diff --git "a/checkpoint-3880/trainer_state.json" "b/checkpoint-3880/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-3880/trainer_state.json"
@@ -0,0 +1,23451 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.999613451874758,
+  "eval_steps": 200,
+  "global_step": 3880,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 6e-06,
+      "loss": 0.4812,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2e-05,
+      "loss": 0.516,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.8e-05,
+      "loss": 0.4684,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4e-05,
+      "loss": 0.7239,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3e-05,
+      "loss": 0.5741,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6e-05,
+      "loss": 0.4694,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2e-05,
+      "loss": 0.3564,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8e-05,
+      "loss": 0.3293,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.4000000000000005e-05,
+      "loss": 0.3084,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6e-05,
+      "loss": 0.4094,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.9999994444092705e-05,
+      "loss": 0.2397,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.9999977776372875e-05,
+      "loss": 0.5541,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.999994999684669e-05,
+      "loss": 0.6926,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.999991110552442e-05,
+      "loss": 0.2365,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.9999861102420494e-05,
+      "loss": 0.5255,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.999979998755342e-05,
+      "loss": 0.6014,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.9999727760945846e-05,
+      "loss": 0.3908,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.999964442262451e-05,
+      "loss": 0.6003,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.999954997262029e-05,
+      "loss": 0.3084,
+      "step": 19
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999944441096816e-05,
+      "loss": 0.5446,
+      "step": 20
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.9999327737707225e-05,
+      "loss": 0.6111,
+      "step": 21
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.99991999528807e-05,
+      "loss": 0.6467,
+      "step": 22
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.9999061056535925e-05,
+      "loss": 0.3795,
+      "step": 23
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999891104872433e-05,
+      "loss": 0.7339,
+      "step": 24
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999874992950148e-05,
+      "loss": 0.7399,
+      "step": 25
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999857769892706e-05,
+      "loss": 0.262,
+      "step": 26
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999839435706486e-05,
+      "loss": 0.2956,
+      "step": 27
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999819990398279e-05,
+      "loss": 0.88,
+      "step": 28
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999799433975287e-05,
+      "loss": 0.4968,
+      "step": 29
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999777766445124e-05,
+      "loss": 0.2416,
+      "step": 30
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999754987815816e-05,
+      "loss": 0.3191,
+      "step": 31
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.999731098095801e-05,
+      "loss": 0.8684,
+      "step": 32
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999706097293925e-05,
+      "loss": 0.4201,
+      "step": 33
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999679985419451e-05,
+      "loss": 0.866,
+      "step": 34
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.9996527624820475e-05,
+      "loss": 0.3832,
+      "step": 35
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.9996244284918004e-05,
+      "loss": 0.293,
+      "step": 36
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.9995949834592026e-05,
+      "loss": 0.5057,
+      "step": 37
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999564427395162e-05,
+      "loss": 0.634,
+      "step": 38
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.9995327603109954e-05,
+      "loss": 0.4579,
+      "step": 39
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999499982218433e-05,
+      "loss": 0.385,
+      "step": 40
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999466093129615e-05,
+      "loss": 0.5858,
+      "step": 41
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.9994310930570924e-05,
+      "loss": 0.5436,
+      "step": 42
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999394982013831e-05,
+      "loss": 0.564,
+      "step": 43
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999357760013205e-05,
+      "loss": 0.6269,
+      "step": 44
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.999319427069002e-05,
+      "loss": 0.3171,
+      "step": 45
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.9992799831954206e-05,
+      "loss": 0.2316,
+      "step": 46
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.999239428407068e-05,
+      "loss": 0.2547,
+      "step": 47
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.999197762718968e-05,
+      "loss": 0.2131,
+      "step": 48
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.999154986146553e-05,
+      "loss": 0.702,
+      "step": 49
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.999111098705666e-05,
+      "loss": 0.6374,
+      "step": 50
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.9990661004125635e-05,
+      "loss": 0.8094,
+      "step": 51
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.999019991283913e-05,
+      "loss": 0.3438,
+      "step": 52
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998972771336792e-05,
+      "loss": 0.6793,
+      "step": 53
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998924440588692e-05,
+      "loss": 0.3126,
+      "step": 54
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.9988749990575126e-05,
+      "loss": 0.2881,
+      "step": 55
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.9988244467615675e-05,
+      "loss": 0.2607,
+      "step": 56
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.99877278371958e-05,
+      "loss": 0.6432,
+      "step": 57
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5.998720009950688e-05,
+      "loss": 0.2747,
+      "step": 58
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.9986661254744366e-05,
+      "loss": 0.5452,
+      "step": 59
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.9986111303107855e-05,
+      "loss": 0.5666,
+      "step": 60
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.9985550244801026e-05,
+      "loss": 0.4007,
+      "step": 61
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.9984978080031714e-05,
+      "loss": 0.258,
+      "step": 62
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.9984394809011845e-05,
+      "loss": 0.3686,
+      "step": 63
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.998380043195744e-05,
+      "loss": 0.5071,
+      "step": 64
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.998319494908867e-05,
+      "loss": 0.6607,
+      "step": 65
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.9982578360629783e-05,
+      "loss": 0.5264,
+      "step": 66
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.998195066680918e-05,
+      "loss": 0.5879,
+      "step": 67
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.998131186785935e-05,
+      "loss": 0.4732,
+      "step": 68
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.998066196401689e-05,
+      "loss": 0.2465,
+      "step": 69
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.998000095552253e-05,
+      "loss": 0.2302,
+      "step": 70
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5.9979328842621105e-05,
+      "loss": 0.5896,
+      "step": 71
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.997864562556155e-05,
+      "loss": 0.3705,
+      "step": 72
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.997795130459694e-05,
+      "loss": 0.317,
+      "step": 73
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.997724587998443e-05,
+      "loss": 0.2679,
+      "step": 74
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.997652935198532e-05,
+      "loss": 0.2633,
+      "step": 75
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.9975801720865e-05,
+      "loss": 0.2387,
+      "step": 76
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.997506298689298e-05,
+      "loss": 0.244,
+      "step": 77
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.997431315034289e-05,
+      "loss": 0.3264,
+      "step": 78
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.9973552211492445e-05,
+      "loss": 0.523,
+      "step": 79
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.9972780170623506e-05,
+      "loss": 0.4393,
+      "step": 80
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.9971997028022045e-05,
+      "loss": 0.3313,
+      "step": 81
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.997120278397811e-05,
+      "loss": 0.4794,
+      "step": 82
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.9970397438785905e-05,
+      "loss": 0.5435,
+      "step": 83
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.996958099274371e-05,
+      "loss": 0.5704,
+      "step": 84
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.996875344615393e-05,
+      "loss": 0.5101,
+      "step": 85
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.996791479932309e-05,
+      "loss": 0.2749,
+      "step": 86
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.9967065052561825e-05,
+      "loss": 0.7018,
+      "step": 87
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.996620420618486e-05,
+      "loss": 0.2563,
+      "step": 88
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.996533226051106e-05,
+      "loss": 0.4292,
+      "step": 89
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.996444921586339e-05,
+      "loss": 0.5187,
+      "step": 90
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.996355507256892e-05,
+      "loss": 0.4349,
+      "step": 91
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.9962649830958825e-05,
+      "loss": 0.7261,
+      "step": 92
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.996173349136841e-05,
+      "loss": 0.5646,
+      "step": 93
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.996080605413708e-05,
+      "loss": 0.9367,
+      "step": 94
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.995986751960836e-05,
+      "loss": 0.4413,
+      "step": 95
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.9958917888129874e-05,
+      "loss": 0.5579,
+      "step": 96
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5.995795716005335e-05,
+      "loss": 0.2298,
+      "step": 97
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.995698533573465e-05,
+      "loss": 0.6869,
+      "step": 98
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.9956002415533715e-05,
+      "loss": 0.3528,
+      "step": 99
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.995500839981463e-05,
+      "loss": 0.7105,
+      "step": 100
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.995400328894556e-05,
+      "loss": 0.5522,
+      "step": 101
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.99529870832988e-05,
+      "loss": 0.3905,
+      "step": 102
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.995195978325073e-05,
+      "loss": 0.3428,
+      "step": 103
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.995092138918188e-05,
+      "loss": 0.4749,
+      "step": 104
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.9949871901476845e-05,
+      "loss": 0.2427,
+      "step": 105
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.9948811320524356e-05,
+      "loss": 0.2339,
+      "step": 106
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.994773964671725e-05,
+      "loss": 0.2152,
+      "step": 107
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.994665688045247e-05,
+      "loss": 0.3904,
+      "step": 108
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5.994556302213104e-05,
+      "loss": 0.3831,
+      "step": 109
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.994445807215816e-05,
+      "loss": 0.5339,
+      "step": 110
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.994334203094306e-05,
+      "loss": 0.3612,
+      "step": 111
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.9942214898899135e-05,
+      "loss": 0.5652,
+      "step": 112
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.994107667644387e-05,
+      "loss": 0.3806,
+      "step": 113
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.993992736399884e-05,
+      "loss": 0.5187,
+      "step": 114
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.9938766961989755e-05,
+      "loss": 0.4882,
+      "step": 115
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.993759547084641e-05,
+      "loss": 0.7215,
+      "step": 116
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.9936412891002736e-05,
+      "loss": 0.7016,
+      "step": 117
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.993521922289674e-05,
+      "loss": 0.4595,
+      "step": 118
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.993401446697054e-05,
+      "loss": 0.318,
+      "step": 119
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.99327986236704e-05,
+      "loss": 0.6752,
+      "step": 120
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.9931571693446624e-05,
+      "loss": 0.5383,
+      "step": 121
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5.993033367675369e-05,
+      "loss": 0.9896,
+      "step": 122
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.992908457405014e-05,
+      "loss": 0.3923,
+      "step": 123
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.992782438579863e-05,
+      "loss": 0.6535,
+      "step": 124
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.992655311246593e-05,
+      "loss": 0.2195,
+      "step": 125
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.9925270754522914e-05,
+      "loss": 0.366,
+      "step": 126
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.9923977312444563e-05,
+      "loss": 0.3465,
+      "step": 127
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.992267278670994e-05,
+      "loss": 0.3339,
+      "step": 128
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.992135717780226e-05,
+      "loss": 0.2815,
+      "step": 129
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.9920030486208795e-05,
+      "loss": 0.456,
+      "step": 130
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.991869271242096e-05,
+      "loss": 0.2442,
+      "step": 131
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.9917343856934235e-05,
+      "loss": 0.2589,
+      "step": 132
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.991598392024825e-05,
+      "loss": 0.3142,
+      "step": 133
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.991461290286672e-05,
+      "loss": 0.3032,
+      "step": 134
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5.991323080529744e-05,
+      "loss": 0.4262,
+      "step": 135
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.9911837628052344e-05,
+      "loss": 0.956,
+      "step": 136
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.991043337164746e-05,
+      "loss": 0.8417,
+      "step": 137
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.99090180366029e-05,
+      "loss": 0.3149,
+      "step": 138
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.9907591623442905e-05,
+      "loss": 0.501,
+      "step": 139
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.9906154132695816e-05,
+      "loss": 0.2584,
+      "step": 140
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.990470556489407e-05,
+      "loss": 0.3148,
+      "step": 141
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.990324592057419e-05,
+      "loss": 0.407,
+      "step": 142
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.990177520027684e-05,
+      "loss": 0.7201,
+      "step": 143
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.990029340454676e-05,
+      "loss": 0.4646,
+      "step": 144
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.9898800533932785e-05,
+      "loss": 0.3544,
+      "step": 145
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.989729658898788e-05,
+      "loss": 0.3569,
+      "step": 146
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.989578157026909e-05,
+      "loss": 0.3161,
+      "step": 147
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5.9894255478337575e-05,
+      "loss": 0.9521,
+      "step": 148
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.989271831375858e-05,
+      "loss": 0.2177,
+      "step": 149
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.989117007710147e-05,
+      "loss": 0.6865,
+      "step": 150
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.98896107689397e-05,
+      "loss": 0.8173,
+      "step": 151
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.9888040389850827e-05,
+      "loss": 0.7679,
+      "step": 152
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.98864589404165e-05,
+      "loss": 0.5417,
+      "step": 153
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.98848664212225e-05,
+      "loss": 0.948,
+      "step": 154
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.988326283285867e-05,
+      "loss": 0.386,
+      "step": 155
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.988164817591898e-05,
+      "loss": 0.6286,
+      "step": 156
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.9880022451001476e-05,
+      "loss": 0.3574,
+      "step": 157
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.987838565870832e-05,
+      "loss": 0.4403,
+      "step": 158
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.987673779964578e-05,
+      "loss": 0.2804,
+      "step": 159
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.98750788744242e-05,
+      "loss": 0.3311,
+      "step": 160
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.9873408883658046e-05,
+      "loss": 0.278,
+      "step": 161
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.987172782796587e-05,
+      "loss": 0.7136,
+      "step": 162
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.987003570797031e-05,
+      "loss": 0.6248,
+      "step": 163
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.986833252429814e-05,
+      "loss": 0.2343,
+      "step": 164
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.986661827758019e-05,
+      "loss": 0.3647,
+      "step": 165
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.9864892968451415e-05,
+      "loss": 0.6634,
+      "step": 166
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.986315659755086e-05,
+      "loss": 0.3824,
+      "step": 167
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.986140916552166e-05,
+      "loss": 0.2113,
+      "step": 168
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.985965067301107e-05,
+      "loss": 0.423,
+      "step": 169
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.98578811206704e-05,
+      "loss": 0.4876,
+      "step": 170
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.9856100509155094e-05,
+      "loss": 0.3362,
+      "step": 171
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.985430883912468e-05,
+      "loss": 0.3751,
+      "step": 172
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.9852506111242775e-05,
+      "loss": 0.4951,
+      "step": 173
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5.985069232617711e-05,
+      "loss": 0.3329,
+      "step": 174
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.9848867484599494e-05,
+      "loss": 0.3924,
+      "step": 175
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.984703158718583e-05,
+      "loss": 0.7844,
+      "step": 176
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.984518463461613e-05,
+      "loss": 0.3063,
+      "step": 177
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.984332662757449e-05,
+      "loss": 0.3657,
+      "step": 178
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.984145756674912e-05,
+      "loss": 0.4005,
+      "step": 179
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.983957745283228e-05,
+      "loss": 0.3735,
+      "step": 180
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.983768628652037e-05,
+      "loss": 0.4974,
+      "step": 181
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.9835784068513876e-05,
+      "loss": 1.0969,
+      "step": 182
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.983387079951735e-05,
+      "loss": 0.2396,
+      "step": 183
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.983194648023946e-05,
+      "loss": 0.8985,
+      "step": 184
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.983001111139297e-05,
+      "loss": 0.6234,
+      "step": 185
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.9828064693694714e-05,
+      "loss": 0.2424,
+      "step": 186
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5.9826107227865645e-05,
+      "loss": 0.5502,
+      "step": 187
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.982413871463079e-05,
+      "loss": 0.483,
+      "step": 188
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.982215915471928e-05,
+      "loss": 0.3774,
+      "step": 189
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.9820168548864325e-05,
+      "loss": 0.4617,
+      "step": 190
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.981816689780324e-05,
+      "loss": 0.7004,
+      "step": 191
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.9816154202277417e-05,
+      "loss": 0.8072,
+      "step": 192
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.9814130463032345e-05,
+      "loss": 0.4487,
+      "step": 193
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.9812095680817615e-05,
+      "loss": 1.002,
+      "step": 194
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.9810049856386894e-05,
+      "loss": 0.7113,
+      "step": 195
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.980799299049794e-05,
+      "loss": 0.2873,
+      "step": 196
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.9805925083912604e-05,
+      "loss": 0.4975,
+      "step": 197
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.980384613739682e-05,
+      "loss": 0.3417,
+      "step": 198
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.980175615172063e-05,
+      "loss": 1.0987,
+      "step": 199
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.979965512765814e-05,
+      "loss": 0.4536,
+      "step": 200
+    },
+    {
+      "epoch": 0.15,
+      "eval_loss": 1.111311912536621,
+      "eval_runtime": 156.5903,
+      "eval_samples_per_second": 0.338,
+      "eval_steps_per_second": 0.172,
+      "step": 200
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.979754306598756e-05,
+      "loss": 0.384,
+      "step": 201
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.979541996749119e-05,
+      "loss": 0.2804,
+      "step": 202
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.97932858329554e-05,
+      "loss": 0.3413,
+      "step": 203
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.979114066317068e-05,
+      "loss": 0.4961,
+      "step": 204
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.978898445893157e-05,
+      "loss": 0.7419,
+      "step": 205
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.978681722103672e-05,
+      "loss": 0.3477,
+      "step": 206
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.9784638950288866e-05,
+      "loss": 0.217,
+      "step": 207
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.978244964749481e-05,
+      "loss": 0.7174,
+      "step": 208
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.9780249313465475e-05,
+      "loss": 0.533,
+      "step": 209
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.977803794901585e-05,
+      "loss": 0.1902,
+      "step": 210
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.9775815554965e-05,
+      "loss": 0.3303,
+      "step": 211
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.9773582132136087e-05,
+      "loss": 0.7741,
+      "step": 212
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.977133768135637e-05,
+      "loss": 0.8236,
+      "step": 213
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.9769082203457166e-05,
+      "loss": 0.6798,
+      "step": 214
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.97668156992739e-05,
+      "loss": 0.2507,
+      "step": 215
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.976453816964606e-05,
+      "loss": 0.3227,
+      "step": 216
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.9762249615417245e-05,
+      "loss": 0.5356,
+      "step": 217
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.97599500374351e-05,
+      "loss": 0.9196,
+      "step": 218
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.975763943655138e-05,
+      "loss": 0.7568,
+      "step": 219
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.975531781362194e-05,
+      "loss": 0.5231,
+      "step": 220
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.975298516950667e-05,
+      "loss": 0.391,
+      "step": 221
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.9750641505069574e-05,
+      "loss": 0.6852,
+      "step": 222
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.974828682117874e-05,
+      "loss": 0.4595,
+      "step": 223
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.9745921118706305e-05,
+      "loss": 0.1695,
+      "step": 224
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.974354439852854e-05,
+      "loss": 0.2466,
+      "step": 225
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.974115666152575e-05,
+      "loss": 0.848,
+      "step": 226
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.973875790858234e-05,
+      "loss": 0.3713,
+      "step": 227
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.973634814058679e-05,
+      "loss": 0.2327,
+      "step": 228
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.973392735843167e-05,
+      "loss": 0.3485,
+      "step": 229
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.973149556301362e-05,
+      "loss": 0.558,
+      "step": 230
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.9729052755233365e-05,
+      "loss": 0.2526,
+      "step": 231
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.97265989359957e-05,
+      "loss": 0.3679,
+      "step": 232
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.97241341062095e-05,
+      "loss": 0.4941,
+      "step": 233
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.9721658266787744e-05,
+      "loss": 0.5728,
+      "step": 234
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.971917141864745e-05,
+      "loss": 0.55,
+      "step": 235
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.971667356270973e-05,
+      "loss": 0.3739,
+      "step": 236
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.9714164699899794e-05,
+      "loss": 0.3676,
+      "step": 237
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.971164483114688e-05,
+      "loss": 0.3829,
+      "step": 238
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.9709113957384355e-05,
+      "loss": 0.2117,
+      "step": 239
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.9706572079549626e-05,
+      "loss": 0.6316,
+      "step": 240
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.97040191985842e-05,
+      "loss": 0.5837,
+      "step": 241
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.970145531543363e-05,
+      "loss": 0.5313,
+      "step": 242
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.969888043104758e-05,
+      "loss": 0.3265,
+      "step": 243
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.969629454637978e-05,
+      "loss": 0.5512,
+      "step": 244
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.969369766238799e-05,
+      "loss": 0.5799,
+      "step": 245
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.969108978003411e-05,
+      "loss": 0.3514,
+      "step": 246
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.9688470900284066e-05,
+      "loss": 0.294,
+      "step": 247
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.96858410241079e-05,
+      "loss": 0.4663,
+      "step": 248
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.968320015247967e-05,
+      "loss": 0.3003,
+      "step": 249
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.968054828637756e-05,
+      "loss": 0.9089,
+      "step": 250
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.9677885426783804e-05,
+      "loss": 0.5817,
+      "step": 251
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5.967521157468469e-05,
+      "loss": 0.7024,
+      "step": 252
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.9672526731070626e-05,
+      "loss": 0.6958,
+      "step": 253
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.966983089693604e-05,
+      "loss": 0.5176,
+      "step": 254
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.966712407327946e-05,
+      "loss": 0.2481,
+      "step": 255
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.966440626110346e-05,
+      "loss": 0.6958,
+      "step": 256
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.966167746141473e-05,
+      "loss": 0.3109,
+      "step": 257
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.9658937675223976e-05,
+      "loss": 0.5295,
+      "step": 258
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.965618690354601e-05,
+      "loss": 0.3527,
+      "step": 259
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.965342514739971e-05,
+      "loss": 0.2053,
+      "step": 260
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.9650652407807994e-05,
+      "loss": 0.2671,
+      "step": 261
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.964786868579787e-05,
+      "loss": 0.2447,
+      "step": 262
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.9645073982400415e-05,
+      "loss": 0.3249,
+      "step": 263
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.9642268298650775e-05,
+      "loss": 0.5172,
+      "step": 264
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5.963945163558816e-05,
+      "loss": 0.7846,
+      "step": 265
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.963662399425584e-05,
+      "loss": 0.4443,
+      "step": 266
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.963378537570115e-05,
+      "loss": 0.3216,
+      "step": 267
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.96309357809755e-05,
+      "loss": 0.288,
+      "step": 268
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.9628075211134364e-05,
+      "loss": 0.2127,
+      "step": 269
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.962520366723728e-05,
+      "loss": 0.6088,
+      "step": 270
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.962232115034786e-05,
+      "loss": 0.7053,
+      "step": 271
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.961942766153375e-05,
+      "loss": 0.3086,
+      "step": 272
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.961652320186669e-05,
+      "loss": 0.2379,
+      "step": 273
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.961360777242248e-05,
+      "loss": 0.8351,
+      "step": 274
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.961068137428096e-05,
+      "loss": 0.3786,
+      "step": 275
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.960774400852607e-05,
+      "loss": 0.4537,
+      "step": 276
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.960479567624579e-05,
+      "loss": 0.6303,
+      "step": 277
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5.960183637853215e-05,
+      "loss": 0.2535,
+      "step": 278
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.9598866116481264e-05,
+      "loss": 0.7446,
+      "step": 279
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.9595884891193294e-05,
+      "loss": 0.2347,
+      "step": 280
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.959289270377247e-05,
+      "loss": 0.6552,
+      "step": 281
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.958988955532708e-05,
+      "loss": 0.4865,
+      "step": 282
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.958687544696948e-05,
+      "loss": 0.4608,
+      "step": 283
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.9583850379816066e-05,
+      "loss": 0.2945,
+      "step": 284
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.95808143549873e-05,
+      "loss": 0.5575,
+      "step": 285
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.957776737360773e-05,
+      "loss": 0.6192,
+      "step": 286
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.9574709436805914e-05,
+      "loss": 0.4336,
+      "step": 287
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.957164054571451e-05,
+      "loss": 0.7521,
+      "step": 288
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.9568560701470194e-05,
+      "loss": 0.2066,
+      "step": 289
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.956546990521375e-05,
+      "loss": 0.3903,
+      "step": 290
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5.956236815808997e-05,
+      "loss": 0.6971,
+      "step": 291
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.9559255461247733e-05,
+      "loss": 0.2922,
+      "step": 292
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.9556131815839945e-05,
+      "loss": 0.6384,
+      "step": 293
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.955299722302361e-05,
+      "loss": 0.5984,
+      "step": 294
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.954985168395975e-05,
+      "loss": 0.7698,
+      "step": 295
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.954669519981344e-05,
+      "loss": 0.2003,
+      "step": 296
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.954352777175385e-05,
+      "loss": 0.556,
+      "step": 297
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.954034940095414e-05,
+      "loss": 0.3751,
+      "step": 298
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.95371600885916e-05,
+      "loss": 0.3835,
+      "step": 299
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.95339598358475e-05,
+      "loss": 0.515,
+      "step": 300
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.953074864390722e-05,
+      "loss": 0.5509,
+      "step": 301
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.952752651396013e-05,
+      "loss": 0.2719,
+      "step": 302
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5.9524293447199714e-05,
+      "loss": 0.3032,
+      "step": 303
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.952104944482347e-05,
+      "loss": 0.5101,
+      "step": 304
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.951779450803297e-05,
+      "loss": 0.2592,
+      "step": 305
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.9514528638033806e-05,
+      "loss": 0.6704,
+      "step": 306
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.951125183603565e-05,
+      "loss": 0.2335,
+      "step": 307
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.9507964103252186e-05,
+      "loss": 0.3625,
+      "step": 308
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.95046654409012e-05,
+      "loss": 0.4951,
+      "step": 309
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.950135585020447e-05,
+      "loss": 0.2304,
+      "step": 310
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.949803533238787e-05,
+      "loss": 0.7244,
+      "step": 311
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.949470388868129e-05,
+      "loss": 0.3666,
+      "step": 312
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.949136152031866e-05,
+      "loss": 0.3138,
+      "step": 313
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.9488008228538e-05,
+      "loss": 0.3038,
+      "step": 314
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.9484644014581344e-05,
+      "loss": 0.3123,
+      "step": 315
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5.9481268879694756e-05,
+      "loss": 0.6086,
+      "step": 316
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.947788282512838e-05,
+      "loss": 0.3519,
+      "step": 317
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.947448585213638e-05,
+      "loss": 0.444,
+      "step": 318
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.9471077961976996e-05,
+      "loss": 0.1887,
+      "step": 319
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.946765915591247e-05,
+      "loss": 0.2224,
+      "step": 320
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.94642294352091e-05,
+      "loss": 0.7657,
+      "step": 321
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.946078880113726e-05,
+      "loss": 0.362,
+      "step": 322
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.94573372549713e-05,
+      "loss": 0.3407,
+      "step": 323
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.94538747979897e-05,
+      "loss": 0.5699,
+      "step": 324
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.9450401431474896e-05,
+      "loss": 0.185,
+      "step": 325
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.944691715671341e-05,
+      "loss": 0.7194,
+      "step": 326
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.9443421974995805e-05,
+      "loss": 0.823,
+      "step": 327
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.943991588761667e-05,
+      "loss": 0.8917,
+      "step": 328
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5.9436398895874636e-05,
+      "loss": 0.7739,
+      "step": 329
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.9432871001072376e-05,
+      "loss": 0.5325,
+      "step": 330
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.9429332204516604e-05,
+      "loss": 0.6008,
+      "step": 331
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.942578250751805e-05,
+      "loss": 0.6468,
+      "step": 332
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.942222191139153e-05,
+      "loss": 0.6676,
+      "step": 333
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.9418650417455855e-05,
+      "loss": 0.3127,
+      "step": 334
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.941506802703387e-05,
+      "loss": 0.3079,
+      "step": 335
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.941147474145248e-05,
+      "loss": 0.5253,
+      "step": 336
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.940787056204262e-05,
+      "loss": 0.3227,
+      "step": 337
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.940425549013926e-05,
+      "loss": 0.1869,
+      "step": 338
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.9400629527081384e-05,
+      "loss": 0.2156,
+      "step": 339
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.939699267421204e-05,
+      "loss": 0.4871,
+      "step": 340
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.939334493287828e-05,
+      "loss": 0.2868,
+      "step": 341
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5.938968630443123e-05,
+      "loss": 0.2815,
+      "step": 342
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.9386016790226e-05,
+      "loss": 0.2657,
+      "step": 343
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.9382336391621774e-05,
+      "loss": 0.7165,
+      "step": 344
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.937864510998173e-05,
+      "loss": 0.5917,
+      "step": 345
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.937494294667311e-05,
+      "loss": 0.6451,
+      "step": 346
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.937122990306717e-05,
+      "loss": 0.4287,
+      "step": 347
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.9367505980539195e-05,
+      "loss": 0.5873,
+      "step": 348
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.93637711804685e-05,
+      "loss": 0.4367,
+      "step": 349
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.936002550423843e-05,
+      "loss": 0.4147,
+      "step": 350
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.935626895323638e-05,
+      "loss": 0.3736,
+      "step": 351
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.9352501528853736e-05,
+      "loss": 0.239,
+      "step": 352
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.9348723232485925e-05,
+      "loss": 0.541,
+      "step": 353
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.934493406553242e-05,
+      "loss": 0.2149,
+      "step": 354
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5.93411340293967e-05,
+      "loss": 0.2656,
+      "step": 355
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.9337323125486266e-05,
+      "loss": 0.2338,
+      "step": 356
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.933350135521265e-05,
+      "loss": 0.2071,
+      "step": 357
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.932966871999144e-05,
+      "loss": 0.6232,
+      "step": 358
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.932582522124218e-05,
+      "loss": 0.3265,
+      "step": 359
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.932197086038851e-05,
+      "loss": 0.245,
+      "step": 360
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.931810563885806e-05,
+      "loss": 0.3178,
+      "step": 361
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.9314229558082465e-05,
+      "loss": 0.3427,
+      "step": 362
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.9310342619497416e-05,
+      "loss": 0.5139,
+      "step": 363
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.93064448245426e-05,
+      "loss": 0.3556,
+      "step": 364
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.930253617466175e-05,
+      "loss": 0.3496,
+      "step": 365
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.92986166713026e-05,
+      "loss": 0.3583,
+      "step": 366
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.929468631591691e-05,
+      "loss": 0.3747,
+      "step": 367
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5.929074510996045e-05,
+      "loss": 0.3768,
+      "step": 368
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.928679305489303e-05,
+      "loss": 0.2898,
+      "step": 369
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.9282830152178464e-05,
+      "loss": 0.328,
+      "step": 370
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.927885640328458e-05,
+      "loss": 0.4419,
+      "step": 371
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.9274871809683244e-05,
+      "loss": 0.3924,
+      "step": 372
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.927087637285031e-05,
+      "loss": 0.8747,
+      "step": 373
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.926687009426567e-05,
+      "loss": 0.8131,
+      "step": 374
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.926285297541323e-05,
+      "loss": 0.3068,
+      "step": 375
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.925882501778089e-05,
+      "loss": 0.9709,
+      "step": 376
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.9254786222860605e-05,
+      "loss": 0.6477,
+      "step": 377
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.9250736592148296e-05,
+      "loss": 0.9228,
+      "step": 378
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.924667612714393e-05,
+      "loss": 0.7027,
+      "step": 379
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.9242604829351485e-05,
+      "loss": 0.4993,
+      "step": 380
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5.923852270027894e-05,
+      "loss": 0.6005,
+      "step": 381
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.923442974143828e-05,
+      "loss": 0.8307,
+      "step": 382
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.9230325954345526e-05,
+      "loss": 0.3705,
+      "step": 383
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.922621134052069e-05,
+      "loss": 0.3479,
+      "step": 384
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.9222085901487797e-05,
+      "loss": 0.4053,
+      "step": 385
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.9217949638774896e-05,
+      "loss": 0.48,
+      "step": 386
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.9213802553914015e-05,
+      "loss": 0.4761,
+      "step": 387
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.920964464844122e-05,
+      "loss": 0.5046,
+      "step": 388
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.9205475923896565e-05,
+      "loss": 0.2626,
+      "step": 389
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.920129638182413e-05,
+      "loss": 0.4622,
+      "step": 390
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.919710602377199e-05,
+      "loss": 0.4898,
+      "step": 391
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.919290485129223e-05,
+      "loss": 0.4509,
+      "step": 392
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.918869286594092e-05,
+      "loss": 0.5007,
+      "step": 393
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5.918447006927817e-05,
+      "loss": 0.3339,
+      "step": 394
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.918023646286808e-05,
+      "loss": 0.7425,
+      "step": 395
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.9175992048278746e-05,
+      "loss": 0.1611,
+      "step": 396
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.917173682708226e-05,
+      "loss": 0.9956,
+      "step": 397
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.916747080085476e-05,
+      "loss": 0.6394,
+      "step": 398
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.916319397117633e-05,
+      "loss": 0.4559,
+      "step": 399
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.915890633963109e-05,
+      "loss": 0.4118,
+      "step": 400
+    },
+    {
+      "epoch": 0.31,
+      "eval_loss": 1.1165653467178345,
+      "eval_runtime": 156.5069,
+      "eval_samples_per_second": 0.339,
+      "eval_steps_per_second": 0.173,
+      "step": 400
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.915460790780715e-05,
+      "loss": 0.2245,
+      "step": 401
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.915029867729663e-05,
+      "loss": 0.4814,
+      "step": 402
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.9145978649695635e-05,
+      "loss": 0.2283,
+      "step": 403
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.914164782660428e-05,
+      "loss": 0.2145,
+      "step": 404
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.9137306209626665e-05,
+      "loss": 0.1971,
+      "step": 405
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.9132953800370914e-05,
+      "loss": 0.2965,
+      "step": 406
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5.912859060044912e-05,
+      "loss": 0.2937,
+      "step": 407
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.91242166114774e-05,
+      "loss": 0.369,
+      "step": 408
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.911983183507583e-05,
+      "loss": 0.3111,
+      "step": 409
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.911543627286853e-05,
+      "loss": 0.3552,
+      "step": 410
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.911102992648357e-05,
+      "loss": 0.3472,
+      "step": 411
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.9106612797553044e-05,
+      "loss": 0.4399,
+      "step": 412
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.910218488771301e-05,
+      "loss": 0.9773,
+      "step": 413
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.909774619860356e-05,
+      "loss": 0.3855,
+      "step": 414
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.9093296731868757e-05,
+      "loss": 0.9498,
+      "step": 415
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.908883648915664e-05,
+      "loss": 0.2086,
+      "step": 416
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.9084365472119276e-05,
+      "loss": 0.5354,
+      "step": 417
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.907988368241269e-05,
+      "loss": 0.5455,
+      "step": 418
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.90753911216969e-05,
+      "loss": 0.3773,
+      "step": 419
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5.907088779163594e-05,
+      "loss": 0.473,
+      "step": 420
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.90663736938978e-05,
+      "loss": 0.4269,
+      "step": 421
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.906184883015448e-05,
+      "loss": 0.7108,
+      "step": 422
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.905731320208197e-05,
+      "loss": 0.817,
+      "step": 423
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.9052766811360234e-05,
+      "loss": 0.4585,
+      "step": 424
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.904820965967322e-05,
+      "loss": 0.4155,
+      "step": 425
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.9043641748708885e-05,
+      "loss": 0.6945,
+      "step": 426
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.903906308015913e-05,
+      "loss": 0.5883,
+      "step": 427
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.903447365571989e-05,
+      "loss": 0.688,
+      "step": 428
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.9029873477091045e-05,
+      "loss": 0.545,
+      "step": 429
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.9025262545976474e-05,
+      "loss": 0.3543,
+      "step": 430
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.9020640864084044e-05,
+      "loss": 0.5038,
+      "step": 431
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.9016008433125596e-05,
+      "loss": 0.3806,
+      "step": 432
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5.9011365254816946e-05,
+      "loss": 0.385,
+      "step": 433
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.9006711330877905e-05,
+      "loss": 0.3369,
+      "step": 434
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.900204666303226e-05,
+      "loss": 0.4913,
+      "step": 435
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.899737125300776e-05,
+      "loss": 0.4832,
+      "step": 436
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.899268510253617e-05,
+      "loss": 0.7796,
+      "step": 437
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.89879882133532e-05,
+      "loss": 0.637,
+      "step": 438
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.898328058719855e-05,
+      "loss": 0.6869,
+      "step": 439
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.8978562225815895e-05,
+      "loss": 0.2504,
+      "step": 440
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.8973833130952884e-05,
+      "loss": 0.785,
+      "step": 441
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.896909330436115e-05,
+      "loss": 0.4622,
+      "step": 442
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.896434274779629e-05,
+      "loss": 0.2809,
+      "step": 443
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.895958146301789e-05,
+      "loss": 0.3632,
+      "step": 444
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.895480945178949e-05,
+      "loss": 0.199,
+      "step": 445
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5.895002671587861e-05,
+      "loss": 0.3994,
+      "step": 446
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.894523325705677e-05,
+      "loss": 0.3558,
+      "step": 447
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.894042907709941e-05,
+      "loss": 0.3103,
+      "step": 448
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.893561417778598e-05,
+      "loss": 0.5171,
+      "step": 449
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.89307885608999e-05,
+      "loss": 0.2363,
+      "step": 450
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.892595222822853e-05,
+      "loss": 0.4868,
+      "step": 451
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.892110518156322e-05,
+      "loss": 0.4602,
+      "step": 452
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.891624742269929e-05,
+      "loss": 0.288,
+      "step": 453
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.891137895343604e-05,
+      "loss": 0.4434,
+      "step": 454
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.8906499775576685e-05,
+      "loss": 0.41,
+      "step": 455
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.8901609890928475e-05,
+      "loss": 0.4501,
+      "step": 456
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.889670930130259e-05,
+      "loss": 1.1347,
+      "step": 457
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.889179800851416e-05,
+      "loss": 0.3077,
+      "step": 458
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5.88868760143823e-05,
+      "loss": 0.4934,
+      "step": 459
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.88819433207301e-05,
+      "loss": 0.2021,
+      "step": 460
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.887699992938459e-05,
+      "loss": 0.639,
+      "step": 461
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.887204584217678e-05,
+      "loss": 0.5605,
+      "step": 462
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.886708106094163e-05,
+      "loss": 0.2656,
+      "step": 463
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.886210558751805e-05,
+      "loss": 0.8034,
+      "step": 464
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.885711942374894e-05,
+      "loss": 0.7848,
+      "step": 465
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.885212257148115e-05,
+      "loss": 0.7135,
+      "step": 466
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.8847115032565465e-05,
+      "loss": 0.2705,
+      "step": 467
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.884209680885665e-05,
+      "loss": 0.709,
+      "step": 468
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.883706790221343e-05,
+      "loss": 0.3881,
+      "step": 469
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.883202831449849e-05,
+      "loss": 0.3763,
+      "step": 470
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.882697804757844e-05,
+      "loss": 0.6709,
+      "step": 471
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5.882191710332389e-05,
+      "loss": 0.2688,
+      "step": 472
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.8816845483609374e-05,
+      "loss": 0.5193,
+      "step": 473
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.881176319031338e-05,
+      "loss": 0.3027,
+      "step": 474
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.8806670225318365e-05,
+      "loss": 0.1892,
+      "step": 475
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.880156659051074e-05,
+      "loss": 0.5579,
+      "step": 476
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.879645228778084e-05,
+      "loss": 0.457,
+      "step": 477
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.879132731902299e-05,
+      "loss": 0.2225,
+      "step": 478
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.878619168613543e-05,
+      "loss": 0.4044,
+      "step": 479
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.878104539102039e-05,
+      "loss": 0.5952,
+      "step": 480
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.8775888435584004e-05,
+      "loss": 0.6356,
+      "step": 481
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.877072082173638e-05,
+      "loss": 0.4809,
+      "step": 482
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.876554255139159e-05,
+      "loss": 0.6026,
+      "step": 483
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.8760353626467606e-05,
+      "loss": 0.6455,
+      "step": 484
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5.875515404888639e-05,
+      "loss": 0.5657,
+      "step": 485
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.874994382057383e-05,
+      "loss": 0.4585,
+      "step": 486
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.8744722943459754e-05,
+      "loss": 1.2231,
+      "step": 487
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.8739491419477954e-05,
+      "loss": 0.5826,
+      "step": 488
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.8734249250566154e-05,
+      "loss": 0.5087,
+      "step": 489
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.872899643866601e-05,
+      "loss": 0.6351,
+      "step": 490
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.872373298572315e-05,
+      "loss": 0.621,
+      "step": 491
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.87184588936871e-05,
+      "loss": 0.5816,
+      "step": 492
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.8713174164511364e-05,
+      "loss": 0.5259,
+      "step": 493
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.870787880015338e-05,
+      "loss": 0.2769,
+      "step": 494
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.8702572802574505e-05,
+      "loss": 0.1949,
+      "step": 495
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.869725617374006e-05,
+      "loss": 0.6682,
+      "step": 496
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5.869192891561928e-05,
+      "loss": 0.2723,
+      "step": 497
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.868659103018536e-05,
+      "loss": 0.5218,
+      "step": 498
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.868124251941541e-05,
+      "loss": 0.6158,
+      "step": 499
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.8675883385290485e-05,
+      "loss": 0.6536,
+      "step": 500
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.867051362979558e-05,
+      "loss": 0.4388,
+      "step": 501
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.8665133254919626e-05,
+      "loss": 0.192,
+      "step": 502
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.865974226265547e-05,
+      "loss": 0.8114,
+      "step": 503
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.86543406549999e-05,
+      "loss": 0.6168,
+      "step": 504
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.8648928433953644e-05,
+      "loss": 0.246,
+      "step": 505
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.864350560152136e-05,
+      "loss": 0.5993,
+      "step": 506
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.863807215971163e-05,
+      "loss": 0.1766,
+      "step": 507
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.863262811053695e-05,
+      "loss": 0.8842,
+      "step": 508
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.862717345601378e-05,
+      "loss": 0.7748,
+      "step": 509
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5.862170819816249e-05,
+      "loss": 0.8039,
+      "step": 510
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.8616232339007363e-05,
+      "loss": 0.2538,
+      "step": 511
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.8610745880576645e-05,
+      "loss": 0.3822,
+      "step": 512
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.860524882490247e-05,
+      "loss": 0.4078,
+      "step": 513
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.859974117402092e-05,
+      "loss": 0.6032,
+      "step": 514
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.859422292997198e-05,
+      "loss": 0.7461,
+      "step": 515
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.858869409479961e-05,
+      "loss": 0.2176,
+      "step": 516
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.8583154670551626e-05,
+      "loss": 0.2107,
+      "step": 517
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.85776046592798e-05,
+      "loss": 0.5857,
+      "step": 518
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.8572044063039825e-05,
+      "loss": 0.3525,
+      "step": 519
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.8566472883891314e-05,
+      "loss": 0.3725,
+      "step": 520
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.856089112389779e-05,
+      "loss": 0.3516,
+      "step": 521
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.855529878512672e-05,
+      "loss": 0.6566,
+      "step": 522
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5.8549695869649444e-05,
+      "loss": 0.5468,
+      "step": 523
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.854408237954128e-05,
+      "loss": 0.2485,
+      "step": 524
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.85384583168814e-05,
+      "loss": 0.299,
+      "step": 525
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.853282368375294e-05,
+      "loss": 0.5085,
+      "step": 526
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.852717848224293e-05,
+      "loss": 0.3165,
+      "step": 527
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.852152271444231e-05,
+      "loss": 0.6489,
+      "step": 528
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.851585638244595e-05,
+      "loss": 0.3654,
+      "step": 529
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.851017948835263e-05,
+      "loss": 0.5046,
+      "step": 530
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.850449203426502e-05,
+      "loss": 0.3597,
+      "step": 531
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.8498794022289726e-05,
+      "loss": 0.3633,
+      "step": 532
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.849308545453726e-05,
+      "loss": 0.4473,
+      "step": 533
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.848736633312203e-05,
+      "loss": 0.6577,
+      "step": 534
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.848163666016238e-05,
+      "loss": 0.2331,
+      "step": 535
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5.847589643778053e-05,
+      "loss": 0.9383,
+      "step": 536
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.847014566810263e-05,
+      "loss": 0.3922,
+      "step": 537
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.846438435325872e-05,
+      "loss": 0.3123,
+      "step": 538
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.845861249538277e-05,
+      "loss": 0.3416,
+      "step": 539
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.845283009661264e-05,
+      "loss": 0.6848,
+      "step": 540
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.844703715909008e-05,
+      "loss": 0.4837,
+      "step": 541
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.844123368496077e-05,
+      "loss": 0.4882,
+      "step": 542
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.843541967637427e-05,
+      "loss": 0.371,
+      "step": 543
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.842959513548407e-05,
+      "loss": 0.3201,
+      "step": 544
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.842376006444754e-05,
+      "loss": 0.3197,
+      "step": 545
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.8417914465425936e-05,
+      "loss": 0.9504,
+      "step": 546
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.841205834058445e-05,
+      "loss": 0.5558,
+      "step": 547
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.840619169209215e-05,
+      "loss": 0.759,
+      "step": 548
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5.840031452212201e-05,
+      "loss": 0.6709,
+      "step": 549
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.839442683285088e-05,
+      "loss": 0.8568,
+      "step": 550
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.8388528626459545e-05,
+      "loss": 0.2504,
+      "step": 551
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.8382619905132656e-05,
+      "loss": 0.2502,
+      "step": 552
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.837670067105876e-05,
+      "loss": 0.2898,
+      "step": 553
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.837077092643032e-05,
+      "loss": 0.4203,
+      "step": 554
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.8364830673443664e-05,
+      "loss": 0.7428,
+      "step": 555
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.835887991429903e-05,
+      "loss": 0.6355,
+      "step": 556
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.835291865120053e-05,
+      "loss": 0.4136,
+      "step": 557
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.834694688635621e-05,
+      "loss": 0.3402,
+      "step": 558
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.8340964621977944e-05,
+      "loss": 0.3439,
+      "step": 559
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.8334971860281536e-05,
+      "loss": 0.4363,
+      "step": 560
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.832896860348667e-05,
+      "loss": 0.754,
+      "step": 561
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5.832295485381692e-05,
+      "loss": 0.3767,
+      "step": 562
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.831693061349973e-05,
+      "loss": 0.3354,
+      "step": 563
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.831089588476645e-05,
+      "loss": 0.2563,
+      "step": 564
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.8304850669852304e-05,
+      "loss": 0.5139,
+      "step": 565
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.8298794970996404e-05,
+      "loss": 0.4523,
+      "step": 566
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.829272879044173e-05,
+      "loss": 0.48,
+      "step": 567
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.8286652130435174e-05,
+      "loss": 0.642,
+      "step": 568
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.8280564993227494e-05,
+      "loss": 0.6824,
+      "step": 569
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.827446738107332e-05,
+      "loss": 0.3662,
+      "step": 570
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.826835929623117e-05,
+      "loss": 0.7231,
+      "step": 571
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.8262240740963445e-05,
+      "loss": 1.0368,
+      "step": 572
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.825611171753642e-05,
+      "loss": 0.3476,
+      "step": 573
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.824997222822024e-05,
+      "loss": 0.2291,
+      "step": 574
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5.824382227528894e-05,
+      "loss": 1.0323,
+      "step": 575
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.823766186102043e-05,
+      "loss": 0.2005,
+      "step": 576
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.8231490987696476e-05,
+      "loss": 0.5418,
+      "step": 577
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.8225309657602734e-05,
+      "loss": 0.1827,
+      "step": 578
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.821911787302874e-05,
+      "loss": 0.2132,
+      "step": 579
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.82129156362679e-05,
+      "loss": 0.3962,
+      "step": 580
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.820670294961746e-05,
+      "loss": 0.4619,
+      "step": 581
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.8200479815378576e-05,
+      "loss": 0.2105,
+      "step": 582
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.8194246235856255e-05,
+      "loss": 0.7321,
+      "step": 583
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.818800221335937e-05,
+      "loss": 1.0181,
+      "step": 584
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.818174775020068e-05,
+      "loss": 0.518,
+      "step": 585
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.8175482848696794e-05,
+      "loss": 0.2236,
+      "step": 586
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.816920751116819e-05,
+      "loss": 0.4044,
+      "step": 587
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5.816292173993923e-05,
+      "loss": 0.6897,
+      "step": 588
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.815662553733809e-05,
+      "loss": 0.3633,
+      "step": 589
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.8150318905696884e-05,
+      "loss": 0.4778,
+      "step": 590
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.8144001847351534e-05,
+      "loss": 0.5024,
+      "step": 591
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.813767436464183e-05,
+      "loss": 0.4672,
+      "step": 592
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.813133645991144e-05,
+      "loss": 0.594,
+      "step": 593
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.812498813550788e-05,
+      "loss": 0.5486,
+      "step": 594
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.811862939378254e-05,
+      "loss": 0.7463,
+      "step": 595
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.811226023709066e-05,
+      "loss": 0.9475,
+      "step": 596
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.8105880667791323e-05,
+      "loss": 0.647,
+      "step": 597
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.8099490688247495e-05,
+      "loss": 1.0189,
+      "step": 598
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.809309030082597e-05,
+      "loss": 0.186,
+      "step": 599
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.808667950789742e-05,
+      "loss": 0.3141,
+      "step": 600
+    },
+    {
+      "epoch": 0.46,
+      "eval_loss": 1.1207194328308105,
+      "eval_runtime": 156.8385,
+      "eval_samples_per_second": 0.338,
+      "eval_steps_per_second": 0.172,
+      "step": 600
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5.8080258311836364e-05,
+      "loss": 0.3218,
+      "step": 601
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.807382671502118e-05,
+      "loss": 0.5193,
+      "step": 602
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.806738471983408e-05,
+      "loss": 0.6695,
+      "step": 603
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.806093232866114e-05,
+      "loss": 0.1825,
+      "step": 604
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.80544695438923e-05,
+      "loss": 0.5766,
+      "step": 605
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.8047996367921323e-05,
+      "loss": 0.5528,
+      "step": 606
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.804151280314583e-05,
+      "loss": 0.4265,
+      "step": 607
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.8035018851967306e-05,
+      "loss": 0.675,
+      "step": 608
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.802851451679106e-05,
+      "loss": 0.3176,
+      "step": 609
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.8021999800026256e-05,
+      "loss": 0.3429,
+      "step": 610
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.8015474704085915e-05,
+      "loss": 0.373,
+      "step": 611
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.800893923138688e-05,
+      "loss": 0.2878,
+      "step": 612
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.8002393384349866e-05,
+      "loss": 0.8264,
+      "step": 613
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5.7995837165399404e-05,
+      "loss": 0.389,
+      "step": 614
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.798927057696388e-05,
+      "loss": 0.248,
+      "step": 615
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.7982693621475504e-05,
+      "loss": 0.8459,
+      "step": 616
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.797610630137036e-05,
+      "loss": 0.2335,
+      "step": 617
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.796950861908835e-05,
+      "loss": 0.8241,
+      "step": 618
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.7962900577073186e-05,
+      "loss": 0.4262,
+      "step": 619
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.795628217777248e-05,
+      "loss": 0.2651,
+      "step": 620
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.794965342363763e-05,
+      "loss": 0.5862,
+      "step": 621
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.794301431712389e-05,
+      "loss": 0.7428,
+      "step": 622
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.793636486069034e-05,
+      "loss": 0.3498,
+      "step": 623
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.7929705056799904e-05,
+      "loss": 0.2017,
+      "step": 624
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.792303490791932e-05,
+      "loss": 0.2475,
+      "step": 625
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.791635441651919e-05,
+      "loss": 0.4353,
+      "step": 626
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5.790966358507392e-05,
+      "loss": 0.2987,
+      "step": 627
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.790296241606173e-05,
+      "loss": 0.476,
+      "step": 628
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.789625091196472e-05,
+      "loss": 0.5607,
+      "step": 629
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.788952907526877e-05,
+      "loss": 0.4446,
+      "step": 630
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.788279690846362e-05,
+      "loss": 0.3187,
+      "step": 631
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.787605441404282e-05,
+      "loss": 0.2175,
+      "step": 632
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.786930159450375e-05,
+      "loss": 0.3142,
+      "step": 633
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.78625384523476e-05,
+      "loss": 0.6462,
+      "step": 634
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.785576499007941e-05,
+      "loss": 0.1823,
+      "step": 635
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.784898121020802e-05,
+      "loss": 0.3077,
+      "step": 636
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.784218711524611e-05,
+      "loss": 0.8487,
+      "step": 637
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.783538270771017e-05,
+      "loss": 0.2494,
+      "step": 638
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.7828567990120495e-05,
+      "loss": 0.2759,
+      "step": 639
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5.782174296500123e-05,
+      "loss": 0.7678,
+      "step": 640
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.781490763488031e-05,
+      "loss": 0.1836,
+      "step": 641
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.780806200228952e-05,
+      "loss": 0.6841,
+      "step": 642
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.7801206069764406e-05,
+      "loss": 0.6677,
+      "step": 643
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.7794339839844406e-05,
+      "loss": 0.4571,
+      "step": 644
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.778746331507269e-05,
+      "loss": 0.5667,
+      "step": 645
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.778057649799629e-05,
+      "loss": 0.3092,
+      "step": 646
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.7773679391166064e-05,
+      "loss": 0.2036,
+      "step": 647
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.7766771997136636e-05,
+      "loss": 0.8114,
+      "step": 648
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.775985431846647e-05,
+      "loss": 0.4029,
+      "step": 649
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.775292635771781e-05,
+      "loss": 0.4355,
+      "step": 650
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.7745988117456764e-05,
+      "loss": 0.1916,
+      "step": 651
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.7739039600253195e-05,
+      "loss": 0.6955,
+      "step": 652
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5.7732080808680794e-05,
+      "loss": 0.5293,
+      "step": 653
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.772511174531705e-05,
+      "loss": 0.4943,
+      "step": 654
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.7718132412743265e-05,
+      "loss": 1.0316,
+      "step": 655
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.7711142813544536e-05,
+      "loss": 0.3226,
+      "step": 656
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.7704142950309774e-05,
+      "loss": 0.474,
+      "step": 657
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.769713282563169e-05,
+      "loss": 0.3364,
+      "step": 658
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.769011244210678e-05,
+      "loss": 0.213,
+      "step": 659
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.768308180233534e-05,
+      "loss": 0.4218,
+      "step": 660
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.767604090892151e-05,
+      "loss": 0.5221,
+      "step": 661
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.766898976447316e-05,
+      "loss": 0.3371,
+      "step": 662
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.766192837160201e-05,
+      "loss": 0.3955,
+      "step": 663
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.7654856732923544e-05,
+      "loss": 0.2206,
+      "step": 664
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.764777485105706e-05,
+      "loss": 0.3485,
+      "step": 665
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5.7640682728625645e-05,
+      "loss": 0.7266,
+      "step": 666
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.763358036825617e-05,
+      "loss": 0.4428,
+      "step": 667
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.7626467772579304e-05,
+      "loss": 0.8318,
+      "step": 668
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.761934494422953e-05,
+      "loss": 0.2175,
+      "step": 669
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.7612211885845075e-05,
+      "loss": 0.5883,
+      "step": 670
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.760506860006799e-05,
+      "loss": 0.7739,
+      "step": 671
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.759791508954411e-05,
+      "loss": 0.3693,
+      "step": 672
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.7590751356923025e-05,
+      "loss": 0.1598,
+      "step": 673
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.758357740485817e-05,
+      "loss": 0.6074,
+      "step": 674
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.757639323600672e-05,
+      "loss": 0.3963,
+      "step": 675
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.756919885302964e-05,
+      "loss": 0.4044,
+      "step": 676
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.7561994258591684e-05,
+      "loss": 0.5734,
+      "step": 677
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.7554779455361405e-05,
+      "loss": 0.5695,
+      "step": 678
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5.7547554446011106e-05,
+      "loss": 0.421,
+      "step": 679
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.7540319233216896e-05,
+      "loss": 0.4516,
+      "step": 680
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.7533073819658644e-05,
+      "loss": 0.5611,
+      "step": 681
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.7525818208020016e-05,
+      "loss": 0.2429,
+      "step": 682
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.751855240098844e-05,
+      "loss": 1.0444,
+      "step": 683
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.751127640125513e-05,
+      "loss": 0.4786,
+      "step": 684
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.750399021151506e-05,
+      "loss": 0.4834,
+      "step": 685
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.7496693834467e-05,
+      "loss": 0.6735,
+      "step": 686
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.7489387272813486e-05,
+      "loss": 0.3527,
+      "step": 687
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.748207052926082e-05,
+      "loss": 0.4532,
+      "step": 688
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.7474743606519065e-05,
+      "loss": 0.2601,
+      "step": 689
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.7467406507302094e-05,
+      "loss": 0.5995,
+      "step": 690
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.74600592343275e-05,
+      "loss": 0.5241,
+      "step": 691
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5.7452701790316674e-05,
+      "loss": 0.2302,
+      "step": 692
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.7445334177994775e-05,
+      "loss": 0.4453,
+      "step": 693
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.7437956400090715e-05,
+      "loss": 0.2946,
+      "step": 694
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.743056845933718e-05,
+      "loss": 0.3186,
+      "step": 695
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.742317035847061e-05,
+      "loss": 0.8952,
+      "step": 696
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.741576210023123e-05,
+      "loss": 0.363,
+      "step": 697
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.7408343687363e-05,
+      "loss": 0.6303,
+      "step": 698
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.7400915122613666e-05,
+      "loss": 0.5031,
+      "step": 699
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.7393476408734705e-05,
+      "loss": 0.5852,
+      "step": 700
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.7386027548481395e-05,
+      "loss": 0.2539,
+      "step": 701
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.737856854461272e-05,
+      "loss": 0.7116,
+      "step": 702
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.737109939989147e-05,
+      "loss": 0.3007,
+      "step": 703
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5.736362011708417e-05,
+      "loss": 0.2393,
+      "step": 704
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.735613069896108e-05,
+      "loss": 0.6819,
+      "step": 705
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.734863114829625e-05,
+      "loss": 0.5014,
+      "step": 706
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.734112146786748e-05,
+      "loss": 0.4315,
+      "step": 707
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.733360166045628e-05,
+      "loss": 0.5662,
+      "step": 708
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.732607172884796e-05,
+      "loss": 0.3189,
+      "step": 709
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.731853167583155e-05,
+      "loss": 0.5436,
+      "step": 710
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.731098150419984e-05,
+      "loss": 0.4548,
+      "step": 711
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.730342121674938e-05,
+      "loss": 0.3814,
+      "step": 712
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.729585081628044e-05,
+      "loss": 0.4693,
+      "step": 713
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.728827030559705e-05,
+      "loss": 0.5534,
+      "step": 714
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.7280679687506995e-05,
+      "loss": 0.2038,
+      "step": 715
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.7273078964821775e-05,
+      "loss": 0.4317,
+      "step": 716
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.726546814035667e-05,
+      "loss": 0.3252,
+      "step": 717
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.725784721693067e-05,
+      "loss": 0.5484,
+      "step": 718
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.725021619736653e-05,
+      "loss": 0.4287,
+      "step": 719
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.7242575084490715e-05,
+      "loss": 0.832,
+      "step": 720
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.7234923881133456e-05,
+      "loss": 0.2756,
+      "step": 721
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.722726259012872e-05,
+      "loss": 0.3473,
+      "step": 722
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.721959121431418e-05,
+      "loss": 0.7344,
+      "step": 723
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.721190975653128e-05,
+      "loss": 0.7256,
+      "step": 724
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.7204218219625194e-05,
+      "loss": 0.3204,
+      "step": 725
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.71965166064448e-05,
+      "loss": 0.7499,
+      "step": 726
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.718880491984274e-05,
+      "loss": 0.5772,
+      "step": 727
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.718108316267537e-05,
+      "loss": 0.2242,
+      "step": 728
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.717335133780279e-05,
+      "loss": 0.4745,
+      "step": 729
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.716560944808881e-05,
+      "loss": 0.5915,
+      "step": 730
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.715785749640098e-05,
+      "loss": 0.3593,
+      "step": 731
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.7150095485610575e-05,
+      "loss": 0.2702,
+      "step": 732
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.7142323418592596e-05,
+      "loss": 0.2364,
+      "step": 733
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.713454129822577e-05,
+      "loss": 0.7629,
+      "step": 734
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.712674912739255e-05,
+      "loss": 0.7182,
+      "step": 735
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.711894690897911e-05,
+      "loss": 0.1881,
+      "step": 736
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.7111134645875323e-05,
+      "loss": 0.3046,
+      "step": 737
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.7103312340974826e-05,
+      "loss": 0.1885,
+      "step": 738
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.7095479997174936e-05,
+      "loss": 0.5289,
+      "step": 739
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.708763761737672e-05,
+      "loss": 0.5108,
+      "step": 740
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.707978520448494e-05,
+      "loss": 0.6085,
+      "step": 741
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.707192276140808e-05,
+      "loss": 0.9777,
+      "step": 742
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.706405029105833e-05,
+      "loss": 0.9324,
+      "step": 743
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.705616779635163e-05,
+      "loss": 0.7302,
+      "step": 744
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.70482752802076e-05,
+      "loss": 0.416,
+      "step": 745
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.704037274554955e-05,
+      "loss": 0.4177,
+      "step": 746
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.7032460195304565e-05,
+      "loss": 0.419,
+      "step": 747
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.7024537632403395e-05,
+      "loss": 0.6588,
+      "step": 748
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.7016605059780495e-05,
+      "loss": 0.6509,
+      "step": 749
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.700866248037406e-05,
+      "loss": 0.2454,
+      "step": 750
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.7000709897125957e-05,
+      "loss": 0.2463,
+      "step": 751
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.699274731298177e-05,
+      "loss": 0.3174,
+      "step": 752
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.698477473089082e-05,
+      "loss": 0.5873,
+      "step": 753
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.697679215380608e-05,
+      "loss": 0.237,
+      "step": 754
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.696879958468424e-05,
+      "loss": 0.5439,
+      "step": 755
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.696079702648572e-05,
+      "loss": 0.5049,
+      "step": 756
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.69527844821746e-05,
+      "loss": 0.6739,
+      "step": 757
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.694476195471868e-05,
+      "loss": 0.2333,
+      "step": 758
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.693672944708946e-05,
+      "loss": 0.4269,
+      "step": 759
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.6928686962262135e-05,
+      "loss": 0.203,
+      "step": 760
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.6920634503215584e-05,
+      "loss": 0.2691,
+      "step": 761
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.691257207293239e-05,
+      "loss": 0.6638,
+      "step": 762
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.6904499674398824e-05,
+      "loss": 0.3526,
+      "step": 763
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.6896417310604853e-05,
+      "loss": 0.62,
+      "step": 764
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.6888324984544146e-05,
+      "loss": 0.7137,
+      "step": 765
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.688022269921404e-05,
+      "loss": 0.7112,
+      "step": 766
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.6872110457615575e-05,
+      "loss": 0.188,
+      "step": 767
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.6863988262753466e-05,
+      "loss": 0.3354,
+      "step": 768
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.685585611763614e-05,
+      "loss": 0.2078,
+      "step": 769
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.684771402527567e-05,
+      "loss": 0.3956,
+      "step": 770
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.683956198868786e-05,
+      "loss": 0.6851,
+      "step": 771
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.6831400010892165e-05,
+      "loss": 0.7625,
+      "step": 772
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.682322809491174e-05,
+      "loss": 0.4788,
+      "step": 773
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.6815046243773395e-05,
+      "loss": 0.3455,
+      "step": 774
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.680685446050765e-05,
+      "loss": 0.5376,
+      "step": 775
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.679865274814868e-05,
+      "loss": 0.5687,
+      "step": 776
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.679044110973436e-05,
+      "loss": 0.52,
+      "step": 777
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.6782219548306225e-05,
+      "loss": 0.1722,
+      "step": 778
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.6773988066909493e-05,
+      "loss": 0.3559,
+      "step": 779
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.6765746668593046e-05,
+      "loss": 1.656,
+      "step": 780
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.6757495356409455e-05,
+      "loss": 0.8508,
+      "step": 781
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.6749234133414955e-05,
+      "loss": 0.3733,
+      "step": 782
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.674096300266945e-05,
+      "loss": 0.2834,
+      "step": 783
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.673268196723651e-05,
+      "loss": 0.815,
+      "step": 784
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.6724391030183386e-05,
+      "loss": 0.6766,
+      "step": 785
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.6716090194580995e-05,
+      "loss": 0.3289,
+      "step": 786
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.67077794635039e-05,
+      "loss": 0.2327,
+      "step": 787
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.6699458840030356e-05,
+      "loss": 0.496,
+      "step": 788
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.6691128327242275e-05,
+      "loss": 0.2559,
+      "step": 789
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.668278792822521e-05,
+      "loss": 0.4663,
+      "step": 790
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.6674437646068406e-05,
+      "loss": 0.342,
+      "step": 791
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.666607748386476e-05,
+      "loss": 0.7126,
+      "step": 792
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.665770744471081e-05,
+      "loss": 0.728,
+      "step": 793
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.664932753170677e-05,
+      "loss": 0.4122,
+      "step": 794
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.664093774795652e-05,
+      "loss": 0.7519,
+      "step": 795
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.6632538096567585e-05,
+      "loss": 0.5428,
+      "step": 796
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.662412858065113e-05,
+      "loss": 0.2798,
+      "step": 797
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.6615709203321984e-05,
+      "loss": 0.6143,
+      "step": 798
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.6607279967698644e-05,
+      "loss": 0.5153,
+      "step": 799
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.659884087690325e-05,
+      "loss": 0.7105,
+      "step": 800
+    },
+    {
+      "epoch": 0.62,
+      "eval_loss": 1.1248592138290405,
+      "eval_runtime": 157.1222,
+      "eval_samples_per_second": 0.337,
+      "eval_steps_per_second": 0.172,
+      "step": 800
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.659039193406158e-05,
+      "loss": 0.1819,
+      "step": 801
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.658193314230307e-05,
+      "loss": 0.196,
+      "step": 802
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.6573464504760816e-05,
+      "loss": 0.3143,
+      "step": 803
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.656498602457154e-05,
+      "loss": 0.7426,
+      "step": 804
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.6556497704875616e-05,
+      "loss": 0.2201,
+      "step": 805
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.654799954881707e-05,
+      "loss": 0.3122,
+      "step": 806
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.6539491559543567e-05,
+      "loss": 0.7059,
+      "step": 807
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.65309737402064e-05,
+      "loss": 0.5913,
+      "step": 808
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.6522446093960544e-05,
+      "loss": 0.5148,
+      "step": 809
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.651390862396456e-05,
+      "loss": 0.282,
+      "step": 810
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.650536133338069e-05,
+      "loss": 0.2132,
+      "step": 811
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.649680422537479e-05,
+      "loss": 0.6906,
+      "step": 812
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.6488237303116365e-05,
+      "loss": 0.3331,
+      "step": 813
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.647966056977854e-05,
+      "loss": 0.6423,
+      "step": 814
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.647107402853811e-05,
+      "loss": 0.3105,
+      "step": 815
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.646247768257544e-05,
+      "loss": 0.3029,
+      "step": 816
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.645387153507459e-05,
+      "loss": 0.7272,
+      "step": 817
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.644525558922321e-05,
+      "loss": 0.5645,
+      "step": 818
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.643662984821261e-05,
+      "loss": 0.4409,
+      "step": 819
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.64279943152377e-05,
+      "loss": 0.754,
+      "step": 820
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.6419348993497014e-05,
+      "loss": 0.7385,
+      "step": 821
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.6410693886192743e-05,
+      "loss": 0.3239,
+      "step": 822
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.640202899653068e-05,
+      "loss": 0.6246,
+      "step": 823
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.6393354327720255e-05,
+      "loss": 0.6375,
+      "step": 824
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.6384669882974504e-05,
+      "loss": 0.6467,
+      "step": 825
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.6375975665510096e-05,
+      "loss": 0.4818,
+      "step": 826
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.6367271678547305e-05,
+      "loss": 0.3294,
+      "step": 827
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.635855792531005e-05,
+      "loss": 0.4038,
+      "step": 828
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.634983440902584e-05,
+      "loss": 0.6227,
+      "step": 829
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.6341101132925806e-05,
+      "loss": 0.753,
+      "step": 830
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.633235810024472e-05,
+      "loss": 0.4929,
+      "step": 831
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.632360531422093e-05,
+      "loss": 0.3151,
+      "step": 832
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.631484277809641e-05,
+      "loss": 0.6028,
+      "step": 833
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.630607049511677e-05,
+      "loss": 0.2165,
+      "step": 834
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.6297288468531194e-05,
+      "loss": 0.5926,
+      "step": 835
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.6288496701592495e-05,
+      "loss": 0.4995,
+      "step": 836
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.627969519755709e-05,
+      "loss": 0.6034,
+      "step": 837
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.6270883959684986e-05,
+      "loss": 0.5017,
+      "step": 838
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.626206299123984e-05,
+      "loss": 0.3112,
+      "step": 839
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.625323229548886e-05,
+      "loss": 0.1844,
+      "step": 840
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.6244391875702887e-05,
+      "loss": 0.4549,
+      "step": 841
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.623554173515637e-05,
+      "loss": 0.1993,
+      "step": 842
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.622668187712733e-05,
+      "loss": 0.3347,
+      "step": 843
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.621781230489742e-05,
+      "loss": 0.9106,
+      "step": 844
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.620893302175186e-05,
+      "loss": 0.5998,
+      "step": 845
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.620004403097948e-05,
+      "loss": 0.6431,
+      "step": 846
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.619114533587272e-05,
+      "loss": 0.2069,
+      "step": 847
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.618223693972761e-05,
+      "loss": 0.5434,
+      "step": 848
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.617331884584373e-05,
+      "loss": 0.7001,
+      "step": 849
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.616439105752432e-05,
+      "loss": 0.681,
+      "step": 850
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.6155453578076154e-05,
+      "loss": 0.5885,
+      "step": 851
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.614650641080964e-05,
+      "loss": 0.7848,
+      "step": 852
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.613754955903875e-05,
+      "loss": 0.8122,
+      "step": 853
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.612858302608103e-05,
+      "loss": 0.563,
+      "step": 854
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.6119606815257645e-05,
+      "loss": 0.7866,
+      "step": 855
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.611062092989332e-05,
+      "loss": 0.4124,
+      "step": 856
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.610162537331637e-05,
+      "loss": 0.451,
+      "step": 857
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.6092620148858695e-05,
+      "loss": 0.3673,
+      "step": 858
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.608360525985579e-05,
+      "loss": 0.4336,
+      "step": 859
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.607458070964669e-05,
+      "loss": 0.3004,
+      "step": 860
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.606554650157405e-05,
+      "loss": 0.3514,
+      "step": 861
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.605650263898407e-05,
+      "loss": 0.34,
+      "step": 862
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.6047449125226546e-05,
+      "loss": 0.3222,
+      "step": 863
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.6038385963654864e-05,
+      "loss": 0.5155,
+      "step": 864
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.602931315762593e-05,
+      "loss": 0.3596,
+      "step": 865
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.602023071050028e-05,
+      "loss": 0.2786,
+      "step": 866
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.6011138625641974e-05,
+      "loss": 0.553,
+      "step": 867
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.600203690641868e-05,
+      "loss": 0.3126,
+      "step": 868
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.599292555620163e-05,
+      "loss": 0.6199,
+      "step": 869
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.5983804578365584e-05,
+      "loss": 0.6395,
+      "step": 870
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.597467397628892e-05,
+      "loss": 1.1085,
+      "step": 871
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.596553375335353e-05,
+      "loss": 0.2408,
+      "step": 872
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.595638391294492e-05,
+      "loss": 0.6306,
+      "step": 873
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.5947224458452135e-05,
+      "loss": 0.4995,
+      "step": 874
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.593805539326776e-05,
+      "loss": 0.4387,
+      "step": 875
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.592887672078798e-05,
+      "loss": 0.9682,
+      "step": 876
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.59196884444125e-05,
+      "loss": 0.5595,
+      "step": 877
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.591049056754461e-05,
+      "loss": 0.4916,
+      "step": 878
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.5901283093591164e-05,
+      "loss": 0.5189,
+      "step": 879
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.589206602596252e-05,
+      "loss": 0.2147,
+      "step": 880
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.588283936807266e-05,
+      "loss": 1.3335,
+      "step": 881
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.5873603123339044e-05,
+      "loss": 0.4986,
+      "step": 882
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.586435729518274e-05,
+      "loss": 0.9419,
+      "step": 883
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.585510188702835e-05,
+      "loss": 0.6237,
+      "step": 884
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.584583690230401e-05,
+      "loss": 0.5348,
+      "step": 885
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.5836562344441404e-05,
+      "loss": 0.5186,
+      "step": 886
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.582727821687579e-05,
+      "loss": 0.7229,
+      "step": 887
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.581798452304595e-05,
+      "loss": 0.3559,
+      "step": 888
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.58086812663942e-05,
+      "loss": 0.6181,
+      "step": 889
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.579936845036642e-05,
+      "loss": 0.3933,
+      "step": 890
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.579004607841201e-05,
+      "loss": 0.8848,
+      "step": 891
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.578071415398391e-05,
+      "loss": 0.373,
+      "step": 892
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.5771372680538636e-05,
+      "loss": 0.3243,
+      "step": 893
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.576202166153619e-05,
+      "loss": 0.8299,
+      "step": 894
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.575266110044014e-05,
+      "loss": 0.4649,
+      "step": 895
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.5743291000717566e-05,
+      "loss": 0.5447,
+      "step": 896
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.5733911365839115e-05,
+      "loss": 0.5269,
+      "step": 897
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.572452219927893e-05,
+      "loss": 0.7487,
+      "step": 898
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.5715123504514725e-05,
+      "loss": 0.2757,
+      "step": 899
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.5705715285027676e-05,
+      "loss": 0.2849,
+      "step": 900
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.569629754430257e-05,
+      "loss": 0.4169,
+      "step": 901
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.568687028582765e-05,
+      "loss": 0.2645,
+      "step": 902
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.567743351309475e-05,
+      "loss": 0.3182,
+      "step": 903
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.566798722959916e-05,
+      "loss": 0.4609,
+      "step": 904
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.565853143883972e-05,
+      "loss": 0.8076,
+      "step": 905
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.564906614431883e-05,
+      "loss": 0.2309,
+      "step": 906
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.563959134954235e-05,
+      "loss": 0.2968,
+      "step": 907
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.56301070580197e-05,
+      "loss": 0.5129,
+      "step": 908
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.562061327326379e-05,
+      "loss": 0.6944,
+      "step": 909
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.5611109998791075e-05,
+      "loss": 0.8313,
+      "step": 910
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.560159723812149e-05,
+      "loss": 0.1838,
+      "step": 911
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.559207499477852e-05,
+      "loss": 0.5128,
+      "step": 912
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.558254327228914e-05,
+      "loss": 0.2337,
+      "step": 913
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.557300207418383e-05,
+      "loss": 0.3289,
+      "step": 914
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.556345140399661e-05,
+      "loss": 0.353,
+      "step": 915
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.555389126526497e-05,
+      "loss": 0.72,
+      "step": 916
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.554432166152995e-05,
+      "loss": 0.3736,
+      "step": 917
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.5534742596336034e-05,
+      "loss": 0.6092,
+      "step": 918
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.5525154073231286e-05,
+      "loss": 0.4441,
+      "step": 919
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.5515556095767226e-05,
+      "loss": 0.3566,
+      "step": 920
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.550594866749887e-05,
+      "loss": 0.2507,
+      "step": 921
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.549633179198477e-05,
+      "loss": 0.7282,
+      "step": 922
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.5486705472786956e-05,
+      "loss": 0.4852,
+      "step": 923
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.547706971347093e-05,
+      "loss": 0.389,
+      "step": 924
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.546742451760575e-05,
+      "loss": 0.6812,
+      "step": 925
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.545776988876392e-05,
+      "loss": 0.2356,
+      "step": 926
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.5448105830521465e-05,
+      "loss": 0.6137,
+      "step": 927
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.5438432346457884e-05,
+      "loss": 0.336,
+      "step": 928
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.542874944015617e-05,
+      "loss": 0.4186,
+      "step": 929
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.541905711520283e-05,
+      "loss": 0.1789,
+      "step": 930
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.540935537518783e-05,
+      "loss": 0.2447,
+      "step": 931
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.539964422370464e-05,
+      "loss": 0.2656,
+      "step": 932
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.538992366435019e-05,
+      "loss": 0.2064,
+      "step": 933
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.538019370072495e-05,
+      "loss": 1.098,
+      "step": 934
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.5370454336432815e-05,
+      "loss": 0.4448,
+      "step": 935
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.536070557508118e-05,
+      "loss": 0.5184,
+      "step": 936
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.535094742028095e-05,
+      "loss": 0.5614,
+      "step": 937
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.534117987564646e-05,
+      "loss": 0.3647,
+      "step": 938
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.5331402944795575e-05,
+      "loss": 0.4893,
+      "step": 939
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.532161663134957e-05,
+      "loss": 0.7035,
+      "step": 940
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.531182093893328e-05,
+      "loss": 0.5578,
+      "step": 941
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.530201587117494e-05,
+      "loss": 0.2917,
+      "step": 942
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.5292201431706295e-05,
+      "loss": 0.6427,
+      "step": 943
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.5282377624162555e-05,
+      "loss": 0.5051,
+      "step": 944
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.5272544452182396e-05,
+      "loss": 0.5343,
+      "step": 945
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.526270191940796e-05,
+      "loss": 0.4871,
+      "step": 946
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.525285002948486e-05,
+      "loss": 0.3707,
+      "step": 947
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.524298878606219e-05,
+      "loss": 0.4218,
+      "step": 948
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.523311819279247e-05,
+      "loss": 0.3268,
+      "step": 949
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.522323825333173e-05,
+      "loss": 0.3386,
+      "step": 950
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.5213348971339417e-05,
+      "loss": 0.3485,
+      "step": 951
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.520345035047848e-05,
+      "loss": 0.5056,
+      "step": 952
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.5193542394415296e-05,
+      "loss": 0.582,
+      "step": 953
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.5183625106819704e-05,
+      "loss": 0.233,
+      "step": 954
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.5173698491365023e-05,
+      "loss": 0.2507,
+      "step": 955
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.516376255172799e-05,
+      "loss": 0.2844,
+      "step": 956
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.515381729158884e-05,
+      "loss": 0.9841,
+      "step": 957
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.5143862714631216e-05,
+      "loss": 0.8945,
+      "step": 958
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.513389882454224e-05,
+      "loss": 0.4222,
+      "step": 959
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.512392562501248e-05,
+      "loss": 0.3932,
+      "step": 960
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.5113943119735925e-05,
+      "loss": 0.4782,
+      "step": 961
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.510395131241006e-05,
+      "loss": 0.4887,
+      "step": 962
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.509395020673578e-05,
+      "loss": 0.6278,
+      "step": 963
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.508393980641742e-05,
+      "loss": 1.0752,
+      "step": 964
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.50739201151628e-05,
+      "loss": 0.3463,
+      "step": 965
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.506389113668312e-05,
+      "loss": 0.374,
+      "step": 966
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.5053852874693066e-05,
+      "loss": 0.2468,
+      "step": 967
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.504380533291075e-05,
+      "loss": 0.2875,
+      "step": 968
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.5033748515057716e-05,
+      "loss": 0.4355,
+      "step": 969
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.502368242485895e-05,
+      "loss": 0.9306,
+      "step": 970
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.501360706604286e-05,
+      "loss": 0.2019,
+      "step": 971
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.500352244234131e-05,
+      "loss": 0.2049,
+      "step": 972
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.499342855748958e-05,
+      "loss": 0.7197,
+      "step": 973
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.498332541522637e-05,
+      "loss": 0.3147,
+      "step": 974
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.4973213019293825e-05,
+      "loss": 0.5956,
+      "step": 975
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.4963091373437525e-05,
+      "loss": 0.6852,
+      "step": 976
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.495296048140646e-05,
+      "loss": 0.7481,
+      "step": 977
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.4942820346953043e-05,
+      "loss": 0.1974,
+      "step": 978
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.493267097383314e-05,
+      "loss": 0.5694,
+      "step": 979
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.4922512365805985e-05,
+      "loss": 0.4697,
+      "step": 980
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.4912344526634285e-05,
+      "loss": 0.7006,
+      "step": 981
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.490216746008413e-05,
+      "loss": 0.7544,
+      "step": 982
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.489198116992505e-05,
+      "loss": 0.7314,
+      "step": 983
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.4881785659929996e-05,
+      "loss": 0.5805,
+      "step": 984
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.4871580933875295e-05,
+      "loss": 0.613,
+      "step": 985
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.486136699554073e-05,
+      "loss": 0.9065,
+      "step": 986
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.4851143848709495e-05,
+      "loss": 0.7141,
+      "step": 987
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.484091149716815e-05,
+      "loss": 0.4406,
+      "step": 988
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.483066994470672e-05,
+      "loss": 0.6628,
+      "step": 989
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.48204191951186e-05,
+      "loss": 0.1914,
+      "step": 990
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.48101592522006e-05,
+      "loss": 0.4072,
+      "step": 991
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.479989011975295e-05,
+      "loss": 0.7959,
+      "step": 992
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.4789611801579274e-05,
+      "loss": 0.7755,
+      "step": 993
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.47793243014866e-05,
+      "loss": 0.2837,
+      "step": 994
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.4769027623285336e-05,
+      "loss": 0.4734,
+      "step": 995
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.475872177078933e-05,
+      "loss": 0.6227,
+      "step": 996
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.474840674781578e-05,
+      "loss": 0.5472,
+      "step": 997
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.4738082558185335e-05,
+      "loss": 0.498,
+      "step": 998
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.4727749205722e-05,
+      "loss": 0.7127,
+      "step": 999
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.471740669425317e-05,
+      "loss": 0.6273,
+      "step": 1000
+    },
+    {
+      "epoch": 0.77,
+      "eval_loss": 1.1267361640930176,
+      "eval_runtime": 156.9648,
+      "eval_samples_per_second": 0.338,
+      "eval_steps_per_second": 0.172,
+      "step": 1000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.4707055027609676e-05,
+      "loss": 0.3553,
+      "step": 1001
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.469669420962569e-05,
+      "loss": 0.3959,
+      "step": 1002
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.46863242441388e-05,
+      "loss": 0.3849,
+      "step": 1003
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.467594513498998e-05,
+      "loss": 1.0125,
+      "step": 1004
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.466555688602358e-05,
+      "loss": 0.2577,
+      "step": 1005
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.4655159501087364e-05,
+      "loss": 0.9823,
+      "step": 1006
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.464475298403243e-05,
+      "loss": 0.5675,
+      "step": 1007
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.4634337338713304e-05,
+      "loss": 0.521,
+      "step": 1008
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.462391256898788e-05,
+      "loss": 0.428,
+      "step": 1009
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.4613478678717426e-05,
+      "loss": 0.7588,
+      "step": 1010
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.4603035671766585e-05,
+      "loss": 0.5969,
+      "step": 1011
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.4592583552003386e-05,
+      "loss": 0.6015,
+      "step": 1012
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.458212232329922e-05,
+      "loss": 0.522,
+      "step": 1013
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.457165198952887e-05,
+      "loss": 0.659,
+      "step": 1014
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.4561172554570494e-05,
+      "loss": 0.409,
+      "step": 1015
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.4550684022305585e-05,
+      "loss": 0.9058,
+      "step": 1016
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.454018639661905e-05,
+      "loss": 0.4236,
+      "step": 1017
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.452967968139915e-05,
+      "loss": 0.1759,
+      "step": 1018
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.451916388053748e-05,
+      "loss": 0.3614,
+      "step": 1019
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.450863899792905e-05,
+      "loss": 0.3195,
+      "step": 1020
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.449810503747221e-05,
+      "loss": 0.2455,
+      "step": 1021
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.448756200306866e-05,
+      "loss": 0.2875,
+      "step": 1022
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.447700989862349e-05,
+      "loss": 0.9915,
+      "step": 1023
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.446644872804512e-05,
+      "loss": 0.3996,
+      "step": 1024
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.445587849524536e-05,
+      "loss": 0.7417,
+      "step": 1025
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.4445299204139354e-05,
+      "loss": 0.5258,
+      "step": 1026
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.443471085864559e-05,
+      "loss": 0.3358,
+      "step": 1027
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.442411346268594e-05,
+      "loss": 0.2117,
+      "step": 1028
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.441350702018561e-05,
+      "loss": 0.1965,
+      "step": 1029
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.440289153507317e-05,
+      "loss": 0.6661,
+      "step": 1030
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.439226701128052e-05,
+      "loss": 0.3252,
+      "step": 1031
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.438163345274292e-05,
+      "loss": 0.5826,
+      "step": 1032
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.437099086339896e-05,
+      "loss": 0.2268,
+      "step": 1033
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.4360339247190616e-05,
+      "loss": 0.5723,
+      "step": 1034
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.4349678608063166e-05,
+      "loss": 0.289,
+      "step": 1035
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.4339008949965256e-05,
+      "loss": 0.2613,
+      "step": 1036
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.432833027684884e-05,
+      "loss": 0.2079,
+      "step": 1037
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.431764259266925e-05,
+      "loss": 0.5204,
+      "step": 1038
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.430694590138513e-05,
+      "loss": 0.213,
+      "step": 1039
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.4296240206958474e-05,
+      "loss": 0.3367,
+      "step": 1040
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.4285525513354606e-05,
+      "loss": 0.719,
+      "step": 1041
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.427480182454218e-05,
+      "loss": 0.7398,
+      "step": 1042
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.426406914449318e-05,
+      "loss": 0.5109,
+      "step": 1043
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.4253327477182934e-05,
+      "loss": 0.232,
+      "step": 1044
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.424257682659007e-05,
+      "loss": 0.4703,
+      "step": 1045
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.423181719669658e-05,
+      "loss": 0.6761,
+      "step": 1046
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.422104859148777e-05,
+      "loss": 0.1604,
+      "step": 1047
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.421027101495224e-05,
+      "loss": 0.5989,
+      "step": 1048
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.419948447108196e-05,
+      "loss": 0.3687,
+      "step": 1049
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.418868896387219e-05,
+      "loss": 0.859,
+      "step": 1050
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.4177884497321515e-05,
+      "loss": 1.2349,
+      "step": 1051
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.416707107543186e-05,
+      "loss": 0.2221,
+      "step": 1052
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.415624870220843e-05,
+      "loss": 0.2181,
+      "step": 1053
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.4145417381659776e-05,
+      "loss": 0.4971,
+      "step": 1054
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.4134577117797745e-05,
+      "loss": 0.3154,
+      "step": 1055
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.412372791463752e-05,
+      "loss": 0.9056,
+      "step": 1056
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.411286977619755e-05,
+      "loss": 0.5873,
+      "step": 1057
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.410200270649964e-05,
+      "loss": 0.4875,
+      "step": 1058
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.40911267095689e-05,
+      "loss": 0.5211,
+      "step": 1059
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.40802417894337e-05,
+      "loss": 0.37,
+      "step": 1060
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.406934795012577e-05,
+      "loss": 0.2767,
+      "step": 1061
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.405844519568012e-05,
+      "loss": 0.9938,
+      "step": 1062
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.4047533530135056e-05,
+      "loss": 0.4033,
+      "step": 1063
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.403661295753219e-05,
+      "loss": 0.2272,
+      "step": 1064
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.4025683481916434e-05,
+      "loss": 0.551,
+      "step": 1065
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.4014745107336e-05,
+      "loss": 0.7068,
+      "step": 1066
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.400379783784241e-05,
+      "loss": 1.1042,
+      "step": 1067
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.3992841677490434e-05,
+      "loss": 0.4266,
+      "step": 1068
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.398187663033819e-05,
+      "loss": 0.6876,
+      "step": 1069
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.397090270044706e-05,
+      "loss": 0.2258,
+      "step": 1070
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.395991989188171e-05,
+      "loss": 1.0326,
+      "step": 1071
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.394892820871012e-05,
+      "loss": 0.5196,
+      "step": 1072
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.393792765500353e-05,
+      "loss": 0.6223,
+      "step": 1073
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.392691823483648e-05,
+      "loss": 0.5421,
+      "step": 1074
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.3915899952286786e-05,
+      "loss": 0.5366,
+      "step": 1075
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.390487281143556e-05,
+      "loss": 0.332,
+      "step": 1076
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.389383681636718e-05,
+      "loss": 0.4843,
+      "step": 1077
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.388279197116931e-05,
+      "loss": 0.673,
+      "step": 1078
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.38717382799329e-05,
+      "loss": 0.4202,
+      "step": 1079
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.386067574675217e-05,
+      "loss": 0.5696,
+      "step": 1080
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.384960437572461e-05,
+      "loss": 0.7408,
+      "step": 1081
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.3838524170950976e-05,
+      "loss": 0.6798,
+      "step": 1082
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.382743513653533e-05,
+      "loss": 0.3016,
+      "step": 1083
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.381633727658496e-05,
+      "loss": 0.2182,
+      "step": 1084
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.380523059521046e-05,
+      "loss": 0.4621,
+      "step": 1085
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.379411509652567e-05,
+      "loss": 0.2048,
+      "step": 1086
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.37829907846477e-05,
+      "loss": 0.4608,
+      "step": 1087
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.377185766369693e-05,
+      "loss": 0.176,
+      "step": 1088
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.3760715737797e-05,
+      "loss": 0.6816,
+      "step": 1089
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.3749565011074795e-05,
+      "loss": 0.2327,
+      "step": 1090
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.3738405487660504e-05,
+      "loss": 0.369,
+      "step": 1091
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.372723717168753e-05,
+      "loss": 0.2542,
+      "step": 1092
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.371606006729254e-05,
+      "loss": 0.5839,
+      "step": 1093
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.370487417861548e-05,
+      "loss": 0.5488,
+      "step": 1094
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.3693679509799525e-05,
+      "loss": 0.2589,
+      "step": 1095
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.368247606499111e-05,
+      "loss": 0.5813,
+      "step": 1096
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.3671263848339934e-05,
+      "loss": 0.6831,
+      "step": 1097
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.366004286399891e-05,
+      "loss": 0.6192,
+      "step": 1098
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.364881311612424e-05,
+      "loss": 0.1838,
+      "step": 1099
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.363757460887535e-05,
+      "loss": 0.3605,
+      "step": 1100
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.362632734641492e-05,
+      "loss": 0.5466,
+      "step": 1101
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.3615071332908846e-05,
+      "loss": 0.5712,
+      "step": 1102
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.360380657252629e-05,
+      "loss": 0.2646,
+      "step": 1103
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.359253306943968e-05,
+      "loss": 0.6681,
+      "step": 1104
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.358125082782461e-05,
+      "loss": 0.7137,
+      "step": 1105
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.356995985185999e-05,
+      "loss": 1.0474,
+      "step": 1106
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.35586601457279e-05,
+      "loss": 0.6482,
+      "step": 1107
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.3547351713613685e-05,
+      "loss": 0.9771,
+      "step": 1108
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.353603455970593e-05,
+      "loss": 1.0409,
+      "step": 1109
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.3524708688196436e-05,
+      "loss": 0.2359,
+      "step": 1110
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.351337410328023e-05,
+      "loss": 0.5244,
+      "step": 1111
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.350203080915558e-05,
+      "loss": 0.2975,
+      "step": 1112
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.349067881002396e-05,
+      "loss": 0.3072,
+      "step": 1113
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.347931811009009e-05,
+      "loss": 0.6718,
+      "step": 1114
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.346794871356191e-05,
+      "loss": 0.6856,
+      "step": 1115
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.345657062465055e-05,
+      "loss": 0.5479,
+      "step": 1116
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.344518384757041e-05,
+      "loss": 0.3224,
+      "step": 1117
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.3433788386539064e-05,
+      "loss": 0.3598,
+      "step": 1118
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.3422384245777324e-05,
+      "loss": 0.7395,
+      "step": 1119
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.341097142950922e-05,
+      "loss": 0.1944,
+      "step": 1120
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.3399549941961986e-05,
+      "loss": 0.5321,
+      "step": 1121
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.338811978736608e-05,
+      "loss": 0.3541,
+      "step": 1122
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.3376680969955126e-05,
+      "loss": 0.8136,
+      "step": 1123
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.336523349396603e-05,
+      "loss": 1.0664,
+      "step": 1124
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.3353777363638855e-05,
+      "loss": 0.5164,
+      "step": 1125
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.334231258321687e-05,
+      "loss": 0.3808,
+      "step": 1126
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.3330839156946566e-05,
+      "loss": 0.2061,
+      "step": 1127
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.331935708907763e-05,
+      "loss": 0.3452,
+      "step": 1128
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.330786638386295e-05,
+      "loss": 0.834,
+      "step": 1129
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.3296367045558605e-05,
+      "loss": 0.5285,
+      "step": 1130
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.328485907842389e-05,
+      "loss": 0.6488,
+      "step": 1131
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.327334248672128e-05,
+      "loss": 0.4344,
+      "step": 1132
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.326181727471645e-05,
+      "loss": 0.7042,
+      "step": 1133
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.325028344667826e-05,
+      "loss": 0.6484,
+      "step": 1134
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.3238741006878776e-05,
+      "loss": 0.6597,
+      "step": 1135
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.322718995959325e-05,
+      "loss": 0.7593,
+      "step": 1136
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.321563030910011e-05,
+      "loss": 0.4066,
+      "step": 1137
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.320406205968098e-05,
+      "loss": 0.3167,
+      "step": 1138
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.3192485215620675e-05,
+      "loss": 0.4071,
+      "step": 1139
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.318089978120717e-05,
+      "loss": 0.3638,
+      "step": 1140
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.316930576073167e-05,
+      "loss": 0.8477,
+      "step": 1141
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.31577031584885e-05,
+      "loss": 0.4141,
+      "step": 1142
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.314609197877519e-05,
+      "loss": 0.7527,
+      "step": 1143
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.313447222589248e-05,
+      "loss": 0.5151,
+      "step": 1144
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.312284390414422e-05,
+      "loss": 0.2228,
+      "step": 1145
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.311120701783749e-05,
+      "loss": 0.4929,
+      "step": 1146
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.309956157128251e-05,
+      "loss": 0.3233,
+      "step": 1147
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.3087907568792697e-05,
+      "loss": 0.2347,
+      "step": 1148
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.30762450146846e-05,
+      "loss": 0.361,
+      "step": 1149
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.306457391327797e-05,
+      "loss": 0.3539,
+      "step": 1150
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.30528942688957e-05,
+      "loss": 0.4971,
+      "step": 1151
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.3041206085863885e-05,
+      "loss": 0.5976,
+      "step": 1152
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.302950936851172e-05,
+      "loss": 0.5734,
+      "step": 1153
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.301780412117161e-05,
+      "loss": 0.2022,
+      "step": 1154
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.300609034817913e-05,
+      "loss": 0.6388,
+      "step": 1155
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.299436805387295e-05,
+      "loss": 0.8052,
+      "step": 1156
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.298263724259496e-05,
+      "loss": 0.5965,
+      "step": 1157
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.2970897918690176e-05,
+      "loss": 0.487,
+      "step": 1158
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.295915008650678e-05,
+      "loss": 0.6199,
+      "step": 1159
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.294739375039607e-05,
+      "loss": 0.3594,
+      "step": 1160
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.293562891471254e-05,
+      "loss": 1.0322,
+      "step": 1161
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.2923855583813813e-05,
+      "loss": 0.3528,
+      "step": 1162
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.291207376206065e-05,
+      "loss": 0.2323,
+      "step": 1163
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.2900283453816976e-05,
+      "loss": 1.0323,
+      "step": 1164
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.288848466344983e-05,
+      "loss": 0.1942,
+      "step": 1165
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.2876677395329425e-05,
+      "loss": 0.2713,
+      "step": 1166
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.28648616538291e-05,
+      "loss": 0.7113,
+      "step": 1167
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.285303744332532e-05,
+      "loss": 0.5827,
+      "step": 1168
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.284120476819772e-05,
+      "loss": 0.2643,
+      "step": 1169
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.282936363282903e-05,
+      "loss": 0.6404,
+      "step": 1170
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.2817514041605146e-05,
+      "loss": 0.704,
+      "step": 1171
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.280565599891509e-05,
+      "loss": 0.3577,
+      "step": 1172
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.279378950915098e-05,
+      "loss": 0.4294,
+      "step": 1173
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.278191457670812e-05,
+      "loss": 0.6226,
+      "step": 1174
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.2770031205984904e-05,
+      "loss": 0.8817,
+      "step": 1175
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.275813940138284e-05,
+      "loss": 0.3103,
+      "step": 1176
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.27462391673066e-05,
+      "loss": 0.5924,
+      "step": 1177
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.273433050816395e-05,
+      "loss": 0.5006,
+      "step": 1178
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.272241342836578e-05,
+      "loss": 0.2673,
+      "step": 1179
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.271048793232611e-05,
+      "loss": 0.5999,
+      "step": 1180
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.2698554024462066e-05,
+      "loss": 0.2999,
+      "step": 1181
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.2686611709193895e-05,
+      "loss": 0.2308,
+      "step": 1182
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.267466099094496e-05,
+      "loss": 0.2563,
+      "step": 1183
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.2662701874141724e-05,
+      "loss": 0.3701,
+      "step": 1184
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.2650734363213786e-05,
+      "loss": 0.3923,
+      "step": 1185
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.263875846259382e-05,
+      "loss": 0.3021,
+      "step": 1186
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.262677417671763e-05,
+      "loss": 0.594,
+      "step": 1187
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.261478151002412e-05,
+      "loss": 0.4287,
+      "step": 1188
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.260278046695532e-05,
+      "loss": 0.7671,
+      "step": 1189
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.2590771051956315e-05,
+      "loss": 0.4122,
+      "step": 1190
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.2578753269475337e-05,
+      "loss": 0.5162,
+      "step": 1191
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.2566727123963685e-05,
+      "loss": 0.5037,
+      "step": 1192
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.255469261987578e-05,
+      "loss": 0.9975,
+      "step": 1193
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.254264976166912e-05,
+      "loss": 0.1973,
+      "step": 1194
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.253059855380431e-05,
+      "loss": 1.2196,
+      "step": 1195
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.251853900074504e-05,
+      "loss": 0.1971,
+      "step": 1196
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.25064711069581e-05,
+      "loss": 0.6289,
+      "step": 1197
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.249439487691334e-05,
+      "loss": 0.3742,
+      "step": 1198
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.248231031508375e-05,
+      "loss": 0.5737,
+      "step": 1199
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.2470217425945376e-05,
+      "loss": 0.7659,
+      "step": 1200
+    },
+    {
+      "epoch": 0.93,
+      "eval_loss": 1.124508261680603,
+      "eval_runtime": 157.1038,
+      "eval_samples_per_second": 0.337,
+      "eval_steps_per_second": 0.172,
+      "step": 1200
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.245811621397733e-05,
+      "loss": 0.9059,
+      "step": 1201
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.2446006683661836e-05,
+      "loss": 0.4814,
+      "step": 1202
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.243388883948419e-05,
+      "loss": 0.509,
+      "step": 1203
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.2421762685932766e-05,
+      "loss": 0.2858,
+      "step": 1204
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.240962822749901e-05,
+      "loss": 0.3293,
+      "step": 1205
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.2397485468677465e-05,
+      "loss": 0.1838,
+      "step": 1206
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.2385334413965724e-05,
+      "loss": 1.1416,
+      "step": 1207
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.237317506786446e-05,
+      "loss": 0.3592,
+      "step": 1208
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.236100743487743e-05,
+      "loss": 0.2955,
+      "step": 1209
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.2348831519511437e-05,
+      "loss": 0.2902,
+      "step": 1210
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.233664732627637e-05,
+      "loss": 0.2106,
+      "step": 1211
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.232445485968518e-05,
+      "loss": 0.3349,
+      "step": 1212
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.231225412425389e-05,
+      "loss": 0.2874,
+      "step": 1213
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.230004512450156e-05,
+      "loss": 0.5994,
+      "step": 1214
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.228782786495033e-05,
+      "loss": 0.233,
+      "step": 1215
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.227560235012541e-05,
+      "loss": 0.6191,
+      "step": 1216
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.226336858455505e-05,
+      "loss": 0.7676,
+      "step": 1217
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.225112657277057e-05,
+      "loss": 0.5631,
+      "step": 1218
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.223887631930631e-05,
+      "loss": 0.2821,
+      "step": 1219
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.222661782869971e-05,
+      "loss": 0.2304,
+      "step": 1220
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.221435110549124e-05,
+      "loss": 0.7935,
+      "step": 1221
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.22020761542244e-05,
+      "loss": 0.4436,
+      "step": 1222
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.2189792979445763e-05,
+      "loss": 0.6695,
+      "step": 1223
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.217750158570496e-05,
+      "loss": 0.7501,
+      "step": 1224
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.216520197755462e-05,
+      "loss": 0.3554,
+      "step": 1225
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.215289415955047e-05,
+      "loss": 0.6012,
+      "step": 1226
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.214057813625122e-05,
+      "loss": 0.7311,
+      "step": 1227
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.212825391221866e-05,
+      "loss": 0.2333,
+      "step": 1228
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.211592149201763e-05,
+      "loss": 0.4106,
+      "step": 1229
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.210358088021595e-05,
+      "loss": 0.3541,
+      "step": 1230
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.209123208138452e-05,
+      "loss": 0.5,
+      "step": 1231
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.207887510009725e-05,
+      "loss": 0.6652,
+      "step": 1232
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.20665099409311e-05,
+      "loss": 0.6271,
+      "step": 1233
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.205413660846605e-05,
+      "loss": 0.2461,
+      "step": 1234
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.2041755107285095e-05,
+      "loss": 0.4145,
+      "step": 1235
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.2029365441974275e-05,
+      "loss": 0.4961,
+      "step": 1236
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.201696761712265e-05,
+      "loss": 0.7363,
+      "step": 1237
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.200456163732228e-05,
+      "loss": 0.4666,
+      "step": 1238
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.199214750716828e-05,
+      "loss": 0.2435,
+      "step": 1239
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.197972523125875e-05,
+      "loss": 0.2195,
+      "step": 1240
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.1967294814194845e-05,
+      "loss": 0.3345,
+      "step": 1241
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.195485626058071e-05,
+      "loss": 0.4727,
+      "step": 1242
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.194240957502349e-05,
+      "loss": 0.2892,
+      "step": 1243
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.1929954762133385e-05,
+      "loss": 0.4445,
+      "step": 1244
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.191749182652356e-05,
+      "loss": 0.5501,
+      "step": 1245
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.190502077281023e-05,
+      "loss": 0.5881,
+      "step": 1246
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.1892541605612576e-05,
+      "loss": 0.4988,
+      "step": 1247
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.1880054329552804e-05,
+      "loss": 0.3994,
+      "step": 1248
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.186755894925615e-05,
+      "loss": 0.8145,
+      "step": 1249
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.18550554693508e-05,
+      "loss": 0.9466,
+      "step": 1250
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.184254389446798e-05,
+      "loss": 0.2658,
+      "step": 1251
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.183002422924189e-05,
+      "loss": 0.4322,
+      "step": 1252
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.1817496478309746e-05,
+      "loss": 0.5787,
+      "step": 1253
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.180496064631174e-05,
+      "loss": 0.2213,
+      "step": 1254
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.179241673789108e-05,
+      "loss": 0.1626,
+      "step": 1255
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.1779864757693926e-05,
+      "loss": 0.2616,
+      "step": 1256
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.176730471036948e-05,
+      "loss": 0.3199,
+      "step": 1257
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.17547366005699e-05,
+      "loss": 0.4169,
+      "step": 1258
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.174216043295033e-05,
+      "loss": 0.572,
+      "step": 1259
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.1729576212168914e-05,
+      "loss": 0.559,
+      "step": 1260
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.171698394288676e-05,
+      "loss": 0.7647,
+      "step": 1261
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.1704383629767966e-05,
+      "loss": 0.6343,
+      "step": 1262
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.169177527747962e-05,
+      "loss": 0.2766,
+      "step": 1263
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.167915889069177e-05,
+      "loss": 0.3793,
+      "step": 1264
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.166653447407745e-05,
+      "loss": 0.4751,
+      "step": 1265
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.165390203231267e-05,
+      "loss": 0.5339,
+      "step": 1266
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.1641261570076406e-05,
+      "loss": 0.4107,
+      "step": 1267
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.16286130920506e-05,
+      "loss": 0.652,
+      "step": 1268
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.161595660292018e-05,
+      "loss": 0.2963,
+      "step": 1269
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.1603292107373026e-05,
+      "loss": 0.677,
+      "step": 1270
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.159061961009998e-05,
+      "loss": 0.304,
+      "step": 1271
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.157793911579487e-05,
+      "loss": 0.732,
+      "step": 1272
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.156525062915447e-05,
+      "loss": 0.8198,
+      "step": 1273
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.155255415487853e-05,
+      "loss": 0.7762,
+      "step": 1274
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.153984969766971e-05,
+      "loss": 0.2926,
+      "step": 1275
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.15271372622337e-05,
+      "loss": 0.5542,
+      "step": 1276
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.151441685327908e-05,
+      "loss": 0.6225,
+      "step": 1277
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.1501688475517435e-05,
+      "loss": 0.225,
+      "step": 1278
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.1488952133663266e-05,
+      "loss": 0.2426,
+      "step": 1279
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.1476207832434015e-05,
+      "loss": 0.3492,
+      "step": 1280
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.1463455576550136e-05,
+      "loss": 0.5778,
+      "step": 1281
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.145069537073495e-05,
+      "loss": 1.22,
+      "step": 1282
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.143792721971477e-05,
+      "loss": 0.5206,
+      "step": 1283
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.1425151128218836e-05,
+      "loss": 0.8969,
+      "step": 1284
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.1412367100979343e-05,
+      "loss": 1.0483,
+      "step": 1285
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.139957514273141e-05,
+      "loss": 0.6419,
+      "step": 1286
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.1386775258213094e-05,
+      "loss": 0.7383,
+      "step": 1287
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.137396745216541e-05,
+      "loss": 0.5142,
+      "step": 1288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.136115172933226e-05,
+      "loss": 0.2054,
+      "step": 1289
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.134832809446054e-05,
+      "loss": 0.5225,
+      "step": 1290
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.133549655230003e-05,
+      "loss": 0.9442,
+      "step": 1291
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.132265710760346e-05,
+      "loss": 0.6469,
+      "step": 1292
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.130980976512648e-05,
+      "loss": 0.8553,
+      "step": 1293
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.129695452962766e-05,
+      "loss": 0.1662,
+      "step": 1294
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.1284091405868496e-05,
+      "loss": 0.3855,
+      "step": 1295
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.1271220398613427e-05,
+      "loss": 0.4631,
+      "step": 1296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.1258341512629777e-05,
+      "loss": 0.1923,
+      "step": 1297
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.124545475268781e-05,
+      "loss": 0.3967,
+      "step": 1298
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.123256012356071e-05,
+      "loss": 0.1455,
+      "step": 1299
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.121965763002457e-05,
+      "loss": 0.1501,
+      "step": 1300
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.120674727685838e-05,
+      "loss": 0.4323,
+      "step": 1301
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.1193829068844063e-05,
+      "loss": 0.4711,
+      "step": 1302
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.1180903010766444e-05,
+      "loss": 0.2094,
+      "step": 1303
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.1167969107413255e-05,
+      "loss": 0.4434,
+      "step": 1304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.115502736357513e-05,
+      "loss": 0.5014,
+      "step": 1305
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.114207778404561e-05,
+      "loss": 0.1922,
+      "step": 1306
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.112912037362115e-05,
+      "loss": 0.1616,
+      "step": 1307
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.111615513710108e-05,
+      "loss": 0.5406,
+      "step": 1308
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.110318207928766e-05,
+      "loss": 0.1891,
+      "step": 1309
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.1090201204986016e-05,
+      "loss": 0.4263,
+      "step": 1310
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.1077212519004184e-05,
+      "loss": 0.299,
+      "step": 1311
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.10642160261531e-05,
+      "loss": 0.3111,
+      "step": 1312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.105121173124659e-05,
+      "loss": 0.5293,
+      "step": 1313
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.1038199639101354e-05,
+      "loss": 0.2241,
+      "step": 1314
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.102517975453699e-05,
+      "loss": 0.3809,
+      "step": 1315
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.101215208237599e-05,
+      "loss": 0.1983,
+      "step": 1316
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.0999116627443716e-05,
+      "loss": 0.1873,
+      "step": 1317
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.0986073394568434e-05,
+      "loss": 0.49,
+      "step": 1318
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.097302238858125e-05,
+      "loss": 0.1789,
+      "step": 1319
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.095996361431621e-05,
+      "loss": 0.4747,
+      "step": 1320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.094689707661017e-05,
+      "loss": 0.2262,
+      "step": 1321
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.0933822780302916e-05,
+      "loss": 0.1482,
+      "step": 1322
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.092074073023708e-05,
+      "loss": 0.1716,
+      "step": 1323
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.090765093125818e-05,
+      "loss": 0.4558,
+      "step": 1324
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.0894553388214584e-05,
+      "loss": 0.4014,
+      "step": 1325
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5.088144810595754e-05,
+      "loss": 0.3944,
+      "step": 1326
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5.0868335089341184e-05,
+      "loss": 0.1944,
+      "step": 1327
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5.085521434322248e-05,
+      "loss": 0.4028,
+      "step": 1328
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5.0842085872461274e-05,
+      "loss": 0.1944,
+      "step": 1329
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5.0828949681920264e-05,
+      "loss": 0.544,
+      "step": 1330
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5.081580577646502e-05,
+      "loss": 0.396,
+      "step": 1331
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5.080265416096397e-05,
+      "loss": 0.3348,
+      "step": 1332
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5.0789494840288386e-05,
+      "loss": 0.5905,
+      "step": 1333
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5.07763278193124e-05,
+      "loss": 0.3371,
+      "step": 1334
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5.0763153102912985e-05,
+      "loss": 0.5343,
+      "step": 1335
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5.074997069596999e-05,
+      "loss": 0.4782,
+      "step": 1336
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5.073678060336609e-05,
+      "loss": 0.1651,
+      "step": 1337
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5.072358282998681e-05,
+      "loss": 0.4555,
+      "step": 1338
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5.0710377380720526e-05,
+      "loss": 0.4876,
+      "step": 1339
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5.069716426045846e-05,
+      "loss": 0.1806,
+      "step": 1340
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5.0683943474094656e-05,
+      "loss": 0.2387,
+      "step": 1341
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5.067071502652603e-05,
+      "loss": 0.2011,
+      "step": 1342
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5.065747892265231e-05,
+      "loss": 0.1342,
+      "step": 1343
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5.064423516737606e-05,
+      "loss": 0.2065,
+      "step": 1344
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5.06309837656027e-05,
+      "loss": 0.1451,
+      "step": 1345
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5.061772472224045e-05,
+      "loss": 0.139,
+      "step": 1346
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5.060445804220039e-05,
+      "loss": 0.141,
+      "step": 1347
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5.0591183730396413e-05,
+      "loss": 0.4356,
+      "step": 1348
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5.057790179174524e-05,
+      "loss": 0.4959,
+      "step": 1349
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5.0564612231166425e-05,
+      "loss": 0.581,
+      "step": 1350
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5.055131505358233e-05,
+      "loss": 0.3189,
+      "step": 1351
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5.053801026391815e-05,
+      "loss": 0.1556,
+      "step": 1352
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5.052469786710191e-05,
+      "loss": 0.2869,
+      "step": 1353
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5.051137786806441e-05,
+      "loss": 0.3157,
+      "step": 1354
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5.049805027173932e-05,
+      "loss": 0.336,
+      "step": 1355
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5.048471508306311e-05,
+      "loss": 0.178,
+      "step": 1356
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5.0471372306975016e-05,
+      "loss": 0.4372,
+      "step": 1357
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5.045802194841714e-05,
+      "loss": 0.3221,
+      "step": 1358
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5.044466401233437e-05,
+      "loss": 0.1757,
+      "step": 1359
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5.0431298503674414e-05,
+      "loss": 0.3013,
+      "step": 1360
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5.041792542738776e-05,
+      "loss": 0.3974,
+      "step": 1361
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5.040454478842771e-05,
+      "loss": 0.1375,
+      "step": 1362
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5.039115659175038e-05,
+      "loss": 0.204,
+      "step": 1363
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5.0377760842314657e-05,
+      "loss": 0.3789,
+      "step": 1364
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5.036435754508228e-05,
+      "loss": 0.1087,
+      "step": 1365
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5.035094670501771e-05,
+      "loss": 0.2179,
+      "step": 1366
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5.033752832708827e-05,
+      "loss": 0.1625,
+      "step": 1367
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5.0324102416264015e-05,
+      "loss": 0.4605,
+      "step": 1368
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5.031066897751784e-05,
+      "loss": 0.1882,
+      "step": 1369
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5.029722801582541e-05,
+      "loss": 0.3365,
+      "step": 1370
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5.028377953616515e-05,
+      "loss": 0.2703,
+      "step": 1371
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5.0270323543518316e-05,
+      "loss": 0.461,
+      "step": 1372
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5.025686004286892e-05,
+      "loss": 0.371,
+      "step": 1373
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5.024338903920375e-05,
+      "loss": 0.39,
+      "step": 1374
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5.0229910537512396e-05,
+      "loss": 0.1489,
+      "step": 1375
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5.0216424542787205e-05,
+      "loss": 0.2114,
+      "step": 1376
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5.0202931060023294e-05,
+      "loss": 0.2169,
+      "step": 1377
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5.01894300942186e-05,
+      "loss": 0.3814,
+      "step": 1378
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5.017592165037376e-05,
+      "loss": 0.3749,
+      "step": 1379
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5.016240573349223e-05,
+      "loss": 0.3143,
+      "step": 1380
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5.014888234858024e-05,
+      "loss": 0.4398,
+      "step": 1381
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5.0135351500646745e-05,
+      "loss": 0.1912,
+      "step": 1382
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5.01218131947035e-05,
+      "loss": 0.5447,
+      "step": 1383
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5.0108267435765e-05,
+      "loss": 0.4238,
+      "step": 1384
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5.009471422884852e-05,
+      "loss": 0.4337,
+      "step": 1385
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5.0081153578974086e-05,
+      "loss": 0.2089,
+      "step": 1386
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5.006758549116447e-05,
+      "loss": 0.2403,
+      "step": 1387
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5.0054009970445204e-05,
+      "loss": 0.1471,
+      "step": 1388
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5.00404270218446e-05,
+      "loss": 0.1182,
+      "step": 1389
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5.0026836650393664e-05,
+      "loss": 0.4018,
+      "step": 1390
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5.0013238861126215e-05,
+      "loss": 0.3141,
+      "step": 1391
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 4.9999633659078774e-05,
+      "loss": 0.2692,
+      "step": 1392
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 4.9986021049290626e-05,
+      "loss": 0.7963,
+      "step": 1393
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 4.997240103680381e-05,
+      "loss": 0.1632,
+      "step": 1394
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 4.995877362666308e-05,
+      "loss": 0.135,
+      "step": 1395
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 4.9945138823915956e-05,
+      "loss": 0.1872,
+      "step": 1396
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 4.993149663361267e-05,
+      "loss": 0.2158,
+      "step": 1397
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 4.991784706080622e-05,
+      "loss": 0.363,
+      "step": 1398
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 4.990419011055231e-05,
+      "loss": 0.152,
+      "step": 1399
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 4.989052578790939e-05,
+      "loss": 0.587,
+      "step": 1400
+    },
+    {
+      "epoch": 1.08,
+      "eval_loss": 1.1660311222076416,
+      "eval_runtime": 157.1816,
+      "eval_samples_per_second": 0.337,
+      "eval_steps_per_second": 0.172,
+      "step": 1400
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 4.987685409793866e-05,
+      "loss": 0.7865,
+      "step": 1401
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 4.986317504570402e-05,
+      "loss": 0.5338,
+      "step": 1402
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 4.984948863627209e-05,
+      "loss": 0.1625,
+      "step": 1403
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 4.983579487471225e-05,
+      "loss": 0.4162,
+      "step": 1404
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 4.982209376609658e-05,
+      "loss": 0.1475,
+      "step": 1405
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 4.9808385315499886e-05,
+      "loss": 0.1346,
+      "step": 1406
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 4.979466952799969e-05,
+      "loss": 0.2249,
+      "step": 1407
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 4.9780946408676235e-05,
+      "loss": 0.577,
+      "step": 1408
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 4.976721596261248e-05,
+      "loss": 0.2926,
+      "step": 1409
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 4.97534781948941e-05,
+      "loss": 0.1914,
+      "step": 1410
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 4.9739733110609483e-05,
+      "loss": 0.1839,
+      "step": 1411
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 4.9725980714849725e-05,
+      "loss": 0.4626,
+      "step": 1412
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 4.971222101270862e-05,
+      "loss": 0.3006,
+      "step": 1413
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 4.969845400928267e-05,
+      "loss": 0.4917,
+      "step": 1414
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 4.9684679709671096e-05,
+      "loss": 0.2917,
+      "step": 1415
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 4.967089811897582e-05,
+      "loss": 0.1477,
+      "step": 1416
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 4.965710924230145e-05,
+      "loss": 0.477,
+      "step": 1417
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 4.964331308475531e-05,
+      "loss": 0.3172,
+      "step": 1418
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 4.9629509651447396e-05,
+      "loss": 0.4378,
+      "step": 1419
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 4.9615698947490426e-05,
+      "loss": 0.2764,
+      "step": 1420
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 4.96018809779998e-05,
+      "loss": 0.3052,
+      "step": 1421
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 4.9588055748093594e-05,
+      "loss": 0.1798,
+      "step": 1422
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 4.957422326289261e-05,
+      "loss": 0.5825,
+      "step": 1423
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 4.95603835275203e-05,
+      "loss": 0.1545,
+      "step": 1424
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 4.954653654710282e-05,
+      "loss": 0.4536,
+      "step": 1425
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 4.9532682326769e-05,
+      "loss": 0.2447,
+      "step": 1426
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 4.951882087165037e-05,
+      "loss": 0.2584,
+      "step": 1427
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 4.950495218688112e-05,
+      "loss": 0.1623,
+      "step": 1428
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 4.949107627759813e-05,
+      "loss": 0.4281,
+      "step": 1429
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 4.947719314894094e-05,
+      "loss": 0.1834,
+      "step": 1430
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 4.946330280605179e-05,
+      "loss": 0.2156,
+      "step": 1431
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 4.9449405254075564e-05,
+      "loss": 0.1361,
+      "step": 1432
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 4.943550049815983e-05,
+      "loss": 0.391,
+      "step": 1433
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 4.942158854345483e-05,
+      "loss": 0.5056,
+      "step": 1434
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 4.9407669395113467e-05,
+      "loss": 0.1457,
+      "step": 1435
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 4.939374305829131e-05,
+      "loss": 0.1824,
+      "step": 1436
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 4.9379809538146576e-05,
+      "loss": 0.1831,
+      "step": 1437
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 4.936586883984016e-05,
+      "loss": 0.2427,
+      "step": 1438
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 4.9351920968535614e-05,
+      "loss": 0.3717,
+      "step": 1439
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 4.933796592939914e-05,
+      "loss": 0.1933,
+      "step": 1440
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 4.9324003727599607e-05,
+      "loss": 0.4872,
+      "step": 1441
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 4.931003436830851e-05,
+      "loss": 0.1696,
+      "step": 1442
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 4.929605785670004e-05,
+      "loss": 0.1328,
+      "step": 1443
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 4.928207419795098e-05,
+      "loss": 0.237,
+      "step": 1444
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 4.926808339724082e-05,
+      "loss": 0.1155,
+      "step": 1445
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 4.9254085459751634e-05,
+      "loss": 0.3289,
+      "step": 1446
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 4.9240080390668206e-05,
+      "loss": 0.1576,
+      "step": 1447
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 4.922606819517789e-05,
+      "loss": 0.322,
+      "step": 1448
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 4.921204887847075e-05,
+      "loss": 0.1991,
+      "step": 1449
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 4.9198022445739446e-05,
+      "loss": 0.2882,
+      "step": 1450
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 4.918398890217928e-05,
+      "loss": 0.1909,
+      "step": 1451
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 4.916994825298817e-05,
+      "loss": 0.1818,
+      "step": 1452
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 4.915590050336672e-05,
+      "loss": 0.3953,
+      "step": 1453
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 4.9141845658518095e-05,
+      "loss": 0.3209,
+      "step": 1454
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 4.912778372364816e-05,
+      "loss": 0.1353,
+      "step": 1455
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 4.911371470396533e-05,
+      "loss": 0.4684,
+      "step": 1456
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 4.909963860468071e-05,
+      "loss": 0.3078,
+      "step": 1457
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 4.9085555431007985e-05,
+      "loss": 0.1763,
+      "step": 1458
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 4.9071465188163494e-05,
+      "loss": 0.4894,
+      "step": 1459
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 4.905736788136616e-05,
+      "loss": 0.4859,
+      "step": 1460
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 4.904326351583754e-05,
+      "loss": 0.1757,
+      "step": 1461
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 4.902915209680181e-05,
+      "loss": 0.2282,
+      "step": 1462
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 4.901503362948574e-05,
+      "loss": 0.3282,
+      "step": 1463
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 4.900090811911874e-05,
+      "loss": 0.1731,
+      "step": 1464
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 4.89867755709328e-05,
+      "loss": 0.2264,
+      "step": 1465
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 4.897263599016254e-05,
+      "loss": 0.2526,
+      "step": 1466
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 4.8958489382045157e-05,
+      "loss": 0.4951,
+      "step": 1467
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 4.894433575182048e-05,
+      "loss": 0.4928,
+      "step": 1468
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 4.893017510473092e-05,
+      "loss": 0.4836,
+      "step": 1469
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 4.89160074460215e-05,
+      "loss": 0.5966,
+      "step": 1470
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 4.8901832780939814e-05,
+      "loss": 0.6066,
+      "step": 1471
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 4.88876511147361e-05,
+      "loss": 0.3777,
+      "step": 1472
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 4.887346245266313e-05,
+      "loss": 0.1671,
+      "step": 1473
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 4.885926679997632e-05,
+      "loss": 0.3247,
+      "step": 1474
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 4.884506416193364e-05,
+      "loss": 0.177,
+      "step": 1475
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 4.883085454379566e-05,
+      "loss": 0.2717,
+      "step": 1476
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 4.881663795082554e-05,
+      "loss": 0.0933,
+      "step": 1477
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 4.8802414388289014e-05,
+      "loss": 0.1717,
+      "step": 1478
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 4.87881838614544e-05,
+      "loss": 0.4947,
+      "step": 1479
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 4.8773946375592606e-05,
+      "loss": 0.2941,
+      "step": 1480
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 4.87597019359771e-05,
+      "loss": 0.1359,
+      "step": 1481
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 4.874545054788394e-05,
+      "loss": 0.661,
+      "step": 1482
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 4.873119221659175e-05,
+      "loss": 0.1803,
+      "step": 1483
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 4.871692694738172e-05,
+      "loss": 0.3092,
+      "step": 1484
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 4.870265474553763e-05,
+      "loss": 0.1388,
+      "step": 1485
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 4.86883756163458e-05,
+      "loss": 0.5941,
+      "step": 1486
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 4.867408956509515e-05,
+      "loss": 0.2171,
+      "step": 1487
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 4.865979659707713e-05,
+      "loss": 0.6448,
+      "step": 1488
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 4.864549671758579e-05,
+      "loss": 0.3269,
+      "step": 1489
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 4.863118993191768e-05,
+      "loss": 0.3005,
+      "step": 1490
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 4.861687624537197e-05,
+      "loss": 0.1825,
+      "step": 1491
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 4.8602555663250364e-05,
+      "loss": 0.1913,
+      "step": 1492
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 4.85882281908571e-05,
+      "loss": 0.1253,
+      "step": 1493
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 4.8573893833499e-05,
+      "loss": 0.5582,
+      "step": 1494
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 4.855955259648541e-05,
+      "loss": 0.5394,
+      "step": 1495
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 4.854520448512825e-05,
+      "loss": 0.3632,
+      "step": 1496
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 4.853084950474196e-05,
+      "loss": 0.5094,
+      "step": 1497
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 4.851648766064354e-05,
+      "loss": 0.2417,
+      "step": 1498
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 4.850211895815253e-05,
+      "loss": 0.4345,
+      "step": 1499
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 4.8487743402590997e-05,
+      "loss": 0.3805,
+      "step": 1500
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 4.847336099928357e-05,
+      "loss": 0.1335,
+      "step": 1501
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 4.84589717535574e-05,
+      "loss": 0.1933,
+      "step": 1502
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 4.8444575670742174e-05,
+      "loss": 0.4076,
+      "step": 1503
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 4.8430172756170116e-05,
+      "loss": 0.3403,
+      "step": 1504
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 4.8415763015175975e-05,
+      "loss": 0.3754,
+      "step": 1505
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 4.840134645309702e-05,
+      "loss": 0.2186,
+      "step": 1506
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 4.838692307527307e-05,
+      "loss": 0.3666,
+      "step": 1507
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 4.837249288704644e-05,
+      "loss": 0.1445,
+      "step": 1508
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 4.8358055893762e-05,
+      "loss": 0.142,
+      "step": 1509
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 4.8343612100767114e-05,
+      "loss": 0.3152,
+      "step": 1510
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 4.832916151341167e-05,
+      "loss": 0.1351,
+      "step": 1511
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 4.831470413704809e-05,
+      "loss": 0.3694,
+      "step": 1512
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 4.830023997703128e-05,
+      "loss": 0.3791,
+      "step": 1513
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 4.828576903871869e-05,
+      "loss": 0.3341,
+      "step": 1514
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 4.827129132747025e-05,
+      "loss": 0.3359,
+      "step": 1515
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 4.825680684864842e-05,
+      "loss": 0.2151,
+      "step": 1516
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 4.8242315607618175e-05,
+      "loss": 0.1612,
+      "step": 1517
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 4.822781760974697e-05,
+      "loss": 0.2277,
+      "step": 1518
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 4.821331286040478e-05,
+      "loss": 0.4105,
+      "step": 1519
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 4.819880136496406e-05,
+      "loss": 0.1641,
+      "step": 1520
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 4.8184283128799795e-05,
+      "loss": 0.273,
+      "step": 1521
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 4.8169758157289444e-05,
+      "loss": 0.159,
+      "step": 1522
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 4.815522645581296e-05,
+      "loss": 0.5423,
+      "step": 1523
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 4.8140688029752797e-05,
+      "loss": 0.1739,
+      "step": 1524
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 4.812614288449391e-05,
+      "loss": 0.1851,
+      "step": 1525
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 4.811159102542372e-05,
+      "loss": 0.3237,
+      "step": 1526
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 4.809703245793215e-05,
+      "loss": 0.1385,
+      "step": 1527
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 4.8082467187411605e-05,
+      "loss": 0.3158,
+      "step": 1528
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 4.8067895219256966e-05,
+      "loss": 0.5921,
+      "step": 1529
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 4.80533165588656e-05,
+      "loss": 0.2657,
+      "step": 1530
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 4.803873121163736e-05,
+      "loss": 0.2677,
+      "step": 1531
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 4.802413918297456e-05,
+      "loss": 0.3159,
+      "step": 1532
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 4.8009540478282e-05,
+      "loss": 0.4432,
+      "step": 1533
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 4.799493510296696e-05,
+      "loss": 0.1113,
+      "step": 1534
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 4.7980323062439164e-05,
+      "loss": 0.2258,
+      "step": 1535
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 4.7965704362110825e-05,
+      "loss": 0.4702,
+      "step": 1536
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 4.795107900739663e-05,
+      "loss": 0.398,
+      "step": 1537
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 4.793644700371372e-05,
+      "loss": 0.391,
+      "step": 1538
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 4.792180835648168e-05,
+      "loss": 0.1796,
+      "step": 1539
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 4.7907163071122585e-05,
+      "loss": 0.185,
+      "step": 1540
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 4.789251115306096e-05,
+      "loss": 0.4434,
+      "step": 1541
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 4.787785260772379e-05,
+      "loss": 0.2889,
+      "step": 1542
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 4.78631874405405e-05,
+      "loss": 0.7114,
+      "step": 1543
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 4.7848515656942994e-05,
+      "loss": 0.2064,
+      "step": 1544
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 4.783383726236558e-05,
+      "loss": 0.1789,
+      "step": 1545
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 4.7819152262245066e-05,
+      "loss": 0.1707,
+      "step": 1546
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 4.780446066202069e-05,
+      "loss": 0.1912,
+      "step": 1547
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 4.7789762467134114e-05,
+      "loss": 0.6474,
+      "step": 1548
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 4.777505768302947e-05,
+      "loss": 0.2589,
+      "step": 1549
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 4.7760346315153316e-05,
+      "loss": 0.5744,
+      "step": 1550
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 4.7745628368954645e-05,
+      "loss": 0.2402,
+      "step": 1551
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 4.77309038498849e-05,
+      "loss": 0.5489,
+      "step": 1552
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 4.771617276339796e-05,
+      "loss": 0.1786,
+      "step": 1553
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 4.770143511495011e-05,
+      "loss": 0.2166,
+      "step": 1554
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 4.7686690910000094e-05,
+      "loss": 0.3809,
+      "step": 1555
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 4.767194015400908e-05,
+      "loss": 0.3765,
+      "step": 1556
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 4.7657182852440635e-05,
+      "loss": 0.241,
+      "step": 1557
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 4.7642419010760796e-05,
+      "loss": 0.1601,
+      "step": 1558
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 4.7627648634438e-05,
+      "loss": 0.239,
+      "step": 1559
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 4.761287172894308e-05,
+      "loss": 0.3755,
+      "step": 1560
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 4.7598088299749335e-05,
+      "loss": 0.1803,
+      "step": 1561
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 4.7583298352332445e-05,
+      "loss": 0.7244,
+      "step": 1562
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 4.75685018921705e-05,
+      "loss": 0.4073,
+      "step": 1563
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 4.7553698924744046e-05,
+      "loss": 0.4278,
+      "step": 1564
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 4.753888945553599e-05,
+      "loss": 0.1208,
+      "step": 1565
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 4.752407349003168e-05,
+      "loss": 0.1952,
+      "step": 1566
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 4.750925103371885e-05,
+      "loss": 0.2739,
+      "step": 1567
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 4.749442209208765e-05,
+      "loss": 0.2894,
+      "step": 1568
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 4.747958667063063e-05,
+      "loss": 0.2516,
+      "step": 1569
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 4.7464744774842735e-05,
+      "loss": 0.2956,
+      "step": 1570
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 4.744989641022131e-05,
+      "loss": 0.3115,
+      "step": 1571
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 4.74350415822661e-05,
+      "loss": 0.3919,
+      "step": 1572
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 4.742018029647924e-05,
+      "loss": 0.163,
+      "step": 1573
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 4.740531255836526e-05,
+      "loss": 0.265,
+      "step": 1574
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 4.739043837343108e-05,
+      "loss": 0.3123,
+      "step": 1575
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 4.7375557747186e-05,
+      "loss": 0.8207,
+      "step": 1576
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 4.736067068514172e-05,
+      "loss": 0.507,
+      "step": 1577
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 4.734577719281231e-05,
+      "loss": 0.148,
+      "step": 1578
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 4.733087727571424e-05,
+      "loss": 0.3597,
+      "step": 1579
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 4.7315970939366325e-05,
+      "loss": 0.6706,
+      "step": 1580
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 4.730105818928979e-05,
+      "loss": 0.2539,
+      "step": 1581
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 4.728613903100824e-05,
+      "loss": 0.8123,
+      "step": 1582
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 4.727121347004761e-05,
+      "loss": 0.2354,
+      "step": 1583
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 4.725628151193626e-05,
+      "loss": 0.4379,
+      "step": 1584
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 4.7241343162204894e-05,
+      "loss": 0.3838,
+      "step": 1585
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 4.722639842638656e-05,
+      "loss": 0.5395,
+      "step": 1586
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 4.7211447310016724e-05,
+      "loss": 0.3968,
+      "step": 1587
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 4.719648981863317e-05,
+      "loss": 0.267,
+      "step": 1588
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 4.718152595777607e-05,
+      "loss": 0.1063,
+      "step": 1589
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 4.716655573298795e-05,
+      "loss": 0.3027,
+      "step": 1590
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 4.715157914981367e-05,
+      "loss": 0.4679,
+      "step": 1591
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 4.713659621380047e-05,
+      "loss": 0.2086,
+      "step": 1592
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 4.712160693049795e-05,
+      "loss": 0.1618,
+      "step": 1593
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 4.7106611305458035e-05,
+      "loss": 0.2109,
+      "step": 1594
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 4.709160934423502e-05,
+      "loss": 0.2449,
+      "step": 1595
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 4.707660105238553e-05,
+      "loss": 0.1461,
+      "step": 1596
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 4.7061586435468554e-05,
+      "loss": 0.2142,
+      "step": 1597
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 4.704656549904541e-05,
+      "loss": 0.2135,
+      "step": 1598
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 4.703153824867974e-05,
+      "loss": 0.2439,
+      "step": 1599
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 4.701650468993758e-05,
+      "loss": 0.8468,
+      "step": 1600
+    },
+    {
+      "epoch": 1.24,
+      "eval_loss": 1.1743203401565552,
+      "eval_runtime": 157.1296,
+      "eval_samples_per_second": 0.337,
+      "eval_steps_per_second": 0.172,
+      "step": 1600
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 4.700146482838725e-05,
+      "loss": 0.3526,
+      "step": 1601
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 4.698641866959941e-05,
+      "loss": 0.4107,
+      "step": 1602
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 4.697136621914708e-05,
+      "loss": 0.3319,
+      "step": 1603
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 4.695630748260559e-05,
+      "loss": 0.2944,
+      "step": 1604
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 4.69412424655526e-05,
+      "loss": 0.3102,
+      "step": 1605
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 4.692617117356811e-05,
+      "loss": 0.4157,
+      "step": 1606
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 4.6911093612234425e-05,
+      "loss": 0.1821,
+      "step": 1607
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 4.6896009787136175e-05,
+      "loss": 0.1417,
+      "step": 1608
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 4.688091970386033e-05,
+      "loss": 0.4915,
+      "step": 1609
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 4.6865823367996145e-05,
+      "loss": 0.4662,
+      "step": 1610
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 4.6850720785135225e-05,
+      "loss": 0.3912,
+      "step": 1611
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 4.683561196087148e-05,
+      "loss": 0.4161,
+      "step": 1612
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 4.6820496900801094e-05,
+      "loss": 0.1517,
+      "step": 1613
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 4.680537561052262e-05,
+      "loss": 0.1895,
+      "step": 1614
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 4.679024809563688e-05,
+      "loss": 0.3604,
+      "step": 1615
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 4.677511436174702e-05,
+      "loss": 0.2097,
+      "step": 1616
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 4.675997441445847e-05,
+      "loss": 0.4231,
+      "step": 1617
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 4.6744828259378975e-05,
+      "loss": 0.1562,
+      "step": 1618
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 4.672967590211858e-05,
+      "loss": 0.2951,
+      "step": 1619
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 4.6714517348289636e-05,
+      "loss": 0.21,
+      "step": 1620
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 4.669935260350676e-05,
+      "loss": 0.3808,
+      "step": 1621
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 4.668418167338688e-05,
+      "loss": 0.2323,
+      "step": 1622
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 4.666900456354923e-05,
+      "loss": 0.3059,
+      "step": 1623
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 4.66538212796153e-05,
+      "loss": 0.6909,
+      "step": 1624
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 4.663863182720891e-05,
+      "loss": 0.4043,
+      "step": 1625
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 4.662343621195611e-05,
+      "loss": 0.2909,
+      "step": 1626
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 4.6608234439485266e-05,
+      "loss": 0.4078,
+      "step": 1627
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 4.659302651542704e-05,
+      "loss": 0.4947,
+      "step": 1628
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 4.657781244541434e-05,
+      "loss": 0.1815,
+      "step": 1629
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 4.656259223508237e-05,
+      "loss": 0.1934,
+      "step": 1630
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 4.6547365890068596e-05,
+      "loss": 0.2339,
+      "step": 1631
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 4.653213341601276e-05,
+      "loss": 0.1622,
+      "step": 1632
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 4.6516894818556884e-05,
+      "loss": 0.1848,
+      "step": 1633
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 4.650165010334524e-05,
+      "loss": 0.1643,
+      "step": 1634
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 4.648639927602439e-05,
+      "loss": 0.4081,
+      "step": 1635
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 4.647114234224313e-05,
+      "loss": 0.1628,
+      "step": 1636
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 4.645587930765255e-05,
+      "loss": 0.167,
+      "step": 1637
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 4.644061017790598e-05,
+      "loss": 0.467,
+      "step": 1638
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 4.642533495865899e-05,
+      "loss": 0.3382,
+      "step": 1639
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 4.641005365556946e-05,
+      "loss": 0.384,
+      "step": 1640
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 4.639476627429747e-05,
+      "loss": 0.174,
+      "step": 1641
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 4.6379472820505375e-05,
+      "loss": 0.1315,
+      "step": 1642
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 4.636417329985778e-05,
+      "loss": 0.1873,
+      "step": 1643
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 4.634886771802153e-05,
+      "loss": 0.207,
+      "step": 1644
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 4.633355608066572e-05,
+      "loss": 0.3335,
+      "step": 1645
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 4.6318238393461675e-05,
+      "loss": 0.7602,
+      "step": 1646
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 4.630291466208298e-05,
+      "loss": 0.7919,
+      "step": 1647
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 4.628758489220546e-05,
+      "loss": 0.4882,
+      "step": 1648
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 4.627224908950715e-05,
+      "loss": 0.5787,
+      "step": 1649
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 4.625690725966836e-05,
+      "loss": 0.4899,
+      "step": 1650
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 4.6241559408371574e-05,
+      "loss": 0.164,
+      "step": 1651
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 4.6226205541301563e-05,
+      "loss": 0.2549,
+      "step": 1652
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 4.621084566414531e-05,
+      "loss": 0.142,
+      "step": 1653
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 4.619547978259201e-05,
+      "loss": 0.248,
+      "step": 1654
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 4.618010790233309e-05,
+      "loss": 0.5244,
+      "step": 1655
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 4.616473002906219e-05,
+      "loss": 0.5871,
+      "step": 1656
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 4.614934616847521e-05,
+      "loss": 0.1451,
+      "step": 1657
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 4.613395632627019e-05,
+      "loss": 0.3652,
+      "step": 1658
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 4.611856050814748e-05,
+      "loss": 0.5853,
+      "step": 1659
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 4.610315871980957e-05,
+      "loss": 0.27,
+      "step": 1660
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 4.608775096696119e-05,
+      "loss": 0.181,
+      "step": 1661
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 4.607233725530928e-05,
+      "loss": 0.3926,
+      "step": 1662
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 4.6056917590562974e-05,
+      "loss": 0.4263,
+      "step": 1663
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 4.604149197843365e-05,
+      "loss": 0.4352,
+      "step": 1664
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 4.6026060424634815e-05,
+      "loss": 0.1649,
+      "step": 1665
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 4.6010622934882254e-05,
+      "loss": 0.121,
+      "step": 1666
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 4.599517951489391e-05,
+      "loss": 0.2989,
+      "step": 1667
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 4.597973017038993e-05,
+      "loss": 0.2072,
+      "step": 1668
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 4.5964274907092644e-05,
+      "loss": 0.151,
+      "step": 1669
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 4.59488137307266e-05,
+      "loss": 0.138,
+      "step": 1670
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 4.5933346647018526e-05,
+      "loss": 0.2177,
+      "step": 1671
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 4.591787366169732e-05,
+      "loss": 0.3373,
+      "step": 1672
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 4.5902394780494084e-05,
+      "loss": 0.4049,
+      "step": 1673
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 4.5886910009142105e-05,
+      "loss": 0.2356,
+      "step": 1674
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 4.587141935337684e-05,
+      "loss": 0.1708,
+      "step": 1675
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 4.5855922818935946e-05,
+      "loss": 0.1559,
+      "step": 1676
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 4.584042041155923e-05,
+      "loss": 0.4135,
+      "step": 1677
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 4.582491213698869e-05,
+      "loss": 0.4111,
+      "step": 1678
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 4.5809398000968494e-05,
+      "loss": 0.1408,
+      "step": 1679
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 4.579387800924499e-05,
+      "loss": 0.3861,
+      "step": 1680
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 4.5778352167566675e-05,
+      "loss": 0.1348,
+      "step": 1681
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 4.576282048168424e-05,
+      "loss": 0.7301,
+      "step": 1682
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 4.574728295735052e-05,
+      "loss": 0.4178,
+      "step": 1683
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 4.573173960032051e-05,
+      "loss": 0.2441,
+      "step": 1684
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 4.5716190416351375e-05,
+      "loss": 0.3204,
+      "step": 1685
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 4.570063541120245e-05,
+      "loss": 0.1754,
+      "step": 1686
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 4.568507459063519e-05,
+      "loss": 0.2553,
+      "step": 1687
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 4.566950796041325e-05,
+      "loss": 0.4379,
+      "step": 1688
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 4.5653935526302386e-05,
+      "loss": 0.3735,
+      "step": 1689
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 4.563835729407057e-05,
+      "loss": 0.4039,
+      "step": 1690
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 4.562277326948785e-05,
+      "loss": 0.1876,
+      "step": 1691
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 4.5607183458326464e-05,
+      "loss": 0.3262,
+      "step": 1692
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 4.5591587866360786e-05,
+      "loss": 0.188,
+      "step": 1693
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 4.5575986499367326e-05,
+      "loss": 0.4432,
+      "step": 1694
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 4.556037936312472e-05,
+      "loss": 0.4442,
+      "step": 1695
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 4.554476646341378e-05,
+      "loss": 0.3282,
+      "step": 1696
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 4.55291478060174e-05,
+      "loss": 0.1616,
+      "step": 1697
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 4.551352339672065e-05,
+      "loss": 0.4991,
+      "step": 1698
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 4.549789324131072e-05,
+      "loss": 0.3436,
+      "step": 1699
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 4.5482257345576905e-05,
+      "loss": 0.1626,
+      "step": 1700
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 4.5466615715310654e-05,
+      "loss": 0.2576,
+      "step": 1701
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 4.5450968356305536e-05,
+      "loss": 0.4767,
+      "step": 1702
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 4.543531527435723e-05,
+      "loss": 0.1965,
+      "step": 1703
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 4.541965647526354e-05,
+      "loss": 0.1896,
+      "step": 1704
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 4.540399196482438e-05,
+      "loss": 0.3592,
+      "step": 1705
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 4.53883217488418e-05,
+      "loss": 0.1599,
+      "step": 1706
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 4.537264583311996e-05,
+      "loss": 0.2298,
+      "step": 1707
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 4.5356964223465094e-05,
+      "loss": 0.2025,
+      "step": 1708
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 4.534127692568561e-05,
+      "loss": 0.1583,
+      "step": 1709
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 4.5325583945591945e-05,
+      "loss": 0.3545,
+      "step": 1710
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 4.530988528899671e-05,
+      "loss": 0.2403,
+      "step": 1711
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 4.52941809617146e-05,
+      "loss": 0.3797,
+      "step": 1712
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 4.527847096956236e-05,
+      "loss": 0.3683,
+      "step": 1713
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 4.526275531835891e-05,
+      "loss": 0.2678,
+      "step": 1714
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 4.524703401392522e-05,
+      "loss": 0.3133,
+      "step": 1715
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 4.523130706208436e-05,
+      "loss": 0.4171,
+      "step": 1716
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 4.521557446866149e-05,
+      "loss": 0.1497,
+      "step": 1717
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 4.519983623948388e-05,
+      "loss": 0.4235,
+      "step": 1718
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 4.518409238038087e-05,
+      "loss": 0.3707,
+      "step": 1719
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 4.5168342897183885e-05,
+      "loss": 0.4167,
+      "step": 1720
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 4.515258779572642e-05,
+      "loss": 0.3375,
+      "step": 1721
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 4.5136827081844096e-05,
+      "loss": 0.1905,
+      "step": 1722
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 4.5121060761374567e-05,
+      "loss": 0.3045,
+      "step": 1723
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 4.5105288840157585e-05,
+      "loss": 0.1273,
+      "step": 1724
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 4.5089511324034974e-05,
+      "loss": 0.3415,
+      "step": 1725
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 4.507372821885062e-05,
+      "loss": 0.355,
+      "step": 1726
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 4.5057939530450495e-05,
+      "loss": 0.4731,
+      "step": 1727
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 4.504214526468264e-05,
+      "loss": 0.1307,
+      "step": 1728
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 4.5026345427397134e-05,
+      "loss": 0.2996,
+      "step": 1729
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 4.5010540024446145e-05,
+      "loss": 0.1463,
+      "step": 1730
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 4.4994729061683905e-05,
+      "loss": 0.1322,
+      "step": 1731
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 4.49789125449667e-05,
+      "loss": 0.3611,
+      "step": 1732
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 4.4963090480152846e-05,
+      "loss": 0.3522,
+      "step": 1733
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 4.494726287310276e-05,
+      "loss": 0.4063,
+      "step": 1734
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 4.49314297296789e-05,
+      "loss": 0.2713,
+      "step": 1735
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 4.491559105574573e-05,
+      "loss": 0.2225,
+      "step": 1736
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 4.489974685716982e-05,
+      "loss": 0.1822,
+      "step": 1737
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 4.488389713981976e-05,
+      "loss": 0.1621,
+      "step": 1738
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 4.486804190956618e-05,
+      "loss": 0.3388,
+      "step": 1739
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 4.485218117228177e-05,
+      "loss": 0.3912,
+      "step": 1740
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 4.483631493384125e-05,
+      "loss": 0.2755,
+      "step": 1741
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 4.4820443200121366e-05,
+      "loss": 0.6218,
+      "step": 1742
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 4.4804565977000916e-05,
+      "loss": 0.3416,
+      "step": 1743
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 4.478868327036071e-05,
+      "loss": 0.1315,
+      "step": 1744
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 4.4772795086083646e-05,
+      "loss": 0.1746,
+      "step": 1745
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 4.4756901430054566e-05,
+      "loss": 0.7351,
+      "step": 1746
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 4.47410023081604e-05,
+      "loss": 0.1735,
+      "step": 1747
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 4.472509772629008e-05,
+      "loss": 0.201,
+      "step": 1748
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 4.470918769033458e-05,
+      "loss": 0.397,
+      "step": 1749
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 4.469327220618685e-05,
+      "loss": 0.2414,
+      "step": 1750
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 4.467735127974192e-05,
+      "loss": 0.4881,
+      "step": 1751
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 4.466142491689677e-05,
+      "loss": 0.4123,
+      "step": 1752
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 4.464549312355045e-05,
+      "loss": 0.3493,
+      "step": 1753
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 4.462955590560398e-05,
+      "loss": 0.3314,
+      "step": 1754
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 4.4613613268960437e-05,
+      "loss": 0.4376,
+      "step": 1755
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 4.459766521952484e-05,
+      "loss": 0.4198,
+      "step": 1756
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 4.458171176320427e-05,
+      "loss": 0.1553,
+      "step": 1757
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 4.456575290590778e-05,
+      "loss": 0.2982,
+      "step": 1758
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 4.454978865354643e-05,
+      "loss": 0.504,
+      "step": 1759
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 4.4533819012033285e-05,
+      "loss": 0.5656,
+      "step": 1760
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 4.451784398728341e-05,
+      "loss": 0.2258,
+      "step": 1761
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 4.4501863585213844e-05,
+      "loss": 0.3457,
+      "step": 1762
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 4.448587781174363e-05,
+      "loss": 0.2058,
+      "step": 1763
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 4.446988667279379e-05,
+      "loss": 0.5471,
+      "step": 1764
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 4.4453890174287374e-05,
+      "loss": 0.3105,
+      "step": 1765
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 4.443788832214935e-05,
+      "loss": 0.2517,
+      "step": 1766
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 4.442188112230673e-05,
+      "loss": 0.1476,
+      "step": 1767
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 4.440586858068847e-05,
+      "loss": 0.1392,
+      "step": 1768
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 4.438985070322552e-05,
+      "loss": 0.1543,
+      "step": 1769
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 4.43738274958508e-05,
+      "loss": 0.4049,
+      "step": 1770
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 4.4357798964499224e-05,
+      "loss": 0.4838,
+      "step": 1771
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 4.434176511510763e-05,
+      "loss": 0.1179,
+      "step": 1772
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 4.4325725953614876e-05,
+      "loss": 0.3871,
+      "step": 1773
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 4.430968148596177e-05,
+      "loss": 0.2346,
+      "step": 1774
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 4.4293631718091084e-05,
+      "loss": 0.3703,
+      "step": 1775
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 4.427757665594754e-05,
+      "loss": 0.574,
+      "step": 1776
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 4.4261516305477845e-05,
+      "loss": 0.4122,
+      "step": 1777
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 4.424545067263065e-05,
+      "loss": 0.2226,
+      "step": 1778
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 4.4229379763356564e-05,
+      "loss": 0.2553,
+      "step": 1779
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 4.421330358360817e-05,
+      "loss": 0.406,
+      "step": 1780
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 4.419722213933995e-05,
+      "loss": 0.4193,
+      "step": 1781
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 4.4181135436508406e-05,
+      "loss": 0.231,
+      "step": 1782
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 4.4165043481071935e-05,
+      "loss": 0.2181,
+      "step": 1783
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 4.414894627899091e-05,
+      "loss": 0.3007,
+      "step": 1784
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 4.4132843836227615e-05,
+      "loss": 0.207,
+      "step": 1785
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 4.4116736158746316e-05,
+      "loss": 0.2042,
+      "step": 1786
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 4.4100623252513185e-05,
+      "loss": 0.624,
+      "step": 1787
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 4.408450512349635e-05,
+      "loss": 0.1797,
+      "step": 1788
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 4.406838177766585e-05,
+      "loss": 0.3221,
+      "step": 1789
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 4.4052253220993704e-05,
+      "loss": 0.1727,
+      "step": 1790
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 4.403611945945381e-05,
+      "loss": 0.5819,
+      "step": 1791
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 4.401998049902201e-05,
+      "loss": 0.4682,
+      "step": 1792
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 4.400383634567608e-05,
+      "loss": 0.2322,
+      "step": 1793
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 4.398768700539572e-05,
+      "loss": 0.2228,
+      "step": 1794
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 4.397153248416254e-05,
+      "loss": 0.5528,
+      "step": 1795
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 4.395537278796007e-05,
+      "loss": 0.2393,
+      "step": 1796
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 4.393920792277377e-05,
+      "loss": 0.1239,
+      "step": 1797
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 4.3923037894591e-05,
+      "loss": 0.2591,
+      "step": 1798
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 4.390686270940105e-05,
+      "loss": 0.397,
+      "step": 1799
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 4.38906823731951e-05,
+      "loss": 0.3071,
+      "step": 1800
+    },
+    {
+      "epoch": 1.39,
+      "eval_loss": 1.18120276927948,
+      "eval_runtime": 157.4049,
+      "eval_samples_per_second": 0.337,
+      "eval_steps_per_second": 0.172,
+      "step": 1800
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 4.3874496891966247e-05,
+      "loss": 0.295,
+      "step": 1801
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 4.385830627170949e-05,
+      "loss": 0.523,
+      "step": 1802
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 4.3842110518421744e-05,
+      "loss": 0.6303,
+      "step": 1803
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 4.38259096381018e-05,
+      "loss": 0.4021,
+      "step": 1804
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 4.380970363675038e-05,
+      "loss": 0.3006,
+      "step": 1805
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 4.3793492520370075e-05,
+      "loss": 0.5278,
+      "step": 1806
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 4.3777276294965386e-05,
+      "loss": 0.2007,
+      "step": 1807
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 4.3761054966542714e-05,
+      "loss": 0.3056,
+      "step": 1808
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 4.3744828541110334e-05,
+      "loss": 0.3093,
+      "step": 1809
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 4.37285970246784e-05,
+      "loss": 0.2004,
+      "step": 1810
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 4.371236042325898e-05,
+      "loss": 0.5074,
+      "step": 1811
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 4.3696118742866e-05,
+      "loss": 0.2254,
+      "step": 1812
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 4.367987198951529e-05,
+      "loss": 0.506,
+      "step": 1813
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 4.366362016922453e-05,
+      "loss": 0.7358,
+      "step": 1814
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 4.364736328801331e-05,
+      "loss": 0.1777,
+      "step": 1815
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 4.3631101351903076e-05,
+      "loss": 0.6243,
+      "step": 1816
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 4.361483436691714e-05,
+      "loss": 0.5977,
+      "step": 1817
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 4.3598562339080695e-05,
+      "loss": 0.2168,
+      "step": 1818
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 4.3582285274420816e-05,
+      "loss": 0.2006,
+      "step": 1819
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 4.356600317896641e-05,
+      "loss": 0.1944,
+      "step": 1820
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 4.354971605874826e-05,
+      "loss": 0.3807,
+      "step": 1821
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 4.353342391979902e-05,
+      "loss": 0.1353,
+      "step": 1822
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 4.351712676815321e-05,
+      "loss": 0.3123,
+      "step": 1823
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 4.3500824609847175e-05,
+      "loss": 0.3863,
+      "step": 1824
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 4.3484517450919155e-05,
+      "loss": 0.2327,
+      "step": 1825
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 4.3468205297409204e-05,
+      "loss": 0.4209,
+      "step": 1826
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 4.345188815535925e-05,
+      "loss": 0.1391,
+      "step": 1827
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 4.3435566030813055e-05,
+      "loss": 0.1891,
+      "step": 1828
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 4.341923892981624e-05,
+      "loss": 0.3643,
+      "step": 1829
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 4.340290685841626e-05,
+      "loss": 0.183,
+      "step": 1830
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 4.338656982266241e-05,
+      "loss": 0.1347,
+      "step": 1831
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 4.337022782860584e-05,
+      "loss": 0.3482,
+      "step": 1832
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 4.3353880882299507e-05,
+      "loss": 0.2857,
+      "step": 1833
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 4.333752898979822e-05,
+      "loss": 1.0175,
+      "step": 1834
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 4.332117215715863e-05,
+      "loss": 0.2115,
+      "step": 1835
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 4.3304810390439213e-05,
+      "loss": 0.2667,
+      "step": 1836
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 4.328844369570025e-05,
+      "loss": 0.4096,
+      "step": 1837
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 4.327207207900386e-05,
+      "loss": 0.1772,
+      "step": 1838
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 4.3255695546414e-05,
+      "loss": 0.4444,
+      "step": 1839
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 4.323931410399643e-05,
+      "loss": 0.2175,
+      "step": 1840
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 4.322292775781874e-05,
+      "loss": 0.3304,
+      "step": 1841
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 4.320653651395033e-05,
+      "loss": 0.1643,
+      "step": 1842
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 4.319014037846242e-05,
+      "loss": 0.277,
+      "step": 1843
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 4.317373935742803e-05,
+      "loss": 0.3795,
+      "step": 1844
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 4.315733345692201e-05,
+      "loss": 0.2385,
+      "step": 1845
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 4.3140922683020974e-05,
+      "loss": 0.4048,
+      "step": 1846
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 4.3124507041803396e-05,
+      "loss": 0.203,
+      "step": 1847
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 4.3108086539349533e-05,
+      "loss": 0.7223,
+      "step": 1848
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 4.3091661181741415e-05,
+      "loss": 0.4489,
+      "step": 1849
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 4.307523097506291e-05,
+      "loss": 0.512,
+      "step": 1850
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 4.3058795925399665e-05,
+      "loss": 0.6486,
+      "step": 1851
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 4.3042356038839116e-05,
+      "loss": 0.324,
+      "step": 1852
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 4.302591132147049e-05,
+      "loss": 0.1855,
+      "step": 1853
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 4.300946177938481e-05,
+      "loss": 0.6003,
+      "step": 1854
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 4.29930074186749e-05,
+      "loss": 0.3603,
+      "step": 1855
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 4.2976548245435334e-05,
+      "loss": 0.3355,
+      "step": 1856
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 4.29600842657625e-05,
+      "loss": 0.413,
+      "step": 1857
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 4.294361548575456e-05,
+      "loss": 0.4099,
+      "step": 1858
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 4.292714191151142e-05,
+      "loss": 0.1893,
+      "step": 1859
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 4.291066354913482e-05,
+      "loss": 0.2177,
+      "step": 1860
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 4.2894180404728245e-05,
+      "loss": 0.1336,
+      "step": 1861
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 4.287769248439692e-05,
+      "loss": 0.3189,
+      "step": 1862
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 4.2861199794247886e-05,
+      "loss": 0.2591,
+      "step": 1863
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 4.2844702340389946e-05,
+      "loss": 0.4344,
+      "step": 1864
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 4.282820012893364e-05,
+      "loss": 0.1444,
+      "step": 1865
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 4.281169316599128e-05,
+      "loss": 0.4593,
+      "step": 1866
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 4.2795181457676954e-05,
+      "loss": 0.2353,
+      "step": 1867
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 4.27786650101065e-05,
+      "loss": 0.1674,
+      "step": 1868
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 4.27621438293975e-05,
+      "loss": 0.4616,
+      "step": 1869
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 4.2745617921669295e-05,
+      "loss": 0.2325,
+      "step": 1870
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 4.272908729304298e-05,
+      "loss": 0.3614,
+      "step": 1871
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 4.271255194964141e-05,
+      "loss": 0.5145,
+      "step": 1872
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 4.2696011897589155e-05,
+      "loss": 0.2016,
+      "step": 1873
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 4.267946714301257e-05,
+      "loss": 0.3335,
+      "step": 1874
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 4.266291769203971e-05,
+      "loss": 0.3706,
+      "step": 1875
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 4.26463635508004e-05,
+      "loss": 0.5925,
+      "step": 1876
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 4.262980472542619e-05,
+      "loss": 0.2543,
+      "step": 1877
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 4.2613241222050364e-05,
+      "loss": 0.3746,
+      "step": 1878
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 4.259667304680794e-05,
+      "loss": 0.1521,
+      "step": 1879
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 4.258010020583567e-05,
+      "loss": 0.2124,
+      "step": 1880
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 4.2563522705272036e-05,
+      "loss": 0.2354,
+      "step": 1881
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 4.2546940551257235e-05,
+      "loss": 0.3945,
+      "step": 1882
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 4.25303537499332e-05,
+      "loss": 0.2083,
+      "step": 1883
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 4.251376230744359e-05,
+      "loss": 0.2037,
+      "step": 1884
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 4.249716622993374e-05,
+      "loss": 0.1429,
+      "step": 1885
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 4.248056552355075e-05,
+      "loss": 0.4022,
+      "step": 1886
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 4.246396019444344e-05,
+      "loss": 0.2749,
+      "step": 1887
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 4.2447350248762294e-05,
+      "loss": 0.2107,
+      "step": 1888
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 4.243073569265954e-05,
+      "loss": 0.3996,
+      "step": 1889
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 4.241411653228913e-05,
+      "loss": 0.3475,
+      "step": 1890
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 4.2397492773806664e-05,
+      "loss": 0.1442,
+      "step": 1891
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 4.23808644233695e-05,
+      "loss": 0.3095,
+      "step": 1892
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 4.236423148713667e-05,
+      "loss": 0.2491,
+      "step": 1893
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 4.234759397126891e-05,
+      "loss": 0.2294,
+      "step": 1894
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 4.2330951881928656e-05,
+      "loss": 0.439,
+      "step": 1895
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 4.231430522528003e-05,
+      "loss": 0.5024,
+      "step": 1896
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 4.229765400748886e-05,
+      "loss": 0.1899,
+      "step": 1897
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 4.228099823472265e-05,
+      "loss": 0.3486,
+      "step": 1898
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 4.226433791315059e-05,
+      "loss": 0.2794,
+      "step": 1899
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 4.224767304894356e-05,
+      "loss": 0.166,
+      "step": 1900
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 4.2231003648274136e-05,
+      "loss": 0.3149,
+      "step": 1901
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 4.2214329717316546e-05,
+      "loss": 0.2402,
+      "step": 1902
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 4.219765126224672e-05,
+      "loss": 0.1665,
+      "step": 1903
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 4.218096828924224e-05,
+      "loss": 0.1555,
+      "step": 1904
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 4.2164280804482394e-05,
+      "loss": 0.1467,
+      "step": 1905
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 4.214758881414812e-05,
+      "loss": 0.4955,
+      "step": 1906
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 4.2130892324422026e-05,
+      "loss": 0.2705,
+      "step": 1907
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 4.211419134148838e-05,
+      "loss": 0.3119,
+      "step": 1908
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 4.209748587153313e-05,
+      "loss": 0.3765,
+      "step": 1909
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 4.2080775920743886e-05,
+      "loss": 0.6719,
+      "step": 1910
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 4.2064061495309905e-05,
+      "loss": 0.1419,
+      "step": 1911
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 4.20473426014221e-05,
+      "loss": 0.3556,
+      "step": 1912
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 4.2030619245273056e-05,
+      "loss": 0.1492,
+      "step": 1913
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 4.201389143305699e-05,
+      "loss": 0.4126,
+      "step": 1914
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 4.1997159170969785e-05,
+      "loss": 0.3155,
+      "step": 1915
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 4.1980422465208976e-05,
+      "loss": 0.4496,
+      "step": 1916
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 4.196368132197373e-05,
+      "loss": 0.1601,
+      "step": 1917
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 4.194693574746485e-05,
+      "loss": 0.4014,
+      "step": 1918
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 4.193018574788481e-05,
+      "loss": 0.3759,
+      "step": 1919
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 4.1913431329437706e-05,
+      "loss": 0.19,
+      "step": 1920
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 4.1896672498329254e-05,
+      "loss": 0.4874,
+      "step": 1921
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 4.187990926076684e-05,
+      "loss": 0.1493,
+      "step": 1922
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 4.186314162295946e-05,
+      "loss": 0.1418,
+      "step": 1923
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 4.184636959111774e-05,
+      "loss": 0.4755,
+      "step": 1924
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 4.1829593171453933e-05,
+      "loss": 0.1965,
+      "step": 1925
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 4.181281237018193e-05,
+      "loss": 0.3353,
+      "step": 1926
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 4.179602719351723e-05,
+      "loss": 0.3211,
+      "step": 1927
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 4.1779237647676956e-05,
+      "loss": 0.3391,
+      "step": 1928
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 4.176244373887987e-05,
+      "loss": 0.4269,
+      "step": 1929
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 4.1745645473346296e-05,
+      "loss": 0.1034,
+      "step": 1930
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 4.172884285729825e-05,
+      "loss": 0.3079,
+      "step": 1931
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 4.1712035896959284e-05,
+      "loss": 0.4713,
+      "step": 1932
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 4.169522459855461e-05,
+      "loss": 0.1563,
+      "step": 1933
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 4.167840896831103e-05,
+      "loss": 0.2403,
+      "step": 1934
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 4.166158901245694e-05,
+      "loss": 0.1806,
+      "step": 1935
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 4.164476473722235e-05,
+      "loss": 0.459,
+      "step": 1936
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 4.162793614883887e-05,
+      "loss": 0.2195,
+      "step": 1937
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 4.16111032535397e-05,
+      "loss": 0.1095,
+      "step": 1938
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 4.1594266057559646e-05,
+      "loss": 0.2661,
+      "step": 1939
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 4.15774245671351e-05,
+      "loss": 0.2225,
+      "step": 1940
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 4.156057878850405e-05,
+      "loss": 0.1837,
+      "step": 1941
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 4.154372872790607e-05,
+      "loss": 0.8236,
+      "step": 1942
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 4.15268743915823e-05,
+      "loss": 0.1597,
+      "step": 1943
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 4.15100157857755e-05,
+      "loss": 0.3873,
+      "step": 1944
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 4.1493152916729996e-05,
+      "loss": 0.1992,
+      "step": 1945
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 4.147628579069168e-05,
+      "loss": 0.2095,
+      "step": 1946
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 4.145941441390802e-05,
+      "loss": 0.3533,
+      "step": 1947
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 4.1442538792628095e-05,
+      "loss": 0.331,
+      "step": 1948
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 4.142565893310252e-05,
+      "loss": 0.3494,
+      "step": 1949
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 4.140877484158348e-05,
+      "loss": 0.127,
+      "step": 1950
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 4.139188652432476e-05,
+      "loss": 0.1374,
+      "step": 1951
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 4.137499398758167e-05,
+      "loss": 0.1769,
+      "step": 1952
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 4.1358097237611103e-05,
+      "loss": 0.3289,
+      "step": 1953
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 4.134119628067152e-05,
+      "loss": 0.7231,
+      "step": 1954
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 4.1324291123022925e-05,
+      "loss": 0.1993,
+      "step": 1955
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 4.1307381770926886e-05,
+      "loss": 0.2314,
+      "step": 1956
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 4.129046823064652e-05,
+      "loss": 0.1558,
+      "step": 1957
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 4.1273550508446516e-05,
+      "loss": 0.6187,
+      "step": 1958
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 4.125662861059306e-05,
+      "loss": 0.1201,
+      "step": 1959
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.123970254335393e-05,
+      "loss": 0.4828,
+      "step": 1960
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.122277231299846e-05,
+      "loss": 0.3971,
+      "step": 1961
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.120583792579748e-05,
+      "loss": 0.2849,
+      "step": 1962
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.1188899388023386e-05,
+      "loss": 0.2311,
+      "step": 1963
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.117195670595011e-05,
+      "loss": 0.368,
+      "step": 1964
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.115500988585311e-05,
+      "loss": 0.1441,
+      "step": 1965
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.113805893400939e-05,
+      "loss": 0.2167,
+      "step": 1966
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.1121103856697475e-05,
+      "loss": 0.1935,
+      "step": 1967
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.110414466019744e-05,
+      "loss": 0.5194,
+      "step": 1968
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.108718135079084e-05,
+      "loss": 0.3225,
+      "step": 1969
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.107021393476078e-05,
+      "loss": 0.1997,
+      "step": 1970
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.1053242418391894e-05,
+      "loss": 0.6301,
+      "step": 1971
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.103626680797034e-05,
+      "loss": 0.3273,
+      "step": 1972
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.101928710978377e-05,
+      "loss": 0.473,
+      "step": 1973
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.100230333012135e-05,
+      "loss": 0.5825,
+      "step": 1974
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.0985315475273775e-05,
+      "loss": 0.7091,
+      "step": 1975
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.096832355153324e-05,
+      "loss": 0.1998,
+      "step": 1976
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.095132756519344e-05,
+      "loss": 0.6227,
+      "step": 1977
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.0934327522549614e-05,
+      "loss": 0.2694,
+      "step": 1978
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.091732342989843e-05,
+      "loss": 0.3005,
+      "step": 1979
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.090031529353813e-05,
+      "loss": 0.3766,
+      "step": 1980
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.0883303119768416e-05,
+      "loss": 0.2048,
+      "step": 1981
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.086628691489048e-05,
+      "loss": 0.2712,
+      "step": 1982
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.084926668520703e-05,
+      "loss": 0.4379,
+      "step": 1983
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.0832242437022245e-05,
+      "loss": 0.4311,
+      "step": 1984
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.081521417664182e-05,
+      "loss": 0.2808,
+      "step": 1985
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.07981819103729e-05,
+      "loss": 0.1928,
+      "step": 1986
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.0781145644524125e-05,
+      "loss": 0.6029,
+      "step": 1987
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.076410538540564e-05,
+      "loss": 0.3585,
+      "step": 1988
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.0747061139329054e-05,
+      "loss": 0.1465,
+      "step": 1989
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.073001291260744e-05,
+      "loss": 0.3593,
+      "step": 1990
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.071296071155536e-05,
+      "loss": 0.4165,
+      "step": 1991
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.0695904542488826e-05,
+      "loss": 0.1568,
+      "step": 1992
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.067884441172537e-05,
+      "loss": 0.6753,
+      "step": 1993
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.066178032558393e-05,
+      "loss": 0.4629,
+      "step": 1994
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.064471229038496e-05,
+      "loss": 0.379,
+      "step": 1995
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.062764031245033e-05,
+      "loss": 0.2073,
+      "step": 1996
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.0610564398103434e-05,
+      "loss": 0.3068,
+      "step": 1997
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.059348455366904e-05,
+      "loss": 0.1872,
+      "step": 1998
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.0576400785473445e-05,
+      "loss": 0.2492,
+      "step": 1999
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.055931309984436e-05,
+      "loss": 0.3447,
+      "step": 2000
+    },
+    {
+      "epoch": 1.55,
+      "eval_loss": 1.1841756105422974,
+      "eval_runtime": 157.173,
+      "eval_samples_per_second": 0.337,
+      "eval_steps_per_second": 0.172,
+      "step": 2000
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.054222150311097e-05,
+      "loss": 0.2177,
+      "step": 2001
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.052512600160389e-05,
+      "loss": 0.5768,
+      "step": 2002
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.050802660165518e-05,
+      "loss": 0.3635,
+      "step": 2003
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.0490923309598354e-05,
+      "loss": 0.395,
+      "step": 2004
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.047381613176838e-05,
+      "loss": 0.3739,
+      "step": 2005
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.045670507450164e-05,
+      "loss": 0.2242,
+      "step": 2006
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.043959014413597e-05,
+      "loss": 0.2321,
+      "step": 2007
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.042247134701063e-05,
+      "loss": 0.207,
+      "step": 2008
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.040534868946631e-05,
+      "loss": 0.1853,
+      "step": 2009
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.038822217784515e-05,
+      "loss": 0.4297,
+      "step": 2010
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.037109181849069e-05,
+      "loss": 0.16,
+      "step": 2011
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.035395761774792e-05,
+      "loss": 0.1413,
+      "step": 2012
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.0336819581963245e-05,
+      "loss": 0.4214,
+      "step": 2013
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.031967771748447e-05,
+      "loss": 0.1725,
+      "step": 2014
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.030253203066086e-05,
+      "loss": 0.1313,
+      "step": 2015
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.028538252784305e-05,
+      "loss": 0.2641,
+      "step": 2016
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.026822921538312e-05,
+      "loss": 0.3787,
+      "step": 2017
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.025107209963455e-05,
+      "loss": 0.1377,
+      "step": 2018
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.0233911186952224e-05,
+      "loss": 0.3715,
+      "step": 2019
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.021674648369245e-05,
+      "loss": 0.51,
+      "step": 2020
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.019957799621291e-05,
+      "loss": 0.3668,
+      "step": 2021
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.0182405730872716e-05,
+      "loss": 0.2586,
+      "step": 2022
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.016522969403238e-05,
+      "loss": 0.3034,
+      "step": 2023
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.014804989205378e-05,
+      "loss": 0.4044,
+      "step": 2024
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.0130866331300226e-05,
+      "loss": 0.1319,
+      "step": 2025
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.0113679018136396e-05,
+      "loss": 0.2254,
+      "step": 2026
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.009648795892836e-05,
+      "loss": 0.386,
+      "step": 2027
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.007929316004358e-05,
+      "loss": 0.363,
+      "step": 2028
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.006209462785091e-05,
+      "loss": 0.1847,
+      "step": 2029
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.0044892368720576e-05,
+      "loss": 0.4387,
+      "step": 2030
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.0027686389024185e-05,
+      "loss": 0.2231,
+      "step": 2031
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.001047669513473e-05,
+      "loss": 0.2416,
+      "step": 2032
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 3.999326329342658e-05,
+      "loss": 0.4893,
+      "step": 2033
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 3.9976046190275466e-05,
+      "loss": 0.3525,
+      "step": 2034
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 3.99588253920585e-05,
+      "loss": 0.3298,
+      "step": 2035
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 3.994160090515415e-05,
+      "loss": 0.1395,
+      "step": 2036
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 3.9924372735942256e-05,
+      "loss": 0.1456,
+      "step": 2037
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 3.990714089080405e-05,
+      "loss": 0.394,
+      "step": 2038
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 3.98899053761221e-05,
+      "loss": 0.0969,
+      "step": 2039
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 3.98726661982803e-05,
+      "loss": 0.3168,
+      "step": 2040
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 3.985542336366396e-05,
+      "loss": 0.1366,
+      "step": 2041
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 3.983817687865973e-05,
+      "loss": 0.1862,
+      "step": 2042
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 3.982092674965557e-05,
+      "loss": 0.1695,
+      "step": 2043
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 3.980367298304084e-05,
+      "loss": 0.1533,
+      "step": 2044
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 3.978641558520623e-05,
+      "loss": 0.4591,
+      "step": 2045
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 3.976915456254377e-05,
+      "loss": 0.1496,
+      "step": 2046
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 3.975188992144684e-05,
+      "loss": 0.3954,
+      "step": 2047
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 3.973462166831014e-05,
+      "loss": 0.179,
+      "step": 2048
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 3.971734980952975e-05,
+      "loss": 0.1192,
+      "step": 2049
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 3.9700074351503045e-05,
+      "loss": 0.2001,
+      "step": 2050
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 3.968279530062874e-05,
+      "loss": 0.3705,
+      "step": 2051
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 3.96655126633069e-05,
+      "loss": 0.4331,
+      "step": 2052
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 3.9648226445938915e-05,
+      "loss": 0.1758,
+      "step": 2053
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 3.963093665492747e-05,
+      "loss": 0.1874,
+      "step": 2054
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 3.9613643296676624e-05,
+      "loss": 0.2735,
+      "step": 2055
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 3.959634637759171e-05,
+      "loss": 0.1956,
+      "step": 2056
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 3.9579045904079395e-05,
+      "loss": 0.4191,
+      "step": 2057
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 3.956174188254769e-05,
+      "loss": 0.1836,
+      "step": 2058
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 3.9544434319405886e-05,
+      "loss": 0.3659,
+      "step": 2059
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 3.9527123221064594e-05,
+      "loss": 0.2379,
+      "step": 2060
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 3.950980859393574e-05,
+      "loss": 0.1924,
+      "step": 2061
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 3.9492490444432564e-05,
+      "loss": 0.1224,
+      "step": 2062
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 3.947516877896959e-05,
+      "loss": 0.4754,
+      "step": 2063
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 3.945784360396265e-05,
+      "loss": 0.1316,
+      "step": 2064
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 3.944051492582891e-05,
+      "loss": 0.2493,
+      "step": 2065
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 3.9423182750986766e-05,
+      "loss": 0.2578,
+      "step": 2066
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 3.9405847085855974e-05,
+      "loss": 0.4146,
+      "step": 2067
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 3.938850793685756e-05,
+      "loss": 0.4159,
+      "step": 2068
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 3.9371165310413836e-05,
+      "loss": 0.6225,
+      "step": 2069
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 3.9353819212948374e-05,
+      "loss": 0.1764,
+      "step": 2070
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 3.93364696508861e-05,
+      "loss": 0.164,
+      "step": 2071
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 3.931911663065317e-05,
+      "loss": 0.1444,
+      "step": 2072
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 3.930176015867703e-05,
+      "loss": 0.1639,
+      "step": 2073
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 3.928440024138641e-05,
+      "loss": 0.5127,
+      "step": 2074
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 3.926703688521134e-05,
+      "loss": 0.2895,
+      "step": 2075
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 3.924967009658306e-05,
+      "loss": 0.1402,
+      "step": 2076
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 3.923229988193415e-05,
+      "loss": 0.5778,
+      "step": 2077
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 3.921492624769843e-05,
+      "loss": 0.4433,
+      "step": 2078
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 3.9197549200310965e-05,
+      "loss": 0.13,
+      "step": 2079
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 3.918016874620812e-05,
+      "loss": 0.3228,
+      "step": 2080
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 3.916278489182751e-05,
+      "loss": 0.2094,
+      "step": 2081
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 3.9145397643608015e-05,
+      "loss": 0.3616,
+      "step": 2082
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 3.9128007007989745e-05,
+      "loss": 0.3973,
+      "step": 2083
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 3.9110612991414095e-05,
+      "loss": 0.2061,
+      "step": 2084
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 3.909321560032369e-05,
+      "loss": 0.3026,
+      "step": 2085
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 3.907581484116244e-05,
+      "loss": 0.3955,
+      "step": 2086
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 3.9058410720375446e-05,
+      "loss": 0.117,
+      "step": 2087
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 3.904100324440912e-05,
+      "loss": 0.4963,
+      "step": 2088
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 3.902359241971106e-05,
+      "loss": 0.1921,
+      "step": 2089
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 3.900617825273014e-05,
+      "loss": 0.2996,
+      "step": 2090
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 3.898876074991646e-05,
+      "loss": 0.2565,
+      "step": 2091
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 3.897133991772134e-05,
+      "loss": 0.3496,
+      "step": 2092
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 3.895391576259737e-05,
+      "loss": 0.6038,
+      "step": 2093
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 3.8936488290998335e-05,
+      "loss": 0.6207,
+      "step": 2094
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 3.891905750937927e-05,
+      "loss": 0.4831,
+      "step": 2095
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 3.890162342419643e-05,
+      "loss": 0.424,
+      "step": 2096
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 3.888418604190728e-05,
+      "loss": 0.3195,
+      "step": 2097
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 3.886674536897053e-05,
+      "loss": 0.4914,
+      "step": 2098
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 3.8849301411846084e-05,
+      "loss": 0.2128,
+      "step": 2099
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 3.8831854176995094e-05,
+      "loss": 0.1552,
+      "step": 2100
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 3.8814403670879903e-05,
+      "loss": 0.4547,
+      "step": 2101
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 3.8796949899964065e-05,
+      "loss": 0.2186,
+      "step": 2102
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 3.877949287071235e-05,
+      "loss": 0.17,
+      "step": 2103
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 3.876203258959074e-05,
+      "loss": 0.5685,
+      "step": 2104
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 3.8744569063066404e-05,
+      "loss": 0.3218,
+      "step": 2105
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 3.872710229760774e-05,
+      "loss": 0.2715,
+      "step": 2106
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 3.8709632299684316e-05,
+      "loss": 0.3199,
+      "step": 2107
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 3.869215907576692e-05,
+      "loss": 0.7672,
+      "step": 2108
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 3.867468263232751e-05,
+      "loss": 0.4878,
+      "step": 2109
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 3.8657202975839274e-05,
+      "loss": 0.138,
+      "step": 2110
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 3.8639720112776565e-05,
+      "loss": 0.515,
+      "step": 2111
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 3.862223404961491e-05,
+      "loss": 0.1967,
+      "step": 2112
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 3.860474479283106e-05,
+      "loss": 0.2171,
+      "step": 2113
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 3.858725234890292e-05,
+      "loss": 0.3763,
+      "step": 2114
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 3.856975672430957e-05,
+      "loss": 0.1229,
+      "step": 2115
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 3.8552257925531295e-05,
+      "loss": 0.3895,
+      "step": 2116
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 3.853475595904955e-05,
+      "loss": 0.3245,
+      "step": 2117
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 3.851725083134694e-05,
+      "loss": 0.2133,
+      "step": 2118
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 3.849974254890726e-05,
+      "loss": 0.584,
+      "step": 2119
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 3.848223111821547e-05,
+      "loss": 0.4731,
+      "step": 2120
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 3.8464716545757695e-05,
+      "loss": 0.2326,
+      "step": 2121
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 3.844719883802123e-05,
+      "loss": 0.1565,
+      "step": 2122
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 3.842967800149452e-05,
+      "loss": 0.4235,
+      "step": 2123
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 3.8412154042667176e-05,
+      "loss": 0.2952,
+      "step": 2124
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 3.839462696802996e-05,
+      "loss": 0.6539,
+      "step": 2125
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 3.8377096784074795e-05,
+      "loss": 0.5554,
+      "step": 2126
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 3.835956349729477e-05,
+      "loss": 0.4983,
+      "step": 2127
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 3.8342027114184074e-05,
+      "loss": 0.3273,
+      "step": 2128
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 3.83244876412381e-05,
+      "loss": 0.3566,
+      "step": 2129
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 3.8306945084953346e-05,
+      "loss": 0.424,
+      "step": 2130
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 3.828939945182747e-05,
+      "loss": 0.6855,
+      "step": 2131
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 3.827185074835927e-05,
+      "loss": 0.1756,
+      "step": 2132
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 3.825429898104868e-05,
+      "loss": 0.1198,
+      "step": 2133
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 3.823674415639676e-05,
+      "loss": 0.1774,
+      "step": 2134
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 3.82191862809057e-05,
+      "loss": 0.3572,
+      "step": 2135
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 3.820162536107884e-05,
+      "loss": 0.6206,
+      "step": 2136
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 3.818406140342065e-05,
+      "loss": 0.3753,
+      "step": 2137
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 3.816649441443668e-05,
+      "loss": 0.2137,
+      "step": 2138
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 3.814892440063365e-05,
+      "loss": 0.3397,
+      "step": 2139
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 3.8131351368519395e-05,
+      "loss": 0.1583,
+      "step": 2140
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 3.811377532460284e-05,
+      "loss": 0.1557,
+      "step": 2141
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 3.809619627539405e-05,
+      "loss": 0.3181,
+      "step": 2142
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 3.807861422740421e-05,
+      "loss": 0.1928,
+      "step": 2143
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 3.806102918714557e-05,
+      "loss": 0.4084,
+      "step": 2144
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 3.8043441161131546e-05,
+      "loss": 0.1933,
+      "step": 2145
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 3.8025850155876635e-05,
+      "loss": 0.4815,
+      "step": 2146
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 3.8008256177896414e-05,
+      "loss": 0.377,
+      "step": 2147
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 3.79906592337076e-05,
+      "loss": 0.3922,
+      "step": 2148
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 3.7973059329827994e-05,
+      "loss": 0.182,
+      "step": 2149
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 3.795545647277649e-05,
+      "loss": 0.2391,
+      "step": 2150
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 3.793785066907308e-05,
+      "loss": 0.5367,
+      "step": 2151
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 3.7920241925238824e-05,
+      "loss": 0.1479,
+      "step": 2152
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 3.790263024779592e-05,
+      "loss": 0.6408,
+      "step": 2153
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 3.7885015643267606e-05,
+      "loss": 0.5057,
+      "step": 2154
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 3.786739811817823e-05,
+      "loss": 0.387,
+      "step": 2155
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 3.7849777679053224e-05,
+      "loss": 0.1078,
+      "step": 2156
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 3.783215433241907e-05,
+      "loss": 0.6569,
+      "step": 2157
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 3.781452808480336e-05,
+      "loss": 0.3089,
+      "step": 2158
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 3.779689894273474e-05,
+      "loss": 0.1777,
+      "step": 2159
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 3.777926691274294e-05,
+      "loss": 0.2317,
+      "step": 2160
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 3.776163200135875e-05,
+      "loss": 0.1744,
+      "step": 2161
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 3.774399421511405e-05,
+      "loss": 0.3447,
+      "step": 2162
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 3.772635356054174e-05,
+      "loss": 0.1892,
+      "step": 2163
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 3.770871004417582e-05,
+      "loss": 0.2156,
+      "step": 2164
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 3.769106367255135e-05,
+      "loss": 0.1536,
+      "step": 2165
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 3.7673414452204425e-05,
+      "loss": 0.3101,
+      "step": 2166
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.765576238967221e-05,
+      "loss": 0.3849,
+      "step": 2167
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.763810749149292e-05,
+      "loss": 1.0938,
+      "step": 2168
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.762044976420583e-05,
+      "loss": 0.348,
+      "step": 2169
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.760278921435123e-05,
+      "loss": 0.4566,
+      "step": 2170
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.758512584847049e-05,
+      "loss": 0.1544,
+      "step": 2171
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.756745967310603e-05,
+      "loss": 0.1338,
+      "step": 2172
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.754979069480126e-05,
+      "loss": 0.7908,
+      "step": 2173
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.753211892010069e-05,
+      "loss": 0.2042,
+      "step": 2174
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.751444435554982e-05,
+      "loss": 0.1461,
+      "step": 2175
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.74967670076952e-05,
+      "loss": 0.304,
+      "step": 2176
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.747908688308441e-05,
+      "loss": 0.573,
+      "step": 2177
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.746140398826606e-05,
+      "loss": 0.4387,
+      "step": 2178
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 3.744371832978979e-05,
+      "loss": 0.2072,
+      "step": 2179
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.7426029914206255e-05,
+      "loss": 0.1673,
+      "step": 2180
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.740833874806713e-05,
+      "loss": 0.1986,
+      "step": 2181
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.739064483792513e-05,
+      "loss": 0.1542,
+      "step": 2182
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.737294819033395e-05,
+      "loss": 0.3245,
+      "step": 2183
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.7355248811848326e-05,
+      "loss": 0.1787,
+      "step": 2184
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.7337546709024004e-05,
+      "loss": 0.5739,
+      "step": 2185
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.7319841888417724e-05,
+      "loss": 0.3069,
+      "step": 2186
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.730213435658725e-05,
+      "loss": 0.2291,
+      "step": 2187
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.728442412009134e-05,
+      "loss": 0.2755,
+      "step": 2188
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.726671118548975e-05,
+      "loss": 0.3493,
+      "step": 2189
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.724899555934324e-05,
+      "loss": 0.7381,
+      "step": 2190
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.723127724821361e-05,
+      "loss": 0.2973,
+      "step": 2191
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.7213556258663545e-05,
+      "loss": 0.4242,
+      "step": 2192
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.719583259725683e-05,
+      "loss": 0.1819,
+      "step": 2193
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.7178106270558206e-05,
+      "loss": 0.1741,
+      "step": 2194
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.7160377285133375e-05,
+      "loss": 0.1473,
+      "step": 2195
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.714264564754905e-05,
+      "loss": 0.5197,
+      "step": 2196
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.712491136437292e-05,
+      "loss": 0.3789,
+      "step": 2197
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.710717444217365e-05,
+      "loss": 0.3907,
+      "step": 2198
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.70894348875209e-05,
+      "loss": 0.4604,
+      "step": 2199
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.7071692706985276e-05,
+      "loss": 0.212,
+      "step": 2200
+    },
+    {
+      "epoch": 1.7,
+      "eval_loss": 1.1778496503829956,
+      "eval_runtime": 157.3588,
+      "eval_samples_per_second": 0.337,
+      "eval_steps_per_second": 0.172,
+      "step": 2200
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.705394790713838e-05,
+      "loss": 0.2984,
+      "step": 2201
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.703620049455278e-05,
+      "loss": 0.1587,
+      "step": 2202
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.701845047580199e-05,
+      "loss": 0.2078,
+      "step": 2203
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.700069785746053e-05,
+      "loss": 0.8882,
+      "step": 2204
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 3.6982942646103847e-05,
+      "loss": 0.4482,
+      "step": 2205
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 3.696518484830836e-05,
+      "loss": 0.1634,
+      "step": 2206
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 3.694742447065145e-05,
+      "loss": 0.1586,
+      "step": 2207
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 3.6929661519711464e-05,
+      "loss": 0.1907,
+      "step": 2208
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 3.691189600206767e-05,
+      "loss": 0.7243,
+      "step": 2209
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 3.689412792430032e-05,
+      "loss": 0.1739,
+      "step": 2210
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 3.6876357292990594e-05,
+      "loss": 0.3183,
+      "step": 2211
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 3.685858411472062e-05,
+      "loss": 0.182,
+      "step": 2212
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 3.6840808396073484e-05,
+      "loss": 0.1557,
+      "step": 2213
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 3.68230301436332e-05,
+      "loss": 0.2972,
+      "step": 2214
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 3.6805249363984706e-05,
+      "loss": 0.2827,
+      "step": 2215
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 3.6787466063713906e-05,
+      "loss": 0.1785,
+      "step": 2216
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 3.676968024940764e-05,
+      "loss": 0.5056,
+      "step": 2217
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 3.6751891927653636e-05,
+      "loss": 0.2025,
+      "step": 2218
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 3.6734101105040584e-05,
+      "loss": 0.272,
+      "step": 2219
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 3.671630778815811e-05,
+      "loss": 0.1595,
+      "step": 2220
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 3.669851198359675e-05,
+      "loss": 0.5141,
+      "step": 2221
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 3.668071369794793e-05,
+      "loss": 0.7421,
+      "step": 2222
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 3.666291293780406e-05,
+      "loss": 0.1346,
+      "step": 2223
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 3.6645109709758414e-05,
+      "loss": 0.193,
+      "step": 2224
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 3.662730402040519e-05,
+      "loss": 0.1439,
+      "step": 2225
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 3.6609495876339524e-05,
+      "loss": 0.3187,
+      "step": 2226
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 3.659168528415744e-05,
+      "loss": 0.3177,
+      "step": 2227
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 3.6573872250455854e-05,
+      "loss": 0.3774,
+      "step": 2228
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 3.655605678183262e-05,
+      "loss": 0.4594,
+      "step": 2229
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 3.653823888488646e-05,
+      "loss": 0.1277,
+      "step": 2230
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 3.652041856621705e-05,
+      "loss": 0.3645,
+      "step": 2231
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.6502595832424886e-05,
+      "loss": 0.1521,
+      "step": 2232
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.648477069011142e-05,
+      "loss": 0.1468,
+      "step": 2233
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.6466943145878954e-05,
+      "loss": 0.4577,
+      "step": 2234
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.6449113206330726e-05,
+      "loss": 0.1833,
+      "step": 2235
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.643128087807081e-05,
+      "loss": 0.2714,
+      "step": 2236
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.641344616770422e-05,
+      "loss": 0.125,
+      "step": 2237
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.6395609081836805e-05,
+      "loss": 0.3501,
+      "step": 2238
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.637776962707531e-05,
+      "loss": 0.1996,
+      "step": 2239
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.635992781002736e-05,
+      "loss": 0.1705,
+      "step": 2240
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.6342083637301456e-05,
+      "loss": 0.2467,
+      "step": 2241
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.6324237115506964e-05,
+      "loss": 0.4677,
+      "step": 2242
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.630638825125414e-05,
+      "loss": 0.2725,
+      "step": 2243
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 3.628853705115408e-05,
+      "loss": 0.3334,
+      "step": 2244
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.6270683521818745e-05,
+      "loss": 0.199,
+      "step": 2245
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.6252827669861004e-05,
+      "loss": 0.518,
+      "step": 2246
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.6234969501894536e-05,
+      "loss": 0.2116,
+      "step": 2247
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.62171090245339e-05,
+      "loss": 0.2162,
+      "step": 2248
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.61992462443945e-05,
+      "loss": 0.2485,
+      "step": 2249
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.618138116809262e-05,
+      "loss": 0.2257,
+      "step": 2250
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.616351380224534e-05,
+      "loss": 0.1203,
+      "step": 2251
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.6145644153470636e-05,
+      "loss": 0.1283,
+      "step": 2252
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.612777222838733e-05,
+      "loss": 0.3596,
+      "step": 2253
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.610989803361506e-05,
+      "loss": 0.3907,
+      "step": 2254
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.6092021575774315e-05,
+      "loss": 0.1719,
+      "step": 2255
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.607414286148644e-05,
+      "loss": 0.4005,
+      "step": 2256
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 3.6056261897373575e-05,
+      "loss": 0.2371,
+      "step": 2257
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 3.603837869005874e-05,
+      "loss": 0.4323,
+      "step": 2258
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 3.602049324616576e-05,
+      "loss": 0.1565,
+      "step": 2259
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 3.6002605572319285e-05,
+      "loss": 0.1804,
+      "step": 2260
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 3.5984715675144804e-05,
+      "loss": 0.165,
+      "step": 2261
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 3.596682356126863e-05,
+      "loss": 0.5502,
+      "step": 2262
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 3.594892923731788e-05,
+      "loss": 0.9075,
+      "step": 2263
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 3.5931032709920505e-05,
+      "loss": 0.1117,
+      "step": 2264
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 3.591313398570527e-05,
+      "loss": 0.2498,
+      "step": 2265
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 3.5895233071301755e-05,
+      "loss": 0.4704,
+      "step": 2266
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 3.587732997334035e-05,
+      "loss": 0.291,
+      "step": 2267
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 3.585942469845223e-05,
+      "loss": 0.4279,
+      "step": 2268
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 3.584151725326942e-05,
+      "loss": 0.234,
+      "step": 2269
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 3.582360764442473e-05,
+      "loss": 0.4396,
+      "step": 2270
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 3.580569587855175e-05,
+      "loss": 0.1898,
+      "step": 2271
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 3.578778196228491e-05,
+      "loss": 0.3028,
+      "step": 2272
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 3.5769865902259396e-05,
+      "loss": 0.8363,
+      "step": 2273
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 3.57519477051112e-05,
+      "loss": 0.1604,
+      "step": 2274
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 3.573402737747714e-05,
+      "loss": 0.5732,
+      "step": 2275
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 3.571610492599477e-05,
+      "loss": 0.4878,
+      "step": 2276
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 3.569818035730246e-05,
+      "loss": 0.112,
+      "step": 2277
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 3.5680253678039353e-05,
+      "loss": 0.2871,
+      "step": 2278
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 3.566232489484541e-05,
+      "loss": 0.2359,
+      "step": 2279
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 3.564439401436131e-05,
+      "loss": 0.2761,
+      "step": 2280
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 3.562646104322855e-05,
+      "loss": 0.7051,
+      "step": 2281
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 3.56085259880894e-05,
+      "loss": 0.1228,
+      "step": 2282
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 3.5590588855586874e-05,
+      "loss": 0.3286,
+      "step": 2283
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 3.55726496523648e-05,
+      "loss": 0.164,
+      "step": 2284
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 3.555470838506772e-05,
+      "loss": 0.3124,
+      "step": 2285
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 3.553676506034099e-05,
+      "loss": 0.3003,
+      "step": 2286
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 3.5518819684830707e-05,
+      "loss": 0.179,
+      "step": 2287
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 3.5500872265183716e-05,
+      "loss": 0.3451,
+      "step": 2288
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 3.548292280804764e-05,
+      "loss": 0.7703,
+      "step": 2289
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 3.5464971320070834e-05,
+      "loss": 0.4635,
+      "step": 2290
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 3.5447017807902425e-05,
+      "loss": 0.3925,
+      "step": 2291
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 3.54290622781923e-05,
+      "loss": 0.1767,
+      "step": 2292
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 3.541110473759105e-05,
+      "loss": 0.379,
+      "step": 2293
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 3.539314519275003e-05,
+      "loss": 0.5675,
+      "step": 2294
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 3.537518365032139e-05,
+      "loss": 0.2746,
+      "step": 2295
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 3.535722011695793e-05,
+      "loss": 0.2063,
+      "step": 2296
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 3.533925459931325e-05,
+      "loss": 0.729,
+      "step": 2297
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 3.532128710404167e-05,
+      "loss": 0.301,
+      "step": 2298
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 3.5303317637798236e-05,
+      "loss": 0.1887,
+      "step": 2299
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 3.528534620723872e-05,
+      "loss": 0.1575,
+      "step": 2300
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 3.526737281901962e-05,
+      "loss": 0.3326,
+      "step": 2301
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 3.5249397479798195e-05,
+      "loss": 0.1898,
+      "step": 2302
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 3.5231420196232385e-05,
+      "loss": 0.5812,
+      "step": 2303
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 3.5213440974980865e-05,
+      "loss": 0.383,
+      "step": 2304
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 3.519545982270304e-05,
+      "loss": 0.4718,
+      "step": 2305
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 3.517747674605899e-05,
+      "loss": 0.1508,
+      "step": 2306
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 3.515949175170955e-05,
+      "loss": 0.4321,
+      "step": 2307
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 3.514150484631627e-05,
+      "loss": 0.1379,
+      "step": 2308
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 3.512351603654136e-05,
+      "loss": 0.1806,
+      "step": 2309
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 3.5105525329047776e-05,
+      "loss": 0.442,
+      "step": 2310
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 3.5087532730499164e-05,
+      "loss": 0.5147,
+      "step": 2311
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 3.5069538247559864e-05,
+      "loss": 0.7097,
+      "step": 2312
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 3.5051541886894935e-05,
+      "loss": 0.2348,
+      "step": 2313
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 3.5033543655170105e-05,
+      "loss": 0.2121,
+      "step": 2314
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 3.501554355905183e-05,
+      "loss": 0.2047,
+      "step": 2315
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 3.4997541605207206e-05,
+      "loss": 0.4924,
+      "step": 2316
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 3.497953780030406e-05,
+      "loss": 0.1765,
+      "step": 2317
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 3.4961532151010885e-05,
+      "loss": 0.1918,
+      "step": 2318
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 3.494352466399686e-05,
+      "loss": 0.1523,
+      "step": 2319
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 3.492551534593185e-05,
+      "loss": 0.3571,
+      "step": 2320
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 3.490750420348641e-05,
+      "loss": 0.2045,
+      "step": 2321
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 3.488949124333174e-05,
+      "loss": 0.3656,
+      "step": 2322
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 3.487147647213973e-05,
+      "loss": 0.1323,
+      "step": 2323
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 3.485345989658295e-05,
+      "loss": 0.188,
+      "step": 2324
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 3.483544152333461e-05,
+      "loss": 0.5212,
+      "step": 2325
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 3.481742135906861e-05,
+      "loss": 0.1897,
+      "step": 2326
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 3.4799399410459516e-05,
+      "loss": 0.2211,
+      "step": 2327
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 3.478137568418255e-05,
+      "loss": 0.1195,
+      "step": 2328
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 3.476335018691357e-05,
+      "loss": 0.6386,
+      "step": 2329
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 3.474532292532912e-05,
+      "loss": 0.1642,
+      "step": 2330
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 3.47272939061064e-05,
+      "loss": 0.3829,
+      "step": 2331
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 3.4709263135923216e-05,
+      "loss": 0.2455,
+      "step": 2332
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 3.469123062145807e-05,
+      "loss": 0.1847,
+      "step": 2333
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 3.4673196369390114e-05,
+      "loss": 0.1818,
+      "step": 2334
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 3.465516038639908e-05,
+      "loss": 0.2241,
+      "step": 2335
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 3.4637122679165424e-05,
+      "loss": 0.4547,
+      "step": 2336
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 3.461908325437019e-05,
+      "loss": 0.4029,
+      "step": 2337
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 3.460104211869506e-05,
+      "loss": 0.2211,
+      "step": 2338
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 3.458299927882236e-05,
+      "loss": 0.3695,
+      "step": 2339
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 3.4564954741435064e-05,
+      "loss": 0.3113,
+      "step": 2340
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 3.4546908513216736e-05,
+      "loss": 0.133,
+      "step": 2341
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 3.452886060085159e-05,
+      "loss": 0.1822,
+      "step": 2342
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 3.451081101102448e-05,
+      "loss": 0.422,
+      "step": 2343
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 3.449275975042084e-05,
+      "loss": 0.2504,
+      "step": 2344
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 3.447470682572676e-05,
+      "loss": 0.2931,
+      "step": 2345
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 3.445665224362892e-05,
+      "loss": 0.4986,
+      "step": 2346
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 3.443859601081465e-05,
+      "loss": 0.1871,
+      "step": 2347
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.442053813397183e-05,
+      "loss": 0.5775,
+      "step": 2348
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.440247861978901e-05,
+      "loss": 0.2035,
+      "step": 2349
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.438441747495532e-05,
+      "loss": 0.2439,
+      "step": 2350
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.43663547061605e-05,
+      "loss": 0.315,
+      "step": 2351
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.4348290320094866e-05,
+      "loss": 0.2436,
+      "step": 2352
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.433022432344939e-05,
+      "loss": 0.4593,
+      "step": 2353
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.431215672291557e-05,
+      "loss": 0.5566,
+      "step": 2354
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.429408752518556e-05,
+      "loss": 0.5427,
+      "step": 2355
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.427601673695206e-05,
+      "loss": 0.3897,
+      "step": 2356
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.4257944364908404e-05,
+      "loss": 0.4746,
+      "step": 2357
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.423987041574846e-05,
+      "loss": 0.4534,
+      "step": 2358
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.422179489616672e-05,
+      "loss": 0.6384,
+      "step": 2359
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 3.420371781285824e-05,
+      "loss": 0.3034,
+      "step": 2360
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.418563917251866e-05,
+      "loss": 0.3065,
+      "step": 2361
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.4167558981844204e-05,
+      "loss": 0.4383,
+      "step": 2362
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.414947724753166e-05,
+      "loss": 0.7766,
+      "step": 2363
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.413139397627839e-05,
+      "loss": 0.1873,
+      "step": 2364
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.411330917478231e-05,
+      "loss": 0.1238,
+      "step": 2365
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.409522284974194e-05,
+      "loss": 0.1451,
+      "step": 2366
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.407713500785635e-05,
+      "loss": 0.1865,
+      "step": 2367
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.405904565582513e-05,
+      "loss": 0.7578,
+      "step": 2368
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.404095480034849e-05,
+      "loss": 0.1928,
+      "step": 2369
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.402286244812717e-05,
+      "loss": 0.3826,
+      "step": 2370
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.400476860586246e-05,
+      "loss": 0.5359,
+      "step": 2371
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.398667328025621e-05,
+      "loss": 0.2032,
+      "step": 2372
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.3968576478010826e-05,
+      "loss": 0.1884,
+      "step": 2373
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.395047820582923e-05,
+      "loss": 0.1764,
+      "step": 2374
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.393237847041492e-05,
+      "loss": 0.1779,
+      "step": 2375
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.391427727847193e-05,
+      "loss": 0.5879,
+      "step": 2376
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.3896174636704825e-05,
+      "loss": 0.3064,
+      "step": 2377
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.3878070551818724e-05,
+      "loss": 0.4641,
+      "step": 2378
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.385996503051925e-05,
+      "loss": 0.1654,
+      "step": 2379
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.3841858079512586e-05,
+      "loss": 0.151,
+      "step": 2380
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.382374970550544e-05,
+      "loss": 0.1306,
+      "step": 2381
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.380563991520502e-05,
+      "loss": 0.2828,
+      "step": 2382
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.378752871531911e-05,
+      "loss": 0.3083,
+      "step": 2383
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.376941611255597e-05,
+      "loss": 0.5293,
+      "step": 2384
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.3751302113624376e-05,
+      "loss": 0.1301,
+      "step": 2385
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 3.373318672523368e-05,
+      "loss": 0.4215,
+      "step": 2386
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.3715069954093704e-05,
+      "loss": 0.1875,
+      "step": 2387
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.3696951806914755e-05,
+      "loss": 0.288,
+      "step": 2388
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.367883229040773e-05,
+      "loss": 0.2118,
+      "step": 2389
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.366071141128394e-05,
+      "loss": 0.2086,
+      "step": 2390
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.3642589176255276e-05,
+      "loss": 0.3006,
+      "step": 2391
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.362446559203409e-05,
+      "loss": 0.2132,
+      "step": 2392
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.360634066533325e-05,
+      "loss": 0.3676,
+      "step": 2393
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.358821440286612e-05,
+      "loss": 0.1723,
+      "step": 2394
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.357008681134655e-05,
+      "loss": 0.3975,
+      "step": 2395
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.355195789748889e-05,
+      "loss": 0.1892,
+      "step": 2396
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.3533827668007975e-05,
+      "loss": 0.2125,
+      "step": 2397
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.351569612961913e-05,
+      "loss": 0.5144,
+      "step": 2398
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 3.349756328903817e-05,
+      "loss": 0.4088,
+      "step": 2399
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.3479429152981376e-05,
+      "loss": 0.6613,
+      "step": 2400
+    },
+    {
+      "epoch": 1.86,
+      "eval_loss": 1.1753171682357788,
+      "eval_runtime": 157.4309,
+      "eval_samples_per_second": 0.337,
+      "eval_steps_per_second": 0.172,
+      "step": 2400
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.346129372816553e-05,
+      "loss": 0.566,
+      "step": 2401
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.344315702130789e-05,
+      "loss": 0.2193,
+      "step": 2402
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.342501903912615e-05,
+      "loss": 0.4148,
+      "step": 2403
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.340687978833854e-05,
+      "loss": 0.6537,
+      "step": 2404
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.3388739275663716e-05,
+      "loss": 0.1294,
+      "step": 2405
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.337059750782079e-05,
+      "loss": 0.1229,
+      "step": 2406
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.335245449152939e-05,
+      "loss": 0.2453,
+      "step": 2407
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.3334310233509566e-05,
+      "loss": 0.2588,
+      "step": 2408
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.331616474048183e-05,
+      "loss": 0.724,
+      "step": 2409
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.329801801916718e-05,
+      "loss": 0.3062,
+      "step": 2410
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.327987007628703e-05,
+      "loss": 0.4649,
+      "step": 2411
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.326172091856328e-05,
+      "loss": 0.3204,
+      "step": 2412
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.324357055271826e-05,
+      "loss": 0.4035,
+      "step": 2413
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.322541898547475e-05,
+      "loss": 0.1923,
+      "step": 2414
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.320726622355598e-05,
+      "loss": 0.6163,
+      "step": 2415
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.3189112273685625e-05,
+      "loss": 0.3185,
+      "step": 2416
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.317095714258779e-05,
+      "loss": 0.3323,
+      "step": 2417
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.3152800836987034e-05,
+      "loss": 0.2249,
+      "step": 2418
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.3134643363608324e-05,
+      "loss": 0.2516,
+      "step": 2419
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.311648472917709e-05,
+      "loss": 0.3172,
+      "step": 2420
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.309832494041918e-05,
+      "loss": 0.6043,
+      "step": 2421
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.3080164004060856e-05,
+      "loss": 0.4875,
+      "step": 2422
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.306200192682881e-05,
+      "loss": 0.2233,
+      "step": 2423
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.304383871545019e-05,
+      "loss": 0.1641,
+      "step": 2424
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 3.302567437665251e-05,
+      "loss": 0.1843,
+      "step": 2425
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.300750891716375e-05,
+      "loss": 0.1982,
+      "step": 2426
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.298934234371227e-05,
+      "loss": 0.3195,
+      "step": 2427
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.297117466302687e-05,
+      "loss": 0.5131,
+      "step": 2428
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.295300588183672e-05,
+      "loss": 0.9012,
+      "step": 2429
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.293483600687145e-05,
+      "loss": 0.4115,
+      "step": 2430
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.291666504486107e-05,
+      "loss": 0.2508,
+      "step": 2431
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.289849300253597e-05,
+      "loss": 0.262,
+      "step": 2432
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.288031988662698e-05,
+      "loss": 0.3705,
+      "step": 2433
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.28621457038653e-05,
+      "loss": 0.3236,
+      "step": 2434
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.284397046098254e-05,
+      "loss": 0.3748,
+      "step": 2435
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.2825794164710703e-05,
+      "loss": 0.4525,
+      "step": 2436
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.2807616821782175e-05,
+      "loss": 0.45,
+      "step": 2437
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 3.278943843892973e-05,
+      "loss": 0.4343,
+      "step": 2438
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.2771259022886515e-05,
+      "loss": 0.7319,
+      "step": 2439
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.275307858038609e-05,
+      "loss": 0.2413,
+      "step": 2440
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.273489711816238e-05,
+      "loss": 0.4171,
+      "step": 2441
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.271671464294967e-05,
+      "loss": 0.2132,
+      "step": 2442
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.2698531161482654e-05,
+      "loss": 0.2573,
+      "step": 2443
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.2680346680496375e-05,
+      "loss": 0.1707,
+      "step": 2444
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.2662161206726245e-05,
+      "loss": 0.199,
+      "step": 2445
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.264397474690806e-05,
+      "loss": 0.3317,
+      "step": 2446
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.2625787307777966e-05,
+      "loss": 0.2063,
+      "step": 2447
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.260759889607248e-05,
+      "loss": 0.2351,
+      "step": 2448
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.2589409518528486e-05,
+      "loss": 0.3447,
+      "step": 2449
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.25712191818832e-05,
+      "loss": 0.1287,
+      "step": 2450
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.255302789287422e-05,
+      "loss": 0.38,
+      "step": 2451
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.2534835658239496e-05,
+      "loss": 0.3425,
+      "step": 2452
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.2516642484717295e-05,
+      "loss": 0.3509,
+      "step": 2453
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.2498448379046294e-05,
+      "loss": 0.7004,
+      "step": 2454
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.2480253347965444e-05,
+      "loss": 0.7204,
+      "step": 2455
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.246205739821408e-05,
+      "loss": 0.3186,
+      "step": 2456
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.244386053653188e-05,
+      "loss": 0.2454,
+      "step": 2457
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.2425662769658825e-05,
+      "loss": 0.5799,
+      "step": 2458
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.240746410433528e-05,
+      "loss": 0.2337,
+      "step": 2459
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.2389264547301916e-05,
+      "loss": 0.205,
+      "step": 2460
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.237106410529972e-05,
+      "loss": 0.1358,
+      "step": 2461
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.235286278507004e-05,
+      "loss": 0.1287,
+      "step": 2462
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.233466059335452e-05,
+      "loss": 0.3839,
+      "step": 2463
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 3.231645753689515e-05,
+      "loss": 0.5571,
+      "step": 2464
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 3.229825362243422e-05,
+      "loss": 0.4426,
+      "step": 2465
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 3.228004885671436e-05,
+      "loss": 0.557,
+      "step": 2466
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 3.2261843246478484e-05,
+      "loss": 0.1896,
+      "step": 2467
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 3.224363679846985e-05,
+      "loss": 0.4123,
+      "step": 2468
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 3.2225429519432015e-05,
+      "loss": 0.4817,
+      "step": 2469
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 3.220722141610885e-05,
+      "loss": 0.2906,
+      "step": 2470
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 3.21890124952445e-05,
+      "loss": 0.3371,
+      "step": 2471
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 3.217080276358346e-05,
+      "loss": 0.3117,
+      "step": 2472
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 3.215259222787049e-05,
+      "loss": 0.1332,
+      "step": 2473
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 3.2134380894850655e-05,
+      "loss": 0.4432,
+      "step": 2474
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 3.211616877126933e-05,
+      "loss": 0.6304,
+      "step": 2475
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 3.209795586387217e-05,
+      "loss": 0.1364,
+      "step": 2476
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 3.2079742179405134e-05,
+      "loss": 0.3121,
+      "step": 2477
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 3.206152772461444e-05,
+      "loss": 0.1771,
+      "step": 2478
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 3.204331250624661e-05,
+      "loss": 0.1476,
+      "step": 2479
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 3.202509653104848e-05,
+      "loss": 0.473,
+      "step": 2480
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 3.200687980576709e-05,
+      "loss": 0.3406,
+      "step": 2481
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 3.198866233714983e-05,
+      "loss": 0.2171,
+      "step": 2482
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 3.197044413194433e-05,
+      "loss": 0.1963,
+      "step": 2483
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 3.19522251968985e-05,
+      "loss": 0.4411,
+      "step": 2484
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 3.193400553876052e-05,
+      "loss": 0.5454,
+      "step": 2485
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 3.1915785164278844e-05,
+      "loss": 0.2958,
+      "step": 2486
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 3.189756408020218e-05,
+      "loss": 0.172,
+      "step": 2487
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 3.187934229327951e-05,
+      "loss": 0.2809,
+      "step": 2488
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 3.186111981026007e-05,
+      "loss": 0.2105,
+      "step": 2489
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 3.1842896637893366e-05,
+      "loss": 0.4377,
+      "step": 2490
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 3.182467278292912e-05,
+      "loss": 0.4509,
+      "step": 2491
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 3.180644825211736e-05,
+      "loss": 0.5389,
+      "step": 2492
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 3.178822305220834e-05,
+      "loss": 0.4461,
+      "step": 2493
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 3.176999718995253e-05,
+      "loss": 0.3679,
+      "step": 2494
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 3.1751770672100716e-05,
+      "loss": 0.246,
+      "step": 2495
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 3.173354350540387e-05,
+      "loss": 0.1105,
+      "step": 2496
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 3.1715315696613223e-05,
+      "loss": 0.2165,
+      "step": 2497
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 3.169708725248024e-05,
+      "loss": 0.4289,
+      "step": 2498
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 3.1678858179756625e-05,
+      "loss": 0.265,
+      "step": 2499
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 3.166062848519433e-05,
+      "loss": 0.6181,
+      "step": 2500
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 3.1642398175545486e-05,
+      "loss": 0.1614,
+      "step": 2501
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 3.162416725756253e-05,
+      "loss": 0.1388,
+      "step": 2502
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 3.1605935737998046e-05,
+      "loss": 0.15,
+      "step": 2503
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 3.158770362360489e-05,
+      "loss": 0.2631,
+      "step": 2504
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 3.156947092113613e-05,
+      "loss": 0.3655,
+      "step": 2505
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 3.155123763734505e-05,
+      "loss": 0.5143,
+      "step": 2506
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 3.153300377898512e-05,
+      "loss": 0.3104,
+      "step": 2507
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 3.151476935281006e-05,
+      "loss": 0.1402,
+      "step": 2508
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 3.1496534365573815e-05,
+      "loss": 0.1962,
+      "step": 2509
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 3.1478298824030474e-05,
+      "loss": 0.2542,
+      "step": 2510
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 3.146006273493438e-05,
+      "loss": 0.225,
+      "step": 2511
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 3.144182610504009e-05,
+      "loss": 0.1246,
+      "step": 2512
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 3.142358894110231e-05,
+      "loss": 0.4931,
+      "step": 2513
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 3.140535124987598e-05,
+      "loss": 0.1804,
+      "step": 2514
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 3.1387113038116244e-05,
+      "loss": 0.3447,
+      "step": 2515
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.136887431257841e-05,
+      "loss": 0.4306,
+      "step": 2516
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.135063508001798e-05,
+      "loss": 0.3624,
+      "step": 2517
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.133239534719068e-05,
+      "loss": 0.2684,
+      "step": 2518
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.1314155120852367e-05,
+      "loss": 0.5036,
+      "step": 2519
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.129591440775912e-05,
+      "loss": 0.1739,
+      "step": 2520
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.127767321466718e-05,
+      "loss": 0.2238,
+      "step": 2521
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.125943154833299e-05,
+      "loss": 0.7494,
+      "step": 2522
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.124118941551313e-05,
+      "loss": 0.3573,
+      "step": 2523
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.1222946822964385e-05,
+      "loss": 0.7626,
+      "step": 2524
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.12047037774437e-05,
+      "loss": 0.4664,
+      "step": 2525
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.118646028570817e-05,
+      "loss": 0.2228,
+      "step": 2526
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.116821635451509e-05,
+      "loss": 0.532,
+      "step": 2527
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.114997199062189e-05,
+      "loss": 0.3582,
+      "step": 2528
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 3.113172720078617e-05,
+      "loss": 0.3315,
+      "step": 2529
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 3.111348199176569e-05,
+      "loss": 0.4453,
+      "step": 2530
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 3.1095236370318365e-05,
+      "loss": 0.1454,
+      "step": 2531
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 3.107699034320226e-05,
+      "loss": 0.372,
+      "step": 2532
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 3.1058743917175575e-05,
+      "loss": 0.501,
+      "step": 2533
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 3.104049709899669e-05,
+      "loss": 0.1513,
+      "step": 2534
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 3.102224989542412e-05,
+      "loss": 0.1207,
+      "step": 2535
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 3.10040023132165e-05,
+      "loss": 0.1631,
+      "step": 2536
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 3.098575435913262e-05,
+      "loss": 0.3999,
+      "step": 2537
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 3.096750603993144e-05,
+      "loss": 0.2556,
+      "step": 2538
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 3.094925736237199e-05,
+      "loss": 0.674,
+      "step": 2539
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 3.093100833321348e-05,
+      "loss": 0.1567,
+      "step": 2540
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 3.0912758959215236e-05,
+      "loss": 0.1225,
+      "step": 2541
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.0894509247136714e-05,
+      "loss": 0.8201,
+      "step": 2542
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.0876259203737494e-05,
+      "loss": 0.4759,
+      "step": 2543
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.0858008835777286e-05,
+      "loss": 0.1201,
+      "step": 2544
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.08397581500159e-05,
+      "loss": 0.1682,
+      "step": 2545
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.082150715321329e-05,
+      "loss": 0.2009,
+      "step": 2546
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.080325585212951e-05,
+      "loss": 0.1965,
+      "step": 2547
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.078500425352473e-05,
+      "loss": 0.1985,
+      "step": 2548
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.076675236415922e-05,
+      "loss": 0.2376,
+      "step": 2549
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.074850019079336e-05,
+      "loss": 0.2491,
+      "step": 2550
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.073024774018767e-05,
+      "loss": 0.3514,
+      "step": 2551
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.071199501910273e-05,
+      "loss": 0.6186,
+      "step": 2552
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.069374203429923e-05,
+      "loss": 0.1723,
+      "step": 2553
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.067548879253798e-05,
+      "loss": 0.3188,
+      "step": 2554
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 3.0657235300579844e-05,
+      "loss": 0.4673,
+      "step": 2555
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 3.0638981565185814e-05,
+      "loss": 0.1918,
+      "step": 2556
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 3.0620727593116976e-05,
+      "loss": 0.4299,
+      "step": 2557
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 3.060247339113446e-05,
+      "loss": 0.3324,
+      "step": 2558
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 3.0584218965999525e-05,
+      "loss": 0.1528,
+      "step": 2559
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 3.0565964324473494e-05,
+      "loss": 0.5624,
+      "step": 2560
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 3.054770947331778e-05,
+      "loss": 0.4548,
+      "step": 2561
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 3.0529454419293866e-05,
+      "loss": 0.6112,
+      "step": 2562
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 3.0511199169163292e-05,
+      "loss": 0.3469,
+      "step": 2563
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 3.0492943729687722e-05,
+      "loss": 0.5332,
+      "step": 2564
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 3.0474688107628833e-05,
+      "loss": 0.1745,
+      "step": 2565
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 3.04564323097484e-05,
+      "loss": 0.1686,
+      "step": 2566
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 3.043817634280826e-05,
+      "loss": 0.5119,
+      "step": 2567
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.041992021357032e-05,
+      "loss": 0.3597,
+      "step": 2568
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.040166392879652e-05,
+      "loss": 0.2871,
+      "step": 2569
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.038340749524889e-05,
+      "loss": 0.3846,
+      "step": 2570
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.0365150919689487e-05,
+      "loss": 0.3755,
+      "step": 2571
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.034689420888044e-05,
+      "loss": 0.4556,
+      "step": 2572
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.032863736958392e-05,
+      "loss": 0.3046,
+      "step": 2573
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.0310380408562164e-05,
+      "loss": 0.5478,
+      "step": 2574
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.0292123332577408e-05,
+      "loss": 0.1595,
+      "step": 2575
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.027386614839198e-05,
+      "loss": 0.4751,
+      "step": 2576
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.025560886276822e-05,
+      "loss": 0.4221,
+      "step": 2577
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.0237351482468526e-05,
+      "loss": 0.3778,
+      "step": 2578
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.0219094014255297e-05,
+      "loss": 0.2671,
+      "step": 2579
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 3.020083646489101e-05,
+      "loss": 0.2433,
+      "step": 2580
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.0182578841138134e-05,
+      "loss": 0.2239,
+      "step": 2581
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.0164321149759182e-05,
+      "loss": 0.1477,
+      "step": 2582
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.0146063397516687e-05,
+      "loss": 0.5676,
+      "step": 2583
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.0127805591173215e-05,
+      "loss": 0.4448,
+      "step": 2584
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.0109547737491345e-05,
+      "loss": 0.1411,
+      "step": 2585
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.0091289843233677e-05,
+      "loss": 0.429,
+      "step": 2586
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.0073031915162805e-05,
+      "loss": 0.245,
+      "step": 2587
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.0054773960041364e-05,
+      "loss": 0.1777,
+      "step": 2588
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.003651598463199e-05,
+      "loss": 0.1928,
+      "step": 2589
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.0018257995697325e-05,
+      "loss": 0.1841,
+      "step": 2590
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3e-05,
+      "loss": 0.163,
+      "step": 2591
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.998174200430268e-05,
+      "loss": 0.1857,
+      "step": 2592
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.9963484015368015e-05,
+      "loss": 0.0856,
+      "step": 2593
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.994522603995864e-05,
+      "loss": 0.2907,
+      "step": 2594
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.99269680848372e-05,
+      "loss": 0.0996,
+      "step": 2595
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.9908710156766328e-05,
+      "loss": 0.0611,
+      "step": 2596
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.989045226250866e-05,
+      "loss": 0.0814,
+      "step": 2597
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.9872194408826786e-05,
+      "loss": 0.4061,
+      "step": 2598
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.9853936602483315e-05,
+      "loss": 0.0682,
+      "step": 2599
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.983567885024083e-05,
+      "loss": 0.0695,
+      "step": 2600
+    },
+    {
+      "epoch": 2.01,
+      "eval_loss": 1.335438847541809,
+      "eval_runtime": 157.3144,
+      "eval_samples_per_second": 0.337,
+      "eval_steps_per_second": 0.172,
+      "step": 2600
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.9817421158861877e-05,
+      "loss": 0.1366,
+      "step": 2601
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.9799163535109002e-05,
+      "loss": 0.2119,
+      "step": 2602
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.9780905985744715e-05,
+      "loss": 0.124,
+      "step": 2603
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.976264851753149e-05,
+      "loss": 0.5529,
+      "step": 2604
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.9744391137231787e-05,
+      "loss": 0.2249,
+      "step": 2605
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.972613385160803e-05,
+      "loss": 0.1151,
+      "step": 2606
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.97078766674226e-05,
+      "loss": 0.0664,
+      "step": 2607
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.9689619591437838e-05,
+      "loss": 0.3593,
+      "step": 2608
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.9671362630416077e-05,
+      "loss": 0.1133,
+      "step": 2609
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.9653105791119556e-05,
+      "loss": 0.1993,
+      "step": 2610
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.963484908031051e-05,
+      "loss": 0.2697,
+      "step": 2611
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.9616592504751116e-05,
+      "loss": 0.1941,
+      "step": 2612
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.9598336071203483e-05,
+      "loss": 0.2864,
+      "step": 2613
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.958007978642968e-05,
+      "loss": 0.0806,
+      "step": 2614
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.956182365719174e-05,
+      "loss": 0.2995,
+      "step": 2615
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.9543567690251603e-05,
+      "loss": 0.0899,
+      "step": 2616
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.9525311892371172e-05,
+      "loss": 0.2613,
+      "step": 2617
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.950705627031228e-05,
+      "loss": 0.1359,
+      "step": 2618
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.948880083083671e-05,
+      "loss": 0.2371,
+      "step": 2619
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 2.9470545580706142e-05,
+      "loss": 0.0938,
+      "step": 2620
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 2.9452290526682222e-05,
+      "loss": 0.0825,
+      "step": 2621
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 2.943403567552651e-05,
+      "loss": 0.5083,
+      "step": 2622
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 2.9415781034000484e-05,
+      "loss": 0.1081,
+      "step": 2623
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 2.9397526608865547e-05,
+      "loss": 0.2707,
+      "step": 2624
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 2.9379272406883032e-05,
+      "loss": 0.2003,
+      "step": 2625
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 2.936101843481419e-05,
+      "loss": 0.1218,
+      "step": 2626
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 2.934276469942016e-05,
+      "loss": 0.0905,
+      "step": 2627
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 2.932451120746203e-05,
+      "loss": 0.4,
+      "step": 2628
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 2.9306257965700776e-05,
+      "loss": 0.0747,
+      "step": 2629
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 2.9288004980897278e-05,
+      "loss": 0.1602,
+      "step": 2630
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 2.9269752259812334e-05,
+      "loss": 0.3759,
+      "step": 2631
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 2.9251499809206645e-05,
+      "loss": 0.0988,
+      "step": 2632
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 2.9233247635840798e-05,
+      "loss": 0.1328,
+      "step": 2633
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 2.9214995746475283e-05,
+      "loss": 0.0779,
+      "step": 2634
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 2.91967441478705e-05,
+      "loss": 0.1418,
+      "step": 2635
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 2.9178492846786718e-05,
+      "loss": 0.1991,
+      "step": 2636
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 2.916024184998411e-05,
+      "loss": 0.0897,
+      "step": 2637
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 2.9141991164222726e-05,
+      "loss": 0.0677,
+      "step": 2638
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 2.9123740796262517e-05,
+      "loss": 0.0735,
+      "step": 2639
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 2.91054907528633e-05,
+      "loss": 0.1536,
+      "step": 2640
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 2.908724104078477e-05,
+      "loss": 0.0883,
+      "step": 2641
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 2.9068991666786523e-05,
+      "loss": 0.1142,
+      "step": 2642
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 2.9050742637628013e-05,
+      "loss": 0.3071,
+      "step": 2643
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 2.9032493960068567e-05,
+      "loss": 0.1363,
+      "step": 2644
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 2.9014245640867368e-05,
+      "loss": 0.155,
+      "step": 2645
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 2.89959976867835e-05,
+      "loss": 0.1454,
+      "step": 2646
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 2.897775010457588e-05,
+      "loss": 0.0655,
+      "step": 2647
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 2.8959502901003307e-05,
+      "loss": 0.3273,
+      "step": 2648
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 2.8941256082824426e-05,
+      "loss": 0.1002,
+      "step": 2649
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 2.892300965679775e-05,
+      "loss": 0.0958,
+      "step": 2650
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 2.8904763629681644e-05,
+      "loss": 0.0994,
+      "step": 2651
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 2.8886518008234313e-05,
+      "loss": 0.2024,
+      "step": 2652
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 2.8868272799213834e-05,
+      "loss": 0.1901,
+      "step": 2653
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 2.8850028009378115e-05,
+      "loss": 0.0862,
+      "step": 2654
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 2.883178364548492e-05,
+      "loss": 0.0827,
+      "step": 2655
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 2.8813539714291834e-05,
+      "loss": 0.0599,
+      "step": 2656
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 2.879529622255631e-05,
+      "loss": 0.0843,
+      "step": 2657
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 2.877705317703562e-05,
+      "loss": 0.2363,
+      "step": 2658
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2.8758810584486873e-05,
+      "loss": 0.19,
+      "step": 2659
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2.8740568451667014e-05,
+      "loss": 0.1867,
+      "step": 2660
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2.8722326785332827e-05,
+      "loss": 0.2186,
+      "step": 2661
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2.8704085592240887e-05,
+      "loss": 0.0741,
+      "step": 2662
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2.8685844879147642e-05,
+      "loss": 0.0878,
+      "step": 2663
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2.866760465280933e-05,
+      "loss": 0.0753,
+      "step": 2664
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2.8649364919982025e-05,
+      "loss": 0.0955,
+      "step": 2665
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2.8631125687421602e-05,
+      "loss": 0.1026,
+      "step": 2666
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2.8612886961883764e-05,
+      "loss": 0.1824,
+      "step": 2667
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2.8594648750124027e-05,
+      "loss": 0.2122,
+      "step": 2668
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2.85764110588977e-05,
+      "loss": 0.067,
+      "step": 2669
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2.855817389495992e-05,
+      "loss": 0.0873,
+      "step": 2670
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2.8539937265065627e-05,
+      "loss": 0.1973,
+      "step": 2671
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2.852170117596954e-05,
+      "loss": 0.1529,
+      "step": 2672
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2.8503465634426204e-05,
+      "loss": 0.109,
+      "step": 2673
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2.8485230647189935e-05,
+      "loss": 0.3222,
+      "step": 2674
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2.8466996221014883e-05,
+      "loss": 0.0856,
+      "step": 2675
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2.8448762362654958e-05,
+      "loss": 0.1352,
+      "step": 2676
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2.843052907886387e-05,
+      "loss": 0.2769,
+      "step": 2677
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2.8412296376395104e-05,
+      "loss": 0.215,
+      "step": 2678
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2.8394064262001952e-05,
+      "loss": 0.2883,
+      "step": 2679
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2.8375832742437478e-05,
+      "loss": 0.1699,
+      "step": 2680
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2.835760182445451e-05,
+      "loss": 0.0642,
+      "step": 2681
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2.8339371514805672e-05,
+      "loss": 0.2097,
+      "step": 2682
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2.8321141820243373e-05,
+      "loss": 0.0615,
+      "step": 2683
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2.8302912747519765e-05,
+      "loss": 0.0812,
+      "step": 2684
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 2.828468430338678e-05,
+      "loss": 0.2516,
+      "step": 2685
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 2.8266456494596132e-05,
+      "loss": 0.1568,
+      "step": 2686
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 2.8248229327899293e-05,
+      "loss": 0.2145,
+      "step": 2687
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 2.8230002810047472e-05,
+      "loss": 0.2721,
+      "step": 2688
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 2.8211776947791674e-05,
+      "loss": 0.0745,
+      "step": 2689
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 2.8193551747882648e-05,
+      "loss": 0.0656,
+      "step": 2690
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 2.8175327217070883e-05,
+      "loss": 0.1612,
+      "step": 2691
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 2.8157103362106643e-05,
+      "loss": 0.3159,
+      "step": 2692
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 2.8138880189739925e-05,
+      "loss": 0.2971,
+      "step": 2693
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 2.8120657706720495e-05,
+      "loss": 0.0713,
+      "step": 2694
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 2.810243591979782e-05,
+      "loss": 0.1245,
+      "step": 2695
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 2.808421483572116e-05,
+      "loss": 0.1247,
+      "step": 2696
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.806599446123949e-05,
+      "loss": 0.1011,
+      "step": 2697
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.804777480310151e-05,
+      "loss": 0.1108,
+      "step": 2698
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.802955586805568e-05,
+      "loss": 0.3804,
+      "step": 2699
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.8011337662850185e-05,
+      "loss": 0.222,
+      "step": 2700
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.799312019423292e-05,
+      "loss": 0.0574,
+      "step": 2701
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.7974903468951535e-05,
+      "loss": 0.1208,
+      "step": 2702
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.7956687493753388e-05,
+      "loss": 0.2133,
+      "step": 2703
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.7938472275385573e-05,
+      "loss": 0.0663,
+      "step": 2704
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.7920257820594877e-05,
+      "loss": 0.3275,
+      "step": 2705
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.7902044136127836e-05,
+      "loss": 0.0894,
+      "step": 2706
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.7883831228730667e-05,
+      "loss": 0.1515,
+      "step": 2707
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.7865619105149344e-05,
+      "loss": 0.1558,
+      "step": 2708
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.7847407772129522e-05,
+      "loss": 0.0822,
+      "step": 2709
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2.7829197236416545e-05,
+      "loss": 0.2721,
+      "step": 2710
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2.7810987504755498e-05,
+      "loss": 0.1222,
+      "step": 2711
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2.7792778583891157e-05,
+      "loss": 0.1614,
+      "step": 2712
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2.7774570480567983e-05,
+      "loss": 0.0554,
+      "step": 2713
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2.7756363201530147e-05,
+      "loss": 0.0706,
+      "step": 2714
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2.7738156753521517e-05,
+      "loss": 0.0816,
+      "step": 2715
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2.771995114328565e-05,
+      "loss": 0.2137,
+      "step": 2716
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2.7701746377565782e-05,
+      "loss": 0.0694,
+      "step": 2717
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2.7683542463104854e-05,
+      "loss": 0.2035,
+      "step": 2718
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2.7665339406645486e-05,
+      "loss": 0.0767,
+      "step": 2719
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2.7647137214929968e-05,
+      "loss": 0.0689,
+      "step": 2720
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2.7628935894700283e-05,
+      "loss": 0.1049,
+      "step": 2721
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2.761073545269809e-05,
+      "loss": 0.3203,
+      "step": 2722
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.7592535895664723e-05,
+      "loss": 0.2265,
+      "step": 2723
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.757433723034118e-05,
+      "loss": 0.1851,
+      "step": 2724
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.755613946346813e-05,
+      "loss": 0.6216,
+      "step": 2725
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.753794260178593e-05,
+      "loss": 0.0753,
+      "step": 2726
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.7519746652034564e-05,
+      "loss": 0.2199,
+      "step": 2727
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.750155162095371e-05,
+      "loss": 0.0728,
+      "step": 2728
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.7483357515282703e-05,
+      "loss": 0.0836,
+      "step": 2729
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.7465164341760512e-05,
+      "loss": 0.0877,
+      "step": 2730
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.744697210712578e-05,
+      "loss": 0.1349,
+      "step": 2731
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.7428780818116804e-05,
+      "loss": 0.4259,
+      "step": 2732
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.741059048147153e-05,
+      "loss": 0.0621,
+      "step": 2733
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.739240110392753e-05,
+      "loss": 0.2448,
+      "step": 2734
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.7374212692222046e-05,
+      "loss": 0.1045,
+      "step": 2735
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.7356025253091956e-05,
+      "loss": 0.4856,
+      "step": 2736
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.7337838793273767e-05,
+      "loss": 0.0928,
+      "step": 2737
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.731965331950364e-05,
+      "loss": 0.1569,
+      "step": 2738
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.7301468838517354e-05,
+      "loss": 0.2061,
+      "step": 2739
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.7283285357050326e-05,
+      "loss": 0.1117,
+      "step": 2740
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.7265102881837626e-05,
+      "loss": 0.3108,
+      "step": 2741
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.724692141961391e-05,
+      "loss": 0.2458,
+      "step": 2742
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.7228740977113483e-05,
+      "loss": 0.0847,
+      "step": 2743
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.7210561561070274e-05,
+      "loss": 0.057,
+      "step": 2744
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.7192383178217827e-05,
+      "loss": 0.0583,
+      "step": 2745
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.717420583528929e-05,
+      "loss": 0.3448,
+      "step": 2746
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.7156029539017454e-05,
+      "loss": 0.0698,
+      "step": 2747
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.7137854296134703e-05,
+      "loss": 0.1725,
+      "step": 2748
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.7119680113373028e-05,
+      "loss": 0.159,
+      "step": 2749
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.7101506997464035e-05,
+      "loss": 0.1498,
+      "step": 2750
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.7083334955138937e-05,
+      "loss": 0.0782,
+      "step": 2751
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.7065163993128555e-05,
+      "loss": 0.1827,
+      "step": 2752
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.7046994118163282e-05,
+      "loss": 0.1389,
+      "step": 2753
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.7028825336973138e-05,
+      "loss": 0.0784,
+      "step": 2754
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.7010657656287736e-05,
+      "loss": 0.0993,
+      "step": 2755
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.6992491082836256e-05,
+      "loss": 0.1358,
+      "step": 2756
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.6974325623347486e-05,
+      "loss": 0.2114,
+      "step": 2757
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.6956161284549817e-05,
+      "loss": 0.3379,
+      "step": 2758
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.693799807317119e-05,
+      "loss": 0.2415,
+      "step": 2759
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.6919835995939152e-05,
+      "loss": 0.3605,
+      "step": 2760
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.6901675059580825e-05,
+      "loss": 0.188,
+      "step": 2761
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.6883515270822914e-05,
+      "loss": 0.2151,
+      "step": 2762
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.6865356636391677e-05,
+      "loss": 0.2545,
+      "step": 2763
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.6847199163012975e-05,
+      "loss": 0.0685,
+      "step": 2764
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.6829042857412216e-05,
+      "loss": 0.0737,
+      "step": 2765
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.6810887726314383e-05,
+      "loss": 0.147,
+      "step": 2766
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.6792733776444026e-05,
+      "loss": 0.1408,
+      "step": 2767
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.6774581014525264e-05,
+      "loss": 0.1227,
+      "step": 2768
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.6756429447281754e-05,
+      "loss": 0.0657,
+      "step": 2769
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.673827908143673e-05,
+      "loss": 0.1261,
+      "step": 2770
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.6720129923712978e-05,
+      "loss": 0.4706,
+      "step": 2771
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.6701981980832837e-05,
+      "loss": 0.0767,
+      "step": 2772
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.6683835259518162e-05,
+      "loss": 0.0668,
+      "step": 2773
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.666568976649044e-05,
+      "loss": 0.1023,
+      "step": 2774
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2.6647545508470614e-05,
+      "loss": 0.1177,
+      "step": 2775
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2.662940249217921e-05,
+      "loss": 0.0841,
+      "step": 2776
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2.6611260724336296e-05,
+      "loss": 0.0684,
+      "step": 2777
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2.659312021166146e-05,
+      "loss": 0.1045,
+      "step": 2778
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2.6574980960873848e-05,
+      "loss": 0.1161,
+      "step": 2779
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2.6556842978692114e-05,
+      "loss": 0.1463,
+      "step": 2780
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2.6538706271834472e-05,
+      "loss": 0.0593,
+      "step": 2781
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2.6520570847018625e-05,
+      "loss": 0.1233,
+      "step": 2782
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2.6502436710961835e-05,
+      "loss": 0.611,
+      "step": 2783
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2.648430387038088e-05,
+      "loss": 0.0842,
+      "step": 2784
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2.6466172331992034e-05,
+      "loss": 0.3038,
+      "step": 2785
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2.6448042102511115e-05,
+      "loss": 0.0972,
+      "step": 2786
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2.6429913188653456e-05,
+      "loss": 0.0839,
+      "step": 2787
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 2.6411785597133887e-05,
+      "loss": 0.0609,
+      "step": 2788
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 2.6393659334666748e-05,
+      "loss": 0.0665,
+      "step": 2789
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 2.6375534407965906e-05,
+      "loss": 0.0736,
+      "step": 2790
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 2.635741082374473e-05,
+      "loss": 0.1233,
+      "step": 2791
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 2.633928858871606e-05,
+      "loss": 0.0765,
+      "step": 2792
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 2.632116770959228e-05,
+      "loss": 0.2324,
+      "step": 2793
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 2.6303048193085243e-05,
+      "loss": 0.4544,
+      "step": 2794
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 2.6284930045906304e-05,
+      "loss": 0.0573,
+      "step": 2795
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 2.6266813274766317e-05,
+      "loss": 0.1565,
+      "step": 2796
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 2.6248697886375625e-05,
+      "loss": 0.0736,
+      "step": 2797
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 2.6230583887444048e-05,
+      "loss": 0.0793,
+      "step": 2798
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 2.6212471284680903e-05,
+      "loss": 0.0959,
+      "step": 2799
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 2.6194360084794983e-05,
+      "loss": 0.0666,
+      "step": 2800
+    },
+    {
+      "epoch": 2.16,
+      "eval_loss": 1.3078112602233887,
+      "eval_runtime": 157.2317,
+      "eval_samples_per_second": 0.337,
+      "eval_steps_per_second": 0.172,
+      "step": 2800
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.6176250294494577e-05,
+      "loss": 0.0786,
+      "step": 2801
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.6158141920487422e-05,
+      "loss": 0.2882,
+      "step": 2802
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.614003496948076e-05,
+      "loss": 0.1665,
+      "step": 2803
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.6121929448181288e-05,
+      "loss": 0.0841,
+      "step": 2804
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.610382536329518e-05,
+      "loss": 0.0976,
+      "step": 2805
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.6085722721528076e-05,
+      "loss": 0.0686,
+      "step": 2806
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.606762152958508e-05,
+      "loss": 0.1333,
+      "step": 2807
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.6049521794170772e-05,
+      "loss": 0.1714,
+      "step": 2808
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.6031423521989176e-05,
+      "loss": 0.0671,
+      "step": 2809
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.6013326719743785e-05,
+      "loss": 0.3498,
+      "step": 2810
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.5995231394137535e-05,
+      "loss": 0.0978,
+      "step": 2811
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.5977137551872824e-05,
+      "loss": 0.0909,
+      "step": 2812
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.595904519965151e-05,
+      "loss": 0.2613,
+      "step": 2813
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2.5940954344174876e-05,
+      "loss": 0.2699,
+      "step": 2814
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2.592286499214366e-05,
+      "loss": 0.213,
+      "step": 2815
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2.5904777150258065e-05,
+      "loss": 0.3899,
+      "step": 2816
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2.5886690825217695e-05,
+      "loss": 0.2692,
+      "step": 2817
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2.586860602372162e-05,
+      "loss": 0.3346,
+      "step": 2818
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2.5850522752468345e-05,
+      "loss": 0.0557,
+      "step": 2819
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2.58324410181558e-05,
+      "loss": 0.1108,
+      "step": 2820
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2.5814360827481342e-05,
+      "loss": 0.1635,
+      "step": 2821
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2.5796282187141765e-05,
+      "loss": 0.1227,
+      "step": 2822
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2.577820510383329e-05,
+      "loss": 0.1378,
+      "step": 2823
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2.5760129584251547e-05,
+      "loss": 0.0666,
+      "step": 2824
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2.5742055635091605e-05,
+      "loss": 0.0871,
+      "step": 2825
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2.5723983263047942e-05,
+      "loss": 0.0737,
+      "step": 2826
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.570591247481445e-05,
+      "loss": 0.1837,
+      "step": 2827
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.5687843277084432e-05,
+      "loss": 0.4125,
+      "step": 2828
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.566977567655062e-05,
+      "loss": 0.2905,
+      "step": 2829
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.565170967990514e-05,
+      "loss": 0.2337,
+      "step": 2830
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.563364529383951e-05,
+      "loss": 0.083,
+      "step": 2831
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.5615582525044685e-05,
+      "loss": 0.1048,
+      "step": 2832
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.5597521380211e-05,
+      "loss": 0.1531,
+      "step": 2833
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.557946186602818e-05,
+      "loss": 0.0955,
+      "step": 2834
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.5561403989185364e-05,
+      "loss": 0.2507,
+      "step": 2835
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.5543347756371082e-05,
+      "loss": 0.1372,
+      "step": 2836
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.5525293174273253e-05,
+      "loss": 0.1665,
+      "step": 2837
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.5507240249579167e-05,
+      "loss": 0.3247,
+      "step": 2838
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.5489188988975527e-05,
+      "loss": 0.2182,
+      "step": 2839
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2.5471139399148404e-05,
+      "loss": 0.0753,
+      "step": 2840
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2.5453091486783266e-05,
+      "loss": 0.1892,
+      "step": 2841
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2.5435045258564938e-05,
+      "loss": 0.0925,
+      "step": 2842
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2.541700072117764e-05,
+      "loss": 0.061,
+      "step": 2843
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2.5398957881304938e-05,
+      "loss": 0.0705,
+      "step": 2844
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2.5380916745629815e-05,
+      "loss": 0.2093,
+      "step": 2845
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2.5362877320834574e-05,
+      "loss": 0.2194,
+      "step": 2846
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2.534483961360091e-05,
+      "loss": 0.1499,
+      "step": 2847
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2.532680363060989e-05,
+      "loss": 0.2568,
+      "step": 2848
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2.5308769378541927e-05,
+      "loss": 0.077,
+      "step": 2849
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2.529073686407679e-05,
+      "loss": 0.2072,
+      "step": 2850
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2.5272706093893608e-05,
+      "loss": 0.0833,
+      "step": 2851
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2.5254677074670882e-05,
+      "loss": 0.09,
+      "step": 2852
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 2.5236649813086434e-05,
+      "loss": 0.2725,
+      "step": 2853
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 2.521862431581746e-05,
+      "loss": 0.2481,
+      "step": 2854
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 2.520060058954049e-05,
+      "loss": 0.2084,
+      "step": 2855
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 2.5182578640931395e-05,
+      "loss": 0.1727,
+      "step": 2856
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 2.51645584766654e-05,
+      "loss": 0.1294,
+      "step": 2857
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 2.5146540103417058e-05,
+      "loss": 0.059,
+      "step": 2858
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 2.5128523527860275e-05,
+      "loss": 0.163,
+      "step": 2859
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 2.5110508756668262e-05,
+      "loss": 0.0712,
+      "step": 2860
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 2.509249579651359e-05,
+      "loss": 0.1971,
+      "step": 2861
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 2.5074484654068147e-05,
+      "loss": 0.3466,
+      "step": 2862
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 2.5056475336003145e-05,
+      "loss": 0.4911,
+      "step": 2863
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 2.5038467848989123e-05,
+      "loss": 0.0668,
+      "step": 2864
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 2.502046219969595e-05,
+      "loss": 0.0883,
+      "step": 2865
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 2.500245839479281e-05,
+      "loss": 0.1612,
+      "step": 2866
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 2.4984456440948185e-05,
+      "loss": 0.0703,
+      "step": 2867
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 2.4966456344829896e-05,
+      "loss": 0.2028,
+      "step": 2868
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 2.4948458113105074e-05,
+      "loss": 0.26,
+      "step": 2869
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 2.4930461752440144e-05,
+      "loss": 0.1548,
+      "step": 2870
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 2.4912467269500848e-05,
+      "loss": 0.2949,
+      "step": 2871
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 2.4894474670952242e-05,
+      "loss": 0.0962,
+      "step": 2872
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 2.487648396345864e-05,
+      "loss": 0.0812,
+      "step": 2873
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 2.4858495153683736e-05,
+      "loss": 0.2167,
+      "step": 2874
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 2.4840508248290445e-05,
+      "loss": 0.2553,
+      "step": 2875
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 2.4822523253941007e-05,
+      "loss": 0.0806,
+      "step": 2876
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 2.480454017729696e-05,
+      "loss": 0.1696,
+      "step": 2877
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 2.478655902501913e-05,
+      "loss": 0.0706,
+      "step": 2878
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 2.476857980376761e-05,
+      "loss": 0.0709,
+      "step": 2879
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 2.47506025202018e-05,
+      "loss": 0.3292,
+      "step": 2880
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 2.473262718098038e-05,
+      "loss": 0.059,
+      "step": 2881
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 2.4714653792761294e-05,
+      "loss": 0.1249,
+      "step": 2882
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 2.4696682362201776e-05,
+      "loss": 0.2049,
+      "step": 2883
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 2.4678712895958337e-05,
+      "loss": 0.0985,
+      "step": 2884
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 2.466074540068675e-05,
+      "loss": 0.1655,
+      "step": 2885
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 2.464277988304207e-05,
+      "loss": 0.0704,
+      "step": 2886
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 2.4624816349678615e-05,
+      "loss": 0.214,
+      "step": 2887
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 2.4606854807249965e-05,
+      "loss": 0.0781,
+      "step": 2888
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 2.4588895262408965e-05,
+      "loss": 0.1426,
+      "step": 2889
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 2.4570937721807708e-05,
+      "loss": 0.4493,
+      "step": 2890
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 2.455298219209758e-05,
+      "loss": 0.136,
+      "step": 2891
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 2.453502867992917e-05,
+      "loss": 0.2229,
+      "step": 2892
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 2.4517077191952367e-05,
+      "loss": 0.2765,
+      "step": 2893
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 2.4499127734816285e-05,
+      "loss": 0.1678,
+      "step": 2894
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 2.44811803151693e-05,
+      "loss": 0.3031,
+      "step": 2895
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 2.446323493965901e-05,
+      "loss": 0.081,
+      "step": 2896
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 2.4445291614932282e-05,
+      "loss": 0.0986,
+      "step": 2897
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 2.442735034763522e-05,
+      "loss": 0.2692,
+      "step": 2898
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 2.4409411144413135e-05,
+      "loss": 0.2796,
+      "step": 2899
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 2.4391474011910615e-05,
+      "loss": 0.1834,
+      "step": 2900
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 2.4373538956771463e-05,
+      "loss": 0.066,
+      "step": 2901
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 2.4355605985638703e-05,
+      "loss": 0.0515,
+      "step": 2902
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 2.43376751051546e-05,
+      "loss": 0.3805,
+      "step": 2903
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 2.4319746321960645e-05,
+      "loss": 0.2,
+      "step": 2904
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 2.4301819642697555e-05,
+      "loss": 0.0755,
+      "step": 2905
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 2.428389507400523e-05,
+      "loss": 0.0857,
+      "step": 2906
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 2.4265972622522868e-05,
+      "loss": 0.221,
+      "step": 2907
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 2.4248052294888792e-05,
+      "loss": 0.231,
+      "step": 2908
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 2.423013409774061e-05,
+      "loss": 0.0674,
+      "step": 2909
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 2.4212218037715095e-05,
+      "loss": 0.104,
+      "step": 2910
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 2.419430412144825e-05,
+      "loss": 0.1078,
+      "step": 2911
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 2.4176392355575272e-05,
+      "loss": 0.357,
+      "step": 2912
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 2.4158482746730582e-05,
+      "loss": 0.1384,
+      "step": 2913
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 2.4140575301547773e-05,
+      "loss": 0.3137,
+      "step": 2914
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 2.4122670026659664e-05,
+      "loss": 0.1006,
+      "step": 2915
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 2.410476692869825e-05,
+      "loss": 0.0951,
+      "step": 2916
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.4086866014294736e-05,
+      "loss": 0.0577,
+      "step": 2917
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.4068967290079503e-05,
+      "loss": 0.2945,
+      "step": 2918
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.4051070762682125e-05,
+      "loss": 0.0651,
+      "step": 2919
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.4033176438731384e-05,
+      "loss": 0.0949,
+      "step": 2920
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.40152843248552e-05,
+      "loss": 0.126,
+      "step": 2921
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.399739442768072e-05,
+      "loss": 0.1068,
+      "step": 2922
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.3979506753834254e-05,
+      "loss": 0.0477,
+      "step": 2923
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.3961621309941267e-05,
+      "loss": 0.1824,
+      "step": 2924
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.3943738102626427e-05,
+      "loss": 0.073,
+      "step": 2925
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.392585713851357e-05,
+      "loss": 0.1703,
+      "step": 2926
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.3907978424225693e-05,
+      "loss": 0.1642,
+      "step": 2927
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.389010196638495e-05,
+      "loss": 0.0783,
+      "step": 2928
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.387222777161268e-05,
+      "loss": 0.1993,
+      "step": 2929
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.3854355846529373e-05,
+      "loss": 0.2087,
+      "step": 2930
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.3836486197754678e-05,
+      "loss": 0.2758,
+      "step": 2931
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.38186188319074e-05,
+      "loss": 0.3023,
+      "step": 2932
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.38007537556055e-05,
+      "loss": 0.1007,
+      "step": 2933
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.3782890975466112e-05,
+      "loss": 0.4228,
+      "step": 2934
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.376503049810547e-05,
+      "loss": 0.2433,
+      "step": 2935
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.3747172330139e-05,
+      "loss": 0.0678,
+      "step": 2936
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.372931647818126e-05,
+      "loss": 0.0717,
+      "step": 2937
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.371146294884594e-05,
+      "loss": 0.0739,
+      "step": 2938
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.3693611748745866e-05,
+      "loss": 0.0938,
+      "step": 2939
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.3675762884493034e-05,
+      "loss": 0.2542,
+      "step": 2940
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.3657916362698545e-05,
+      "loss": 0.4047,
+      "step": 2941
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.3640072189972645e-05,
+      "loss": 0.0704,
+      "step": 2942
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.3622230372924694e-05,
+      "loss": 0.0901,
+      "step": 2943
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.3604390918163196e-05,
+      "loss": 0.2623,
+      "step": 2944
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.3586553832295774e-05,
+      "loss": 0.0945,
+      "step": 2945
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.3568719121929185e-05,
+      "loss": 0.0726,
+      "step": 2946
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.355088679366928e-05,
+      "loss": 0.1178,
+      "step": 2947
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.3533056854121044e-05,
+      "loss": 0.1265,
+      "step": 2948
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.3515229309888594e-05,
+      "loss": 0.0866,
+      "step": 2949
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.349740416757512e-05,
+      "loss": 0.2318,
+      "step": 2950
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.3479581433782956e-05,
+      "loss": 0.0691,
+      "step": 2951
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.3461761115113536e-05,
+      "loss": 0.4023,
+      "step": 2952
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.3443943218167388e-05,
+      "loss": 0.2885,
+      "step": 2953
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.3426127749544147e-05,
+      "loss": 0.0836,
+      "step": 2954
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.3408314715842566e-05,
+      "loss": 0.0729,
+      "step": 2955
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.3390504123660478e-05,
+      "loss": 0.0991,
+      "step": 2956
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.337269597959481e-05,
+      "loss": 0.1652,
+      "step": 2957
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.335489029024159e-05,
+      "loss": 0.2315,
+      "step": 2958
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.3337087062195947e-05,
+      "loss": 0.32,
+      "step": 2959
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.3319286302052068e-05,
+      "loss": 0.2266,
+      "step": 2960
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.330148801640326e-05,
+      "loss": 0.1463,
+      "step": 2961
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.3283692211841887e-05,
+      "loss": 0.0821,
+      "step": 2962
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.326589889495942e-05,
+      "loss": 0.1516,
+      "step": 2963
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.3248108072346372e-05,
+      "loss": 0.138,
+      "step": 2964
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.323031975059237e-05,
+      "loss": 0.122,
+      "step": 2965
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.32125339362861e-05,
+      "loss": 0.4482,
+      "step": 2966
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.3194750636015302e-05,
+      "loss": 0.1657,
+      "step": 2967
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.3176969856366814e-05,
+      "loss": 0.183,
+      "step": 2968
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.3159191603926525e-05,
+      "loss": 0.2936,
+      "step": 2969
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.3141415885279388e-05,
+      "loss": 0.2359,
+      "step": 2970
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.3123642707009414e-05,
+      "loss": 0.4265,
+      "step": 2971
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.310587207569968e-05,
+      "loss": 0.0848,
+      "step": 2972
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.3088103997932326e-05,
+      "loss": 0.0592,
+      "step": 2973
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.307033848028853e-05,
+      "loss": 0.1835,
+      "step": 2974
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.3052575529348546e-05,
+      "loss": 0.274,
+      "step": 2975
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.303481515169164e-05,
+      "loss": 0.1245,
+      "step": 2976
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.3017057353896155e-05,
+      "loss": 0.1155,
+      "step": 2977
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.299930214253948e-05,
+      "loss": 0.6974,
+      "step": 2978
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.2981549524198012e-05,
+      "loss": 0.0777,
+      "step": 2979
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.296379950544723e-05,
+      "loss": 0.2531,
+      "step": 2980
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.294605209286163e-05,
+      "loss": 0.0918,
+      "step": 2981
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.2928307293014732e-05,
+      "loss": 0.067,
+      "step": 2982
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.2910565112479106e-05,
+      "loss": 0.1085,
+      "step": 2983
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.289282555782635e-05,
+      "loss": 0.0718,
+      "step": 2984
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.287508863562709e-05,
+      "loss": 0.073,
+      "step": 2985
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.285735435245096e-05,
+      "loss": 0.1006,
+      "step": 2986
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.2839622714866633e-05,
+      "loss": 0.3068,
+      "step": 2987
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.2821893729441806e-05,
+      "loss": 0.3124,
+      "step": 2988
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.280416740274317e-05,
+      "loss": 0.2031,
+      "step": 2989
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.278644374133646e-05,
+      "loss": 0.3403,
+      "step": 2990
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.2768722751786404e-05,
+      "loss": 0.1748,
+      "step": 2991
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.2751004440656755e-05,
+      "loss": 0.3148,
+      "step": 2992
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.2733288814510255e-05,
+      "loss": 0.7694,
+      "step": 2993
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.271557587990867e-05,
+      "loss": 0.0811,
+      "step": 2994
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.269786564341276e-05,
+      "loss": 0.1983,
+      "step": 2995
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.2680158111582288e-05,
+      "loss": 0.144,
+      "step": 2996
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.2662453290976004e-05,
+      "loss": 0.321,
+      "step": 2997
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.2644751188151686e-05,
+      "loss": 0.2437,
+      "step": 2998
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.2627051809666063e-05,
+      "loss": 0.1453,
+      "step": 2999
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.2609355162074883e-05,
+      "loss": 0.0961,
+      "step": 3000
+    },
+    {
+      "epoch": 2.32,
+      "eval_loss": 1.3273593187332153,
+      "eval_runtime": 157.4881,
+      "eval_samples_per_second": 0.337,
+      "eval_steps_per_second": 0.171,
+      "step": 3000
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.259166125193287e-05,
+      "loss": 0.1723,
+      "step": 3001
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.2573970085793754e-05,
+      "loss": 0.1552,
+      "step": 3002
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.2556281670210216e-05,
+      "loss": 0.2298,
+      "step": 3003
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.2538596011733948e-05,
+      "loss": 0.2557,
+      "step": 3004
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.252091311691559e-05,
+      "loss": 0.1939,
+      "step": 3005
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.25032329923048e-05,
+      "loss": 0.2718,
+      "step": 3006
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.2485555644450185e-05,
+      "loss": 0.1911,
+      "step": 3007
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.246788107989931e-05,
+      "loss": 0.2414,
+      "step": 3008
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.2450209305198733e-05,
+      "loss": 0.2775,
+      "step": 3009
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.2432540326893977e-05,
+      "loss": 0.1044,
+      "step": 3010
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.2414874151529505e-05,
+      "loss": 0.2079,
+      "step": 3011
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.2397210785648775e-05,
+      "loss": 0.3887,
+      "step": 3012
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.2379550235794174e-05,
+      "loss": 0.1086,
+      "step": 3013
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.2361892508507084e-05,
+      "loss": 0.068,
+      "step": 3014
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.234423761032779e-05,
+      "loss": 0.0768,
+      "step": 3015
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.2326585547795577e-05,
+      "loss": 0.1619,
+      "step": 3016
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.2308936327448657e-05,
+      "loss": 0.1565,
+      "step": 3017
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.229128995582418e-05,
+      "loss": 0.2929,
+      "step": 3018
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.2273646439458266e-05,
+      "loss": 0.0762,
+      "step": 3019
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.225600578488596e-05,
+      "loss": 0.0694,
+      "step": 3020
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.223836799864125e-05,
+      "loss": 0.0903,
+      "step": 3021
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.2220733087257064e-05,
+      "loss": 0.0897,
+      "step": 3022
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.2203101057265262e-05,
+      "loss": 0.2005,
+      "step": 3023
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.218547191519665e-05,
+      "loss": 0.0601,
+      "step": 3024
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.2167845667580935e-05,
+      "loss": 0.4054,
+      "step": 3025
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.2150222320946784e-05,
+      "loss": 0.1748,
+      "step": 3026
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.2132601881821778e-05,
+      "loss": 0.065,
+      "step": 3027
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.2114984356732402e-05,
+      "loss": 0.0983,
+      "step": 3028
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.209736975220409e-05,
+      "loss": 0.1382,
+      "step": 3029
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.207975807476118e-05,
+      "loss": 0.184,
+      "step": 3030
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.206214933092694e-05,
+      "loss": 0.0866,
+      "step": 3031
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.2044543527223518e-05,
+      "loss": 0.2523,
+      "step": 3032
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.202694067017201e-05,
+      "loss": 0.3453,
+      "step": 3033
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.200934076629241e-05,
+      "loss": 0.102,
+      "step": 3034
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.19917438221036e-05,
+      "loss": 0.1282,
+      "step": 3035
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.197414984412338e-05,
+      "loss": 0.0709,
+      "step": 3036
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.195655883886847e-05,
+      "loss": 0.1099,
+      "step": 3037
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.193897081285443e-05,
+      "loss": 0.2374,
+      "step": 3038
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.19213857725958e-05,
+      "loss": 0.1072,
+      "step": 3039
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.1903803724605946e-05,
+      "loss": 0.1142,
+      "step": 3040
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.1886224675397156e-05,
+      "loss": 0.1405,
+      "step": 3041
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.1868648631480603e-05,
+      "loss": 0.0771,
+      "step": 3042
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.1851075599366344e-05,
+      "loss": 0.1038,
+      "step": 3043
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.1833505585563323e-05,
+      "loss": 0.3828,
+      "step": 3044
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.1815938596579355e-05,
+      "loss": 0.0654,
+      "step": 3045
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.179837463892116e-05,
+      "loss": 0.5922,
+      "step": 3046
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.17808137190943e-05,
+      "loss": 0.0854,
+      "step": 3047
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.176325584360325e-05,
+      "loss": 0.0912,
+      "step": 3048
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.1745701018951326e-05,
+      "loss": 0.0748,
+      "step": 3049
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.1728149251640733e-05,
+      "loss": 0.1441,
+      "step": 3050
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.1710600548172532e-05,
+      "loss": 0.2416,
+      "step": 3051
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.169305491504666e-05,
+      "loss": 0.0882,
+      "step": 3052
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.1675512358761908e-05,
+      "loss": 0.0783,
+      "step": 3053
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.1657972885815928e-05,
+      "loss": 0.1622,
+      "step": 3054
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.1640436502705237e-05,
+      "loss": 0.351,
+      "step": 3055
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.1622903215925203e-05,
+      "loss": 0.0639,
+      "step": 3056
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.1605373031970043e-05,
+      "loss": 0.212,
+      "step": 3057
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.1587845957332825e-05,
+      "loss": 0.1296,
+      "step": 3058
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.157032199850548e-05,
+      "loss": 0.0957,
+      "step": 3059
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.1552801161978775e-05,
+      "loss": 0.2007,
+      "step": 3060
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.153528345424231e-05,
+      "loss": 0.2163,
+      "step": 3061
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.1517768881784532e-05,
+      "loss": 0.1819,
+      "step": 3062
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.150025745109275e-05,
+      "loss": 0.2169,
+      "step": 3063
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.1482749168653066e-05,
+      "loss": 0.3431,
+      "step": 3064
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.1465244040950454e-05,
+      "loss": 0.1702,
+      "step": 3065
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.144774207446871e-05,
+      "loss": 0.2981,
+      "step": 3066
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.143024327569044e-05,
+      "loss": 0.1777,
+      "step": 3067
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.1412747651097094e-05,
+      "loss": 0.1696,
+      "step": 3068
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.1395255207168946e-05,
+      "loss": 0.0747,
+      "step": 3069
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.13777659503851e-05,
+      "loss": 0.4287,
+      "step": 3070
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.1360279887223436e-05,
+      "loss": 0.2883,
+      "step": 3071
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.1342797024160724e-05,
+      "loss": 0.1705,
+      "step": 3072
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.1325317367672486e-05,
+      "loss": 0.3439,
+      "step": 3073
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.1307840924233084e-05,
+      "loss": 0.253,
+      "step": 3074
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.129036770031569e-05,
+      "loss": 0.1127,
+      "step": 3075
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.1272897702392266e-05,
+      "loss": 0.2546,
+      "step": 3076
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.125543093693359e-05,
+      "loss": 0.0698,
+      "step": 3077
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.1237967410409268e-05,
+      "loss": 0.3331,
+      "step": 3078
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.122050712928765e-05,
+      "loss": 0.1651,
+      "step": 3079
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.1203050100035936e-05,
+      "loss": 0.0993,
+      "step": 3080
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.1185596329120095e-05,
+      "loss": 0.2234,
+      "step": 3081
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.1168145823004907e-05,
+      "loss": 0.097,
+      "step": 3082
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.115069858815392e-05,
+      "loss": 0.1905,
+      "step": 3083
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.113325463102948e-05,
+      "loss": 0.1497,
+      "step": 3084
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.1115813958092734e-05,
+      "loss": 0.0992,
+      "step": 3085
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.109837657580358e-05,
+      "loss": 0.2294,
+      "step": 3086
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.1080942490620736e-05,
+      "loss": 0.0588,
+      "step": 3087
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.1063511709001664e-05,
+      "loss": 0.2288,
+      "step": 3088
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.1046084237402635e-05,
+      "loss": 0.226,
+      "step": 3089
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.1028660082278658e-05,
+      "loss": 0.0719,
+      "step": 3090
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.1011239250083543e-05,
+      "loss": 0.2841,
+      "step": 3091
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.0993821747269864e-05,
+      "loss": 0.2446,
+      "step": 3092
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.097640758028894e-05,
+      "loss": 0.0873,
+      "step": 3093
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.095899675559088e-05,
+      "loss": 0.0988,
+      "step": 3094
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.094158927962456e-05,
+      "loss": 0.0895,
+      "step": 3095
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.0924185158837574e-05,
+      "loss": 0.3384,
+      "step": 3096
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.0906784399676315e-05,
+      "loss": 0.23,
+      "step": 3097
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.0889387008585917e-05,
+      "loss": 0.2289,
+      "step": 3098
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.0871992992010267e-05,
+      "loss": 0.1357,
+      "step": 3099
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.0854602356391997e-05,
+      "loss": 0.1437,
+      "step": 3100
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.0837215108172493e-05,
+      "loss": 0.0784,
+      "step": 3101
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.081983125379189e-05,
+      "loss": 0.0924,
+      "step": 3102
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.0802450799689046e-05,
+      "loss": 0.1043,
+      "step": 3103
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.0785073752301583e-05,
+      "loss": 0.0828,
+      "step": 3104
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.076770011806585e-05,
+      "loss": 0.1374,
+      "step": 3105
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.075032990341694e-05,
+      "loss": 0.185,
+      "step": 3106
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.073296311478867e-05,
+      "loss": 0.0751,
+      "step": 3107
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.0715599758613583e-05,
+      "loss": 0.1192,
+      "step": 3108
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.069823984132297e-05,
+      "loss": 0.2434,
+      "step": 3109
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.0680883369346834e-05,
+      "loss": 0.1324,
+      "step": 3110
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2.06635303491139e-05,
+      "loss": 0.0576,
+      "step": 3111
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2.0646180787051625e-05,
+      "loss": 0.0769,
+      "step": 3112
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2.0628834689586176e-05,
+      "loss": 0.0684,
+      "step": 3113
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2.0611492063142443e-05,
+      "loss": 0.1483,
+      "step": 3114
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2.0594152914144024e-05,
+      "loss": 0.1853,
+      "step": 3115
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2.0576817249013235e-05,
+      "loss": 0.1718,
+      "step": 3116
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2.0559485074171098e-05,
+      "loss": 0.235,
+      "step": 3117
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2.054215639603735e-05,
+      "loss": 0.0955,
+      "step": 3118
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2.0524831221030417e-05,
+      "loss": 0.0791,
+      "step": 3119
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2.050750955556744e-05,
+      "loss": 0.2232,
+      "step": 3120
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2.049019140606426e-05,
+      "loss": 0.2791,
+      "step": 3121
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2.0472876778935408e-05,
+      "loss": 0.0786,
+      "step": 3122
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2.0455565680594116e-05,
+      "loss": 0.1062,
+      "step": 3123
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 2.043825811745231e-05,
+      "loss": 0.075,
+      "step": 3124
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 2.0420954095920607e-05,
+      "loss": 0.0555,
+      "step": 3125
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 2.04036536224083e-05,
+      "loss": 0.1225,
+      "step": 3126
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 2.038635670332338e-05,
+      "loss": 0.0776,
+      "step": 3127
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 2.0369063345072532e-05,
+      "loss": 0.0453,
+      "step": 3128
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 2.0351773554061093e-05,
+      "loss": 0.2103,
+      "step": 3129
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 2.03344873366931e-05,
+      "loss": 0.0896,
+      "step": 3130
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 2.031720469937127e-05,
+      "loss": 0.0809,
+      "step": 3131
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 2.029992564849697e-05,
+      "loss": 0.0831,
+      "step": 3132
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 2.0282650190470257e-05,
+      "loss": 0.068,
+      "step": 3133
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 2.0265378331689863e-05,
+      "loss": 0.0933,
+      "step": 3134
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 2.0248110078553176e-05,
+      "loss": 0.3988,
+      "step": 3135
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 2.023084543745624e-05,
+      "loss": 0.2042,
+      "step": 3136
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2.021358441479377e-05,
+      "loss": 0.3622,
+      "step": 3137
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2.019632701695916e-05,
+      "loss": 0.2203,
+      "step": 3138
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2.017907325034443e-05,
+      "loss": 0.5253,
+      "step": 3139
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2.016182312134028e-05,
+      "loss": 0.0786,
+      "step": 3140
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2.0144576636336037e-05,
+      "loss": 0.3342,
+      "step": 3141
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2.01273338017197e-05,
+      "loss": 0.0764,
+      "step": 3142
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2.0110094623877912e-05,
+      "loss": 0.0923,
+      "step": 3143
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2.0092859109195943e-05,
+      "loss": 0.1382,
+      "step": 3144
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2.0075627264057736e-05,
+      "loss": 0.2533,
+      "step": 3145
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2.0058399094845855e-05,
+      "loss": 0.0798,
+      "step": 3146
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2.004117460794151e-05,
+      "loss": 0.0745,
+      "step": 3147
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2.002395380972454e-05,
+      "loss": 0.1892,
+      "step": 3148
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2.0006736706573422e-05,
+      "loss": 0.1921,
+      "step": 3149
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.9989523304865273e-05,
+      "loss": 0.2171,
+      "step": 3150
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.9972313610975816e-05,
+      "loss": 0.1315,
+      "step": 3151
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.995510763127943e-05,
+      "loss": 0.1611,
+      "step": 3152
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.9937905372149097e-05,
+      "loss": 0.6706,
+      "step": 3153
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.9920706839956424e-05,
+      "loss": 0.0748,
+      "step": 3154
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.9903512041071645e-05,
+      "loss": 0.14,
+      "step": 3155
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.9886320981863605e-05,
+      "loss": 0.1479,
+      "step": 3156
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.9869133668699775e-05,
+      "loss": 0.2619,
+      "step": 3157
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.9851950107946216e-05,
+      "loss": 0.1033,
+      "step": 3158
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.983477030596762e-05,
+      "loss": 0.0841,
+      "step": 3159
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.981759426912729e-05,
+      "loss": 0.3937,
+      "step": 3160
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.98004220037871e-05,
+      "loss": 0.116,
+      "step": 3161
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.9783253516307563e-05,
+      "loss": 0.1884,
+      "step": 3162
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.976608881304779e-05,
+      "loss": 0.2441,
+      "step": 3163
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.974892790036547e-05,
+      "loss": 0.2476,
+      "step": 3164
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.9731770784616893e-05,
+      "loss": 0.0729,
+      "step": 3165
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.971461747215696e-05,
+      "loss": 0.2304,
+      "step": 3166
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.9697467969339157e-05,
+      "loss": 0.1317,
+      "step": 3167
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.9680322282515536e-05,
+      "loss": 0.0801,
+      "step": 3168
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.9663180418036764e-05,
+      "loss": 0.2121,
+      "step": 3169
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.964604238225209e-05,
+      "loss": 0.1045,
+      "step": 3170
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.9628908181509308e-05,
+      "loss": 0.2674,
+      "step": 3171
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.9611777822154856e-05,
+      "loss": 0.1462,
+      "step": 3172
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.9594651310533693e-05,
+      "loss": 0.2202,
+      "step": 3173
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.9577528652989375e-05,
+      "loss": 0.1902,
+      "step": 3174
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.956040985586403e-05,
+      "loss": 0.2109,
+      "step": 3175
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.954329492549836e-05,
+      "loss": 0.2508,
+      "step": 3176
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.952618386823162e-05,
+      "loss": 0.1905,
+      "step": 3177
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.9509076690401638e-05,
+      "loss": 0.0755,
+      "step": 3178
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.9491973398344828e-05,
+      "loss": 0.0855,
+      "step": 3179
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.947487399839612e-05,
+      "loss": 0.2709,
+      "step": 3180
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.9457778496889034e-05,
+      "loss": 0.0929,
+      "step": 3181
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.9440686900155645e-05,
+      "loss": 0.2903,
+      "step": 3182
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.9423599214526563e-05,
+      "loss": 0.3561,
+      "step": 3183
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.9406515446330966e-05,
+      "loss": 0.0733,
+      "step": 3184
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.9389435601896578e-05,
+      "loss": 0.0808,
+      "step": 3185
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.937235968754967e-05,
+      "loss": 0.2955,
+      "step": 3186
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.935528770961505e-05,
+      "loss": 0.0737,
+      "step": 3187
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.933821967441607e-05,
+      "loss": 0.3318,
+      "step": 3188
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.9321155588274642e-05,
+      "loss": 0.051,
+      "step": 3189
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.930409545751118e-05,
+      "loss": 0.1765,
+      "step": 3190
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.9287039288444653e-05,
+      "loss": 0.0829,
+      "step": 3191
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.926998708739257e-05,
+      "loss": 0.1437,
+      "step": 3192
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.925293886067095e-05,
+      "loss": 0.1366,
+      "step": 3193
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.9235894614594357e-05,
+      "loss": 0.0848,
+      "step": 3194
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.9218854355475873e-05,
+      "loss": 0.2512,
+      "step": 3195
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.9201818089627113e-05,
+      "loss": 0.091,
+      "step": 3196
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.9184785823358192e-05,
+      "loss": 0.2447,
+      "step": 3197
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.9167757562977757e-05,
+      "loss": 0.0862,
+      "step": 3198
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.9150733314792984e-05,
+      "loss": 0.1373,
+      "step": 3199
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.9133713085109533e-05,
+      "loss": 0.3409,
+      "step": 3200
+    },
+    {
+      "epoch": 2.47,
+      "eval_loss": 1.3230520486831665,
+      "eval_runtime": 157.2251,
+      "eval_samples_per_second": 0.337,
+      "eval_steps_per_second": 0.172,
+      "step": 3200
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.9116696880231596e-05,
+      "loss": 0.1011,
+      "step": 3201
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.9099684706461876e-05,
+      "loss": 0.0778,
+      "step": 3202
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.9082676570101577e-05,
+      "loss": 0.4023,
+      "step": 3203
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.9065672477450385e-05,
+      "loss": 0.248,
+      "step": 3204
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.904867243480655e-05,
+      "loss": 0.1485,
+      "step": 3205
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.903167644846676e-05,
+      "loss": 0.0714,
+      "step": 3206
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.9014684524726224e-05,
+      "loss": 0.1055,
+      "step": 3207
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.899769666987865e-05,
+      "loss": 0.2482,
+      "step": 3208
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.8980712890216234e-05,
+      "loss": 0.0787,
+      "step": 3209
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.8963733192029657e-05,
+      "loss": 0.077,
+      "step": 3210
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.8946757581608104e-05,
+      "loss": 0.2437,
+      "step": 3211
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.8929786065239228e-05,
+      "loss": 0.3075,
+      "step": 3212
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.891281864920917e-05,
+      "loss": 0.1919,
+      "step": 3213
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.8895855339802568e-05,
+      "loss": 0.0977,
+      "step": 3214
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.8878896143302524e-05,
+      "loss": 0.0916,
+      "step": 3215
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.8861941065990614e-05,
+      "loss": 0.1722,
+      "step": 3216
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.8844990114146894e-05,
+      "loss": 0.0864,
+      "step": 3217
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.8828043294049904e-05,
+      "loss": 0.1843,
+      "step": 3218
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.8811100611976622e-05,
+      "loss": 0.0699,
+      "step": 3219
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.879416207420253e-05,
+      "loss": 0.1021,
+      "step": 3220
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.877722768700155e-05,
+      "loss": 0.0781,
+      "step": 3221
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.8760297456646066e-05,
+      "loss": 0.1235,
+      "step": 3222
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.874337138940695e-05,
+      "loss": 0.3711,
+      "step": 3223
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.8726449491553495e-05,
+      "loss": 0.3873,
+      "step": 3224
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.8709531769353483e-05,
+      "loss": 0.2503,
+      "step": 3225
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.8692618229073115e-05,
+      "loss": 0.1566,
+      "step": 3226
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.867570887697708e-05,
+      "loss": 0.083,
+      "step": 3227
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.8658803719328487e-05,
+      "loss": 0.1876,
+      "step": 3228
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.8641902762388905e-05,
+      "loss": 0.094,
+      "step": 3229
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.8625006012418345e-05,
+      "loss": 0.1023,
+      "step": 3230
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.8608113475675253e-05,
+      "loss": 0.077,
+      "step": 3231
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.859122515841653e-05,
+      "loss": 0.0864,
+      "step": 3232
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.8574341066897495e-05,
+      "loss": 0.0478,
+      "step": 3233
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.8557461207371914e-05,
+      "loss": 0.4015,
+      "step": 3234
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.8540585586091995e-05,
+      "loss": 0.2079,
+      "step": 3235
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.852371420930834e-05,
+      "loss": 0.0686,
+      "step": 3236
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.850684708327001e-05,
+      "loss": 0.0666,
+      "step": 3237
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.8489984214224496e-05,
+      "loss": 0.1152,
+      "step": 3238
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.8473125608417694e-05,
+      "loss": 0.073,
+      "step": 3239
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.8456271272093934e-05,
+      "loss": 0.1897,
+      "step": 3240
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.8439421211495947e-05,
+      "loss": 0.06,
+      "step": 3241
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.8422575432864895e-05,
+      "loss": 0.0819,
+      "step": 3242
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.840573394244035e-05,
+      "loss": 0.2292,
+      "step": 3243
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.8388896746460302e-05,
+      "loss": 0.3146,
+      "step": 3244
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.8372063851161134e-05,
+      "loss": 0.2472,
+      "step": 3245
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.8355235262777653e-05,
+      "loss": 0.2725,
+      "step": 3246
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.833841098754307e-05,
+      "loss": 0.2335,
+      "step": 3247
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.832159103168898e-05,
+      "loss": 0.1585,
+      "step": 3248
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.830477540144539e-05,
+      "loss": 0.0876,
+      "step": 3249
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.828796410304072e-05,
+      "loss": 0.2563,
+      "step": 3250
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.8271157142701756e-05,
+      "loss": 0.2114,
+      "step": 3251
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.8254354526653702e-05,
+      "loss": 0.4051,
+      "step": 3252
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.8237556261120138e-05,
+      "loss": 0.516,
+      "step": 3253
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.822076235232305e-05,
+      "loss": 0.1063,
+      "step": 3254
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.8203972806482777e-05,
+      "loss": 0.1182,
+      "step": 3255
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.8187187629818072e-05,
+      "loss": 0.082,
+      "step": 3256
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.817040682854607e-05,
+      "loss": 0.1716,
+      "step": 3257
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.8153630408882266e-05,
+      "loss": 0.0932,
+      "step": 3258
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.8136858377040542e-05,
+      "loss": 0.1958,
+      "step": 3259
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.8120090739233157e-05,
+      "loss": 0.0496,
+      "step": 3260
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.810332750167075e-05,
+      "loss": 0.0691,
+      "step": 3261
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.8086568670562306e-05,
+      "loss": 0.2984,
+      "step": 3262
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.8069814252115195e-05,
+      "loss": 0.0616,
+      "step": 3263
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.8053064252535158e-05,
+      "loss": 0.3767,
+      "step": 3264
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.8036318678026287e-05,
+      "loss": 0.0739,
+      "step": 3265
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.8019577534791032e-05,
+      "loss": 0.2489,
+      "step": 3266
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.8002840829030223e-05,
+      "loss": 0.0566,
+      "step": 3267
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.798610856694302e-05,
+      "loss": 0.0488,
+      "step": 3268
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.796938075472696e-05,
+      "loss": 0.1294,
+      "step": 3269
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.7952657398577903e-05,
+      "loss": 0.0782,
+      "step": 3270
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.7935938504690097e-05,
+      "loss": 0.0945,
+      "step": 3271
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.791922407925611e-05,
+      "loss": 0.1697,
+      "step": 3272
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.7902514128466866e-05,
+      "loss": 0.0736,
+      "step": 3273
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.7885808658511622e-05,
+      "loss": 0.2839,
+      "step": 3274
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.786910767557798e-05,
+      "loss": 0.2255,
+      "step": 3275
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.7852411185851882e-05,
+      "loss": 0.0594,
+      "step": 3276
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.7835719195517608e-05,
+      "loss": 0.1362,
+      "step": 3277
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.781903171075776e-05,
+      "loss": 0.0903,
+      "step": 3278
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.7802348737753295e-05,
+      "loss": 0.0996,
+      "step": 3279
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.7785670282683463e-05,
+      "loss": 0.6917,
+      "step": 3280
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.7768996351725872e-05,
+      "loss": 0.2735,
+      "step": 3281
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.775232695105644e-05,
+      "loss": 0.1431,
+      "step": 3282
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.773566208684942e-05,
+      "loss": 0.1124,
+      "step": 3283
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.7719001765277358e-05,
+      "loss": 0.2111,
+      "step": 3284
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.7702345992511142e-05,
+      "loss": 0.0774,
+      "step": 3285
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.7685694774719975e-05,
+      "loss": 0.0884,
+      "step": 3286
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.7669048118071346e-05,
+      "loss": 0.1639,
+      "step": 3287
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.765240602873109e-05,
+      "loss": 0.2004,
+      "step": 3288
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.763576851286333e-05,
+      "loss": 0.3478,
+      "step": 3289
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.7619135576630507e-05,
+      "loss": 0.1315,
+      "step": 3290
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.760250722619334e-05,
+      "loss": 0.2456,
+      "step": 3291
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.7585883467710876e-05,
+      "loss": 0.0815,
+      "step": 3292
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.7569264307340456e-05,
+      "loss": 0.0696,
+      "step": 3293
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.755264975123771e-05,
+      "loss": 0.2947,
+      "step": 3294
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.7536039805556572e-05,
+      "loss": 0.1581,
+      "step": 3295
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.7519434476449257e-05,
+      "loss": 0.1296,
+      "step": 3296
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.7502833770066277e-05,
+      "loss": 0.2138,
+      "step": 3297
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.7486237692556438e-05,
+      "loss": 0.3733,
+      "step": 3298
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.7469646250066805e-05,
+      "loss": 0.0977,
+      "step": 3299
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.7453059448742773e-05,
+      "loss": 0.2693,
+      "step": 3300
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.743647729472798e-05,
+      "loss": 0.1173,
+      "step": 3301
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.741989979416434e-05,
+      "loss": 0.184,
+      "step": 3302
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.7403326953192062e-05,
+      "loss": 0.3727,
+      "step": 3303
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.7386758777949644e-05,
+      "loss": 0.0421,
+      "step": 3304
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.737019527457381e-05,
+      "loss": 0.0613,
+      "step": 3305
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.73536364491996e-05,
+      "loss": 0.2354,
+      "step": 3306
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.7337082307960293e-05,
+      "loss": 0.0949,
+      "step": 3307
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.732053285698743e-05,
+      "loss": 0.0669,
+      "step": 3308
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.7303988102410843e-05,
+      "loss": 0.1827,
+      "step": 3309
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.7287448050358597e-05,
+      "loss": 0.0744,
+      "step": 3310
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.7270912706957015e-05,
+      "loss": 0.1904,
+      "step": 3311
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.7254382078330713e-05,
+      "loss": 0.2492,
+      "step": 3312
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.723785617060251e-05,
+      "loss": 0.1279,
+      "step": 3313
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.72213349898935e-05,
+      "loss": 0.1661,
+      "step": 3314
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.720481854232305e-05,
+      "loss": 0.2589,
+      "step": 3315
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.7188306834008717e-05,
+      "loss": 0.2248,
+      "step": 3316
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.7171799871066366e-05,
+      "loss": 0.2255,
+      "step": 3317
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.7155297659610062e-05,
+      "loss": 0.1073,
+      "step": 3318
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.713880020575211e-05,
+      "loss": 0.2789,
+      "step": 3319
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.7122307515603086e-05,
+      "loss": 0.064,
+      "step": 3320
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.7105819595271774e-05,
+      "loss": 0.1793,
+      "step": 3321
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.7089336450865176e-05,
+      "loss": 0.0611,
+      "step": 3322
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.707285808848858e-05,
+      "loss": 0.2666,
+      "step": 3323
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.7056384514245453e-05,
+      "loss": 0.0812,
+      "step": 3324
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.7039915734237497e-05,
+      "loss": 0.0617,
+      "step": 3325
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.7023451754564668e-05,
+      "loss": 0.1498,
+      "step": 3326
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.700699258132511e-05,
+      "loss": 0.1946,
+      "step": 3327
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.699053822061519e-05,
+      "loss": 0.0961,
+      "step": 3328
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.6974088678529523e-05,
+      "loss": 0.2279,
+      "step": 3329
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.69576439611609e-05,
+      "loss": 0.0762,
+      "step": 3330
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.694120407460034e-05,
+      "loss": 0.0895,
+      "step": 3331
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.69247690249371e-05,
+      "loss": 0.284,
+      "step": 3332
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.6908338818258587e-05,
+      "loss": 0.0804,
+      "step": 3333
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.6891913460650478e-05,
+      "loss": 0.0552,
+      "step": 3334
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.6875492958196612e-05,
+      "loss": 0.1173,
+      "step": 3335
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.685907731697903e-05,
+      "loss": 0.1714,
+      "step": 3336
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.6842666543077997e-05,
+      "loss": 0.0677,
+      "step": 3337
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.682626064257197e-05,
+      "loss": 0.0778,
+      "step": 3338
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.6809859621537585e-05,
+      "loss": 0.1278,
+      "step": 3339
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.6793463486049663e-05,
+      "loss": 0.227,
+      "step": 3340
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.677707224218126e-05,
+      "loss": 0.0912,
+      "step": 3341
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.676068589600358e-05,
+      "loss": 0.2227,
+      "step": 3342
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.6744304453586004e-05,
+      "loss": 0.3174,
+      "step": 3343
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 1.6727927920996152e-05,
+      "loss": 0.2406,
+      "step": 3344
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 1.6711556304299755e-05,
+      "loss": 0.2413,
+      "step": 3345
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 1.669518960956079e-05,
+      "loss": 0.2974,
+      "step": 3346
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 1.6678827842841365e-05,
+      "loss": 0.1321,
+      "step": 3347
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 1.6662471010201772e-05,
+      "loss": 0.1902,
+      "step": 3348
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 1.6646119117700495e-05,
+      "loss": 0.1185,
+      "step": 3349
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 1.662977217139417e-05,
+      "loss": 0.2181,
+      "step": 3350
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 1.6613430177337587e-05,
+      "loss": 0.2532,
+      "step": 3351
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 1.6597093141583744e-05,
+      "loss": 0.0955,
+      "step": 3352
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 1.6580761070183766e-05,
+      "loss": 0.088,
+      "step": 3353
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 1.6564433969186947e-05,
+      "loss": 0.0716,
+      "step": 3354
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 1.654811184464076e-05,
+      "loss": 0.3373,
+      "step": 3355
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 1.6531794702590804e-05,
+      "loss": 0.1323,
+      "step": 3356
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1.6515482549080847e-05,
+      "loss": 0.099,
+      "step": 3357
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1.649917539015283e-05,
+      "loss": 0.0955,
+      "step": 3358
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1.6482873231846803e-05,
+      "loss": 0.0633,
+      "step": 3359
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1.6466576080200983e-05,
+      "loss": 0.0993,
+      "step": 3360
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1.645028394125175e-05,
+      "loss": 0.0981,
+      "step": 3361
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1.6433996821033597e-05,
+      "loss": 0.0853,
+      "step": 3362
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1.6417714725579192e-05,
+      "loss": 0.1906,
+      "step": 3363
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1.640143766091931e-05,
+      "loss": 0.1009,
+      "step": 3364
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1.6385165633082865e-05,
+      "loss": 0.1372,
+      "step": 3365
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1.6368898648096932e-05,
+      "loss": 0.1429,
+      "step": 3366
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1.63526367119867e-05,
+      "loss": 0.1065,
+      "step": 3367
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1.6336379830775472e-05,
+      "loss": 0.0712,
+      "step": 3368
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1.6320128010484712e-05,
+      "loss": 0.1571,
+      "step": 3369
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.6303881257134005e-05,
+      "loss": 0.078,
+      "step": 3370
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.628763957674103e-05,
+      "loss": 0.1891,
+      "step": 3371
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.62714029753216e-05,
+      "loss": 0.1615,
+      "step": 3372
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.625517145888967e-05,
+      "loss": 0.1249,
+      "step": 3373
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.623894503345728e-05,
+      "loss": 0.1303,
+      "step": 3374
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.6222723705034605e-05,
+      "loss": 0.1561,
+      "step": 3375
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.620650747962993e-05,
+      "loss": 0.1183,
+      "step": 3376
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.6190296363249624e-05,
+      "loss": 0.1086,
+      "step": 3377
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.6174090361898206e-05,
+      "loss": 0.0906,
+      "step": 3378
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.615788948157827e-05,
+      "loss": 0.0899,
+      "step": 3379
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.6141693728290515e-05,
+      "loss": 0.0841,
+      "step": 3380
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.6125503108033762e-05,
+      "loss": 0.2699,
+      "step": 3381
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.6109317626804908e-05,
+      "loss": 0.0825,
+      "step": 3382
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 1.609313729059895e-05,
+      "loss": 0.1161,
+      "step": 3383
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 1.6076962105409003e-05,
+      "loss": 0.1864,
+      "step": 3384
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 1.606079207722624e-05,
+      "loss": 0.0722,
+      "step": 3385
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 1.6044627212039936e-05,
+      "loss": 0.0784,
+      "step": 3386
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 1.602846751583747e-05,
+      "loss": 0.0833,
+      "step": 3387
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 1.6012312994604293e-05,
+      "loss": 0.098,
+      "step": 3388
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 1.5996163654323922e-05,
+      "loss": 0.0593,
+      "step": 3389
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 1.5980019500977997e-05,
+      "loss": 0.453,
+      "step": 3390
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 1.596388054054619e-05,
+      "loss": 0.5214,
+      "step": 3391
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 1.59477467790063e-05,
+      "loss": 0.0774,
+      "step": 3392
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 1.593161822233415e-05,
+      "loss": 0.0879,
+      "step": 3393
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 1.5915494876503657e-05,
+      "loss": 0.06,
+      "step": 3394
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 1.5899376747486823e-05,
+      "loss": 0.2209,
+      "step": 3395
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 1.5883263841253702e-05,
+      "loss": 0.1389,
+      "step": 3396
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 1.5867156163772393e-05,
+      "loss": 0.1577,
+      "step": 3397
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 1.5851053721009105e-05,
+      "loss": 0.0708,
+      "step": 3398
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 1.5834956518928077e-05,
+      "loss": 0.072,
+      "step": 3399
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 1.58188645634916e-05,
+      "loss": 0.0947,
+      "step": 3400
+    },
+    {
+      "epoch": 2.63,
+      "eval_loss": 1.3287853002548218,
+      "eval_runtime": 157.468,
+      "eval_samples_per_second": 0.337,
+      "eval_steps_per_second": 0.171,
+      "step": 3400
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 1.5802777860660055e-05,
+      "loss": 0.3136,
+      "step": 3401
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 1.5786696416391837e-05,
+      "loss": 0.0783,
+      "step": 3402
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 1.5770620236643427e-05,
+      "loss": 0.0828,
+      "step": 3403
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 1.5754549327369347e-05,
+      "loss": 0.2324,
+      "step": 3404
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 1.5738483694522153e-05,
+      "loss": 0.0846,
+      "step": 3405
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 1.5722423344052454e-05,
+      "loss": 0.1588,
+      "step": 3406
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 1.5706368281908918e-05,
+      "loss": 0.2533,
+      "step": 3407
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 1.569031851403823e-05,
+      "loss": 0.1628,
+      "step": 3408
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 1.567427404638512e-05,
+      "loss": 0.207,
+      "step": 3409
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 1.565823488489237e-05,
+      "loss": 0.1033,
+      "step": 3410
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 1.564220103550079e-05,
+      "loss": 0.1558,
+      "step": 3411
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 1.5626172504149195e-05,
+      "loss": 0.122,
+      "step": 3412
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 1.5610149296774482e-05,
+      "loss": 0.0602,
+      "step": 3413
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 1.5594131419311543e-05,
+      "loss": 0.0948,
+      "step": 3414
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 1.557811887769327e-05,
+      "loss": 0.1055,
+      "step": 3415
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 1.5562111677850657e-05,
+      "loss": 0.1492,
+      "step": 3416
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 1.5546109825712645e-05,
+      "loss": 0.0761,
+      "step": 3417
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 1.5530113327206206e-05,
+      "loss": 0.0528,
+      "step": 3418
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 1.5514122188256384e-05,
+      "loss": 0.12,
+      "step": 3419
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 1.5498136414786158e-05,
+      "loss": 0.2034,
+      "step": 3420
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 1.5482156012716592e-05,
+      "loss": 0.3674,
+      "step": 3421
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 1.5466180987966717e-05,
+      "loss": 0.3642,
+      "step": 3422
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 1.545021134645357e-05,
+      "loss": 0.1668,
+      "step": 3423
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 1.5434247094092225e-05,
+      "loss": 0.4341,
+      "step": 3424
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 1.541828823679574e-05,
+      "loss": 0.088,
+      "step": 3425
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 1.540233478047516e-05,
+      "loss": 0.1685,
+      "step": 3426
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 1.5386386731039572e-05,
+      "loss": 0.3558,
+      "step": 3427
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 1.5370444094396023e-05,
+      "loss": 0.0497,
+      "step": 3428
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 1.5354506876449554e-05,
+      "loss": 0.2313,
+      "step": 3429
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 1.5338575083103237e-05,
+      "loss": 0.2295,
+      "step": 3430
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 1.53226487202581e-05,
+      "loss": 0.1659,
+      "step": 3431
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 1.5306727793813153e-05,
+      "loss": 0.2008,
+      "step": 3432
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 1.5290812309665432e-05,
+      "loss": 0.1445,
+      "step": 3433
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 1.527490227370993e-05,
+      "loss": 0.5688,
+      "step": 3434
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 1.5258997691839594e-05,
+      "loss": 0.0784,
+      "step": 3435
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 1.5243098569945434e-05,
+      "loss": 0.1616,
+      "step": 3436
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 1.5227204913916358e-05,
+      "loss": 0.0833,
+      "step": 3437
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 1.5211316729639273e-05,
+      "loss": 0.1616,
+      "step": 3438
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 1.5195434022999086e-05,
+      "loss": 0.2477,
+      "step": 3439
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 1.5179556799878637e-05,
+      "loss": 0.0812,
+      "step": 3440
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 1.516368506615875e-05,
+      "loss": 0.0936,
+      "step": 3441
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 1.5147818827718228e-05,
+      "loss": 0.0685,
+      "step": 3442
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 1.5131958090433822e-05,
+      "loss": 0.0779,
+      "step": 3443
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 1.5116102860180242e-05,
+      "loss": 0.0759,
+      "step": 3444
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 1.5100253142830187e-05,
+      "loss": 0.0772,
+      "step": 3445
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 1.508440894425428e-05,
+      "loss": 0.0949,
+      "step": 3446
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 1.5068570270321112e-05,
+      "loss": 0.0699,
+      "step": 3447
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 1.5052737126897242e-05,
+      "loss": 0.2139,
+      "step": 3448
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 1.503690951984715e-05,
+      "loss": 0.3656,
+      "step": 3449
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 1.5021087455033306e-05,
+      "loss": 0.0677,
+      "step": 3450
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 1.5005270938316102e-05,
+      "loss": 0.0733,
+      "step": 3451
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 1.4989459975553855e-05,
+      "loss": 0.1765,
+      "step": 3452
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 1.4973654572602874e-05,
+      "loss": 0.3897,
+      "step": 3453
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 1.4957854735317376e-05,
+      "loss": 0.2022,
+      "step": 3454
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 1.4942060469549505e-05,
+      "loss": 0.3213,
+      "step": 3455
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 1.4926271781149385e-05,
+      "loss": 0.1359,
+      "step": 3456
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 1.4910488675965041e-05,
+      "loss": 0.3397,
+      "step": 3457
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 1.489471115984242e-05,
+      "loss": 0.2301,
+      "step": 3458
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 1.4878939238625442e-05,
+      "loss": 0.3092,
+      "step": 3459
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 1.4863172918155916e-05,
+      "loss": 0.1339,
+      "step": 3460
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 1.4847412204273582e-05,
+      "loss": 0.1983,
+      "step": 3461
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 1.4831657102816132e-05,
+      "loss": 0.1647,
+      "step": 3462
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 1.4815907619619145e-05,
+      "loss": 0.0713,
+      "step": 3463
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 1.4800163760516123e-05,
+      "loss": 0.0952,
+      "step": 3464
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 1.478442553133852e-05,
+      "loss": 0.1731,
+      "step": 3465
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 1.476869293791566e-05,
+      "loss": 0.1661,
+      "step": 3466
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 1.4752965986074791e-05,
+      "loss": 0.2658,
+      "step": 3467
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 1.4737244681641091e-05,
+      "loss": 0.0696,
+      "step": 3468
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 1.4721529030437642e-05,
+      "loss": 0.4072,
+      "step": 3469
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 1.4705819038285406e-05,
+      "loss": 0.1426,
+      "step": 3470
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 1.4690114711003288e-05,
+      "loss": 0.1779,
+      "step": 3471
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 1.4674416054408059e-05,
+      "loss": 0.0615,
+      "step": 3472
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 1.4658723074314395e-05,
+      "loss": 0.2154,
+      "step": 3473
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 1.4643035776534904e-05,
+      "loss": 0.1561,
+      "step": 3474
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 1.4627354166880052e-05,
+      "loss": 0.2022,
+      "step": 3475
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 1.4611678251158197e-05,
+      "loss": 0.2369,
+      "step": 3476
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 1.4596008035175626e-05,
+      "loss": 0.0845,
+      "step": 3477
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 1.4580343524736466e-05,
+      "loss": 0.1885,
+      "step": 3478
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 1.4564684725642778e-05,
+      "loss": 0.176,
+      "step": 3479
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 1.4549031643694471e-05,
+      "loss": 0.3801,
+      "step": 3480
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 1.4533384284689346e-05,
+      "loss": 0.1702,
+      "step": 3481
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 1.4517742654423101e-05,
+      "loss": 0.149,
+      "step": 3482
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 1.450210675868929e-05,
+      "loss": 0.0651,
+      "step": 3483
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 1.4486476603279349e-05,
+      "loss": 0.2946,
+      "step": 3484
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 1.4470852193982603e-05,
+      "loss": 0.0805,
+      "step": 3485
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 1.4455233536586233e-05,
+      "loss": 0.0697,
+      "step": 3486
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 1.4439620636875279e-05,
+      "loss": 0.2757,
+      "step": 3487
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 1.4424013500632683e-05,
+      "loss": 0.1541,
+      "step": 3488
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 1.4408412133639221e-05,
+      "loss": 0.1439,
+      "step": 3489
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 1.4392816541673537e-05,
+      "loss": 0.0814,
+      "step": 3490
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 1.4377226730512157e-05,
+      "loss": 0.1788,
+      "step": 3491
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 1.4361642705929443e-05,
+      "loss": 0.1159,
+      "step": 3492
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 1.4346064473697614e-05,
+      "loss": 0.1463,
+      "step": 3493
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 1.4330492039586765e-05,
+      "loss": 0.181,
+      "step": 3494
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 1.4314925409364824e-05,
+      "loss": 0.3379,
+      "step": 3495
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 1.4299364588797563e-05,
+      "loss": 0.3075,
+      "step": 3496
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 1.4283809583648637e-05,
+      "loss": 0.1468,
+      "step": 3497
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 1.4268260399679497e-05,
+      "loss": 0.2002,
+      "step": 3498
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.425271704264949e-05,
+      "loss": 0.4158,
+      "step": 3499
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.4237179518315767e-05,
+      "loss": 0.0678,
+      "step": 3500
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.4221647832433324e-05,
+      "loss": 0.387,
+      "step": 3501
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.4206121990755006e-05,
+      "loss": 0.2781,
+      "step": 3502
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.4190601999031504e-05,
+      "loss": 0.1218,
+      "step": 3503
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.4175087863011318e-05,
+      "loss": 0.072,
+      "step": 3504
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.4159579588440771e-05,
+      "loss": 0.0647,
+      "step": 3505
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.4144077181064058e-05,
+      "loss": 0.153,
+      "step": 3506
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.4128580646623149e-05,
+      "loss": 0.1347,
+      "step": 3507
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.4113089990857894e-05,
+      "loss": 0.0602,
+      "step": 3508
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.409760521950592e-05,
+      "loss": 0.2157,
+      "step": 3509
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.408212633830268e-05,
+      "loss": 0.3833,
+      "step": 3510
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.406665335298148e-05,
+      "loss": 0.1348,
+      "step": 3511
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.4051186269273404e-05,
+      "loss": 0.0788,
+      "step": 3512
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.4035725092907353e-05,
+      "loss": 0.3406,
+      "step": 3513
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.4020269829610078e-05,
+      "loss": 0.1245,
+      "step": 3514
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.4004820485106097e-05,
+      "loss": 0.4501,
+      "step": 3515
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.3989377065117744e-05,
+      "loss": 0.3725,
+      "step": 3516
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.3973939575365188e-05,
+      "loss": 0.422,
+      "step": 3517
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.3958508021566366e-05,
+      "loss": 0.0964,
+      "step": 3518
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.3943082409437022e-05,
+      "loss": 0.0867,
+      "step": 3519
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.3927662744690725e-05,
+      "loss": 0.3798,
+      "step": 3520
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.3912249033038821e-05,
+      "loss": 0.1425,
+      "step": 3521
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.3896841280190433e-05,
+      "loss": 0.2776,
+      "step": 3522
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.3881439491852528e-05,
+      "loss": 0.1562,
+      "step": 3523
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.3866043673729815e-05,
+      "loss": 0.0695,
+      "step": 3524
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.3850653831524801e-05,
+      "loss": 0.1264,
+      "step": 3525
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.3835269970937814e-05,
+      "loss": 0.2196,
+      "step": 3526
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.3819892097666915e-05,
+      "loss": 0.0669,
+      "step": 3527
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.3804520217408e-05,
+      "loss": 0.0676,
+      "step": 3528
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.37891543358547e-05,
+      "loss": 0.0806,
+      "step": 3529
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.3773794458698436e-05,
+      "loss": 0.1069,
+      "step": 3530
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.3758440591628436e-05,
+      "loss": 0.1744,
+      "step": 3531
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.374309274033166e-05,
+      "loss": 0.2737,
+      "step": 3532
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.372775091049285e-05,
+      "loss": 0.126,
+      "step": 3533
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.3712415107794547e-05,
+      "loss": 0.0643,
+      "step": 3534
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.3697085337917017e-05,
+      "loss": 0.0932,
+      "step": 3535
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.3681761606538332e-05,
+      "loss": 0.0867,
+      "step": 3536
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.3666443919334283e-05,
+      "loss": 0.085,
+      "step": 3537
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.3651132281978474e-05,
+      "loss": 0.0594,
+      "step": 3538
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.3635826700142217e-05,
+      "loss": 0.0879,
+      "step": 3539
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.3620527179494625e-05,
+      "loss": 0.1425,
+      "step": 3540
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.3605233725702533e-05,
+      "loss": 0.0701,
+      "step": 3541
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.3589946344430538e-05,
+      "loss": 0.1619,
+      "step": 3542
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.3574665041341008e-05,
+      "loss": 0.166,
+      "step": 3543
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.3559389822094033e-05,
+      "loss": 0.1703,
+      "step": 3544
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.3544120692347451e-05,
+      "loss": 0.0895,
+      "step": 3545
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.352885765775687e-05,
+      "loss": 0.0471,
+      "step": 3546
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.3513600723975622e-05,
+      "loss": 0.3293,
+      "step": 3547
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.349834989665476e-05,
+      "loss": 0.1864,
+      "step": 3548
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.3483105181443123e-05,
+      "loss": 0.0786,
+      "step": 3549
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.3467866583987249e-05,
+      "loss": 0.1675,
+      "step": 3550
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.3452634109931407e-05,
+      "loss": 0.1028,
+      "step": 3551
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.3437407764917637e-05,
+      "loss": 0.209,
+      "step": 3552
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.3422187554585667e-05,
+      "loss": 0.0483,
+      "step": 3553
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.3406973484572962e-05,
+      "loss": 0.2117,
+      "step": 3554
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.339176556051474e-05,
+      "loss": 0.0563,
+      "step": 3555
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.3376563788043896e-05,
+      "loss": 0.0796,
+      "step": 3556
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.3361368172791102e-05,
+      "loss": 0.1,
+      "step": 3557
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.3346178720384704e-05,
+      "loss": 0.0834,
+      "step": 3558
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.3330995436450773e-05,
+      "loss": 0.1484,
+      "step": 3559
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.3315818326613124e-05,
+      "loss": 0.4027,
+      "step": 3560
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.3300647396493252e-05,
+      "loss": 0.0872,
+      "step": 3561
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.3285482651710371e-05,
+      "loss": 0.0694,
+      "step": 3562
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.3270324097881423e-05,
+      "loss": 0.2678,
+      "step": 3563
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.3255171740621037e-05,
+      "loss": 0.5103,
+      "step": 3564
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.3240025585541537e-05,
+      "loss": 0.2115,
+      "step": 3565
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.322488563825299e-05,
+      "loss": 0.063,
+      "step": 3566
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.320975190436313e-05,
+      "loss": 0.0763,
+      "step": 3567
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.3194624389477375e-05,
+      "loss": 0.1907,
+      "step": 3568
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.3179503099198905e-05,
+      "loss": 0.0747,
+      "step": 3569
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.3164388039128532e-05,
+      "loss": 0.1058,
+      "step": 3570
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.3149279214864766e-05,
+      "loss": 0.0768,
+      "step": 3571
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.3134176632003852e-05,
+      "loss": 0.2263,
+      "step": 3572
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.3119080296139679e-05,
+      "loss": 0.4745,
+      "step": 3573
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.3103990212863821e-05,
+      "loss": 0.1754,
+      "step": 3574
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.3088906387765577e-05,
+      "loss": 0.1578,
+      "step": 3575
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.3073828826431893e-05,
+      "loss": 0.1786,
+      "step": 3576
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.3058757534447395e-05,
+      "loss": 0.1002,
+      "step": 3577
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.3043692517394412e-05,
+      "loss": 0.111,
+      "step": 3578
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.3028633780852926e-05,
+      "loss": 0.5632,
+      "step": 3579
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.301358133040059e-05,
+      "loss": 0.1362,
+      "step": 3580
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.299853517161276e-05,
+      "loss": 0.1333,
+      "step": 3581
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.2983495310062428e-05,
+      "loss": 0.09,
+      "step": 3582
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.2968461751320257e-05,
+      "loss": 0.1713,
+      "step": 3583
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.2953434500954602e-05,
+      "loss": 0.0925,
+      "step": 3584
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.2938413564531446e-05,
+      "loss": 0.0756,
+      "step": 3585
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.292339894761447e-05,
+      "loss": 0.0526,
+      "step": 3586
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.2908390655764987e-05,
+      "loss": 0.1361,
+      "step": 3587
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.2893388694541965e-05,
+      "loss": 0.284,
+      "step": 3588
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.2878393069502055e-05,
+      "loss": 0.0988,
+      "step": 3589
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.2863403786199535e-05,
+      "loss": 0.2622,
+      "step": 3590
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.2848420850186338e-05,
+      "loss": 0.3211,
+      "step": 3591
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.2833444267012061e-05,
+      "loss": 0.2843,
+      "step": 3592
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.2818474042223938e-05,
+      "loss": 0.1818,
+      "step": 3593
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.280351018136683e-05,
+      "loss": 0.0919,
+      "step": 3594
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.2788552689983285e-05,
+      "loss": 0.104,
+      "step": 3595
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.277360157361345e-05,
+      "loss": 0.3268,
+      "step": 3596
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.2758656837795114e-05,
+      "loss": 0.0727,
+      "step": 3597
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.2743718488063747e-05,
+      "loss": 0.0772,
+      "step": 3598
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.2728786529952398e-05,
+      "loss": 0.2047,
+      "step": 3599
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.2713860968991769e-05,
+      "loss": 0.104,
+      "step": 3600
+    },
+    {
+      "epoch": 2.78,
+      "eval_loss": 1.3151249885559082,
+      "eval_runtime": 157.3472,
+      "eval_samples_per_second": 0.337,
+      "eval_steps_per_second": 0.172,
+      "step": 3600
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.2698941810710204e-05,
+      "loss": 0.2686,
+      "step": 3601
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.2684029060633678e-05,
+      "loss": 0.0669,
+      "step": 3602
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.2669122724285761e-05,
+      "loss": 0.1398,
+      "step": 3603
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.2654222807187686e-05,
+      "loss": 0.0518,
+      "step": 3604
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.263932931485828e-05,
+      "loss": 0.0656,
+      "step": 3605
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.2624442252813992e-05,
+      "loss": 0.0671,
+      "step": 3606
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.2609561626568921e-05,
+      "loss": 0.0995,
+      "step": 3607
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.2594687441634747e-05,
+      "loss": 0.158,
+      "step": 3608
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.2579819703520761e-05,
+      "loss": 0.1505,
+      "step": 3609
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.2564958417733904e-05,
+      "loss": 0.0917,
+      "step": 3610
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.2550103589778699e-05,
+      "loss": 0.3163,
+      "step": 3611
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.2535255225157267e-05,
+      "loss": 0.1443,
+      "step": 3612
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.2520413329369375e-05,
+      "loss": 0.3753,
+      "step": 3613
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.2505577907912347e-05,
+      "loss": 0.0549,
+      "step": 3614
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.2490748966281151e-05,
+      "loss": 0.1646,
+      "step": 3615
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.2475926509968325e-05,
+      "loss": 0.7631,
+      "step": 3616
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.2461110544464006e-05,
+      "loss": 0.0819,
+      "step": 3617
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.2446301075255958e-05,
+      "loss": 0.085,
+      "step": 3618
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.2431498107829504e-05,
+      "loss": 0.3685,
+      "step": 3619
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.2416701647667562e-05,
+      "loss": 0.1286,
+      "step": 3620
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.2401911700250671e-05,
+      "loss": 0.0723,
+      "step": 3621
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.2387128271056925e-05,
+      "loss": 0.0611,
+      "step": 3622
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.2372351365562003e-05,
+      "loss": 0.0968,
+      "step": 3623
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.2357580989239203e-05,
+      "loss": 0.1992,
+      "step": 3624
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.2342817147559372e-05,
+      "loss": 0.0703,
+      "step": 3625
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.2328059845990932e-05,
+      "loss": 0.0688,
+      "step": 3626
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.2313309089999915e-05,
+      "loss": 0.0703,
+      "step": 3627
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.2298564885049905e-05,
+      "loss": 0.2117,
+      "step": 3628
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.2283827236602048e-05,
+      "loss": 0.0957,
+      "step": 3629
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.2269096150115104e-05,
+      "loss": 0.2873,
+      "step": 3630
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.2254371631045356e-05,
+      "loss": 0.0747,
+      "step": 3631
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.2239653684846694e-05,
+      "loss": 0.2058,
+      "step": 3632
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.2224942316970539e-05,
+      "loss": 0.1774,
+      "step": 3633
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.2210237532865888e-05,
+      "loss": 0.2102,
+      "step": 3634
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.2195539337979308e-05,
+      "loss": 0.147,
+      "step": 3635
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.2180847737754928e-05,
+      "loss": 0.3738,
+      "step": 3636
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.2166162737634424e-05,
+      "loss": 0.3279,
+      "step": 3637
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.2151484343057009e-05,
+      "loss": 0.5393,
+      "step": 3638
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.21368125594595e-05,
+      "loss": 0.1296,
+      "step": 3639
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.2122147392276215e-05,
+      "loss": 0.0579,
+      "step": 3640
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.2107488846939038e-05,
+      "loss": 0.0939,
+      "step": 3641
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.2092836928877422e-05,
+      "loss": 0.1912,
+      "step": 3642
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.2078191643518324e-05,
+      "loss": 0.0746,
+      "step": 3643
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.206355299628629e-05,
+      "loss": 0.156,
+      "step": 3644
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.2048920992603374e-05,
+      "loss": 0.1814,
+      "step": 3645
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.203429563788917e-05,
+      "loss": 0.3025,
+      "step": 3646
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.2019676937560843e-05,
+      "loss": 0.33,
+      "step": 3647
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.2005064897033048e-05,
+      "loss": 0.0691,
+      "step": 3648
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.1990459521717995e-05,
+      "loss": 0.0804,
+      "step": 3649
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.1975860817025442e-05,
+      "loss": 0.0761,
+      "step": 3650
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.1961268788362645e-05,
+      "loss": 0.1135,
+      "step": 3651
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.1946683441134397e-05,
+      "loss": 0.1921,
+      "step": 3652
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.1932104780743037e-05,
+      "loss": 0.0682,
+      "step": 3653
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.1917532812588403e-05,
+      "loss": 0.2612,
+      "step": 3654
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.190296754206785e-05,
+      "loss": 0.085,
+      "step": 3655
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.1888408974576285e-05,
+      "loss": 0.087,
+      "step": 3656
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.1873857115506097e-05,
+      "loss": 0.347,
+      "step": 3657
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.1859311970247205e-05,
+      "loss": 0.1961,
+      "step": 3658
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.1844773544187052e-05,
+      "loss": 0.1655,
+      "step": 3659
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.1830241842710565e-05,
+      "loss": 0.0765,
+      "step": 3660
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.1815716871200213e-05,
+      "loss": 0.2064,
+      "step": 3661
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.180119863503595e-05,
+      "loss": 0.1809,
+      "step": 3662
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.1786687139595227e-05,
+      "loss": 0.7007,
+      "step": 3663
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.1772182390253037e-05,
+      "loss": 0.0846,
+      "step": 3664
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.1757684392381835e-05,
+      "loss": 0.1831,
+      "step": 3665
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.1743193151351583e-05,
+      "loss": 0.082,
+      "step": 3666
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.1728708672529754e-05,
+      "loss": 0.2317,
+      "step": 3667
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.171423096128132e-05,
+      "loss": 0.3106,
+      "step": 3668
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.1699760022968728e-05,
+      "loss": 0.1135,
+      "step": 3669
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.1685295862951913e-05,
+      "loss": 0.3601,
+      "step": 3670
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.1670838486588328e-05,
+      "loss": 0.0924,
+      "step": 3671
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.1656387899232883e-05,
+      "loss": 0.3221,
+      "step": 3672
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.1641944106237998e-05,
+      "loss": 0.0643,
+      "step": 3673
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.1627507112953562e-05,
+      "loss": 0.0504,
+      "step": 3674
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.1613076924726933e-05,
+      "loss": 0.1356,
+      "step": 3675
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.1598653546902981e-05,
+      "loss": 0.18,
+      "step": 3676
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.1584236984824035e-05,
+      "loss": 0.0798,
+      "step": 3677
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.1569827243829882e-05,
+      "loss": 0.1767,
+      "step": 3678
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.1555424329257824e-05,
+      "loss": 0.3907,
+      "step": 3679
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.1541028246442603e-05,
+      "loss": 0.2094,
+      "step": 3680
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.152663900071643e-05,
+      "loss": 0.065,
+      "step": 3681
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.1512256597409008e-05,
+      "loss": 0.1342,
+      "step": 3682
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.1497881041847485e-05,
+      "loss": 0.0693,
+      "step": 3683
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.1483512339356465e-05,
+      "loss": 0.0838,
+      "step": 3684
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.1469150495258045e-05,
+      "loss": 0.0743,
+      "step": 3685
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.1454795514871757e-05,
+      "loss": 0.3241,
+      "step": 3686
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.1440447403514588e-05,
+      "loss": 0.2549,
+      "step": 3687
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.1426106166501004e-05,
+      "loss": 0.3354,
+      "step": 3688
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.1411771809142899e-05,
+      "loss": 0.143,
+      "step": 3689
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.1397444336749641e-05,
+      "loss": 0.2411,
+      "step": 3690
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.1383123754628035e-05,
+      "loss": 0.211,
+      "step": 3691
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.136881006808232e-05,
+      "loss": 0.0825,
+      "step": 3692
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.1354503282414223e-05,
+      "loss": 0.0929,
+      "step": 3693
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.1340203402922875e-05,
+      "loss": 0.2203,
+      "step": 3694
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.1325910434904852e-05,
+      "loss": 0.2216,
+      "step": 3695
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.1311624383654203e-05,
+      "loss": 0.4275,
+      "step": 3696
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.1297345254462385e-05,
+      "loss": 0.295,
+      "step": 3697
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.1283073052618288e-05,
+      "loss": 0.0928,
+      "step": 3698
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.1268807783408264e-05,
+      "loss": 0.0629,
+      "step": 3699
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.1254549452116061e-05,
+      "loss": 0.2809,
+      "step": 3700
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.1240298064022891e-05,
+      "loss": 0.0848,
+      "step": 3701
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.122605362440739e-05,
+      "loss": 0.0659,
+      "step": 3702
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.1211816138545597e-05,
+      "loss": 0.1622,
+      "step": 3703
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.119758561171098e-05,
+      "loss": 0.2235,
+      "step": 3704
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.1183362049174459e-05,
+      "loss": 0.0601,
+      "step": 3705
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.1169145456204343e-05,
+      "loss": 0.3188,
+      "step": 3706
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.1154935838066358e-05,
+      "loss": 0.0904,
+      "step": 3707
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.1140733200023683e-05,
+      "loss": 0.1495,
+      "step": 3708
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.1126537547336874e-05,
+      "loss": 0.1358,
+      "step": 3709
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.1112348885263905e-05,
+      "loss": 0.1042,
+      "step": 3710
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.1098167219060189e-05,
+      "loss": 0.2413,
+      "step": 3711
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.1083992553978517e-05,
+      "loss": 0.2102,
+      "step": 3712
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.1069824895269083e-05,
+      "loss": 0.0658,
+      "step": 3713
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.1055664248179525e-05,
+      "loss": 0.0701,
+      "step": 3714
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.1041510617954852e-05,
+      "loss": 0.0867,
+      "step": 3715
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.1027364009837465e-05,
+      "loss": 0.2121,
+      "step": 3716
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.10132244290672e-05,
+      "loss": 0.0746,
+      "step": 3717
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.0999091880881258e-05,
+      "loss": 0.2236,
+      "step": 3718
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.098496637051426e-05,
+      "loss": 0.0483,
+      "step": 3719
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.0970847903198198e-05,
+      "loss": 0.1732,
+      "step": 3720
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.0956736484162462e-05,
+      "loss": 0.345,
+      "step": 3721
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.0942632118633844e-05,
+      "loss": 0.2127,
+      "step": 3722
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.0928534811836513e-05,
+      "loss": 0.1631,
+      "step": 3723
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.0914444568992011e-05,
+      "loss": 0.066,
+      "step": 3724
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.0900361395319296e-05,
+      "loss": 0.0887,
+      "step": 3725
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.088628529603468e-05,
+      "loss": 0.1001,
+      "step": 3726
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.0872216276351849e-05,
+      "loss": 0.1402,
+      "step": 3727
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.0858154341481908e-05,
+      "loss": 0.1576,
+      "step": 3728
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.0844099496633297e-05,
+      "loss": 0.0715,
+      "step": 3729
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.0830051747011832e-05,
+      "loss": 0.0635,
+      "step": 3730
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.0816011097820734e-05,
+      "loss": 0.0965,
+      "step": 3731
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.0801977554260562e-05,
+      "loss": 0.2713,
+      "step": 3732
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.0787951121529236e-05,
+      "loss": 0.0697,
+      "step": 3733
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.07739318048221e-05,
+      "loss": 0.0793,
+      "step": 3734
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.0759919609331803e-05,
+      "loss": 0.1079,
+      "step": 3735
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.074591454024836e-05,
+      "loss": 0.0689,
+      "step": 3736
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.0731916602759188e-05,
+      "loss": 0.1018,
+      "step": 3737
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.0717925802049023e-05,
+      "loss": 0.0812,
+      "step": 3738
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.0703942143299959e-05,
+      "loss": 0.1627,
+      "step": 3739
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.0689965631691487e-05,
+      "loss": 0.1246,
+      "step": 3740
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.0675996272400398e-05,
+      "loss": 0.5067,
+      "step": 3741
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.0662034070600855e-05,
+      "loss": 0.0828,
+      "step": 3742
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.0648079031464389e-05,
+      "loss": 0.0596,
+      "step": 3743
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.0634131160159848e-05,
+      "loss": 0.0768,
+      "step": 3744
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.0620190461853428e-05,
+      "loss": 0.1537,
+      "step": 3745
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.0606256941708701e-05,
+      "loss": 0.1627,
+      "step": 3746
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.059233060488653e-05,
+      "loss": 0.5241,
+      "step": 3747
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.0578411456545174e-05,
+      "loss": 0.3447,
+      "step": 3748
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.056449950184018e-05,
+      "loss": 0.204,
+      "step": 3749
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.0550594745924443e-05,
+      "loss": 0.2402,
+      "step": 3750
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.053669719394822e-05,
+      "loss": 0.0723,
+      "step": 3751
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.052280685105907e-05,
+      "loss": 0.2425,
+      "step": 3752
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.0508923722401875e-05,
+      "loss": 0.1764,
+      "step": 3753
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.0495047813118884e-05,
+      "loss": 0.0791,
+      "step": 3754
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.0481179128349635e-05,
+      "loss": 0.0984,
+      "step": 3755
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.0467317673230998e-05,
+      "loss": 0.0793,
+      "step": 3756
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.0453463452897187e-05,
+      "loss": 0.1092,
+      "step": 3757
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.043961647247971e-05,
+      "loss": 0.0807,
+      "step": 3758
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.0425776737107394e-05,
+      "loss": 0.1584,
+      "step": 3759
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.0411944251906407e-05,
+      "loss": 0.0492,
+      "step": 3760
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.0398119022000214e-05,
+      "loss": 0.072,
+      "step": 3761
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.0384301052509577e-05,
+      "loss": 0.0918,
+      "step": 3762
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.0370490348552612e-05,
+      "loss": 0.511,
+      "step": 3763
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.0356686915244704e-05,
+      "loss": 0.2408,
+      "step": 3764
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.0342890757698552e-05,
+      "loss": 0.0678,
+      "step": 3765
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.032910188102418e-05,
+      "loss": 0.2749,
+      "step": 3766
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.0315320290328905e-05,
+      "loss": 0.1175,
+      "step": 3767
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.0301545990717329e-05,
+      "loss": 0.0619,
+      "step": 3768
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.0287778987291388e-05,
+      "loss": 0.2852,
+      "step": 3769
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.0274019285150281e-05,
+      "loss": 0.0821,
+      "step": 3770
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.0260266889390511e-05,
+      "loss": 0.1308,
+      "step": 3771
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.0246521805105899e-05,
+      "loss": 0.0626,
+      "step": 3772
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.0232784037387525e-05,
+      "loss": 0.2195,
+      "step": 3773
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.021905359132377e-05,
+      "loss": 0.3468,
+      "step": 3774
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.020533047200032e-05,
+      "loss": 0.2775,
+      "step": 3775
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.0191614684500128e-05,
+      "loss": 0.2559,
+      "step": 3776
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.0177906233903425e-05,
+      "loss": 0.2722,
+      "step": 3777
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.0164205125287756e-05,
+      "loss": 0.3367,
+      "step": 3778
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.015051136372791e-05,
+      "loss": 0.1191,
+      "step": 3779
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.0136824954295989e-05,
+      "loss": 0.2287,
+      "step": 3780
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.0123145902061343e-05,
+      "loss": 0.4363,
+      "step": 3781
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.0109474212090605e-05,
+      "loss": 0.0689,
+      "step": 3782
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.0095809889447697e-05,
+      "loss": 0.1624,
+      "step": 3783
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.0082152939193791e-05,
+      "loss": 0.1508,
+      "step": 3784
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.0068503366387332e-05,
+      "loss": 0.1339,
+      "step": 3785
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.0054861176084053e-05,
+      "loss": 0.3231,
+      "step": 3786
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.0041226373336926e-05,
+      "loss": 0.0866,
+      "step": 3787
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.0027598963196189e-05,
+      "loss": 0.0841,
+      "step": 3788
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.0013978950709373e-05,
+      "loss": 0.0901,
+      "step": 3789
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.0000366340921236e-05,
+      "loss": 0.0967,
+      "step": 3790
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 9.986761138873792e-06,
+      "loss": 0.2936,
+      "step": 3791
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 9.973163349606342e-06,
+      "loss": 0.0791,
+      "step": 3792
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 9.959572978155418e-06,
+      "loss": 0.0887,
+      "step": 3793
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 9.945990029554798e-06,
+      "loss": 0.0685,
+      "step": 3794
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 9.932414508835541e-06,
+      "loss": 0.1636,
+      "step": 3795
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 9.918846421025919e-06,
+      "loss": 0.1121,
+      "step": 3796
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 9.905285771151484e-06,
+      "loss": 0.2241,
+      "step": 3797
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 9.891732564235007e-06,
+      "loss": 0.0828,
+      "step": 3798
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 9.878186805296506e-06,
+      "loss": 0.1156,
+      "step": 3799
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 9.864648499353251e-06,
+      "loss": 0.1057,
+      "step": 3800
+    },
+    {
+      "epoch": 2.94,
+      "eval_loss": 1.3326809406280518,
+      "eval_runtime": 157.2548,
+      "eval_samples_per_second": 0.337,
+      "eval_steps_per_second": 0.172,
+      "step": 3800
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 9.851117651419761e-06,
+      "loss": 0.1258,
+      "step": 3801
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 9.837594266507769e-06,
+      "loss": 0.1323,
+      "step": 3802
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 9.824078349626242e-06,
+      "loss": 0.1143,
+      "step": 3803
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 9.810569905781405e-06,
+      "loss": 0.0819,
+      "step": 3804
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 9.797068939976704e-06,
+      "loss": 0.0578,
+      "step": 3805
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 9.7835754572128e-06,
+      "loss": 0.1509,
+      "step": 3806
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 9.77008946248761e-06,
+      "loss": 0.2242,
+      "step": 3807
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 9.756610960796246e-06,
+      "loss": 0.215,
+      "step": 3808
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 9.743139957131084e-06,
+      "loss": 0.1712,
+      "step": 3809
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 9.72967645648169e-06,
+      "loss": 0.1476,
+      "step": 3810
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 9.71622046383485e-06,
+      "loss": 0.1083,
+      "step": 3811
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 9.7027719841746e-06,
+      "loss": 0.0454,
+      "step": 3812
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 9.689331022482164e-06,
+      "loss": 0.0848,
+      "step": 3813
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 9.675897583735981e-06,
+      "loss": 0.0643,
+      "step": 3814
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 9.662471672911737e-06,
+      "loss": 0.0647,
+      "step": 3815
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 9.64905329498229e-06,
+      "loss": 0.3458,
+      "step": 3816
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 9.635642454917724e-06,
+      "loss": 0.093,
+      "step": 3817
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 9.622239157685338e-06,
+      "loss": 0.0572,
+      "step": 3818
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 9.608843408249634e-06,
+      "loss": 0.0728,
+      "step": 3819
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 9.595455211572296e-06,
+      "loss": 0.3482,
+      "step": 3820
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 9.582074572612253e-06,
+      "loss": 0.1352,
+      "step": 3821
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 9.568701496325598e-06,
+      "loss": 0.2157,
+      "step": 3822
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 9.55533598766563e-06,
+      "loss": 0.1091,
+      "step": 3823
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 9.541978051582864e-06,
+      "loss": 0.12,
+      "step": 3824
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 9.528627693024988e-06,
+      "loss": 0.136,
+      "step": 3825
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 9.515284916936902e-06,
+      "loss": 0.3178,
+      "step": 3826
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 9.501949728260682e-06,
+      "loss": 0.0683,
+      "step": 3827
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 9.48862213193559e-06,
+      "loss": 0.0919,
+      "step": 3828
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 9.475302132898103e-06,
+      "loss": 0.2656,
+      "step": 3829
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 9.461989736081858e-06,
+      "loss": 0.0741,
+      "step": 3830
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 9.448684946417672e-06,
+      "loss": 0.0817,
+      "step": 3831
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 9.435387768833574e-06,
+      "loss": 0.4048,
+      "step": 3832
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 9.422098208254756e-06,
+      "loss": 0.2043,
+      "step": 3833
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 9.40881626960359e-06,
+      "loss": 0.1802,
+      "step": 3834
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 9.395541957799606e-06,
+      "loss": 0.1183,
+      "step": 3835
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 9.38227527775955e-06,
+      "loss": 0.0681,
+      "step": 3836
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 9.369016234397298e-06,
+      "loss": 0.1706,
+      "step": 3837
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 9.355764832623937e-06,
+      "loss": 0.1848,
+      "step": 3838
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 9.342521077347694e-06,
+      "loss": 0.0996,
+      "step": 3839
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 9.329284973473963e-06,
+      "loss": 0.0744,
+      "step": 3840
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 9.316056525905344e-06,
+      "loss": 0.1789,
+      "step": 3841
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 9.30283573954155e-06,
+      "loss": 0.2029,
+      "step": 3842
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 9.289622619279473e-06,
+      "loss": 0.3038,
+      "step": 3843
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 9.276417170013195e-06,
+      "loss": 0.0925,
+      "step": 3844
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 9.26321939663392e-06,
+      "loss": 0.1215,
+      "step": 3845
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 9.25002930403001e-06,
+      "loss": 0.0553,
+      "step": 3846
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 9.236846897087014e-06,
+      "loss": 0.2317,
+      "step": 3847
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 9.223672180687608e-06,
+      "loss": 0.0977,
+      "step": 3848
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 9.210505159711613e-06,
+      "loss": 0.0724,
+      "step": 3849
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 9.197345839036029e-06,
+      "loss": 0.0498,
+      "step": 3850
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 9.184194223534983e-06,
+      "loss": 0.1865,
+      "step": 3851
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 9.17105031807974e-06,
+      "loss": 0.0796,
+      "step": 3852
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 9.157914127538736e-06,
+      "loss": 0.0798,
+      "step": 3853
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 9.144785656777521e-06,
+      "loss": 0.0934,
+      "step": 3854
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 9.131664910658819e-06,
+      "loss": 0.186,
+      "step": 3855
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 9.118551894042461e-06,
+      "loss": 0.1257,
+      "step": 3856
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 9.105446611785421e-06,
+      "loss": 0.184,
+      "step": 3857
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 9.09234906874183e-06,
+      "loss": 0.2357,
+      "step": 3858
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 9.079259269762928e-06,
+      "loss": 0.3227,
+      "step": 3859
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 9.066177219697089e-06,
+      "loss": 0.0962,
+      "step": 3860
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 9.053102923389837e-06,
+      "loss": 0.2694,
+      "step": 3861
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 9.040036385683806e-06,
+      "loss": 0.1014,
+      "step": 3862
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 9.026977611418753e-06,
+      "loss": 0.0947,
+      "step": 3863
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 9.013926605431581e-06,
+      "loss": 0.0941,
+      "step": 3864
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 9.000883372556282e-06,
+      "loss": 0.1839,
+      "step": 3865
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 8.987847917624006e-06,
+      "loss": 0.0812,
+      "step": 3866
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 8.97482024546301e-06,
+      "loss": 0.2203,
+      "step": 3867
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 8.96180036089865e-06,
+      "loss": 0.1364,
+      "step": 3868
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 8.948788268753405e-06,
+      "loss": 0.2639,
+      "step": 3869
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 8.935783973846897e-06,
+      "loss": 0.0566,
+      "step": 3870
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 8.92278748099582e-06,
+      "loss": 0.3168,
+      "step": 3871
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 8.909798795013986e-06,
+      "loss": 0.1379,
+      "step": 3872
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 8.896817920712346e-06,
+      "loss": 0.0752,
+      "step": 3873
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 8.883844862898923e-06,
+      "loss": 0.1626,
+      "step": 3874
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 8.870879626378851e-06,
+      "loss": 0.1928,
+      "step": 3875
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 8.857922215954393e-06,
+      "loss": 0.0889,
+      "step": 3876
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 8.844972636424879e-06,
+      "loss": 0.1981,
+      "step": 3877
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 8.832030892586748e-06,
+      "loss": 0.1856,
+      "step": 3878
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 8.819096989233563e-06,
+      "loss": 0.058,
+      "step": 3879
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 8.806170931155947e-06,
+      "loss": 0.0921,
+      "step": 3880
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 5172,
+  "num_train_epochs": 4,
+  "save_steps": 500,
+  "total_flos": 7.600436406376464e+18,
+  "trial_name": null,
+  "trial_params": null
+}