.6922526315789473,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00011210682736926713,
+      "loss": 3.0889,
+      "step": 16441
+    },
+    {
+      "epoch": 0.6922947368421053,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00011207868161453796,
+      "loss": 3.1656,
+      "step": 16442
+    },
+    {
+      "epoch": 0.6923368421052631,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00011205053837257326,
+      "loss": 3.1241,
+      "step": 16443
+    },
+    {
+      "epoch": 0.6923789473684211,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00011202239764388567,
+      "loss": 2.8511,
+      "step": 16444
+    },
+    {
+      "epoch": 0.6924210526315789,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00011199425942898774,
+      "loss": 3.0178,
+      "step": 16445
+    },
+    {
+      "epoch": 0.6924631578947369,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00011196612372839249,
+      "loss": 3.0678,
+      "step": 16446
+    },
+    {
+      "epoch": 0.6925052631578947,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00011193799054261205,
+      "loss": 3.2099,
+      "step": 16447
+    },
+    {
+      "epoch": 0.6925473684210526,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00011190985987215943,
+      "loss": 3.0619,
+      "step": 16448
+    },
+    {
+      "epoch": 0.6925894736842105,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0001118817317175467,
+      "loss": 3.1331,
+      "step": 16449
+    },
+    {
+      "epoch": 0.6926315789473684,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00011185360607928665,
+      "loss": 3.1066,
+      "step": 16450
+    },
+    {
+      "epoch": 0.6926736842105263,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00011182548295789158,
+      "loss": 3.1472,
+      "step": 16451
+    },
+    {
+      "epoch": 0.6927157894736842,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00011179736235387378,
+      "loss": 2.87,
+      "step": 16452
+    },
+    {
+      "epoch": 0.6927578947368421,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00011176924426774582,
+      "loss": 2.8794,
+      "step": 16453
+    },
+    {
+      "epoch": 0.6928,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00011174112870001962,
+      "loss": 2.951,
+      "step": 16454
+    },
+    {
+      "epoch": 0.6928421052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00011171301565120781,
+      "loss": 3.0599,
+      "step": 16455
+    },
+    {
+      "epoch": 0.6928842105263158,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00011168490512182219,
+      "loss": 3.0366,
+      "step": 16456
+    },
+    {
+      "epoch": 0.6929263157894737,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0001116567971123752,
+      "loss": 2.6173,
+      "step": 16457
+    },
+    {
+      "epoch": 0.6929684210526316,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00011162869162337878,
+      "loss": 3.2064,
+      "step": 16458
+    },
+    {
+      "epoch": 0.6930105263157895,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00011160058865534502,
+      "loss": 3.0706,
+      "step": 16459
+    },
+    {
+      "epoch": 0.6930526315789474,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00011157248820878596,
+      "loss": 3.1579,
+      "step": 16460
+    },
+    {
+      "epoch": 0.6930947368421052,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0001115443902842134,
+      "loss": 3.3192,
+      "step": 16461
+    },
+    {
+      "epoch": 0.6931368421052632,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00011151629488213955,
+      "loss": 3.4768,
+      "step": 16462
+    },
+    {
+      "epoch": 0.693178947368421,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00011148820200307594,
+      "loss": 3.3651,
+      "step": 16463
+    },
+    {
+      "epoch": 0.693221052631579,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00011146011164753461,
+      "loss": 2.7716,
+      "step": 16464
+    },
+    {
+      "epoch": 0.6932631578947368,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00011143202381602726,
+      "loss": 3.1964,
+      "step": 16465
+    },
+    {
+      "epoch": 0.6933052631578948,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00011140393850906566,
+      "loss": 3.1017,
+      "step": 16466
+    },
+    {
+      "epoch": 0.6933473684210526,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00011137585572716144,
+      "loss": 3.4841,
+      "step": 16467
+    },
+    {
+      "epoch": 0.6933894736842106,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00011134777547082626,
+      "loss": 3.0423,
+      "step": 16468
+    },
+    {
+      "epoch": 0.6934315789473684,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00011131969774057166,
+      "loss": 3.0539,
+      "step": 16469
+    },
+    {
+      "epoch": 0.6934736842105264,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00011129162253690916,
+      "loss": 3.3442,
+      "step": 16470
+    },
+    {
+      "epoch": 0.6935157894736842,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00011126354986035042,
+      "loss": 3.2052,
+      "step": 16471
+    },
+    {
+      "epoch": 0.6935578947368422,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00011123547971140679,
+      "loss": 2.8121,
+      "step": 16472
+    },
+    {
+      "epoch": 0.6936,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00011120741209058969,
+      "loss": 3.225,
+      "step": 16473
+    },
+    {
+      "epoch": 0.693642105263158,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00011117934699841045,
+      "loss": 3.1394,
+      "step": 16474
+    },
+    {
+      "epoch": 0.6936842105263158,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00011115128443538042,
+      "loss": 2.7482,
+      "step": 16475
+    },
+    {
+      "epoch": 0.6937263157894736,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00011112322440201081,
+      "loss": 2.7058,
+      "step": 16476
+    },
+    {
+      "epoch": 0.6937684210526316,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00011109516689881282,
+      "loss": 3.2368,
+      "step": 16477
+    },
+    {
+      "epoch": 0.6938105263157894,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00011106711192629776,
+      "loss": 2.9772,
+      "step": 16478
+    },
+    {
+      "epoch": 0.6938526315789474,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0001110390594849767,
+      "loss": 2.3507,
+      "step": 16479
+    },
+    {
+      "epoch": 0.6938947368421052,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00011101100957536067,
+      "loss": 3.2843,
+      "step": 16480
+    },
+    {
+      "epoch": 0.6939368421052632,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00011098296219796075,
+      "loss": 2.9698,
+      "step": 16481
+    },
+    {
+      "epoch": 0.693978947368421,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00011095491735328795,
+      "loss": 3.3932,
+      "step": 16482
+    },
+    {
+      "epoch": 0.694021052631579,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00011092687504185307,
+      "loss": 2.9106,
+      "step": 16483
+    },
+    {
+      "epoch": 0.6940631578947368,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0001108988352641673,
+      "loss": 3.0251,
+      "step": 16484
+    },
+    {
+      "epoch": 0.6941052631578948,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00011087079802074112,
+      "loss": 3.283,
+      "step": 16485
+    },
+    {
+      "epoch": 0.6941473684210526,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00011084276331208565,
+      "loss": 3.0841,
+      "step": 16486
+    },
+    {
+      "epoch": 0.6941894736842106,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00011081473113871152,
+      "loss": 3.01,
+      "step": 16487
+    },
+    {
+      "epoch": 0.6942315789473684,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00011078670150112946,
+      "loss": 2.869,
+      "step": 16488
+    },
+    {
+      "epoch": 0.6942736842105263,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00011075867439985013,
+      "loss": 3.5004,
+      "step": 16489
+    },
+    {
+      "epoch": 0.6943157894736842,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00011073064983538402,
+      "loss": 2.6513,
+      "step": 16490
+    },
+    {
+      "epoch": 0.6943578947368421,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00011070262780824205,
+      "loss": 2.7632,
+      "step": 16491
+    },
+    {
+      "epoch": 0.6944,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00011067460831893436,
+      "loss": 2.9029,
+      "step": 16492
+    },
+    {
+      "epoch": 0.6944421052631579,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001106465913679717,
+      "loss": 2.8965,
+      "step": 16493
+    },
+    {
+      "epoch": 0.6944842105263158,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00011061857695586438,
+      "loss": 3.4039,
+      "step": 16494
+    },
+    {
+      "epoch": 0.6945263157894737,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00011059056508312284,
+      "loss": 3.1164,
+      "step": 16495
+    },
+    {
+      "epoch": 0.6945684210526316,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00011056255575025742,
+      "loss": 3.1529,
+      "step": 16496
+    },
+    {
+      "epoch": 0.6946105263157895,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00011053454895777829,
+      "loss": 2.9859,
+      "step": 16497
+    },
+    {
+      "epoch": 0.6946526315789474,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.000110506544706196,
+      "loss": 3.0734,
+      "step": 16498
+    },
+    {
+      "epoch": 0.6946947368421053,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00011047854299602039,
+      "loss": 2.6992,
+      "step": 16499
+    },
+    {
+      "epoch": 0.6947368421052632,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00011045054382776199,
+      "loss": 2.7945,
+      "step": 16500
+    },
+    {
+      "epoch": 0.694778947368421,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00011042254720193052,
+      "loss": 3.1227,
+      "step": 16501
+    },
+    {
+      "epoch": 0.6948210526315789,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00011039455311903632,
+      "loss": 2.9842,
+      "step": 16502
+    },
+    {
+      "epoch": 0.6948631578947368,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00011036656157958936,
+      "loss": 3.3261,
+      "step": 16503
+    },
+    {
+      "epoch": 0.6949052631578947,
+      "grad_norm": 3.84375,
+      "learning_rate": 0.00011033857258409946,
+      "loss": 3.4224,
+      "step": 16504
+    },
+    {
+      "epoch": 0.6949473684210526,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00011031058613307687,
+      "loss": 2.8688,
+      "step": 16505
+    },
+    {
+      "epoch": 0.6949894736842105,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0001102826022270311,
+      "loss": 3.2516,
+      "step": 16506
+    },
+    {
+      "epoch": 0.6950315789473684,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00011025462086647233,
+      "loss": 3.2864,
+      "step": 16507
+    },
+    {
+      "epoch": 0.6950736842105263,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00011022664205191002,
+      "loss": 3.4576,
+      "step": 16508
+    },
+    {
+      "epoch": 0.6951157894736842,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00011019866578385412,
+      "loss": 3.2072,
+      "step": 16509
+    },
+    {
+      "epoch": 0.6951578947368421,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00011017069206281427,
+      "loss": 3.1646,
+      "step": 16510
+    },
+    {
+      "epoch": 0.6952,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00011014272088930013,
+      "loss": 3.0003,
+      "step": 16511
+    },
+    {
+      "epoch": 0.6952421052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00011011475226382131,
+      "loss": 3.0334,
+      "step": 16512
+    },
+    {
+      "epoch": 0.6952842105263158,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00011008678618688723,
+      "loss": 3.2111,
+      "step": 16513
+    },
+    {
+      "epoch": 0.6953263157894737,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00011005882265900768,
+      "loss": 2.942,
+      "step": 16514
+    },
+    {
+      "epoch": 0.6953684210526315,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0001100308616806918,
+      "loss": 2.5132,
+      "step": 16515
+    },
+    {
+      "epoch": 0.6954105263157895,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00011000290325244935,
+      "loss": 2.8922,
+      "step": 16516
+    },
+    {
+      "epoch": 0.6954526315789473,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00010997494737478931,
+      "loss": 3.2324,
+      "step": 16517
+    },
+    {
+      "epoch": 0.6954947368421053,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00010994699404822131,
+      "loss": 2.7703,
+      "step": 16518
+    },
+    {
+      "epoch": 0.6955368421052631,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00010991904327325453,
+      "loss": 3.0457,
+      "step": 16519
+    },
+    {
+      "epoch": 0.6955789473684211,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00010989109505039808,
+      "loss": 3.0829,
+      "step": 16520
+    },
+    {
+      "epoch": 0.6956210526315789,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00010986314938016145,
+      "loss": 2.9194,
+      "step": 16521
+    },
+    {
+      "epoch": 0.6956631578947369,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00010983520626305341,
+      "loss": 3.1658,
+      "step": 16522
+    },
+    {
+      "epoch": 0.6957052631578947,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0001098072656995834,
+      "loss": 3.1023,
+      "step": 16523
+    },
+    {
+      "epoch": 0.6957473684210527,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001097793276902601,
+      "loss": 3.1457,
+      "step": 16524
+    },
+    {
+      "epoch": 0.6957894736842105,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001097513922355928,
+      "loss": 2.7371,
+      "step": 16525
+    },
+    {
+      "epoch": 0.6958315789473685,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00010972345933609034,
+      "loss": 3.0368,
+      "step": 16526
+    },
+    {
+      "epoch": 0.6958736842105263,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00010969552899226166,
+      "loss": 3.023,
+      "step": 16527
+    },
+    {
+      "epoch": 0.6959157894736843,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00010966760120461556,
+      "loss": 2.8969,
+      "step": 16528
+    },
+    {
+      "epoch": 0.6959578947368421,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00010963967597366078,
+      "loss": 3.2524,
+      "step": 16529
+    },
+    {
+      "epoch": 0.696,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0001096117532999063,
+      "loss": 3.1935,
+      "step": 16530
+    },
+    {
+      "epoch": 0.6960421052631579,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00010958383318386073,
+      "loss": 3.4846,
+      "step": 16531
+    },
+    {
+      "epoch": 0.6960842105263157,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00010955591562603273,
+      "loss": 2.7995,
+      "step": 16532
+    },
+    {
+      "epoch": 0.6961263157894737,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00010952800062693095,
+      "loss": 2.3446,
+      "step": 16533
+    },
+    {
+      "epoch": 0.6961684210526315,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00010950008818706391,
+      "loss": 3.0496,
+      "step": 16534
+    },
+    {
+      "epoch": 0.6962105263157895,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00010947217830694023,
+      "loss": 2.975,
+      "step": 16535
+    },
+    {
+      "epoch": 0.6962526315789473,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00010944427098706821,
+      "loss": 3.2035,
+      "step": 16536
+    },
+    {
+      "epoch": 0.6962947368421053,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00010941636622795655,
+      "loss": 3.3659,
+      "step": 16537
+    },
+    {
+      "epoch": 0.6963368421052631,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00010938846403011352,
+      "loss": 3.1652,
+      "step": 16538
+    },
+    {
+      "epoch": 0.6963789473684211,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00010936056439404746,
+      "loss": 3.0609,
+      "step": 16539
+    },
+    {
+      "epoch": 0.6964210526315789,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00010933266732026665,
+      "loss": 3.314,
+      "step": 16540
+    },
+    {
+      "epoch": 0.6964631578947369,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00010930477280927939,
+      "loss": 3.3458,
+      "step": 16541
+    },
+    {
+      "epoch": 0.6965052631578947,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00010927688086159385,
+      "loss": 3.1161,
+      "step": 16542
+    },
+    {
+      "epoch": 0.6965473684210526,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00010924899147771819,
+      "loss": 2.9989,
+      "step": 16543
+    },
+    {
+      "epoch": 0.6965894736842105,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00010922110465816044,
+      "loss": 3.3675,
+      "step": 16544
+    },
+    {
+      "epoch": 0.6966315789473684,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00010919322040342886,
+      "loss": 2.9962,
+      "step": 16545
+    },
+    {
+      "epoch": 0.6966736842105263,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00010916533871403134,
+      "loss": 2.9618,
+      "step": 16546
+    },
+    {
+      "epoch": 0.6967157894736842,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00010913745959047589,
+      "loss": 3.24,
+      "step": 16547
+    },
+    {
+      "epoch": 0.6967578947368421,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00010910958303327043,
+      "loss": 3.0315,
+      "step": 16548
+    },
+    {
+      "epoch": 0.6968,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0001090817090429228,
+      "loss": 2.5952,
+      "step": 16549
+    },
+    {
+      "epoch": 0.6968421052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00010905383761994087,
+      "loss": 3.202,
+      "step": 16550
+    },
+    {
+      "epoch": 0.6968842105263158,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00010902596876483235,
+      "loss": 2.9901,
+      "step": 16551
+    },
+    {
+      "epoch": 0.6969263157894737,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0001089981024781051,
+      "loss": 3.1672,
+      "step": 16552
+    },
+    {
+      "epoch": 0.6969684210526316,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00010897023876026673,
+      "loss": 2.943,
+      "step": 16553
+    },
+    {
+      "epoch": 0.6970105263157895,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00010894237761182496,
+      "loss": 3.2603,
+      "step": 16554
+    },
+    {
+      "epoch": 0.6970526315789474,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0001089145190332873,
+      "loss": 3.3624,
+      "step": 16555
+    },
+    {
+      "epoch": 0.6970947368421052,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00010888666302516137,
+      "loss": 2.8968,
+      "step": 16556
+    },
+    {
+      "epoch": 0.6971368421052632,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001088588095879546,
+      "loss": 3.1117,
+      "step": 16557
+    },
+    {
+      "epoch": 0.697178947368421,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00010883095872217442,
+      "loss": 3.0859,
+      "step": 16558
+    },
+    {
+      "epoch": 0.697221052631579,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00010880311042832846,
+      "loss": 2.9567,
+      "step": 16559
+    },
+    {
+      "epoch": 0.6972631578947368,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00010877526470692378,
+      "loss": 3.006,
+      "step": 16560
+    },
+    {
+      "epoch": 0.6973052631578948,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00010874742155846792,
+      "loss": 3.0657,
+      "step": 16561
+    },
+    {
+      "epoch": 0.6973473684210526,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00010871958098346807,
+      "loss": 3.3572,
+      "step": 16562
+    },
+    {
+      "epoch": 0.6973894736842106,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00010869174298243146,
+      "loss": 2.764,
+      "step": 16563
+    },
+    {
+      "epoch": 0.6974315789473684,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00010866390755586524,
+      "loss": 2.7574,
+      "step": 16564
+    },
+    {
+      "epoch": 0.6974736842105264,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00010863607470427653,
+      "loss": 2.8312,
+      "step": 16565
+    },
+    {
+      "epoch": 0.6975157894736842,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00010860824442817258,
+      "loss": 3.2233,
+      "step": 16566
+    },
+    {
+      "epoch": 0.6975578947368422,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0001085804167280601,
+      "loss": 2.9248,
+      "step": 16567
+    },
+    {
+      "epoch": 0.6976,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001085525916044464,
+      "loss": 2.6604,
+      "step": 16568
+    },
+    {
+      "epoch": 0.6976421052631578,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00010852476905783823,
+      "loss": 3.0337,
+      "step": 16569
+    },
+    {
+      "epoch": 0.6976842105263158,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00010849694908874259,
+      "loss": 3.2376,
+      "step": 16570
+    },
+    {
+      "epoch": 0.6977263157894736,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00010846913169766626,
+      "loss": 3.2844,
+      "step": 16571
+    },
+    {
+      "epoch": 0.6977684210526316,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00010844131688511594,
+      "loss": 3.1323,
+      "step": 16572
+    },
+    {
+      "epoch": 0.6978105263157894,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00010841350465159871,
+      "loss": 3.3473,
+      "step": 16573
+    },
+    {
+      "epoch": 0.6978526315789474,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00010838569499762088,
+      "loss": 3.2903,
+      "step": 16574
+    },
+    {
+      "epoch": 0.6978947368421052,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0001083578879236895,
+      "loss": 2.8586,
+      "step": 16575
+    },
+    {
+      "epoch": 0.6979368421052632,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00010833008343031078,
+      "loss": 3.3354,
+      "step": 16576
+    },
+    {
+      "epoch": 0.697978947368421,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00010830228151799157,
+      "loss": 3.1322,
+      "step": 16577
+    },
+    {
+      "epoch": 0.698021052631579,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00010827448218723831,
+      "loss": 3.2725,
+      "step": 16578
+    },
+    {
+      "epoch": 0.6980631578947368,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00010824668543855737,
+      "loss": 3.2174,
+      "step": 16579
+    },
+    {
+      "epoch": 0.6981052631578948,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00010821889127245546,
+      "loss": 2.7626,
+      "step": 16580
+    },
+    {
+      "epoch": 0.6981473684210526,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00010819109968943858,
+      "loss": 3.3108,
+      "step": 16581
+    },
+    {
+      "epoch": 0.6981894736842105,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00010816331069001342,
+      "loss": 3.48,
+      "step": 16582
+    },
+    {
+      "epoch": 0.6982315789473684,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00010813552427468593,
+      "loss": 3.2027,
+      "step": 16583
+    },
+    {
+      "epoch": 0.6982736842105263,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00010810774044396257,
+      "loss": 3.1309,
+      "step": 16584
+    },
+    {
+      "epoch": 0.6983157894736842,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00010807995919834946,
+      "loss": 2.8929,
+      "step": 16585
+    },
+    {
+      "epoch": 0.6983578947368421,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00010805218053835276,
+      "loss": 3.2349,
+      "step": 16586
+    },
+    {
+      "epoch": 0.6984,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00010802440446447853,
+      "loss": 3.4102,
+      "step": 16587
+    },
+    {
+      "epoch": 0.6984421052631579,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00010799663097723276,
+      "loss": 2.8622,
+      "step": 16588
+    },
+    {
+      "epoch": 0.6984842105263158,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0001079688600771217,
+      "loss": 3.0378,
+      "step": 16589
+    },
+    {
+      "epoch": 0.6985263157894737,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00010794109176465095,
+      "loss": 3.0569,
+      "step": 16590
+    },
+    {
+      "epoch": 0.6985684210526316,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00010791332604032669,
+      "loss": 3.3399,
+      "step": 16591
+    },
+    {
+      "epoch": 0.6986105263157895,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00010788556290465468,
+      "loss": 3.3416,
+      "step": 16592
+    },
+    {
+      "epoch": 0.6986526315789474,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00010785780235814074,
+      "loss": 3.1937,
+      "step": 16593
+    },
+    {
+      "epoch": 0.6986947368421053,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00010783004440129063,
+      "loss": 3.0957,
+      "step": 16594
+    },
+    {
+      "epoch": 0.6987368421052632,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00010780228903460998,
+      "loss": 3.5741,
+      "step": 16595
+    },
+    {
+      "epoch": 0.698778947368421,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00010777453625860473,
+      "loss": 3.1415,
+      "step": 16596
+    },
+    {
+      "epoch": 0.6988210526315789,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00010774678607378013,
+      "loss": 2.5353,
+      "step": 16597
+    },
+    {
+      "epoch": 0.6988631578947369,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00010771903848064204,
+      "loss": 3.1803,
+      "step": 16598
+    },
+    {
+      "epoch": 0.6989052631578947,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00010769129347969589,
+      "loss": 3.0482,
+      "step": 16599
+    },
+    {
+      "epoch": 0.6989473684210527,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00010766355107144718,
+      "loss": 3.4186,
+      "step": 16600
+    },
+    {
+      "epoch": 0.6989894736842105,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00010763581125640134,
+      "loss": 3.096,
+      "step": 16601
+    },
+    {
+      "epoch": 0.6990315789473684,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00010760807403506373,
+      "loss": 2.8746,
+      "step": 16602
+    },
+    {
+      "epoch": 0.6990736842105263,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00010758033940793971,
+      "loss": 3.2631,
+      "step": 16603
+    },
+    {
+      "epoch": 0.6991157894736842,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00010755260737553449,
+      "loss": 3.1139,
+      "step": 16604
+    },
+    {
+      "epoch": 0.6991578947368421,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00010752487793835347,
+      "loss": 3.4771,
+      "step": 16605
+    },
+    {
+      "epoch": 0.6992,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00010749715109690179,
+      "loss": 3.1211,
+      "step": 16606
+    },
+    {
+      "epoch": 0.6992421052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00010746942685168457,
+      "loss": 3.1463,
+      "step": 16607
+    },
+    {
+      "epoch": 0.6992842105263158,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00010744170520320692,
+      "loss": 3.243,
+      "step": 16608
+    },
+    {
+      "epoch": 0.6993263157894737,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00010741398615197392,
+      "loss": 3.1078,
+      "step": 16609
+    },
+    {
+      "epoch": 0.6993684210526315,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00010738626969849056,
+      "loss": 3.1656,
+      "step": 16610
+    },
+    {
+      "epoch": 0.6994105263157895,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0001073585558432617,
+      "loss": 3.5144,
+      "step": 16611
+    },
+    {
+      "epoch": 0.6994526315789473,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00010733084458679246,
+      "loss": 2.9088,
+      "step": 16612
+    },
+    {
+      "epoch": 0.6994947368421053,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00010730313592958757,
+      "loss": 3.1005,
+      "step": 16613
+    },
+    {
+      "epoch": 0.6995368421052631,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00010727542987215192,
+      "loss": 2.8828,
+      "step": 16614
+    },
+    {
+      "epoch": 0.6995789473684211,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00010724772641499022,
+      "loss": 3.2217,
+      "step": 16615
+    },
+    {
+      "epoch": 0.6996210526315789,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00010722002555860721,
+      "loss": 3.4868,
+      "step": 16616
+    },
+    {
+      "epoch": 0.6996631578947369,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00010719232730350762,
+      "loss": 3.2557,
+      "step": 16617
+    },
+    {
+      "epoch": 0.6997052631578947,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.000107164631650196,
+      "loss": 2.9334,
+      "step": 16618
+    },
+    {
+      "epoch": 0.6997473684210527,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00010713693859917687,
+      "loss": 3.3321,
+      "step": 16619
+    },
+    {
+      "epoch": 0.6997894736842105,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00010710924815095496,
+      "loss": 2.9559,
+      "step": 16620
+    },
+    {
+      "epoch": 0.6998315789473685,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00010708156030603467,
+      "loss": 3.4068,
+      "step": 16621
+    },
+    {
+      "epoch": 0.6998736842105263,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00010705387506492042,
+      "loss": 2.6694,
+      "step": 16622
+    },
+    {
+      "epoch": 0.6999157894736842,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00010702619242811659,
+      "loss": 2.721,
+      "step": 16623
+    },
+    {
+      "epoch": 0.6999578947368421,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001069985123961276,
+      "loss": 2.7245,
+      "step": 16624
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00010697083496945764,
+      "loss": 3.3076,
+      "step": 16625
+    },
+    {
+      "epoch": 0.7000421052631579,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00010694316014861093,
+      "loss": 3.3274,
+      "step": 16626
+    },
+    {
+      "epoch": 0.7000842105263158,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001069154879340919,
+      "loss": 3.111,
+      "step": 16627
+    },
+    {
+      "epoch": 0.7001263157894737,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00010688781832640452,
+      "loss": 2.4551,
+      "step": 16628
+    },
+    {
+      "epoch": 0.7001684210526316,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00010686015132605296,
+      "loss": 2.5352,
+      "step": 16629
+    },
+    {
+      "epoch": 0.7002105263157895,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00010683248693354123,
+      "loss": 3.0144,
+      "step": 16630
+    },
+    {
+      "epoch": 0.7002526315789473,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00010680482514937328,
+      "loss": 3.0624,
+      "step": 16631
+    },
+    {
+      "epoch": 0.7002947368421053,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0001067771659740534,
+      "loss": 3.3789,
+      "step": 16632
+    },
+    {
+      "epoch": 0.7003368421052631,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00010674950940808505,
+      "loss": 3.3257,
+      "step": 16633
+    },
+    {
+      "epoch": 0.7003789473684211,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00010672185545197252,
+      "loss": 3.298,
+      "step": 16634
+    },
+    {
+      "epoch": 0.7004210526315789,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00010669420410621928,
+      "loss": 3.125,
+      "step": 16635
+    },
+    {
+      "epoch": 0.7004631578947368,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00010666655537132933,
+      "loss": 2.8726,
+      "step": 16636
+    },
+    {
+      "epoch": 0.7005052631578947,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00010663890924780634,
+      "loss": 3.3646,
+      "step": 16637
+    },
+    {
+      "epoch": 0.7005473684210526,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00010661126573615388,
+      "loss": 2.6596,
+      "step": 16638
+    },
+    {
+      "epoch": 0.7005894736842105,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00010658362483687586,
+      "loss": 3.4095,
+      "step": 16639
+    },
+    {
+      "epoch": 0.7006315789473684,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00010655598655047552,
+      "loss": 2.8012,
+      "step": 16640
+    },
+    {
+      "epoch": 0.7006736842105263,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00010652835087745674,
+      "loss": 2.7025,
+      "step": 16641
+    },
+    {
+      "epoch": 0.7007157894736842,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00010650071781832266,
+      "loss": 2.8998,
+      "step": 16642
+    },
+    {
+      "epoch": 0.7007578947368421,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00010647308737357697,
+      "loss": 3.3097,
+      "step": 16643
+    },
+    {
+      "epoch": 0.7008,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00010644545954372304,
+      "loss": 2.5418,
+      "step": 16644
+    },
+    {
+      "epoch": 0.7008421052631579,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0001064178343292641,
+      "loss": 3.0269,
+      "step": 16645
+    },
+    {
+      "epoch": 0.7008842105263158,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00010639021173070357,
+      "loss": 2.8906,
+      "step": 16646
+    },
+    {
+      "epoch": 0.7009263157894737,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00010636259174854454,
+      "loss": 2.9361,
+      "step": 16647
+    },
+    {
+      "epoch": 0.7009684210526316,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00010633497438329049,
+      "loss": 2.8861,
+      "step": 16648
+    },
+    {
+      "epoch": 0.7010105263157895,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00010630735963544425,
+      "loss": 3.0437,
+      "step": 16649
+    },
+    {
+      "epoch": 0.7010526315789474,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00010627974750550922,
+      "loss": 2.6417,
+      "step": 16650
+    },
+    {
+      "epoch": 0.7010947368421052,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001062521379939882,
+      "loss": 3.2037,
+      "step": 16651
+    },
+    {
+      "epoch": 0.7011368421052632,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00010622453110138439,
+      "loss": 3.2096,
+      "step": 16652
+    },
+    {
+      "epoch": 0.701178947368421,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0001061969268282007,
+      "loss": 3.1297,
+      "step": 16653
+    },
+    {
+      "epoch": 0.701221052631579,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00010616932517493996,
+      "loss": 2.9453,
+      "step": 16654
+    },
+    {
+      "epoch": 0.7012631578947368,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00010614172614210529,
+      "loss": 3.1269,
+      "step": 16655
+    },
+    {
+      "epoch": 0.7013052631578948,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00010611412973019915,
+      "loss": 3.0827,
+      "step": 16656
+    },
+    {
+      "epoch": 0.7013473684210526,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00010608653593972473,
+      "loss": 3.1703,
+      "step": 16657
+    },
+    {
+      "epoch": 0.7013894736842106,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00010605894477118431,
+      "loss": 3.0012,
+      "step": 16658
+    },
+    {
+      "epoch": 0.7014315789473684,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00010603135622508089,
+      "loss": 3.0868,
+      "step": 16659
+    },
+    {
+      "epoch": 0.7014736842105264,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00010600377030191701,
+      "loss": 3.0522,
+      "step": 16660
+    },
+    {
+      "epoch": 0.7015157894736842,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00010597618700219527,
+      "loss": 3.1718,
+      "step": 16661
+    },
+    {
+      "epoch": 0.7015578947368422,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00010594860632641817,
+      "loss": 2.8979,
+      "step": 16662
+    },
+    {
+      "epoch": 0.7016,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0001059210282750881,
+      "loss": 3.0655,
+      "step": 16663
+    },
+    {
+      "epoch": 0.7016421052631578,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001058934528487078,
+      "loss": 2.7588,
+      "step": 16664
+    },
+    {
+      "epoch": 0.7016842105263158,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00010586588004777931,
+      "loss": 3.0551,
+      "step": 16665
+    },
+    {
+      "epoch": 0.7017263157894736,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00010583830987280519,
+      "loss": 3.1109,
+      "step": 16666
+    },
+    {
+      "epoch": 0.7017684210526316,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0001058107423242877,
+      "loss": 2.6443,
+      "step": 16667
+    },
+    {
+      "epoch": 0.7018105263157894,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00010578317740272905,
+      "loss": 3.0563,
+      "step": 16668
+    },
+    {
+      "epoch": 0.7018526315789474,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00010575561510863149,
+      "loss": 2.9679,
+      "step": 16669
+    },
+    {
+      "epoch": 0.7018947368421052,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00010572805544249703,
+      "loss": 2.9888,
+      "step": 16670
+    },
+    {
+      "epoch": 0.7019368421052632,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00010570049840482809,
+      "loss": 2.8357,
+      "step": 16671
+    },
+    {
+      "epoch": 0.701978947368421,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00010567294399612632,
+      "loss": 3.021,
+      "step": 16672
+    },
+    {
+      "epoch": 0.702021052631579,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00010564539221689404,
+      "loss": 3.0833,
+      "step": 16673
+    },
+    {
+      "epoch": 0.7020631578947368,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00010561784306763308,
+      "loss": 3.4316,
+      "step": 16674
+    },
+    {
+      "epoch": 0.7021052631578948,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00010559029654884541,
+      "loss": 2.9175,
+      "step": 16675
+    },
+    {
+      "epoch": 0.7021473684210526,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00010556275266103285,
+      "loss": 3.4217,
+      "step": 16676
+    },
+    {
+      "epoch": 0.7021894736842105,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00010553521140469721,
+      "loss": 2.6792,
+      "step": 16677
+    },
+    {
+      "epoch": 0.7022315789473684,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00010550767278034031,
+      "loss": 2.9779,
+      "step": 16678
+    },
+    {
+      "epoch": 0.7022736842105263,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00010548013678846375,
+      "loss": 2.9548,
+      "step": 16679
+    },
+    {
+      "epoch": 0.7023157894736842,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00010545260342956937,
+      "loss": 2.6873,
+      "step": 16680
+    },
+    {
+      "epoch": 0.7023578947368421,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00010542507270415872,
+      "loss": 3.3689,
+      "step": 16681
+    },
+    {
+      "epoch": 0.7024,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00010539754461273338,
+      "loss": 3.3119,
+      "step": 16682
+    },
+    {
+      "epoch": 0.7024421052631579,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00010537001915579486,
+      "loss": 3.1748,
+      "step": 16683
+    },
+    {
+      "epoch": 0.7024842105263158,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00010534249633384465,
+      "loss": 2.8221,
+      "step": 16684
+    },
+    {
+      "epoch": 0.7025263157894737,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00010531497614738414,
+      "loss": 2.5808,
+      "step": 16685
+    },
+    {
+      "epoch": 0.7025684210526316,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00010528745859691482,
+      "loss": 3.0362,
+      "step": 16686
+    },
+    {
+      "epoch": 0.7026105263157895,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.000105259943682938,
+      "loss": 3.2403,
+      "step": 16687
+    },
+    {
+      "epoch": 0.7026526315789474,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00010523243140595496,
+      "loss": 3.1517,
+      "step": 16688
+    },
+    {
+      "epoch": 0.7026947368421053,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0001052049217664669,
+      "loss": 2.9609,
+      "step": 16689
+    },
+    {
+      "epoch": 0.7027368421052631,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00010517741476497502,
+      "loss": 3.1796,
+      "step": 16690
+    },
+    {
+      "epoch": 0.7027789473684211,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00010514991040198052,
+      "loss": 3.1659,
+      "step": 16691
+    },
+    {
+      "epoch": 0.7028210526315789,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00010512240867798434,
+      "loss": 3.3833,
+      "step": 16692
+    },
+    {
+      "epoch": 0.7028631578947369,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00010509490959348786,
+      "loss": 2.8139,
+      "step": 16693
+    },
+    {
+      "epoch": 0.7029052631578947,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00010506741314899165,
+      "loss": 3.3989,
+      "step": 16694
+    },
+    {
+      "epoch": 0.7029473684210527,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00010503991934499702,
+      "loss": 3.3479,
+      "step": 16695
+    },
+    {
+      "epoch": 0.7029894736842105,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00010501242818200471,
+      "loss": 2.8311,
+      "step": 16696
+    },
+    {
+      "epoch": 0.7030315789473685,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00010498493966051561,
+      "loss": 3.0758,
+      "step": 16697
+    },
+    {
+      "epoch": 0.7030736842105263,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00010495745378103052,
+      "loss": 2.6687,
+      "step": 16698
+    },
+    {
+      "epoch": 0.7031157894736842,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00010492997054405013,
+      "loss": 2.9243,
+      "step": 16699
+    },
+    {
+      "epoch": 0.7031578947368421,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001049024899500754,
+      "loss": 3.3541,
+      "step": 16700
+    },
+    {
+      "epoch": 0.7032,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00010487501199960662,
+      "loss": 2.8791,
+      "step": 16701
+    },
+    {
+      "epoch": 0.7032421052631579,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001048475366931447,
+      "loss": 3.3868,
+      "step": 16702
+    },
+    {
+      "epoch": 0.7032842105263158,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00010482006403119016,
+      "loss": 2.4396,
+      "step": 16703
+    },
+    {
+      "epoch": 0.7033263157894737,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00010479259401424343,
+      "loss": 2.8902,
+      "step": 16704
+    },
+    {
+      "epoch": 0.7033684210526315,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00010476512664280503,
+      "loss": 3.3431,
+      "step": 16705
+    },
+    {
+      "epoch": 0.7034105263157895,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001047376619173753,
+      "loss": 3.2178,
+      "step": 16706
+    },
+    {
+      "epoch": 0.7034526315789473,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00010471019983845487,
+      "loss": 2.7693,
+      "step": 16707
+    },
+    {
+      "epoch": 0.7034947368421053,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00010468274040654369,
+      "loss": 3.2433,
+      "step": 16708
+    },
+    {
+      "epoch": 0.7035368421052631,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00010465528362214243,
+      "loss": 3.004,
+      "step": 16709
+    },
+    {
+      "epoch": 0.7035789473684211,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00010462782948575097,
+      "loss": 3.0806,
+      "step": 16710
+    },
+    {
+      "epoch": 0.7036210526315789,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00010460037799786973,
+      "loss": 3.3326,
+      "step": 16711
+    },
+    {
+      "epoch": 0.7036631578947369,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00010457292915899877,
+      "loss": 3.0864,
+      "step": 16712
+    },
+    {
+      "epoch": 0.7037052631578947,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00010454548296963806,
+      "loss": 3.4395,
+      "step": 16713
+    },
+    {
+      "epoch": 0.7037473684210527,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00010451803943028795,
+      "loss": 3.0074,
+      "step": 16714
+    },
+    {
+      "epoch": 0.7037894736842105,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00010449059854144805,
+      "loss": 2.716,
+      "step": 16715
+    },
+    {
+      "epoch": 0.7038315789473685,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00010446316030361867,
+      "loss": 2.7013,
+      "step": 16716
+    },
+    {
+      "epoch": 0.7038736842105263,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00010443572471729929,
+      "loss": 2.702,
+      "step": 16717
+    },
+    {
+      "epoch": 0.7039157894736842,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00010440829178299008,
+      "loss": 3.0986,
+      "step": 16718
+    },
+    {
+      "epoch": 0.7039578947368421,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00010438086150119073,
+      "loss": 3.1983,
+      "step": 16719
+    },
+    {
+      "epoch": 0.704,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00010435343387240099,
+      "loss": 2.7745,
+      "step": 16720
+    },
+    {
+      "epoch": 0.7040421052631579,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00010432600889712057,
+      "loss": 2.8786,
+      "step": 16721
+    },
+    {
+      "epoch": 0.7040842105263158,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.000104298586575849,
+      "loss": 2.8894,
+      "step": 16722
+    },
+    {
+      "epoch": 0.7041263157894737,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00010427116690908617,
+      "loss": 2.9818,
+      "step": 16723
+    },
+    {
+      "epoch": 0.7041684210526316,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0001042437498973313,
+      "loss": 3.048,
+      "step": 16724
+    },
+    {
+      "epoch": 0.7042105263157895,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001042163355410842,
+      "loss": 3.4527,
+      "step": 16725
+    },
+    {
+      "epoch": 0.7042526315789474,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00010418892384084402,
+      "loss": 3.247,
+      "step": 16726
+    },
+    {
+      "epoch": 0.7042947368421053,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00010416151479711042,
+      "loss": 3.0934,
+      "step": 16727
+    },
+    {
+      "epoch": 0.7043368421052632,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00010413410841038264,
+      "loss": 2.9772,
+      "step": 16728
+    },
+    {
+      "epoch": 0.7043789473684211,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00010410670468115996,
+      "loss": 2.6052,
+      "step": 16729
+    },
+    {
+      "epoch": 0.704421052631579,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00010407930360994187,
+      "loss": 2.9569,
+      "step": 16730
+    },
+    {
+      "epoch": 0.7044631578947368,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0001040519051972272,
+      "loss": 3.1986,
+      "step": 16731
+    },
+    {
+      "epoch": 0.7045052631578947,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001040245094435155,
+      "loss": 2.6324,
+      "step": 16732
+    },
+    {
+      "epoch": 0.7045473684210526,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00010399711634930567,
+      "loss": 2.819,
+      "step": 16733
+    },
+    {
+      "epoch": 0.7045894736842105,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00010396972591509685,
+      "loss": 3.4435,
+      "step": 16734
+    },
+    {
+      "epoch": 0.7046315789473684,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00010394233814138803,
+      "loss": 3.2124,
+      "step": 16735
+    },
+    {
+      "epoch": 0.7046736842105263,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001039149530286782,
+      "loss": 3.3302,
+      "step": 16736
+    },
+    {
+      "epoch": 0.7047157894736842,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00010388757057746631,
+      "loss": 3.1718,
+      "step": 16737
+    },
+    {
+      "epoch": 0.7047578947368421,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001038601907882511,
+      "loss": 3.1592,
+      "step": 16738
+    },
+    {
+      "epoch": 0.7048,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0001038328136615316,
+      "loss": 3.2752,
+      "step": 16739
+    },
+    {
+      "epoch": 0.7048421052631579,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0001038054391978065,
+      "loss": 2.9082,
+      "step": 16740
+    },
+    {
+      "epoch": 0.7048842105263158,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00010377806739757454,
+      "loss": 3.2233,
+      "step": 16741
+    },
+    {
+      "epoch": 0.7049263157894737,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00010375069826133435,
+      "loss": 3.0101,
+      "step": 16742
+    },
+    {
+      "epoch": 0.7049684210526316,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00010372333178958463,
+      "loss": 3.1636,
+      "step": 16743
+    },
+    {
+      "epoch": 0.7050105263157894,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00010369596798282393,
+      "loss": 2.7833,
+      "step": 16744
+    },
+    {
+      "epoch": 0.7050526315789474,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00010366860684155072,
+      "loss": 3.2063,
+      "step": 16745
+    },
+    {
+      "epoch": 0.7050947368421052,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00010364124836626362,
+      "loss": 3.6625,
+      "step": 16746
+    },
+    {
+      "epoch": 0.7051368421052632,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00010361389255746103,
+      "loss": 3.1599,
+      "step": 16747
+    },
+    {
+      "epoch": 0.705178947368421,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00010358653941564131,
+      "loss": 2.9522,
+      "step": 16748
+    },
+    {
+      "epoch": 0.705221052631579,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00010355918894130282,
+      "loss": 2.8161,
+      "step": 16749
+    },
+    {
+      "epoch": 0.7052631578947368,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00010353184113494385,
+      "loss": 3.1789,
+      "step": 16750
+    },
+    {
+      "epoch": 0.7053052631578948,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00010350449599706263,
+      "loss": 3.0638,
+      "step": 16751
+    },
+    {
+      "epoch": 0.7053473684210526,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00010347715352815738,
+      "loss": 2.812,
+      "step": 16752
+    },
+    {
+      "epoch": 0.7053894736842106,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00010344981372872614,
+      "loss": 3.1208,
+      "step": 16753
+    },
+    {
+      "epoch": 0.7054315789473684,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00010342247659926717,
+      "loss": 2.6972,
+      "step": 16754
+    },
+    {
+      "epoch": 0.7054736842105264,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001033951421402785,
+      "loss": 3.346,
+      "step": 16755
+    },
+    {
+      "epoch": 0.7055157894736842,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00010336781035225806,
+      "loss": 3.0928,
+      "step": 16756
+    },
+    {
+      "epoch": 0.705557894736842,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00010334048123570383,
+      "loss": 2.5824,
+      "step": 16757
+    },
+    {
+      "epoch": 0.7056,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00010331315479111373,
+      "loss": 2.865,
+      "step": 16758
+    },
+    {
+      "epoch": 0.7056421052631578,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00010328583101898558,
+      "loss": 3.2539,
+      "step": 16759
+    },
+    {
+      "epoch": 0.7056842105263158,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00010325850991981714,
+      "loss": 3.1026,
+      "step": 16760
+    },
+    {
+      "epoch": 0.7057263157894736,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001032311914941064,
+      "loss": 3.4005,
+      "step": 16761
+    },
+    {
+      "epoch": 0.7057684210526316,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00010320387574235071,
+      "loss": 3.1221,
+      "step": 16762
+    },
+    {
+      "epoch": 0.7058105263157894,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00010317656266504805,
+      "loss": 3.0309,
+      "step": 16763
+    },
+    {
+      "epoch": 0.7058526315789474,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00010314925226269589,
+      "loss": 3.3726,
+      "step": 16764
+    },
+    {
+      "epoch": 0.7058947368421052,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0001031219445357918,
+      "loss": 3.3472,
+      "step": 16765
+    },
+    {
+      "epoch": 0.7059368421052632,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00010309463948483333,
+      "loss": 2.7601,
+      "step": 16766
+    },
+    {
+      "epoch": 0.705978947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00010306733711031782,
+      "loss": 3.2505,
+      "step": 16767
+    },
+    {
+      "epoch": 0.706021052631579,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00010304003741274295,
+      "loss": 3.1844,
+      "step": 16768
+    },
+    {
+      "epoch": 0.7060631578947368,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00010301274039260578,
+      "loss": 3.8285,
+      "step": 16769
+    },
+    {
+      "epoch": 0.7061052631578948,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00010298544605040385,
+      "loss": 2.7247,
+      "step": 16770
+    },
+    {
+      "epoch": 0.7061473684210526,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00010295815438663436,
+      "loss": 2.8939,
+      "step": 16771
+    },
+    {
+      "epoch": 0.7061894736842105,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00010293086540179453,
+      "loss": 2.368,
+      "step": 16772
+    },
+    {
+      "epoch": 0.7062315789473684,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00010290357909638154,
+      "loss": 3.0345,
+      "step": 16773
+    },
+    {
+      "epoch": 0.7062736842105263,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00010287629547089239,
+      "loss": 3.2272,
+      "step": 16774
+    },
+    {
+      "epoch": 0.7063157894736842,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00010284901452582449,
+      "loss": 2.9205,
+      "step": 16775
+    },
+    {
+      "epoch": 0.7063578947368421,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00010282173626167445,
+      "loss": 2.8684,
+      "step": 16776
+    },
+    {
+      "epoch": 0.7064,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00010279446067893964,
+      "loss": 3.116,
+      "step": 16777
+    },
+    {
+      "epoch": 0.7064421052631579,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00010276718777811661,
+      "loss": 2.9388,
+      "step": 16778
+    },
+    {
+      "epoch": 0.7064842105263158,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001027399175597025,
+      "loss": 3.6932,
+      "step": 16779
+    },
+    {
+      "epoch": 0.7065263157894737,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00010271265002419409,
+      "loss": 3.127,
+      "step": 16780
+    },
+    {
+      "epoch": 0.7065684210526316,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00010268538517208806,
+      "loss": 3.1645,
+      "step": 16781
+    },
+    {
+      "epoch": 0.7066105263157895,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00010265812300388139,
+      "loss": 2.7272,
+      "step": 16782
+    },
+    {
+      "epoch": 0.7066526315789474,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00010263086352007042,
+      "loss": 3.2609,
+      "step": 16783
+    },
+    {
+      "epoch": 0.7066947368421053,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00010260360672115213,
+      "loss": 3.1046,
+      "step": 16784
+    },
+    {
+      "epoch": 0.7067368421052631,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0001025763526076228,
+      "loss": 2.7069,
+      "step": 16785
+    },
+    {
+      "epoch": 0.7067789473684211,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00010254910117997915,
+      "loss": 2.3588,
+      "step": 16786
+    },
+    {
+      "epoch": 0.7068210526315789,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00010252185243871767,
+      "loss": 2.6971,
+      "step": 16787
+    },
+    {
+      "epoch": 0.7068631578947369,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00010249460638433467,
+      "loss": 3.2409,
+      "step": 16788
+    },
+    {
+      "epoch": 0.7069052631578947,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00010246736301732679,
+      "loss": 2.9812,
+      "step": 16789
+    },
+    {
+      "epoch": 0.7069473684210527,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00010244012233819002,
+      "loss": 2.8533,
+      "step": 16790
+    },
+    {
+      "epoch": 0.7069894736842105,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00010241288434742105,
+      "loss": 3.3971,
+      "step": 16791
+    },
+    {
+      "epoch": 0.7070315789473685,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00010238564904551573,
+      "loss": 3.5716,
+      "step": 16792
+    },
+    {
+      "epoch": 0.7070736842105263,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00010235841643297053,
+      "loss": 3.0048,
+      "step": 16793
+    },
+    {
+      "epoch": 0.7071157894736843,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00010233118651028154,
+      "loss": 3.353,
+      "step": 16794
+    },
+    {
+      "epoch": 0.7071578947368421,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00010230395927794478,
+      "loss": 3.1493,
+      "step": 16795
+    },
+    {
+      "epoch": 0.7072,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00010227673473645638,
+      "loss": 3.3531,
+      "step": 16796
+    },
+    {
+      "epoch": 0.7072421052631579,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00010224951288631221,
+      "loss": 3.0967,
+      "step": 16797
+    },
+    {
+      "epoch": 0.7072842105263157,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00010222229372800848,
+      "loss": 2.6947,
+      "step": 16798
+    },
+    {
+      "epoch": 0.7073263157894737,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00010219507726204078,
+      "loss": 3.2281,
+      "step": 16799
+    },
+    {
+      "epoch": 0.7073684210526315,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00010216786348890516,
+      "loss": 3.3852,
+      "step": 16800
+    },
+    {
+      "epoch": 0.7074105263157895,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00010214065240909739,
+      "loss": 3.1668,
+      "step": 16801
+    },
+    {
+      "epoch": 0.7074526315789473,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001021134440231132,
+      "loss": 2.7059,
+      "step": 16802
+    },
+    {
+      "epoch": 0.7074947368421053,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00010208623833144831,
+      "loss": 3.3877,
+      "step": 16803
+    },
+    {
+      "epoch": 0.7075368421052631,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00010205903533459837,
+      "loss": 2.8692,
+      "step": 16804
+    },
+    {
+      "epoch": 0.7075789473684211,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00010203183503305896,
+      "loss": 2.8384,
+      "step": 16805
+    },
+    {
+      "epoch": 0.7076210526315789,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001020046374273256,
+      "loss": 3.067,
+      "step": 16806
+    },
+    {
+      "epoch": 0.7076631578947369,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00010197744251789393,
+      "loss": 3.3237,
+      "step": 16807
+    },
+    {
+      "epoch": 0.7077052631578947,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00010195025030525935,
+      "loss": 3.2135,
+      "step": 16808
+    },
+    {
+      "epoch": 0.7077473684210527,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00010192306078991723,
+      "loss": 2.6191,
+      "step": 16809
+    },
+    {
+      "epoch": 0.7077894736842105,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.000101895873972363,
+      "loss": 2.8221,
+      "step": 16810
+    },
+    {
+      "epoch": 0.7078315789473684,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0001018686898530919,
+      "loss": 3.3585,
+      "step": 16811
+    },
+    {
+      "epoch": 0.7078736842105263,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00010184150843259923,
+      "loss": 2.7731,
+      "step": 16812
+    },
+    {
+      "epoch": 0.7079157894736842,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0001018143297113801,
+      "loss": 2.7212,
+      "step": 16813
+    },
+    {
+      "epoch": 0.7079578947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00010178715368992989,
+      "loss": 2.8268,
+      "step": 16814
+    },
+    {
+      "epoch": 0.708,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00010175998036874357,
+      "loss": 3.4696,
+      "step": 16815
+    },
+    {
+      "epoch": 0.7080421052631579,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00010173280974831625,
+      "loss": 2.6628,
+      "step": 16816
+    },
+    {
+      "epoch": 0.7080842105263158,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00010170564182914291,
+      "loss": 3.7094,
+      "step": 16817
+    },
+    {
+      "epoch": 0.7081263157894737,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00010167847661171855,
+      "loss": 3.027,
+      "step": 16818
+    },
+    {
+      "epoch": 0.7081684210526316,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00010165131409653808,
+      "loss": 2.8369,
+      "step": 16819
+    },
+    {
+      "epoch": 0.7082105263157895,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00010162415428409635,
+      "loss": 3.4182,
+      "step": 16820
+    },
+    {
+      "epoch": 0.7082526315789474,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00010159699717488813,
+      "loss": 2.8661,
+      "step": 16821
+    },
+    {
+      "epoch": 0.7082947368421053,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00010156984276940831,
+      "loss": 2.8583,
+      "step": 16822
+    },
+    {
+      "epoch": 0.7083368421052632,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001015426910681516,
+      "loss": 3.0961,
+      "step": 16823
+    },
+    {
+      "epoch": 0.7083789473684211,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00010151554207161259,
+      "loss": 3.1078,
+      "step": 16824
+    },
+    {
+      "epoch": 0.708421052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00010148839578028596,
+      "loss": 3.5337,
+      "step": 16825
+    },
+    {
+      "epoch": 0.7084631578947368,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00010146125219466626,
+      "loss": 2.8675,
+      "step": 16826
+    },
+    {
+      "epoch": 0.7085052631578947,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00010143411131524801,
+      "loss": 3.1385,
+      "step": 16827
+    },
+    {
+      "epoch": 0.7085473684210526,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00010140697314252561,
+      "loss": 3.0397,
+      "step": 16828
+    },
+    {
+      "epoch": 0.7085894736842105,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00010137983767699365,
+      "loss": 2.4184,
+      "step": 16829
+    },
+    {
+      "epoch": 0.7086315789473684,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00010135270491914644,
+      "loss": 3.1687,
+      "step": 16830
+    },
+    {
+      "epoch": 0.7086736842105263,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0001013255748694783,
+      "loss": 3.2474,
+      "step": 16831
+    },
+    {
+      "epoch": 0.7087157894736842,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001012984475284835,
+      "loss": 3.0533,
+      "step": 16832
+    },
+    {
+      "epoch": 0.7087578947368421,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00010127132289665616,
+      "loss": 3.0474,
+      "step": 16833
+    },
+    {
+      "epoch": 0.7088,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00010124420097449077,
+      "loss": 2.8751,
+      "step": 16834
+    },
+    {
+      "epoch": 0.7088421052631579,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00010121708176248107,
+      "loss": 3.0923,
+      "step": 16835
+    },
+    {
+      "epoch": 0.7088842105263158,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001011899652611215,
+      "loss": 2.9358,
+      "step": 16836
+    },
+    {
+      "epoch": 0.7089263157894737,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00010116285147090574,
+      "loss": 3.2144,
+      "step": 16837
+    },
+    {
+      "epoch": 0.7089684210526316,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00010113574039232804,
+      "loss": 3.0642,
+      "step": 16838
+    },
+    {
+      "epoch": 0.7090105263157894,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00010110863202588224,
+      "loss": 2.6987,
+      "step": 16839
+    },
+    {
+      "epoch": 0.7090526315789474,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00010108152637206214,
+      "loss": 3.0445,
+      "step": 16840
+    },
+    {
+      "epoch": 0.7090947368421052,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00010105442343136184,
+      "loss": 2.546,
+      "step": 16841
+    },
+    {
+      "epoch": 0.7091368421052632,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00010102732320427477,
+      "loss": 2.9341,
+      "step": 16842
+    },
+    {
+      "epoch": 0.709178947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.000101000225691295,
+      "loss": 3.1991,
+      "step": 16843
+    },
+    {
+      "epoch": 0.709221052631579,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00010097313089291587,
+      "loss": 3.4971,
+      "step": 16844
+    },
+    {
+      "epoch": 0.7092631578947368,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001009460388096313,
+      "loss": 2.8716,
+      "step": 16845
+    },
+    {
+      "epoch": 0.7093052631578948,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00010091894944193475,
+      "loss": 3.3469,
+      "step": 16846
+    },
+    {
+      "epoch": 0.7093473684210526,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00010089186279031979,
+      "loss": 2.8271,
+      "step": 16847
+    },
+    {
+      "epoch": 0.7093894736842106,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00010086477885527989,
+      "loss": 3.139,
+      "step": 16848
+    },
+    {
+      "epoch": 0.7094315789473684,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001008376976373084,
+      "loss": 3.1309,
+      "step": 16849
+    },
+    {
+      "epoch": 0.7094736842105264,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00010081061913689898,
+      "loss": 3.1511,
+      "step": 16850
+    },
+    {
+      "epoch": 0.7095157894736842,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00010078354335454463,
+      "loss": 3.1276,
+      "step": 16851
+    },
+    {
+      "epoch": 0.709557894736842,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00010075647029073895,
+      "loss": 3.2364,
+      "step": 16852
+    },
+    {
+      "epoch": 0.7096,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00010072939994597487,
+      "loss": 3.1303,
+      "step": 16853
+    },
+    {
+      "epoch": 0.7096421052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00010070233232074583,
+      "loss": 3.3399,
+      "step": 16854
+    },
+    {
+      "epoch": 0.7096842105263158,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00010067526741554486,
+      "loss": 3.0039,
+      "step": 16855
+    },
+    {
+      "epoch": 0.7097263157894736,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00010064820523086499,
+      "loss": 3.0008,
+      "step": 16856
+    },
+    {
+      "epoch": 0.7097684210526316,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001006211457671995,
+      "loss": 3.0404,
+      "step": 16857
+    },
+    {
+      "epoch": 0.7098105263157894,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00010059408902504105,
+      "loss": 3.3679,
+      "step": 16858
+    },
+    {
+      "epoch": 0.7098526315789474,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00010056703500488296,
+      "loss": 3.4215,
+      "step": 16859
+    },
+    {
+      "epoch": 0.7098947368421052,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0001005399837072177,
+      "loss": 3.1663,
+      "step": 16860
+    },
+    {
+      "epoch": 0.7099368421052632,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00010051293513253842,
+      "loss": 3.4876,
+      "step": 16861
+    },
+    {
+      "epoch": 0.709978947368421,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00010048588928133784,
+      "loss": 3.0212,
+      "step": 16862
+    },
+    {
+      "epoch": 0.710021052631579,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00010045884615410869,
+      "loss": 3.3315,
+      "step": 16863
+    },
+    {
+      "epoch": 0.7100631578947368,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00010043180575134365,
+      "loss": 3.072,
+      "step": 16864
+    },
+    {
+      "epoch": 0.7101052631578947,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0001004047680735353,
+      "loss": 2.8151,
+      "step": 16865
+    },
+    {
+      "epoch": 0.7101473684210526,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00010037773312117646,
+      "loss": 2.5591,
+      "step": 16866
+    },
+    {
+      "epoch": 0.7101894736842105,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00010035070089475935,
+      "loss": 3.6372,
+      "step": 16867
+    },
+    {
+      "epoch": 0.7102315789473684,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00010032367139477674,
+      "loss": 3.0567,
+      "step": 16868
+    },
+    {
+      "epoch": 0.7102736842105263,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00010029664462172097,
+      "loss": 2.8877,
+      "step": 16869
+    },
+    {
+      "epoch": 0.7103157894736842,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00010026962057608444,
+      "loss": 2.9324,
+      "step": 16870
+    },
+    {
+      "epoch": 0.7103578947368421,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0001002425992583595,
+      "loss": 3.0796,
+      "step": 16871
+    },
+    {
+      "epoch": 0.7104,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00010021558066903835,
+      "loss": 2.9856,
+      "step": 16872
+    },
+    {
+      "epoch": 0.7104421052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00010018856480861349,
+      "loss": 3.1319,
+      "step": 16873
+    },
+    {
+      "epoch": 0.7104842105263158,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00010016155167757682,
+      "loss": 3.0465,
+      "step": 16874
+    },
+    {
+      "epoch": 0.7105263157894737,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00010013454127642068,
+      "loss": 3.1929,
+      "step": 16875
+    },
+    {
+      "epoch": 0.7105684210526316,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00010010753360563713,
+      "loss": 3.2114,
+      "step": 16876
+    },
+    {
+      "epoch": 0.7106105263157895,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00010008052866571821,
+      "loss": 3.0843,
+      "step": 16877
+    },
+    {
+      "epoch": 0.7106526315789473,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001000535264571559,
+      "loss": 2.9095,
+      "step": 16878
+    },
+    {
+      "epoch": 0.7106947368421053,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00010002652698044215,
+      "loss": 3.1875,
+      "step": 16879
+    },
+    {
+      "epoch": 0.7107368421052631,
+      "grad_norm": 0.4453125,
+      "learning_rate": 9.999953023606879e-05,
+      "loss": 3.4109,
+      "step": 16880
+    },
+    {
+      "epoch": 0.7107789473684211,
+      "grad_norm": 0.4140625,
+      "learning_rate": 9.997253622452782e-05,
+      "loss": 3.0866,
+      "step": 16881
+    },
+    {
+      "epoch": 0.7108210526315789,
+      "grad_norm": 0.412109375,
+      "learning_rate": 9.994554494631097e-05,
+      "loss": 3.0554,
+      "step": 16882
+    },
+    {
+      "epoch": 0.7108631578947369,
+      "grad_norm": 0.412109375,
+      "learning_rate": 9.991855640191e-05,
+      "loss": 3.4006,
+      "step": 16883
+    },
+    {
+      "epoch": 0.7109052631578947,
+      "grad_norm": 0.494140625,
+      "learning_rate": 9.989157059181659e-05,
+      "loss": 3.1484,
+      "step": 16884
+    },
+    {
+      "epoch": 0.7109473684210527,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.986458751652239e-05,
+      "loss": 2.9575,
+      "step": 16885
+    },
+    {
+      "epoch": 0.7109894736842105,
+      "grad_norm": 0.43359375,
+      "learning_rate": 9.9837607176519e-05,
+      "loss": 2.8618,
+      "step": 16886
+    },
+    {
+      "epoch": 0.7110315789473685,
+      "grad_norm": 0.4375,
+      "learning_rate": 9.98106295722979e-05,
+      "loss": 3.0809,
+      "step": 16887
+    },
+    {
+      "epoch": 0.7110736842105263,
+      "grad_norm": 0.482421875,
+      "learning_rate": 9.978365470435071e-05,
+      "loss": 3.3581,
+      "step": 16888
+    },
+    {
+      "epoch": 0.7111157894736843,
+      "grad_norm": 0.453125,
+      "learning_rate": 9.975668257316886e-05,
+      "loss": 2.9859,
+      "step": 16889
+    },
+    {
+      "epoch": 0.7111578947368421,
+      "grad_norm": 0.443359375,
+      "learning_rate": 9.972971317924374e-05,
+      "loss": 2.9541,
+      "step": 16890
+    },
+    {
+      "epoch": 0.7112,
+      "grad_norm": 0.421875,
+      "learning_rate": 9.970274652306666e-05,
+      "loss": 2.8482,
+      "step": 16891
+    },
+    {
+      "epoch": 0.7112421052631579,
+      "grad_norm": 0.4453125,
+      "learning_rate": 9.967578260512897e-05,
+      "loss": 3.0412,
+      "step": 16892
+    },
+    {
+      "epoch": 0.7112842105263157,
+      "grad_norm": 0.439453125,
+      "learning_rate": 9.964882142592188e-05,
+      "loss": 3.1506,
+      "step": 16893
+    },
+    {
+      "epoch": 0.7113263157894737,
+      "grad_norm": 0.43359375,
+      "learning_rate": 9.962186298593654e-05,
+      "loss": 3.5694,
+      "step": 16894
+    },
+    {
+      "epoch": 0.7113684210526315,
+      "grad_norm": 0.4375,
+      "learning_rate": 9.959490728566433e-05,
+      "loss": 3.4042,
+      "step": 16895
+    },
+    {
+      "epoch": 0.7114105263157895,
+      "grad_norm": 0.404296875,
+      "learning_rate": 9.956795432559601e-05,
+      "loss": 3.2916,
+      "step": 16896
+    },
+    {
+      "epoch": 0.7114526315789473,
+      "grad_norm": 0.4140625,
+      "learning_rate": 9.95410041062229e-05,
+      "loss": 3.0627,
+      "step": 16897
+    },
+    {
+      "epoch": 0.7114947368421053,
+      "grad_norm": 0.4296875,
+      "learning_rate": 9.951405662803592e-05,
+      "loss": 3.3171,
+      "step": 16898
+    },
+    {
+      "epoch": 0.7115368421052631,
+      "grad_norm": 0.443359375,
+      "learning_rate": 9.948711189152601e-05,
+      "loss": 3.1674,
+      "step": 16899
+    },
+    {
+      "epoch": 0.7115789473684211,
+      "grad_norm": 0.41796875,
+      "learning_rate": 9.946016989718409e-05,
+      "loss": 2.9179,
+      "step": 16900
+    },
+    {
+      "epoch": 0.7116210526315789,
+      "grad_norm": 0.44921875,
+      "learning_rate": 9.943323064550089e-05,
+      "loss": 2.9274,
+      "step": 16901
+    },
+    {
+      "epoch": 0.7116631578947369,
+      "grad_norm": 0.4921875,
+      "learning_rate": 9.940629413696753e-05,
+      "loss": 3.222,
+      "step": 16902
+    },
+    {
+      "epoch": 0.7117052631578947,
+      "grad_norm": 0.423828125,
+      "learning_rate": 9.937936037207434e-05,
+      "loss": 3.0021,
+      "step": 16903
+    },
+    {
+      "epoch": 0.7117473684210527,
+      "grad_norm": 0.451171875,
+      "learning_rate": 9.935242935131234e-05,
+      "loss": 3.306,
+      "step": 16904
+    },
+    {
+      "epoch": 0.7117894736842105,
+      "grad_norm": 0.4375,
+      "learning_rate": 9.932550107517205e-05,
+      "loss": 3.1326,
+      "step": 16905
+    },
+    {
+      "epoch": 0.7118315789473684,
+      "grad_norm": 0.419921875,
+      "learning_rate": 9.929857554414409e-05,
+      "loss": 3.1925,
+      "step": 16906
+    },
+    {
+      "epoch": 0.7118736842105263,
+      "grad_norm": 0.431640625,
+      "learning_rate": 9.927165275871904e-05,
+      "loss": 2.9996,
+      "step": 16907
+    },
+    {
+      "epoch": 0.7119157894736842,
+      "grad_norm": 0.412109375,
+      "learning_rate": 9.924473271938731e-05,
+      "loss": 3.0013,
+      "step": 16908
+    },
+    {
+      "epoch": 0.7119578947368421,
+      "grad_norm": 0.4296875,
+      "learning_rate": 9.921781542663955e-05,
+      "loss": 3.5803,
+      "step": 16909
+    },
+    {
+      "epoch": 0.712,
+      "grad_norm": 0.419921875,
+      "learning_rate": 9.919090088096589e-05,
+      "loss": 2.9812,
+      "step": 16910
+    },
+    {
+      "epoch": 0.7120421052631579,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.916398908285698e-05,
+      "loss": 3.3846,
+      "step": 16911
+    },
+    {
+      "epoch": 0.7120842105263158,
+      "grad_norm": 0.4375,
+      "learning_rate": 9.913708003280278e-05,
+      "loss": 3.0663,
+      "step": 16912
+    },
+    {
+      "epoch": 0.7121263157894737,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.911017373129383e-05,
+      "loss": 2.9371,
+      "step": 16913
+    },
+    {
+      "epoch": 0.7121684210526316,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.908327017882021e-05,
+      "loss": 2.9263,
+      "step": 16914
+    },
+    {
+      "epoch": 0.7122105263157895,
+      "grad_norm": 0.4296875,
+      "learning_rate": 9.905636937587203e-05,
+      "loss": 2.782,
+      "step": 16915
+    },
+    {
+      "epoch": 0.7122526315789474,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.90294713229396e-05,
+      "loss": 3.0034,
+      "step": 16916
+    },
+    {
+      "epoch": 0.7122947368421053,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.900257602051266e-05,
+      "loss": 3.2003,
+      "step": 16917
+    },
+    {
+      "epoch": 0.7123368421052632,
+      "grad_norm": 0.400390625,
+      "learning_rate": 9.897568346908154e-05,
+      "loss": 2.9292,
+      "step": 16918
+    },
+    {
+      "epoch": 0.712378947368421,
+      "grad_norm": 0.474609375,
+      "learning_rate": 9.894879366913581e-05,
+      "loss": 2.8176,
+      "step": 16919
+    },
+    {
+      "epoch": 0.712421052631579,
+      "grad_norm": 0.408203125,
+      "learning_rate": 9.892190662116571e-05,
+      "loss": 2.5378,
+      "step": 16920
+    },
+    {
+      "epoch": 0.7124631578947368,
+      "grad_norm": 0.45703125,
+      "learning_rate": 9.889502232566094e-05,
+      "loss": 2.9101,
+      "step": 16921
+    },
+    {
+      "epoch": 0.7125052631578948,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.886814078311135e-05,
+      "loss": 3.0179,
+      "step": 16922
+    },
+    {
+      "epoch": 0.7125473684210526,
+      "grad_norm": 0.421875,
+      "learning_rate": 9.884126199400665e-05,
+      "loss": 2.9971,
+      "step": 16923
+    },
+    {
+      "epoch": 0.7125894736842106,
+      "grad_norm": 0.404296875,
+      "learning_rate": 9.881438595883646e-05,
+      "loss": 2.6038,
+      "step": 16924
+    },
+    {
+      "epoch": 0.7126315789473684,
+      "grad_norm": 0.4140625,
+      "learning_rate": 9.87875126780907e-05,
+      "loss": 2.7271,
+      "step": 16925
+    },
+    {
+      "epoch": 0.7126736842105263,
+      "grad_norm": 0.447265625,
+      "learning_rate": 9.876064215225861e-05,
+      "loss": 2.7675,
+      "step": 16926
+    },
+    {
+      "epoch": 0.7127157894736842,
+      "grad_norm": 0.4375,
+      "learning_rate": 9.873377438183001e-05,
+      "loss": 3.4314,
+      "step": 16927
+    },
+    {
+      "epoch": 0.7127578947368421,
+      "grad_norm": 0.40625,
+      "learning_rate": 9.870690936729432e-05,
+      "loss": 3.6537,
+      "step": 16928
+    },
+    {
+      "epoch": 0.7128,
+      "grad_norm": 0.421875,
+      "learning_rate": 9.868004710914097e-05,
+      "loss": 3.0912,
+      "step": 16929
+    },
+    {
+      "epoch": 0.7128421052631579,
+      "grad_norm": 0.4296875,
+      "learning_rate": 9.865318760785937e-05,
+      "loss": 3.1416,
+      "step": 16930
+    },
+    {
+      "epoch": 0.7128842105263158,
+      "grad_norm": 0.4140625,
+      "learning_rate": 9.862633086393877e-05,
+      "loss": 3.2296,
+      "step": 16931
+    },
+    {
+      "epoch": 0.7129263157894736,
+      "grad_norm": 0.439453125,
+      "learning_rate": 9.859947687786877e-05,
+      "loss": 3.1374,
+      "step": 16932
+    },
+    {
+      "epoch": 0.7129684210526316,
+      "grad_norm": 0.4453125,
+      "learning_rate": 9.857262565013819e-05,
+      "loss": 3.3868,
+      "step": 16933
+    },
+    {
+      "epoch": 0.7130105263157894,
+      "grad_norm": 0.4296875,
+      "learning_rate": 9.854577718123659e-05,
+      "loss": 3.0208,
+      "step": 16934
+    },
+    {
+      "epoch": 0.7130526315789474,
+      "grad_norm": 0.4375,
+      "learning_rate": 9.851893147165295e-05,
+      "loss": 3.3971,
+      "step": 16935
+    },
+    {
+      "epoch": 0.7130947368421052,
+      "grad_norm": 0.4140625,
+      "learning_rate": 9.849208852187638e-05,
+      "loss": 3.0257,
+      "step": 16936
+    },
+    {
+      "epoch": 0.7131368421052632,
+      "grad_norm": 0.421875,
+      "learning_rate": 9.846524833239597e-05,
+      "loss": 3.1356,
+      "step": 16937
+    },
+    {
+      "epoch": 0.713178947368421,
+      "grad_norm": 0.41796875,
+      "learning_rate": 9.843841090370071e-05,
+      "loss": 2.7626,
+      "step": 16938
+    },
+    {
+      "epoch": 0.713221052631579,
+      "grad_norm": 0.400390625,
+      "learning_rate": 9.841157623627947e-05,
+      "loss": 2.4258,
+      "step": 16939
+    },
+    {
+      "epoch": 0.7132631578947368,
+      "grad_norm": 0.435546875,
+      "learning_rate": 9.838474433062114e-05,
+      "loss": 3.1583,
+      "step": 16940
+    },
+    {
+      "epoch": 0.7133052631578948,
+      "grad_norm": 0.4375,
+      "learning_rate": 9.835791518721474e-05,
+      "loss": 2.873,
+      "step": 16941
+    },
+    {
+      "epoch": 0.7133473684210526,
+      "grad_norm": 0.455078125,
+      "learning_rate": 9.833108880654892e-05,
+      "loss": 3.0108,
+      "step": 16942
+    },
+    {
+      "epoch": 0.7133894736842106,
+      "grad_norm": 0.462890625,
+      "learning_rate": 9.830426518911248e-05,
+      "loss": 2.9234,
+      "step": 16943
+    },
+    {
+      "epoch": 0.7134315789473684,
+      "grad_norm": 0.466796875,
+      "learning_rate": 9.827744433539406e-05,
+      "loss": 2.8782,
+      "step": 16944
+    },
+    {
+      "epoch": 0.7134736842105264,
+      "grad_norm": 0.419921875,
+      "learning_rate": 9.825062624588238e-05,
+      "loss": 3.3175,
+      "step": 16945
+    },
+    {
+      "epoch": 0.7135157894736842,
+      "grad_norm": 0.439453125,
+      "learning_rate": 9.822381092106597e-05,
+      "loss": 3.2667,
+      "step": 16946
+    },
+    {
+      "epoch": 0.7135578947368421,
+      "grad_norm": 0.421875,
+      "learning_rate": 9.819699836143331e-05,
+      "loss": 2.9353,
+      "step": 16947
+    },
+    {
+      "epoch": 0.7136,
+      "grad_norm": 0.43359375,
+      "learning_rate": 9.817018856747307e-05,
+      "loss": 2.937,
+      "step": 16948
+    },
+    {
+      "epoch": 0.7136421052631579,
+      "grad_norm": 0.43359375,
+      "learning_rate": 9.814338153967359e-05,
+      "loss": 3.0775,
+      "step": 16949
+    },
+    {
+      "epoch": 0.7136842105263158,
+      "grad_norm": 0.458984375,
+      "learning_rate": 9.811657727852328e-05,
+      "loss": 3.2195,
+      "step": 16950
+    },
+    {
+      "epoch": 0.7137263157894737,
+      "grad_norm": 0.4453125,
+      "learning_rate": 9.808977578451047e-05,
+      "loss": 3.1562,
+      "step": 16951
+    },
+    {
+      "epoch": 0.7137684210526316,
+      "grad_norm": 0.439453125,
+      "learning_rate": 9.806297705812348e-05,
+      "loss": 2.9758,
+      "step": 16952
+    },
+    {
+      "epoch": 0.7138105263157895,
+      "grad_norm": 0.48046875,
+      "learning_rate": 9.80361810998505e-05,
+      "loss": 2.7876,
+      "step": 16953
+    },
+    {
+      "epoch": 0.7138526315789474,
+      "grad_norm": 0.4140625,
+      "learning_rate": 9.800938791017977e-05,
+      "loss": 3.2477,
+      "step": 16954
+    },
+    {
+      "epoch": 0.7138947368421052,
+      "grad_norm": 0.458984375,
+      "learning_rate": 9.798259748959931e-05,
+      "loss": 2.933,
+      "step": 16955
+    },
+    {
+      "epoch": 0.7139368421052632,
+      "grad_norm": 0.419921875,
+      "learning_rate": 9.795580983859742e-05,
+      "loss": 3.3745,
+      "step": 16956
+    },
+    {
+      "epoch": 0.713978947368421,
+      "grad_norm": 0.43359375,
+      "learning_rate": 9.792902495766201e-05,
+      "loss": 3.1138,
+      "step": 16957
+    },
+    {
+      "epoch": 0.714021052631579,
+      "grad_norm": 0.4140625,
+      "learning_rate": 9.790224284728107e-05,
+      "loss": 2.6987,
+      "step": 16958
+    },
+    {
+      "epoch": 0.7140631578947368,
+      "grad_norm": 0.46484375,
+      "learning_rate": 9.78754635079426e-05,
+      "loss": 3.5947,
+      "step": 16959
+    },
+    {
+      "epoch": 0.7141052631578947,
+      "grad_norm": 0.400390625,
+      "learning_rate": 9.784868694013443e-05,
+      "loss": 2.6584,
+      "step": 16960
+    },
+    {
+      "epoch": 0.7141473684210526,
+      "grad_norm": 0.427734375,
+      "learning_rate": 9.782191314434441e-05,
+      "loss": 2.9984,
+      "step": 16961
+    },
+    {
+      "epoch": 0.7141894736842105,
+      "grad_norm": 0.443359375,
+      "learning_rate": 9.779514212106025e-05,
+      "loss": 3.096,
+      "step": 16962
+    },
+    {
+      "epoch": 0.7142315789473684,
+      "grad_norm": 0.421875,
+      "learning_rate": 9.776837387076986e-05,
+      "loss": 3.4277,
+      "step": 16963
+    },
+    {
+      "epoch": 0.7142736842105263,
+      "grad_norm": 0.4296875,
+      "learning_rate": 9.774160839396082e-05,
+      "loss": 2.7251,
+      "step": 16964
+    },
+    {
+      "epoch": 0.7143157894736842,
+      "grad_norm": 0.41796875,
+      "learning_rate": 9.771484569112079e-05,
+      "loss": 3.1881,
+      "step": 16965
+    },
+    {
+      "epoch": 0.7143578947368421,
+      "grad_norm": 0.4140625,
+      "learning_rate": 9.768808576273733e-05,
+      "loss": 3.2741,
+      "step": 16966
+    },
+    {
+      "epoch": 0.7144,
+      "grad_norm": 0.419921875,
+      "learning_rate": 9.766132860929799e-05,
+      "loss": 3.158,
+      "step": 16967
+    },
+    {
+      "epoch": 0.7144421052631579,
+      "grad_norm": 0.462890625,
+      "learning_rate": 9.763457423129027e-05,
+      "loss": 2.6538,
+      "step": 16968
+    },
+    {
+      "epoch": 0.7144842105263158,
+      "grad_norm": 0.4375,
+      "learning_rate": 9.760782262920148e-05,
+      "loss": 3.0077,
+      "step": 16969
+    },
+    {
+      "epoch": 0.7145263157894737,
+      "grad_norm": 0.4296875,
+      "learning_rate": 9.75810738035193e-05,
+      "loss": 3.1471,
+      "step": 16970
+    },
+    {
+      "epoch": 0.7145684210526316,
+      "grad_norm": 0.458984375,
+      "learning_rate": 9.755432775473067e-05,
+      "loss": 3.1062,
+      "step": 16971
+    },
+    {
+      "epoch": 0.7146105263157895,
+      "grad_norm": 0.408203125,
+      "learning_rate": 9.752758448332317e-05,
+      "loss": 2.8726,
+      "step": 16972
+    },
+    {
+      "epoch": 0.7146526315789473,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.750084398978393e-05,
+      "loss": 3.1971,
+      "step": 16973
+    },
+    {
+      "epoch": 0.7146947368421053,
+      "grad_norm": 0.470703125,
+      "learning_rate": 9.747410627460016e-05,
+      "loss": 3.1966,
+      "step": 16974
+    },
+    {
+      "epoch": 0.7147368421052631,
+      "grad_norm": 0.423828125,
+      "learning_rate": 9.744737133825893e-05,
+      "loss": 2.4889,
+      "step": 16975
+    },
+    {
+      "epoch": 0.7147789473684211,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.742063918124728e-05,
+      "loss": 3.0717,
+      "step": 16976
+    },
+    {
+      "epoch": 0.7148210526315789,
+      "grad_norm": 0.4296875,
+      "learning_rate": 9.73939098040525e-05,
+      "loss": 3.141,
+      "step": 16977
+    },
+    {
+      "epoch": 0.7148631578947369,
+      "grad_norm": 0.451171875,
+      "learning_rate": 9.736718320716117e-05,
+      "loss": 3.0969,
+      "step": 16978
+    },
+    {
+      "epoch": 0.7149052631578947,
+      "grad_norm": 0.4453125,
+      "learning_rate": 9.734045939106054e-05,
+      "loss": 2.9496,
+      "step": 16979
+    },
+    {
+      "epoch": 0.7149473684210527,
+      "grad_norm": 0.45703125,
+      "learning_rate": 9.731373835623735e-05,
+      "loss": 3.3816,
+      "step": 16980
+    },
+    {
+      "epoch": 0.7149894736842105,
+      "grad_norm": 0.41015625,
+      "learning_rate": 9.728702010317847e-05,
+      "loss": 3.1532,
+      "step": 16981
+    },
+    {
+      "epoch": 0.7150315789473685,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.726030463237065e-05,
+      "loss": 3.2694,
+      "step": 16982
+    },
+    {
+      "epoch": 0.7150736842105263,
+      "grad_norm": 0.41796875,
+      "learning_rate": 9.723359194430051e-05,
+      "loss": 3.2479,
+      "step": 16983
+    },
+    {
+      "epoch": 0.7151157894736843,
+      "grad_norm": 0.41796875,
+      "learning_rate": 9.720688203945501e-05,
+      "loss": 3.1115,
+      "step": 16984
+    },
+    {
+      "epoch": 0.7151578947368421,
+      "grad_norm": 0.451171875,
+      "learning_rate": 9.718017491832046e-05,
+      "loss": 3.0338,
+      "step": 16985
+    },
+    {
+      "epoch": 0.7152,
+      "grad_norm": 0.443359375,
+      "learning_rate": 9.715347058138371e-05,
+      "loss": 3.0961,
+      "step": 16986
+    },
+    {
+      "epoch": 0.7152421052631579,
+      "grad_norm": 0.388671875,
+      "learning_rate": 9.712676902913095e-05,
+      "loss": 2.9653,
+      "step": 16987
+    },
+    {
+      "epoch": 0.7152842105263157,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.710007026204895e-05,
+      "loss": 2.9415,
+      "step": 16988
+    },
+    {
+      "epoch": 0.7153263157894737,
+      "grad_norm": 0.4296875,
+      "learning_rate": 9.707337428062401e-05,
+      "loss": 2.3012,
+      "step": 16989
+    },
+    {
+      "epoch": 0.7153684210526315,
+      "grad_norm": 0.4609375,
+      "learning_rate": 9.704668108534243e-05,
+      "loss": 3.1839,
+      "step": 16990
+    },
+    {
+      "epoch": 0.7154105263157895,
+      "grad_norm": 0.435546875,
+      "learning_rate": 9.701999067669079e-05,
+      "loss": 2.948,
+      "step": 16991
+    },
+    {
+      "epoch": 0.7154526315789473,
+      "grad_norm": 0.44140625,
+      "learning_rate": 9.699330305515497e-05,
+      "loss": 3.2594,
+      "step": 16992
+    },
+    {
+      "epoch": 0.7154947368421053,
+      "grad_norm": 0.4140625,
+      "learning_rate": 9.69666182212216e-05,
+      "loss": 3.2303,
+      "step": 16993
+    },
+    {
+      "epoch": 0.7155368421052631,
+      "grad_norm": 0.419921875,
+      "learning_rate": 9.693993617537644e-05,
+      "loss": 3.3469,
+      "step": 16994
+    },
+    {
+      "epoch": 0.7155789473684211,
+      "grad_norm": 0.4453125,
+      "learning_rate": 9.69132569181059e-05,
+      "loss": 3.1142,
+      "step": 16995
+    },
+    {
+      "epoch": 0.7156210526315789,
+      "grad_norm": 0.439453125,
+      "learning_rate": 9.688658044989596e-05,
+      "loss": 3.2032,
+      "step": 16996
+    },
+    {
+      "epoch": 0.7156631578947369,
+      "grad_norm": 0.41015625,
+      "learning_rate": 9.685990677123263e-05,
+      "loss": 3.1566,
+      "step": 16997
+    },
+    {
+      "epoch": 0.7157052631578947,
+      "grad_norm": 0.41796875,
+      "learning_rate": 9.683323588260185e-05,
+      "loss": 3.0318,
+      "step": 16998
+    },
+    {
+      "epoch": 0.7157473684210526,
+      "grad_norm": 0.43359375,
+      "learning_rate": 9.680656778448946e-05,
+      "loss": 3.6259,
+      "step": 16999
+    },
+    {
+      "epoch": 0.7157894736842105,
+      "grad_norm": 0.416015625,
+      "learning_rate": 9.677990247738158e-05,
+      "loss": 3.1723,
+      "step": 17000
+    },
+    {
+      "epoch": 0.7158315789473684,
+      "grad_norm": 0.431640625,
+      "learning_rate": 9.675323996176369e-05,
+      "loss": 2.608,
+      "step": 17001
+    },
+    {
+      "epoch": 0.7158736842105263,
+      "grad_norm": 0.412109375,
+      "learning_rate": 9.672658023812179e-05,
+      "loss": 3.4809,
+      "step": 17002
+    },
+    {
+      "epoch": 0.7159157894736842,
+      "grad_norm": 0.43359375,
+      "learning_rate": 9.669992330694153e-05,
+      "loss": 3.2492,
+      "step": 17003
+    },
+    {
+      "epoch": 0.7159578947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 9.66732691687085e-05,
+      "loss": 2.7765,
+      "step": 17004
+    },
+    {
+      "epoch": 0.716,
+      "grad_norm": 0.44140625,
+      "learning_rate": 9.66466178239084e-05,
+      "loss": 3.43,
+      "step": 17005
+    },
+    {
+      "epoch": 0.7160421052631579,
+      "grad_norm": 0.4140625,
+      "learning_rate": 9.661996927302663e-05,
+      "loss": 2.7444,
+      "step": 17006
+    },
+    {
+      "epoch": 0.7160842105263158,
+      "grad_norm": 0.41796875,
+      "learning_rate": 9.659332351654899e-05,
+      "loss": 2.959,
+      "step": 17007
+    },
+    {
+      "epoch": 0.7161263157894737,
+      "grad_norm": 0.4296875,
+      "learning_rate": 9.656668055496057e-05,
+      "loss": 3.1073,
+      "step": 17008
+    },
+    {
+      "epoch": 0.7161684210526316,
+      "grad_norm": 0.421875,
+      "learning_rate": 9.654004038874703e-05,
+      "loss": 3.2833,
+      "step": 17009
+    },
+    {
+      "epoch": 0.7162105263157895,
+      "grad_norm": 0.43359375,
+      "learning_rate": 9.651340301839367e-05,
+      "loss": 3.2678,
+      "step": 17010
+    },
+    {
+      "epoch": 0.7162526315789474,
+      "grad_norm": 0.419921875,
+      "learning_rate": 9.648676844438576e-05,
+      "loss": 3.0995,
+      "step": 17011
+    },
+    {
+      "epoch": 0.7162947368421053,
+      "grad_norm": 0.4453125,
+      "learning_rate": 9.646013666720854e-05,
+      "loss": 3.4039,
+      "step": 17012
+    },
+    {
+      "epoch": 0.7163368421052632,
+      "grad_norm": 0.4609375,
+      "learning_rate": 9.643350768734727e-05,
+      "loss": 3.4853,
+      "step": 17013
+    },
+    {
+      "epoch": 0.716378947368421,
+      "grad_norm": 0.421875,
+      "learning_rate": 9.6406881505287e-05,
+      "loss": 3.2767,
+      "step": 17014
+    },
+    {
+      "epoch": 0.716421052631579,
+      "grad_norm": 0.41796875,
+      "learning_rate": 9.638025812151282e-05,
+      "loss": 3.3066,
+      "step": 17015
+    },
+    {
+      "epoch": 0.7164631578947368,
+      "grad_norm": 0.392578125,
+      "learning_rate": 9.635363753650994e-05,
+      "loss": 2.878,
+      "step": 17016
+    },
+    {
+      "epoch": 0.7165052631578948,
+      "grad_norm": 0.421875,
+      "learning_rate": 9.632701975076327e-05,
+      "loss": 3.0943,
+      "step": 17017
+    },
+    {
+      "epoch": 0.7165473684210526,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.63004047647577e-05,
+      "loss": 3.1208,
+      "step": 17018
+    },
+    {
+      "epoch": 0.7165894736842106,
+      "grad_norm": 0.423828125,
+      "learning_rate": 9.627379257897823e-05,
+      "loss": 3.2549,
+      "step": 17019
+    },
+    {
+      "epoch": 0.7166315789473684,
+      "grad_norm": 0.4296875,
+      "learning_rate": 9.624718319390957e-05,
+      "loss": 3.2786,
+      "step": 17020
+    },
+    {
+      "epoch": 0.7166736842105264,
+      "grad_norm": 0.439453125,
+      "learning_rate": 9.622057661003661e-05,
+      "loss": 3.5141,
+      "step": 17021
+    },
+    {
+      "epoch": 0.7167157894736842,
+      "grad_norm": 0.4375,
+      "learning_rate": 9.619397282784398e-05,
+      "loss": 2.9217,
+      "step": 17022
+    },
+    {
+      "epoch": 0.7167578947368421,
+      "grad_norm": 0.4609375,
+      "learning_rate": 9.616737184781654e-05,
+      "loss": 3.0516,
+      "step": 17023
+    },
+    {
+      "epoch": 0.7168,
+      "grad_norm": 0.44140625,
+      "learning_rate": 9.614077367043883e-05,
+      "loss": 3.3174,
+      "step": 17024
+    },
+    {
+      "epoch": 0.716842105263158,
+      "grad_norm": 0.44140625,
+      "learning_rate": 9.611417829619543e-05,
+      "loss": 2.8571,
+      "step": 17025
+    },
+    {
+      "epoch": 0.7168842105263158,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.608758572557091e-05,
+      "loss": 3.3848,
+      "step": 17026
+    },
+    {
+      "epoch": 0.7169263157894736,
+      "grad_norm": 0.421875,
+      "learning_rate": 9.606099595904974e-05,
+      "loss": 3.1613,
+      "step": 17027
+    },
+    {
+      "epoch": 0.7169684210526316,
+      "grad_norm": 0.4453125,
+      "learning_rate": 9.603440899711627e-05,
+      "loss": 2.915,
+      "step": 17028
+    },
+    {
+      "epoch": 0.7170105263157894,
+      "grad_norm": 0.4296875,
+      "learning_rate": 9.600782484025513e-05,
+      "loss": 3.3253,
+      "step": 17029
+    },
+    {
+      "epoch": 0.7170526315789474,
+      "grad_norm": 0.40625,
+      "learning_rate": 9.598124348895032e-05,
+      "loss": 2.6367,
+      "step": 17030
+    },
+    {
+      "epoch": 0.7170947368421052,
+      "grad_norm": 0.421875,
+      "learning_rate": 9.595466494368634e-05,
+      "loss": 3.4796,
+      "step": 17031
+    },
+    {
+      "epoch": 0.7171368421052632,
+      "grad_norm": 0.44140625,
+      "learning_rate": 9.592808920494742e-05,
+      "loss": 3.4416,
+      "step": 17032
+    },
+    {
+      "epoch": 0.717178947368421,
+      "grad_norm": 0.4296875,
+      "learning_rate": 9.590151627321766e-05,
+      "loss": 2.9603,
+      "step": 17033
+    },
+    {
+      "epoch": 0.717221052631579,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.587494614898124e-05,
+      "loss": 2.5606,
+      "step": 17034
+    },
+    {
+      "epoch": 0.7172631578947368,
+      "grad_norm": 0.423828125,
+      "learning_rate": 9.58483788327221e-05,
+      "loss": 2.9297,
+      "step": 17035
+    },
+    {
+      "epoch": 0.7173052631578948,
+      "grad_norm": 0.4140625,
+      "learning_rate": 9.582181432492457e-05,
+      "loss": 3.0156,
+      "step": 17036
+    },
+    {
+      "epoch": 0.7173473684210526,
+      "grad_norm": 0.4921875,
+      "learning_rate": 9.579525262607224e-05,
+      "loss": 2.9094,
+      "step": 17037
+    },
+    {
+      "epoch": 0.7173894736842106,
+      "grad_norm": 0.4140625,
+      "learning_rate": 9.576869373664931e-05,
+      "loss": 2.6853,
+      "step": 17038
+    },
+    {
+      "epoch": 0.7174315789473684,
+      "grad_norm": 0.4140625,
+      "learning_rate": 9.574213765713957e-05,
+      "loss": 3.0654,
+      "step": 17039
+    },
+    {
+      "epoch": 0.7174736842105263,
+      "grad_norm": 0.4609375,
+      "learning_rate": 9.571558438802685e-05,
+      "loss": 3.0479,
+      "step": 17040
+    },
+    {
+      "epoch": 0.7175157894736842,
+      "grad_norm": 0.421875,
+      "learning_rate": 9.568903392979488e-05,
+      "loss": 3.2583,
+      "step": 17041
+    },
+    {
+      "epoch": 0.7175578947368421,
+      "grad_norm": 0.435546875,
+      "learning_rate": 9.566248628292734e-05,
+      "loss": 3.2792,
+      "step": 17042
+    },
+    {
+      "epoch": 0.7176,
+      "grad_norm": 0.4453125,
+      "learning_rate": 9.563594144790813e-05,
+      "loss": 2.8641,
+      "step": 17043
+    },
+    {
+      "epoch": 0.7176421052631579,
+      "grad_norm": 0.44921875,
+      "learning_rate": 9.560939942522055e-05,
+      "loss": 3.2026,
+      "step": 17044
+    },
+    {
+      "epoch": 0.7176842105263158,
+      "grad_norm": 0.4375,
+      "learning_rate": 9.558286021534842e-05,
+      "loss": 3.4816,
+      "step": 17045
+    },
+    {
+      "epoch": 0.7177263157894737,
+      "grad_norm": 0.50390625,
+      "learning_rate": 9.555632381877502e-05,
+      "loss": 2.7224,
+      "step": 17046
+    },
+    {
+      "epoch": 0.7177684210526316,
+      "grad_norm": 0.453125,
+      "learning_rate": 9.552979023598402e-05,
+      "loss": 3.4709,
+      "step": 17047
+    },
+    {
+      "epoch": 0.7178105263157895,
+      "grad_norm": 0.419921875,
+      "learning_rate": 9.550325946745872e-05,
+      "loss": 3.1153,
+      "step": 17048
+    },
+    {
+      "epoch": 0.7178526315789474,
+      "grad_norm": 0.40234375,
+      "learning_rate": 9.547673151368244e-05,
+      "loss": 3.1406,
+      "step": 17049
+    },
+    {
+      "epoch": 0.7178947368421053,
+      "grad_norm": 0.447265625,
+      "learning_rate": 9.545020637513871e-05,
+      "loss": 2.9752,
+      "step": 17050
+    },
+    {
+      "epoch": 0.7179368421052632,
+      "grad_norm": 0.447265625,
+      "learning_rate": 9.542368405231047e-05,
+      "loss": 3.0054,
+      "step": 17051
+    },
+    {
+      "epoch": 0.717978947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.539716454568129e-05,
+      "loss": 3.3265,
+      "step": 17052
+    },
+    {
+      "epoch": 0.7180210526315789,
+      "grad_norm": 0.431640625,
+      "learning_rate": 9.53706478557339e-05,
+      "loss": 3.4637,
+      "step": 17053
+    },
+    {
+      "epoch": 0.7180631578947368,
+      "grad_norm": 0.416015625,
+      "learning_rate": 9.534413398295184e-05,
+      "loss": 3.163,
+      "step": 17054
+    },
+    {
+      "epoch": 0.7181052631578947,
+      "grad_norm": 0.4296875,
+      "learning_rate": 9.531762292781774e-05,
+      "loss": 2.8401,
+      "step": 17055
+    },
+    {
+      "epoch": 0.7181473684210526,
+      "grad_norm": 0.44140625,
+      "learning_rate": 9.529111469081489e-05,
+      "loss": 2.6626,
+      "step": 17056
+    },
+    {
+      "epoch": 0.7181894736842105,
+      "grad_norm": 0.41015625,
+      "learning_rate": 9.526460927242617e-05,
+      "loss": 3.2765,
+      "step": 17057
+    },
+    {
+      "epoch": 0.7182315789473684,
+      "grad_norm": 0.400390625,
+      "learning_rate": 9.523810667313437e-05,
+      "loss": 2.6277,
+      "step": 17058
+    },
+    {
+      "epoch": 0.7182736842105263,
+      "grad_norm": 0.427734375,
+      "learning_rate": 9.52116068934226e-05,
+      "loss": 3.2643,
+      "step": 17059
+    },
+    {
+      "epoch": 0.7183157894736842,
+      "grad_norm": 0.443359375,
+      "learning_rate": 9.51851099337733e-05,
+      "loss": 2.7817,
+      "step": 17060
+    },
+    {
+      "epoch": 0.7183578947368421,
+      "grad_norm": 0.4140625,
+      "learning_rate": 9.515861579466961e-05,
+      "loss": 3.1166,
+      "step": 17061
+    },
+    {
+      "epoch": 0.7184,
+      "grad_norm": 0.40234375,
+      "learning_rate": 9.513212447659381e-05,
+      "loss": 3.1691,
+      "step": 17062
+    },
+    {
+      "epoch": 0.7184421052631579,
+      "grad_norm": 0.416015625,
+      "learning_rate": 9.51056359800288e-05,
+      "loss": 2.6046,
+      "step": 17063
+    },
+    {
+      "epoch": 0.7184842105263158,
+      "grad_norm": 0.44140625,
+      "learning_rate": 9.507915030545716e-05,
+      "loss": 3.1064,
+      "step": 17064
+    },
+    {
+      "epoch": 0.7185263157894737,
+      "grad_norm": 0.46875,
+      "learning_rate": 9.505266745336124e-05,
+      "loss": 3.3288,
+      "step": 17065
+    },
+    {
+      "epoch": 0.7185684210526316,
+      "grad_norm": 0.412109375,
+      "learning_rate": 9.50261874242239e-05,
+      "loss": 3.1016,
+      "step": 17066
+    },
+    {
+      "epoch": 0.7186105263157895,
+      "grad_norm": 0.44140625,
+      "learning_rate": 9.49997102185271e-05,
+      "loss": 2.8839,
+      "step": 17067
+    },
+    {
+      "epoch": 0.7186526315789473,
+      "grad_norm": 0.423828125,
+      "learning_rate": 9.497323583675366e-05,
+      "loss": 2.9743,
+      "step": 17068
+    },
+    {
+      "epoch": 0.7186947368421053,
+      "grad_norm": 0.435546875,
+      "learning_rate": 9.494676427938556e-05,
+      "loss": 3.0364,
+      "step": 17069
+    },
+    {
+      "epoch": 0.7187368421052631,
+      "grad_norm": 0.443359375,
+      "learning_rate": 9.492029554690529e-05,
+      "loss": 3.1935,
+      "step": 17070
+    },
+    {
+      "epoch": 0.7187789473684211,
+      "grad_norm": 0.435546875,
+      "learning_rate": 9.489382963979504e-05,
+      "loss": 3.477,
+      "step": 17071
+    },
+    {
+      "epoch": 0.7188210526315789,
+      "grad_norm": 0.4453125,
+      "learning_rate": 9.486736655853698e-05,
+      "loss": 3.2681,
+      "step": 17072
+    },
+    {
+      "epoch": 0.7188631578947369,
+      "grad_norm": 0.431640625,
+      "learning_rate": 9.484090630361322e-05,
+      "loss": 2.7928,
+      "step": 17073
+    },
+    {
+      "epoch": 0.7189052631578947,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.481444887550575e-05,
+      "loss": 3.314,
+      "step": 17074
+    },
+    {
+      "epoch": 0.7189473684210527,
+      "grad_norm": 0.421875,
+      "learning_rate": 9.478799427469686e-05,
+      "loss": 3.1794,
+      "step": 17075
+    },
+    {
+      "epoch": 0.7189894736842105,
+      "grad_norm": 0.43359375,
+      "learning_rate": 9.476154250166819e-05,
+      "loss": 3.1201,
+      "step": 17076
+    },
+    {
+      "epoch": 0.7190315789473685,
+      "grad_norm": 0.41796875,
+      "learning_rate": 9.473509355690188e-05,
+      "loss": 3.118,
+      "step": 17077
+    },
+    {
+      "epoch": 0.7190736842105263,
+      "grad_norm": 0.45703125,
+      "learning_rate": 9.470864744087976e-05,
+      "loss": 2.9984,
+      "step": 17078
+    },
+    {
+      "epoch": 0.7191157894736843,
+      "grad_norm": 0.435546875,
+      "learning_rate": 9.46822041540836e-05,
+      "loss": 2.9723,
+      "step": 17079
+    },
+    {
+      "epoch": 0.7191578947368421,
+      "grad_norm": 0.431640625,
+      "learning_rate": 9.465576369699522e-05,
+      "loss": 2.8231,
+      "step": 17080
+    },
+    {
+      "epoch": 0.7192,
+      "grad_norm": 0.40234375,
+      "learning_rate": 9.46293260700962e-05,
+      "loss": 2.7715,
+      "step": 17081
+    },
+    {
+      "epoch": 0.7192421052631579,
+      "grad_norm": 0.51171875,
+      "learning_rate": 9.46028912738684e-05,
+      "loss": 3.0794,
+      "step": 17082
+    },
+    {
+      "epoch": 0.7192842105263157,
+      "grad_norm": 0.435546875,
+      "learning_rate": 9.457645930879333e-05,
+      "loss": 3.0214,
+      "step": 17083
+    },
+    {
+      "epoch": 0.7193263157894737,
+      "grad_norm": 0.419921875,
+      "learning_rate": 9.455003017535255e-05,
+      "loss": 3.108,
+      "step": 17084
+    },
+    {
+      "epoch": 0.7193684210526315,
+      "grad_norm": 0.462890625,
+      "learning_rate": 9.452360387402755e-05,
+      "loss": 3.1912,
+      "step": 17085
+    },
+    {
+      "epoch": 0.7194105263157895,
+      "grad_norm": 0.443359375,
+      "learning_rate": 9.449718040529987e-05,
+      "loss": 2.9224,
+      "step": 17086
+    },
+    {
+      "epoch": 0.7194526315789473,
+      "grad_norm": 0.4375,
+      "learning_rate": 9.447075976965081e-05,
+      "loss": 3.287,
+      "step": 17087
+    },
+    {
+      "epoch": 0.7194947368421053,
+      "grad_norm": 0.44140625,
+      "learning_rate": 9.44443419675618e-05,
+      "loss": 3.1698,
+      "step": 17088
+    },
+    {
+      "epoch": 0.7195368421052631,
+      "grad_norm": 0.416015625,
+      "learning_rate": 9.441792699951399e-05,
+      "loss": 3.3985,
+      "step": 17089
+    },
+    {
+      "epoch": 0.7195789473684211,
+      "grad_norm": 0.45703125,
+      "learning_rate": 9.439151486598884e-05,
+      "loss": 3.7491,
+      "step": 17090
+    },
+    {
+      "epoch": 0.7196210526315789,
+      "grad_norm": 0.451171875,
+      "learning_rate": 9.436510556746747e-05,
+      "loss": 3.0031,
+      "step": 17091
+    },
+    {
+      "epoch": 0.7196631578947369,
+      "grad_norm": 0.45703125,
+      "learning_rate": 9.433869910443102e-05,
+      "loss": 3.1601,
+      "step": 17092
+    },
+    {
+      "epoch": 0.7197052631578947,
+      "grad_norm": 0.443359375,
+      "learning_rate": 9.431229547736056e-05,
+      "loss": 3.3263,
+      "step": 17093
+    },
+    {
+      "epoch": 0.7197473684210526,
+      "grad_norm": 0.466796875,
+      "learning_rate": 9.428589468673718e-05,
+      "loss": 3.1814,
+      "step": 17094
+    },
+    {
+      "epoch": 0.7197894736842105,
+      "grad_norm": 0.416015625,
+      "learning_rate": 9.425949673304179e-05,
+      "loss": 3.1573,
+      "step": 17095
+    },
+    {
+      "epoch": 0.7198315789473684,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.423310161675532e-05,
+      "loss": 2.5584,
+      "step": 17096
+    },
+    {
+      "epoch": 0.7198736842105263,
+      "grad_norm": 0.45703125,
+      "learning_rate": 9.420670933835892e-05,
+      "loss": 3.0186,
+      "step": 17097
+    },
+    {
+      "epoch": 0.7199157894736842,
+      "grad_norm": 0.439453125,
+      "learning_rate": 9.418031989833304e-05,
+      "loss": 2.8352,
+      "step": 17098
+    },
+    {
+      "epoch": 0.7199578947368421,
+      "grad_norm": 0.431640625,
+      "learning_rate": 9.415393329715872e-05,
+      "loss": 2.9753,
+      "step": 17099
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.451171875,
+      "learning_rate": 9.412754953531663e-05,
+      "loss": 3.0613,
+      "step": 17100
+    },
+    {
+      "epoch": 0.7200421052631579,
+      "grad_norm": 0.51171875,
+      "learning_rate": 9.410116861328746e-05,
+      "loss": 2.9335,
+      "step": 17101
+    },
+    {
+      "epoch": 0.7200842105263158,
+      "grad_norm": 0.412109375,
+      "learning_rate": 9.40747905315518e-05,
+      "loss": 2.9339,
+      "step": 17102
+    },
+    {
+      "epoch": 0.7201263157894737,
+      "grad_norm": 0.443359375,
+      "learning_rate": 9.404841529059018e-05,
+      "loss": 3.0512,
+      "step": 17103
+    },
+    {
+      "epoch": 0.7201684210526316,
+      "grad_norm": 0.4375,
+      "learning_rate": 9.402204289088337e-05,
+      "loss": 3.3844,
+      "step": 17104
+    },
+    {
+      "epoch": 0.7202105263157895,
+      "grad_norm": 0.4453125,
+      "learning_rate": 9.39956733329115e-05,
+      "loss": 3.4266,
+      "step": 17105
+    },
+    {
+      "epoch": 0.7202526315789474,
+      "grad_norm": 0.41796875,
+      "learning_rate": 9.396930661715524e-05,
+      "loss": 2.9153,
+      "step": 17106
+    },
+    {
+      "epoch": 0.7202947368421052,
+      "grad_norm": 0.44921875,
+      "learning_rate": 9.39429427440949e-05,
+      "loss": 3.3782,
+      "step": 17107
+    },
+    {
+      "epoch": 0.7203368421052632,
+      "grad_norm": 0.412109375,
+      "learning_rate": 9.391658171421078e-05,
+      "loss": 3.1637,
+      "step": 17108
+    },
+    {
+      "epoch": 0.720378947368421,
+      "grad_norm": 0.435546875,
+      "learning_rate": 9.389022352798315e-05,
+      "loss": 2.6935,
+      "step": 17109
+    },
+    {
+      "epoch": 0.720421052631579,
+      "grad_norm": 0.412109375,
+      "learning_rate": 9.386386818589212e-05,
+      "loss": 2.9736,
+      "step": 17110
+    },
+    {
+      "epoch": 0.7204631578947368,
+      "grad_norm": 0.44921875,
+      "learning_rate": 9.383751568841817e-05,
+      "loss": 3.0077,
+      "step": 17111
+    },
+    {
+      "epoch": 0.7205052631578948,
+      "grad_norm": 0.43359375,
+      "learning_rate": 9.381116603604101e-05,
+      "loss": 3.0732,
+      "step": 17112
+    },
+    {
+      "epoch": 0.7205473684210526,
+      "grad_norm": 0.44140625,
+      "learning_rate": 9.378481922924107e-05,
+      "loss": 3.6222,
+      "step": 17113
+    },
+    {
+      "epoch": 0.7205894736842106,
+      "grad_norm": 0.421875,
+      "learning_rate": 9.375847526849798e-05,
+      "loss": 3.396,
+      "step": 17114
+    },
+    {
+      "epoch": 0.7206315789473684,
+      "grad_norm": 0.427734375,
+      "learning_rate": 9.3732134154292e-05,
+      "loss": 2.8817,
+      "step": 17115
+    },
+    {
+      "epoch": 0.7206736842105264,
+      "grad_norm": 0.40625,
+      "learning_rate": 9.370579588710292e-05,
+      "loss": 3.5487,
+      "step": 17116
+    },
+    {
+      "epoch": 0.7207157894736842,
+      "grad_norm": 0.453125,
+      "learning_rate": 9.367946046741052e-05,
+      "loss": 2.7398,
+      "step": 17117
+    },
+    {
+      "epoch": 0.7207578947368422,
+      "grad_norm": 0.40625,
+      "learning_rate": 9.365312789569485e-05,
+      "loss": 3.0182,
+      "step": 17118
+    },
+    {
+      "epoch": 0.7208,
+      "grad_norm": 0.423828125,
+      "learning_rate": 9.362679817243533e-05,
+      "loss": 2.675,
+      "step": 17119
+    },
+    {
+      "epoch": 0.7208421052631578,
+      "grad_norm": 0.470703125,
+      "learning_rate": 9.360047129811197e-05,
+      "loss": 3.1011,
+      "step": 17120
+    },
+    {
+      "epoch": 0.7208842105263158,
+      "grad_norm": 0.439453125,
+      "learning_rate": 9.35741472732041e-05,
+      "loss": 3.2048,
+      "step": 17121
+    },
+    {
+      "epoch": 0.7209263157894736,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.354782609819157e-05,
+      "loss": 3.2017,
+      "step": 17122
+    },
+    {
+      "epoch": 0.7209684210526316,
+      "grad_norm": 0.41015625,
+      "learning_rate": 9.352150777355381e-05,
+      "loss": 3.137,
+      "step": 17123
+    },
+    {
+      "epoch": 0.7210105263157894,
+      "grad_norm": 0.466796875,
+      "learning_rate": 9.349519229977033e-05,
+      "loss": 2.791,
+      "step": 17124
+    },
+    {
+      "epoch": 0.7210526315789474,
+      "grad_norm": 0.435546875,
+      "learning_rate": 9.346887967732056e-05,
+      "loss": 3.3161,
+      "step": 17125
+    },
+    {
+      "epoch": 0.7210947368421052,
+      "grad_norm": 0.443359375,
+      "learning_rate": 9.344256990668379e-05,
+      "loss": 3.0221,
+      "step": 17126
+    },
+    {
+      "epoch": 0.7211368421052632,
+      "grad_norm": 0.43359375,
+      "learning_rate": 9.341626298833963e-05,
+      "loss": 3.2131,
+      "step": 17127
+    },
+    {
+      "epoch": 0.721178947368421,
+      "grad_norm": 0.453125,
+      "learning_rate": 9.3389958922767e-05,
+      "loss": 3.4351,
+      "step": 17128
+    },
+    {
+      "epoch": 0.721221052631579,
+      "grad_norm": 0.447265625,
+      "learning_rate": 9.336365771044541e-05,
+      "loss": 2.3956,
+      "step": 17129
+    },
+    {
+      "epoch": 0.7212631578947368,
+      "grad_norm": 0.443359375,
+      "learning_rate": 9.333735935185391e-05,
+      "loss": 2.6837,
+      "step": 17130
+    },
+    {
+      "epoch": 0.7213052631578948,
+      "grad_norm": 0.40234375,
+      "learning_rate": 9.331106384747168e-05,
+      "loss": 3.2801,
+      "step": 17131
+    },
+    {
+      "epoch": 0.7213473684210526,
+      "grad_norm": 0.447265625,
+      "learning_rate": 9.328477119777775e-05,
+      "loss": 3.4917,
+      "step": 17132
+    },
+    {
+      "epoch": 0.7213894736842106,
+      "grad_norm": 0.423828125,
+      "learning_rate": 9.325848140325108e-05,
+      "loss": 2.9655,
+      "step": 17133
+    },
+    {
+      "epoch": 0.7214315789473684,
+      "grad_norm": 0.435546875,
+      "learning_rate": 9.32321944643709e-05,
+      "loss": 2.9106,
+      "step": 17134
+    },
+    {
+      "epoch": 0.7214736842105263,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.320591038161574e-05,
+      "loss": 2.7763,
+      "step": 17135
+    },
+    {
+      "epoch": 0.7215157894736842,
+      "grad_norm": 0.44140625,
+      "learning_rate": 9.317962915546477e-05,
+      "loss": 3.1005,
+      "step": 17136
+    },
+    {
+      "epoch": 0.7215578947368421,
+      "grad_norm": 0.43359375,
+      "learning_rate": 9.31533507863967e-05,
+      "loss": 2.7304,
+      "step": 17137
+    },
+    {
+      "epoch": 0.7216,
+      "grad_norm": 0.431640625,
+      "learning_rate": 9.312707527489029e-05,
+      "loss": 3.3067,
+      "step": 17138
+    },
+    {
+      "epoch": 0.7216421052631579,
+      "grad_norm": 0.41015625,
+      "learning_rate": 9.310080262142422e-05,
+      "loss": 2.9438,
+      "step": 17139
+    },
+    {
+      "epoch": 0.7216842105263158,
+      "grad_norm": 0.431640625,
+      "learning_rate": 9.307453282647723e-05,
+      "loss": 3.3292,
+      "step": 17140
+    },
+    {
+      "epoch": 0.7217263157894737,
+      "grad_norm": 0.435546875,
+      "learning_rate": 9.304826589052784e-05,
+      "loss": 3.1027,
+      "step": 17141
+    },
+    {
+      "epoch": 0.7217684210526316,
+      "grad_norm": 0.4296875,
+      "learning_rate": 9.302200181405457e-05,
+      "loss": 2.7469,
+      "step": 17142
+    },
+    {
+      "epoch": 0.7218105263157895,
+      "grad_norm": 0.44921875,
+      "learning_rate": 9.299574059753604e-05,
+      "loss": 2.9746,
+      "step": 17143
+    },
+    {
+      "epoch": 0.7218526315789474,
+      "grad_norm": 0.51953125,
+      "learning_rate": 9.296948224145065e-05,
+      "loss": 3.4325,
+      "step": 17144
+    },
+    {
+      "epoch": 0.7218947368421053,
+      "grad_norm": 0.44921875,
+      "learning_rate": 9.29432267462768e-05,
+      "loss": 2.8044,
+      "step": 17145
+    },
+    {
+      "epoch": 0.7219368421052632,
+      "grad_norm": 0.44140625,
+      "learning_rate": 9.291697411249283e-05,
+      "loss": 3.1655,
+      "step": 17146
+    },
+    {
+      "epoch": 0.721978947368421,
+      "grad_norm": 0.453125,
+      "learning_rate": 9.289072434057705e-05,
+      "loss": 3.0306,
+      "step": 17147
+    },
+    {
+      "epoch": 0.7220210526315789,
+      "grad_norm": 0.419921875,
+      "learning_rate": 9.286447743100762e-05,
+      "loss": 3.2504,
+      "step": 17148
+    },
+    {
+      "epoch": 0.7220631578947369,
+      "grad_norm": 0.40625,
+      "learning_rate": 9.283823338426276e-05,
+      "loss": 2.8076,
+      "step": 17149
+    },
+    {
+      "epoch": 0.7221052631578947,
+      "grad_norm": 0.482421875,
+      "learning_rate": 9.281199220082068e-05,
+      "loss": 2.9643,
+      "step": 17150
+    },
+    {
+      "epoch": 0.7221473684210526,
+      "grad_norm": 0.443359375,
+      "learning_rate": 9.278575388115939e-05,
+      "loss": 3.3393,
+      "step": 17151
+    },
+    {
+      "epoch": 0.7221894736842105,
+      "grad_norm": 0.451171875,
+      "learning_rate": 9.2759518425757e-05,
+      "loss": 3.103,
+      "step": 17152
+    },
+    {
+      "epoch": 0.7222315789473684,
+      "grad_norm": 0.4296875,
+      "learning_rate": 9.273328583509139e-05,
+      "loss": 3.4286,
+      "step": 17153
+    },
+    {
+      "epoch": 0.7222736842105263,
+      "grad_norm": 0.435546875,
+      "learning_rate": 9.270705610964055e-05,
+      "loss": 3.5033,
+      "step": 17154
+    },
+    {
+      "epoch": 0.7223157894736842,
+      "grad_norm": 0.431640625,
+      "learning_rate": 9.268082924988233e-05,
+      "loss": 3.0154,
+      "step": 17155
+    },
+    {
+      "epoch": 0.7223578947368421,
+      "grad_norm": 0.427734375,
+      "learning_rate": 9.265460525629457e-05,
+      "loss": 2.8436,
+      "step": 17156
+    },
+    {
+      "epoch": 0.7224,
+      "grad_norm": 0.41015625,
+      "learning_rate": 9.262838412935492e-05,
+      "loss": 3.1102,
+      "step": 17157
+    },
+    {
+      "epoch": 0.7224421052631579,
+      "grad_norm": 0.408203125,
+      "learning_rate": 9.260216586954132e-05,
+      "loss": 2.54,
+      "step": 17158
+    },
+    {
+      "epoch": 0.7224842105263158,
+      "grad_norm": 0.388671875,
+      "learning_rate": 9.257595047733134e-05,
+      "loss": 2.9371,
+      "step": 17159
+    },
+    {
+      "epoch": 0.7225263157894737,
+      "grad_norm": 0.4296875,
+      "learning_rate": 9.254973795320254e-05,
+      "loss": 3.3212,
+      "step": 17160
+    },
+    {
+      "epoch": 0.7225684210526315,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.252352829763252e-05,
+      "loss": 3.1005,
+      "step": 17161
+    },
+    {
+      "epoch": 0.7226105263157895,
+      "grad_norm": 0.421875,
+      "learning_rate": 9.24973215110988e-05,
+      "loss": 3.0158,
+      "step": 17162
+    },
+    {
+      "epoch": 0.7226526315789473,
+      "grad_norm": 0.4765625,
+      "learning_rate": 9.24711175940788e-05,
+      "loss": 3.2439,
+      "step": 17163
+    },
+    {
+      "epoch": 0.7226947368421053,
+      "grad_norm": 0.416015625,
+      "learning_rate": 9.24449165470499e-05,
+      "loss": 2.9197,
+      "step": 17164
+    },
+    {
+      "epoch": 0.7227368421052631,
+      "grad_norm": 0.4375,
+      "learning_rate": 9.241871837048957e-05,
+      "loss": 3.1942,
+      "step": 17165
+    },
+    {
+      "epoch": 0.7227789473684211,
+      "grad_norm": 0.421875,
+      "learning_rate": 9.239252306487503e-05,
+      "loss": 3.3429,
+      "step": 17166
+    },
+    {
+      "epoch": 0.7228210526315789,
+      "grad_norm": 0.408203125,
+      "learning_rate": 9.236633063068356e-05,
+      "loss": 3.2426,
+      "step": 17167
+    },
+    {
+      "epoch": 0.7228631578947369,
+      "grad_norm": 0.421875,
+      "learning_rate": 9.23401410683923e-05,
+      "loss": 2.9097,
+      "step": 17168
+    },
+    {
+      "epoch": 0.7229052631578947,
+      "grad_norm": 0.421875,
+      "learning_rate": 9.231395437847845e-05,
+      "loss": 3.4837,
+      "step": 17169
+    },
+    {
+      "epoch": 0.7229473684210527,
+      "grad_norm": 0.41796875,
+      "learning_rate": 9.22877705614191e-05,
+      "loss": 3.2581,
+      "step": 17170
+    },
+    {
+      "epoch": 0.7229894736842105,
+      "grad_norm": 0.412109375,
+      "learning_rate": 9.226158961769115e-05,
+      "loss": 2.9161,
+      "step": 17171
+    },
+    {
+      "epoch": 0.7230315789473685,
+      "grad_norm": 0.447265625,
+      "learning_rate": 9.223541154777187e-05,
+      "loss": 3.7349,
+      "step": 17172
+    },
+    {
+      "epoch": 0.7230736842105263,
+      "grad_norm": 0.455078125,
+      "learning_rate": 9.220923635213788e-05,
+      "loss": 2.7016,
+      "step": 17173
+    },
+    {
+      "epoch": 0.7231157894736842,
+      "grad_norm": 0.44140625,
+      "learning_rate": 9.218306403126627e-05,
+      "loss": 3.1164,
+      "step": 17174
+    },
+    {
+      "epoch": 0.7231578947368421,
+      "grad_norm": 0.431640625,
+      "learning_rate": 9.21568945856338e-05,
+      "loss": 3.0624,
+      "step": 17175
+    },
+    {
+      "epoch": 0.7232,
+      "grad_norm": 0.4296875,
+      "learning_rate": 9.21307280157172e-05,
+      "loss": 3.6387,
+      "step": 17176
+    },
+    {
+      "epoch": 0.7232421052631579,
+      "grad_norm": 0.431640625,
+      "learning_rate": 9.210456432199338e-05,
+      "loss": 3.3541,
+      "step": 17177
+    },
+    {
+      "epoch": 0.7232842105263158,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.207840350493874e-05,
+      "loss": 3.2809,
+      "step": 17178
+    },
+    {
+      "epoch": 0.7233263157894737,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.205224556503019e-05,
+      "loss": 2.5804,
+      "step": 17179
+    },
+    {
+      "epoch": 0.7233684210526315,
+      "grad_norm": 0.388671875,
+      "learning_rate": 9.202609050274399e-05,
+      "loss": 3.0256,
+      "step": 17180
+    },
+    {
+      "epoch": 0.7234105263157895,
+      "grad_norm": 0.435546875,
+      "learning_rate": 9.199993831855688e-05,
+      "loss": 3.4492,
+      "step": 17181
+    },
+    {
+      "epoch": 0.7234526315789473,
+      "grad_norm": 0.4375,
+      "learning_rate": 9.197378901294525e-05,
+      "loss": 3.411,
+      "step": 17182
+    },
+    {
+      "epoch": 0.7234947368421053,
+      "grad_norm": 0.423828125,
+      "learning_rate": 9.194764258638552e-05,
+      "loss": 3.0743,
+      "step": 17183
+    },
+    {
+      "epoch": 0.7235368421052631,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.192149903935403e-05,
+      "loss": 3.2384,
+      "step": 17184
+    },
+    {
+      "epoch": 0.7235789473684211,
+      "grad_norm": 0.44921875,
+      "learning_rate": 9.189535837232701e-05,
+      "loss": 3.2272,
+      "step": 17185
+    },
+    {
+      "epoch": 0.7236210526315789,
+      "grad_norm": 0.443359375,
+      "learning_rate": 9.186922058578099e-05,
+      "loss": 3.1117,
+      "step": 17186
+    },
+    {
+      "epoch": 0.7236631578947369,
+      "grad_norm": 0.4375,
+      "learning_rate": 9.184308568019176e-05,
+      "loss": 3.0523,
+      "step": 17187
+    },
+    {
+      "epoch": 0.7237052631578947,
+      "grad_norm": 0.4453125,
+      "learning_rate": 9.181695365603587e-05,
+      "loss": 3.0491,
+      "step": 17188
+    },
+    {
+      "epoch": 0.7237473684210526,
+      "grad_norm": 0.443359375,
+      "learning_rate": 9.179082451378906e-05,
+      "loss": 3.2683,
+      "step": 17189
+    },
+    {
+      "epoch": 0.7237894736842105,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.176469825392763e-05,
+      "loss": 2.7752,
+      "step": 17190
+    },
+    {
+      "epoch": 0.7238315789473684,
+      "grad_norm": 0.41015625,
+      "learning_rate": 9.173857487692747e-05,
+      "loss": 3.1661,
+      "step": 17191
+    },
+    {
+      "epoch": 0.7238736842105263,
+      "grad_norm": 0.419921875,
+      "learning_rate": 9.171245438326445e-05,
+      "loss": 2.9622,
+      "step": 17192
+    },
+    {
+      "epoch": 0.7239157894736842,
+      "grad_norm": 0.4296875,
+      "learning_rate": 9.16863367734147e-05,
+      "loss": 3.0648,
+      "step": 17193
+    },
+    {
+      "epoch": 0.7239578947368421,
+      "grad_norm": 0.404296875,
+      "learning_rate": 9.166022204785371e-05,
+      "loss": 3.04,
+      "step": 17194
+    },
+    {
+      "epoch": 0.724,
+      "grad_norm": 0.40625,
+      "learning_rate": 9.163411020705762e-05,
+      "loss": 2.9497,
+      "step": 17195
+    },
+    {
+      "epoch": 0.7240421052631579,
+      "grad_norm": 0.427734375,
+      "learning_rate": 9.160800125150176e-05,
+      "loss": 3.2913,
+      "step": 17196
+    },
+    {
+      "epoch": 0.7240842105263158,
+      "grad_norm": 0.443359375,
+      "learning_rate": 9.158189518166213e-05,
+      "loss": 2.4641,
+      "step": 17197
+    },
+    {
+      "epoch": 0.7241263157894737,
+      "grad_norm": 0.474609375,
+      "learning_rate": 9.155579199801422e-05,
+      "loss": 2.8075,
+      "step": 17198
+    },
+    {
+      "epoch": 0.7241684210526316,
+      "grad_norm": 0.447265625,
+      "learning_rate": 9.152969170103362e-05,
+      "loss": 2.8622,
+      "step": 17199
+    },
+    {
+      "epoch": 0.7242105263157895,
+      "grad_norm": 0.40234375,
+      "learning_rate": 9.150359429119587e-05,
+      "loss": 3.0946,
+      "step": 17200
+    },
+    {
+      "epoch": 0.7242526315789474,
+      "grad_norm": 0.466796875,
+      "learning_rate": 9.147749976897629e-05,
+      "loss": 2.7788,
+      "step": 17201
+    },
+    {
+      "epoch": 0.7242947368421052,
+      "grad_norm": 0.4140625,
+      "learning_rate": 9.145140813485058e-05,
+      "loss": 3.1854,
+      "step": 17202
+    },
+    {
+      "epoch": 0.7243368421052632,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.142531938929377e-05,
+      "loss": 2.6205,
+      "step": 17203
+    },
+    {
+      "epoch": 0.724378947368421,
+      "grad_norm": 0.4296875,
+      "learning_rate": 9.139923353278141e-05,
+      "loss": 2.8062,
+      "step": 17204
+    },
+    {
+      "epoch": 0.724421052631579,
+      "grad_norm": 0.431640625,
+      "learning_rate": 9.137315056578863e-05,
+      "loss": 2.8614,
+      "step": 17205
+    },
+    {
+      "epoch": 0.7244631578947368,
+      "grad_norm": 0.439453125,
+      "learning_rate": 9.134707048879071e-05,
+      "loss": 2.8881,
+      "step": 17206
+    },
+    {
+      "epoch": 0.7245052631578948,
+      "grad_norm": 0.41796875,
+      "learning_rate": 9.132099330226273e-05,
+      "loss": 2.9783,
+      "step": 17207
+    },
+    {
+      "epoch": 0.7245473684210526,
+      "grad_norm": 0.44140625,
+      "learning_rate": 9.129491900667974e-05,
+      "loss": 2.9735,
+      "step": 17208
+    },
+    {
+      "epoch": 0.7245894736842106,
+      "grad_norm": 0.4296875,
+      "learning_rate": 9.126884760251703e-05,
+      "loss": 3.0815,
+      "step": 17209
+    },
+    {
+      "epoch": 0.7246315789473684,
+      "grad_norm": 0.412109375,
+      "learning_rate": 9.124277909024923e-05,
+      "loss": 3.0697,
+      "step": 17210
+    },
+    {
+      "epoch": 0.7246736842105264,
+      "grad_norm": 0.40625,
+      "learning_rate": 9.121671347035157e-05,
+      "loss": 2.9519,
+      "step": 17211
+    },
+    {
+      "epoch": 0.7247157894736842,
+      "grad_norm": 0.423828125,
+      "learning_rate": 9.119065074329879e-05,
+      "loss": 3.4233,
+      "step": 17212
+    },
+    {
+      "epoch": 0.7247578947368422,
+      "grad_norm": 0.400390625,
+      "learning_rate": 9.116459090956581e-05,
+      "loss": 3.2722,
+      "step": 17213
+    },
+    {
+      "epoch": 0.7248,
+      "grad_norm": 0.427734375,
+      "learning_rate": 9.113853396962732e-05,
+      "loss": 3.2711,
+      "step": 17214
+    },
+    {
+      "epoch": 0.7248421052631578,
+      "grad_norm": 0.421875,
+      "learning_rate": 9.111247992395811e-05,
+      "loss": 3.095,
+      "step": 17215
+    },
+    {
+      "epoch": 0.7248842105263158,
+      "grad_norm": 0.43359375,
+      "learning_rate": 9.10864287730328e-05,
+      "loss": 3.3393,
+      "step": 17216
+    },
+    {
+      "epoch": 0.7249263157894736,
+      "grad_norm": 0.4375,
+      "learning_rate": 9.106038051732598e-05,
+      "loss": 3.1158,
+      "step": 17217
+    },
+    {
+      "epoch": 0.7249684210526316,
+      "grad_norm": 0.47265625,
+      "learning_rate": 9.103433515731238e-05,
+      "loss": 2.8447,
+      "step": 17218
+    },
+    {
+      "epoch": 0.7250105263157894,
+      "grad_norm": 0.47265625,
+      "learning_rate": 9.10082926934664e-05,
+      "loss": 3.0099,
+      "step": 17219
+    },
+    {
+      "epoch": 0.7250526315789474,
+      "grad_norm": 0.416015625,
+      "learning_rate": 9.098225312626252e-05,
+      "loss": 3.02,
+      "step": 17220
+    },
+    {
+      "epoch": 0.7250947368421052,
+      "grad_norm": 0.4375,
+      "learning_rate": 9.095621645617513e-05,
+      "loss": 2.9472,
+      "step": 17221
+    },
+    {
+      "epoch": 0.7251368421052632,
+      "grad_norm": 0.4375,
+      "learning_rate": 9.093018268367863e-05,
+      "loss": 3.4754,
+      "step": 17222
+    },
+    {
+      "epoch": 0.725178947368421,
+      "grad_norm": 0.4453125,
+      "learning_rate": 9.09041518092473e-05,
+      "loss": 3.0367,
+      "step": 17223
+    },
+    {
+      "epoch": 0.725221052631579,
+      "grad_norm": 0.48828125,
+      "learning_rate": 9.08781238333553e-05,
+      "loss": 3.1135,
+      "step": 17224
+    },
+    {
+      "epoch": 0.7252631578947368,
+      "grad_norm": 0.431640625,
+      "learning_rate": 9.085209875647704e-05,
+      "loss": 3.2696,
+      "step": 17225
+    },
+    {
+      "epoch": 0.7253052631578948,
+      "grad_norm": 0.4140625,
+      "learning_rate": 9.082607657908651e-05,
+      "loss": 2.9587,
+      "step": 17226
+    },
+    {
+      "epoch": 0.7253473684210526,
+      "grad_norm": 0.453125,
+      "learning_rate": 9.080005730165786e-05,
+      "loss": 2.5119,
+      "step": 17227
+    },
+    {
+      "epoch": 0.7253894736842105,
+      "grad_norm": 0.419921875,
+      "learning_rate": 9.077404092466513e-05,
+      "loss": 3.1148,
+      "step": 17228
+    },
+    {
+      "epoch": 0.7254315789473684,
+      "grad_norm": 0.44921875,
+      "learning_rate": 9.074802744858227e-05,
+      "loss": 3.1082,
+      "step": 17229
+    },
+    {
+      "epoch": 0.7254736842105263,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.072201687388315e-05,
+      "loss": 2.963,
+      "step": 17230
+    },
+    {
+      "epoch": 0.7255157894736842,
+      "grad_norm": 0.439453125,
+      "learning_rate": 9.06960092010419e-05,
+      "loss": 3.2361,
+      "step": 17231
+    },
+    {
+      "epoch": 0.7255578947368421,
+      "grad_norm": 0.41015625,
+      "learning_rate": 9.067000443053206e-05,
+      "loss": 2.7243,
+      "step": 17232
+    },
+    {
+      "epoch": 0.7256,
+      "grad_norm": 0.4453125,
+      "learning_rate": 9.064400256282756e-05,
+      "loss": 3.2412,
+      "step": 17233
+    },
+    {
+      "epoch": 0.7256421052631579,
+      "grad_norm": 0.421875,
+      "learning_rate": 9.061800359840214e-05,
+      "loss": 2.4901,
+      "step": 17234
+    },
+    {
+      "epoch": 0.7256842105263158,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.059200753772942e-05,
+      "loss": 3.2408,
+      "step": 17235
+    },
+    {
+      "epoch": 0.7257263157894737,
+      "grad_norm": 0.439453125,
+      "learning_rate": 9.056601438128303e-05,
+      "loss": 2.5786,
+      "step": 17236
+    },
+    {
+      "epoch": 0.7257684210526316,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.054002412953643e-05,
+      "loss": 3.2745,
+      "step": 17237
+    },
+    {
+      "epoch": 0.7258105263157895,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.051403678296342e-05,
+      "loss": 2.9857,
+      "step": 17238
+    },
+    {
+      "epoch": 0.7258526315789474,
+      "grad_norm": 0.451171875,
+      "learning_rate": 9.048805234203708e-05,
+      "loss": 3.4904,
+      "step": 17239
+    },
+    {
+      "epoch": 0.7258947368421053,
+      "grad_norm": 0.45703125,
+      "learning_rate": 9.046207080723112e-05,
+      "loss": 3.1568,
+      "step": 17240
+    },
+    {
+      "epoch": 0.7259368421052631,
+      "grad_norm": 0.423828125,
+      "learning_rate": 9.043609217901877e-05,
+      "loss": 3.0911,
+      "step": 17241
+    },
+    {
+      "epoch": 0.7259789473684211,
+      "grad_norm": 0.439453125,
+      "learning_rate": 9.041011645787334e-05,
+      "loss": 2.9859,
+      "step": 17242
+    },
+    {
+      "epoch": 0.7260210526315789,
+      "grad_norm": 0.44140625,
+      "learning_rate": 9.038414364426806e-05,
+      "loss": 2.7096,
+      "step": 17243
+    },
+    {
+      "epoch": 0.7260631578947369,
+      "grad_norm": 0.453125,
+      "learning_rate": 9.035817373867608e-05,
+      "loss": 3.2808,
+      "step": 17244
+    },
+    {
+      "epoch": 0.7261052631578947,
+      "grad_norm": 0.416015625,
+      "learning_rate": 9.033220674157075e-05,
+      "loss": 3.2713,
+      "step": 17245
+    },
+    {
+      "epoch": 0.7261473684210527,
+      "grad_norm": 0.4453125,
+      "learning_rate": 9.030624265342486e-05,
+      "loss": 3.3782,
+      "step": 17246
+    },
+    {
+      "epoch": 0.7261894736842105,
+      "grad_norm": 0.431640625,
+      "learning_rate": 9.028028147471177e-05,
+      "loss": 2.887,
+      "step": 17247
+    },
+    {
+      "epoch": 0.7262315789473685,
+      "grad_norm": 0.419921875,
+      "learning_rate": 9.02543232059041e-05,
+      "loss": 2.6477,
+      "step": 17248
+    },
+    {
+      "epoch": 0.7262736842105263,
+      "grad_norm": 0.427734375,
+      "learning_rate": 9.022836784747507e-05,
+      "loss": 2.5102,
+      "step": 17249
+    },
+    {
+      "epoch": 0.7263157894736842,
+      "grad_norm": 0.423828125,
+      "learning_rate": 9.020241539989746e-05,
+      "loss": 2.9786,
+      "step": 17250
+    },
+    {
+      "epoch": 0.7263578947368421,
+      "grad_norm": 0.5078125,
+      "learning_rate": 9.017646586364398e-05,
+      "loss": 2.9606,
+      "step": 17251
+    },
+    {
+      "epoch": 0.7264,
+      "grad_norm": 0.482421875,
+      "learning_rate": 9.015051923918768e-05,
+      "loss": 3.2284,
+      "step": 17252
+    },
+    {
+      "epoch": 0.7264421052631579,
+      "grad_norm": 0.396484375,
+      "learning_rate": 9.012457552700095e-05,
+      "loss": 2.8222,
+      "step": 17253
+    },
+    {
+      "epoch": 0.7264842105263158,
+      "grad_norm": 0.44140625,
+      "learning_rate": 9.009863472755677e-05,
+      "loss": 3.0449,
+      "step": 17254
+    },
+    {
+      "epoch": 0.7265263157894737,
+      "grad_norm": 0.455078125,
+      "learning_rate": 9.007269684132741e-05,
+      "loss": 2.7784,
+      "step": 17255
+    },
+    {
+      "epoch": 0.7265684210526315,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.004676186878571e-05,
+      "loss": 3.2042,
+      "step": 17256
+    },
+    {
+      "epoch": 0.7266105263157895,
+      "grad_norm": 0.408203125,
+      "learning_rate": 9.002082981040405e-05,
+      "loss": 3.128,
+      "step": 17257
+    },
+    {
+      "epoch": 0.7266526315789473,
+      "grad_norm": 0.44921875,
+      "learning_rate": 8.999490066665491e-05,
+      "loss": 3.1982,
+      "step": 17258
+    },
+    {
+      "epoch": 0.7266947368421053,
+      "grad_norm": 0.474609375,
+      "learning_rate": 8.996897443801069e-05,
+      "loss": 3.3585,
+      "step": 17259
+    },
+    {
+      "epoch": 0.7267368421052631,
+      "grad_norm": 0.439453125,
+      "learning_rate": 8.994305112494364e-05,
+      "loss": 3.3966,
+      "step": 17260
+    },
+    {
+      "epoch": 0.7267789473684211,
+      "grad_norm": 0.427734375,
+      "learning_rate": 8.991713072792631e-05,
+      "loss": 3.1515,
+      "step": 17261
+    },
+    {
+      "epoch": 0.7268210526315789,
+      "grad_norm": 0.42578125,
+      "learning_rate": 8.98912132474306e-05,
+      "loss": 3.3688,
+      "step": 17262
+    },
+    {
+      "epoch": 0.7268631578947369,
+      "grad_norm": 0.404296875,
+      "learning_rate": 8.986529868392904e-05,
+      "loss": 2.9692,
+      "step": 17263
+    },
+    {
+      "epoch": 0.7269052631578947,
+      "grad_norm": 0.42578125,
+      "learning_rate": 8.983938703789337e-05,
+      "loss": 3.3511,
+      "step": 17264
+    },
+    {
+      "epoch": 0.7269473684210527,
+      "grad_norm": 0.478515625,
+      "learning_rate": 8.981347830979603e-05,
+      "loss": 3.0733,
+      "step": 17265
+    },
+    {
+      "epoch": 0.7269894736842105,
+      "grad_norm": 0.431640625,
+      "learning_rate": 8.978757250010888e-05,
+      "loss": 2.7986,
+      "step": 17266
+    },
+    {
+      "epoch": 0.7270315789473685,
+      "grad_norm": 0.4765625,
+      "learning_rate": 8.976166960930385e-05,
+      "loss": 2.9375,
+      "step": 17267
+    },
+    {
+      "epoch": 0.7270736842105263,
+      "grad_norm": 0.4140625,
+      "learning_rate": 8.973576963785308e-05,
+      "loss": 3.3754,
+      "step": 17268
+    },
+    {
+      "epoch": 0.7271157894736842,
+      "grad_norm": 0.408203125,
+      "learning_rate": 8.970987258622812e-05,
+      "loss": 3.2108,
+      "step": 17269
+    },
+    {
+      "epoch": 0.7271578947368421,
+      "grad_norm": 0.439453125,
+      "learning_rate": 8.968397845490111e-05,
+      "loss": 3.1547,
+      "step": 17270
+    },
+    {
+      "epoch": 0.7272,
+      "grad_norm": 0.462890625,
+      "learning_rate": 8.965808724434349e-05,
+      "loss": 2.6631,
+      "step": 17271
+    },
+    {
+      "epoch": 0.7272421052631579,
+      "grad_norm": 0.423828125,
+      "learning_rate": 8.963219895502719e-05,
+      "loss": 2.826,
+      "step": 17272
+    },
+    {
+      "epoch": 0.7272842105263158,
+      "grad_norm": 0.45703125,
+      "learning_rate": 8.960631358742383e-05,
+      "loss": 3.4719,
+      "step": 17273
+    },
+    {
+      "epoch": 0.7273263157894737,
+      "grad_norm": 0.4375,
+      "learning_rate": 8.958043114200496e-05,
+      "loss": 3.2278,
+      "step": 17274
+    },
+    {
+      "epoch": 0.7273684210526316,
+      "grad_norm": 0.431640625,
+      "learning_rate": 8.955455161924216e-05,
+      "loss": 2.8796,
+      "step": 17275
+    },
+    {
+      "epoch": 0.7274105263157895,
+      "grad_norm": 0.458984375,
+      "learning_rate": 8.952867501960682e-05,
+      "loss": 2.7719,
+      "step": 17276
+    },
+    {
+      "epoch": 0.7274526315789474,
+      "grad_norm": 0.40234375,
+      "learning_rate": 8.950280134357059e-05,
+      "loss": 2.3865,
+      "step": 17277
+    },
+    {
+      "epoch": 0.7274947368421053,
+      "grad_norm": 0.439453125,
+      "learning_rate": 8.947693059160469e-05,
+      "loss": 3.2381,
+      "step": 17278
+    },
+    {
+      "epoch": 0.7275368421052631,
+      "grad_norm": 0.443359375,
+      "learning_rate": 8.945106276418055e-05,
+      "loss": 3.184,
+      "step": 17279
+    },
+    {
+      "epoch": 0.7275789473684211,
+      "grad_norm": 0.439453125,
+      "learning_rate": 8.942519786176939e-05,
+      "loss": 3.1452,
+      "step": 17280
+    },
+    {
+      "epoch": 0.727621052631579,
+      "grad_norm": 0.408203125,
+      "learning_rate": 8.939933588484247e-05,
+      "loss": 2.9437,
+      "step": 17281
+    },
+    {
+      "epoch": 0.7276631578947368,
+      "grad_norm": 0.412109375,
+      "learning_rate": 8.937347683387096e-05,
+      "loss": 3.0372,
+      "step": 17282
+    },
+    {
+      "epoch": 0.7277052631578947,
+      "grad_norm": 0.43359375,
+      "learning_rate": 8.934762070932587e-05,
+      "loss": 2.8693,
+      "step": 17283
+    },
+    {
+      "epoch": 0.7277473684210526,
+      "grad_norm": 0.45703125,
+      "learning_rate": 8.932176751167845e-05,
+      "loss": 3.6909,
+      "step": 17284
+    },
+    {
+      "epoch": 0.7277894736842105,
+      "grad_norm": 0.4375,
+      "learning_rate": 8.929591724139966e-05,
+      "loss": 3.1891,
+      "step": 17285
+    },
+    {
+      "epoch": 0.7278315789473684,
+      "grad_norm": 0.416015625,
+      "learning_rate": 8.927006989896042e-05,
+      "loss": 3.0754,
+      "step": 17286
+    },
+    {
+      "epoch": 0.7278736842105263,
+      "grad_norm": 0.447265625,
+      "learning_rate": 8.924422548483165e-05,
+      "loss": 3.2261,
+      "step": 17287
+    },
+    {
+      "epoch": 0.7279157894736842,
+      "grad_norm": 0.416015625,
+      "learning_rate": 8.921838399948423e-05,
+      "loss": 2.528,
+      "step": 17288
+    },
+    {
+      "epoch": 0.7279578947368421,
+      "grad_norm": 0.431640625,
+      "learning_rate": 8.919254544338895e-05,
+      "loss": 3.0695,
+      "step": 17289
+    },
+    {
+      "epoch": 0.728,
+      "grad_norm": 0.431640625,
+      "learning_rate": 8.916670981701655e-05,
+      "loss": 3.2911,
+      "step": 17290
+    },
+    {
+      "epoch": 0.7280421052631579,
+      "grad_norm": 0.45703125,
+      "learning_rate": 8.914087712083762e-05,
+      "loss": 3.251,
+      "step": 17291
+    },
+    {
+      "epoch": 0.7280842105263158,
+      "grad_norm": 0.439453125,
+      "learning_rate": 8.9115047355323e-05,
+      "loss": 2.9643,
+      "step": 17292
+    },
+    {
+      "epoch": 0.7281263157894737,
+      "grad_norm": 0.408203125,
+      "learning_rate": 8.908922052094318e-05,
+      "loss": 3.0655,
+      "step": 17293
+    },
+    {
+      "epoch": 0.7281684210526316,
+      "grad_norm": 0.435546875,
+      "learning_rate": 8.90633966181687e-05,
+      "loss": 3.4648,
+      "step": 17294
+    },
+    {
+      "epoch": 0.7282105263157894,
+      "grad_norm": 0.466796875,
+      "learning_rate": 8.903757564747006e-05,
+      "loss": 3.0558,
+      "step": 17295
+    },
+    {
+      "epoch": 0.7282526315789474,
+      "grad_norm": 0.41796875,
+      "learning_rate": 8.901175760931765e-05,
+      "loss": 2.9578,
+      "step": 17296
+    },
+    {
+      "epoch": 0.7282947368421052,
+      "grad_norm": 0.4296875,
+      "learning_rate": 8.898594250418188e-05,
+      "loss": 3.3682,
+      "step": 17297
+    },
+    {
+      "epoch": 0.7283368421052632,
+      "grad_norm": 0.412109375,
+      "learning_rate": 8.896013033253295e-05,
+      "loss": 2.8451,
+      "step": 17298
+    },
+    {
+      "epoch": 0.728378947368421,
+      "grad_norm": 0.4140625,
+      "learning_rate": 8.893432109484135e-05,
+      "loss": 2.5537,
+      "step": 17299
+    },
+    {
+      "epoch": 0.728421052631579,
+      "grad_norm": 0.4140625,
+      "learning_rate": 8.890851479157713e-05,
+      "loss": 3.04,
+      "step": 17300
+    },
+    {
+      "epoch": 0.7284631578947368,
+      "grad_norm": 0.44140625,
+      "learning_rate": 8.888271142321052e-05,
+      "loss": 3.4336,
+      "step": 17301
+    },
+    {
+      "epoch": 0.7285052631578948,
+      "grad_norm": 0.431640625,
+      "learning_rate": 8.885691099021159e-05,
+      "loss": 3.5854,
+      "step": 17302
+    },
+    {
+      "epoch": 0.7285473684210526,
+      "grad_norm": 0.431640625,
+      "learning_rate": 8.883111349305042e-05,
+      "loss": 3.2895,
+      "step": 17303
+    },
+    {
+      "epoch": 0.7285894736842106,
+      "grad_norm": 0.43359375,
+      "learning_rate": 8.880531893219698e-05,
+      "loss": 3.4168,
+      "step": 17304
+    },
+    {
+      "epoch": 0.7286315789473684,
+      "grad_norm": 0.40234375,
+      "learning_rate": 8.877952730812117e-05,
+      "loss": 2.5948,
+      "step": 17305
+    },
+    {
+      "epoch": 0.7286736842105264,
+      "grad_norm": 0.44921875,
+      "learning_rate": 8.87537386212931e-05,
+      "loss": 3.1602,
+      "step": 17306
+    },
+    {
+      "epoch": 0.7287157894736842,
+      "grad_norm": 0.4296875,
+      "learning_rate": 8.87279528721823e-05,
+      "loss": 3.2576,
+      "step": 17307
+    },
+    {
+      "epoch": 0.7287578947368422,
+      "grad_norm": 0.43359375,
+      "learning_rate": 8.870217006125877e-05,
+      "loss": 2.5548,
+      "step": 17308
+    },
+    {
+      "epoch": 0.7288,
+      "grad_norm": 0.443359375,
+      "learning_rate": 8.86763901889922e-05,
+      "loss": 2.7953,
+      "step": 17309
+    },
+    {
+      "epoch": 0.7288421052631578,
+      "grad_norm": 0.5,
+      "learning_rate": 8.865061325585227e-05,
+      "loss": 3.199,
+      "step": 17310
+    },
+    {
+      "epoch": 0.7288842105263158,
+      "grad_norm": 0.41796875,
+      "learning_rate": 8.862483926230857e-05,
+      "loss": 2.7377,
+      "step": 17311
+    },
+    {
+      "epoch": 0.7289263157894736,
+      "grad_norm": 0.41015625,
+      "learning_rate": 8.859906820883063e-05,
+      "loss": 3.3574,
+      "step": 17312
+    },
+    {
+      "epoch": 0.7289684210526316,
+      "grad_norm": 0.408203125,
+      "learning_rate": 8.857330009588816e-05,
+      "loss": 3.0753,
+      "step": 17313
+    },
+    {
+      "epoch": 0.7290105263157894,
+      "grad_norm": 0.41015625,
+      "learning_rate": 8.854753492395034e-05,
+      "loss": 3.0465,
+      "step": 17314
+    },
+    {
+      "epoch": 0.7290526315789474,
+      "grad_norm": 0.498046875,
+      "learning_rate": 8.852177269348682e-05,
+      "loss": 3.3348,
+      "step": 17315
+    },
+    {
+      "epoch": 0.7290947368421052,
+      "grad_norm": 0.494140625,
+      "learning_rate": 8.849601340496688e-05,
+      "loss": 3.3594,
+      "step": 17316
+    },
+    {
+      "epoch": 0.7291368421052632,
+      "grad_norm": 0.404296875,
+      "learning_rate": 8.847025705885983e-05,
+      "loss": 2.9826,
+      "step": 17317
+    },
+    {
+      "epoch": 0.729178947368421,
+      "grad_norm": 0.451171875,
+      "learning_rate": 8.84445036556349e-05,
+      "loss": 3.0789,
+      "step": 17318
+    },
+    {
+      "epoch": 0.729221052631579,
+      "grad_norm": 0.42578125,
+      "learning_rate": 8.841875319576118e-05,
+      "loss": 3.0824,
+      "step": 17319
+    },
+    {
+      "epoch": 0.7292631578947368,
+      "grad_norm": 0.427734375,
+      "learning_rate": 8.839300567970809e-05,
+      "loss": 3.3129,
+      "step": 17320
+    },
+    {
+      "epoch": 0.7293052631578948,
+      "grad_norm": 0.42578125,
+      "learning_rate": 8.836726110794443e-05,
+      "loss": 2.7362,
+      "step": 17321
+    },
+    {
+      "epoch": 0.7293473684210526,
+      "grad_norm": 0.439453125,
+      "learning_rate": 8.834151948093952e-05,
+      "loss": 3.0541,
+      "step": 17322
+    },
+    {
+      "epoch": 0.7293894736842105,
+      "grad_norm": 0.43359375,
+      "learning_rate": 8.8315780799162e-05,
+      "loss": 2.881,
+      "step": 17323
+    },
+    {
+      "epoch": 0.7294315789473684,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.829004506308106e-05,
+      "loss": 2.8521,
+      "step": 17324
+    },
+    {
+      "epoch": 0.7294736842105263,
+      "grad_norm": 0.4375,
+      "learning_rate": 8.826431227316553e-05,
+      "loss": 3.1212,
+      "step": 17325
+    },
+    {
+      "epoch": 0.7295157894736842,
+      "grad_norm": 0.443359375,
+      "learning_rate": 8.823858242988408e-05,
+      "loss": 2.97,
+      "step": 17326
+    },
+    {
+      "epoch": 0.7295578947368421,
+      "grad_norm": 0.41796875,
+      "learning_rate": 8.821285553370578e-05,
+      "loss": 2.902,
+      "step": 17327
+    },
+    {
+      "epoch": 0.7296,
+      "grad_norm": 0.4140625,
+      "learning_rate": 8.8187131585099e-05,
+      "loss": 3.1538,
+      "step": 17328
+    },
+    {
+      "epoch": 0.7296421052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 8.816141058453272e-05,
+      "loss": 3.1456,
+      "step": 17329
+    },
+    {
+      "epoch": 0.7296842105263158,
+      "grad_norm": 0.41796875,
+      "learning_rate": 8.813569253247522e-05,
+      "loss": 2.7389,
+      "step": 17330
+    },
+    {
+      "epoch": 0.7297263157894737,
+      "grad_norm": 0.43359375,
+      "learning_rate": 8.810997742939531e-05,
+      "loss": 3.4947,
+      "step": 17331
+    },
+    {
+      "epoch": 0.7297684210526316,
+      "grad_norm": 0.44140625,
+      "learning_rate": 8.808426527576138e-05,
+      "loss": 3.0299,
+      "step": 17332
+    },
+    {
+      "epoch": 0.7298105263157895,
+      "grad_norm": 0.4140625,
+      "learning_rate": 8.805855607204192e-05,
+      "loss": 2.8992,
+      "step": 17333
+    },
+    {
+      "epoch": 0.7298526315789474,
+      "grad_norm": 0.431640625,
+      "learning_rate": 8.803284981870527e-05,
+      "loss": 3.1772,
+      "step": 17334
+    },
+    {
+      "epoch": 0.7298947368421053,
+      "grad_norm": 0.4296875,
+      "learning_rate": 8.800714651621972e-05,
+      "loss": 3.0338,
+      "step": 17335
+    },
+    {
+      "epoch": 0.7299368421052631,
+      "grad_norm": 0.439453125,
+      "learning_rate": 8.798144616505377e-05,
+      "loss": 3.018,
+      "step": 17336
+    },
+    {
+      "epoch": 0.7299789473684211,
+      "grad_norm": 0.43359375,
+      "learning_rate": 8.795574876567539e-05,
+      "loss": 2.7872,
+      "step": 17337
+    },
+    {
+      "epoch": 0.7300210526315789,
+      "grad_norm": 0.443359375,
+      "learning_rate": 8.793005431855292e-05,
+      "loss": 3.0086,
+      "step": 17338
+    },
+    {
+      "epoch": 0.7300631578947369,
+      "grad_norm": 0.4453125,
+      "learning_rate": 8.790436282415443e-05,
+      "loss": 3.4956,
+      "step": 17339
+    },
+    {
+      "epoch": 0.7301052631578947,
+      "grad_norm": 0.443359375,
+      "learning_rate": 8.787867428294801e-05,
+      "loss": 3.4822,
+      "step": 17340
+    },
+    {
+      "epoch": 0.7301473684210527,
+      "grad_norm": 0.3984375,
+      "learning_rate": 8.785298869540167e-05,
+      "loss": 3.0295,
+      "step": 17341
+    },
+    {
+      "epoch": 0.7301894736842105,
+      "grad_norm": 0.443359375,
+      "learning_rate": 8.782730606198325e-05,
+      "loss": 3.0432,
+      "step": 17342
+    },
+    {
+      "epoch": 0.7302315789473685,
+      "grad_norm": 0.431640625,
+      "learning_rate": 8.780162638316097e-05,
+      "loss": 3.2298,
+      "step": 17343
+    },
+    {
+      "epoch": 0.7302736842105263,
+      "grad_norm": 0.41796875,
+      "learning_rate": 8.777594965940231e-05,
+      "loss": 3.3024,
+      "step": 17344
+    },
+    {
+      "epoch": 0.7303157894736843,
+      "grad_norm": 0.4375,
+      "learning_rate": 8.77502758911753e-05,
+      "loss": 3.5554,
+      "step": 17345
+    },
+    {
+      "epoch": 0.7303578947368421,
+      "grad_norm": 0.431640625,
+      "learning_rate": 8.772460507894764e-05,
+      "loss": 3.0164,
+      "step": 17346
+    },
+    {
+      "epoch": 0.7304,
+      "grad_norm": 0.416015625,
+      "learning_rate": 8.7698937223187e-05,
+      "loss": 3.1862,
+      "step": 17347
+    },
+    {
+      "epoch": 0.7304421052631579,
+      "grad_norm": 0.443359375,
+      "learning_rate": 8.767327232436103e-05,
+      "loss": 3.5387,
+      "step": 17348
+    },
+    {
+      "epoch": 0.7304842105263157,
+      "grad_norm": 0.427734375,
+      "learning_rate": 8.764761038293733e-05,
+      "loss": 2.7363,
+      "step": 17349
+    },
+    {
+      "epoch": 0.7305263157894737,
+      "grad_norm": 0.421875,
+      "learning_rate": 8.762195139938342e-05,
+      "loss": 2.899,
+      "step": 17350
+    },
+    {
+      "epoch": 0.7305684210526315,
+      "grad_norm": 0.41796875,
+      "learning_rate": 8.759629537416666e-05,
+      "loss": 2.8102,
+      "step": 17351
+    },
+    {
+      "epoch": 0.7306105263157895,
+      "grad_norm": 0.423828125,
+      "learning_rate": 8.757064230775466e-05,
+      "loss": 3.327,
+      "step": 17352
+    },
+    {
+      "epoch": 0.7306526315789473,
+      "grad_norm": 0.4609375,
+      "learning_rate": 8.754499220061473e-05,
+      "loss": 3.1067,
+      "step": 17353
+    },
+    {
+      "epoch": 0.7306947368421053,
+      "grad_norm": 0.40234375,
+      "learning_rate": 8.751934505321415e-05,
+      "loss": 2.5714,
+      "step": 17354
+    },
+    {
+      "epoch": 0.7307368421052631,
+      "grad_norm": 0.400390625,
+      "learning_rate": 8.74937008660202e-05,
+      "loss": 2.8484,
+      "step": 17355
+    },
+    {
+      "epoch": 0.7307789473684211,
+      "grad_norm": 0.412109375,
+      "learning_rate": 8.746805963950011e-05,
+      "loss": 3.0872,
+      "step": 17356
+    },
+    {
+      "epoch": 0.7308210526315789,
+      "grad_norm": 0.4375,
+      "learning_rate": 8.744242137412097e-05,
+      "loss": 2.6025,
+      "step": 17357
+    },
+    {
+      "epoch": 0.7308631578947369,
+      "grad_norm": 0.419921875,
+      "learning_rate": 8.741678607034986e-05,
+      "loss": 2.9304,
+      "step": 17358
+    },
+    {
+      "epoch": 0.7309052631578947,
+      "grad_norm": 0.421875,
+      "learning_rate": 8.739115372865394e-05,
+      "loss": 2.837,
+      "step": 17359
+    },
+    {
+      "epoch": 0.7309473684210527,
+      "grad_norm": 0.423828125,
+      "learning_rate": 8.736552434950015e-05,
+      "loss": 2.8668,
+      "step": 17360
+    },
+    {
+      "epoch": 0.7309894736842105,
+      "grad_norm": 0.4140625,
+      "learning_rate": 8.733989793335542e-05,
+      "loss": 2.9285,
+      "step": 17361
+    },
+    {
+      "epoch": 0.7310315789473684,
+      "grad_norm": 0.435546875,
+      "learning_rate": 8.731427448068666e-05,
+      "loss": 3.2715,
+      "step": 17362
+    },
+    {
+      "epoch": 0.7310736842105263,
+      "grad_norm": 0.453125,
+      "learning_rate": 8.728865399196062e-05,
+      "loss": 3.0269,
+      "step": 17363
+    },
+    {
+      "epoch": 0.7311157894736842,
+      "grad_norm": 0.419921875,
+      "learning_rate": 8.726303646764417e-05,
+      "loss": 2.8758,
+      "step": 17364
+    },
+    {
+      "epoch": 0.7311578947368421,
+      "grad_norm": 0.4140625,
+      "learning_rate": 8.723742190820396e-05,
+      "loss": 3.4082,
+      "step": 17365
+    },
+    {
+      "epoch": 0.7312,
+      "grad_norm": 0.396484375,
+      "learning_rate": 8.72118103141066e-05,
+      "loss": 3.1981,
+      "step": 17366
+    },
+    {
+      "epoch": 0.7312421052631579,
+      "grad_norm": 0.41015625,
+      "learning_rate": 8.71862016858189e-05,
+      "loss": 3.0439,
+      "step": 17367
+    },
+    {
+      "epoch": 0.7312842105263158,
+      "grad_norm": 0.431640625,
+      "learning_rate": 8.716059602380727e-05,
+      "loss": 3.0945,
+      "step": 17368
+    },
+    {
+      "epoch": 0.7313263157894737,
+      "grad_norm": 0.4296875,
+      "learning_rate": 8.713499332853828e-05,
+      "loss": 2.9448,
+      "step": 17369
+    },
+    {
+      "epoch": 0.7313684210526316,
+      "grad_norm": 0.455078125,
+      "learning_rate": 8.710939360047834e-05,
+      "loss": 3.3814,
+      "step": 17370
+    },
+    {
+      "epoch": 0.7314105263157895,
+      "grad_norm": 0.46484375,
+      "learning_rate": 8.708379684009387e-05,
+      "loss": 3.5724,
+      "step": 17371
+    },
+    {
+      "epoch": 0.7314526315789474,
+      "grad_norm": 0.41015625,
+      "learning_rate": 8.70582030478512e-05,
+      "loss": 3.1196,
+      "step": 17372
+    },
+    {
+      "epoch": 0.7314947368421053,
+      "grad_norm": 0.43359375,
+      "learning_rate": 8.703261222421652e-05,
+      "loss": 3.4754,
+      "step": 17373
+    },
+    {
+      "epoch": 0.7315368421052632,
+      "grad_norm": 0.427734375,
+      "learning_rate": 8.700702436965635e-05,
+      "loss": 3.0133,
+      "step": 17374
+    },
+    {
+      "epoch": 0.7315789473684211,
+      "grad_norm": 0.443359375,
+      "learning_rate": 8.698143948463652e-05,
+      "loss": 3.3963,
+      "step": 17375
+    },
+    {
+      "epoch": 0.731621052631579,
+      "grad_norm": 0.447265625,
+      "learning_rate": 8.69558575696234e-05,
+      "loss": 3.1036,
+      "step": 17376
+    },
+    {
+      "epoch": 0.7316631578947368,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.693027862508299e-05,
+      "loss": 2.8572,
+      "step": 17377
+    },
+    {
+      "epoch": 0.7317052631578947,
+      "grad_norm": 0.435546875,
+      "learning_rate": 8.690470265148124e-05,
+      "loss": 3.2866,
+      "step": 17378
+    },
+    {
+      "epoch": 0.7317473684210526,
+      "grad_norm": 0.4296875,
+      "learning_rate": 8.687912964928432e-05,
+      "loss": 2.8592,
+      "step": 17379
+    },
+    {
+      "epoch": 0.7317894736842105,
+      "grad_norm": 0.4375,
+      "learning_rate": 8.685355961895783e-05,
+      "loss": 3.1077,
+      "step": 17380
+    },
+    {
+      "epoch": 0.7318315789473684,
+      "grad_norm": 0.42578125,
+      "learning_rate": 8.682799256096796e-05,
+      "loss": 2.8482,
+      "step": 17381
+    },
+    {
+      "epoch": 0.7318736842105263,
+      "grad_norm": 0.40234375,
+      "learning_rate": 8.68024284757802e-05,
+      "loss": 2.9467,
+      "step": 17382
+    },
+    {
+      "epoch": 0.7319157894736842,
+      "grad_norm": 0.466796875,
+      "learning_rate": 8.67768673638605e-05,
+      "loss": 3.4864,
+      "step": 17383
+    },
+    {
+      "epoch": 0.7319578947368421,
+      "grad_norm": 0.419921875,
+      "learning_rate": 8.675130922567454e-05,
+      "loss": 2.9723,
+      "step": 17384
+    },
+    {
+      "epoch": 0.732,
+      "grad_norm": 0.427734375,
+      "learning_rate": 8.672575406168781e-05,
+      "loss": 3.2359,
+      "step": 17385
+    },
+    {
+      "epoch": 0.7320421052631579,
+      "grad_norm": 0.44140625,
+      "learning_rate": 8.670020187236616e-05,
+      "loss": 2.9765,
+      "step": 17386
+    },
+    {
+      "epoch": 0.7320842105263158,
+      "grad_norm": 0.42578125,
+      "learning_rate": 8.667465265817478e-05,
+      "loss": 3.1162,
+      "step": 17387
+    },
+    {
+      "epoch": 0.7321263157894737,
+      "grad_norm": 0.44140625,
+      "learning_rate": 8.664910641957955e-05,
+      "loss": 2.7987,
+      "step": 17388
+    },
+    {
+      "epoch": 0.7321684210526316,
+      "grad_norm": 0.435546875,
+      "learning_rate": 8.662356315704545e-05,
+      "loss": 3.2022,
+      "step": 17389
+    },
+    {
+      "epoch": 0.7322105263157894,
+      "grad_norm": 0.4296875,
+      "learning_rate": 8.659802287103827e-05,
+      "loss": 3.2669,
+      "step": 17390
+    },
+    {
+      "epoch": 0.7322526315789474,
+      "grad_norm": 0.427734375,
+      "learning_rate": 8.657248556202293e-05,
+      "loss": 2.9025,
+      "step": 17391
+    },
+    {
+      "epoch": 0.7322947368421052,
+      "grad_norm": 0.478515625,
+      "learning_rate": 8.6546951230465e-05,
+      "loss": 3.0792,
+      "step": 17392
+    },
+    {
+      "epoch": 0.7323368421052632,
+      "grad_norm": 0.439453125,
+      "learning_rate": 8.652141987682955e-05,
+      "loss": 2.8577,
+      "step": 17393
+    },
+    {
+      "epoch": 0.732378947368421,
+      "grad_norm": 0.435546875,
+      "learning_rate": 8.649589150158165e-05,
+      "loss": 3.2648,
+      "step": 17394
+    },
+    {
+      "epoch": 0.732421052631579,
+      "grad_norm": 0.421875,
+      "learning_rate": 8.647036610518671e-05,
+      "loss": 3.1647,
+      "step": 17395
+    },
+    {
+      "epoch": 0.7324631578947368,
+      "grad_norm": 0.453125,
+      "learning_rate": 8.644484368810937e-05,
+      "loss": 2.877,
+      "step": 17396
+    },
+    {
+      "epoch": 0.7325052631578948,
+      "grad_norm": 0.44921875,
+      "learning_rate": 8.641932425081497e-05,
+      "loss": 3.1518,
+      "step": 17397
+    },
+    {
+      "epoch": 0.7325473684210526,
+      "grad_norm": 0.427734375,
+      "learning_rate": 8.63938077937681e-05,
+      "loss": 2.8324,
+      "step": 17398
+    },
+    {
+      "epoch": 0.7325894736842106,
+      "grad_norm": 0.423828125,
+      "learning_rate": 8.636829431743393e-05,
+      "loss": 2.6568,
+      "step": 17399
+    },
+    {
+      "epoch": 0.7326315789473684,
+      "grad_norm": 0.455078125,
+      "learning_rate": 8.634278382227717e-05,
+      "loss": 3.2829,
+      "step": 17400
+    },
+    {
+      "epoch": 0.7326736842105264,
+      "grad_norm": 0.423828125,
+      "learning_rate": 8.631727630876251e-05,
+      "loss": 2.9695,
+      "step": 17401
+    },
+    {
+      "epoch": 0.7327157894736842,
+      "grad_norm": 0.4140625,
+      "learning_rate": 8.629177177735495e-05,
+      "loss": 2.8694,
+      "step": 17402
+    },
+    {
+      "epoch": 0.732757894736842,
+      "grad_norm": 0.44921875,
+      "learning_rate": 8.626627022851876e-05,
+      "loss": 2.9942,
+      "step": 17403
+    },
+    {
+      "epoch": 0.7328,
+      "grad_norm": 0.419921875,
+      "learning_rate": 8.624077166271893e-05,
+      "loss": 3.0147,
+      "step": 17404
+    },
+    {
+      "epoch": 0.7328421052631579,
+      "grad_norm": 0.4296875,
+      "learning_rate": 8.621527608041968e-05,
+      "loss": 3.1181,
+      "step": 17405
+    },
+    {
+      "epoch": 0.7328842105263158,
+      "grad_norm": 0.408203125,
+      "learning_rate": 8.618978348208573e-05,
+      "loss": 2.9844,
+      "step": 17406
+    },
+    {
+      "epoch": 0.7329263157894736,
+      "grad_norm": 0.443359375,
+      "learning_rate": 8.616429386818145e-05,
+      "loss": 2.9553,
+      "step": 17407
+    },
+    {
+      "epoch": 0.7329684210526316,
+      "grad_norm": 0.421875,
+      "learning_rate": 8.613880723917123e-05,
+      "loss": 3.0633,
+      "step": 17408
+    },
+    {
+      "epoch": 0.7330105263157894,
+      "grad_norm": 0.41796875,
+      "learning_rate": 8.611332359551943e-05,
+      "loss": 3.2271,
+      "step": 17409
+    },
+    {
+      "epoch": 0.7330526315789474,
+      "grad_norm": 0.42578125,
+      "learning_rate": 8.608784293769023e-05,
+      "loss": 3.3197,
+      "step": 17410
+    },
+    {
+      "epoch": 0.7330947368421052,
+      "grad_norm": 0.4296875,
+      "learning_rate": 8.60623652661481e-05,
+      "loss": 3.4767,
+      "step": 17411
+    },
+    {
+      "epoch": 0.7331368421052632,
+      "grad_norm": 0.451171875,
+      "learning_rate": 8.603689058135688e-05,
+      "loss": 3.3598,
+      "step": 17412
+    },
+    {
+      "epoch": 0.733178947368421,
+      "grad_norm": 0.4765625,
+      "learning_rate": 8.601141888378097e-05,
+      "loss": 3.1161,
+      "step": 17413
+    },
+    {
+      "epoch": 0.733221052631579,
+      "grad_norm": 0.44140625,
+      "learning_rate": 8.598595017388433e-05,
+      "loss": 3.0147,
+      "step": 17414
+    },
+    {
+      "epoch": 0.7332631578947368,
+      "grad_norm": 0.4375,
+      "learning_rate": 8.596048445213096e-05,
+      "loss": 2.6017,
+      "step": 17415
+    },
+    {
+      "epoch": 0.7333052631578947,
+      "grad_norm": 0.412109375,
+      "learning_rate": 8.593502171898482e-05,
+      "loss": 3.2113,
+      "step": 17416
+    },
+    {
+      "epoch": 0.7333473684210526,
+      "grad_norm": 0.453125,
+      "learning_rate": 8.590956197490981e-05,
+      "loss": 3.2438,
+      "step": 17417
+    },
+    {
+      "epoch": 0.7333894736842105,
+      "grad_norm": 0.453125,
+      "learning_rate": 8.588410522036979e-05,
+      "loss": 3.2906,
+      "step": 17418
+    },
+    {
+      "epoch": 0.7334315789473684,
+      "grad_norm": 0.40234375,
+      "learning_rate": 8.585865145582847e-05,
+      "loss": 3.1117,
+      "step": 17419
+    },
+    {
+      "epoch": 0.7334736842105263,
+      "grad_norm": 0.51953125,
+      "learning_rate": 8.583320068174974e-05,
+      "loss": 3.1481,
+      "step": 17420
+    },
+    {
+      "epoch": 0.7335157894736842,
+      "grad_norm": 0.427734375,
+      "learning_rate": 8.580775289859721e-05,
+      "loss": 3.1312,
+      "step": 17421
+    },
+    {
+      "epoch": 0.7335578947368421,
+      "grad_norm": 0.431640625,
+      "learning_rate": 8.578230810683449e-05,
+      "loss": 3.1325,
+      "step": 17422
+    },
+    {
+      "epoch": 0.7336,
+      "grad_norm": 0.44921875,
+      "learning_rate": 8.57568663069252e-05,
+      "loss": 2.9373,
+      "step": 17423
+    },
+    {
+      "epoch": 0.7336421052631579,
+      "grad_norm": 0.46484375,
+      "learning_rate": 8.57314274993328e-05,
+      "loss": 2.9642,
+      "step": 17424
+    },
+    {
+      "epoch": 0.7336842105263158,
+      "grad_norm": 0.4140625,
+      "learning_rate": 8.57059916845207e-05,
+      "loss": 2.8016,
+      "step": 17425
+    },
+    {
+      "epoch": 0.7337263157894737,
+      "grad_norm": 0.45703125,
+      "learning_rate": 8.568055886295248e-05,
+      "loss": 3.4876,
+      "step": 17426
+    },
+    {
+      "epoch": 0.7337684210526316,
+      "grad_norm": 0.453125,
+      "learning_rate": 8.565512903509143e-05,
+      "loss": 3.1125,
+      "step": 17427
+    },
+    {
+      "epoch": 0.7338105263157895,
+      "grad_norm": 0.455078125,
+      "learning_rate": 8.562970220140079e-05,
+      "loss": 2.6941,
+      "step": 17428
+    },
+    {
+      "epoch": 0.7338526315789474,
+      "grad_norm": 0.421875,
+      "learning_rate": 8.560427836234388e-05,
+      "loss": 3.1189,
+      "step": 17429
+    },
+    {
+      "epoch": 0.7338947368421053,
+      "grad_norm": 0.458984375,
+      "learning_rate": 8.557885751838387e-05,
+      "loss": 3.0854,
+      "step": 17430
+    },
+    {
+      "epoch": 0.7339368421052631,
+      "grad_norm": 0.41796875,
+      "learning_rate": 8.555343966998385e-05,
+      "loss": 3.1352,
+      "step": 17431
+    },
+    {
+      "epoch": 0.7339789473684211,
+      "grad_norm": 0.453125,
+      "learning_rate": 8.552802481760691e-05,
+      "loss": 3.4583,
+      "step": 17432
+    },
+    {
+      "epoch": 0.7340210526315789,
+      "grad_norm": 0.451171875,
+      "learning_rate": 8.550261296171624e-05,
+      "loss": 3.0273,
+      "step": 17433
+    },
+    {
+      "epoch": 0.7340631578947369,
+      "grad_norm": 0.435546875,
+      "learning_rate": 8.547720410277454e-05,
+      "loss": 2.8165,
+      "step": 17434
+    },
+    {
+      "epoch": 0.7341052631578947,
+      "grad_norm": 0.439453125,
+      "learning_rate": 8.545179824124493e-05,
+      "loss": 2.7775,
+      "step": 17435
+    },
+    {
+      "epoch": 0.7341473684210527,
+      "grad_norm": 0.427734375,
+      "learning_rate": 8.542639537759025e-05,
+      "loss": 3.1867,
+      "step": 17436
+    },
+    {
+      "epoch": 0.7341894736842105,
+      "grad_norm": 0.43359375,
+      "learning_rate": 8.540099551227326e-05,
+      "loss": 2.9713,
+      "step": 17437
+    },
+    {
+      "epoch": 0.7342315789473685,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.537559864575675e-05,
+      "loss": 2.9672,
+      "step": 17438
+    },
+    {
+      "epoch": 0.7342736842105263,
+      "grad_norm": 0.458984375,
+      "learning_rate": 8.535020477850333e-05,
+      "loss": 2.9943,
+      "step": 17439
+    },
+    {
+      "epoch": 0.7343157894736843,
+      "grad_norm": 0.42578125,
+      "learning_rate": 8.532481391097587e-05,
+      "loss": 3.2691,
+      "step": 17440
+    },
+    {
+      "epoch": 0.7343578947368421,
+      "grad_norm": 0.4609375,
+      "learning_rate": 8.529942604363666e-05,
+      "loss": 3.1355,
+      "step": 17441
+    },
+    {
+      "epoch": 0.7344,
+      "grad_norm": 0.427734375,
+      "learning_rate": 8.527404117694849e-05,
+      "loss": 3.1819,
+      "step": 17442
+    },
+    {
+      "epoch": 0.7344421052631579,
+      "grad_norm": 0.43359375,
+      "learning_rate": 8.52486593113737e-05,
+      "loss": 3.0089,
+      "step": 17443
+    },
+    {
+      "epoch": 0.7344842105263157,
+      "grad_norm": 0.427734375,
+      "learning_rate": 8.522328044737479e-05,
+      "loss": 3.1019,
+      "step": 17444
+    },
+    {
+      "epoch": 0.7345263157894737,
+      "grad_norm": 0.4453125,
+      "learning_rate": 8.519790458541409e-05,
+      "loss": 3.2983,
+      "step": 17445
+    },
+    {
+      "epoch": 0.7345684210526315,
+      "grad_norm": 0.42578125,
+      "learning_rate": 8.517253172595388e-05,
+      "loss": 3.4151,
+      "step": 17446
+    },
+    {
+      "epoch": 0.7346105263157895,
+      "grad_norm": 0.439453125,
+      "learning_rate": 8.514716186945664e-05,
+      "loss": 3.4249,
+      "step": 17447
+    },
+    {
+      "epoch": 0.7346526315789473,
+      "grad_norm": 0.416015625,
+      "learning_rate": 8.512179501638423e-05,
+      "loss": 2.7593,
+      "step": 17448
+    },
+    {
+      "epoch": 0.7346947368421053,
+      "grad_norm": 0.4296875,
+      "learning_rate": 8.509643116719918e-05,
+      "loss": 2.776,
+      "step": 17449
+    },
+    {
+      "epoch": 0.7347368421052631,
+      "grad_norm": 0.412109375,
+      "learning_rate": 8.507107032236322e-05,
+      "loss": 3.0074,
+      "step": 17450
+    },
+    {
+      "epoch": 0.7347789473684211,
+      "grad_norm": 0.404296875,
+      "learning_rate": 8.504571248233867e-05,
+      "loss": 2.7538,
+      "step": 17451
+    },
+    {
+      "epoch": 0.7348210526315789,
+      "grad_norm": 0.4296875,
+      "learning_rate": 8.502035764758739e-05,
+      "loss": 3.3589,
+      "step": 17452
+    },
+    {
+      "epoch": 0.7348631578947369,
+      "grad_norm": 0.44921875,
+      "learning_rate": 8.499500581857131e-05,
+      "loss": 3.1937,
+      "step": 17453
+    },
+    {
+      "epoch": 0.7349052631578947,
+      "grad_norm": 0.4296875,
+      "learning_rate": 8.496965699575249e-05,
+      "loss": 2.8483,
+      "step": 17454
+    },
+    {
+      "epoch": 0.7349473684210527,
+      "grad_norm": 0.458984375,
+      "learning_rate": 8.494431117959247e-05,
+      "loss": 3.2585,
+      "step": 17455
+    },
+    {
+      "epoch": 0.7349894736842105,
+      "grad_norm": 0.421875,
+      "learning_rate": 8.491896837055333e-05,
+      "loss": 2.7329,
+      "step": 17456
+    },
+    {
+      "epoch": 0.7350315789473684,
+      "grad_norm": 0.47265625,
+      "learning_rate": 8.489362856909643e-05,
+      "loss": 3.1627,
+      "step": 17457
+    },
+    {
+      "epoch": 0.7350736842105263,
+      "grad_norm": 0.41796875,
+      "learning_rate": 8.486829177568375e-05,
+      "loss": 3.3207,
+      "step": 17458
+    },
+    {
+      "epoch": 0.7351157894736842,
+      "grad_norm": 0.44921875,
+      "learning_rate": 8.484295799077674e-05,
+      "loss": 2.8375,
+      "step": 17459
+    },
+    {
+      "epoch": 0.7351578947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 8.481762721483699e-05,
+      "loss": 2.9576,
+      "step": 17460
+    },
+    {
+      "epoch": 0.7352,
+      "grad_norm": 0.4609375,
+      "learning_rate": 8.479229944832601e-05,
+      "loss": 2.9608,
+      "step": 17461
+    },
+    {
+      "epoch": 0.7352421052631579,
+      "grad_norm": 0.4375,
+      "learning_rate": 8.476697469170514e-05,
+      "loss": 2.7885,
+      "step": 17462
+    },
+    {
+      "epoch": 0.7352842105263158,
+      "grad_norm": 0.43359375,
+      "learning_rate": 8.474165294543598e-05,
+      "loss": 3.1125,
+      "step": 17463
+    },
+    {
+      "epoch": 0.7353263157894737,
+      "grad_norm": 0.408203125,
+      "learning_rate": 8.471633420997959e-05,
+      "loss": 3.2301,
+      "step": 17464
+    },
+    {
+      "epoch": 0.7353684210526316,
+      "grad_norm": 0.4140625,
+      "learning_rate": 8.469101848579757e-05,
+      "loss": 3.1992,
+      "step": 17465
+    },
+    {
+      "epoch": 0.7354105263157895,
+      "grad_norm": 0.435546875,
+      "learning_rate": 8.466570577335078e-05,
+      "loss": 2.7817,
+      "step": 17466
+    },
+    {
+      "epoch": 0.7354526315789474,
+      "grad_norm": 0.42578125,
+      "learning_rate": 8.464039607310067e-05,
+      "loss": 2.8892,
+      "step": 17467
+    },
+    {
+      "epoch": 0.7354947368421053,
+      "grad_norm": 0.435546875,
+      "learning_rate": 8.461508938550825e-05,
+      "loss": 2.4723,
+      "step": 17468
+    },
+    {
+      "epoch": 0.7355368421052632,
+      "grad_norm": 0.4140625,
+      "learning_rate": 8.458978571103448e-05,
+      "loss": 3.1733,
+      "step": 17469
+    },
+    {
+      "epoch": 0.735578947368421,
+      "grad_norm": 0.427734375,
+      "learning_rate": 8.456448505014064e-05,
+      "loss": 3.1467,
+      "step": 17470
+    },
+    {
+      "epoch": 0.735621052631579,
+      "grad_norm": 0.4609375,
+      "learning_rate": 8.453918740328733e-05,
+      "loss": 3.214,
+      "step": 17471
+    },
+    {
+      "epoch": 0.7356631578947368,
+      "grad_norm": 0.4140625,
+      "learning_rate": 8.45138927709358e-05,
+      "loss": 3.3429,
+      "step": 17472
+    },
+    {
+      "epoch": 0.7357052631578948,
+      "grad_norm": 0.42578125,
+      "learning_rate": 8.448860115354653e-05,
+      "loss": 3.1089,
+      "step": 17473
+    },
+    {
+      "epoch": 0.7357473684210526,
+      "grad_norm": 0.484375,
+      "learning_rate": 8.446331255158057e-05,
+      "loss": 2.9917,
+      "step": 17474
+    },
+    {
+      "epoch": 0.7357894736842105,
+      "grad_norm": 0.431640625,
+      "learning_rate": 8.443802696549857e-05,
+      "loss": 3.214,
+      "step": 17475
+    },
+    {
+      "epoch": 0.7358315789473684,
+      "grad_norm": 0.42578125,
+      "learning_rate": 8.441274439576115e-05,
+      "loss": 2.8693,
+      "step": 17476
+    },
+    {
+      "epoch": 0.7358736842105263,
+      "grad_norm": 0.44140625,
+      "learning_rate": 8.4387464842829e-05,
+      "loss": 3.1647,
+      "step": 17477
+    },
+    {
+      "epoch": 0.7359157894736842,
+      "grad_norm": 0.408203125,
+      "learning_rate": 8.436218830716258e-05,
+      "loss": 2.6969,
+      "step": 17478
+    },
+    {
+      "epoch": 0.7359578947368421,
+      "grad_norm": 0.43359375,
+      "learning_rate": 8.433691478922254e-05,
+      "loss": 2.813,
+      "step": 17479
+    },
+    {
+      "epoch": 0.736,
+      "grad_norm": 0.427734375,
+      "learning_rate": 8.431164428946927e-05,
+      "loss": 3.3372,
+      "step": 17480
+    },
+    {
+      "epoch": 0.7360421052631579,
+      "grad_norm": 0.458984375,
+      "learning_rate": 8.428637680836316e-05,
+      "loss": 3.085,
+      "step": 17481
+    },
+    {
+      "epoch": 0.7360842105263158,
+      "grad_norm": 0.4375,
+      "learning_rate": 8.426111234636455e-05,
+      "loss": 3.0584,
+      "step": 17482
+    },
+    {
+      "epoch": 0.7361263157894737,
+      "grad_norm": 0.435546875,
+      "learning_rate": 8.423585090393376e-05,
+      "loss": 3.6094,
+      "step": 17483
+    },
+    {
+      "epoch": 0.7361684210526316,
+      "grad_norm": 0.439453125,
+      "learning_rate": 8.421059248153098e-05,
+      "loss": 2.8345,
+      "step": 17484
+    },
+    {
+      "epoch": 0.7362105263157894,
+      "grad_norm": 0.4296875,
+      "learning_rate": 8.418533707961634e-05,
+      "loss": 3.0573,
+      "step": 17485
+    },
+    {
+      "epoch": 0.7362526315789474,
+      "grad_norm": 0.443359375,
+      "learning_rate": 8.416008469865013e-05,
+      "loss": 3.5585,
+      "step": 17486
+    },
+    {
+      "epoch": 0.7362947368421052,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.413483533909231e-05,
+      "loss": 3.2086,
+      "step": 17487
+    },
+    {
+      "epoch": 0.7363368421052632,
+      "grad_norm": 0.423828125,
+      "learning_rate": 8.41095890014029e-05,
+      "loss": 3.4099,
+      "step": 17488
+    },
+    {
+      "epoch": 0.736378947368421,
+      "grad_norm": 0.435546875,
+      "learning_rate": 8.408434568604187e-05,
+      "loss": 3.4943,
+      "step": 17489
+    },
+    {
+      "epoch": 0.736421052631579,
+      "grad_norm": 0.423828125,
+      "learning_rate": 8.405910539346911e-05,
+      "loss": 3.2247,
+      "step": 17490
+    },
+    {
+      "epoch": 0.7364631578947368,
+      "grad_norm": 0.451171875,
+      "learning_rate": 8.403386812414448e-05,
+      "loss": 3.1184,
+      "step": 17491
+    },
+    {
+      "epoch": 0.7365052631578948,
+      "grad_norm": 0.439453125,
+      "learning_rate": 8.400863387852778e-05,
+      "loss": 3.0231,
+      "step": 17492
+    },
+    {
+      "epoch": 0.7365473684210526,
+      "grad_norm": 0.4140625,
+      "learning_rate": 8.398340265707866e-05,
+      "loss": 2.3049,
+      "step": 17493
+    },
+    {
+      "epoch": 0.7365894736842106,
+      "grad_norm": 0.431640625,
+      "learning_rate": 8.395817446025698e-05,
+      "loss": 2.8502,
+      "step": 17494
+    },
+    {
+      "epoch": 0.7366315789473684,
+      "grad_norm": 0.435546875,
+      "learning_rate": 8.393294928852224e-05,
+      "loss": 3.0826,
+      "step": 17495
+    },
+    {
+      "epoch": 0.7366736842105264,
+      "grad_norm": 0.462890625,
+      "learning_rate": 8.390772714233405e-05,
+      "loss": 2.8338,
+      "step": 17496
+    },
+    {
+      "epoch": 0.7367157894736842,
+      "grad_norm": 0.4453125,
+      "learning_rate": 8.388250802215191e-05,
+      "loss": 2.6943,
+      "step": 17497
+    },
+    {
+      "epoch": 0.7367578947368421,
+      "grad_norm": 0.3984375,
+      "learning_rate": 8.38572919284353e-05,
+      "loss": 2.7742,
+      "step": 17498
+    },
+    {
+      "epoch": 0.7368,
+      "grad_norm": 0.455078125,
+      "learning_rate": 8.383207886164365e-05,
+      "loss": 3.2979,
+      "step": 17499
+    },
+    {
+      "epoch": 0.7368421052631579,
+      "grad_norm": 0.439453125,
+      "learning_rate": 8.380686882223618e-05,
+      "loss": 3.0861,
+      "step": 17500
+    },
+    {
+      "epoch": 0.7368842105263158,
+      "grad_norm": 0.4375,
+      "learning_rate": 8.378166181067238e-05,
+      "loss": 2.9101,
+      "step": 17501
+    },
+    {
+      "epoch": 0.7369263157894737,
+      "grad_norm": 0.44140625,
+      "learning_rate": 8.375645782741137e-05,
+      "loss": 3.4054,
+      "step": 17502
+    },
+    {
+      "epoch": 0.7369684210526316,
+      "grad_norm": 0.419921875,
+      "learning_rate": 8.373125687291241e-05,
+      "loss": 3.0726,
+      "step": 17503
+    },
+    {
+      "epoch": 0.7370105263157894,
+      "grad_norm": 0.412109375,
+      "learning_rate": 8.370605894763459e-05,
+      "loss": 2.8369,
+      "step": 17504
+    },
+    {
+      "epoch": 0.7370526315789474,
+      "grad_norm": 0.4296875,
+      "learning_rate": 8.368086405203695e-05,
+      "loss": 3.2243,
+      "step": 17505
+    },
+    {
+      "epoch": 0.7370947368421052,
+      "grad_norm": 0.41015625,
+      "learning_rate": 8.36556721865786e-05,
+      "loss": 2.6013,
+      "step": 17506
+    },
+    {
+      "epoch": 0.7371368421052632,
+      "grad_norm": 0.423828125,
+      "learning_rate": 8.363048335171833e-05,
+      "loss": 3.0765,
+      "step": 17507
+    },
+    {
+      "epoch": 0.737178947368421,
+      "grad_norm": 0.4296875,
+      "learning_rate": 8.360529754791538e-05,
+      "loss": 3.3717,
+      "step": 17508
+    },
+    {
+      "epoch": 0.737221052631579,
+      "grad_norm": 0.40625,
+      "learning_rate": 8.35801147756282e-05,
+      "loss": 2.8376,
+      "step": 17509
+    },
+    {
+      "epoch": 0.7372631578947368,
+      "grad_norm": 0.451171875,
+      "learning_rate": 8.355493503531592e-05,
+      "loss": 2.9544,
+      "step": 17510
+    },
+    {
+      "epoch": 0.7373052631578947,
+      "grad_norm": 0.416015625,
+      "learning_rate": 8.352975832743712e-05,
+      "loss": 3.192,
+      "step": 17511
+    },
+    {
+      "epoch": 0.7373473684210526,
+      "grad_norm": 0.451171875,
+      "learning_rate": 8.350458465245056e-05,
+      "loss": 3.2392,
+      "step": 17512
+    },
+    {
+      "epoch": 0.7373894736842105,
+      "grad_norm": 0.43359375,
+      "learning_rate": 8.347941401081483e-05,
+      "loss": 3.1794,
+      "step": 17513
+    },
+    {
+      "epoch": 0.7374315789473684,
+      "grad_norm": 0.43359375,
+      "learning_rate": 8.345424640298843e-05,
+      "loss": 2.9829,
+      "step": 17514
+    },
+    {
+      "epoch": 0.7374736842105263,
+      "grad_norm": 0.423828125,
+      "learning_rate": 8.342908182943016e-05,
+      "loss": 2.9809,
+      "step": 17515
+    },
+    {
+      "epoch": 0.7375157894736842,
+      "grad_norm": 0.435546875,
+      "learning_rate": 8.340392029059812e-05,
+      "loss": 2.8339,
+      "step": 17516
+    },
+    {
+      "epoch": 0.7375578947368421,
+      "grad_norm": 0.482421875,
+      "learning_rate": 8.3378761786951e-05,
+      "loss": 3.1362,
+      "step": 17517
+    },
+    {
+      "epoch": 0.7376,
+      "grad_norm": 0.416015625,
+      "learning_rate": 8.335360631894709e-05,
+      "loss": 2.7523,
+      "step": 17518
+    },
+    {
+      "epoch": 0.7376421052631579,
+      "grad_norm": 0.421875,
+      "learning_rate": 8.332845388704466e-05,
+      "loss": 3.086,
+      "step": 17519
+    },
+    {
+      "epoch": 0.7376842105263158,
+      "grad_norm": 0.431640625,
+      "learning_rate": 8.330330449170195e-05,
+      "loss": 3.1042,
+      "step": 17520
+    },
+    {
+      "epoch": 0.7377263157894737,
+      "grad_norm": 0.4140625,
+      "learning_rate": 8.32781581333771e-05,
+      "loss": 2.8653,
+      "step": 17521
+    },
+    {
+      "epoch": 0.7377684210526316,
+      "grad_norm": 0.416015625,
+      "learning_rate": 8.325301481252848e-05,
+      "loss": 2.6848,
+      "step": 17522
+    },
+    {
+      "epoch": 0.7378105263157895,
+      "grad_norm": 0.4453125,
+      "learning_rate": 8.322787452961387e-05,
+      "loss": 3.2506,
+      "step": 17523
+    },
+    {
+      "epoch": 0.7378526315789473,
+      "grad_norm": 0.42578125,
+      "learning_rate": 8.320273728509159e-05,
+      "loss": 2.9591,
+      "step": 17524
+    },
+    {
+      "epoch": 0.7378947368421053,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.317760307941927e-05,
+      "loss": 2.7788,
+      "step": 17525
+    },
+    {
+      "epoch": 0.7379368421052631,
+      "grad_norm": 0.44140625,
+      "learning_rate": 8.315247191305511e-05,
+      "loss": 3.026,
+      "step": 17526
+    },
+    {
+      "epoch": 0.7379789473684211,
+      "grad_norm": 0.412109375,
+      "learning_rate": 8.31273437864569e-05,
+      "loss": 2.8661,
+      "step": 17527
+    },
+    {
+      "epoch": 0.7380210526315789,
+      "grad_norm": 0.451171875,
+      "learning_rate": 8.31022187000823e-05,
+      "loss": 3.0473,
+      "step": 17528
+    },
+    {
+      "epoch": 0.7380631578947369,
+      "grad_norm": 0.427734375,
+      "learning_rate": 8.307709665438937e-05,
+      "loss": 3.0359,
+      "step": 17529
+    },
+    {
+      "epoch": 0.7381052631578947,
+      "grad_norm": 0.44140625,
+      "learning_rate": 8.305197764983541e-05,
+      "loss": 2.9506,
+      "step": 17530
+    },
+    {
+      "epoch": 0.7381473684210527,
+      "grad_norm": 0.451171875,
+      "learning_rate": 8.302686168687845e-05,
+      "loss": 3.1856,
+      "step": 17531
+    },
+    {
+      "epoch": 0.7381894736842105,
+      "grad_norm": 0.400390625,
+      "learning_rate": 8.300174876597575e-05,
+      "loss": 3.0519,
+      "step": 17532
+    },
+    {
+      "epoch": 0.7382315789473685,
+      "grad_norm": 0.443359375,
+      "learning_rate": 8.2976638887585e-05,
+      "loss": 3.1356,
+      "step": 17533
+    },
+    {
+      "epoch": 0.7382736842105263,
+      "grad_norm": 0.43359375,
+      "learning_rate": 8.295153205216369e-05,
+      "loss": 2.8951,
+      "step": 17534
+    },
+    {
+      "epoch": 0.7383157894736843,
+      "grad_norm": 0.4375,
+      "learning_rate": 8.292642826016916e-05,
+      "loss": 2.9037,
+      "step": 17535
+    },
+    {
+      "epoch": 0.7383578947368421,
+      "grad_norm": 0.439453125,
+      "learning_rate": 8.290132751205881e-05,
+      "loss": 3.191,
+      "step": 17536
+    },
+    {
+      "epoch": 0.7384,
+      "grad_norm": 0.431640625,
+      "learning_rate": 8.287622980828988e-05,
+      "loss": 3.085,
+      "step": 17537
+    },
+    {
+      "epoch": 0.7384421052631579,
+      "grad_norm": 0.412109375,
+      "learning_rate": 8.285113514931983e-05,
+      "loss": 3.3099,
+      "step": 17538
+    },
+    {
+      "epoch": 0.7384842105263157,
+      "grad_norm": 0.412109375,
+      "learning_rate": 8.282604353560552e-05,
+      "loss": 3.357,
+      "step": 17539
+    },
+    {
+      "epoch": 0.7385263157894737,
+      "grad_norm": 0.4296875,
+      "learning_rate": 8.280095496760439e-05,
+      "loss": 3.0421,
+      "step": 17540
+    },
+    {
+      "epoch": 0.7385684210526315,
+      "grad_norm": 0.42578125,
+      "learning_rate": 8.27758694457734e-05,
+      "loss": 3.1583,
+      "step": 17541
+    },
+    {
+      "epoch": 0.7386105263157895,
+      "grad_norm": 0.416015625,
+      "learning_rate": 8.27507869705696e-05,
+      "loss": 2.7443,
+      "step": 17542
+    },
+    {
+      "epoch": 0.7386526315789473,
+      "grad_norm": 0.46484375,
+      "learning_rate": 8.272570754244993e-05,
+      "loss": 3.1097,
+      "step": 17543
+    },
+    {
+      "epoch": 0.7386947368421053,
+      "grad_norm": 0.439453125,
+      "learning_rate": 8.270063116187127e-05,
+      "loss": 3.0966,
+      "step": 17544
+    },
+    {
+      "epoch": 0.7387368421052631,
+      "grad_norm": 0.416015625,
+      "learning_rate": 8.267555782929067e-05,
+      "loss": 3.2179,
+      "step": 17545
+    },
+    {
+      "epoch": 0.7387789473684211,
+      "grad_norm": 0.453125,
+      "learning_rate": 8.265048754516466e-05,
+      "loss": 2.6872,
+      "step": 17546
+    },
+    {
+      "epoch": 0.7388210526315789,
+      "grad_norm": 0.41796875,
+      "learning_rate": 8.262542030995024e-05,
+      "loss": 3.0463,
+      "step": 17547
+    },
+    {
+      "epoch": 0.7388631578947369,
+      "grad_norm": 0.40625,
+      "learning_rate": 8.260035612410399e-05,
+      "loss": 3.2567,
+      "step": 17548
+    },
+    {
+      "epoch": 0.7389052631578947,
+      "grad_norm": 0.431640625,
+      "learning_rate": 8.257529498808256e-05,
+      "loss": 2.7769,
+      "step": 17549
+    },
+    {
+      "epoch": 0.7389473684210527,
+      "grad_norm": 0.4296875,
+      "learning_rate": 8.255023690234257e-05,
+      "loss": 3.2597,
+      "step": 17550
+    },
+    {
+      "epoch": 0.7389894736842105,
+      "grad_norm": 0.42578125,
+      "learning_rate": 8.252518186734048e-05,
+      "loss": 3.1473,
+      "step": 17551
+    },
+    {
+      "epoch": 0.7390315789473684,
+      "grad_norm": 0.455078125,
+      "learning_rate": 8.250012988353282e-05,
+      "loss": 2.8691,
+      "step": 17552
+    },
+    {
+      "epoch": 0.7390736842105263,
+      "grad_norm": 0.427734375,
+      "learning_rate": 8.24750809513759e-05,
+      "loss": 3.0088,
+      "step": 17553
+    },
+    {
+      "epoch": 0.7391157894736842,
+      "grad_norm": 0.41796875,
+      "learning_rate": 8.245003507132625e-05,
+      "loss": 3.2294,
+      "step": 17554
+    },
+    {
+      "epoch": 0.7391578947368421,
+      "grad_norm": 0.4375,
+      "learning_rate": 8.24249922438401e-05,
+      "loss": 2.9939,
+      "step": 17555
+    },
+    {
+      "epoch": 0.7392,
+      "grad_norm": 0.447265625,
+      "learning_rate": 8.239995246937371e-05,
+      "loss": 2.546,
+      "step": 17556
+    },
+    {
+      "epoch": 0.7392421052631579,
+      "grad_norm": 0.453125,
+      "learning_rate": 8.237491574838327e-05,
+      "loss": 3.105,
+      "step": 17557
+    },
+    {
+      "epoch": 0.7392842105263158,
+      "grad_norm": 0.431640625,
+      "learning_rate": 8.234988208132487e-05,
+      "loss": 3.1455,
+      "step": 17558
+    },
+    {
+      "epoch": 0.7393263157894737,
+      "grad_norm": 0.427734375,
+      "learning_rate": 8.232485146865468e-05,
+      "loss": 2.7145,
+      "step": 17559
+    },
+    {
+      "epoch": 0.7393684210526316,
+      "grad_norm": 0.41015625,
+      "learning_rate": 8.22998239108286e-05,
+      "loss": 2.9565,
+      "step": 17560
+    },
+    {
+      "epoch": 0.7394105263157895,
+      "grad_norm": 0.41796875,
+      "learning_rate": 8.227479940830277e-05,
+      "loss": 2.9454,
+      "step": 17561
+    },
+    {
+      "epoch": 0.7394526315789474,
+      "grad_norm": 0.419921875,
+      "learning_rate": 8.224977796153302e-05,
+      "loss": 2.2829,
+      "step": 17562
+    },
+    {
+      "epoch": 0.7394947368421053,
+      "grad_norm": 0.43359375,
+      "learning_rate": 8.222475957097523e-05,
+      "loss": 3.337,
+      "step": 17563
+    },
+    {
+      "epoch": 0.7395368421052632,
+      "grad_norm": 0.427734375,
+      "learning_rate": 8.219974423708514e-05,
+      "loss": 3.1087,
+      "step": 17564
+    },
+    {
+      "epoch": 0.739578947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 8.21747319603186e-05,
+      "loss": 2.9359,
+      "step": 17565
+    },
+    {
+      "epoch": 0.739621052631579,
+      "grad_norm": 0.40625,
+      "learning_rate": 8.214972274113125e-05,
+      "loss": 2.619,
+      "step": 17566
+    },
+    {
+      "epoch": 0.7396631578947368,
+      "grad_norm": 0.388671875,
+      "learning_rate": 8.212471657997875e-05,
+      "loss": 2.6439,
+      "step": 17567
+    },
+    {
+      "epoch": 0.7397052631578948,
+      "grad_norm": 0.421875,
+      "learning_rate": 8.209971347731656e-05,
+      "loss": 2.7097,
+      "step": 17568
+    },
+    {
+      "epoch": 0.7397473684210526,
+      "grad_norm": 0.45703125,
+      "learning_rate": 8.20747134336004e-05,
+      "loss": 3.1776,
+      "step": 17569
+    },
+    {
+      "epoch": 0.7397894736842106,
+      "grad_norm": 0.453125,
+      "learning_rate": 8.20497164492857e-05,
+      "loss": 3.1628,
+      "step": 17570
+    },
+    {
+      "epoch": 0.7398315789473684,
+      "grad_norm": 0.4453125,
+      "learning_rate": 8.20247225248278e-05,
+      "loss": 2.8349,
+      "step": 17571
+    },
+    {
+      "epoch": 0.7398736842105264,
+      "grad_norm": 0.40625,
+      "learning_rate": 8.199973166068206e-05,
+      "loss": 2.7887,
+      "step": 17572
+    },
+    {
+      "epoch": 0.7399157894736842,
+      "grad_norm": 0.416015625,
+      "learning_rate": 8.197474385730377e-05,
+      "loss": 2.8789,
+      "step": 17573
+    },
+    {
+      "epoch": 0.7399578947368421,
+      "grad_norm": 0.40625,
+      "learning_rate": 8.19497591151484e-05,
+      "loss": 2.6945,
+      "step": 17574
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4140625,
+      "learning_rate": 8.192477743467078e-05,
+      "loss": 2.8433,
+      "step": 17575
+    },
+    {
+      "epoch": 0.740042105263158,
+      "grad_norm": 0.4765625,
+      "learning_rate": 8.189979881632634e-05,
+      "loss": 2.8476,
+      "step": 17576
+    },
+    {
+      "epoch": 0.7400842105263158,
+      "grad_norm": 0.431640625,
+      "learning_rate": 8.187482326057002e-05,
+      "loss": 2.9385,
+      "step": 17577
+    },
+    {
+      "epoch": 0.7401263157894736,
+      "grad_norm": 0.41796875,
+      "learning_rate": 8.184985076785692e-05,
+      "loss": 3.172,
+      "step": 17578
+    },
+    {
+      "epoch": 0.7401684210526316,
+      "grad_norm": 0.4140625,
+      "learning_rate": 8.182488133864197e-05,
+      "loss": 2.9247,
+      "step": 17579
+    },
+    {
+      "epoch": 0.7402105263157894,
+      "grad_norm": 0.4609375,
+      "learning_rate": 8.179991497337996e-05,
+      "loss": 3.6467,
+      "step": 17580
+    },
+    {
+      "epoch": 0.7402526315789474,
+      "grad_norm": 0.4140625,
+      "learning_rate": 8.177495167252608e-05,
+      "loss": 2.848,
+      "step": 17581
+    },
+    {
+      "epoch": 0.7402947368421052,
+      "grad_norm": 0.43359375,
+      "learning_rate": 8.174999143653475e-05,
+      "loss": 3.0746,
+      "step": 17582
+    },
+    {
+      "epoch": 0.7403368421052632,
+      "grad_norm": 0.42578125,
+      "learning_rate": 8.172503426586103e-05,
+      "loss": 3.3865,
+      "step": 17583
+    },
+    {
+      "epoch": 0.740378947368421,
+      "grad_norm": 0.451171875,
+      "learning_rate": 8.170008016095934e-05,
+      "loss": 2.9487,
+      "step": 17584
+    },
+    {
+      "epoch": 0.740421052631579,
+      "grad_norm": 0.41015625,
+      "learning_rate": 8.16751291222845e-05,
+      "loss": 3.0877,
+      "step": 17585
+    },
+    {
+      "epoch": 0.7404631578947368,
+      "grad_norm": 0.439453125,
+      "learning_rate": 8.165018115029105e-05,
+      "loss": 3.1764,
+      "step": 17586
+    },
+    {
+      "epoch": 0.7405052631578948,
+      "grad_norm": 0.44140625,
+      "learning_rate": 8.162523624543342e-05,
+      "loss": 2.9305,
+      "step": 17587
+    },
+    {
+      "epoch": 0.7405473684210526,
+      "grad_norm": 0.455078125,
+      "learning_rate": 8.16002944081663e-05,
+      "loss": 2.6253,
+      "step": 17588
+    },
+    {
+      "epoch": 0.7405894736842106,
+      "grad_norm": 0.43359375,
+      "learning_rate": 8.157535563894381e-05,
+      "loss": 3.064,
+      "step": 17589
+    },
+    {
+      "epoch": 0.7406315789473684,
+      "grad_norm": 0.44921875,
+      "learning_rate": 8.15504199382206e-05,
+      "loss": 2.7958,
+      "step": 17590
+    },
+    {
+      "epoch": 0.7406736842105263,
+      "grad_norm": 0.3984375,
+      "learning_rate": 8.152548730645065e-05,
+      "loss": 2.5647,
+      "step": 17591
+    },
+    {
+      "epoch": 0.7407157894736842,
+      "grad_norm": 0.45703125,
+      "learning_rate": 8.150055774408847e-05,
+      "loss": 3.0793,
+      "step": 17592
+    },
+    {
+      "epoch": 0.7407578947368421,
+      "grad_norm": 0.46875,
+      "learning_rate": 8.147563125158814e-05,
+      "loss": 3.1509,
+      "step": 17593
+    },
+    {
+      "epoch": 0.7408,
+      "grad_norm": 0.439453125,
+      "learning_rate": 8.145070782940377e-05,
+      "loss": 2.9246,
+      "step": 17594
+    },
+    {
+      "epoch": 0.7408421052631579,
+      "grad_norm": 0.4375,
+      "learning_rate": 8.142578747798951e-05,
+      "loss": 3.0869,
+      "step": 17595
+    },
+    {
+      "epoch": 0.7408842105263158,
+      "grad_norm": 0.423828125,
+      "learning_rate": 8.140087019779924e-05,
+      "loss": 3.3136,
+      "step": 17596
+    },
+    {
+      "epoch": 0.7409263157894737,
+      "grad_norm": 0.43359375,
+      "learning_rate": 8.137595598928718e-05,
+      "loss": 3.2361,
+      "step": 17597
+    },
+    {
+      "epoch": 0.7409684210526316,
+      "grad_norm": 0.4296875,
+      "learning_rate": 8.135104485290693e-05,
+      "loss": 3.0041,
+      "step": 17598
+    },
+    {
+      "epoch": 0.7410105263157895,
+      "grad_norm": 0.439453125,
+      "learning_rate": 8.132613678911263e-05,
+      "loss": 2.9811,
+      "step": 17599
+    },
+    {
+      "epoch": 0.7410526315789474,
+      "grad_norm": 0.4140625,
+      "learning_rate": 8.130123179835777e-05,
+      "loss": 3.1249,
+      "step": 17600
+    },
+    {
+      "epoch": 0.7410947368421053,
+      "grad_norm": 0.427734375,
+      "learning_rate": 8.127632988109634e-05,
+      "loss": 3.2064,
+      "step": 17601
+    },
+    {
+      "epoch": 0.7411368421052632,
+      "grad_norm": 0.4296875,
+      "learning_rate": 8.125143103778196e-05,
+      "loss": 2.7307,
+      "step": 17602
+    },
+    {
+      "epoch": 0.741178947368421,
+      "grad_norm": 0.408203125,
+      "learning_rate": 8.122653526886814e-05,
+      "loss": 3.3496,
+      "step": 17603
+    },
+    {
+      "epoch": 0.741221052631579,
+      "grad_norm": 0.412109375,
+      "learning_rate": 8.12016425748087e-05,
+      "loss": 2.8468,
+      "step": 17604
+    },
+    {
+      "epoch": 0.7412631578947368,
+      "grad_norm": 0.427734375,
+      "learning_rate": 8.117675295605684e-05,
+      "loss": 2.8568,
+      "step": 17605
+    },
+    {
+      "epoch": 0.7413052631578947,
+      "grad_norm": 0.439453125,
+      "learning_rate": 8.115186641306636e-05,
+      "loss": 3.268,
+      "step": 17606
+    },
+    {
+      "epoch": 0.7413473684210526,
+      "grad_norm": 0.404296875,
+      "learning_rate": 8.112698294629032e-05,
+      "loss": 3.4852,
+      "step": 17607
+    },
+    {
+      "epoch": 0.7413894736842105,
+      "grad_norm": 0.4375,
+      "learning_rate": 8.110210255618231e-05,
+      "loss": 3.6075,
+      "step": 17608
+    },
+    {
+      "epoch": 0.7414315789473684,
+      "grad_norm": 0.423828125,
+      "learning_rate": 8.107722524319552e-05,
+      "loss": 3.1941,
+      "step": 17609
+    },
+    {
+      "epoch": 0.7414736842105263,
+      "grad_norm": 0.43359375,
+      "learning_rate": 8.105235100778325e-05,
+      "loss": 3.1158,
+      "step": 17610
+    },
+    {
+      "epoch": 0.7415157894736842,
+      "grad_norm": 0.431640625,
+      "learning_rate": 8.10274798503986e-05,
+      "loss": 3.2473,
+      "step": 17611
+    },
+    {
+      "epoch": 0.7415578947368421,
+      "grad_norm": 0.458984375,
+      "learning_rate": 8.100261177149468e-05,
+      "loss": 3.1386,
+      "step": 17612
+    },
+    {
+      "epoch": 0.7416,
+      "grad_norm": 0.435546875,
+      "learning_rate": 8.097774677152475e-05,
+      "loss": 3.4269,
+      "step": 17613
+    },
+    {
+      "epoch": 0.7416421052631579,
+      "grad_norm": 0.419921875,
+      "learning_rate": 8.095288485094155e-05,
+      "loss": 3.2482,
+      "step": 17614
+    },
+    {
+      "epoch": 0.7416842105263158,
+      "grad_norm": 0.42578125,
+      "learning_rate": 8.092802601019822e-05,
+      "loss": 2.6936,
+      "step": 17615
+    },
+    {
+      "epoch": 0.7417263157894737,
+      "grad_norm": 0.431640625,
+      "learning_rate": 8.09031702497476e-05,
+      "loss": 3.1977,
+      "step": 17616
+    },
+    {
+      "epoch": 0.7417684210526316,
+      "grad_norm": 0.4296875,
+      "learning_rate": 8.087831757004258e-05,
+      "loss": 3.1649,
+      "step": 17617
+    },
+    {
+      "epoch": 0.7418105263157895,
+      "grad_norm": 0.412109375,
+      "learning_rate": 8.085346797153586e-05,
+      "loss": 3.0625,
+      "step": 17618
+    },
+    {
+      "epoch": 0.7418526315789473,
+      "grad_norm": 0.40625,
+      "learning_rate": 8.082862145468015e-05,
+      "loss": 3.4446,
+      "step": 17619
+    },
+    {
+      "epoch": 0.7418947368421053,
+      "grad_norm": 0.427734375,
+      "learning_rate": 8.080377801992835e-05,
+      "loss": 3.2164,
+      "step": 17620
+    },
+    {
+      "epoch": 0.7419368421052631,
+      "grad_norm": 0.451171875,
+      "learning_rate": 8.077893766773276e-05,
+      "loss": 2.9525,
+      "step": 17621
+    },
+    {
+      "epoch": 0.7419789473684211,
+      "grad_norm": 0.482421875,
+      "learning_rate": 8.075410039854616e-05,
+      "loss": 3.4812,
+      "step": 17622
+    },
+    {
+      "epoch": 0.7420210526315789,
+      "grad_norm": 0.4296875,
+      "learning_rate": 8.072926621282104e-05,
+      "loss": 3.3057,
+      "step": 17623
+    },
+    {
+      "epoch": 0.7420631578947369,
+      "grad_norm": 0.423828125,
+      "learning_rate": 8.070443511100975e-05,
+      "loss": 3.3029,
+      "step": 17624
+    },
+    {
+      "epoch": 0.7421052631578947,
+      "grad_norm": 0.44140625,
+      "learning_rate": 8.067960709356478e-05,
+      "loss": 3.3961,
+      "step": 17625
+    },
+    {
+      "epoch": 0.7421473684210527,
+      "grad_norm": 4.75,
+      "learning_rate": 8.065478216093841e-05,
+      "loss": 3.2198,
+      "step": 17626
+    },
+    {
+      "epoch": 0.7421894736842105,
+      "grad_norm": 0.4375,
+      "learning_rate": 8.062996031358286e-05,
+      "loss": 3.3833,
+      "step": 17627
+    },
+    {
+      "epoch": 0.7422315789473685,
+      "grad_norm": 0.4453125,
+      "learning_rate": 8.060514155195048e-05,
+      "loss": 2.692,
+      "step": 17628
+    },
+    {
+      "epoch": 0.7422736842105263,
+      "grad_norm": 0.43359375,
+      "learning_rate": 8.058032587649341e-05,
+      "loss": 3.2869,
+      "step": 17629
+    },
+    {
+      "epoch": 0.7423157894736843,
+      "grad_norm": 0.43359375,
+      "learning_rate": 8.055551328766377e-05,
+      "loss": 3.0522,
+      "step": 17630
+    },
+    {
+      "epoch": 0.7423578947368421,
+      "grad_norm": 0.44921875,
+      "learning_rate": 8.053070378591357e-05,
+      "loss": 3.1317,
+      "step": 17631
+    },
+    {
+      "epoch": 0.7424,
+      "grad_norm": 0.421875,
+      "learning_rate": 8.050589737169483e-05,
+      "loss": 2.8788,
+      "step": 17632
+    },
+    {
+      "epoch": 0.7424421052631579,
+      "grad_norm": 0.4453125,
+      "learning_rate": 8.048109404545948e-05,
+      "loss": 3.075,
+      "step": 17633
+    },
+    {
+      "epoch": 0.7424842105263157,
+      "grad_norm": 0.427734375,
+      "learning_rate": 8.045629380765934e-05,
+      "loss": 3.1869,
+      "step": 17634
+    },
+    {
+      "epoch": 0.7425263157894737,
+      "grad_norm": 0.423828125,
+      "learning_rate": 8.043149665874641e-05,
+      "loss": 2.9315,
+      "step": 17635
+    },
+    {
+      "epoch": 0.7425684210526315,
+      "grad_norm": 0.435546875,
+      "learning_rate": 8.040670259917238e-05,
+      "loss": 3.0078,
+      "step": 17636
+    },
+    {
+      "epoch": 0.7426105263157895,
+      "grad_norm": 0.431640625,
+      "learning_rate": 8.038191162938896e-05,
+      "loss": 3.2411,
+      "step": 17637
+    },
+    {
+      "epoch": 0.7426526315789473,
+      "grad_norm": 0.443359375,
+      "learning_rate": 8.035712374984783e-05,
+      "loss": 3.0164,
+      "step": 17638
+    },
+    {
+      "epoch": 0.7426947368421053,
+      "grad_norm": 0.4140625,
+      "learning_rate": 8.033233896100056e-05,
+      "loss": 3.2539,
+      "step": 17639
+    },
+    {
+      "epoch": 0.7427368421052631,
+      "grad_norm": 0.4375,
+      "learning_rate": 8.030755726329874e-05,
+      "loss": 2.764,
+      "step": 17640
+    },
+    {
+      "epoch": 0.7427789473684211,
+      "grad_norm": 0.427734375,
+      "learning_rate": 8.028277865719378e-05,
+      "loss": 2.9396,
+      "step": 17641
+    },
+    {
+      "epoch": 0.7428210526315789,
+      "grad_norm": 0.419921875,
+      "learning_rate": 8.02580031431373e-05,
+      "loss": 2.8634,
+      "step": 17642
+    },
+    {
+      "epoch": 0.7428631578947369,
+      "grad_norm": 0.42578125,
+      "learning_rate": 8.023323072158042e-05,
+      "loss": 2.7683,
+      "step": 17643
+    },
+    {
+      "epoch": 0.7429052631578947,
+      "grad_norm": 0.43359375,
+      "learning_rate": 8.020846139297472e-05,
+      "loss": 3.4987,
+      "step": 17644
+    },
+    {
+      "epoch": 0.7429473684210526,
+      "grad_norm": 0.41796875,
+      "learning_rate": 8.018369515777133e-05,
+      "loss": 2.975,
+      "step": 17645
+    },
+    {
+      "epoch": 0.7429894736842105,
+      "grad_norm": 0.4140625,
+      "learning_rate": 8.015893201642149e-05,
+      "loss": 2.9077,
+      "step": 17646
+    },
+    {
+      "epoch": 0.7430315789473684,
+      "grad_norm": 0.46484375,
+      "learning_rate": 8.013417196937636e-05,
+      "loss": 3.2693,
+      "step": 17647
+    },
+    {
+      "epoch": 0.7430736842105263,
+      "grad_norm": 0.43359375,
+      "learning_rate": 8.010941501708693e-05,
+      "loss": 3.349,
+      "step": 17648
+    },
+    {
+      "epoch": 0.7431157894736842,
+      "grad_norm": 0.419921875,
+      "learning_rate": 8.008466116000454e-05,
+      "loss": 3.3587,
+      "step": 17649
+    },
+    {
+      "epoch": 0.7431578947368421,
+      "grad_norm": 0.419921875,
+      "learning_rate": 8.005991039857979e-05,
+      "loss": 3.1469,
+      "step": 17650
+    },
+    {
+      "epoch": 0.7432,
+      "grad_norm": 0.43359375,
+      "learning_rate": 8.003516273326392e-05,
+      "loss": 3.2724,
+      "step": 17651
+    },
+    {
+      "epoch": 0.7432421052631579,
+      "grad_norm": 0.4140625,
+      "learning_rate": 8.001041816450763e-05,
+      "loss": 2.8888,
+      "step": 17652
+    },
+    {
+      "epoch": 0.7432842105263158,
+      "grad_norm": 0.4296875,
+      "learning_rate": 7.998567669276183e-05,
+      "loss": 2.745,
+      "step": 17653
+    },
+    {
+      "epoch": 0.7433263157894737,
+      "grad_norm": 0.45703125,
+      "learning_rate": 7.996093831847723e-05,
+      "loss": 2.9117,
+      "step": 17654
+    },
+    {
+      "epoch": 0.7433684210526316,
+      "grad_norm": 0.421875,
+      "learning_rate": 7.993620304210447e-05,
+      "loss": 3.3102,
+      "step": 17655
+    },
+    {
+      "epoch": 0.7434105263157895,
+      "grad_norm": 0.451171875,
+      "learning_rate": 7.991147086409445e-05,
+      "loss": 3.4329,
+      "step": 17656
+    },
+    {
+      "epoch": 0.7434526315789474,
+      "grad_norm": 0.439453125,
+      "learning_rate": 7.988674178489741e-05,
+      "loss": 3.1856,
+      "step": 17657
+    },
+    {
+      "epoch": 0.7434947368421052,
+      "grad_norm": 0.458984375,
+      "learning_rate": 7.986201580496422e-05,
+      "loss": 3.6201,
+      "step": 17658
+    },
+    {
+      "epoch": 0.7435368421052632,
+      "grad_norm": 0.3984375,
+      "learning_rate": 7.983729292474504e-05,
+      "loss": 2.9935,
+      "step": 17659
+    },
+    {
+      "epoch": 0.743578947368421,
+      "grad_norm": 0.44921875,
+      "learning_rate": 7.981257314469051e-05,
+      "loss": 3.1357,
+      "step": 17660
+    },
+    {
+      "epoch": 0.743621052631579,
+      "grad_norm": 0.4375,
+      "learning_rate": 7.978785646525097e-05,
+      "loss": 3.3248,
+      "step": 17661
+    },
+    {
+      "epoch": 0.7436631578947368,
+      "grad_norm": 0.43359375,
+      "learning_rate": 7.976314288687658e-05,
+      "loss": 2.3929,
+      "step": 17662
+    },
+    {
+      "epoch": 0.7437052631578948,
+      "grad_norm": 0.4375,
+      "learning_rate": 7.973843241001788e-05,
+      "loss": 3.0709,
+      "step": 17663
+    },
+    {
+      "epoch": 0.7437473684210526,
+      "grad_norm": 0.408203125,
+      "learning_rate": 7.971372503512475e-05,
+      "loss": 2.5831,
+      "step": 17664
+    },
+    {
+      "epoch": 0.7437894736842106,
+      "grad_norm": 0.458984375,
+      "learning_rate": 7.968902076264761e-05,
+      "loss": 2.8337,
+      "step": 17665
+    },
+    {
+      "epoch": 0.7438315789473684,
+      "grad_norm": 0.431640625,
+      "learning_rate": 7.966431959303624e-05,
+      "loss": 3.4095,
+      "step": 17666
+    },
+    {
+      "epoch": 0.7438736842105264,
+      "grad_norm": 0.423828125,
+      "learning_rate": 7.9639621526741e-05,
+      "loss": 2.9965,
+      "step": 17667
+    },
+    {
+      "epoch": 0.7439157894736842,
+      "grad_norm": 0.4140625,
+      "learning_rate": 7.961492656421155e-05,
+      "loss": 3.3854,
+      "step": 17668
+    },
+    {
+      "epoch": 0.7439578947368422,
+      "grad_norm": 0.43359375,
+      "learning_rate": 7.959023470589799e-05,
+      "loss": 2.9396,
+      "step": 17669
+    },
+    {
+      "epoch": 0.744,
+      "grad_norm": 0.419921875,
+      "learning_rate": 7.956554595225015e-05,
+      "loss": 3.1431,
+      "step": 17670
+    },
+    {
+      "epoch": 0.744042105263158,
+      "grad_norm": 0.44140625,
+      "learning_rate": 7.954086030371774e-05,
+      "loss": 3.0147,
+      "step": 17671
+    },
+    {
+      "epoch": 0.7440842105263158,
+      "grad_norm": 0.427734375,
+      "learning_rate": 7.95161777607507e-05,
+      "loss": 2.9955,
+      "step": 17672
+    },
+    {
+      "epoch": 0.7441263157894736,
+      "grad_norm": 0.4296875,
+      "learning_rate": 7.949149832379846e-05,
+      "loss": 2.6265,
+      "step": 17673
+    },
+    {
+      "epoch": 0.7441684210526316,
+      "grad_norm": 0.431640625,
+      "learning_rate": 7.946682199331088e-05,
+      "loss": 3.0518,
+      "step": 17674
+    },
+    {
+      "epoch": 0.7442105263157894,
+      "grad_norm": 0.451171875,
+      "learning_rate": 7.944214876973738e-05,
+      "loss": 3.3039,
+      "step": 17675
+    },
+    {
+      "epoch": 0.7442526315789474,
+      "grad_norm": 0.4453125,
+      "learning_rate": 7.941747865352758e-05,
+      "loss": 2.5217,
+      "step": 17676
+    },
+    {
+      "epoch": 0.7442947368421052,
+      "grad_norm": 0.41015625,
+      "learning_rate": 7.939281164513088e-05,
+      "loss": 3.2892,
+      "step": 17677
+    },
+    {
+      "epoch": 0.7443368421052632,
+      "grad_norm": 0.416015625,
+      "learning_rate": 7.936814774499659e-05,
+      "loss": 2.758,
+      "step": 17678
+    },
+    {
+      "epoch": 0.744378947368421,
+      "grad_norm": 0.4296875,
+      "learning_rate": 7.934348695357435e-05,
+      "loss": 3.3586,
+      "step": 17679
+    },
+    {
+      "epoch": 0.744421052631579,
+      "grad_norm": 0.447265625,
+      "learning_rate": 7.931882927131309e-05,
+      "loss": 2.8103,
+      "step": 17680
+    },
+    {
+      "epoch": 0.7444631578947368,
+      "grad_norm": 0.455078125,
+      "learning_rate": 7.929417469866232e-05,
+      "loss": 3.2136,
+      "step": 17681
+    },
+    {
+      "epoch": 0.7445052631578948,
+      "grad_norm": 0.4375,
+      "learning_rate": 7.926952323607109e-05,
+      "loss": 2.8392,
+      "step": 17682
+    },
+    {
+      "epoch": 0.7445473684210526,
+      "grad_norm": 0.439453125,
+      "learning_rate": 7.924487488398857e-05,
+      "loss": 3.2488,
+      "step": 17683
+    },
+    {
+      "epoch": 0.7445894736842106,
+      "grad_norm": 0.412109375,
+      "learning_rate": 7.92202296428638e-05,
+      "loss": 2.9607,
+      "step": 17684
+    },
+    {
+      "epoch": 0.7446315789473684,
+      "grad_norm": 0.4296875,
+      "learning_rate": 7.919558751314576e-05,
+      "loss": 3.4319,
+      "step": 17685
+    },
+    {
+      "epoch": 0.7446736842105263,
+      "grad_norm": 0.41015625,
+      "learning_rate": 7.917094849528344e-05,
+      "loss": 3.2068,
+      "step": 17686
+    },
+    {
+      "epoch": 0.7447157894736842,
+      "grad_norm": 0.42578125,
+      "learning_rate": 7.914631258972566e-05,
+      "loss": 3.3719,
+      "step": 17687
+    },
+    {
+      "epoch": 0.7447578947368421,
+      "grad_norm": 0.443359375,
+      "learning_rate": 7.91216797969214e-05,
+      "loss": 3.1213,
+      "step": 17688
+    },
+    {
+      "epoch": 0.7448,
+      "grad_norm": 0.45703125,
+      "learning_rate": 7.909705011731936e-05,
+      "loss": 2.9213,
+      "step": 17689
+    },
+    {
+      "epoch": 0.7448421052631579,
+      "grad_norm": 0.416015625,
+      "learning_rate": 7.907242355136823e-05,
+      "loss": 2.9728,
+      "step": 17690
+    },
+    {
+      "epoch": 0.7448842105263158,
+      "grad_norm": 0.458984375,
+      "learning_rate": 7.904780009951673e-05,
+      "loss": 2.916,
+      "step": 17691
+    },
+    {
+      "epoch": 0.7449263157894737,
+      "grad_norm": 0.41015625,
+      "learning_rate": 7.902317976221346e-05,
+      "loss": 3.3855,
+      "step": 17692
+    },
+    {
+      "epoch": 0.7449684210526316,
+      "grad_norm": 0.4453125,
+      "learning_rate": 7.899856253990698e-05,
+      "loss": 3.0976,
+      "step": 17693
+    },
+    {
+      "epoch": 0.7450105263157895,
+      "grad_norm": 0.431640625,
+      "learning_rate": 7.897394843304567e-05,
+      "loss": 3.2046,
+      "step": 17694
+    },
+    {
+      "epoch": 0.7450526315789474,
+      "grad_norm": 0.42578125,
+      "learning_rate": 7.894933744207816e-05,
+      "loss": 3.3728,
+      "step": 17695
+    },
+    {
+      "epoch": 0.7450947368421053,
+      "grad_norm": 0.427734375,
+      "learning_rate": 7.892472956745272e-05,
+      "loss": 3.4017,
+      "step": 17696
+    },
+    {
+      "epoch": 0.7451368421052632,
+      "grad_norm": 0.421875,
+      "learning_rate": 7.890012480961773e-05,
+      "loss": 3.0002,
+      "step": 17697
+    },
+    {
+      "epoch": 0.745178947368421,
+      "grad_norm": 0.453125,
+      "learning_rate": 7.887552316902144e-05,
+      "loss": 3.3021,
+      "step": 17698
+    },
+    {
+      "epoch": 0.7452210526315789,
+      "grad_norm": 0.416015625,
+      "learning_rate": 7.885092464611204e-05,
+      "loss": 2.9502,
+      "step": 17699
+    },
+    {
+      "epoch": 0.7452631578947368,
+      "grad_norm": 0.431640625,
+      "learning_rate": 7.88263292413377e-05,
+      "loss": 3.3809,
+      "step": 17700
+    },
+    {
+      "epoch": 0.7453052631578947,
+      "grad_norm": 0.443359375,
+      "learning_rate": 7.880173695514653e-05,
+      "loss": 3.1285,
+      "step": 17701
+    },
+    {
+      "epoch": 0.7453473684210526,
+      "grad_norm": 0.443359375,
+      "learning_rate": 7.877714778798645e-05,
+      "loss": 2.9476,
+      "step": 17702
+    },
+    {
+      "epoch": 0.7453894736842105,
+      "grad_norm": 0.412109375,
+      "learning_rate": 7.875256174030567e-05,
+      "loss": 3.11,
+      "step": 17703
+    },
+    {
+      "epoch": 0.7454315789473684,
+      "grad_norm": 0.41796875,
+      "learning_rate": 7.8727978812552e-05,
+      "loss": 3.4689,
+      "step": 17704
+    },
+    {
+      "epoch": 0.7454736842105263,
+      "grad_norm": 0.43359375,
+      "learning_rate": 7.870339900517331e-05,
+      "loss": 3.4221,
+      "step": 17705
+    },
+    {
+      "epoch": 0.7455157894736842,
+      "grad_norm": 0.404296875,
+      "learning_rate": 7.867882231861745e-05,
+      "loss": 3.1046,
+      "step": 17706
+    },
+    {
+      "epoch": 0.7455578947368421,
+      "grad_norm": 0.44140625,
+      "learning_rate": 7.865424875333216e-05,
+      "loss": 3.3729,
+      "step": 17707
+    },
+    {
+      "epoch": 0.7456,
+      "grad_norm": 0.419921875,
+      "learning_rate": 7.862967830976511e-05,
+      "loss": 2.9193,
+      "step": 17708
+    },
+    {
+      "epoch": 0.7456421052631579,
+      "grad_norm": 0.44921875,
+      "learning_rate": 7.860511098836387e-05,
+      "loss": 3.326,
+      "step": 17709
+    },
+    {
+      "epoch": 0.7456842105263158,
+      "grad_norm": 0.443359375,
+      "learning_rate": 7.858054678957633e-05,
+      "loss": 3.0627,
+      "step": 17710
+    },
+    {
+      "epoch": 0.7457263157894737,
+      "grad_norm": 0.4140625,
+      "learning_rate": 7.855598571384964e-05,
+      "loss": 2.9963,
+      "step": 17711
+    },
+    {
+      "epoch": 0.7457684210526315,
+      "grad_norm": 0.44140625,
+      "learning_rate": 7.853142776163152e-05,
+      "loss": 3.0309,
+      "step": 17712
+    },
+    {
+      "epoch": 0.7458105263157895,
+      "grad_norm": 0.427734375,
+      "learning_rate": 7.850687293336933e-05,
+      "loss": 3.2342,
+      "step": 17713
+    },
+    {
+      "epoch": 0.7458526315789473,
+      "grad_norm": 0.4296875,
+      "learning_rate": 7.848232122951041e-05,
+      "loss": 3.0369,
+      "step": 17714
+    },
+    {
+      "epoch": 0.7458947368421053,
+      "grad_norm": 0.419921875,
+      "learning_rate": 7.845777265050208e-05,
+      "loss": 3.0712,
+      "step": 17715
+    },
+    {
+      "epoch": 0.7459368421052631,
+      "grad_norm": 0.42578125,
+      "learning_rate": 7.843322719679149e-05,
+      "loss": 3.3188,
+      "step": 17716
+    },
+    {
+      "epoch": 0.7459789473684211,
+      "grad_norm": 0.421875,
+      "learning_rate": 7.84086848688261e-05,
+      "loss": 2.8848,
+      "step": 17717
+    },
+    {
+      "epoch": 0.7460210526315789,
+      "grad_norm": 0.412109375,
+      "learning_rate": 7.838414566705268e-05,
+      "loss": 2.8543,
+      "step": 17718
+    },
+    {
+      "epoch": 0.7460631578947369,
+      "grad_norm": 0.625,
+      "learning_rate": 7.83596095919186e-05,
+      "loss": 3.1319,
+      "step": 17719
+    },
+    {
+      "epoch": 0.7461052631578947,
+      "grad_norm": 0.423828125,
+      "learning_rate": 7.833507664387071e-05,
+      "loss": 3.0337,
+      "step": 17720
+    },
+    {
+      "epoch": 0.7461473684210527,
+      "grad_norm": 0.41796875,
+      "learning_rate": 7.831054682335598e-05,
+      "loss": 2.845,
+      "step": 17721
+    },
+    {
+      "epoch": 0.7461894736842105,
+      "grad_norm": 0.396484375,
+      "learning_rate": 7.82860201308215e-05,
+      "loss": 2.5879,
+      "step": 17722
+    },
+    {
+      "epoch": 0.7462315789473685,
+      "grad_norm": 0.431640625,
+      "learning_rate": 7.826149656671386e-05,
+      "loss": 2.9478,
+      "step": 17723
+    },
+    {
+      "epoch": 0.7462736842105263,
+      "grad_norm": 0.431640625,
+      "learning_rate": 7.823697613148009e-05,
+      "loss": 3.3218,
+      "step": 17724
+    },
+    {
+      "epoch": 0.7463157894736843,
+      "grad_norm": 0.42578125,
+      "learning_rate": 7.821245882556665e-05,
+      "loss": 3.0221,
+      "step": 17725
+    },
+    {
+      "epoch": 0.7463578947368421,
+      "grad_norm": 0.412109375,
+      "learning_rate": 7.818794464942052e-05,
+      "loss": 3.1068,
+      "step": 17726
+    },
+    {
+      "epoch": 0.7464,
+      "grad_norm": 0.43359375,
+      "learning_rate": 7.816343360348801e-05,
+      "loss": 2.767,
+      "step": 17727
+    },
+    {
+      "epoch": 0.7464421052631579,
+      "grad_norm": 0.451171875,
+      "learning_rate": 7.813892568821593e-05,
+      "loss": 3.6161,
+      "step": 17728
+    },
+    {
+      "epoch": 0.7464842105263158,
+      "grad_norm": 0.427734375,
+      "learning_rate": 7.811442090405069e-05,
+      "loss": 3.2526,
+      "step": 17729
+    },
+    {
+      "epoch": 0.7465263157894737,
+      "grad_norm": 0.4296875,
+      "learning_rate": 7.808991925143869e-05,
+      "loss": 3.4836,
+      "step": 17730
+    },
+    {
+      "epoch": 0.7465684210526315,
+      "grad_norm": 0.419921875,
+      "learning_rate": 7.806542073082649e-05,
+      "loss": 3.4092,
+      "step": 17731
+    },
+    {
+      "epoch": 0.7466105263157895,
+      "grad_norm": 0.421875,
+      "learning_rate": 7.804092534266017e-05,
+      "loss": 2.8611,
+      "step": 17732
+    },
+    {
+      "epoch": 0.7466526315789473,
+      "grad_norm": 0.43359375,
+      "learning_rate": 7.801643308738631e-05,
+      "loss": 3.3969,
+      "step": 17733
+    },
+    {
+      "epoch": 0.7466947368421053,
+      "grad_norm": 0.400390625,
+      "learning_rate": 7.799194396545082e-05,
+      "loss": 2.8292,
+      "step": 17734
+    },
+    {
+      "epoch": 0.7467368421052631,
+      "grad_norm": 0.455078125,
+      "learning_rate": 7.796745797730007e-05,
+      "loss": 3.3956,
+      "step": 17735
+    },
+    {
+      "epoch": 0.7467789473684211,
+      "grad_norm": 0.416015625,
+      "learning_rate": 7.794297512338008e-05,
+      "loss": 3.2819,
+      "step": 17736
+    },
+    {
+      "epoch": 0.7468210526315789,
+      "grad_norm": 0.44140625,
+      "learning_rate": 7.791849540413698e-05,
+      "loss": 3.1499,
+      "step": 17737
+    },
+    {
+      "epoch": 0.7468631578947369,
+      "grad_norm": 0.4375,
+      "learning_rate": 7.789401882001668e-05,
+      "loss": 3.3349,
+      "step": 17738
+    },
+    {
+      "epoch": 0.7469052631578947,
+      "grad_norm": 0.41796875,
+      "learning_rate": 7.786954537146504e-05,
+      "loss": 3.4044,
+      "step": 17739
+    },
+    {
+      "epoch": 0.7469473684210526,
+      "grad_norm": 0.4296875,
+      "learning_rate": 7.784507505892824e-05,
+      "loss": 3.0916,
+      "step": 17740
+    },
+    {
+      "epoch": 0.7469894736842105,
+      "grad_norm": 0.451171875,
+      "learning_rate": 7.782060788285169e-05,
+      "loss": 2.8171,
+      "step": 17741
+    },
+    {
+      "epoch": 0.7470315789473684,
+      "grad_norm": 0.421875,
+      "learning_rate": 7.779614384368147e-05,
+      "loss": 2.8135,
+      "step": 17742
+    },
+    {
+      "epoch": 0.7470736842105263,
+      "grad_norm": 0.43359375,
+      "learning_rate": 7.777168294186317e-05,
+      "loss": 3.1717,
+      "step": 17743
+    },
+    {
+      "epoch": 0.7471157894736842,
+      "grad_norm": 0.431640625,
+      "learning_rate": 7.774722517784246e-05,
+      "loss": 3.1439,
+      "step": 17744
+    },
+    {
+      "epoch": 0.7471578947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 7.772277055206492e-05,
+      "loss": 3.2358,
+      "step": 17745
+    },
+    {
+      "epoch": 0.7472,
+      "grad_norm": 0.431640625,
+      "learning_rate": 7.769831906497602e-05,
+      "loss": 3.0483,
+      "step": 17746
+    },
+    {
+      "epoch": 0.7472421052631579,
+      "grad_norm": 0.41796875,
+      "learning_rate": 7.767387071702142e-05,
+      "loss": 2.9274,
+      "step": 17747
+    },
+    {
+      "epoch": 0.7472842105263158,
+      "grad_norm": 0.439453125,
+      "learning_rate": 7.764942550864629e-05,
+      "loss": 3.3444,
+      "step": 17748
+    },
+    {
+      "epoch": 0.7473263157894737,
+      "grad_norm": 0.466796875,
+      "learning_rate": 7.762498344029617e-05,
+      "loss": 2.9817,
+      "step": 17749
+    },
+    {
+      "epoch": 0.7473684210526316,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.760054451241636e-05,
+      "loss": 2.7918,
+      "step": 17750
+    },
+    {
+      "epoch": 0.7474105263157895,
+      "grad_norm": 0.419921875,
+      "learning_rate": 7.757610872545204e-05,
+      "loss": 3.4325,
+      "step": 17751
+    },
+    {
+      "epoch": 0.7474526315789474,
+      "grad_norm": 0.42578125,
+      "learning_rate": 7.755167607984845e-05,
+      "loss": 3.4448,
+      "step": 17752
+    },
+    {
+      "epoch": 0.7474947368421052,
+      "grad_norm": 0.408203125,
+      "learning_rate": 7.752724657605068e-05,
+      "loss": 3.2694,
+      "step": 17753
+    },
+    {
+      "epoch": 0.7475368421052632,
+      "grad_norm": 0.419921875,
+      "learning_rate": 7.750282021450384e-05,
+      "loss": 3.0342,
+      "step": 17754
+    },
+    {
+      "epoch": 0.747578947368421,
+      "grad_norm": 0.44140625,
+      "learning_rate": 7.747839699565287e-05,
+      "loss": 3.2658,
+      "step": 17755
+    },
+    {
+      "epoch": 0.747621052631579,
+      "grad_norm": 0.470703125,
+      "learning_rate": 7.745397691994285e-05,
+      "loss": 3.0521,
+      "step": 17756
+    },
+    {
+      "epoch": 0.7476631578947368,
+      "grad_norm": 0.451171875,
+      "learning_rate": 7.742955998781867e-05,
+      "loss": 3.279,
+      "step": 17757
+    },
+    {
+      "epoch": 0.7477052631578948,
+      "grad_norm": 0.421875,
+      "learning_rate": 7.740514619972513e-05,
+      "loss": 3.0341,
+      "step": 17758
+    },
+    {
+      "epoch": 0.7477473684210526,
+      "grad_norm": 0.427734375,
+      "learning_rate": 7.738073555610705e-05,
+      "loss": 3.1511,
+      "step": 17759
+    },
+    {
+      "epoch": 0.7477894736842106,
+      "grad_norm": 0.4296875,
+      "learning_rate": 7.735632805740914e-05,
+      "loss": 3.4554,
+      "step": 17760
+    },
+    {
+      "epoch": 0.7478315789473684,
+      "grad_norm": 0.412109375,
+      "learning_rate": 7.733192370407607e-05,
+      "loss": 3.0547,
+      "step": 17761
+    },
+    {
+      "epoch": 0.7478736842105264,
+      "grad_norm": 0.462890625,
+      "learning_rate": 7.730752249655242e-05,
+      "loss": 3.2991,
+      "step": 17762
+    },
+    {
+      "epoch": 0.7479157894736842,
+      "grad_norm": 0.408203125,
+      "learning_rate": 7.728312443528288e-05,
+      "loss": 3.0757,
+      "step": 17763
+    },
+    {
+      "epoch": 0.7479578947368422,
+      "grad_norm": 0.42578125,
+      "learning_rate": 7.725872952071186e-05,
+      "loss": 2.4171,
+      "step": 17764
+    },
+    {
+      "epoch": 0.748,
+      "grad_norm": 0.42578125,
+      "learning_rate": 7.723433775328384e-05,
+      "loss": 3.2037,
+      "step": 17765
+    },
+    {
+      "epoch": 0.7480421052631578,
+      "grad_norm": 0.44140625,
+      "learning_rate": 7.72099491334432e-05,
+      "loss": 3.2609,
+      "step": 17766
+    },
+    {
+      "epoch": 0.7480842105263158,
+      "grad_norm": 0.44921875,
+      "learning_rate": 7.718556366163427e-05,
+      "loss": 3.1533,
+      "step": 17767
+    },
+    {
+      "epoch": 0.7481263157894736,
+      "grad_norm": 0.44921875,
+      "learning_rate": 7.71611813383013e-05,
+      "loss": 3.0509,
+      "step": 17768
+    },
+    {
+      "epoch": 0.7481684210526316,
+      "grad_norm": 0.435546875,
+      "learning_rate": 7.713680216388855e-05,
+      "loss": 3.2013,
+      "step": 17769
+    },
+    {
+      "epoch": 0.7482105263157894,
+      "grad_norm": 0.4296875,
+      "learning_rate": 7.711242613884007e-05,
+      "loss": 3.1478,
+      "step": 17770
+    },
+    {
+      "epoch": 0.7482526315789474,
+      "grad_norm": 0.427734375,
+      "learning_rate": 7.708805326360013e-05,
+      "loss": 3.1851,
+      "step": 17771
+    },
+    {
+      "epoch": 0.7482947368421052,
+      "grad_norm": 0.412109375,
+      "learning_rate": 7.706368353861268e-05,
+      "loss": 2.7668,
+      "step": 17772
+    },
+    {
+      "epoch": 0.7483368421052632,
+      "grad_norm": 0.490234375,
+      "learning_rate": 7.703931696432174e-05,
+      "loss": 3.357,
+      "step": 17773
+    },
+    {
+      "epoch": 0.748378947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 7.701495354117121e-05,
+      "loss": 3.493,
+      "step": 17774
+    },
+    {
+      "epoch": 0.748421052631579,
+      "grad_norm": 0.41015625,
+      "learning_rate": 7.699059326960492e-05,
+      "loss": 2.7973,
+      "step": 17775
+    },
+    {
+      "epoch": 0.7484631578947368,
+      "grad_norm": 0.435546875,
+      "learning_rate": 7.69662361500669e-05,
+      "loss": 2.5367,
+      "step": 17776
+    },
+    {
+      "epoch": 0.7485052631578948,
+      "grad_norm": 0.431640625,
+      "learning_rate": 7.694188218300061e-05,
+      "loss": 3.1258,
+      "step": 17777
+    },
+    {
+      "epoch": 0.7485473684210526,
+      "grad_norm": 0.439453125,
+      "learning_rate": 7.691753136884993e-05,
+      "loss": 3.0488,
+      "step": 17778
+    },
+    {
+      "epoch": 0.7485894736842105,
+      "grad_norm": 0.453125,
+      "learning_rate": 7.689318370805847e-05,
+      "loss": 2.9457,
+      "step": 17779
+    },
+    {
+      "epoch": 0.7486315789473684,
+      "grad_norm": 0.435546875,
+      "learning_rate": 7.686883920106982e-05,
+      "loss": 2.7233,
+      "step": 17780
+    },
+    {
+      "epoch": 0.7486736842105263,
+      "grad_norm": 0.431640625,
+      "learning_rate": 7.68444978483275e-05,
+      "loss": 3.1506,
+      "step": 17781
+    },
+    {
+      "epoch": 0.7487157894736842,
+      "grad_norm": 0.40625,
+      "learning_rate": 7.682015965027492e-05,
+      "loss": 3.1776,
+      "step": 17782
+    },
+    {
+      "epoch": 0.7487578947368421,
+      "grad_norm": 0.408203125,
+      "learning_rate": 7.67958246073557e-05,
+      "loss": 3.2016,
+      "step": 17783
+    },
+    {
+      "epoch": 0.7488,
+      "grad_norm": 0.455078125,
+      "learning_rate": 7.677149272001288e-05,
+      "loss": 3.0499,
+      "step": 17784
+    },
+    {
+      "epoch": 0.7488421052631579,
+      "grad_norm": 0.400390625,
+      "learning_rate": 7.674716398869009e-05,
+      "loss": 3.0218,
+      "step": 17785
+    },
+    {
+      "epoch": 0.7488842105263158,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.672283841383027e-05,
+      "loss": 3.1776,
+      "step": 17786
+    },
+    {
+      "epoch": 0.7489263157894737,
+      "grad_norm": 0.3984375,
+      "learning_rate": 7.66985159958768e-05,
+      "loss": 2.9598,
+      "step": 17787
+    },
+    {
+      "epoch": 0.7489684210526316,
+      "grad_norm": 0.51171875,
+      "learning_rate": 7.667419673527273e-05,
+      "loss": 3.0317,
+      "step": 17788
+    },
+    {
+      "epoch": 0.7490105263157895,
+      "grad_norm": 0.42578125,
+      "learning_rate": 7.66498806324611e-05,
+      "loss": 2.7939,
+      "step": 17789
+    },
+    {
+      "epoch": 0.7490526315789474,
+      "grad_norm": 0.431640625,
+      "learning_rate": 7.662556768788512e-05,
+      "loss": 3.1224,
+      "step": 17790
+    },
+    {
+      "epoch": 0.7490947368421053,
+      "grad_norm": 0.4375,
+      "learning_rate": 7.660125790198741e-05,
+      "loss": 3.3367,
+      "step": 17791
+    },
+    {
+      "epoch": 0.7491368421052632,
+      "grad_norm": 0.439453125,
+      "learning_rate": 7.657695127521122e-05,
+      "loss": 2.8424,
+      "step": 17792
+    },
+    {
+      "epoch": 0.749178947368421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 7.655264780799907e-05,
+      "loss": 3.4951,
+      "step": 17793
+    },
+    {
+      "epoch": 0.7492210526315789,
+      "grad_norm": 0.462890625,
+      "learning_rate": 7.652834750079394e-05,
+      "loss": 3.3853,
+      "step": 17794
+    },
+    {
+      "epoch": 0.7492631578947369,
+      "grad_norm": 0.419921875,
+      "learning_rate": 7.650405035403848e-05,
+      "loss": 3.1308,
+      "step": 17795
+    },
+    {
+      "epoch": 0.7493052631578947,
+      "grad_norm": 0.41796875,
+      "learning_rate": 7.647975636817542e-05,
+      "loss": 3.2264,
+      "step": 17796
+    },
+    {
+      "epoch": 0.7493473684210527,
+      "grad_norm": 0.44921875,
+      "learning_rate": 7.645546554364729e-05,
+      "loss": 3.1742,
+      "step": 17797
+    },
+    {
+      "epoch": 0.7493894736842105,
+      "grad_norm": 0.423828125,
+      "learning_rate": 7.643117788089659e-05,
+      "loss": 3.0633,
+      "step": 17798
+    },
+    {
+      "epoch": 0.7494315789473684,
+      "grad_norm": 0.439453125,
+      "learning_rate": 7.640689338036608e-05,
+      "loss": 3.2137,
+      "step": 17799
+    },
+    {
+      "epoch": 0.7494736842105263,
+      "grad_norm": 0.447265625,
+      "learning_rate": 7.638261204249783e-05,
+      "loss": 2.8879,
+      "step": 17800
+    },
+    {
+      "epoch": 0.7495157894736842,
+      "grad_norm": 0.412109375,
+      "learning_rate": 7.635833386773456e-05,
+      "loss": 3.1504,
+      "step": 17801
+    },
+    {
+      "epoch": 0.7495578947368421,
+      "grad_norm": 0.44921875,
+      "learning_rate": 7.633405885651826e-05,
+      "loss": 2.754,
+      "step": 17802
+    },
+    {
+      "epoch": 0.7496,
+      "grad_norm": 0.46875,
+      "learning_rate": 7.630978700929145e-05,
+      "loss": 3.0729,
+      "step": 17803
+    },
+    {
+      "epoch": 0.7496421052631579,
+      "grad_norm": 0.4296875,
+      "learning_rate": 7.628551832649625e-05,
+      "loss": 3.2097,
+      "step": 17804
+    },
+    {
+      "epoch": 0.7496842105263158,
+      "grad_norm": 0.451171875,
+      "learning_rate": 7.626125280857471e-05,
+      "loss": 3.1063,
+      "step": 17805
+    },
+    {
+      "epoch": 0.7497263157894737,
+      "grad_norm": 0.419921875,
+      "learning_rate": 7.623699045596916e-05,
+      "loss": 3.2728,
+      "step": 17806
+    },
+    {
+      "epoch": 0.7497684210526315,
+      "grad_norm": 0.443359375,
+      "learning_rate": 7.621273126912134e-05,
+      "loss": 2.7823,
+      "step": 17807
+    },
+    {
+      "epoch": 0.7498105263157895,
+      "grad_norm": 0.451171875,
+      "learning_rate": 7.618847524847349e-05,
+      "loss": 3.0244,
+      "step": 17808
+    },
+    {
+      "epoch": 0.7498526315789473,
+      "grad_norm": 0.466796875,
+      "learning_rate": 7.616422239446727e-05,
+      "loss": 2.8898,
+      "step": 17809
+    },
+    {
+      "epoch": 0.7498947368421053,
+      "grad_norm": 0.451171875,
+      "learning_rate": 7.613997270754474e-05,
+      "loss": 2.9765,
+      "step": 17810
+    },
+    {
+      "epoch": 0.7499368421052631,
+      "grad_norm": 0.44140625,
+      "learning_rate": 7.611572618814763e-05,
+      "loss": 3.5159,
+      "step": 17811
+    },
+    {
+      "epoch": 0.7499789473684211,
+      "grad_norm": 0.41796875,
+      "learning_rate": 7.609148283671771e-05,
+      "loss": 2.993,
+      "step": 17812
+    },
+    {
+      "epoch": 0.7500210526315789,
+      "grad_norm": 0.421875,
+      "learning_rate": 7.60672426536966e-05,
+      "loss": 3.0904,
+      "step": 17813
+    },
+    {
+      "epoch": 0.7500631578947369,
+      "grad_norm": 0.490234375,
+      "learning_rate": 7.60430056395259e-05,
+      "loss": 3.0397,
+      "step": 17814
+    },
+    {
+      "epoch": 0.7501052631578947,
+      "grad_norm": 0.412109375,
+      "learning_rate": 7.601877179464741e-05,
+      "loss": 3.0567,
+      "step": 17815
+    },
+    {
+      "epoch": 0.7501473684210527,
+      "grad_norm": 0.412109375,
+      "learning_rate": 7.599454111950232e-05,
+      "loss": 3.0329,
+      "step": 17816
+    },
+    {
+      "epoch": 0.7501894736842105,
+      "grad_norm": 0.42578125,
+      "learning_rate": 7.597031361453233e-05,
+      "loss": 3.4886,
+      "step": 17817
+    },
+    {
+      "epoch": 0.7502315789473685,
+      "grad_norm": 0.4296875,
+      "learning_rate": 7.594608928017874e-05,
+      "loss": 3.253,
+      "step": 17818
+    },
+    {
+      "epoch": 0.7502736842105263,
+      "grad_norm": 0.4296875,
+      "learning_rate": 7.59218681168829e-05,
+      "loss": 3.0556,
+      "step": 17819
+    },
+    {
+      "epoch": 0.7503157894736842,
+      "grad_norm": 0.44140625,
+      "learning_rate": 7.589765012508607e-05,
+      "loss": 3.2216,
+      "step": 17820
+    },
+    {
+      "epoch": 0.7503578947368421,
+      "grad_norm": 0.42578125,
+      "learning_rate": 7.587343530522944e-05,
+      "loss": 3.2909,
+      "step": 17821
+    },
+    {
+      "epoch": 0.7504,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.584922365775429e-05,
+      "loss": 3.2148,
+      "step": 17822
+    },
+    {
+      "epoch": 0.7504421052631579,
+      "grad_norm": 0.4453125,
+      "learning_rate": 7.582501518310167e-05,
+      "loss": 2.8767,
+      "step": 17823
+    },
+    {
+      "epoch": 0.7504842105263158,
+      "grad_norm": 0.45703125,
+      "learning_rate": 7.580080988171265e-05,
+      "loss": 3.209,
+      "step": 17824
+    },
+    {
+      "epoch": 0.7505263157894737,
+      "grad_norm": 0.431640625,
+      "learning_rate": 7.577660775402818e-05,
+      "loss": 3.1463,
+      "step": 17825
+    },
+    {
+      "epoch": 0.7505684210526316,
+      "grad_norm": 0.416015625,
+      "learning_rate": 7.57524088004892e-05,
+      "loss": 2.9476,
+      "step": 17826
+    },
+    {
+      "epoch": 0.7506105263157895,
+      "grad_norm": 0.458984375,
+      "learning_rate": 7.572821302153663e-05,
+      "loss": 3.1166,
+      "step": 17827
+    },
+    {
+      "epoch": 0.7506526315789473,
+      "grad_norm": 0.431640625,
+      "learning_rate": 7.570402041761121e-05,
+      "loss": 3.5198,
+      "step": 17828
+    },
+    {
+      "epoch": 0.7506947368421053,
+      "grad_norm": 0.447265625,
+      "learning_rate": 7.567983098915371e-05,
+      "loss": 3.1533,
+      "step": 17829
+    },
+    {
+      "epoch": 0.7507368421052631,
+      "grad_norm": 0.42578125,
+      "learning_rate": 7.565564473660494e-05,
+      "loss": 3.3258,
+      "step": 17830
+    },
+    {
+      "epoch": 0.7507789473684211,
+      "grad_norm": 0.42578125,
+      "learning_rate": 7.563146166040547e-05,
+      "loss": 3.378,
+      "step": 17831
+    },
+    {
+      "epoch": 0.750821052631579,
+      "grad_norm": 0.44921875,
+      "learning_rate": 7.56072817609959e-05,
+      "loss": 2.9641,
+      "step": 17832
+    },
+    {
+      "epoch": 0.7508631578947368,
+      "grad_norm": 0.41015625,
+      "learning_rate": 7.558310503881674e-05,
+      "loss": 2.9044,
+      "step": 17833
+    },
+    {
+      "epoch": 0.7509052631578947,
+      "grad_norm": 0.4140625,
+      "learning_rate": 7.555893149430851e-05,
+      "loss": 2.7828,
+      "step": 17834
+    },
+    {
+      "epoch": 0.7509473684210526,
+      "grad_norm": 0.41796875,
+      "learning_rate": 7.553476112791155e-05,
+      "loss": 2.9097,
+      "step": 17835
+    },
+    {
+      "epoch": 0.7509894736842105,
+      "grad_norm": 0.453125,
+      "learning_rate": 7.551059394006621e-05,
+      "loss": 2.7156,
+      "step": 17836
+    },
+    {
+      "epoch": 0.7510315789473684,
+      "grad_norm": 0.490234375,
+      "learning_rate": 7.548642993121288e-05,
+      "loss": 3.0724,
+      "step": 17837
+    },
+    {
+      "epoch": 0.7510736842105263,
+      "grad_norm": 0.453125,
+      "learning_rate": 7.546226910179177e-05,
+      "loss": 2.9513,
+      "step": 17838
+    },
+    {
+      "epoch": 0.7511157894736842,
+      "grad_norm": 0.43359375,
+      "learning_rate": 7.543811145224303e-05,
+      "loss": 3.4351,
+      "step": 17839
+    },
+    {
+      "epoch": 0.7511578947368421,
+      "grad_norm": 0.439453125,
+      "learning_rate": 7.541395698300679e-05,
+      "loss": 2.6673,
+      "step": 17840
+    },
+    {
+      "epoch": 0.7512,
+      "grad_norm": 0.400390625,
+      "learning_rate": 7.538980569452314e-05,
+      "loss": 3.2137,
+      "step": 17841
+    },
+    {
+      "epoch": 0.7512421052631579,
+      "grad_norm": 0.396484375,
+      "learning_rate": 7.536565758723207e-05,
+      "loss": 2.8238,
+      "step": 17842
+    },
+    {
+      "epoch": 0.7512842105263158,
+      "grad_norm": 0.419921875,
+      "learning_rate": 7.534151266157347e-05,
+      "loss": 3.0127,
+      "step": 17843
+    },
+    {
+      "epoch": 0.7513263157894737,
+      "grad_norm": 0.421875,
+      "learning_rate": 7.531737091798743e-05,
+      "loss": 3.286,
+      "step": 17844
+    },
+    {
+      "epoch": 0.7513684210526316,
+      "grad_norm": 0.40234375,
+      "learning_rate": 7.52932323569135e-05,
+      "loss": 2.7039,
+      "step": 17845
+    },
+    {
+      "epoch": 0.7514105263157895,
+      "grad_norm": 0.451171875,
+      "learning_rate": 7.526909697879167e-05,
+      "loss": 3.3882,
+      "step": 17846
+    },
+    {
+      "epoch": 0.7514526315789474,
+      "grad_norm": 0.462890625,
+      "learning_rate": 7.524496478406159e-05,
+      "loss": 2.9403,
+      "step": 17847
+    },
+    {
+      "epoch": 0.7514947368421052,
+      "grad_norm": 0.421875,
+      "learning_rate": 7.522083577316293e-05,
+      "loss": 3.196,
+      "step": 17848
+    },
+    {
+      "epoch": 0.7515368421052632,
+      "grad_norm": 0.4375,
+      "learning_rate": 7.519670994653529e-05,
+      "loss": 2.6317,
+      "step": 17849
+    },
+    {
+      "epoch": 0.751578947368421,
+      "grad_norm": 0.419921875,
+      "learning_rate": 7.517258730461815e-05,
+      "loss": 2.562,
+      "step": 17850
+    },
+    {
+      "epoch": 0.751621052631579,
+      "grad_norm": 0.423828125,
+      "learning_rate": 7.514846784785118e-05,
+      "loss": 3.1515,
+      "step": 17851
+    },
+    {
+      "epoch": 0.7516631578947368,
+      "grad_norm": 0.404296875,
+      "learning_rate": 7.512435157667354e-05,
+      "loss": 3.0127,
+      "step": 17852
+    },
+    {
+      "epoch": 0.7517052631578948,
+      "grad_norm": 0.42578125,
+      "learning_rate": 7.510023849152484e-05,
+      "loss": 3.2391,
+      "step": 17853
+    },
+    {
+      "epoch": 0.7517473684210526,
+      "grad_norm": 0.44140625,
+      "learning_rate": 7.507612859284427e-05,
+      "loss": 3.0148,
+      "step": 17854
+    },
+    {
+      "epoch": 0.7517894736842106,
+      "grad_norm": 0.3984375,
+      "learning_rate": 7.50520218810711e-05,
+      "loss": 3.3489,
+      "step": 17855
+    },
+    {
+      "epoch": 0.7518315789473684,
+      "grad_norm": 0.4296875,
+      "learning_rate": 7.502791835664458e-05,
+      "loss": 3.0919,
+      "step": 17856
+    },
+    {
+      "epoch": 0.7518736842105264,
+      "grad_norm": 0.447265625,
+      "learning_rate": 7.500381802000369e-05,
+      "loss": 3.1156,
+      "step": 17857
+    },
+    {
+      "epoch": 0.7519157894736842,
+      "grad_norm": 0.419921875,
+      "learning_rate": 7.497972087158778e-05,
+      "loss": 3.1994,
+      "step": 17858
+    },
+    {
+      "epoch": 0.7519578947368422,
+      "grad_norm": 0.40234375,
+      "learning_rate": 7.495562691183558e-05,
+      "loss": 2.8095,
+      "step": 17859
+    },
+    {
+      "epoch": 0.752,
+      "grad_norm": 0.431640625,
+      "learning_rate": 7.493153614118633e-05,
+      "loss": 3.2132,
+      "step": 17860
+    },
+    {
+      "epoch": 0.7520421052631578,
+      "grad_norm": 0.462890625,
+      "learning_rate": 7.490744856007864e-05,
+      "loss": 2.8838,
+      "step": 17861
+    },
+    {
+      "epoch": 0.7520842105263158,
+      "grad_norm": 0.4296875,
+      "learning_rate": 7.488336416895158e-05,
+      "loss": 2.9682,
+      "step": 17862
+    },
+    {
+      "epoch": 0.7521263157894736,
+      "grad_norm": 0.443359375,
+      "learning_rate": 7.48592829682439e-05,
+      "loss": 3.5944,
+      "step": 17863
+    },
+    {
+      "epoch": 0.7521684210526316,
+      "grad_norm": 0.39453125,
+      "learning_rate": 7.48352049583942e-05,
+      "loss": 2.9618,
+      "step": 17864
+    },
+    {
+      "epoch": 0.7522105263157894,
+      "grad_norm": 0.423828125,
+      "learning_rate": 7.481113013984142e-05,
+      "loss": 3.3649,
+      "step": 17865
+    },
+    {
+      "epoch": 0.7522526315789474,
+      "grad_norm": 0.490234375,
+      "learning_rate": 7.478705851302384e-05,
+      "loss": 2.653,
+      "step": 17866
+    },
+    {
+      "epoch": 0.7522947368421052,
+      "grad_norm": 0.431640625,
+      "learning_rate": 7.476299007838036e-05,
+      "loss": 3.1838,
+      "step": 17867
+    },
+    {
+      "epoch": 0.7523368421052632,
+      "grad_norm": 0.41015625,
+      "learning_rate": 7.473892483634915e-05,
+      "loss": 2.7595,
+      "step": 17868
+    },
+    {
+      "epoch": 0.752378947368421,
+      "grad_norm": 0.419921875,
+      "learning_rate": 7.471486278736888e-05,
+      "loss": 2.4145,
+      "step": 17869
+    },
+    {
+      "epoch": 0.752421052631579,
+      "grad_norm": 0.443359375,
+      "learning_rate": 7.469080393187786e-05,
+      "loss": 3.1352,
+      "step": 17870
+    },
+    {
+      "epoch": 0.7524631578947368,
+      "grad_norm": 0.46484375,
+      "learning_rate": 7.466674827031441e-05,
+      "loss": 2.647,
+      "step": 17871
+    },
+    {
+      "epoch": 0.7525052631578948,
+      "grad_norm": 0.42578125,
+      "learning_rate": 7.46426958031168e-05,
+      "loss": 3.1138,
+      "step": 17872
+    },
+    {
+      "epoch": 0.7525473684210526,
+      "grad_norm": 0.40234375,
+      "learning_rate": 7.461864653072317e-05,
+      "loss": 2.7776,
+      "step": 17873
+    },
+    {
+      "epoch": 0.7525894736842105,
+      "grad_norm": 0.431640625,
+      "learning_rate": 7.459460045357186e-05,
+      "loss": 3.0042,
+      "step": 17874
+    },
+    {
+      "epoch": 0.7526315789473684,
+      "grad_norm": 0.44140625,
+      "learning_rate": 7.457055757210071e-05,
+      "loss": 2.9169,
+      "step": 17875
+    },
+    {
+      "epoch": 0.7526736842105263,
+      "grad_norm": 0.408203125,
+      "learning_rate": 7.454651788674796e-05,
+      "loss": 3.1255,
+      "step": 17876
+    },
+    {
+      "epoch": 0.7527157894736842,
+      "grad_norm": 0.4140625,
+      "learning_rate": 7.452248139795148e-05,
+      "loss": 3.1371,
+      "step": 17877
+    },
+    {
+      "epoch": 0.7527578947368421,
+      "grad_norm": 0.44921875,
+      "learning_rate": 7.449844810614922e-05,
+      "loss": 3.0477,
+      "step": 17878
+    },
+    {
+      "epoch": 0.7528,
+      "grad_norm": 0.45703125,
+      "learning_rate": 7.447441801177903e-05,
+      "loss": 2.9108,
+      "step": 17879
+    },
+    {
+      "epoch": 0.7528421052631579,
+      "grad_norm": 0.439453125,
+      "learning_rate": 7.445039111527865e-05,
+      "loss": 3.1906,
+      "step": 17880
+    },
+    {
+      "epoch": 0.7528842105263158,
+      "grad_norm": 0.4296875,
+      "learning_rate": 7.442636741708602e-05,
+      "loss": 2.8447,
+      "step": 17881
+    },
+    {
+      "epoch": 0.7529263157894737,
+      "grad_norm": 0.50390625,
+      "learning_rate": 7.440234691763853e-05,
+      "loss": 3.03,
+      "step": 17882
+    },
+    {
+      "epoch": 0.7529684210526316,
+      "grad_norm": 0.439453125,
+      "learning_rate": 7.437832961737403e-05,
+      "loss": 2.5647,
+      "step": 17883
+    },
+    {
+      "epoch": 0.7530105263157895,
+      "grad_norm": 0.443359375,
+      "learning_rate": 7.435431551673005e-05,
+      "loss": 3.2236,
+      "step": 17884
+    },
+    {
+      "epoch": 0.7530526315789474,
+      "grad_norm": 0.423828125,
+      "learning_rate": 7.433030461614404e-05,
+      "loss": 3.224,
+      "step": 17885
+    },
+    {
+      "epoch": 0.7530947368421053,
+      "grad_norm": 0.41796875,
+      "learning_rate": 7.430629691605347e-05,
+      "loss": 3.074,
+      "step": 17886
+    },
+    {
+      "epoch": 0.7531368421052631,
+      "grad_norm": 0.44921875,
+      "learning_rate": 7.428229241689577e-05,
+      "loss": 3.2722,
+      "step": 17887
+    },
+    {
+      "epoch": 0.7531789473684211,
+      "grad_norm": 0.45703125,
+      "learning_rate": 7.42582911191082e-05,
+      "loss": 2.7538,
+      "step": 17888
+    },
+    {
+      "epoch": 0.7532210526315789,
+      "grad_norm": 0.421875,
+      "learning_rate": 7.423429302312804e-05,
+      "loss": 3.3023,
+      "step": 17889
+    },
+    {
+      "epoch": 0.7532631578947369,
+      "grad_norm": 0.447265625,
+      "learning_rate": 7.421029812939259e-05,
+      "loss": 3.398,
+      "step": 17890
+    },
+    {
+      "epoch": 0.7533052631578947,
+      "grad_norm": 0.419921875,
+      "learning_rate": 7.418630643833899e-05,
+      "loss": 3.2064,
+      "step": 17891
+    },
+    {
+      "epoch": 0.7533473684210527,
+      "grad_norm": 0.41796875,
+      "learning_rate": 7.416231795040429e-05,
+      "loss": 2.9016,
+      "step": 17892
+    },
+    {
+      "epoch": 0.7533894736842105,
+      "grad_norm": 0.42578125,
+      "learning_rate": 7.413833266602557e-05,
+      "loss": 3.1228,
+      "step": 17893
+    },
+    {
+      "epoch": 0.7534315789473685,
+      "grad_norm": 0.427734375,
+      "learning_rate": 7.411435058563981e-05,
+      "loss": 2.9197,
+      "step": 17894
+    },
+    {
+      "epoch": 0.7534736842105263,
+      "grad_norm": 0.427734375,
+      "learning_rate": 7.409037170968391e-05,
+      "loss": 3.3076,
+      "step": 17895
+    },
+    {
+      "epoch": 0.7535157894736842,
+      "grad_norm": 0.478515625,
+      "learning_rate": 7.406639603859467e-05,
+      "loss": 2.915,
+      "step": 17896
+    },
+    {
+      "epoch": 0.7535578947368421,
+      "grad_norm": 0.47265625,
+      "learning_rate": 7.404242357280908e-05,
+      "loss": 3.2305,
+      "step": 17897
+    },
+    {
+      "epoch": 0.7536,
+      "grad_norm": 0.447265625,
+      "learning_rate": 7.401845431276377e-05,
+      "loss": 3.1816,
+      "step": 17898
+    },
+    {
+      "epoch": 0.7536421052631579,
+      "grad_norm": 0.40625,
+      "learning_rate": 7.399448825889546e-05,
+      "loss": 2.893,
+      "step": 17899
+    },
+    {
+      "epoch": 0.7536842105263157,
+      "grad_norm": 0.41015625,
+      "learning_rate": 7.397052541164079e-05,
+      "loss": 3.0306,
+      "step": 17900
+    },
+    {
+      "epoch": 0.7537263157894737,
+      "grad_norm": 0.419921875,
+      "learning_rate": 7.394656577143629e-05,
+      "loss": 3.0031,
+      "step": 17901
+    },
+    {
+      "epoch": 0.7537684210526315,
+      "grad_norm": 0.4453125,
+      "learning_rate": 7.392260933871852e-05,
+      "loss": 3.2557,
+      "step": 17902
+    },
+    {
+      "epoch": 0.7538105263157895,
+      "grad_norm": 0.431640625,
+      "learning_rate": 7.389865611392394e-05,
+      "loss": 3.0523,
+      "step": 17903
+    },
+    {
+      "epoch": 0.7538526315789473,
+      "grad_norm": 0.421875,
+      "learning_rate": 7.387470609748883e-05,
+      "loss": 2.7884,
+      "step": 17904
+    },
+    {
+      "epoch": 0.7538947368421053,
+      "grad_norm": 0.419921875,
+      "learning_rate": 7.385075928984975e-05,
+      "loss": 3.1049,
+      "step": 17905
+    },
+    {
+      "epoch": 0.7539368421052631,
+      "grad_norm": 0.4140625,
+      "learning_rate": 7.382681569144284e-05,
+      "loss": 2.9251,
+      "step": 17906
+    },
+    {
+      "epoch": 0.7539789473684211,
+      "grad_norm": 0.4609375,
+      "learning_rate": 7.380287530270435e-05,
+      "loss": 3.2891,
+      "step": 17907
+    },
+    {
+      "epoch": 0.7540210526315789,
+      "grad_norm": 0.4609375,
+      "learning_rate": 7.377893812407046e-05,
+      "loss": 3.272,
+      "step": 17908
+    },
+    {
+      "epoch": 0.7540631578947369,
+      "grad_norm": 0.4453125,
+      "learning_rate": 7.375500415597727e-05,
+      "loss": 2.7189,
+      "step": 17909
+    },
+    {
+      "epoch": 0.7541052631578947,
+      "grad_norm": 0.458984375,
+      "learning_rate": 7.373107339886084e-05,
+      "loss": 2.7377,
+      "step": 17910
+    },
+    {
+      "epoch": 0.7541473684210527,
+      "grad_norm": 0.421875,
+      "learning_rate": 7.370714585315704e-05,
+      "loss": 3.4069,
+      "step": 17911
+    },
+    {
+      "epoch": 0.7541894736842105,
+      "grad_norm": 0.416015625,
+      "learning_rate": 7.368322151930201e-05,
+      "loss": 2.7286,
+      "step": 17912
+    },
+    {
+      "epoch": 0.7542315789473685,
+      "grad_norm": 0.423828125,
+      "learning_rate": 7.36593003977315e-05,
+      "loss": 2.8868,
+      "step": 17913
+    },
+    {
+      "epoch": 0.7542736842105263,
+      "grad_norm": 0.4296875,
+      "learning_rate": 7.363538248888136e-05,
+      "loss": 2.8195,
+      "step": 17914
+    },
+    {
+      "epoch": 0.7543157894736842,
+      "grad_norm": 0.416015625,
+      "learning_rate": 7.361146779318731e-05,
+      "loss": 2.8003,
+      "step": 17915
+    },
+    {
+      "epoch": 0.7543578947368421,
+      "grad_norm": 0.427734375,
+      "learning_rate": 7.358755631108507e-05,
+      "loss": 2.9281,
+      "step": 17916
+    },
+    {
+      "epoch": 0.7544,
+      "grad_norm": 0.458984375,
+      "learning_rate": 7.35636480430103e-05,
+      "loss": 2.9806,
+      "step": 17917
+    },
+    {
+      "epoch": 0.7544421052631579,
+      "grad_norm": 0.4140625,
+      "learning_rate": 7.353974298939847e-05,
+      "loss": 2.6939,
+      "step": 17918
+    },
+    {
+      "epoch": 0.7544842105263158,
+      "grad_norm": 0.44921875,
+      "learning_rate": 7.351584115068535e-05,
+      "loss": 2.8462,
+      "step": 17919
+    },
+    {
+      "epoch": 0.7545263157894737,
+      "grad_norm": 0.416015625,
+      "learning_rate": 7.349194252730607e-05,
+      "loss": 3.0892,
+      "step": 17920
+    },
+    {
+      "epoch": 0.7545684210526316,
+      "grad_norm": 0.427734375,
+      "learning_rate": 7.346804711969632e-05,
+      "loss": 3.2774,
+      "step": 17921
+    },
+    {
+      "epoch": 0.7546105263157895,
+      "grad_norm": 0.4375,
+      "learning_rate": 7.344415492829129e-05,
+      "loss": 3.2071,
+      "step": 17922
+    },
+    {
+      "epoch": 0.7546526315789474,
+      "grad_norm": 0.427734375,
+      "learning_rate": 7.342026595352628e-05,
+      "loss": 3.2088,
+      "step": 17923
+    },
+    {
+      "epoch": 0.7546947368421053,
+      "grad_norm": 0.416015625,
+      "learning_rate": 7.339638019583666e-05,
+      "loss": 3.3859,
+      "step": 17924
+    },
+    {
+      "epoch": 0.7547368421052632,
+      "grad_norm": 0.423828125,
+      "learning_rate": 7.337249765565737e-05,
+      "loss": 2.9565,
+      "step": 17925
+    },
+    {
+      "epoch": 0.7547789473684211,
+      "grad_norm": 0.43359375,
+      "learning_rate": 7.33486183334238e-05,
+      "loss": 3.3,
+      "step": 17926
+    },
+    {
+      "epoch": 0.754821052631579,
+      "grad_norm": 0.4609375,
+      "learning_rate": 7.33247422295707e-05,
+      "loss": 3.1472,
+      "step": 17927
+    },
+    {
+      "epoch": 0.7548631578947368,
+      "grad_norm": 0.423828125,
+      "learning_rate": 7.330086934453326e-05,
+      "loss": 3.2403,
+      "step": 17928
+    },
+    {
+      "epoch": 0.7549052631578947,
+      "grad_norm": 0.404296875,
+      "learning_rate": 7.327699967874638e-05,
+      "loss": 3.327,
+      "step": 17929
+    },
+    {
+      "epoch": 0.7549473684210526,
+      "grad_norm": 0.435546875,
+      "learning_rate": 7.325313323264493e-05,
+      "loss": 2.9435,
+      "step": 17930
+    },
+    {
+      "epoch": 0.7549894736842105,
+      "grad_norm": 0.431640625,
+      "learning_rate": 7.322927000666371e-05,
+      "loss": 3.4752,
+      "step": 17931
+    },
+    {
+      "epoch": 0.7550315789473684,
+      "grad_norm": 0.43359375,
+      "learning_rate": 7.320541000123743e-05,
+      "loss": 3.0975,
+      "step": 17932
+    },
+    {
+      "epoch": 0.7550736842105263,
+      "grad_norm": 0.4296875,
+      "learning_rate": 7.3181553216801e-05,
+      "loss": 3.533,
+      "step": 17933
+    },
+    {
+      "epoch": 0.7551157894736842,
+      "grad_norm": 0.443359375,
+      "learning_rate": 7.315769965378877e-05,
+      "loss": 3.1239,
+      "step": 17934
+    },
+    {
+      "epoch": 0.7551578947368421,
+      "grad_norm": 0.451171875,
+      "learning_rate": 7.313384931263565e-05,
+      "loss": 3.1477,
+      "step": 17935
+    },
+    {
+      "epoch": 0.7552,
+      "grad_norm": 0.427734375,
+      "learning_rate": 7.311000219377578e-05,
+      "loss": 3.0842,
+      "step": 17936
+    },
+    {
+      "epoch": 0.7552421052631579,
+      "grad_norm": 0.427734375,
+      "learning_rate": 7.308615829764395e-05,
+      "loss": 2.9766,
+      "step": 17937
+    },
+    {
+      "epoch": 0.7552842105263158,
+      "grad_norm": 0.408203125,
+      "learning_rate": 7.306231762467445e-05,
+      "loss": 3.3122,
+      "step": 17938
+    },
+    {
+      "epoch": 0.7553263157894737,
+      "grad_norm": 0.4375,
+      "learning_rate": 7.303848017530151e-05,
+      "loss": 3.195,
+      "step": 17939
+    },
+    {
+      "epoch": 0.7553684210526316,
+      "grad_norm": 0.435546875,
+      "learning_rate": 7.301464594995974e-05,
+      "loss": 2.8319,
+      "step": 17940
+    },
+    {
+      "epoch": 0.7554105263157894,
+      "grad_norm": 0.4296875,
+      "learning_rate": 7.299081494908299e-05,
+      "loss": 2.7748,
+      "step": 17941
+    },
+    {
+      "epoch": 0.7554526315789474,
+      "grad_norm": 0.40234375,
+      "learning_rate": 7.29669871731058e-05,
+      "loss": 2.7599,
+      "step": 17942
+    },
+    {
+      "epoch": 0.7554947368421052,
+      "grad_norm": 0.416015625,
+      "learning_rate": 7.294316262246189e-05,
+      "loss": 3.0617,
+      "step": 17943
+    },
+    {
+      "epoch": 0.7555368421052632,
+      "grad_norm": 0.412109375,
+      "learning_rate": 7.291934129758565e-05,
+      "loss": 3.6084,
+      "step": 17944
+    },
+    {
+      "epoch": 0.755578947368421,
+      "grad_norm": 0.427734375,
+      "learning_rate": 7.289552319891094e-05,
+      "loss": 3.1954,
+      "step": 17945
+    },
+    {
+      "epoch": 0.755621052631579,
+      "grad_norm": 0.416015625,
+      "learning_rate": 7.28717083268717e-05,
+      "loss": 2.8634,
+      "step": 17946
+    },
+    {
+      "epoch": 0.7556631578947368,
+      "grad_norm": 0.40625,
+      "learning_rate": 7.284789668190181e-05,
+      "loss": 2.6953,
+      "step": 17947
+    },
+    {
+      "epoch": 0.7557052631578948,
+      "grad_norm": 0.416015625,
+      "learning_rate": 7.282408826443505e-05,
+      "loss": 3.3015,
+      "step": 17948
+    },
+    {
+      "epoch": 0.7557473684210526,
+      "grad_norm": 0.4140625,
+      "learning_rate": 7.280028307490535e-05,
+      "loss": 3.1275,
+      "step": 17949
+    },
+    {
+      "epoch": 0.7557894736842106,
+      "grad_norm": 0.45703125,
+      "learning_rate": 7.277648111374616e-05,
+      "loss": 3.4385,
+      "step": 17950
+    },
+    {
+      "epoch": 0.7558315789473684,
+      "grad_norm": 0.400390625,
+      "learning_rate": 7.275268238139133e-05,
+      "loss": 2.8682,
+      "step": 17951
+    },
+    {
+      "epoch": 0.7558736842105264,
+      "grad_norm": 0.4140625,
+      "learning_rate": 7.272888687827434e-05,
+      "loss": 3.1425,
+      "step": 17952
+    },
+    {
+      "epoch": 0.7559157894736842,
+      "grad_norm": 0.443359375,
+      "learning_rate": 7.270509460482878e-05,
+      "loss": 2.4862,
+      "step": 17953
+    },
+    {
+      "epoch": 0.755957894736842,
+      "grad_norm": 0.435546875,
+      "learning_rate": 7.268130556148805e-05,
+      "loss": 3.1172,
+      "step": 17954
+    },
+    {
+      "epoch": 0.756,
+      "grad_norm": 0.4375,
+      "learning_rate": 7.265751974868553e-05,
+      "loss": 3.3407,
+      "step": 17955
+    },
+    {
+      "epoch": 0.7560421052631578,
+      "grad_norm": 0.4140625,
+      "learning_rate": 7.263373716685481e-05,
+      "loss": 2.8891,
+      "step": 17956
+    },
+    {
+      "epoch": 0.7560842105263158,
+      "grad_norm": 0.4375,
+      "learning_rate": 7.260995781642881e-05,
+      "loss": 3.3673,
+      "step": 17957
+    },
+    {
+      "epoch": 0.7561263157894736,
+      "grad_norm": 0.42578125,
+      "learning_rate": 7.258618169784106e-05,
+      "loss": 3.0176,
+      "step": 17958
+    },
+    {
+      "epoch": 0.7561684210526316,
+      "grad_norm": 0.439453125,
+      "learning_rate": 7.256240881152462e-05,
+      "loss": 3.1554,
+      "step": 17959
+    },
+    {
+      "epoch": 0.7562105263157894,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.253863915791262e-05,
+      "loss": 3.3399,
+      "step": 17960
+    },
+    {
+      "epoch": 0.7562526315789474,
+      "grad_norm": 0.400390625,
+      "learning_rate": 7.25148727374381e-05,
+      "loss": 2.9801,
+      "step": 17961
+    },
+    {
+      "epoch": 0.7562947368421052,
+      "grad_norm": 0.412109375,
+      "learning_rate": 7.249110955053406e-05,
+      "loss": 3.0449,
+      "step": 17962
+    },
+    {
+      "epoch": 0.7563368421052632,
+      "grad_norm": 0.42578125,
+      "learning_rate": 7.246734959763343e-05,
+      "loss": 3.2464,
+      "step": 17963
+    },
+    {
+      "epoch": 0.756378947368421,
+      "grad_norm": 0.416015625,
+      "learning_rate": 7.244359287916901e-05,
+      "loss": 3.0581,
+      "step": 17964
+    },
+    {
+      "epoch": 0.756421052631579,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.241983939557381e-05,
+      "loss": 3.0583,
+      "step": 17965
+    },
+    {
+      "epoch": 0.7564631578947368,
+      "grad_norm": 0.412109375,
+      "learning_rate": 7.239608914728049e-05,
+      "loss": 3.0835,
+      "step": 17966
+    },
+    {
+      "epoch": 0.7565052631578948,
+      "grad_norm": 0.4296875,
+      "learning_rate": 7.237234213472174e-05,
+      "loss": 2.927,
+      "step": 17967
+    },
+    {
+      "epoch": 0.7565473684210526,
+      "grad_norm": 0.423828125,
+      "learning_rate": 7.234859835833022e-05,
+      "loss": 2.768,
+      "step": 17968
+    },
+    {
+      "epoch": 0.7565894736842105,
+      "grad_norm": 0.455078125,
+      "learning_rate": 7.232485781853849e-05,
+      "loss": 3.6237,
+      "step": 17969
+    },
+    {
+      "epoch": 0.7566315789473684,
+      "grad_norm": 0.431640625,
+      "learning_rate": 7.2301120515779e-05,
+      "loss": 3.1102,
+      "step": 17970
+    },
+    {
+      "epoch": 0.7566736842105263,
+      "grad_norm": 0.423828125,
+      "learning_rate": 7.227738645048443e-05,
+      "loss": 3.2453,
+      "step": 17971
+    },
+    {
+      "epoch": 0.7567157894736842,
+      "grad_norm": 0.404296875,
+      "learning_rate": 7.225365562308703e-05,
+      "loss": 3.1052,
+      "step": 17972
+    },
+    {
+      "epoch": 0.7567578947368421,
+      "grad_norm": 0.43359375,
+      "learning_rate": 7.222992803401918e-05,
+      "loss": 3.3128,
+      "step": 17973
+    },
+    {
+      "epoch": 0.7568,
+      "grad_norm": 0.4140625,
+      "learning_rate": 7.22062036837132e-05,
+      "loss": 3.2558,
+      "step": 17974
+    },
+    {
+      "epoch": 0.7568421052631579,
+      "grad_norm": 0.447265625,
+      "learning_rate": 7.218248257260127e-05,
+      "loss": 2.6788,
+      "step": 17975
+    },
+    {
+      "epoch": 0.7568842105263158,
+      "grad_norm": 0.431640625,
+      "learning_rate": 7.215876470111558e-05,
+      "loss": 3.2525,
+      "step": 17976
+    },
+    {
+      "epoch": 0.7569263157894737,
+      "grad_norm": 0.427734375,
+      "learning_rate": 7.213505006968815e-05,
+      "loss": 3.3588,
+      "step": 17977
+    },
+    {
+      "epoch": 0.7569684210526316,
+      "grad_norm": 0.419921875,
+      "learning_rate": 7.21113386787513e-05,
+      "loss": 3.3014,
+      "step": 17978
+    },
+    {
+      "epoch": 0.7570105263157895,
+      "grad_norm": 0.427734375,
+      "learning_rate": 7.208763052873667e-05,
+      "loss": 2.8016,
+      "step": 17979
+    },
+    {
+      "epoch": 0.7570526315789474,
+      "grad_norm": 0.431640625,
+      "learning_rate": 7.20639256200765e-05,
+      "loss": 3.109,
+      "step": 17980
+    },
+    {
+      "epoch": 0.7570947368421053,
+      "grad_norm": 0.451171875,
+      "learning_rate": 7.204022395320251e-05,
+      "loss": 3.0281,
+      "step": 17981
+    },
+    {
+      "epoch": 0.7571368421052631,
+      "grad_norm": 0.40625,
+      "learning_rate": 7.201652552854654e-05,
+      "loss": 2.9049,
+      "step": 17982
+    },
+    {
+      "epoch": 0.7571789473684211,
+      "grad_norm": 0.435546875,
+      "learning_rate": 7.199283034654036e-05,
+      "loss": 3.2437,
+      "step": 17983
+    },
+    {
+      "epoch": 0.7572210526315789,
+      "grad_norm": 0.404296875,
+      "learning_rate": 7.196913840761557e-05,
+      "loss": 3.3204,
+      "step": 17984
+    },
+    {
+      "epoch": 0.7572631578947369,
+      "grad_norm": 0.421875,
+      "learning_rate": 7.194544971220407e-05,
+      "loss": 2.6717,
+      "step": 17985
+    },
+    {
+      "epoch": 0.7573052631578947,
+      "grad_norm": 0.470703125,
+      "learning_rate": 7.19217642607371e-05,
+      "loss": 2.8856,
+      "step": 17986
+    },
+    {
+      "epoch": 0.7573473684210527,
+      "grad_norm": 0.439453125,
+      "learning_rate": 7.189808205364645e-05,
+      "loss": 3.191,
+      "step": 17987
+    },
+    {
+      "epoch": 0.7573894736842105,
+      "grad_norm": 0.455078125,
+      "learning_rate": 7.187440309136348e-05,
+      "loss": 3.4122,
+      "step": 17988
+    },
+    {
+      "epoch": 0.7574315789473685,
+      "grad_norm": 0.4765625,
+      "learning_rate": 7.185072737431957e-05,
+      "loss": 3.0196,
+      "step": 17989
+    },
+    {
+      "epoch": 0.7574736842105263,
+      "grad_norm": 0.421875,
+      "learning_rate": 7.182705490294614e-05,
+      "loss": 3.4915,
+      "step": 17990
+    },
+    {
+      "epoch": 0.7575157894736843,
+      "grad_norm": 0.396484375,
+      "learning_rate": 7.18033856776743e-05,
+      "loss": 2.9076,
+      "step": 17991
+    },
+    {
+      "epoch": 0.7575578947368421,
+      "grad_norm": 0.41796875,
+      "learning_rate": 7.17797196989356e-05,
+      "loss": 3.1533,
+      "step": 17992
+    },
+    {
+      "epoch": 0.7576,
+      "grad_norm": 0.431640625,
+      "learning_rate": 7.175605696716082e-05,
+      "loss": 2.7835,
+      "step": 17993
+    },
+    {
+      "epoch": 0.7576421052631579,
+      "grad_norm": 0.408203125,
+      "learning_rate": 7.17323974827814e-05,
+      "loss": 3.0588,
+      "step": 17994
+    },
+    {
+      "epoch": 0.7576842105263157,
+      "grad_norm": 0.421875,
+      "learning_rate": 7.170874124622812e-05,
+      "loss": 3.0316,
+      "step": 17995
+    },
+    {
+      "epoch": 0.7577263157894737,
+      "grad_norm": 0.42578125,
+      "learning_rate": 7.168508825793216e-05,
+      "loss": 3.2328,
+      "step": 17996
+    },
+    {
+      "epoch": 0.7577684210526315,
+      "grad_norm": 0.435546875,
+      "learning_rate": 7.166143851832434e-05,
+      "loss": 3.0991,
+      "step": 17997
+    },
+    {
+      "epoch": 0.7578105263157895,
+      "grad_norm": 0.427734375,
+      "learning_rate": 7.163779202783555e-05,
+      "loss": 3.1678,
+      "step": 17998
+    },
+    {
+      "epoch": 0.7578526315789473,
+      "grad_norm": 0.423828125,
+      "learning_rate": 7.161414878689677e-05,
+      "loss": 3.1859,
+      "step": 17999
+    },
+    {
+      "epoch": 0.7578947368421053,
+      "grad_norm": 0.41015625,
+      "learning_rate": 7.159050879593843e-05,
+      "loss": 3.1153,
+      "step": 18000
+    },
+    {
+      "epoch": 0.7578947368421053,
+      "eval_loss": 3.08333420753479,
+      "eval_runtime": 335.6342,
+      "eval_samples_per_second": 44.692,
+      "eval_steps_per_second": 5.586,
+      "step": 18000
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 23750,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 3000,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 2,
+        "early_stopping_threshold": 0.0
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 0
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.28297586098176e+17,
+  "train_batch_size": 12,
+  "trial_name": null,
+  "trial_params": null
+}