diff --git "a/checkpoint-1002/trainer_state.json" "b/checkpoint-1002/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1002/trainer_state.json"
@@ -0,0 +1,6431 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9932785660941,
+  "eval_steps": 20,
+  "global_step": 1002,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 2e-05,
+      "loss": 1.5197,
+      "step": 1
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4e-05,
+      "loss": 1.5435,
+      "step": 2
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6e-05,
+      "loss": 1.3967,
+      "step": 3
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8e-05,
+      "loss": 1.368,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0001,
+      "loss": 1.2766,
+      "step": 5
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00012,
+      "loss": 1.1676,
+      "step": 6
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00014,
+      "loss": 1.0507,
+      "step": 7
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00016,
+      "loss": 1.0016,
+      "step": 8
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00018,
+      "loss": 0.9427,
+      "step": 9
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0002,
+      "loss": 0.9294,
+      "step": 10
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00019999949852874652,
+      "loss": 0.9023,
+      "step": 11
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019999799412001546,
+      "loss": 0.8257,
+      "step": 12
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001999954867888952,
+      "loss": 0.9064,
+      "step": 13
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019999197656053288,
+      "loss": 0.8408,
+      "step": 14
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00019998746347013398,
+      "loss": 0.9058,
+      "step": 15
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001999819475629623,
+      "loss": 0.8419,
+      "step": 16
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019997542889433915,
+      "loss": 0.8535,
+      "step": 17
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019996790752964305,
+      "loss": 0.8121,
+      "step": 18
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019995938354430898,
+      "loss": 0.7725,
+      "step": 19
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019994985702382758,
+      "loss": 0.7661,
+      "step": 20
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 1.517298698425293,
+      "eval_runtime": 15.7157,
+      "eval_samples_per_second": 37.097,
+      "eval_steps_per_second": 18.58,
+      "step": 20
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0001999393280637444,
+      "loss": 0.7878,
+      "step": 21
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019992779676965885,
+      "loss": 0.7435,
+      "step": 22
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019991526325722315,
+      "loss": 0.8219,
+      "step": 23
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019990172765214128,
+      "loss": 0.7624,
+      "step": 24
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019988719009016753,
+      "loss": 0.7527,
+      "step": 25
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00019987165071710527,
+      "loss": 0.7699,
+      "step": 26
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00019985510968880557,
+      "loss": 0.7536,
+      "step": 27
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00019983756717116536,
+      "loss": 0.7863,
+      "step": 28
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.000199819023340126,
+      "loss": 0.7166,
+      "step": 29
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001997994783816715,
+      "loss": 0.7459,
+      "step": 30
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00019977893249182652,
+      "loss": 0.7376,
+      "step": 31
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00019975738587665456,
+      "loss": 0.7624,
+      "step": 32
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00019973483875225574,
+      "loss": 0.7329,
+      "step": 33
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00019971129134476473,
+      "loss": 0.7512,
+      "step": 34
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0001996867438903486,
+      "loss": 0.7211,
+      "step": 35
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00019966119663520412,
+      "loss": 0.7467,
+      "step": 36
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00019963464983555556,
+      "loss": 0.7186,
+      "step": 37
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001996071037576521,
+      "loss": 0.7244,
+      "step": 38
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00019957855867776508,
+      "loss": 0.7071,
+      "step": 39
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00019954901488218515,
+      "loss": 0.7681,
+      "step": 40
+    },
+    {
+      "epoch": 0.12,
+      "eval_loss": 1.2323185205459595,
+      "eval_runtime": 15.7174,
+      "eval_samples_per_second": 37.093,
+      "eval_steps_per_second": 18.578,
+      "step": 40
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0001995184726672197,
+      "loss": 0.708,
+      "step": 41
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00019948693233918952,
+      "loss": 0.7443,
+      "step": 42
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.000199454394214426,
+      "loss": 0.7079,
+      "step": 43
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001994208586192678,
+      "loss": 0.6812,
+      "step": 44
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001993863258900577,
+      "loss": 0.6792,
+      "step": 45
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00019935079637313906,
+      "loss": 0.6553,
+      "step": 46
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001993142704248525,
+      "loss": 0.6936,
+      "step": 47
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00019927674841153237,
+      "loss": 0.74,
+      "step": 48
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00019923823070950282,
+      "loss": 0.7154,
+      "step": 49
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001991987177050743,
+      "loss": 0.6902,
+      "step": 50
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0001991582097945395,
+      "loss": 0.7009,
+      "step": 51
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00019911670738416947,
+      "loss": 0.699,
+      "step": 52
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00019907421089020955,
+      "loss": 0.681,
+      "step": 53
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00019903072073887507,
+      "loss": 0.7097,
+      "step": 54
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00019898623736634737,
+      "loss": 0.6861,
+      "step": 55
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.000198940761218769,
+      "loss": 0.7279,
+      "step": 56
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00019889429275223957,
+      "loss": 0.6784,
+      "step": 57
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00019884683243281116,
+      "loss": 0.7,
+      "step": 58
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00019879838073648341,
+      "loss": 0.6634,
+      "step": 59
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00019874893814919906,
+      "loss": 0.6647,
+      "step": 60
+    },
+    {
+      "epoch": 0.18,
+      "eval_loss": 1.1306297779083252,
+      "eval_runtime": 15.73,
+      "eval_samples_per_second": 37.063,
+      "eval_steps_per_second": 18.563,
+      "step": 60
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00019869850516683875,
+      "loss": 0.7377,
+      "step": 61
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00019864708229521636,
+      "loss": 0.6794,
+      "step": 62
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00019859467005007368,
+      "loss": 0.6744,
+      "step": 63
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001985412689570754,
+      "loss": 0.6921,
+      "step": 64
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00019848687955180386,
+      "loss": 0.7389,
+      "step": 65
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00019843150237975344,
+      "loss": 0.6795,
+      "step": 66
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00019837513799632534,
+      "loss": 0.6917,
+      "step": 67
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00019831778696682194,
+      "loss": 0.66,
+      "step": 68
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00019825944986644114,
+      "loss": 0.6926,
+      "step": 69
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00019820012728027044,
+      "loss": 0.7265,
+      "step": 70
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00019813981980328129,
+      "loss": 0.733,
+      "step": 71
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00019807852804032305,
+      "loss": 0.6543,
+      "step": 72
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00019801625260611685,
+      "loss": 0.673,
+      "step": 73
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00019795299412524945,
+      "loss": 0.6773,
+      "step": 74
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0001978887532321671,
+      "loss": 0.6469,
+      "step": 75
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.000197823530571169,
+      "loss": 0.7272,
+      "step": 76
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00019775732679640093,
+      "loss": 0.7015,
+      "step": 77
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001976901425718487,
+      "loss": 0.6973,
+      "step": 78
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00019762197857133148,
+      "loss": 0.6831,
+      "step": 79
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00019755283547849494,
+      "loss": 0.6742,
+      "step": 80
+    },
+    {
+      "epoch": 0.24,
+      "eval_loss": 1.0846511125564575,
+      "eval_runtime": 15.6885,
+      "eval_samples_per_second": 37.161,
+      "eval_steps_per_second": 18.612,
+      "step": 80
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001974827139868046,
+      "loss": 0.683,
+      "step": 81
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001974116147995387,
+      "loss": 0.7063,
+      "step": 82
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00019733953862978123,
+      "loss": 0.6963,
+      "step": 83
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00019726648620041468,
+      "loss": 0.7179,
+      "step": 84
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00019719245824411296,
+      "loss": 0.7315,
+      "step": 85
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001971174555033339,
+      "loss": 0.6839,
+      "step": 86
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00019704147873031186,
+      "loss": 0.6917,
+      "step": 87
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00019696452868705024,
+      "loss": 0.6828,
+      "step": 88
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00019688660614531367,
+      "loss": 0.6965,
+      "step": 89
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00019680771188662044,
+      "loss": 0.6522,
+      "step": 90
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00019672784670223464,
+      "loss": 0.6939,
+      "step": 91
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0001966470113931582,
+      "loss": 0.7037,
+      "step": 92
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00019656520677012264,
+      "loss": 0.6903,
+      "step": 93
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00019648243365358146,
+      "loss": 0.706,
+      "step": 94
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0001963986928737014,
+      "loss": 0.6253,
+      "step": 95
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00019631398527035422,
+      "loss": 0.732,
+      "step": 96
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0001962283116931086,
+      "loss": 0.6935,
+      "step": 97
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00019614167300122126,
+      "loss": 0.731,
+      "step": 98
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001960540700636284,
+      "loss": 0.6916,
+      "step": 99
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001959655037589372,
+      "loss": 0.6995,
+      "step": 100
+    },
+    {
+      "epoch": 0.3,
+      "eval_loss": 1.0572510957717896,
+      "eval_runtime": 15.7086,
+      "eval_samples_per_second": 37.113,
+      "eval_steps_per_second": 18.588,
+      "step": 100
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001958759749754167,
+      "loss": 0.7313,
+      "step": 101
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00019578548461098914,
+      "loss": 0.7337,
+      "step": 102
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001956940335732209,
+      "loss": 0.6721,
+      "step": 103
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00019560162277931325,
+      "loss": 0.6907,
+      "step": 104
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0001955082531560934,
+      "loss": 0.6599,
+      "step": 105
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00019541392564000488,
+      "loss": 0.6941,
+      "step": 106
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00019531864117709855,
+      "loss": 0.7037,
+      "step": 107
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00019522240072302274,
+      "loss": 0.6819,
+      "step": 108
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00019512520524301387,
+      "loss": 0.6729,
+      "step": 109
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00019502705571188672,
+      "loss": 0.713,
+      "step": 110
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00019492795311402473,
+      "loss": 0.6638,
+      "step": 111
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001948278984433699,
+      "loss": 0.647,
+      "step": 112
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00019472689270341303,
+      "loss": 0.614,
+      "step": 113
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001946249369071837,
+      "loss": 0.6783,
+      "step": 114
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0001945220320772399,
+      "loss": 0.7093,
+      "step": 115
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00019441817924565786,
+      "loss": 0.7179,
+      "step": 116
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00019431337945402185,
+      "loss": 0.7241,
+      "step": 117
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0001942076337534135,
+      "loss": 0.7168,
+      "step": 118
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00019410094320440138,
+      "loss": 0.7127,
+      "step": 119
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00019399330887703037,
+      "loss": 0.6883,
+      "step": 120
+    },
+    {
+      "epoch": 0.36,
+      "eval_loss": 1.0412259101867676,
+      "eval_runtime": 15.6448,
+      "eval_samples_per_second": 37.265,
+      "eval_steps_per_second": 18.664,
+      "step": 120
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00019388473185081087,
+      "loss": 0.6663,
+      "step": 121
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00019377521321470805,
+      "loss": 0.7319,
+      "step": 122
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00019366475406713086,
+      "loss": 0.6942,
+      "step": 123
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00019355335551592105,
+      "loss": 0.6847,
+      "step": 124
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.000193441018678342,
+      "loss": 0.6732,
+      "step": 125
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00019332774468106768,
+      "loss": 0.6661,
+      "step": 126
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00019321353466017108,
+      "loss": 0.7044,
+      "step": 127
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00019309838976111311,
+      "loss": 0.6794,
+      "step": 128
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00019298231113873087,
+      "loss": 0.7275,
+      "step": 129
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00019286529995722623,
+      "loss": 0.685,
+      "step": 130
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00019274735739015404,
+      "loss": 0.6825,
+      "step": 131
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00019262848462041045,
+      "loss": 0.6857,
+      "step": 132
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.000192508682840221,
+      "loss": 0.7139,
+      "step": 133
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001923879532511287,
+      "loss": 0.6413,
+      "step": 134
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00019226629706398186,
+      "loss": 0.7257,
+      "step": 135
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001921437154989221,
+      "loss": 0.6365,
+      "step": 136
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00019202020978537207,
+      "loss": 0.66,
+      "step": 137
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00019189578116202307,
+      "loss": 0.6525,
+      "step": 138
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00019177043087682265,
+      "loss": 0.6438,
+      "step": 139
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00019164416018696207,
+      "loss": 0.6437,
+      "step": 140
+    },
+    {
+      "epoch": 0.42,
+      "eval_loss": 1.0374538898468018,
+      "eval_runtime": 15.6903,
+      "eval_samples_per_second": 37.157,
+      "eval_steps_per_second": 18.61,
+      "step": 140
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00019151697035886382,
+      "loss": 0.6878,
+      "step": 141
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00019138886266816866,
+      "loss": 0.7072,
+      "step": 142
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00019125983839972317,
+      "loss": 0.69,
+      "step": 143
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00019112989884756653,
+      "loss": 0.7666,
+      "step": 144
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00019099904531491774,
+      "loss": 0.689,
+      "step": 145
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001908672791141625,
+      "loss": 0.7394,
+      "step": 146
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001907346015668401,
+      "loss": 0.6882,
+      "step": 147
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00019060101400362998,
+      "loss": 0.6394,
+      "step": 148
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00019046651776433862,
+      "loss": 0.6932,
+      "step": 149
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00019033111419788597,
+      "loss": 0.6676,
+      "step": 150
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00019019480466229201,
+      "loss": 0.6818,
+      "step": 151
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.000190057590524663,
+      "loss": 0.6695,
+      "step": 152
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001899194731611778,
+      "loss": 0.7082,
+      "step": 153
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00018978045395707418,
+      "loss": 0.6626,
+      "step": 154
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00018964053430663483,
+      "loss": 0.6708,
+      "step": 155
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001894997156131734,
+      "loss": 0.7864,
+      "step": 156
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00018935799928902045,
+      "loss": 0.6899,
+      "step": 157
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001892153867555092,
+      "loss": 0.6574,
+      "step": 158
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00018907187944296137,
+      "loss": 0.7157,
+      "step": 159
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00018892747879067286,
+      "loss": 0.6331,
+      "step": 160
+    },
+    {
+      "epoch": 0.48,
+      "eval_loss": 1.0186299085617065,
+      "eval_runtime": 15.6809,
+      "eval_samples_per_second": 37.179,
+      "eval_steps_per_second": 18.621,
+      "step": 160
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001887821862468991,
+      "loss": 0.7219,
+      "step": 161
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00018863600326884082,
+      "loss": 0.7203,
+      "step": 162
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00018848893132262926,
+      "loss": 0.6722,
+      "step": 163
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00018834097188331143,
+      "loss": 0.7223,
+      "step": 164
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001881921264348355,
+      "loss": 0.659,
+      "step": 165
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00018804239647003573,
+      "loss": 0.6763,
+      "step": 166
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00018789178349061753,
+      "loss": 0.649,
+      "step": 167
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00018774028900714256,
+      "loss": 0.6762,
+      "step": 168
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00018758791453901336,
+      "loss": 0.6547,
+      "step": 169
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00018743466161445823,
+      "loss": 0.6516,
+      "step": 170
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0001872805317705159,
+      "loss": 0.6479,
+      "step": 171
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0001871255265530201,
+      "loss": 0.6986,
+      "step": 172
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00018696964751658404,
+      "loss": 0.7402,
+      "step": 173
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00018681289622458485,
+      "loss": 0.6629,
+      "step": 174
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00018665527424914787,
+      "loss": 0.6761,
+      "step": 175
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00018649678317113084,
+      "loss": 0.6746,
+      "step": 176
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001863374245801082,
+      "loss": 0.67,
+      "step": 177
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00018617720007435497,
+      "loss": 0.7406,
+      "step": 178
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00018601611126083087,
+      "loss": 0.6735,
+      "step": 179
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.000185854159755164,
+      "loss": 0.6686,
+      "step": 180
+    },
+    {
+      "epoch": 0.54,
+      "eval_loss": 1.0153461694717407,
+      "eval_runtime": 15.6769,
+      "eval_samples_per_second": 37.189,
+      "eval_steps_per_second": 18.626,
+      "step": 180
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00018569134718163496,
+      "loss": 0.6416,
+      "step": 181
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00018552767517316022,
+      "loss": 0.6456,
+      "step": 182
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00018536314537127586,
+      "loss": 0.6794,
+      "step": 183
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00018519775942612128,
+      "loss": 0.6754,
+      "step": 184
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00018503151899642236,
+      "loss": 0.6251,
+      "step": 185
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00018486442574947511,
+      "loss": 0.7354,
+      "step": 186
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00018469648136112867,
+      "loss": 0.6457,
+      "step": 187
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001845276875157687,
+      "loss": 0.6823,
+      "step": 188
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00018435804590630043,
+      "loss": 0.6891,
+      "step": 189
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001841875582341317,
+      "loss": 0.6731,
+      "step": 190
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00018401622620915583,
+      "loss": 0.6857,
+      "step": 191
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001838440515497345,
+      "loss": 0.6807,
+      "step": 192
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001836710359826806,
+      "loss": 0.6622,
+      "step": 193
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00018349718124324076,
+      "loss": 0.6628,
+      "step": 194
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00018332248907507807,
+      "loss": 0.724,
+      "step": 195
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00018314696123025454,
+      "loss": 0.675,
+      "step": 196
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00018297059946921355,
+      "loss": 0.677,
+      "step": 197
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00018279340556076216,
+      "loss": 0.6555,
+      "step": 198
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001826153812820534,
+      "loss": 0.6273,
+      "step": 199
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001824365284185684,
+      "loss": 0.6767,
+      "step": 200
+    },
+    {
+      "epoch": 0.6,
+      "eval_loss": 1.004173994064331,
+      "eval_runtime": 15.726,
+      "eval_samples_per_second": 37.072,
+      "eval_steps_per_second": 18.568,
+      "step": 200
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00018225684876409863,
+      "loss": 0.6719,
+      "step": 201
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00018207634412072764,
+      "loss": 0.6522,
+      "step": 202
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001818950162988133,
+      "loss": 0.6957,
+      "step": 203
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00018171286711696934,
+      "loss": 0.7013,
+      "step": 204
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00018152989840204735,
+      "loss": 0.6603,
+      "step": 205
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001813461119891184,
+      "loss": 0.6788,
+      "step": 206
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001811615097214545,
+      "loss": 0.6795,
+      "step": 207
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00018097609345051025,
+      "loss": 0.6458,
+      "step": 208
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00018078986503590428,
+      "loss": 0.6741,
+      "step": 209
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00018060282634540053,
+      "loss": 0.6537,
+      "step": 210
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001804149792548895,
+      "loss": 0.6501,
+      "step": 211
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00018022632564836948,
+      "loss": 0.673,
+      "step": 212
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00018003686741792772,
+      "loss": 0.6746,
+      "step": 213
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0001798466064637214,
+      "loss": 0.6329,
+      "step": 214
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00017965554469395837,
+      "loss": 0.69,
+      "step": 215
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00017946368402487845,
+      "loss": 0.6513,
+      "step": 216
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00017927102638073383,
+      "loss": 0.6758,
+      "step": 217
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00017907757369376985,
+      "loss": 0.6776,
+      "step": 218
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00017888332790420584,
+      "loss": 0.6487,
+      "step": 219
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00017868829096021527,
+      "loss": 0.7037,
+      "step": 220
+    },
+    {
+      "epoch": 0.66,
+      "eval_loss": 1.002336025238037,
+      "eval_runtime": 17.4124,
+      "eval_samples_per_second": 33.482,
+      "eval_steps_per_second": 16.77,
+      "step": 220
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00017849246481790664,
+      "loss": 0.71,
+      "step": 221
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00017829585144130356,
+      "loss": 0.6695,
+      "step": 222
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00017809845280232514,
+      "loss": 0.6645,
+      "step": 223
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0001779002708807662,
+      "loss": 0.6983,
+      "step": 224
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0001777013076642776,
+      "loss": 0.6785,
+      "step": 225
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0001775015651483459,
+      "loss": 0.6791,
+      "step": 226
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0001773010453362737,
+      "loss": 0.695,
+      "step": 227
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00017709975023915949,
+      "loss": 0.6352,
+      "step": 228
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00017689768187587732,
+      "loss": 0.6911,
+      "step": 229
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0001766948422730567,
+      "loss": 0.7189,
+      "step": 230
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00017649123346506223,
+      "loss": 0.6838,
+      "step": 231
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0001762868574939732,
+      "loss": 0.7209,
+      "step": 232
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00017608171640956313,
+      "loss": 0.6719,
+      "step": 233
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001758758122692791,
+      "loss": 0.6624,
+      "step": 234
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0001756691471382213,
+      "loss": 0.637,
+      "step": 235
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00017546172308912213,
+      "loss": 0.653,
+      "step": 236
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0001752535422023256,
+      "loss": 0.6368,
+      "step": 237
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00017504460656576627,
+      "loss": 0.6713,
+      "step": 238
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00017483491827494847,
+      "loss": 0.6614,
+      "step": 239
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0001746244794329252,
+      "loss": 0.6994,
+      "step": 240
+    },
+    {
+      "epoch": 0.72,
+      "eval_loss": 1.0014034509658813,
+      "eval_runtime": 15.6575,
+      "eval_samples_per_second": 37.235,
+      "eval_steps_per_second": 18.649,
+      "step": 240
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00017441329215027706,
+      "loss": 0.6644,
+      "step": 241
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0001742013585450911,
+      "loss": 0.6374,
+      "step": 242
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00017398868074293948,
+      "loss": 0.6627,
+      "step": 243
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00017377526087685832,
+      "loss": 0.6936,
+      "step": 244
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00017356110108732617,
+      "loss": 0.6419,
+      "step": 245
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0001733462035222426,
+      "loss": 0.6402,
+      "step": 246
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001731305703369066,
+      "loss": 0.7028,
+      "step": 247
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001729142036939951,
+      "loss": 0.6146,
+      "step": 248
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0001726971057635411,
+      "loss": 0.6702,
+      "step": 249
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.000172479278722912,
+      "loss": 0.6971,
+      "step": 250
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00017226072475678782,
+      "loss": 0.6289,
+      "step": 251
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0001720414460571392,
+      "loss": 0.6786,
+      "step": 252
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0001718214448232054,
+      "loss": 0.6599,
+      "step": 253
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0001716007232614723,
+      "loss": 0.656,
+      "step": 254
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00017137928358565028,
+      "loss": 0.7092,
+      "step": 255
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.000171157128016652,
+      "loss": 0.6596,
+      "step": 256
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00017093425878257005,
+      "loss": 0.6937,
+      "step": 257
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00017071067811865476,
+      "loss": 0.7027,
+      "step": 258
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00017048638826729163,
+      "loss": 0.6776,
+      "step": 259
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0001702613914779789,
+      "loss": 0.7012,
+      "step": 260
+    },
+    {
+      "epoch": 0.78,
+      "eval_loss": 0.9995876550674438,
+      "eval_runtime": 15.664,
+      "eval_samples_per_second": 37.219,
+      "eval_steps_per_second": 18.641,
+      "step": 260
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00017003569000730498,
+      "loss": 0.6786,
+      "step": 261
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0001698092861189259,
+      "loss": 0.7071,
+      "step": 262
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00016958218208354252,
+      "loss": 0.6556,
+      "step": 263
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00016935438017887772,
+      "loss": 0.6418,
+      "step": 264
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0001691258826896536,
+      "loss": 0.6726,
+      "step": 265
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00016889669190756868,
+      "loss": 0.6514,
+      "step": 266
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00016866681013127466,
+      "loss": 0.6914,
+      "step": 267
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00016843623966635366,
+      "loss": 0.6884,
+      "step": 268
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00016820498282529479,
+      "loss": 0.6624,
+      "step": 269
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0001679730419274713,
+      "loss": 0.7241,
+      "step": 270
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00016774041929911698,
+      "loss": 0.6923,
+      "step": 271
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0001675071172733031,
+      "loss": 0.6817,
+      "step": 272
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00016727313818991487,
+      "loss": 0.7042,
+      "step": 273
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00016703848439562785,
+      "loss": 0.66,
+      "step": 274
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00016680315824388483,
+      "loss": 0.662,
+      "step": 275
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00016656716209487174,
+      "loss": 0.6174,
+      "step": 276
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00016633049831549424,
+      "loss": 0.6684,
+      "step": 277
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0001660931692793541,
+      "loss": 0.6581,
+      "step": 278
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00016585517736672498,
+      "loss": 0.6835,
+      "step": 279
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.000165616524964529,
+      "loss": 0.6599,
+      "step": 280
+    },
+    {
+      "epoch": 0.84,
+      "eval_loss": 0.9926242828369141,
+      "eval_runtime": 15.669,
+      "eval_samples_per_second": 37.207,
+      "eval_steps_per_second": 18.636,
+      "step": 280
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00016537721446631252,
+      "loss": 0.6619,
+      "step": 281
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00016513724827222227,
+      "loss": 0.6557,
+      "step": 282
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00016489662878898113,
+      "loss": 0.6577,
+      "step": 283
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00016465535842986434,
+      "loss": 0.6423,
+      "step": 284
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00016441343961467471,
+      "loss": 0.6472,
+      "step": 285
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.000164170874769719,
+      "loss": 0.6434,
+      "step": 286
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00016392766632778312,
+      "loss": 0.6461,
+      "step": 287
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00016368381672810786,
+      "loss": 0.6857,
+      "step": 288
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00016343932841636456,
+      "loss": 0.6927,
+      "step": 289
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0001631942038446304,
+      "loss": 0.6263,
+      "step": 290
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00016294844547136395,
+      "loss": 0.6857,
+      "step": 291
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00016270205576138032,
+      "loss": 0.682,
+      "step": 292
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0001624550371858267,
+      "loss": 0.6964,
+      "step": 293
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00016220739222215738,
+      "loss": 0.6261,
+      "step": 294
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00016195912335410895,
+      "loss": 0.6848,
+      "step": 295
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00016171023307167545,
+      "loss": 0.635,
+      "step": 296
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0001614607238710833,
+      "loss": 0.6941,
+      "step": 297
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0001612105982547663,
+      "loss": 0.6339,
+      "step": 298
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0001609598587313406,
+      "loss": 0.6298,
+      "step": 299
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00016070850781557948,
+      "loss": 0.6401,
+      "step": 300
+    },
+    {
+      "epoch": 0.9,
+      "eval_loss": 0.9913042187690735,
+      "eval_runtime": 15.6715,
+      "eval_samples_per_second": 37.201,
+      "eval_steps_per_second": 18.633,
+      "step": 300
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00016045654802838812,
+      "loss": 0.6483,
+      "step": 301
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0001602039818967783,
+      "loss": 0.6352,
+      "step": 302
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00015995081195384315,
+      "loss": 0.6344,
+      "step": 303
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00015969704073873157,
+      "loss": 0.6317,
+      "step": 304
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.000159442670796623,
+      "loss": 0.6979,
+      "step": 305
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0001591877046787017,
+      "loss": 0.6429,
+      "step": 306
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00015893214494213132,
+      "loss": 0.6493,
+      "step": 307
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00015867599415002895,
+      "loss": 0.6765,
+      "step": 308
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00015841925487143987,
+      "loss": 0.6871,
+      "step": 309
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00015816192968131138,
+      "loss": 0.6389,
+      "step": 310
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00015790402116046726,
+      "loss": 0.6002,
+      "step": 311
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0001576455318955816,
+      "loss": 0.6025,
+      "step": 312
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001573864644791532,
+      "loss": 0.6692,
+      "step": 313
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00015712682150947923,
+      "loss": 0.7079,
+      "step": 314
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00015686660559062945,
+      "loss": 0.6622,
+      "step": 315
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00015660581933241993,
+      "loss": 0.6402,
+      "step": 316
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00015634446535038688,
+      "loss": 0.7251,
+      "step": 317
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00015608254626576048,
+      "loss": 0.6806,
+      "step": 318
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00015582006470543855,
+      "loss": 0.6475,
+      "step": 319
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00015555702330196023,
+      "loss": 0.6665,
+      "step": 320
+    },
+    {
+      "epoch": 0.96,
+      "eval_loss": 0.9910463094711304,
+      "eval_runtime": 15.7019,
+      "eval_samples_per_second": 37.129,
+      "eval_steps_per_second": 18.596,
+      "step": 320
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00015529342469347958,
+      "loss": 0.6674,
+      "step": 321
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00015502927152373914,
+      "loss": 0.6873,
+      "step": 322
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00015476456644204328,
+      "loss": 0.6425,
+      "step": 323
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0001544993121032318,
+      "loss": 0.6565,
+      "step": 324
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00015423351116765317,
+      "loss": 0.6886,
+      "step": 325
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.000153967166301138,
+      "loss": 0.6904,
+      "step": 326
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00015370028017497217,
+      "loss": 0.6281,
+      "step": 327
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00015343285546587013,
+      "loss": 0.6836,
+      "step": 328
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00015316489485594782,
+      "loss": 0.697,
+      "step": 329
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00015289640103269625,
+      "loss": 0.656,
+      "step": 330
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00015262737668895405,
+      "loss": 0.6649,
+      "step": 331
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00015235782452288068,
+      "loss": 0.6775,
+      "step": 332
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00015208774723792943,
+      "loss": 0.6637,
+      "step": 333
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.0001518171475428202,
+      "loss": 0.7175,
+      "step": 334
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00015154602815151233,
+      "loss": 0.6345,
+      "step": 335
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00015127439178317745,
+      "loss": 0.6214,
+      "step": 336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00015100224116217217,
+      "loss": 0.6026,
+      "step": 337
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00015072957901801076,
+      "loss": 0.6383,
+      "step": 338
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00015045640808533772,
+      "loss": 0.6026,
+      "step": 339
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.0001501827311039005,
+      "loss": 0.5771,
+      "step": 340
+    },
+    {
+      "epoch": 1.02,
+      "eval_loss": 0.9907026886940002,
+      "eval_runtime": 15.6789,
+      "eval_samples_per_second": 37.184,
+      "eval_steps_per_second": 18.624,
+      "step": 340
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00014990855081852185,
+      "loss": 0.6087,
+      "step": 341
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.0001496338699790724,
+      "loss": 0.6411,
+      "step": 342
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00014935869134044308,
+      "loss": 0.6074,
+      "step": 343
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00014908301766251739,
+      "loss": 0.5877,
+      "step": 344
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00014880685171014382,
+      "loss": 0.5731,
+      "step": 345
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00014853019625310813,
+      "loss": 0.6113,
+      "step": 346
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00014825305406610545,
+      "loss": 0.5965,
+      "step": 347
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00014797542792871265,
+      "loss": 0.6175,
+      "step": 348
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.0001476973206253602,
+      "loss": 0.6018,
+      "step": 349
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.0001474187349453045,
+      "loss": 0.6039,
+      "step": 350
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.0001471396736825998,
+      "loss": 0.5692,
+      "step": 351
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00014686013963607,
+      "loss": 0.6234,
+      "step": 352
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00014658013560928096,
+      "loss": 0.5961,
+      "step": 353
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00014629966441051208,
+      "loss": 0.6315,
+      "step": 354
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00014601872885272824,
+      "loss": 0.6223,
+      "step": 355
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.0001457373317535515,
+      "loss": 0.6557,
+      "step": 356
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.00014545547593523307,
+      "loss": 0.6012,
+      "step": 357
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.0001451731642246247,
+      "loss": 0.6405,
+      "step": 358
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.0001448903994531506,
+      "loss": 0.6106,
+      "step": 359
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00014460718445677876,
+      "loss": 0.6286,
+      "step": 360
+    },
+    {
+      "epoch": 1.08,
+      "eval_loss": 0.9830257892608643,
+      "eval_runtime": 15.7168,
+      "eval_samples_per_second": 37.094,
+      "eval_steps_per_second": 18.579,
+      "step": 360
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00014432352207599286,
+      "loss": 0.626,
+      "step": 361
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00014403941515576344,
+      "loss": 0.6169,
+      "step": 362
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.0001437548665455196,
+      "loss": 0.572,
+      "step": 363
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00014346987909912023,
+      "loss": 0.6255,
+      "step": 364
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.0001431844556748257,
+      "loss": 0.6036,
+      "step": 365
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00014289859913526874,
+      "loss": 0.5804,
+      "step": 366
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.00014261231234742619,
+      "loss": 0.5898,
+      "step": 367
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.00014232559818258984,
+      "loss": 0.6358,
+      "step": 368
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.000142038459516338,
+      "loss": 0.5652,
+      "step": 369
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00014175089922850633,
+      "loss": 0.6259,
+      "step": 370
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00014146292020315924,
+      "loss": 0.6309,
+      "step": 371
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00014117452532856083,
+      "loss": 0.6325,
+      "step": 372
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.0001408857174971458,
+      "loss": 0.5984,
+      "step": 373
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.0001405964996054907,
+      "loss": 0.5981,
+      "step": 374
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.00014030687455428474,
+      "loss": 0.6407,
+      "step": 375
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.00014001684524830057,
+      "loss": 0.5686,
+      "step": 376
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.0001397264145963655,
+      "loss": 0.6151,
+      "step": 377
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.00013943558551133186,
+      "loss": 0.5925,
+      "step": 378
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.00013914436091004827,
+      "loss": 0.5872,
+      "step": 379
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00013885274371333,
+      "loss": 0.6064,
+      "step": 380
+    },
+    {
+      "epoch": 1.14,
+      "eval_loss": 0.9865307807922363,
+      "eval_runtime": 15.7209,
+      "eval_samples_per_second": 37.084,
+      "eval_steps_per_second": 18.574,
+      "step": 380
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00013856073684592995,
+      "loss": 0.6151,
+      "step": 381
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.000138268343236509,
+      "loss": 0.6262,
+      "step": 382
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.000137975565817607,
+      "loss": 0.5727,
+      "step": 383
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.00013768240752561314,
+      "loss": 0.6486,
+      "step": 384
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.00013738887130073654,
+      "loss": 0.6102,
+      "step": 385
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.0001370949600869768,
+      "loss": 0.591,
+      "step": 386
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00013680067683209438,
+      "loss": 0.6336,
+      "step": 387
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00013650602448758112,
+      "loss": 0.6699,
+      "step": 388
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00013621100600863065,
+      "loss": 0.6317,
+      "step": 389
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.0001359156243541087,
+      "loss": 0.6129,
+      "step": 390
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.00013561988248652346,
+      "loss": 0.6984,
+      "step": 391
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.00013532378337199582,
+      "loss": 0.5925,
+      "step": 392
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.00013502732998022966,
+      "loss": 0.5835,
+      "step": 393
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00013473052528448201,
+      "loss": 0.5919,
+      "step": 394
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00013443337226153344,
+      "loss": 0.621,
+      "step": 395
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00013413587389165784,
+      "loss": 0.6363,
+      "step": 396
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.0001338380331585928,
+      "loss": 0.6383,
+      "step": 397
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00013353985304950973,
+      "loss": 0.5803,
+      "step": 398
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.0001332413365549836,
+      "loss": 0.6378,
+      "step": 399
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00013294248666896328,
+      "loss": 0.5976,
+      "step": 400
+    },
+    {
+      "epoch": 1.19,
+      "eval_loss": 0.980188250541687,
+      "eval_runtime": 15.7209,
+      "eval_samples_per_second": 37.084,
+      "eval_steps_per_second": 18.574,
+      "step": 400
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.00013264330638874128,
+      "loss": 0.6282,
+      "step": 401
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.0001323437987149238,
+      "loss": 0.6146,
+      "step": 402
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.00013204396665140067,
+      "loss": 0.6089,
+      "step": 403
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.00013174381320531505,
+      "loss": 0.5657,
+      "step": 404
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.0001314433413870334,
+      "loss": 0.6638,
+      "step": 405
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.0001311425542101154,
+      "loss": 0.6129,
+      "step": 406
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.0001308414546912834,
+      "loss": 0.5862,
+      "step": 407
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.00013054004585039258,
+      "loss": 0.5759,
+      "step": 408
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.00013023833071040025,
+      "loss": 0.6275,
+      "step": 409
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.00012993631229733582,
+      "loss": 0.6095,
+      "step": 410
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.0001296339936402703,
+      "loss": 0.6187,
+      "step": 411
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00012933137777128607,
+      "loss": 0.603,
+      "step": 412
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00012902846772544624,
+      "loss": 0.5911,
+      "step": 413
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.0001287252665407645,
+      "loss": 0.5979,
+      "step": 414
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.00012842177725817434,
+      "loss": 0.6157,
+      "step": 415
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.0001281180029214988,
+      "loss": 0.5796,
+      "step": 416
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00012781394657741989,
+      "loss": 0.6215,
+      "step": 417
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.0001275096112754478,
+      "loss": 0.5927,
+      "step": 418
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.0001272050000678908,
+      "loss": 0.628,
+      "step": 419
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.000126900116009824,
+      "loss": 0.5512,
+      "step": 420
+    },
+    {
+      "epoch": 1.25,
+      "eval_loss": 0.9817403554916382,
+      "eval_runtime": 15.7224,
+      "eval_samples_per_second": 37.081,
+      "eval_steps_per_second": 18.572,
+      "step": 420
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.00012659496215905938,
+      "loss": 0.6477,
+      "step": 421
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.0001262895415761145,
+      "loss": 0.6304,
+      "step": 422
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.00012598385732418224,
+      "loss": 0.6261,
+      "step": 423
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.00012567791246909994,
+      "loss": 0.6472,
+      "step": 424
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.00012537171007931858,
+      "loss": 0.6299,
+      "step": 425
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.00012506525322587207,
+      "loss": 0.605,
+      "step": 426
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.00012475854498234648,
+      "loss": 0.5852,
+      "step": 427
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.0001244515884248491,
+      "loss": 0.5736,
+      "step": 428
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.0001241443866319778,
+      "loss": 0.5795,
+      "step": 429
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.00012383694268478993,
+      "loss": 0.6088,
+      "step": 430
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00012352925966677147,
+      "loss": 0.6547,
+      "step": 431
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.0001232213406638062,
+      "loss": 0.5549,
+      "step": 432
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00012291318876414477,
+      "loss": 0.6398,
+      "step": 433
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.0001226048070583735,
+      "loss": 0.6377,
+      "step": 434
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.0001222961986393836,
+      "loss": 0.6207,
+      "step": 435
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.00012198736660234009,
+      "loss": 0.5999,
+      "step": 436
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00012167831404465078,
+      "loss": 0.6088,
+      "step": 437
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00012136904406593507,
+      "loss": 0.6525,
+      "step": 438
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00012105955976799314,
+      "loss": 0.5816,
+      "step": 439
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00012074986425477445,
+      "loss": 0.6333,
+      "step": 440
+    },
+    {
+      "epoch": 1.31,
+      "eval_loss": 0.9809613227844238,
+      "eval_runtime": 15.7148,
+      "eval_samples_per_second": 37.099,
+      "eval_steps_per_second": 18.581,
+      "step": 440
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00012043996063234706,
+      "loss": 0.5735,
+      "step": 441
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00012012985200886602,
+      "loss": 0.5884,
+      "step": 442
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00011981954149454262,
+      "loss": 0.5983,
+      "step": 443
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.00011950903220161285,
+      "loss": 0.6161,
+      "step": 444
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.00011919832724430642,
+      "loss": 0.5827,
+      "step": 445
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.00011888742973881543,
+      "loss": 0.5955,
+      "step": 446
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00011857634280326311,
+      "loss": 0.602,
+      "step": 447
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00011826506955767258,
+      "loss": 0.5978,
+      "step": 448
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00011795361312393549,
+      "loss": 0.6562,
+      "step": 449
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00011764197662578086,
+      "loss": 0.6053,
+      "step": 450
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00011733016318874357,
+      "loss": 0.6176,
+      "step": 451
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00011701817594013312,
+      "loss": 0.6388,
+      "step": 452
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00011670601800900225,
+      "loss": 0.599,
+      "step": 453
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00011639369252611552,
+      "loss": 0.6294,
+      "step": 454
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00011608120262391801,
+      "loss": 0.5844,
+      "step": 455
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00011576855143650371,
+      "loss": 0.6244,
+      "step": 456
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00011545574209958432,
+      "loss": 0.6068,
+      "step": 457
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00011514277775045768,
+      "loss": 0.5585,
+      "step": 458
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.0001148296615279762,
+      "loss": 0.5998,
+      "step": 459
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00011451639657251563,
+      "loss": 0.5883,
+      "step": 460
+    },
+    {
+      "epoch": 1.37,
+      "eval_loss": 0.9817206859588623,
+      "eval_runtime": 15.8012,
+      "eval_samples_per_second": 36.896,
+      "eval_steps_per_second": 18.48,
+      "step": 460
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00011420298602594332,
+      "loss": 0.6029,
+      "step": 461
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00011388943303158693,
+      "loss": 0.6066,
+      "step": 462
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00011357574073420265,
+      "loss": 0.6229,
+      "step": 463
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00011326191227994391,
+      "loss": 0.62,
+      "step": 464
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00011294795081632967,
+      "loss": 0.6268,
+      "step": 465
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00011263385949221295,
+      "loss": 0.6215,
+      "step": 466
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00011231964145774905,
+      "loss": 0.6612,
+      "step": 467
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.0001120052998643643,
+      "loss": 0.6015,
+      "step": 468
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00011169083786472408,
+      "loss": 0.6247,
+      "step": 469
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00011137625861270151,
+      "loss": 0.6472,
+      "step": 470
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.00011106156526334558,
+      "loss": 0.6283,
+      "step": 471
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.00011074676097284973,
+      "loss": 0.6026,
+      "step": 472
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.00011043184889851992,
+      "loss": 0.669,
+      "step": 473
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00011011683219874323,
+      "loss": 0.6302,
+      "step": 474
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.0001098017140329561,
+      "loss": 0.6662,
+      "step": 475
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00010948649756161246,
+      "loss": 0.6014,
+      "step": 476
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00010917118594615237,
+      "loss": 0.5959,
+      "step": 477
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00010885578234897003,
+      "loss": 0.5983,
+      "step": 478
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00010854028993338221,
+      "loss": 0.5428,
+      "step": 479
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00010822471186359639,
+      "loss": 0.5822,
+      "step": 480
+    },
+    {
+      "epoch": 1.43,
+      "eval_loss": 0.978273868560791,
+      "eval_runtime": 15.7427,
+      "eval_samples_per_second": 37.033,
+      "eval_steps_per_second": 18.548,
+      "step": 480
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00010790905130467923,
+      "loss": 0.635,
+      "step": 481
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00010759331142252462,
+      "loss": 0.5787,
+      "step": 482
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00010727749538382211,
+      "loss": 0.617,
+      "step": 483
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00010696160635602487,
+      "loss": 0.5646,
+      "step": 484
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00010664564750731839,
+      "loss": 0.6083,
+      "step": 485
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00010632962200658815,
+      "loss": 0.6098,
+      "step": 486
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.0001060135330233883,
+      "loss": 0.5669,
+      "step": 487
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00010569738372790956,
+      "loss": 0.6169,
+      "step": 488
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00010538117729094766,
+      "loss": 0.595,
+      "step": 489
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00010506491688387127,
+      "loss": 0.6104,
+      "step": 490
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.00010474860567859057,
+      "loss": 0.5904,
+      "step": 491
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.000104432246847525,
+      "loss": 0.5679,
+      "step": 492
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.00010411584356357178,
+      "loss": 0.5984,
+      "step": 493
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00010379939900007393,
+      "loss": 0.6033,
+      "step": 494
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.0001034829163307885,
+      "loss": 0.6357,
+      "step": 495
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00010316639872985472,
+      "loss": 0.6547,
+      "step": 496
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00010284984937176212,
+      "loss": 0.6082,
+      "step": 497
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00010253327143131879,
+      "loss": 0.6176,
+      "step": 498
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00010221666808361947,
+      "loss": 0.5942,
+      "step": 499
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00010190004250401368,
+      "loss": 0.5878,
+      "step": 500
+    },
+    {
+      "epoch": 1.49,
+      "eval_loss": 0.9756883978843689,
+      "eval_runtime": 15.7172,
+      "eval_samples_per_second": 37.093,
+      "eval_steps_per_second": 18.578,
+      "step": 500
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.00010158339786807401,
+      "loss": 0.6009,
+      "step": 501
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.00010126673735156402,
+      "loss": 0.6042,
+      "step": 502
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.0001009500641304067,
+      "loss": 0.6045,
+      "step": 503
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.00010063338138065234,
+      "loss": 0.5939,
+      "step": 504
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.00010031669227844694,
+      "loss": 0.6093,
+      "step": 505
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.0001,
+      "loss": 0.6265,
+      "step": 506
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 9.968330772155311e-05,
+      "loss": 0.5842,
+      "step": 507
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 9.936661861934765e-05,
+      "loss": 0.6032,
+      "step": 508
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 9.904993586959332e-05,
+      "loss": 0.5963,
+      "step": 509
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 9.8733262648436e-05,
+      "loss": 0.5725,
+      "step": 510
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 9.841660213192605e-05,
+      "loss": 0.6323,
+      "step": 511
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 9.809995749598632e-05,
+      "loss": 0.6186,
+      "step": 512
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 9.778333191638055e-05,
+      "loss": 0.6437,
+      "step": 513
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 9.746672856868123e-05,
+      "loss": 0.608,
+      "step": 514
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 9.715015062823789e-05,
+      "loss": 0.6013,
+      "step": 515
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 9.683360127014529e-05,
+      "loss": 0.6431,
+      "step": 516
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 9.651708366921152e-05,
+      "loss": 0.5867,
+      "step": 517
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.620060099992609e-05,
+      "loss": 0.574,
+      "step": 518
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.588415643642826e-05,
+      "loss": 0.5995,
+      "step": 519
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.556775315247501e-05,
+      "loss": 0.5951,
+      "step": 520
+    },
+    {
+      "epoch": 1.55,
+      "eval_loss": 0.9753234386444092,
+      "eval_runtime": 15.77,
+      "eval_samples_per_second": 36.969,
+      "eval_steps_per_second": 18.516,
+      "step": 520
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.525139432140945e-05,
+      "loss": 0.5802,
+      "step": 521
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.493508311612874e-05,
+      "loss": 0.6022,
+      "step": 522
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.461882270905241e-05,
+      "loss": 0.5621,
+      "step": 523
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.430261627209044e-05,
+      "loss": 0.5843,
+      "step": 524
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.398646697661172e-05,
+      "loss": 0.647,
+      "step": 525
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.367037799341187e-05,
+      "loss": 0.614,
+      "step": 526
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.335435249268165e-05,
+      "loss": 0.6347,
+      "step": 527
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.303839364397511e-05,
+      "loss": 0.6657,
+      "step": 528
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.272250461617793e-05,
+      "loss": 0.6394,
+      "step": 529
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.24066885774754e-05,
+      "loss": 0.5967,
+      "step": 530
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.209094869532081e-05,
+      "loss": 0.6504,
+      "step": 531
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.177528813640362e-05,
+      "loss": 0.5948,
+      "step": 532
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.145971006661782e-05,
+      "loss": 0.6202,
+      "step": 533
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.114421765102999e-05,
+      "loss": 0.6011,
+      "step": 534
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.082881405384765e-05,
+      "loss": 0.6236,
+      "step": 535
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.051350243838756e-05,
+      "loss": 0.6033,
+      "step": 536
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.019828596704394e-05,
+      "loss": 0.587,
+      "step": 537
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 8.98831678012568e-05,
+      "loss": 0.5835,
+      "step": 538
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 8.956815110148012e-05,
+      "loss": 0.5779,
+      "step": 539
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 8.925323902715031e-05,
+      "loss": 0.6466,
+      "step": 540
+    },
+    {
+      "epoch": 1.61,
+      "eval_loss": 0.9719182252883911,
+      "eval_runtime": 15.7304,
+      "eval_samples_per_second": 37.062,
+      "eval_steps_per_second": 18.563,
+      "step": 540
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 8.893843473665443e-05,
+      "loss": 0.6011,
+      "step": 541
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 8.862374138729853e-05,
+      "loss": 0.6446,
+      "step": 542
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 8.830916213527592e-05,
+      "loss": 0.6285,
+      "step": 543
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 8.799470013563573e-05,
+      "loss": 0.6259,
+      "step": 544
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 8.768035854225097e-05,
+      "loss": 0.6143,
+      "step": 545
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 8.73661405077871e-05,
+      "loss": 0.6058,
+      "step": 546
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 8.705204918367031e-05,
+      "loss": 0.5803,
+      "step": 547
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 8.67380877200561e-05,
+      "loss": 0.6259,
+      "step": 548
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 8.642425926579739e-05,
+      "loss": 0.5894,
+      "step": 549
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 8.611056696841312e-05,
+      "loss": 0.6112,
+      "step": 550
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.579701397405667e-05,
+      "loss": 0.6618,
+      "step": 551
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.54836034274844e-05,
+      "loss": 0.6115,
+      "step": 552
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.517033847202382e-05,
+      "loss": 0.6285,
+      "step": 553
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.485722224954237e-05,
+      "loss": 0.6507,
+      "step": 554
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.454425790041567e-05,
+      "loss": 0.6125,
+      "step": 555
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.423144856349631e-05,
+      "loss": 0.6536,
+      "step": 556
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.391879737608201e-05,
+      "loss": 0.6476,
+      "step": 557
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.36063074738845e-05,
+      "loss": 0.6355,
+      "step": 558
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.329398199099778e-05,
+      "loss": 0.6266,
+      "step": 559
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.298182405986689e-05,
+      "loss": 0.6246,
+      "step": 560
+    },
+    {
+      "epoch": 1.67,
+      "eval_loss": 0.9681297540664673,
+      "eval_runtime": 15.7562,
+      "eval_samples_per_second": 37.001,
+      "eval_steps_per_second": 18.532,
+      "step": 560
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.266983681125647e-05,
+      "loss": 0.6251,
+      "step": 561
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.235802337421919e-05,
+      "loss": 0.6299,
+      "step": 562
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.20463868760645e-05,
+      "loss": 0.6117,
+      "step": 563
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.173493044232745e-05,
+      "loss": 0.6027,
+      "step": 564
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.142365719673693e-05,
+      "loss": 0.6211,
+      "step": 565
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.11125702611846e-05,
+      "loss": 0.6076,
+      "step": 566
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.080167275569359e-05,
+      "loss": 0.635,
+      "step": 567
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.049096779838719e-05,
+      "loss": 0.603,
+      "step": 568
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.018045850545743e-05,
+      "loss": 0.6406,
+      "step": 569
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 7.987014799113397e-05,
+      "loss": 0.5941,
+      "step": 570
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 7.956003936765297e-05,
+      "loss": 0.5542,
+      "step": 571
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 7.925013574522557e-05,
+      "loss": 0.5938,
+      "step": 572
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 7.894044023200691e-05,
+      "loss": 0.5701,
+      "step": 573
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 7.863095593406491e-05,
+      "loss": 0.6276,
+      "step": 574
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 7.832168595534924e-05,
+      "loss": 0.6352,
+      "step": 575
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 7.801263339765994e-05,
+      "loss": 0.5628,
+      "step": 576
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 7.770380136061643e-05,
+      "loss": 0.6285,
+      "step": 577
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 7.739519294162652e-05,
+      "loss": 0.6409,
+      "step": 578
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 7.708681123585525e-05,
+      "loss": 0.6944,
+      "step": 579
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 7.677865933619379e-05,
+      "loss": 0.627,
+      "step": 580
+    },
+    {
+      "epoch": 1.73,
+      "eval_loss": 0.9704866409301758,
+      "eval_runtime": 15.7519,
+      "eval_samples_per_second": 37.011,
+      "eval_steps_per_second": 18.537,
+      "step": 580
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 7.647074033322856e-05,
+      "loss": 0.656,
+      "step": 581
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 7.616305731521008e-05,
+      "loss": 0.6058,
+      "step": 582
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 7.585561336802221e-05,
+      "loss": 0.5871,
+      "step": 583
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 7.554841157515092e-05,
+      "loss": 0.6347,
+      "step": 584
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.524145501765357e-05,
+      "loss": 0.5847,
+      "step": 585
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.493474677412794e-05,
+      "loss": 0.5988,
+      "step": 586
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.462828992068143e-05,
+      "loss": 0.5942,
+      "step": 587
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.432208753090009e-05,
+      "loss": 0.6388,
+      "step": 588
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.401614267581779e-05,
+      "loss": 0.6184,
+      "step": 589
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.371045842388552e-05,
+      "loss": 0.5469,
+      "step": 590
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.340503784094066e-05,
+      "loss": 0.591,
+      "step": 591
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.309988399017602e-05,
+      "loss": 0.64,
+      "step": 592
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.279499993210928e-05,
+      "loss": 0.6149,
+      "step": 593
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.24903887245522e-05,
+      "loss": 0.7015,
+      "step": 594
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.218605342258015e-05,
+      "loss": 0.6292,
+      "step": 595
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.188199707850122e-05,
+      "loss": 0.6177,
+      "step": 596
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.15782227418257e-05,
+      "loss": 0.5794,
+      "step": 597
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.127473345923554e-05,
+      "loss": 0.6237,
+      "step": 598
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.097153227455379e-05,
+      "loss": 0.6117,
+      "step": 599
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.066862222871397e-05,
+      "loss": 0.6214,
+      "step": 600
+    },
+    {
+      "epoch": 1.79,
+      "eval_loss": 0.9691401720046997,
+      "eval_runtime": 15.7684,
+      "eval_samples_per_second": 36.973,
+      "eval_steps_per_second": 18.518,
+      "step": 600
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.03660063597297e-05,
+      "loss": 0.5839,
+      "step": 601
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.006368770266421e-05,
+      "loss": 0.5862,
+      "step": 602
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 6.976166928959977e-05,
+      "loss": 0.5901,
+      "step": 603
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 6.945995414960744e-05,
+      "loss": 0.6055,
+      "step": 604
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 6.915854530871658e-05,
+      "loss": 0.5828,
+      "step": 605
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 6.885744578988463e-05,
+      "loss": 0.5737,
+      "step": 606
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 6.855665861296661e-05,
+      "loss": 0.5896,
+      "step": 607
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 6.825618679468502e-05,
+      "loss": 0.5914,
+      "step": 608
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 6.795603334859935e-05,
+      "loss": 0.6071,
+      "step": 609
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 6.765620128507619e-05,
+      "loss": 0.6257,
+      "step": 610
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 6.735669361125873e-05,
+      "loss": 0.6182,
+      "step": 611
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 6.705751333103675e-05,
+      "loss": 0.5525,
+      "step": 612
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 6.67586634450164e-05,
+      "loss": 0.5787,
+      "step": 613
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 6.64601469504903e-05,
+      "loss": 0.5781,
+      "step": 614
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.61619668414072e-05,
+      "loss": 0.5784,
+      "step": 615
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.586412610834221e-05,
+      "loss": 0.6093,
+      "step": 616
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.556662773846658e-05,
+      "loss": 0.5732,
+      "step": 617
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.526947471551798e-05,
+      "loss": 0.6163,
+      "step": 618
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.497267001977037e-05,
+      "loss": 0.5984,
+      "step": 619
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.46762166280042e-05,
+      "loss": 0.6558,
+      "step": 620
+    },
+    {
+      "epoch": 1.85,
+      "eval_loss": 0.9709066152572632,
+      "eval_runtime": 15.7265,
+      "eval_samples_per_second": 37.071,
+      "eval_steps_per_second": 18.567,
+      "step": 620
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.438011751347655e-05,
+      "loss": 0.596,
+      "step": 621
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.40843756458913e-05,
+      "loss": 0.5747,
+      "step": 622
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.378899399136938e-05,
+      "loss": 0.6224,
+      "step": 623
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.349397551241894e-05,
+      "loss": 0.6119,
+      "step": 624
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.319932316790565e-05,
+      "loss": 0.632,
+      "step": 625
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.290503991302324e-05,
+      "loss": 0.6042,
+      "step": 626
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.261112869926347e-05,
+      "loss": 0.5992,
+      "step": 627
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.231759247438689e-05,
+      "loss": 0.6207,
+      "step": 628
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.202443418239302e-05,
+      "loss": 0.6841,
+      "step": 629
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.173165676349103e-05,
+      "loss": 0.616,
+      "step": 630
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.14392631540701e-05,
+      "loss": 0.61,
+      "step": 631
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.114725628666998e-05,
+      "loss": 0.587,
+      "step": 632
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.0855639089951735e-05,
+      "loss": 0.6161,
+      "step": 633
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.0564414488668165e-05,
+      "loss": 0.6047,
+      "step": 634
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.027358540363457e-05,
+      "loss": 0.6326,
+      "step": 635
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5.998315475169942e-05,
+      "loss": 0.6319,
+      "step": 636
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5.969312544571529e-05,
+      "loss": 0.5422,
+      "step": 637
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5.94035003945093e-05,
+      "loss": 0.5769,
+      "step": 638
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5.911428250285421e-05,
+      "loss": 0.6539,
+      "step": 639
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5.88254746714392e-05,
+      "loss": 0.5736,
+      "step": 640
+    },
+    {
+      "epoch": 1.91,
+      "eval_loss": 0.9673944711685181,
+      "eval_runtime": 22.4205,
+      "eval_samples_per_second": 26.003,
+      "eval_steps_per_second": 13.024,
+      "step": 640
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5.8537079796840765e-05,
+      "loss": 0.6077,
+      "step": 641
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5.824910077149371e-05,
+      "loss": 0.6218,
+      "step": 642
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5.796154048366208e-05,
+      "loss": 0.6631,
+      "step": 643
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5.767440181741019e-05,
+      "loss": 0.596,
+      "step": 644
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.7387687652573865e-05,
+      "loss": 0.6009,
+      "step": 645
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.710140086473129e-05,
+      "loss": 0.6687,
+      "step": 646
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.6815544325174355e-05,
+      "loss": 0.6074,
+      "step": 647
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.653012090087977e-05,
+      "loss": 0.6322,
+      "step": 648
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.6245133454480446e-05,
+      "loss": 0.6062,
+      "step": 649
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.596058484423656e-05,
+      "loss": 0.6011,
+      "step": 650
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.5676477924007166e-05,
+      "loss": 0.6329,
+      "step": 651
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.5392815543221254e-05,
+      "loss": 0.6324,
+      "step": 652
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.5109600546849416e-05,
+      "loss": 0.638,
+      "step": 653
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.4826835775375285e-05,
+      "loss": 0.6016,
+      "step": 654
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.454452406476694e-05,
+      "loss": 0.6221,
+      "step": 655
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.4262668246448475e-05,
+      "loss": 0.6496,
+      "step": 656
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.398127114727178e-05,
+      "loss": 0.6118,
+      "step": 657
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.3700335589487925e-05,
+      "loss": 0.6216,
+      "step": 658
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.341986439071907e-05,
+      "loss": 0.6238,
+      "step": 659
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.3139860363929996e-05,
+      "loss": 0.6188,
+      "step": 660
+    },
+    {
+      "epoch": 1.97,
+      "eval_loss": 0.9674209356307983,
+      "eval_runtime": 15.74,
+      "eval_samples_per_second": 37.039,
+      "eval_steps_per_second": 18.552,
+      "step": 660
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.286032631740023e-05,
+      "loss": 0.6111,
+      "step": 661
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.2581265054695494e-05,
+      "loss": 0.6158,
+      "step": 662
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.230267937463977e-05,
+      "loss": 0.6135,
+      "step": 663
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.202457207128736e-05,
+      "loss": 0.5729,
+      "step": 664
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.174694593389455e-05,
+      "loss": 0.5397,
+      "step": 665
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.146980374689192e-05,
+      "loss": 0.6445,
+      "step": 666
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.119314828985619e-05,
+      "loss": 0.6251,
+      "step": 667
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.0916982337482644e-05,
+      "loss": 0.5926,
+      "step": 668
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.0641308659556964e-05,
+      "loss": 0.6072,
+      "step": 669
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.0366130020927624e-05,
+      "loss": 0.5806,
+      "step": 670
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.0091449181478164e-05,
+      "loss": 0.527,
+      "step": 671
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.981726889609952e-05,
+      "loss": 0.5451,
+      "step": 672
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.9543591914662314e-05,
+      "loss": 0.5558,
+      "step": 673
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.9270420981989294e-05,
+      "loss": 0.5638,
+      "step": 674
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.899775883782783e-05,
+      "loss": 0.5371,
+      "step": 675
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.872560821682256e-05,
+      "loss": 0.5877,
+      "step": 676
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.84539718484877e-05,
+      "loss": 0.5567,
+      "step": 677
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 4.818285245717984e-05,
+      "loss": 0.5194,
+      "step": 678
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 4.7912252762070576e-05,
+      "loss": 0.543,
+      "step": 679
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 4.764217547711934e-05,
+      "loss": 0.5293,
+      "step": 680
+    },
+    {
+      "epoch": 2.03,
+      "eval_loss": 0.9742249846458435,
+      "eval_runtime": 15.7393,
+      "eval_samples_per_second": 37.041,
+      "eval_steps_per_second": 18.552,
+      "step": 680
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 4.737262331104599e-05,
+      "loss": 0.5678,
+      "step": 681
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.710359896730379e-05,
+      "loss": 0.5677,
+      "step": 682
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.6835105144052185e-05,
+      "loss": 0.5456,
+      "step": 683
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.656714453412993e-05,
+      "loss": 0.5543,
+      "step": 684
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.6299719825027857e-05,
+      "loss": 0.5008,
+      "step": 685
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.6032833698862044e-05,
+      "loss": 0.5208,
+      "step": 686
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.5766488832346856e-05,
+      "loss": 0.5567,
+      "step": 687
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.5500687896768256e-05,
+      "loss": 0.5198,
+      "step": 688
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.5235433557956765e-05,
+      "loss": 0.5325,
+      "step": 689
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.497072847626087e-05,
+      "loss": 0.5146,
+      "step": 690
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.4706575306520415e-05,
+      "loss": 0.5634,
+      "step": 691
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.444297669803981e-05,
+      "loss": 0.5535,
+      "step": 692
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.4179935294561484e-05,
+      "loss": 0.5256,
+      "step": 693
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.3917453734239566e-05,
+      "loss": 0.5381,
+      "step": 694
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.365553464961313e-05,
+      "loss": 0.5325,
+      "step": 695
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.339418066758008e-05,
+      "loss": 0.5329,
+      "step": 696
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.3133394409370555e-05,
+      "loss": 0.5864,
+      "step": 697
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.287317849052075e-05,
+      "loss": 0.5089,
+      "step": 698
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.261353552084682e-05,
+      "loss": 0.558,
+      "step": 699
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.235446810441841e-05,
+      "loss": 0.5463,
+      "step": 700
+    },
+    {
+      "epoch": 2.09,
+      "eval_loss": 0.9766025543212891,
+      "eval_runtime": 15.7985,
+      "eval_samples_per_second": 36.902,
+      "eval_steps_per_second": 18.483,
+      "step": 700
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.209597883953278e-05,
+      "loss": 0.5534,
+      "step": 701
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.1838070318688604e-05,
+      "loss": 0.5145,
+      "step": 702
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.1580745128560136e-05,
+      "loss": 0.5339,
+      "step": 703
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.132400584997106e-05,
+      "loss": 0.5314,
+      "step": 704
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.106785505786872e-05,
+      "loss": 0.5625,
+      "step": 705
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.081229532129827e-05,
+      "loss": 0.5475,
+      "step": 706
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.055732920337699e-05,
+      "loss": 0.5275,
+      "step": 707
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.030295926126845e-05,
+      "loss": 0.5508,
+      "step": 708
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.0049188046156895e-05,
+      "loss": 0.4934,
+      "step": 709
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 3.979601810322169e-05,
+      "loss": 0.5456,
+      "step": 710
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 3.954345197161188e-05,
+      "loss": 0.5255,
+      "step": 711
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 3.929149218442052e-05,
+      "loss": 0.5487,
+      "step": 712
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 3.904014126865943e-05,
+      "loss": 0.5608,
+      "step": 713
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 3.878940174523371e-05,
+      "loss": 0.5703,
+      "step": 714
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 3.8539276128916734e-05,
+      "loss": 0.5448,
+      "step": 715
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 3.828976692832458e-05,
+      "loss": 0.5262,
+      "step": 716
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 3.8040876645891074e-05,
+      "loss": 0.578,
+      "step": 717
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 3.779260777784263e-05,
+      "loss": 0.5423,
+      "step": 718
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.754496281417331e-05,
+      "loss": 0.5093,
+      "step": 719
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.7297944238619706e-05,
+      "loss": 0.5184,
+      "step": 720
+    },
+    {
+      "epoch": 2.15,
+      "eval_loss": 0.9775969982147217,
+      "eval_runtime": 15.7755,
+      "eval_samples_per_second": 36.956,
+      "eval_steps_per_second": 18.51,
+      "step": 720
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.705155452863611e-05,
+      "loss": 0.4866,
+      "step": 721
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.680579615536961e-05,
+      "loss": 0.4973,
+      "step": 722
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.6560671583635467e-05,
+      "loss": 0.5039,
+      "step": 723
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.631618327189218e-05,
+      "loss": 0.5312,
+      "step": 724
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.607233367221691e-05,
+      "loss": 0.5225,
+      "step": 725
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.582912523028101e-05,
+      "loss": 0.5097,
+      "step": 726
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.5586560385325316e-05,
+      "loss": 0.4833,
+      "step": 727
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.534464157013574e-05,
+      "loss": 0.5754,
+      "step": 728
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.510337121101888e-05,
+      "loss": 0.5616,
+      "step": 729
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.4862751727777797e-05,
+      "loss": 0.5643,
+      "step": 730
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.462278553368753e-05,
+      "loss": 0.5297,
+      "step": 731
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.438347503547102e-05,
+      "loss": 0.5031,
+      "step": 732
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.414482263327504e-05,
+      "loss": 0.5497,
+      "step": 733
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.390683072064594e-05,
+      "loss": 0.5243,
+      "step": 734
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.366950168450573e-05,
+      "loss": 0.5568,
+      "step": 735
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.343283790512829e-05,
+      "loss": 0.503,
+      "step": 736
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.3196841756115174e-05,
+      "loss": 0.5319,
+      "step": 737
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.296151560437214e-05,
+      "loss": 0.5542,
+      "step": 738
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.272686181008516e-05,
+      "loss": 0.5373,
+      "step": 739
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.249288272669691e-05,
+      "loss": 0.5349,
+      "step": 740
+    },
+    {
+      "epoch": 2.21,
+      "eval_loss": 0.9783470034599304,
+      "eval_runtime": 15.7891,
+      "eval_samples_per_second": 36.924,
+      "eval_steps_per_second": 18.494,
+      "step": 740
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.2259580700883007e-05,
+      "loss": 0.5125,
+      "step": 741
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.202695807252871e-05,
+      "loss": 0.5274,
+      "step": 742
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.1795017174705224e-05,
+      "loss": 0.5413,
+      "step": 743
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.1563760333646395e-05,
+      "loss": 0.5015,
+      "step": 744
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.133318986872534e-05,
+      "loss": 0.4963,
+      "step": 745
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.110330809243134e-05,
+      "loss": 0.5475,
+      "step": 746
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.0874117310346403e-05,
+      "loss": 0.5404,
+      "step": 747
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.064561982112232e-05,
+      "loss": 0.5321,
+      "step": 748
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.041781791645748e-05,
+      "loss": 0.4827,
+      "step": 749
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.0190713881074105e-05,
+      "loss": 0.534,
+      "step": 750
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 2.996430999269505e-05,
+      "loss": 0.5661,
+      "step": 751
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 2.9738608522021173e-05,
+      "loss": 0.5479,
+      "step": 752
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 2.95136117327084e-05,
+      "loss": 0.5586,
+      "step": 753
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 2.9289321881345254e-05,
+      "loss": 0.5203,
+      "step": 754
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.9065741217429963e-05,
+      "loss": 0.5038,
+      "step": 755
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.8842871983347998e-05,
+      "loss": 0.5114,
+      "step": 756
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.8620716414349713e-05,
+      "loss": 0.5797,
+      "step": 757
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.8399276738527714e-05,
+      "loss": 0.5483,
+      "step": 758
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.817855517679464e-05,
+      "loss": 0.4896,
+      "step": 759
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.795855394286081e-05,
+      "loss": 0.5536,
+      "step": 760
+    },
+    {
+      "epoch": 2.27,
+      "eval_loss": 0.9793931245803833,
+      "eval_runtime": 15.8241,
+      "eval_samples_per_second": 36.843,
+      "eval_steps_per_second": 18.453,
+      "step": 760
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.773927524321218e-05,
+      "loss": 0.5497,
+      "step": 761
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.7520721277088024e-05,
+      "loss": 0.5666,
+      "step": 762
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.7302894236458943e-05,
+      "loss": 0.5539,
+      "step": 763
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.7085796306004906e-05,
+      "loss": 0.5784,
+      "step": 764
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.6869429663093403e-05,
+      "loss": 0.5333,
+      "step": 765
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.6653796477757432e-05,
+      "loss": 0.5448,
+      "step": 766
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.643889891267386e-05,
+      "loss": 0.5364,
+      "step": 767
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.6224739123141684e-05,
+      "loss": 0.5721,
+      "step": 768
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.6011319257060542e-05,
+      "loss": 0.5308,
+      "step": 769
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.5798641454908944e-05,
+      "loss": 0.5233,
+      "step": 770
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.558670784972298e-05,
+      "loss": 0.5204,
+      "step": 771
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.537552056707483e-05,
+      "loss": 0.5356,
+      "step": 772
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.5165081725051565e-05,
+      "loss": 0.5117,
+      "step": 773
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.4955393434233754e-05,
+      "loss": 0.5232,
+      "step": 774
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.4746457797674437e-05,
+      "loss": 0.5374,
+      "step": 775
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.45382769108779e-05,
+      "loss": 0.5726,
+      "step": 776
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.433085286177872e-05,
+      "loss": 0.5895,
+      "step": 777
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.4124187730720917e-05,
+      "loss": 0.5585,
+      "step": 778
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.3918283590436897e-05,
+      "loss": 0.5616,
+      "step": 779
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.3713142506026786e-05,
+      "loss": 0.5016,
+      "step": 780
+    },
+    {
+      "epoch": 2.33,
+      "eval_loss": 0.9822394847869873,
+      "eval_runtime": 15.7819,
+      "eval_samples_per_second": 36.941,
+      "eval_steps_per_second": 18.502,
+      "step": 780
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.3508766534937777e-05,
+      "loss": 0.5262,
+      "step": 781
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.3305157726943327e-05,
+      "loss": 0.4952,
+      "step": 782
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.3102318124122714e-05,
+      "loss": 0.5434,
+      "step": 783
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.290024976084052e-05,
+      "loss": 0.5562,
+      "step": 784
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.26989546637263e-05,
+      "loss": 0.5688,
+      "step": 785
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.2498434851654126e-05,
+      "loss": 0.515,
+      "step": 786
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.2298692335722405e-05,
+      "loss": 0.5532,
+      "step": 787
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.209972911923377e-05,
+      "loss": 0.528,
+      "step": 788
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.1901547197674886e-05,
+      "loss": 0.5528,
+      "step": 789
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.170414855869647e-05,
+      "loss": 0.5592,
+      "step": 790
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.1507535182093353e-05,
+      "loss": 0.5189,
+      "step": 791
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.1311709039784734e-05,
+      "loss": 0.5341,
+      "step": 792
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.11166720957942e-05,
+      "loss": 0.561,
+      "step": 793
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.092242630623016e-05,
+      "loss": 0.5406,
+      "step": 794
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.0728973619266178e-05,
+      "loss": 0.5595,
+      "step": 795
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.0536315975121544e-05,
+      "loss": 0.5085,
+      "step": 796
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.0344455306041632e-05,
+      "loss": 0.5427,
+      "step": 797
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.0153393536278653e-05,
+      "loss": 0.5149,
+      "step": 798
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 1.9963132582072276e-05,
+      "loss": 0.5064,
+      "step": 799
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 1.9773674351630545e-05,
+      "loss": 0.5075,
+      "step": 800
+    },
+    {
+      "epoch": 2.39,
+      "eval_loss": 0.9795373678207397,
+      "eval_runtime": 15.8228,
+      "eval_samples_per_second": 36.846,
+      "eval_steps_per_second": 18.454,
+      "step": 800
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 1.9585020745110548e-05,
+      "loss": 0.5385,
+      "step": 801
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.939717365459952e-05,
+      "loss": 0.5714,
+      "step": 802
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.9210134964095727e-05,
+      "loss": 0.585,
+      "step": 803
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.9023906549489767e-05,
+      "loss": 0.557,
+      "step": 804
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.883849027854553e-05,
+      "loss": 0.5037,
+      "step": 805
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.8653888010881637e-05,
+      "loss": 0.5276,
+      "step": 806
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.847010159795265e-05,
+      "loss": 0.5429,
+      "step": 807
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.82871328830307e-05,
+      "loss": 0.5263,
+      "step": 808
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.810498370118675e-05,
+      "loss": 0.5811,
+      "step": 809
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.7923655879272393e-05,
+      "loss": 0.502,
+      "step": 810
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.7743151235901402e-05,
+      "loss": 0.5171,
+      "step": 811
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.7563471581431624e-05,
+      "loss": 0.5251,
+      "step": 812
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.7384618717946656e-05,
+      "loss": 0.5244,
+      "step": 813
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.7206594439237865e-05,
+      "loss": 0.5929,
+      "step": 814
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.7029400530786478e-05,
+      "loss": 0.5313,
+      "step": 815
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.6853038769745467e-05,
+      "loss": 0.5314,
+      "step": 816
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.667751092492196e-05,
+      "loss": 0.5609,
+      "step": 817
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.6502818756759276e-05,
+      "loss": 0.5649,
+      "step": 818
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.6328964017319415e-05,
+      "loss": 0.5616,
+      "step": 819
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.61559484502655e-05,
+      "loss": 0.5529,
+      "step": 820
+    },
+    {
+      "epoch": 2.45,
+      "eval_loss": 0.9808087944984436,
+      "eval_runtime": 15.8,
+      "eval_samples_per_second": 36.899,
+      "eval_steps_per_second": 18.481,
+      "step": 820
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.598377379084419e-05,
+      "loss": 0.5312,
+      "step": 821
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.5812441765868292e-05,
+      "loss": 0.5136,
+      "step": 822
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.564195409369956e-05,
+      "loss": 0.542,
+      "step": 823
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.547231248423132e-05,
+      "loss": 0.5519,
+      "step": 824
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.530351863887136e-05,
+      "loss": 0.5249,
+      "step": 825
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.5135574250524897e-05,
+      "loss": 0.5617,
+      "step": 826
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.4968481003577629e-05,
+      "loss": 0.54,
+      "step": 827
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.4802240573878733e-05,
+      "loss": 0.4986,
+      "step": 828
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.463685462872415e-05,
+      "loss": 0.6019,
+      "step": 829
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.447232482683979e-05,
+      "loss": 0.5179,
+      "step": 830
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.4308652818365032e-05,
+      "loss": 0.5581,
+      "step": 831
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.4145840244835983e-05,
+      "loss": 0.5481,
+      "step": 832
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.398388873916916e-05,
+      "loss": 0.5319,
+      "step": 833
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.3822799925645036e-05,
+      "loss": 0.5239,
+      "step": 834
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.3662575419891822e-05,
+      "loss": 0.5148,
+      "step": 835
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.3503216828869192e-05,
+      "loss": 0.5335,
+      "step": 836
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.3344725750852183e-05,
+      "loss": 0.5306,
+      "step": 837
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.3187103775415156e-05,
+      "loss": 0.5706,
+      "step": 838
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.3030352483415964e-05,
+      "loss": 0.5218,
+      "step": 839
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.2874473446979918e-05,
+      "loss": 0.5168,
+      "step": 840
+    },
+    {
+      "epoch": 2.51,
+      "eval_loss": 0.978410005569458,
+      "eval_runtime": 15.7774,
+      "eval_samples_per_second": 36.952,
+      "eval_steps_per_second": 18.508,
+      "step": 840
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.2719468229484132e-05,
+      "loss": 0.5246,
+      "step": 841
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.2565338385541792e-05,
+      "loss": 0.5115,
+      "step": 842
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.2412085460986667e-05,
+      "loss": 0.5317,
+      "step": 843
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.2259710992857465e-05,
+      "loss": 0.5319,
+      "step": 844
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.2108216509382486e-05,
+      "loss": 0.6125,
+      "step": 845
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.195760352996429e-05,
+      "loss": 0.5218,
+      "step": 846
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.1807873565164506e-05,
+      "loss": 0.5266,
+      "step": 847
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.1659028116688575e-05,
+      "loss": 0.5449,
+      "step": 848
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.1511068677370752e-05,
+      "loss": 0.536,
+      "step": 849
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.1363996731159188e-05,
+      "loss": 0.5259,
+      "step": 850
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.1217813753100914e-05,
+      "loss": 0.5571,
+      "step": 851
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.107252120932717e-05,
+      "loss": 0.5138,
+      "step": 852
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.0928120557038613e-05,
+      "loss": 0.5398,
+      "step": 853
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.0784613244490816e-05,
+      "loss": 0.5787,
+      "step": 854
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.0642000710979572e-05,
+      "loss": 0.531,
+      "step": 855
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.0500284386826597e-05,
+      "loss": 0.4998,
+      "step": 856
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.035946569336519e-05,
+      "loss": 0.5416,
+      "step": 857
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.0219546042925843e-05,
+      "loss": 0.5065,
+      "step": 858
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.0080526838822214e-05,
+      "loss": 0.5334,
+      "step": 859
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 9.942409475337012e-06,
+      "loss": 0.5416,
+      "step": 860
+    },
+    {
+      "epoch": 2.57,
+      "eval_loss": 0.9793261289596558,
+      "eval_runtime": 15.8253,
+      "eval_samples_per_second": 36.84,
+      "eval_steps_per_second": 18.451,
+      "step": 860
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 9.805195337707962e-06,
+      "loss": 0.5044,
+      "step": 861
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 9.668885802114003e-06,
+      "loss": 0.5659,
+      "step": 862
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 9.533482235661395e-06,
+      "loss": 0.5259,
+      "step": 863
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 9.398985996370058e-06,
+      "loss": 0.5165,
+      "step": 864
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 9.265398433159933e-06,
+      "loss": 0.5081,
+      "step": 865
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.13272088583751e-06,
+      "loss": 0.4909,
+      "step": 866
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.000954685082286e-06,
+      "loss": 0.5524,
+      "step": 867
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 8.870101152433497e-06,
+      "loss": 0.5237,
+      "step": 868
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 8.74016160027683e-06,
+      "loss": 0.5485,
+      "step": 869
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 8.611137331831331e-06,
+      "loss": 0.5421,
+      "step": 870
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 8.483029641136208e-06,
+      "loss": 0.555,
+      "step": 871
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 8.355839813037936e-06,
+      "loss": 0.5149,
+      "step": 872
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 8.229569123177361e-06,
+      "loss": 0.5044,
+      "step": 873
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 8.10421883797694e-06,
+      "loss": 0.5369,
+      "step": 874
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 7.979790214627936e-06,
+      "loss": 0.5227,
+      "step": 875
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 7.856284501077926e-06,
+      "loss": 0.5643,
+      "step": 876
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 7.733702936018162e-06,
+      "loss": 0.5377,
+      "step": 877
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 7.612046748871327e-06,
+      "loss": 0.5318,
+      "step": 878
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 7.49131715977901e-06,
+      "loss": 0.5399,
+      "step": 879
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 7.371515379589555e-06,
+      "loss": 0.4845,
+      "step": 880
+    },
+    {
+      "epoch": 2.63,
+      "eval_loss": 0.9803735017776489,
+      "eval_runtime": 15.8,
+      "eval_samples_per_second": 36.899,
+      "eval_steps_per_second": 18.481,
+      "step": 880
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 7.252642609845972e-06,
+      "loss": 0.5182,
+      "step": 881
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 7.13470004277379e-06,
+      "loss": 0.5725,
+      "step": 882
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 7.017688861269145e-06,
+      "loss": 0.5472,
+      "step": 883
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 6.901610238886891e-06,
+      "loss": 0.5251,
+      "step": 884
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 6.786465339828918e-06,
+      "loss": 0.531,
+      "step": 885
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 6.672255318932341e-06,
+      "loss": 0.5927,
+      "step": 886
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 6.558981321658009e-06,
+      "loss": 0.5668,
+      "step": 887
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 6.4466444840789674e-06,
+      "loss": 0.5329,
+      "step": 888
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.335245932869149e-06,
+      "loss": 0.5379,
+      "step": 889
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.22478678529197e-06,
+      "loss": 0.5399,
+      "step": 890
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.115268149189146e-06,
+      "loss": 0.6011,
+      "step": 891
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.006691122969643e-06,
+      "loss": 0.5089,
+      "step": 892
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5.899056795598634e-06,
+      "loss": 0.5548,
+      "step": 893
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5.792366246586511e-06,
+      "loss": 0.5679,
+      "step": 894
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 5.686620545978161e-06,
+      "loss": 0.5909,
+      "step": 895
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.581820754342137e-06,
+      "loss": 0.5387,
+      "step": 896
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.477967922760141e-06,
+      "loss": 0.5588,
+      "step": 897
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.375063092816313e-06,
+      "loss": 0.5631,
+      "step": 898
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.273107296586988e-06,
+      "loss": 0.5144,
+      "step": 899
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.172101556630149e-06,
+      "loss": 0.5487,
+      "step": 900
+    },
+    {
+      "epoch": 2.69,
+      "eval_loss": 0.9801115989685059,
+      "eval_runtime": 15.7823,
+      "eval_samples_per_second": 36.94,
+      "eval_steps_per_second": 18.502,
+      "step": 900
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.07204688597529e-06,
+      "loss": 0.5086,
+      "step": 901
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 4.972944288113268e-06,
+      "loss": 0.5251,
+      "step": 902
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 4.874794756986145e-06,
+      "loss": 0.5155,
+      "step": 903
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 4.777599276977263e-06,
+      "loss": 0.5472,
+      "step": 904
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 4.6813588229014604e-06,
+      "loss": 0.5663,
+      "step": 905
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.586074359995119e-06,
+      "loss": 0.5339,
+      "step": 906
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.4917468439066305e-06,
+      "loss": 0.5613,
+      "step": 907
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.398377220686745e-06,
+      "loss": 0.5674,
+      "step": 908
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.305966426779118e-06,
+      "loss": 0.514,
+      "step": 909
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.214515389010865e-06,
+      "loss": 0.4964,
+      "step": 910
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.124025024583322e-06,
+      "loss": 0.5457,
+      "step": 911
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.034496241062824e-06,
+      "loss": 0.546,
+      "step": 912
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 3.945929936371595e-06,
+      "loss": 0.5478,
+      "step": 913
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 3.858326998778761e-06,
+      "loss": 0.5576,
+      "step": 914
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 3.7716883068913857e-06,
+      "loss": 0.5314,
+      "step": 915
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 3.6860147296457816e-06,
+      "loss": 0.5498,
+      "step": 916
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 3.6013071262986477e-06,
+      "loss": 0.5194,
+      "step": 917
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 3.5175663464185436e-06,
+      "loss": 0.558,
+      "step": 918
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.4347932298773534e-06,
+      "loss": 0.5827,
+      "step": 919
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.3529886068418447e-06,
+      "loss": 0.5313,
+      "step": 920
+    },
+    {
+      "epoch": 2.75,
+      "eval_loss": 0.9796981811523438,
+      "eval_runtime": 15.7992,
+      "eval_samples_per_second": 36.901,
+      "eval_steps_per_second": 18.482,
+      "step": 920
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.272153297765368e-06,
+      "loss": 0.5035,
+      "step": 921
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.1922881133795825e-06,
+      "loss": 0.4851,
+      "step": 922
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.113393854686353e-06,
+      "loss": 0.5566,
+      "step": 923
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.035471312949778e-06,
+      "loss": 0.5742,
+      "step": 924
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 2.9585212696881305e-06,
+      "loss": 0.5369,
+      "step": 925
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 2.8825444966661063e-06,
+      "loss": 0.5257,
+      "step": 926
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 2.807541755887033e-06,
+      "loss": 0.5616,
+      "step": 927
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 2.7335137995853188e-06,
+      "loss": 0.5439,
+      "step": 928
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.660461370218792e-06,
+      "loss": 0.5283,
+      "step": 929
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.5883852004613074e-06,
+      "loss": 0.5282,
+      "step": 930
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.517286013195397e-06,
+      "loss": 0.4983,
+      "step": 931
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.4471645215050743e-06,
+      "loss": 0.5368,
+      "step": 932
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.378021428668553e-06,
+      "loss": 0.5151,
+      "step": 933
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.3098574281513185e-06,
+      "loss": 0.5227,
+      "step": 934
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.242673203599077e-06,
+      "loss": 0.5189,
+      "step": 935
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.1764694288310184e-06,
+      "loss": 0.513,
+      "step": 936
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.1112467678329194e-06,
+      "loss": 0.5464,
+      "step": 937
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.0470058747505516e-06,
+      "loss": 0.523,
+      "step": 938
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.983747393883173e-06,
+      "loss": 0.5822,
+      "step": 939
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.921471959676957e-06,
+      "loss": 0.5449,
+      "step": 940
+    },
+    {
+      "epoch": 2.81,
+      "eval_loss": 0.9790334105491638,
+      "eval_runtime": 15.8611,
+      "eval_samples_per_second": 36.757,
+      "eval_steps_per_second": 18.41,
+      "step": 940
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.8601801967187082e-06,
+      "loss": 0.5143,
+      "step": 941
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.7998727197295784e-06,
+      "loss": 0.5507,
+      "step": 942
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.7405501335588825e-06,
+      "loss": 0.5496,
+      "step": 943
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.6822130331780484e-06,
+      "loss": 0.5647,
+      "step": 944
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.6248620036746764e-06,
+      "loss": 0.5253,
+      "step": 945
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.5684976202465784e-06,
+      "loss": 0.5047,
+      "step": 946
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.513120448196159e-06,
+      "loss": 0.5724,
+      "step": 947
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.4587310429245882e-06,
+      "loss": 0.52,
+      "step": 948
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.4053299499263373e-06,
+      "loss": 0.5329,
+      "step": 949
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.3529177047836627e-06,
+      "loss": 0.5219,
+      "step": 950
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.3014948331612653e-06,
+      "loss": 0.5593,
+      "step": 951
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.2510618508009608e-06,
+      "loss": 0.5455,
+      "step": 952
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.2016192635165956e-06,
+      "loss": 0.5117,
+      "step": 953
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.1531675671888619e-06,
+      "loss": 0.5259,
+      "step": 954
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.1057072477604346e-06,
+      "loss": 0.5116,
+      "step": 955
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.0592387812310311e-06,
+      "loss": 0.5504,
+      "step": 956
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.0137626336526597e-06,
+      "loss": 0.5314,
+      "step": 957
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 9.692792611249224e-07,
+      "loss": 0.5401,
+      "step": 958
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 9.257891097904758e-07,
+      "loss": 0.5175,
+      "step": 959
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 8.832926158305444e-07,
+      "loss": 0.5303,
+      "step": 960
+    },
+    {
+      "epoch": 2.87,
+      "eval_loss": 0.9794579148292542,
+      "eval_runtime": 15.7952,
+      "eval_samples_per_second": 36.91,
+      "eval_steps_per_second": 18.487,
+      "step": 960
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 8.417902054605132e-07,
+      "loss": 0.5114,
+      "step": 961
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 8.012822949256982e-07,
+      "loss": 0.4935,
+      "step": 962
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 7.617692904971719e-07,
+      "loss": 0.4967,
+      "step": 963
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 7.232515884676328e-07,
+      "loss": 0.5157,
+      "step": 964
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 6.85729575147498e-07,
+      "loss": 0.5303,
+      "step": 965
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 6.492036268609725e-07,
+      "loss": 0.5105,
+      "step": 966
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 6.136741099423416e-07,
+      "loss": 0.5506,
+      "step": 967
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 5.791413807322066e-07,
+      "loss": 0.5549,
+      "step": 968
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 5.456057855740216e-07,
+      "loss": 0.5616,
+      "step": 969
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5.130676608104845e-07,
+      "loss": 0.5393,
+      "step": 970
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 4.815273327803182e-07,
+      "loss": 0.5504,
+      "step": 971
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 4.509851178148505e-07,
+      "loss": 0.5233,
+      "step": 972
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 4.2144132223495e-07,
+      "loss": 0.527,
+      "step": 973
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 3.9289624234790656e-07,
+      "loss": 0.5097,
+      "step": 974
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 3.653501644444446e-07,
+      "loss": 0.499,
+      "step": 975
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 3.3880336479590325e-07,
+      "loss": 0.5475,
+      "step": 976
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 3.1325610965141636e-07,
+      "loss": 0.4882,
+      "step": 977
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 2.8870865523525915e-07,
+      "loss": 0.4904,
+      "step": 978
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 2.651612477442833e-07,
+      "loss": 0.5755,
+      "step": 979
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.4261412334546376e-07,
+      "loss": 0.5599,
+      "step": 980
+    },
+    {
+      "epoch": 2.93,
+      "eval_loss": 0.9795173406600952,
+      "eval_runtime": 15.7999,
+      "eval_samples_per_second": 36.899,
+      "eval_steps_per_second": 18.481,
+      "step": 980
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.210675081734892e-07,
+      "loss": 0.5297,
+      "step": 981
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.0052161832850856e-07,
+      "loss": 0.522,
+      "step": 982
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.8097665987399925e-07,
+      "loss": 0.5462,
+      "step": 983
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.624328288346577e-07,
+      "loss": 0.5331,
+      "step": 984
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.448903111944455e-07,
+      "loss": 0.5207,
+      "step": 985
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.2834928289472416e-07,
+      "loss": 0.5298,
+      "step": 986
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.1280990983248974e-07,
+      "loss": 0.5063,
+      "step": 987
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 9.827234785874107e-08,
+      "loss": 0.5693,
+      "step": 988
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 8.473674277685861e-08,
+      "loss": 0.5298,
+      "step": 989
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 7.220323034117238e-08,
+      "loss": 0.5077,
+      "step": 990
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 6.067193625560741e-08,
+      "loss": 0.5235,
+      "step": 991
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 5.0142976172429246e-08,
+      "loss": 0.5291,
+      "step": 992
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 4.0616455691033785e-08,
+      "loss": 0.5285,
+      "step": 993
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 3.2092470356948066e-08,
+      "loss": 0.5281,
+      "step": 994
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 2.4571105660853298e-08,
+      "loss": 0.5467,
+      "step": 995
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.8052437037707758e-08,
+      "loss": 0.4965,
+      "step": 996
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.2536529866014058e-08,
+      "loss": 0.5113,
+      "step": 997
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 8.023439467141902e-09,
+      "loss": 0.515,
+      "step": 998
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 4.51321110479519e-09,
+      "loss": 0.5373,
+      "step": 999
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 2.005879984556813e-09,
+      "loss": 0.544,
+      "step": 1000
+    },
+    {
+      "epoch": 2.99,
+      "eval_loss": 0.9790173768997192,
+      "eval_runtime": 15.7885,
+      "eval_samples_per_second": 36.926,
+      "eval_steps_per_second": 18.494,
+      "step": 1000
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 5.014712535111876e-10,
+      "loss": 0.5006,
+      "step": 1001
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 0.0,
+      "loss": 0.5376,
+      "step": 1002
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1002,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "total_flos": 1.41733434524015e+18,
+  "trial_name": null,
+  "trial_params": null
+}