diff --git "a/checkpoint-1293/trainer_state.json" "b/checkpoint-1293/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1293/trainer_state.json"
@@ -0,0 +1,7825 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9996134518747584,
+  "eval_steps": 200,
+  "global_step": 1293,
+  "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
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 5172,
+  "num_train_epochs": 4,
+  "save_steps": 500,
+  "total_flos": 2.5243420840805007e+18,
+  "trial_name": null,
+  "trial_params": null
+}