diff --git "a/checkpoint-2821/trainer_state.json" "b/checkpoint-2821/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-2821/trainer_state.json"
@@ -0,0 +1,18073 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 20,
+  "global_step": 2821,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9999999999999997e-05,
+      "loss": 0.9379,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5.9999999999999995e-05,
+      "loss": 0.9398,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.999999999999999e-05,
+      "loss": 1.1066,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.00011999999999999999,
+      "loss": 0.9642,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.00015,
+      "loss": 0.972,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.00017999999999999998,
+      "loss": 0.9888,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.00020999999999999998,
+      "loss": 0.8477,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00023999999999999998,
+      "loss": 0.893,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00027,
+      "loss": 0.8521,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0003,
+      "loss": 0.7497,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00029999995843420544,
+      "loss": 0.8077,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00029999983373684477,
+      "loss": 0.856,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00029999962590798713,
+      "loss": 0.7678,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0002999993349477477,
+      "loss": 0.7234,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0002999989608562877,
+      "loss": 0.704,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0002999985036338145,
+      "loss": 0.8246,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0002999979632805815,
+      "loss": 0.8171,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0002999973397968881,
+      "loss": 0.708,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00029999663318307993,
+      "loss": 0.8104,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0002999958434395485,
+      "loss": 0.7494,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "eval_loss": 1.6390020847320557,
+      "eval_runtime": 42.541,
+      "eval_samples_per_second": 13.704,
+      "eval_steps_per_second": 13.704,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00029999497056673157,
+      "loss": 0.7424,
+      "step": 21
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00029999401456511294,
+      "loss": 0.7952,
+      "step": 22
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0002999929754352223,
+      "loss": 0.8398,
+      "step": 23
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0002999918531776357,
+      "loss": 0.652,
+      "step": 24
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00029999064779297495,
+      "loss": 0.8793,
+      "step": 25
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0002999893592819082,
+      "loss": 0.8673,
+      "step": 26
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00029998798764514957,
+      "loss": 0.7446,
+      "step": 27
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0002999865328834591,
+      "loss": 0.7848,
+      "step": 28
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0002999849949976432,
+      "loss": 0.7745,
+      "step": 29
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0002999833739885541,
+      "loss": 0.8539,
+      "step": 30
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0002999816698570901,
+      "loss": 0.7206,
+      "step": 31
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0002999798826041958,
+      "loss": 0.6571,
+      "step": 32
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0002999780122308616,
+      "loss": 0.6546,
+      "step": 33
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0002999760587381242,
+      "loss": 0.7666,
+      "step": 34
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00029997402212706613,
+      "loss": 0.7754,
+      "step": 35
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0002999719023988162,
+      "loss": 0.7621,
+      "step": 36
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0002999696995545491,
+      "loss": 0.8395,
+      "step": 37
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0002999674135954857,
+      "loss": 0.8533,
+      "step": 38
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0002999650445228928,
+      "loss": 0.6957,
+      "step": 39
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00029996259233808366,
+      "loss": 0.8382,
+      "step": 40
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 1.3849129676818848,
+      "eval_runtime": 46.2711,
+      "eval_samples_per_second": 12.6,
+      "eval_steps_per_second": 12.6,
+      "step": 40
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00029996005704241695,
+      "loss": 0.8747,
+      "step": 41
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00029995743863729803,
+      "loss": 0.7024,
+      "step": 42
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0002999547371241779,
+      "loss": 0.7601,
+      "step": 43
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0002999519525045538,
+      "loss": 0.761,
+      "step": 44
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0002999490847799691,
+      "loss": 0.7525,
+      "step": 45
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0002999461339520129,
+      "loss": 0.6742,
+      "step": 46
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0002999431000223208,
+      "loss": 0.6688,
+      "step": 47
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0002999399829925741,
+      "loss": 0.7927,
+      "step": 48
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0002999367828645003,
+      "loss": 0.7954,
+      "step": 49
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00029993349963987304,
+      "loss": 0.683,
+      "step": 50
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00029993013332051187,
+      "loss": 0.7543,
+      "step": 51
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002999266839082824,
+      "loss": 0.731,
+      "step": 52
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002999231514050963,
+      "loss": 0.7042,
+      "step": 53
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002999195358129115,
+      "loss": 0.736,
+      "step": 54
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002999158371337316,
+      "loss": 0.6867,
+      "step": 55
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002999120553696065,
+      "loss": 0.7932,
+      "step": 56
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00029990819052263216,
+      "loss": 0.7664,
+      "step": 57
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002999042425949505,
+      "loss": 0.7371,
+      "step": 58
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002999002115887495,
+      "loss": 0.8349,
+      "step": 59
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00029989609750626306,
+      "loss": 0.7429,
+      "step": 60
+    },
+    {
+      "epoch": 0.04,
+      "eval_loss": 1.3155401945114136,
+      "eval_runtime": 47.3148,
+      "eval_samples_per_second": 12.322,
+      "eval_steps_per_second": 12.322,
+      "step": 60
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00029989190034977146,
+      "loss": 0.76,
+      "step": 61
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002998876201216007,
+      "loss": 0.6611,
+      "step": 62
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002998832568241229,
+      "loss": 0.7162,
+      "step": 63
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0002998788104597563,
+      "loss": 0.6956,
+      "step": 64
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00029987428103096504,
+      "loss": 0.8829,
+      "step": 65
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0002998696685402595,
+      "loss": 0.7843,
+      "step": 66
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0002998649729901959,
+      "loss": 0.7435,
+      "step": 67
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0002998601943833766,
+      "loss": 0.7351,
+      "step": 68
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0002998553327224499,
+      "loss": 0.6892,
+      "step": 69
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00029985038801011015,
+      "loss": 0.7294,
+      "step": 70
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0002998453602490979,
+      "loss": 0.797,
+      "step": 71
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0002998402494421995,
+      "loss": 0.8509,
+      "step": 72
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0002998350555922474,
+      "loss": 0.5215,
+      "step": 73
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0002998297787021202,
+      "loss": 0.7482,
+      "step": 74
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0002998244187747422,
+      "loss": 0.7166,
+      "step": 75
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0002998189758130841,
+      "loss": 0.7486,
+      "step": 76
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0002998134498201624,
+      "loss": 0.7109,
+      "step": 77
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0002998078407990397,
+      "loss": 0.6915,
+      "step": 78
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00029980214875282446,
+      "loss": 0.7674,
+      "step": 79
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0002997963736846714,
+      "loss": 0.757,
+      "step": 80
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 1.2954022884368896,
+      "eval_runtime": 47.843,
+      "eval_samples_per_second": 12.186,
+      "eval_steps_per_second": 12.186,
+      "step": 80
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0002997905155977811,
+      "loss": 0.6575,
+      "step": 81
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0002997845744954001,
+      "loss": 0.7656,
+      "step": 82
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00029977855038082105,
+      "loss": 0.6814,
+      "step": 83
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0002997724432573827,
+      "loss": 0.685,
+      "step": 84
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0002997662531284695,
+      "loss": 0.6517,
+      "step": 85
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00029975997999751215,
+      "loss": 0.7845,
+      "step": 86
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00029975362386798737,
+      "loss": 0.7487,
+      "step": 87
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00029974718474341767,
+      "loss": 0.7445,
+      "step": 88
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00029974066262737175,
+      "loss": 0.745,
+      "step": 89
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0002997340575234642,
+      "loss": 0.7934,
+      "step": 90
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00029972736943535575,
+      "loss": 0.7288,
+      "step": 91
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0002997205983667528,
+      "loss": 0.7701,
+      "step": 92
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0002997137443214082,
+      "loss": 0.7277,
+      "step": 93
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00029970680730312036,
+      "loss": 0.7406,
+      "step": 94
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0002996997873157338,
+      "loss": 0.731,
+      "step": 95
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00029969268436313924,
+      "loss": 0.8182,
+      "step": 96
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0002996854984492732,
+      "loss": 0.7659,
+      "step": 97
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00029967822957811796,
+      "loss": 0.7462,
+      "step": 98
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00029967087775370233,
+      "loss": 0.8292,
+      "step": 99
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00029966344298010055,
+      "loss": 0.7559,
+      "step": 100
+    },
+    {
+      "epoch": 0.07,
+      "eval_loss": 1.2370859384536743,
+      "eval_runtime": 48.0849,
+      "eval_samples_per_second": 12.124,
+      "eval_steps_per_second": 12.124,
+      "step": 100
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0002996559252614331,
+      "loss": 0.6714,
+      "step": 101
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0002996483246018664,
+      "loss": 0.7865,
+      "step": 102
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0002996406410056128,
+      "loss": 0.7562,
+      "step": 103
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0002996328744769306,
+      "loss": 0.7267,
+      "step": 104
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0002996250250201242,
+      "loss": 0.6713,
+      "step": 105
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00029961709263954373,
+      "loss": 0.7054,
+      "step": 106
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00029960907733958547,
+      "loss": 0.8033,
+      "step": 107
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00029960097912469155,
+      "loss": 0.7956,
+      "step": 108
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0002995927979993501,
+      "loss": 0.7944,
+      "step": 109
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0002995845339680952,
+      "loss": 0.6649,
+      "step": 110
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0002995761870355069,
+      "loss": 0.7295,
+      "step": 111
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.000299567757206211,
+      "loss": 0.7191,
+      "step": 112
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0002995592444848796,
+      "loss": 0.7253,
+      "step": 113
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0002995506488762304,
+      "loss": 0.8292,
+      "step": 114
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0002995419703850272,
+      "loss": 0.7349,
+      "step": 115
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00029953320901607983,
+      "loss": 0.8984,
+      "step": 116
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0002995243647742438,
+      "loss": 0.7994,
+      "step": 117
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0002995154376644207,
+      "loss": 0.7883,
+      "step": 118
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0002995064276915581,
+      "loss": 0.7352,
+      "step": 119
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0002994973348606494,
+      "loss": 0.7796,
+      "step": 120
+    },
+    {
+      "epoch": 0.09,
+      "eval_loss": 1.2445749044418335,
+      "eval_runtime": 48.3442,
+      "eval_samples_per_second": 12.059,
+      "eval_steps_per_second": 12.059,
+      "step": 120
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0002994881591767338,
+      "loss": 0.6947,
+      "step": 121
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00029947890064489676,
+      "loss": 0.7908,
+      "step": 122
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0002994695592702694,
+      "loss": 0.7952,
+      "step": 123
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0002994601350580287,
+      "loss": 0.7262,
+      "step": 124
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00029945062801339784,
+      "loss": 0.64,
+      "step": 125
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00029944103814164556,
+      "loss": 0.6402,
+      "step": 126
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0002994313654480868,
+      "loss": 0.7248,
+      "step": 127
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00029942160993808216,
+      "loss": 0.776,
+      "step": 128
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0002994117716170383,
+      "loss": 0.7079,
+      "step": 129
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00029940185049040773,
+      "loss": 0.7007,
+      "step": 130
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0002993918465636888,
+      "loss": 0.7695,
+      "step": 131
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00029938175984242584,
+      "loss": 0.6742,
+      "step": 132
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00029937159033220905,
+      "loss": 0.7125,
+      "step": 133
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00029936133803867436,
+      "loss": 0.7216,
+      "step": 134
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00029935100296750374,
+      "loss": 0.6479,
+      "step": 135
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00029934058512442514,
+      "loss": 0.7285,
+      "step": 136
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00029933008451521204,
+      "loss": 0.6967,
+      "step": 137
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00029931950114568407,
+      "loss": 0.6871,
+      "step": 138
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0002993088350217067,
+      "loss": 0.7207,
+      "step": 139
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0002992980861491911,
+      "loss": 0.6594,
+      "step": 140
+    },
+    {
+      "epoch": 0.1,
+      "eval_loss": 1.2339797019958496,
+      "eval_runtime": 46.3761,
+      "eval_samples_per_second": 12.571,
+      "eval_steps_per_second": 12.571,
+      "step": 140
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0002992872545340945,
+      "loss": 0.9022,
+      "step": 141
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00029927634018241984,
+      "loss": 0.7138,
+      "step": 142
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00029926534310021605,
+      "loss": 0.7419,
+      "step": 143
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00029925426329357774,
+      "loss": 0.6271,
+      "step": 144
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0002992431007686454,
+      "loss": 0.8016,
+      "step": 145
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00029923185553160564,
+      "loss": 0.6689,
+      "step": 146
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0002992205275886905,
+      "loss": 0.786,
+      "step": 147
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0002992091169461781,
+      "loss": 0.7337,
+      "step": 148
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00029919762361039234,
+      "loss": 0.7181,
+      "step": 149
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00029918604758770293,
+      "loss": 0.8497,
+      "step": 150
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00029917438888452547,
+      "loss": 0.7384,
+      "step": 151
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0002991626475073213,
+      "loss": 0.6907,
+      "step": 152
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0002991508234625977,
+      "loss": 0.793,
+      "step": 153
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0002991389167569075,
+      "loss": 0.7325,
+      "step": 154
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00029912692739684967,
+      "loss": 0.7067,
+      "step": 155
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00029911485538906883,
+      "loss": 0.8082,
+      "step": 156
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00029910270074025536,
+      "loss": 0.8204,
+      "step": 157
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0002990904634571455,
+      "loss": 0.61,
+      "step": 158
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00029907814354652133,
+      "loss": 0.7118,
+      "step": 159
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0002990657410152106,
+      "loss": 0.7793,
+      "step": 160
+    },
+    {
+      "epoch": 0.11,
+      "eval_loss": 1.2419909238815308,
+      "eval_runtime": 46.3939,
+      "eval_samples_per_second": 12.566,
+      "eval_steps_per_second": 12.566,
+      "step": 160
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.000299053255870087,
+      "loss": 0.7281,
+      "step": 161
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00029904068811806994,
+      "loss": 0.7379,
+      "step": 162
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00029902803776612455,
+      "loss": 0.7301,
+      "step": 163
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00029901530482126173,
+      "loss": 0.7148,
+      "step": 164
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0002990024892905384,
+      "loss": 0.6958,
+      "step": 165
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00029898959118105685,
+      "loss": 0.6493,
+      "step": 166
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0002989766104999655,
+      "loss": 0.776,
+      "step": 167
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00029896354725445824,
+      "loss": 0.6049,
+      "step": 168
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00029895040145177506,
+      "loss": 0.6609,
+      "step": 169
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0002989371730992013,
+      "loss": 0.7357,
+      "step": 170
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00029892386220406843,
+      "loss": 0.7805,
+      "step": 171
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0002989104687737533,
+      "loss": 0.6194,
+      "step": 172
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0002988969928156789,
+      "loss": 0.7679,
+      "step": 173
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00029888343433731357,
+      "loss": 0.6083,
+      "step": 174
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00029886979334617164,
+      "loss": 0.6812,
+      "step": 175
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0002988560698498131,
+      "loss": 0.7854,
+      "step": 176
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00029884226385584365,
+      "loss": 0.6972,
+      "step": 177
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0002988283753719147,
+      "loss": 0.7034,
+      "step": 178
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00029881440440572343,
+      "loss": 0.686,
+      "step": 179
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00029880035096501264,
+      "loss": 0.8207,
+      "step": 180
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 1.241647720336914,
+      "eval_runtime": 48.3192,
+      "eval_samples_per_second": 12.066,
+      "eval_steps_per_second": 12.066,
+      "step": 180
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0002987862150575709,
+      "loss": 0.6293,
+      "step": 181
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00029877199669123256,
+      "loss": 0.794,
+      "step": 182
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00029875769587387754,
+      "loss": 0.7091,
+      "step": 183
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0002987433126134315,
+      "loss": 0.6735,
+      "step": 184
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00029872884691786576,
+      "loss": 0.7532,
+      "step": 185
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0002987142987951974,
+      "loss": 0.6701,
+      "step": 186
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00029869966825348915,
+      "loss": 0.7529,
+      "step": 187
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00029868495530084945,
+      "loss": 0.5815,
+      "step": 188
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0002986701599454323,
+      "loss": 0.5967,
+      "step": 189
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0002986552821954374,
+      "loss": 0.619,
+      "step": 190
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0002986403220591103,
+      "loss": 0.7565,
+      "step": 191
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.000298625279544742,
+      "loss": 0.5599,
+      "step": 192
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0002986101546606692,
+      "loss": 0.5309,
+      "step": 193
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00029859494741527433,
+      "loss": 0.662,
+      "step": 194
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0002985796578169853,
+      "loss": 0.6789,
+      "step": 195
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0002985642858742758,
+      "loss": 0.5575,
+      "step": 196
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00029854883159566526,
+      "loss": 0.7487,
+      "step": 197
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0002985332949897184,
+      "loss": 0.7211,
+      "step": 198
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00029851767606504595,
+      "loss": 0.7628,
+      "step": 199
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00029850197483030394,
+      "loss": 0.7325,
+      "step": 200
+    },
+    {
+      "epoch": 0.14,
+      "eval_loss": 1.1919026374816895,
+      "eval_runtime": 48.2679,
+      "eval_samples_per_second": 12.078,
+      "eval_steps_per_second": 12.078,
+      "step": 200
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00029848619129419423,
+      "loss": 0.668,
+      "step": 201
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0002984703254654642,
+      "loss": 0.599,
+      "step": 202
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0002984543773529069,
+      "loss": 0.7292,
+      "step": 203
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.000298438346965361,
+      "loss": 0.7393,
+      "step": 204
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002984222343117105,
+      "loss": 0.7279,
+      "step": 205
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002984060394008854,
+      "loss": 0.6017,
+      "step": 206
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.000298389762241861,
+      "loss": 0.6933,
+      "step": 207
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00029837340284365833,
+      "loss": 0.6825,
+      "step": 208
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002983569612153439,
+      "loss": 0.7295,
+      "step": 209
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002983404373660298,
+      "loss": 0.7763,
+      "step": 210
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00029832383130487387,
+      "loss": 0.7473,
+      "step": 211
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002983071430410792,
+      "loss": 0.645,
+      "step": 212
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002982903725838946,
+      "loss": 0.6389,
+      "step": 213
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002982735199426146,
+      "loss": 0.7145,
+      "step": 214
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00029825658512657897,
+      "loss": 0.5955,
+      "step": 215
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002982395681451732,
+      "loss": 0.769,
+      "step": 216
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00029822246900782835,
+      "loss": 0.6127,
+      "step": 217
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00029820528772402083,
+      "loss": 0.5397,
+      "step": 218
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00029818802430327273,
+      "loss": 0.8308,
+      "step": 219
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00029817067875515164,
+      "loss": 0.7186,
+      "step": 220
+    },
+    {
+      "epoch": 0.16,
+      "eval_loss": 1.1844402551651,
+      "eval_runtime": 48.1289,
+      "eval_samples_per_second": 12.113,
+      "eval_steps_per_second": 12.113,
+      "step": 220
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00029815325108927063,
+      "loss": 0.6424,
+      "step": 221
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0002981357413152883,
+      "loss": 0.6085,
+      "step": 222
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00029811814944290883,
+      "loss": 0.6565,
+      "step": 223
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00029810047548188166,
+      "loss": 0.7385,
+      "step": 224
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00029808271944200204,
+      "loss": 0.6982,
+      "step": 225
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00029806488133311045,
+      "loss": 0.8407,
+      "step": 226
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.000298046961165093,
+      "loss": 0.8323,
+      "step": 227
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00029802895894788127,
+      "loss": 0.7296,
+      "step": 228
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0002980108746914522,
+      "loss": 0.6523,
+      "step": 229
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0002979927084058283,
+      "loss": 0.7354,
+      "step": 230
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0002979744601010777,
+      "loss": 0.6686,
+      "step": 231
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0002979561297873135,
+      "loss": 0.7228,
+      "step": 232
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0002979377174746947,
+      "loss": 0.7499,
+      "step": 233
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00029791922317342565,
+      "loss": 0.6288,
+      "step": 234
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00029790064689375604,
+      "loss": 0.7871,
+      "step": 235
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.000297881988645981,
+      "loss": 0.7477,
+      "step": 236
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00029786324844044117,
+      "loss": 0.5985,
+      "step": 237
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0002978444262875226,
+      "loss": 0.6406,
+      "step": 238
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00029782552219765663,
+      "loss": 0.6528,
+      "step": 239
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0002978065361813202,
+      "loss": 0.7146,
+      "step": 240
+    },
+    {
+      "epoch": 0.17,
+      "eval_loss": 1.1583514213562012,
+      "eval_runtime": 46.8088,
+      "eval_samples_per_second": 12.455,
+      "eval_steps_per_second": 12.455,
+      "step": 240
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00029778746824903556,
+      "loss": 0.7482,
+      "step": 241
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0002977683184113703,
+      "loss": 0.7816,
+      "step": 242
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0002977490866789375,
+      "loss": 0.638,
+      "step": 243
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00029772977306239556,
+      "loss": 0.645,
+      "step": 244
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00029771037757244837,
+      "loss": 0.5682,
+      "step": 245
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00029769090021984504,
+      "loss": 0.657,
+      "step": 246
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0002976713410153801,
+      "loss": 0.6963,
+      "step": 247
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00029765169996989353,
+      "loss": 0.7383,
+      "step": 248
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00029763197709427064,
+      "loss": 0.6571,
+      "step": 249
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.000297612172399442,
+      "loss": 0.7635,
+      "step": 250
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00029759228589638356,
+      "loss": 0.7174,
+      "step": 251
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00029757231759611663,
+      "loss": 0.7353,
+      "step": 252
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00029755226750970793,
+      "loss": 0.6811,
+      "step": 253
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00029753213564826935,
+      "loss": 0.6433,
+      "step": 254
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0002975119220229582,
+      "loss": 0.8396,
+      "step": 255
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0002974916266449771,
+      "loss": 0.601,
+      "step": 256
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.000297471249525574,
+      "loss": 0.697,
+      "step": 257
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0002974507906760421,
+      "loss": 0.6707,
+      "step": 258
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0002974302501077198,
+      "loss": 0.7132,
+      "step": 259
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00029740962783199104,
+      "loss": 0.7288,
+      "step": 260
+    },
+    {
+      "epoch": 0.18,
+      "eval_loss": 1.1574283838272095,
+      "eval_runtime": 46.7635,
+      "eval_samples_per_second": 12.467,
+      "eval_steps_per_second": 12.467,
+      "step": 260
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00029738892386028487,
+      "loss": 0.6959,
+      "step": 261
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00029736813820407563,
+      "loss": 0.7881,
+      "step": 262
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0002973472708748829,
+      "loss": 0.615,
+      "step": 263
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00029732632188427167,
+      "loss": 0.5948,
+      "step": 264
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00029730529124385205,
+      "loss": 0.8122,
+      "step": 265
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0002972841789652794,
+      "loss": 0.7612,
+      "step": 266
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00029726298506025446,
+      "loss": 0.5298,
+      "step": 267
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0002972417095405231,
+      "loss": 0.6862,
+      "step": 268
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00029722035241787635,
+      "loss": 0.8032,
+      "step": 269
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0002971989137041507,
+      "loss": 0.6059,
+      "step": 270
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0002971773934112275,
+      "loss": 0.8086,
+      "step": 271
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0002971557915510337,
+      "loss": 0.7664,
+      "step": 272
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0002971341081355412,
+      "loss": 0.6734,
+      "step": 273
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0002971123431767673,
+      "loss": 0.6785,
+      "step": 274
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0002970904966867742,
+      "loss": 0.6737,
+      "step": 275
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00029706856867766957,
+      "loss": 0.6802,
+      "step": 276
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0002970465591616061,
+      "loss": 0.7044,
+      "step": 277
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0002970244681507817,
+      "loss": 0.7564,
+      "step": 278
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00029700229565743944,
+      "loss": 0.7577,
+      "step": 279
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0002969800416938676,
+      "loss": 0.7928,
+      "step": 280
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 1.1896907091140747,
+      "eval_runtime": 46.8345,
+      "eval_samples_per_second": 12.448,
+      "eval_steps_per_second": 12.448,
+      "step": 280
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00029695770627239943,
+      "loss": 0.6624,
+      "step": 281
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00029693528940541366,
+      "loss": 0.6706,
+      "step": 282
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0002969127911053338,
+      "loss": 0.6857,
+      "step": 283
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00029689021138462867,
+      "loss": 0.6765,
+      "step": 284
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0002968675502558122,
+      "loss": 0.6533,
+      "step": 285
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00029684480773144346,
+      "loss": 0.7015,
+      "step": 286
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00029682198382412656,
+      "loss": 0.8835,
+      "step": 287
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0002967990785465108,
+      "loss": 0.6116,
+      "step": 288
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0002967760919112905,
+      "loss": 0.8762,
+      "step": 289
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.000296753023931205,
+      "loss": 0.5218,
+      "step": 290
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.000296729874619039,
+      "loss": 0.7398,
+      "step": 291
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.000296706643987622,
+      "loss": 0.6898,
+      "step": 292
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00029668333204982866,
+      "loss": 0.7513,
+      "step": 293
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00029665993881857866,
+      "loss": 0.7106,
+      "step": 294
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00029663646430683687,
+      "loss": 0.8256,
+      "step": 295
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0002966129085276131,
+      "loss": 0.7573,
+      "step": 296
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0002965892714939621,
+      "loss": 0.6883,
+      "step": 297
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00029656555321898396,
+      "loss": 0.6114,
+      "step": 298
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00029654175371582336,
+      "loss": 0.7118,
+      "step": 299
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0002965178729976704,
+      "loss": 0.6242,
+      "step": 300
+    },
+    {
+      "epoch": 0.21,
+      "eval_loss": 1.1463284492492676,
+      "eval_runtime": 46.7167,
+      "eval_samples_per_second": 12.479,
+      "eval_steps_per_second": 12.479,
+      "step": 300
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00029649391107776005,
+      "loss": 0.7476,
+      "step": 301
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00029646986796937215,
+      "loss": 0.7118,
+      "step": 302
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0002964457436858317,
+      "loss": 0.6603,
+      "step": 303
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00029642153824050864,
+      "loss": 0.7919,
+      "step": 304
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00029639725164681786,
+      "loss": 0.7787,
+      "step": 305
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00029637288391821926,
+      "loss": 0.7042,
+      "step": 306
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00029634843506821766,
+      "loss": 0.6558,
+      "step": 307
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00029632390511036294,
+      "loss": 0.7338,
+      "step": 308
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00029629929405824984,
+      "loss": 0.7057,
+      "step": 309
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.000296274601925518,
+      "loss": 0.6143,
+      "step": 310
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0002962498287258522,
+      "loss": 0.7085,
+      "step": 311
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00029622497447298184,
+      "loss": 0.6643,
+      "step": 312
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0002962000391806816,
+      "loss": 0.702,
+      "step": 313
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0002961750228627707,
+      "loss": 0.6915,
+      "step": 314
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00029614992553311353,
+      "loss": 0.681,
+      "step": 315
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00029612474720561925,
+      "loss": 0.8107,
+      "step": 316
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00029609948789424207,
+      "loss": 0.8707,
+      "step": 317
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0002960741476129809,
+      "loss": 0.6956,
+      "step": 318
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0002960487263758795,
+      "loss": 0.6908,
+      "step": 319
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0002960232241970268,
+      "loss": 0.8006,
+      "step": 320
+    },
+    {
+      "epoch": 0.23,
+      "eval_loss": 1.139107346534729,
+      "eval_runtime": 46.7127,
+      "eval_samples_per_second": 12.481,
+      "eval_steps_per_second": 12.481,
+      "step": 320
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0002959976410905561,
+      "loss": 0.7075,
+      "step": 321
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00029597197707064614,
+      "loss": 0.692,
+      "step": 322
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00029594623215151996,
+      "loss": 0.6238,
+      "step": 323
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00029592040634744574,
+      "loss": 0.5914,
+      "step": 324
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00029589449967273645,
+      "loss": 0.7838,
+      "step": 325
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00029586851214174977,
+      "loss": 0.8163,
+      "step": 326
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00029584244376888827,
+      "loss": 0.7545,
+      "step": 327
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00029581629456859945,
+      "loss": 0.7105,
+      "step": 328
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0002957900645553753,
+      "loss": 0.76,
+      "step": 329
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0002957637537437529,
+      "loss": 0.6947,
+      "step": 330
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0002957373621483139,
+      "loss": 0.7698,
+      "step": 331
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00029571088978368486,
+      "loss": 0.7208,
+      "step": 332
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.000295684336664537,
+      "loss": 0.6456,
+      "step": 333
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0002956577028055863,
+      "loss": 0.6778,
+      "step": 334
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00029563098822159365,
+      "loss": 0.695,
+      "step": 335
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00029560419292736444,
+      "loss": 0.8475,
+      "step": 336
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00029557731693774904,
+      "loss": 0.6868,
+      "step": 337
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00029555036026764227,
+      "loss": 0.7934,
+      "step": 338
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00029552332293198384,
+      "loss": 0.6624,
+      "step": 339
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00029549620494575814,
+      "loss": 0.7201,
+      "step": 340
+    },
+    {
+      "epoch": 0.24,
+      "eval_loss": 1.1492714881896973,
+      "eval_runtime": 46.9733,
+      "eval_samples_per_second": 12.411,
+      "eval_steps_per_second": 12.411,
+      "step": 340
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00029546900632399425,
+      "loss": 0.744,
+      "step": 341
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00029544172708176596,
+      "loss": 0.7018,
+      "step": 342
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00029541436723419165,
+      "loss": 0.7419,
+      "step": 343
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0002953869267964345,
+      "loss": 0.6734,
+      "step": 344
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00029535940578370224,
+      "loss": 0.6631,
+      "step": 345
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00029533180421124737,
+      "loss": 0.7635,
+      "step": 346
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0002953041220943669,
+      "loss": 0.6861,
+      "step": 347
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0002952763594484026,
+      "loss": 0.7354,
+      "step": 348
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0002952485162887408,
+      "loss": 0.8062,
+      "step": 349
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0002952205926308125,
+      "loss": 0.7205,
+      "step": 350
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00029519258849009325,
+      "loss": 0.7025,
+      "step": 351
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0002951645038821033,
+      "loss": 0.6339,
+      "step": 352
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0002951363388224074,
+      "loss": 0.6724,
+      "step": 353
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00029510809332661486,
+      "loss": 0.8575,
+      "step": 354
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0002950797674103798,
+      "loss": 0.6966,
+      "step": 355
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00029505136108940057,
+      "loss": 0.7674,
+      "step": 356
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00029502287437942036,
+      "loss": 0.7754,
+      "step": 357
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00029499430729622674,
+      "loss": 0.6445,
+      "step": 358
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0002949656598556519,
+      "loss": 0.6201,
+      "step": 359
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0002949369320735726,
+      "loss": 0.7163,
+      "step": 360
+    },
+    {
+      "epoch": 0.26,
+      "eval_loss": 1.1293282508850098,
+      "eval_runtime": 47.0594,
+      "eval_samples_per_second": 12.389,
+      "eval_steps_per_second": 12.389,
+      "step": 360
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0002949081239659101,
+      "loss": 0.7057,
+      "step": 361
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0002948792355486301,
+      "loss": 0.7228,
+      "step": 362
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00029485026683774286,
+      "loss": 0.7344,
+      "step": 363
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0002948212178493032,
+      "loss": 0.7575,
+      "step": 364
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0002947920885994103,
+      "loss": 0.6831,
+      "step": 365
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.000294762879104208,
+      "loss": 0.7405,
+      "step": 366
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0002947335893798844,
+      "loss": 0.648,
+      "step": 367
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0002947042194426723,
+      "loss": 0.8346,
+      "step": 368
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0002946747693088487,
+      "loss": 0.6275,
+      "step": 369
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00029464523899473525,
+      "loss": 0.7413,
+      "step": 370
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00029461562851669797,
+      "loss": 0.8493,
+      "step": 371
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00029458593789114727,
+      "loss": 0.7284,
+      "step": 372
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.000294556167134538,
+      "loss": 0.5514,
+      "step": 373
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00029452631626336953,
+      "loss": 0.5366,
+      "step": 374
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0002944963852941854,
+      "loss": 0.5933,
+      "step": 375
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00029446637424357375,
+      "loss": 0.8107,
+      "step": 376
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00029443628312816693,
+      "loss": 0.7095,
+      "step": 377
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00029440611196464186,
+      "loss": 0.5933,
+      "step": 378
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00029437586076971965,
+      "loss": 0.7379,
+      "step": 379
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0002943455295601659,
+      "loss": 0.7432,
+      "step": 380
+    },
+    {
+      "epoch": 0.27,
+      "eval_loss": 1.1209006309509277,
+      "eval_runtime": 47.5543,
+      "eval_samples_per_second": 12.26,
+      "eval_steps_per_second": 12.26,
+      "step": 380
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0002943151183527904,
+      "loss": 0.7478,
+      "step": 381
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00029428462716444745,
+      "loss": 0.5829,
+      "step": 382
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00029425405601203554,
+      "loss": 0.6084,
+      "step": 383
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00029422340491249755,
+      "loss": 0.7695,
+      "step": 384
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00029419267388282056,
+      "loss": 0.7295,
+      "step": 385
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0002941618629400362,
+      "loss": 0.7138,
+      "step": 386
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0002941309721012201,
+      "loss": 0.6638,
+      "step": 387
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00029410000138349227,
+      "loss": 0.7753,
+      "step": 388
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0002940689508040171,
+      "loss": 0.6573,
+      "step": 389
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.000294037820380003,
+      "loss": 0.7546,
+      "step": 390
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00029400661012870296,
+      "loss": 0.7861,
+      "step": 391
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00029397532006741396,
+      "loss": 0.827,
+      "step": 392
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0002939439502134773,
+      "loss": 0.6356,
+      "step": 393
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0002939125005842784,
+      "loss": 0.7878,
+      "step": 394
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0002938809711972471,
+      "loss": 0.5794,
+      "step": 395
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0002938493620698572,
+      "loss": 0.7031,
+      "step": 396
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00029381767321962695,
+      "loss": 0.6507,
+      "step": 397
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00029378590466411854,
+      "loss": 0.6403,
+      "step": 398
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00029375405642093847,
+      "loss": 0.6967,
+      "step": 399
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0002937221285077374,
+      "loss": 0.7462,
+      "step": 400
+    },
+    {
+      "epoch": 0.28,
+      "eval_loss": 1.1184884309768677,
+      "eval_runtime": 47.3404,
+      "eval_samples_per_second": 12.315,
+      "eval_steps_per_second": 12.315,
+      "step": 400
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00029369012094221007,
+      "loss": 0.7789,
+      "step": 401
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00029365803374209546,
+      "loss": 0.7364,
+      "step": 402
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00029362586692517655,
+      "loss": 0.7206,
+      "step": 403
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00029359362050928066,
+      "loss": 0.6432,
+      "step": 404
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.000293561294512279,
+      "loss": 0.7095,
+      "step": 405
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.000293528888952087,
+      "loss": 0.5924,
+      "step": 406
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0002934964038466642,
+      "loss": 0.6324,
+      "step": 407
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00029346383921401414,
+      "loss": 0.8299,
+      "step": 408
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00029343119507218446,
+      "loss": 0.6723,
+      "step": 409
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.000293398471439267,
+      "loss": 0.8623,
+      "step": 410
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0002933656683333975,
+      "loss": 0.5535,
+      "step": 411
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0002933327857727557,
+      "loss": 0.7685,
+      "step": 412
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0002932998237755656,
+      "loss": 0.6614,
+      "step": 413
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.000293266782360095,
+      "loss": 0.8172,
+      "step": 414
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0002932336615446558,
+      "loss": 0.6812,
+      "step": 415
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0002932004613476039,
+      "loss": 0.6152,
+      "step": 416
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0002931671817873393,
+      "loss": 0.6711,
+      "step": 417
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0002931338228823058,
+      "loss": 0.7777,
+      "step": 418
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00029310038465099126,
+      "loss": 0.6359,
+      "step": 419
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0002930668671119275,
+      "loss": 0.8705,
+      "step": 420
+    },
+    {
+      "epoch": 0.3,
+      "eval_loss": 1.1302189826965332,
+      "eval_runtime": 47.1611,
+      "eval_samples_per_second": 12.362,
+      "eval_steps_per_second": 12.362,
+      "step": 420
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00029303327028369034,
+      "loss": 0.7305,
+      "step": 421
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0002929995941848995,
+      "loss": 0.8036,
+      "step": 422
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0002929658388342185,
+      "loss": 0.7765,
+      "step": 423
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0002929320042503551,
+      "loss": 0.857,
+      "step": 424
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00029289809045206065,
+      "loss": 0.9673,
+      "step": 425
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00029286409745813064,
+      "loss": 0.7616,
+      "step": 426
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0002928300252874042,
+      "loss": 0.5909,
+      "step": 427
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00029279587395876467,
+      "loss": 0.6926,
+      "step": 428
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00029276164349113895,
+      "loss": 0.7008,
+      "step": 429
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00029272733390349795,
+      "loss": 0.6738,
+      "step": 430
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00029269294521485647,
+      "loss": 0.6401,
+      "step": 431
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00029265847744427303,
+      "loss": 0.7389,
+      "step": 432
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00029262393061085,
+      "loss": 0.66,
+      "step": 433
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0002925893047337337,
+      "loss": 0.7043,
+      "step": 434
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00029255459983211404,
+      "loss": 0.766,
+      "step": 435
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00029251981592522494,
+      "loss": 0.6643,
+      "step": 436
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00029248495303234393,
+      "loss": 0.6664,
+      "step": 437
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0002924500111727925,
+      "loss": 0.7073,
+      "step": 438
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0002924149903659357,
+      "loss": 0.6508,
+      "step": 439
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00029237989063118247,
+      "loss": 0.7279,
+      "step": 440
+    },
+    {
+      "epoch": 0.31,
+      "eval_loss": 1.1231439113616943,
+      "eval_runtime": 46.6881,
+      "eval_samples_per_second": 12.487,
+      "eval_steps_per_second": 12.487,
+      "step": 440
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0002923447119879855,
+      "loss": 0.6644,
+      "step": 441
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0002923094544558411,
+      "loss": 0.8197,
+      "step": 442
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00029227411805428935,
+      "loss": 0.6729,
+      "step": 443
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0002922387028029142,
+      "loss": 0.709,
+      "step": 444
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00029220320872134294,
+      "loss": 0.7294,
+      "step": 445
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0002921676358292469,
+      "loss": 0.6705,
+      "step": 446
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00029213198414634097,
+      "loss": 0.7851,
+      "step": 447
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00029209625369238366,
+      "loss": 0.8442,
+      "step": 448
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0002920604444871771,
+      "loss": 0.8135,
+      "step": 449
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0002920245565505673,
+      "loss": 0.7192,
+      "step": 450
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00029198858990244354,
+      "loss": 0.697,
+      "step": 451
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00029195254456273903,
+      "loss": 0.7238,
+      "step": 452
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00029191642055143046,
+      "loss": 0.7011,
+      "step": 453
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0002918802178885381,
+      "loss": 0.5832,
+      "step": 454
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00029184393659412595,
+      "loss": 0.7514,
+      "step": 455
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0002918075766883014,
+      "loss": 0.5981,
+      "step": 456
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00029177113819121544,
+      "loss": 0.6991,
+      "step": 457
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00029173462112306285,
+      "loss": 0.6004,
+      "step": 458
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0002916980255040816,
+      "loss": 0.8339,
+      "step": 459
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00029166135135455344,
+      "loss": 0.7801,
+      "step": 460
+    },
+    {
+      "epoch": 0.33,
+      "eval_loss": 1.1127840280532837,
+      "eval_runtime": 46.8611,
+      "eval_samples_per_second": 12.441,
+      "eval_steps_per_second": 12.441,
+      "step": 460
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00029162459869480357,
+      "loss": 0.8793,
+      "step": 461
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0002915877675452008,
+      "loss": 0.6741,
+      "step": 462
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00029155085792615716,
+      "loss": 0.7363,
+      "step": 463
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0002915138698581285,
+      "loss": 0.6421,
+      "step": 464
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0002914768033616139,
+      "loss": 0.6674,
+      "step": 465
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00029143965845715604,
+      "loss": 0.6903,
+      "step": 466
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00029140243516534106,
+      "loss": 0.6025,
+      "step": 467
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0002913651335067984,
+      "loss": 0.6277,
+      "step": 468
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0002913277535022012,
+      "loss": 0.6974,
+      "step": 469
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00029129029517226574,
+      "loss": 0.7364,
+      "step": 470
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00029125275853775185,
+      "loss": 0.686,
+      "step": 471
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00029121514361946267,
+      "loss": 0.6251,
+      "step": 472
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00029117745043824486,
+      "loss": 0.5641,
+      "step": 473
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0002911396790149884,
+      "loss": 0.6443,
+      "step": 474
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00029110182937062653,
+      "loss": 0.6824,
+      "step": 475
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0002910639015261359,
+      "loss": 0.6971,
+      "step": 476
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00029102589550253675,
+      "loss": 0.8168,
+      "step": 477
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0002909878113208921,
+      "loss": 0.6618,
+      "step": 478
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00029094964900230885,
+      "loss": 0.5732,
+      "step": 479
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00029091140856793684,
+      "loss": 0.8211,
+      "step": 480
+    },
+    {
+      "epoch": 0.34,
+      "eval_loss": 1.1107630729675293,
+      "eval_runtime": 47.8856,
+      "eval_samples_per_second": 12.175,
+      "eval_steps_per_second": 12.175,
+      "step": 480
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0002908730900389694,
+      "loss": 0.7548,
+      "step": 481
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00029083469343664303,
+      "loss": 0.7715,
+      "step": 482
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0002907962187822375,
+      "loss": 0.6367,
+      "step": 483
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00029075766609707593,
+      "loss": 0.6814,
+      "step": 484
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00029071903540252455,
+      "loss": 0.5869,
+      "step": 485
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.000290680326719993,
+      "loss": 0.7688,
+      "step": 486
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00029064154007093395,
+      "loss": 0.8299,
+      "step": 487
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00029060267547684345,
+      "loss": 0.7443,
+      "step": 488
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0002905637329592606,
+      "loss": 0.8312,
+      "step": 489
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0002905247125397678,
+      "loss": 0.6397,
+      "step": 490
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00029048561423999057,
+      "loss": 0.6695,
+      "step": 491
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0002904464380815976,
+      "loss": 0.686,
+      "step": 492
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00029040718408630074,
+      "loss": 0.7094,
+      "step": 493
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0002903678522758549,
+      "loss": 0.7117,
+      "step": 494
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0002903284426720583,
+      "loss": 0.686,
+      "step": 495
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0002902889552967521,
+      "loss": 0.6969,
+      "step": 496
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00029024939017182065,
+      "loss": 0.6636,
+      "step": 497
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0002902097473191913,
+      "loss": 0.8073,
+      "step": 498
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0002901700267608346,
+      "loss": 0.7871,
+      "step": 499
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0002901302285187641,
+      "loss": 0.8281,
+      "step": 500
+    },
+    {
+      "epoch": 0.35,
+      "eval_loss": 1.12065589427948,
+      "eval_runtime": 47.8917,
+      "eval_samples_per_second": 12.173,
+      "eval_steps_per_second": 12.173,
+      "step": 500
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0002900903526150364,
+      "loss": 0.6801,
+      "step": 501
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0002900503990717511,
+      "loss": 0.6792,
+      "step": 502
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00029001036791105096,
+      "loss": 0.7119,
+      "step": 503
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0002899702591551216,
+      "loss": 0.6401,
+      "step": 504
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0002899300728261918,
+      "loss": 0.6256,
+      "step": 505
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0002898898089465331,
+      "loss": 0.6614,
+      "step": 506
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0002898494675384603,
+      "loss": 0.8029,
+      "step": 507
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00028980904862433105,
+      "loss": 0.8261,
+      "step": 508
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00028976855222654583,
+      "loss": 0.7787,
+      "step": 509
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00028972797836754826,
+      "loss": 0.7987,
+      "step": 510
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0002896873270698248,
+      "loss": 0.7587,
+      "step": 511
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00028964659835590477,
+      "loss": 0.7955,
+      "step": 512
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00028960579224836046,
+      "loss": 0.7699,
+      "step": 513
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00028956490876980715,
+      "loss": 0.7786,
+      "step": 514
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0002895239479429028,
+      "loss": 0.7223,
+      "step": 515
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00028948290979034834,
+      "loss": 0.7023,
+      "step": 516
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00028944179433488756,
+      "loss": 0.6807,
+      "step": 517
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0002894006015993071,
+      "loss": 0.6974,
+      "step": 518
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0002893593316064364,
+      "loss": 0.7256,
+      "step": 519
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0002893179843791478,
+      "loss": 0.83,
+      "step": 520
+    },
+    {
+      "epoch": 0.37,
+      "eval_loss": 1.1054948568344116,
+      "eval_runtime": 47.744,
+      "eval_samples_per_second": 12.211,
+      "eval_steps_per_second": 12.211,
+      "step": 520
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0002892765599403562,
+      "loss": 0.7897,
+      "step": 521
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0002892350583130196,
+      "loss": 0.66,
+      "step": 522
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0002891934795201386,
+      "loss": 0.6994,
+      "step": 523
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0002891518235847566,
+      "loss": 0.6678,
+      "step": 524
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00028911009052995975,
+      "loss": 0.7256,
+      "step": 525
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00028906828037887703,
+      "loss": 0.6447,
+      "step": 526
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00028902639315468,
+      "loss": 0.6021,
+      "step": 527
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.000288984428880583,
+      "loss": 0.7714,
+      "step": 528
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00028894238757984304,
+      "loss": 0.7238,
+      "step": 529
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00028890026927576,
+      "loss": 0.7712,
+      "step": 530
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0002888580739916761,
+      "loss": 0.6383,
+      "step": 531
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0002888158017509766,
+      "loss": 0.7202,
+      "step": 532
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.000288773452577089,
+      "loss": 0.6737,
+      "step": 533
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0002887310264934839,
+      "loss": 0.7196,
+      "step": 534
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00028868852352367417,
+      "loss": 0.7153,
+      "step": 535
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00028864594369121537,
+      "loss": 0.7359,
+      "step": 536
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00028860328701970576,
+      "loss": 0.6066,
+      "step": 537
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0002885605535327861,
+      "loss": 0.7727,
+      "step": 538
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00028851774325413973,
+      "loss": 0.8027,
+      "step": 539
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0002884748562074926,
+      "loss": 0.739,
+      "step": 540
+    },
+    {
+      "epoch": 0.38,
+      "eval_loss": 1.1061346530914307,
+      "eval_runtime": 47.7127,
+      "eval_samples_per_second": 12.219,
+      "eval_steps_per_second": 12.219,
+      "step": 540
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0002884318924166131,
+      "loss": 0.7713,
+      "step": 541
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0002883888519053123,
+      "loss": 0.7146,
+      "step": 542
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0002883457346974436,
+      "loss": 0.7098,
+      "step": 543
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0002883025408169031,
+      "loss": 0.7934,
+      "step": 544
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0002882592702876292,
+      "loss": 0.6203,
+      "step": 545
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00028821592313360306,
+      "loss": 0.6232,
+      "step": 546
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00028817249937884796,
+      "loss": 0.6947,
+      "step": 547
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00028812899904743,
+      "loss": 0.7616,
+      "step": 548
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0002880854221634573,
+      "loss": 0.707,
+      "step": 549
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00028804176875108075,
+      "loss": 0.7191,
+      "step": 550
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00028799803883449354,
+      "loss": 0.6821,
+      "step": 551
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00028795423243793125,
+      "loss": 0.7286,
+      "step": 552
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00028791034958567177,
+      "loss": 0.7009,
+      "step": 553
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0002878663903020356,
+      "loss": 0.5568,
+      "step": 554
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00028782235461138526,
+      "loss": 0.6504,
+      "step": 555
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0002877782425381259,
+      "loss": 0.8145,
+      "step": 556
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0002877340541067049,
+      "loss": 0.6267,
+      "step": 557
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0002876897893416119,
+      "loss": 0.826,
+      "step": 558
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.000287645448267379,
+      "loss": 0.6047,
+      "step": 559
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0002876010309085804,
+      "loss": 0.6173,
+      "step": 560
+    },
+    {
+      "epoch": 0.4,
+      "eval_loss": 1.1094388961791992,
+      "eval_runtime": 47.8481,
+      "eval_samples_per_second": 12.184,
+      "eval_steps_per_second": 12.184,
+      "step": 560
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0002875565372898327,
+      "loss": 0.5894,
+      "step": 561
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0002875119674357947,
+      "loss": 0.7049,
+      "step": 562
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00028746732137116754,
+      "loss": 0.7313,
+      "step": 563
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0002874225991206946,
+      "loss": 0.6332,
+      "step": 564
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00028737780070916126,
+      "loss": 0.7909,
+      "step": 565
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00028733292616139536,
+      "loss": 0.8344,
+      "step": 566
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00028728797550226686,
+      "loss": 0.7055,
+      "step": 567
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00028724294875668796,
+      "loss": 0.5849,
+      "step": 568
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0002871978459496128,
+      "loss": 0.7472,
+      "step": 569
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00028715266710603794,
+      "loss": 0.6677,
+      "step": 570
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00028710741225100193,
+      "loss": 0.6309,
+      "step": 571
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00028706208140958554,
+      "loss": 0.7547,
+      "step": 572
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0002870166746069116,
+      "loss": 0.6209,
+      "step": 573
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.000286971191868145,
+      "loss": 0.6612,
+      "step": 574
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0002869256332184927,
+      "loss": 0.6833,
+      "step": 575
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0002868799986832039,
+      "loss": 0.7647,
+      "step": 576
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00028683428828756973,
+      "loss": 0.7335,
+      "step": 577
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00028678850205692333,
+      "loss": 0.6485,
+      "step": 578
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00028674264001663993,
+      "loss": 0.5999,
+      "step": 579
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0002866967021921367,
+      "loss": 0.6947,
+      "step": 580
+    },
+    {
+      "epoch": 0.41,
+      "eval_loss": 1.114338755607605,
+      "eval_runtime": 47.8407,
+      "eval_samples_per_second": 12.186,
+      "eval_steps_per_second": 12.186,
+      "step": 580
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00028665068860887305,
+      "loss": 0.6997,
+      "step": 581
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00028660459929234994,
+      "loss": 0.6043,
+      "step": 582
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0002865584342681108,
+      "loss": 0.6274,
+      "step": 583
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0002865121935617406,
+      "loss": 0.6358,
+      "step": 584
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0002864658771988665,
+      "loss": 0.5953,
+      "step": 585
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0002864194852051575,
+      "loss": 0.7286,
+      "step": 586
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0002863730176063246,
+      "loss": 0.7562,
+      "step": 587
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00028632647442812054,
+      "loss": 0.6712,
+      "step": 588
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0002862798556963401,
+      "loss": 0.7069,
+      "step": 589
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00028623316143681985,
+      "loss": 0.7425,
+      "step": 590
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0002861863916754383,
+      "loss": 0.706,
+      "step": 591
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00028613954643811566,
+      "loss": 0.7399,
+      "step": 592
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00028609262575081406,
+      "loss": 0.8231,
+      "step": 593
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0002860456296395375,
+      "loss": 0.7751,
+      "step": 594
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0002859985581303318,
+      "loss": 0.7628,
+      "step": 595
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0002859514112492843,
+      "loss": 0.7591,
+      "step": 596
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0002859041890225244,
+      "loss": 0.6452,
+      "step": 597
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00028585689147622316,
+      "loss": 0.7025,
+      "step": 598
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00028580951863659335,
+      "loss": 0.765,
+      "step": 599
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00028576207052988955,
+      "loss": 0.7193,
+      "step": 600
+    },
+    {
+      "epoch": 0.43,
+      "eval_loss": 1.10276460647583,
+      "eval_runtime": 47.5133,
+      "eval_samples_per_second": 12.27,
+      "eval_steps_per_second": 12.27,
+      "step": 600
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.000285714547182408,
+      "loss": 0.6013,
+      "step": 601
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00028566694862048654,
+      "loss": 0.5383,
+      "step": 602
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00028561927487050497,
+      "loss": 0.8297,
+      "step": 603
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00028557152595888444,
+      "loss": 0.7171,
+      "step": 604
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00028552370191208795,
+      "loss": 0.8499,
+      "step": 605
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0002854758027566201,
+      "loss": 0.677,
+      "step": 606
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00028542782851902714,
+      "loss": 0.7218,
+      "step": 607
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00028537977922589685,
+      "loss": 0.7971,
+      "step": 608
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0002853316549038587,
+      "loss": 0.6954,
+      "step": 609
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0002852834555795836,
+      "loss": 0.786,
+      "step": 610
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00028523518127978426,
+      "loss": 0.7941,
+      "step": 611
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00028518683203121475,
+      "loss": 0.7294,
+      "step": 612
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00028513840786067074,
+      "loss": 0.6224,
+      "step": 613
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0002850899087949894,
+      "loss": 0.6609,
+      "step": 614
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00028504133486104946,
+      "loss": 0.6309,
+      "step": 615
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.000284992686085771,
+      "loss": 0.7685,
+      "step": 616
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0002849439624961159,
+      "loss": 0.7025,
+      "step": 617
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00028489516411908715,
+      "loss": 0.7522,
+      "step": 618
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0002848462909817293,
+      "loss": 0.716,
+      "step": 619
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00028479734311112837,
+      "loss": 0.7011,
+      "step": 620
+    },
+    {
+      "epoch": 0.44,
+      "eval_loss": 1.115025281906128,
+      "eval_runtime": 50.6455,
+      "eval_samples_per_second": 11.511,
+      "eval_steps_per_second": 11.511,
+      "step": 620
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00028474832053441187,
+      "loss": 0.6432,
+      "step": 621
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00028469922327874854,
+      "loss": 0.7601,
+      "step": 622
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0002846500513713487,
+      "loss": 0.7209,
+      "step": 623
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0002846008048394638,
+      "loss": 0.6631,
+      "step": 624
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.000284551483710387,
+      "loss": 0.691,
+      "step": 625
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00028450208801145235,
+      "loss": 0.7647,
+      "step": 626
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00028445261777003563,
+      "loss": 0.623,
+      "step": 627
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0002844030730135538,
+      "loss": 0.72,
+      "step": 628
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00028435345376946487,
+      "loss": 0.7505,
+      "step": 629
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0002843037600652686,
+      "loss": 0.7589,
+      "step": 630
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00028425399192850563,
+      "loss": 0.7482,
+      "step": 631
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00028420414938675804,
+      "loss": 0.6772,
+      "step": 632
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0002841542324676491,
+      "loss": 0.7684,
+      "step": 633
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0002841042411988432,
+      "loss": 0.6256,
+      "step": 634
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0002840541756080461,
+      "loss": 0.6878,
+      "step": 635
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00028400403572300477,
+      "loss": 0.7954,
+      "step": 636
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00028395382157150715,
+      "loss": 0.7832,
+      "step": 637
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00028390353318138246,
+      "loss": 0.6216,
+      "step": 638
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0002838531705805011,
+      "loss": 0.7863,
+      "step": 639
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0002838027337967746,
+      "loss": 0.6402,
+      "step": 640
+    },
+    {
+      "epoch": 0.45,
+      "eval_loss": 1.1150087118148804,
+      "eval_runtime": 49.696,
+      "eval_samples_per_second": 11.731,
+      "eval_steps_per_second": 11.731,
+      "step": 640
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0002837522228581554,
+      "loss": 0.7054,
+      "step": 641
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00028370163779263736,
+      "loss": 0.7034,
+      "step": 642
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00028365097862825513,
+      "loss": 0.6374,
+      "step": 643
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0002836002453930847,
+      "loss": 0.7538,
+      "step": 644
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0002835494381152429,
+      "loss": 0.6173,
+      "step": 645
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0002834985568228876,
+      "loss": 0.7161,
+      "step": 646
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00028344760154421774,
+      "loss": 0.6281,
+      "step": 647
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00028339657230747343,
+      "loss": 0.6305,
+      "step": 648
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0002833454691409355,
+      "loss": 0.6052,
+      "step": 649
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00028329429207292586,
+      "loss": 0.7426,
+      "step": 650
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00028324304113180746,
+      "loss": 0.8593,
+      "step": 651
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.000283191716345984,
+      "loss": 0.7423,
+      "step": 652
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00028314031774390034,
+      "loss": 0.6713,
+      "step": 653
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0002830888453540421,
+      "loss": 0.7218,
+      "step": 654
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00028303729920493576,
+      "loss": 0.6097,
+      "step": 655
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00028298567932514873,
+      "loss": 0.7131,
+      "step": 656
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0002829339857432894,
+      "loss": 0.6935,
+      "step": 657
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0002828822184880069,
+      "loss": 0.6513,
+      "step": 658
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00028283037758799107,
+      "loss": 0.6987,
+      "step": 659
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00028277846307197277,
+      "loss": 0.7993,
+      "step": 660
+    },
+    {
+      "epoch": 0.47,
+      "eval_loss": 1.1041227579116821,
+      "eval_runtime": 47.6426,
+      "eval_samples_per_second": 12.237,
+      "eval_steps_per_second": 12.237,
+      "step": 660
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00028272647496872355,
+      "loss": 0.6675,
+      "step": 661
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00028267441330705575,
+      "loss": 0.6138,
+      "step": 662
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00028262227811582253,
+      "loss": 0.7104,
+      "step": 663
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0002825700694239178,
+      "loss": 0.6558,
+      "step": 664
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00028251778726027605,
+      "loss": 0.7418,
+      "step": 665
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0002824654316538727,
+      "loss": 0.6263,
+      "step": 666
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00028241300263372383,
+      "loss": 0.6406,
+      "step": 667
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00028236050022888604,
+      "loss": 0.789,
+      "step": 668
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00028230792446845684,
+      "loss": 0.8013,
+      "step": 669
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0002822552753815741,
+      "loss": 0.6006,
+      "step": 670
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0002822025529974167,
+      "loss": 0.5781,
+      "step": 671
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00028214975734520384,
+      "loss": 0.6542,
+      "step": 672
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0002820968884541954,
+      "loss": 0.8975,
+      "step": 673
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0002820439463536919,
+      "loss": 0.8279,
+      "step": 674
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0002819909310730345,
+      "loss": 0.7185,
+      "step": 675
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0002819378426416047,
+      "loss": 0.6912,
+      "step": 676
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0002818846810888248,
+      "loss": 0.6774,
+      "step": 677
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00028183144644415743,
+      "loss": 0.686,
+      "step": 678
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00028177813873710573,
+      "loss": 0.8467,
+      "step": 679
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0002817247579972135,
+      "loss": 0.8079,
+      "step": 680
+    },
+    {
+      "epoch": 0.48,
+      "eval_loss": 1.101340413093567,
+      "eval_runtime": 47.4427,
+      "eval_samples_per_second": 12.289,
+      "eval_steps_per_second": 12.289,
+      "step": 680
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00028167130425406486,
+      "loss": 0.7355,
+      "step": 681
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0002816177775372845,
+      "loss": 0.6765,
+      "step": 682
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0002815641778765374,
+      "loss": 0.5628,
+      "step": 683
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00028151050530152914,
+      "loss": 0.6542,
+      "step": 684
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0002814567598420056,
+      "loss": 0.7913,
+      "step": 685
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00028140294152775305,
+      "loss": 0.5453,
+      "step": 686
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0002813490503885982,
+      "loss": 0.7442,
+      "step": 687
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0002812950864544081,
+      "loss": 0.756,
+      "step": 688
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0002812410497550901,
+      "loss": 0.6421,
+      "step": 689
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00028118694032059195,
+      "loss": 0.6835,
+      "step": 690
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0002811327581809017,
+      "loss": 0.6907,
+      "step": 691
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0002810785033660476,
+      "loss": 0.7085,
+      "step": 692
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0002810241759060982,
+      "loss": 0.6816,
+      "step": 693
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0002809697758311625,
+      "loss": 0.7595,
+      "step": 694
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0002809153031713895,
+      "loss": 0.6809,
+      "step": 695
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00028086075795696853,
+      "loss": 0.7767,
+      "step": 696
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0002808061402181292,
+      "loss": 0.5776,
+      "step": 697
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0002807514499851411,
+      "loss": 0.7094,
+      "step": 698
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0002806966872883142,
+      "loss": 0.6096,
+      "step": 699
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00028064185215799866,
+      "loss": 0.6824,
+      "step": 700
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 1.100295901298523,
+      "eval_runtime": 47.5595,
+      "eval_samples_per_second": 12.258,
+      "eval_steps_per_second": 12.258,
+      "step": 700
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0002805869446245846,
+      "loss": 0.662,
+      "step": 701
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0002805319647185023,
+      "loss": 0.6709,
+      "step": 702
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0002804769124702223,
+      "loss": 0.7764,
+      "step": 703
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00028042178791025515,
+      "loss": 0.6812,
+      "step": 704
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0002803665910691514,
+      "loss": 0.6575,
+      "step": 705
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00028031132197750173,
+      "loss": 0.5974,
+      "step": 706
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00028025598066593694,
+      "loss": 0.7167,
+      "step": 707
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00028020056716512767,
+      "loss": 0.747,
+      "step": 708
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00028014508150578474,
+      "loss": 0.6804,
+      "step": 709
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00028008952371865885,
+      "loss": 0.79,
+      "step": 710
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00028003389383454073,
+      "loss": 0.6611,
+      "step": 711
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.000279978191884261,
+      "loss": 0.7175,
+      "step": 712
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0002799224178986903,
+      "loss": 0.6185,
+      "step": 713
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00027986657190873926,
+      "loss": 0.5929,
+      "step": 714
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0002798106539453582,
+      "loss": 0.6961,
+      "step": 715
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0002797546640395375,
+      "loss": 0.693,
+      "step": 716
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00027969860222230733,
+      "loss": 0.5824,
+      "step": 717
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00027964246852473775,
+      "loss": 0.6474,
+      "step": 718
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0002795862629779387,
+      "loss": 0.7317,
+      "step": 719
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00027952998561305985,
+      "loss": 0.7006,
+      "step": 720
+    },
+    {
+      "epoch": 0.51,
+      "eval_loss": 1.096726655960083,
+      "eval_runtime": 47.332,
+      "eval_samples_per_second": 12.317,
+      "eval_steps_per_second": 12.317,
+      "step": 720
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00027947363646129065,
+      "loss": 0.7285,
+      "step": 721
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0002794172155538605,
+      "loss": 0.6518,
+      "step": 722
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0002793607229220384,
+      "loss": 0.8634,
+      "step": 723
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0002793041585971332,
+      "loss": 0.7818,
+      "step": 724
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00027924752261049345,
+      "loss": 0.5937,
+      "step": 725
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00027919081499350743,
+      "loss": 0.7178,
+      "step": 726
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.000279134035777603,
+      "loss": 0.5108,
+      "step": 727
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0002790771849942479,
+      "loss": 0.6577,
+      "step": 728
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0002790202626749494,
+      "loss": 0.7627,
+      "step": 729
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0002789632688512545,
+      "loss": 0.7037,
+      "step": 730
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0002789062035547497,
+      "loss": 0.7505,
+      "step": 731
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00027884906681706123,
+      "loss": 0.7455,
+      "step": 732
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0002787918586698549,
+      "loss": 0.6208,
+      "step": 733
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0002787345791448361,
+      "loss": 0.6808,
+      "step": 734
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0002786772282737496,
+      "loss": 0.7474,
+      "step": 735
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00027861980608838004,
+      "loss": 0.6529,
+      "step": 736
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0002785623126205513,
+      "loss": 0.6979,
+      "step": 737
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0002785047479021269,
+      "loss": 0.627,
+      "step": 738
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00027844711196500977,
+      "loss": 0.5631,
+      "step": 739
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00027838940484114247,
+      "loss": 0.617,
+      "step": 740
+    },
+    {
+      "epoch": 0.52,
+      "eval_loss": 1.0972075462341309,
+      "eval_runtime": 47.4747,
+      "eval_samples_per_second": 12.28,
+      "eval_steps_per_second": 12.28,
+      "step": 740
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0002783316265625068,
+      "loss": 0.8491,
+      "step": 741
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00027827377716112413,
+      "loss": 0.6763,
+      "step": 742
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00027821585666905514,
+      "loss": 0.7408,
+      "step": 743
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0002781578651184001,
+      "loss": 0.7222,
+      "step": 744
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00027809980254129847,
+      "loss": 0.7733,
+      "step": 745
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0002780416689699292,
+      "loss": 0.7639,
+      "step": 746
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00027798346443651045,
+      "loss": 0.6721,
+      "step": 747
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00027792518897329984,
+      "loss": 0.666,
+      "step": 748
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0002778668426125942,
+      "loss": 0.8278,
+      "step": 749
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0002778084253867298,
+      "loss": 0.6156,
+      "step": 750
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.000277749937328082,
+      "loss": 0.7021,
+      "step": 751
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00027769137846906554,
+      "loss": 0.6826,
+      "step": 752
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00027763274884213434,
+      "loss": 0.5538,
+      "step": 753
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00027757404847978155,
+      "loss": 0.6514,
+      "step": 754
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00027751527741453957,
+      "loss": 0.6617,
+      "step": 755
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0002774564356789799,
+      "loss": 0.7249,
+      "step": 756
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00027739752330571335,
+      "loss": 0.6983,
+      "step": 757
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0002773385403273897,
+      "loss": 0.6446,
+      "step": 758
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0002772794867766979,
+      "loss": 0.6629,
+      "step": 759
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0002772203626863661,
+      "loss": 0.5866,
+      "step": 760
+    },
+    {
+      "epoch": 0.54,
+      "eval_loss": 1.1026842594146729,
+      "eval_runtime": 47.2306,
+      "eval_samples_per_second": 12.344,
+      "eval_steps_per_second": 12.344,
+      "step": 760
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00027716116808916157,
+      "loss": 0.7702,
+      "step": 761
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0002771019030178905,
+      "loss": 0.629,
+      "step": 762
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0002770425675053982,
+      "loss": 0.6873,
+      "step": 763
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00027698316158456904,
+      "loss": 0.6425,
+      "step": 764
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00027692368528832656,
+      "loss": 0.6908,
+      "step": 765
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.000276864138649633,
+      "loss": 0.6587,
+      "step": 766
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00027680452170148975,
+      "loss": 0.68,
+      "step": 767
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0002767448344769372,
+      "loss": 0.6749,
+      "step": 768
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00027668507700905466,
+      "loss": 0.7203,
+      "step": 769
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00027662524933096024,
+      "loss": 0.6384,
+      "step": 770
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00027656535147581123,
+      "loss": 0.8131,
+      "step": 771
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0002765053834768035,
+      "loss": 0.6777,
+      "step": 772
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0002764453453671721,
+      "loss": 0.5776,
+      "step": 773
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0002763852371801907,
+      "loss": 0.6377,
+      "step": 774
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0002763250589491719,
+      "loss": 0.5442,
+      "step": 775
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0002762648107074671,
+      "loss": 0.6873,
+      "step": 776
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00027620449248846657,
+      "loss": 0.7153,
+      "step": 777
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00027614410432559924,
+      "loss": 0.8343,
+      "step": 778
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0002760836462523329,
+      "loss": 0.7094,
+      "step": 779
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00027602311830217407,
+      "loss": 0.7698,
+      "step": 780
+    },
+    {
+      "epoch": 0.55,
+      "eval_loss": 1.097829818725586,
+      "eval_runtime": 50.6812,
+      "eval_samples_per_second": 11.503,
+      "eval_steps_per_second": 11.503,
+      "step": 780
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00027596252050866794,
+      "loss": 0.6536,
+      "step": 781
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0002759018529053985,
+      "loss": 0.7275,
+      "step": 782
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00027584111552598833,
+      "loss": 0.6781,
+      "step": 783
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00027578030840409875,
+      "loss": 0.5599,
+      "step": 784
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0002757194315734298,
+      "loss": 0.7198,
+      "step": 785
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0002756584850677199,
+      "loss": 0.7308,
+      "step": 786
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00027559746892074635,
+      "loss": 0.5907,
+      "step": 787
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0002755363831663249,
+      "loss": 0.7657,
+      "step": 788
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00027547522783831,
+      "loss": 0.7023,
+      "step": 789
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0002754140029705945,
+      "loss": 0.7979,
+      "step": 790
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00027535270859710984,
+      "loss": 0.5099,
+      "step": 791
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00027529134475182614,
+      "loss": 0.6479,
+      "step": 792
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00027522991146875176,
+      "loss": 0.7195,
+      "step": 793
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00027516840878193376,
+      "loss": 0.7089,
+      "step": 794
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0002751068367254576,
+      "loss": 0.7375,
+      "step": 795
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00027504519533344703,
+      "loss": 0.5998,
+      "step": 796
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0002749834846400646,
+      "loss": 0.7867,
+      "step": 797
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0002749217046795108,
+      "loss": 0.6969,
+      "step": 798
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00027485985548602476,
+      "loss": 0.78,
+      "step": 799
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0002747979370938841,
+      "loss": 0.6299,
+      "step": 800
+    },
+    {
+      "epoch": 0.57,
+      "eval_loss": 1.0970231294631958,
+      "eval_runtime": 48.7327,
+      "eval_samples_per_second": 11.963,
+      "eval_steps_per_second": 11.963,
+      "step": 800
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00027473594953740456,
+      "loss": 0.8134,
+      "step": 801
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0002746738928509403,
+      "loss": 0.5842,
+      "step": 802
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00027461176706888385,
+      "loss": 0.6936,
+      "step": 803
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0002745495722256659,
+      "loss": 0.7654,
+      "step": 804
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0002744873083557555,
+      "loss": 0.7603,
+      "step": 805
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00027442497549366,
+      "loss": 0.6689,
+      "step": 806
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0002743625736739249,
+      "loss": 0.7336,
+      "step": 807
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00027430010293113394,
+      "loss": 0.6761,
+      "step": 808
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00027423756329990903,
+      "loss": 0.6938,
+      "step": 809
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0002741749548149104,
+      "loss": 0.591,
+      "step": 810
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00027411227751083627,
+      "loss": 0.7281,
+      "step": 811
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0002740495314224231,
+      "loss": 0.772,
+      "step": 812
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0002739867165844454,
+      "loss": 0.6564,
+      "step": 813
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00027392383303171585,
+      "loss": 0.6472,
+      "step": 814
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0002738608807990851,
+      "loss": 0.8273,
+      "step": 815
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0002737978599214421,
+      "loss": 0.6571,
+      "step": 816
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00027373477043371355,
+      "loss": 0.6657,
+      "step": 817
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0002736716123708643,
+      "loss": 0.7594,
+      "step": 818
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0002736083857678974,
+      "loss": 0.7199,
+      "step": 819
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00027354509065985347,
+      "loss": 0.7457,
+      "step": 820
+    },
+    {
+      "epoch": 0.58,
+      "eval_loss": 1.099078893661499,
+      "eval_runtime": 47.9418,
+      "eval_samples_per_second": 12.161,
+      "eval_steps_per_second": 12.161,
+      "step": 820
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0002734817270818115,
+      "loss": 0.6324,
+      "step": 821
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00027341829506888815,
+      "loss": 0.7262,
+      "step": 822
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0002733547946562382,
+      "loss": 0.6998,
+      "step": 823
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00027329122587905417,
+      "loss": 0.6699,
+      "step": 824
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0002732275887725666,
+      "loss": 0.6911,
+      "step": 825
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00027316388337204385,
+      "loss": 0.702,
+      "step": 826
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0002731001097127921,
+      "loss": 0.7905,
+      "step": 827
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00027303626783015545,
+      "loss": 0.6803,
+      "step": 828
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00027297235775951566,
+      "loss": 0.69,
+      "step": 829
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0002729083795362924,
+      "loss": 0.6352,
+      "step": 830
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0002728443331959431,
+      "loss": 0.74,
+      "step": 831
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00027278021877396295,
+      "loss": 0.6481,
+      "step": 832
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0002727160363058847,
+      "loss": 0.5609,
+      "step": 833
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0002726517858272791,
+      "loss": 0.6673,
+      "step": 834
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00027258746737375447,
+      "loss": 0.7734,
+      "step": 835
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00027252308098095663,
+      "loss": 0.6137,
+      "step": 836
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00027245862668456936,
+      "loss": 0.7611,
+      "step": 837
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00027239410452031376,
+      "loss": 0.728,
+      "step": 838
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0002723295145239488,
+      "loss": 0.5532,
+      "step": 839
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00027226485673127087,
+      "loss": 0.8171,
+      "step": 840
+    },
+    {
+      "epoch": 0.6,
+      "eval_loss": 1.1081715822219849,
+      "eval_runtime": 47.33,
+      "eval_samples_per_second": 12.318,
+      "eval_steps_per_second": 12.318,
+      "step": 840
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00027220013117811406,
+      "loss": 0.7426,
+      "step": 841
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00027213533790034996,
+      "loss": 0.7445,
+      "step": 842
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0002720704769338877,
+      "loss": 0.5527,
+      "step": 843
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00027200554831467385,
+      "loss": 0.7341,
+      "step": 844
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0002719405520786926,
+      "loss": 0.6549,
+      "step": 845
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0002718754882619655,
+      "loss": 0.7067,
+      "step": 846
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0002718103569005517,
+      "loss": 0.5433,
+      "step": 847
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0002717451580305475,
+      "loss": 0.7509,
+      "step": 848
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.000271679891688087,
+      "loss": 0.8112,
+      "step": 849
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00027161455790934144,
+      "loss": 0.6866,
+      "step": 850
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00027154915673051946,
+      "loss": 0.7054,
+      "step": 851
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00027148368818786714,
+      "loss": 0.7772,
+      "step": 852
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0002714181523176678,
+      "loss": 0.5794,
+      "step": 853
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0002713525491562421,
+      "loss": 0.5791,
+      "step": 854
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.000271286878739948,
+      "loss": 0.8285,
+      "step": 855
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00027122114110518085,
+      "loss": 0.8201,
+      "step": 856
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.000271155336288373,
+      "loss": 0.6588,
+      "step": 857
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0002710894643259943,
+      "loss": 0.6127,
+      "step": 858
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00027102352525455156,
+      "loss": 0.7655,
+      "step": 859
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.000270957519110589,
+      "loss": 0.765,
+      "step": 860
+    },
+    {
+      "epoch": 0.61,
+      "eval_loss": 1.099799633026123,
+      "eval_runtime": 47.4691,
+      "eval_samples_per_second": 12.282,
+      "eval_steps_per_second": 12.282,
+      "step": 860
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0002708914459306879,
+      "loss": 0.6746,
+      "step": 861
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0002708253057514667,
+      "loss": 0.5929,
+      "step": 862
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00027075909860958103,
+      "loss": 0.6837,
+      "step": 863
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0002706928245417236,
+      "loss": 0.6992,
+      "step": 864
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0002706264835846241,
+      "loss": 0.6892,
+      "step": 865
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0002705600757750495,
+      "loss": 0.7222,
+      "step": 866
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00027049360114980367,
+      "loss": 0.6983,
+      "step": 867
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0002704270597457276,
+      "loss": 0.6219,
+      "step": 868
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00027036045159969913,
+      "loss": 0.7201,
+      "step": 869
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0002702937767486333,
+      "loss": 0.5794,
+      "step": 870
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00027022703522948193,
+      "loss": 0.5765,
+      "step": 871
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00027016022707923397,
+      "loss": 0.6944,
+      "step": 872
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00027009335233491515,
+      "loss": 0.7201,
+      "step": 873
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00027002641103358817,
+      "loss": 0.6937,
+      "step": 874
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0002699594032123527,
+      "loss": 0.7625,
+      "step": 875
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.000269892328908345,
+      "loss": 0.6588,
+      "step": 876
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00026982518815873844,
+      "loss": 0.6357,
+      "step": 877
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00026975798100074323,
+      "loss": 0.7822,
+      "step": 878
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0002696907074716062,
+      "loss": 0.7428,
+      "step": 879
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.000269623367608611,
+      "loss": 0.5401,
+      "step": 880
+    },
+    {
+      "epoch": 0.62,
+      "eval_loss": 1.0921635627746582,
+      "eval_runtime": 47.2722,
+      "eval_samples_per_second": 12.333,
+      "eval_steps_per_second": 12.333,
+      "step": 880
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0002695559614490783,
+      "loss": 0.6914,
+      "step": 881
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.000269488489030365,
+      "loss": 0.7636,
+      "step": 882
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0002694209503898653,
+      "loss": 0.6765,
+      "step": 883
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0002693533455650097,
+      "loss": 0.7419,
+      "step": 884
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00026928567459326555,
+      "loss": 0.6286,
+      "step": 885
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00026921793751213675,
+      "loss": 0.6524,
+      "step": 886
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00026915013435916405,
+      "loss": 0.7304,
+      "step": 887
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00026908226517192454,
+      "loss": 0.7269,
+      "step": 888
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0002690143299880321,
+      "loss": 0.6344,
+      "step": 889
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00026894632884513715,
+      "loss": 0.6777,
+      "step": 890
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0002688782617809266,
+      "loss": 0.8047,
+      "step": 891
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00026881012883312393,
+      "loss": 0.7549,
+      "step": 892
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00026874193003948924,
+      "loss": 0.6478,
+      "step": 893
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00026867366543781887,
+      "loss": 0.5166,
+      "step": 894
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0002686053350659459,
+      "loss": 0.611,
+      "step": 895
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0002685369389617397,
+      "loss": 0.6614,
+      "step": 896
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0002684684771631061,
+      "loss": 0.7312,
+      "step": 897
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00026839994970798744,
+      "loss": 0.6198,
+      "step": 898
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00026833135663436225,
+      "loss": 0.8031,
+      "step": 899
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0002682626979802456,
+      "loss": 0.7338,
+      "step": 900
+    },
+    {
+      "epoch": 0.64,
+      "eval_loss": 1.088255524635315,
+      "eval_runtime": 47.2687,
+      "eval_samples_per_second": 12.334,
+      "eval_steps_per_second": 12.334,
+      "step": 900
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00026819397378368884,
+      "loss": 0.7108,
+      "step": 901
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0002681251840827797,
+      "loss": 0.7178,
+      "step": 902
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00026805632891564203,
+      "loss": 0.7131,
+      "step": 903
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0002679874083204362,
+      "loss": 0.5557,
+      "step": 904
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00026791842233535865,
+      "loss": 0.7235,
+      "step": 905
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00026784937099864225,
+      "loss": 0.6942,
+      "step": 906
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0002677802543485559,
+      "loss": 0.7861,
+      "step": 907
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0002677110724234048,
+      "loss": 0.6738,
+      "step": 908
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00026764182526153033,
+      "loss": 0.718,
+      "step": 909
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00026757251290131,
+      "loss": 0.6923,
+      "step": 910
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0002675031353811574,
+      "loss": 0.7414,
+      "step": 911
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00026743369273952235,
+      "loss": 0.6,
+      "step": 912
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0002673641850148906,
+      "loss": 0.6809,
+      "step": 913
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0002672946122457842,
+      "loss": 0.7223,
+      "step": 914
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0002672249744707611,
+      "loss": 0.7747,
+      "step": 915
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00026715527172841515,
+      "loss": 0.7385,
+      "step": 916
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0002670855040573765,
+      "loss": 0.6039,
+      "step": 917
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0002670156714963111,
+      "loss": 0.7158,
+      "step": 918
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0002669457740839208,
+      "loss": 0.6413,
+      "step": 919
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0002668758118589436,
+      "loss": 0.8229,
+      "step": 920
+    },
+    {
+      "epoch": 0.65,
+      "eval_loss": 1.0963973999023438,
+      "eval_runtime": 47.339,
+      "eval_samples_per_second": 12.315,
+      "eval_steps_per_second": 12.315,
+      "step": 920
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0002668057848601533,
+      "loss": 0.7845,
+      "step": 921
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0002667356931263595,
+      "loss": 0.713,
+      "step": 922
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0002666655366964079,
+      "loss": 0.629,
+      "step": 923
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0002665953156091799,
+      "loss": 0.7912,
+      "step": 924
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0002665250299035927,
+      "loss": 0.7193,
+      "step": 925
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00026645467961859947,
+      "loss": 0.7289,
+      "step": 926
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00026638426479318905,
+      "loss": 0.7735,
+      "step": 927
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0002663137854663861,
+      "loss": 0.6906,
+      "step": 928
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.000266243241677251,
+      "loss": 0.7589,
+      "step": 929
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00026617263346487985,
+      "loss": 0.7471,
+      "step": 930
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0002661019608684044,
+      "loss": 0.6947,
+      "step": 931
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00026603122392699236,
+      "loss": 0.6659,
+      "step": 932
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0002659604226798467,
+      "loss": 0.7943,
+      "step": 933
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00026588955716620635,
+      "loss": 0.6711,
+      "step": 934
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0002658186274253456,
+      "loss": 0.6181,
+      "step": 935
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0002657476334965745,
+      "loss": 0.7655,
+      "step": 936
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00026567657541923874,
+      "loss": 0.7305,
+      "step": 937
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0002656054532327194,
+      "loss": 0.694,
+      "step": 938
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0002655342669764331,
+      "loss": 0.7927,
+      "step": 939
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00026546301668983204,
+      "loss": 0.6952,
+      "step": 940
+    },
+    {
+      "epoch": 0.67,
+      "eval_loss": 1.0913337469100952,
+      "eval_runtime": 47.9753,
+      "eval_samples_per_second": 12.152,
+      "eval_steps_per_second": 12.152,
+      "step": 940
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00026539170241240387,
+      "loss": 0.6878,
+      "step": 941
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00026532032418367176,
+      "loss": 0.7091,
+      "step": 942
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0002652488820431943,
+      "loss": 0.777,
+      "step": 943
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0002651773760305654,
+      "loss": 0.5893,
+      "step": 944
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00026510580618541456,
+      "loss": 0.6003,
+      "step": 945
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0002650341725474064,
+      "loss": 0.6618,
+      "step": 946
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00026496247515624116,
+      "loss": 0.7763,
+      "step": 947
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0002648907140516543,
+      "loss": 0.6872,
+      "step": 948
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00026481888927341646,
+      "loss": 0.7189,
+      "step": 949
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0002647470008613338,
+      "loss": 0.6247,
+      "step": 950
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0002646750488552476,
+      "loss": 0.6981,
+      "step": 951
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0002646030332950345,
+      "loss": 0.8851,
+      "step": 952
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0002645309542206062,
+      "loss": 0.732,
+      "step": 953
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0002644588116719097,
+      "loss": 0.5475,
+      "step": 954
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00026438660568892715,
+      "loss": 0.691,
+      "step": 955
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0002643143363116759,
+      "loss": 0.6416,
+      "step": 956
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0002642420035802084,
+      "loss": 0.6631,
+      "step": 957
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00026416960753461227,
+      "loss": 0.6988,
+      "step": 958
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00026409714821501006,
+      "loss": 0.814,
+      "step": 959
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0002640246256615596,
+      "loss": 0.6717,
+      "step": 960
+    },
+    {
+      "epoch": 0.68,
+      "eval_loss": 1.0884716510772705,
+      "eval_runtime": 48.5515,
+      "eval_samples_per_second": 12.008,
+      "eval_steps_per_second": 12.008,
+      "step": 960
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0002639520399144535,
+      "loss": 0.6198,
+      "step": 961
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0002638793910139197,
+      "loss": 0.6536,
+      "step": 962
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.000263806679000221,
+      "loss": 0.7975,
+      "step": 963
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00026373390391365507,
+      "loss": 0.53,
+      "step": 964
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00026366106579455464,
+      "loss": 0.6796,
+      "step": 965
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00026358816468328744,
+      "loss": 0.7322,
+      "step": 966
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00026351520062025595,
+      "loss": 0.7159,
+      "step": 967
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00026344217364589775,
+      "loss": 0.5632,
+      "step": 968
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0002633690838006851,
+      "loss": 0.7722,
+      "step": 969
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00026329593112512506,
+      "loss": 0.8052,
+      "step": 970
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0002632227156597598,
+      "loss": 0.6485,
+      "step": 971
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00026314943744516593,
+      "loss": 0.7204,
+      "step": 972
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0002630760965219552,
+      "loss": 0.8512,
+      "step": 973
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00026300269293077377,
+      "loss": 0.7081,
+      "step": 974
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0002629292267123028,
+      "loss": 0.6648,
+      "step": 975
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0002628556979072579,
+      "loss": 0.6304,
+      "step": 976
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00026278210655638967,
+      "loss": 0.66,
+      "step": 977
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00026270845270048317,
+      "loss": 0.6961,
+      "step": 978
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0002626347363803581,
+      "loss": 0.7981,
+      "step": 979
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0002625609576368689,
+      "loss": 0.618,
+      "step": 980
+    },
+    {
+      "epoch": 0.69,
+      "eval_loss": 1.1006025075912476,
+      "eval_runtime": 47.6266,
+      "eval_samples_per_second": 12.241,
+      "eval_steps_per_second": 12.241,
+      "step": 980
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00026248711651090444,
+      "loss": 0.682,
+      "step": 981
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00026241321304338835,
+      "loss": 0.674,
+      "step": 982
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0002623392472752786,
+      "loss": 0.7887,
+      "step": 983
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0002622652192475679,
+      "loss": 0.4768,
+      "step": 984
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00026219112900128334,
+      "loss": 0.6787,
+      "step": 985
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00026211697657748653,
+      "loss": 0.6714,
+      "step": 986
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0002620427620172735,
+      "loss": 0.6751,
+      "step": 987
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0002619684853617747,
+      "loss": 0.6296,
+      "step": 988
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0002618941466521552,
+      "loss": 0.7707,
+      "step": 989
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00026181974592961413,
+      "loss": 0.571,
+      "step": 990
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00026174528323538525,
+      "loss": 0.6615,
+      "step": 991
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00026167075861073655,
+      "loss": 0.7241,
+      "step": 992
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0002615961720969704,
+      "loss": 0.7075,
+      "step": 993
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0002615215237354234,
+      "loss": 0.5905,
+      "step": 994
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0002614468135674664,
+      "loss": 0.6275,
+      "step": 995
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0002613720416345047,
+      "loss": 0.7542,
+      "step": 996
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0002612972079779776,
+      "loss": 0.6398,
+      "step": 997
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0002612223126393587,
+      "loss": 0.6024,
+      "step": 998
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00026114735566015593,
+      "loss": 0.8033,
+      "step": 999
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00026107233708191105,
+      "loss": 0.5702,
+      "step": 1000
+    },
+    {
+      "epoch": 0.71,
+      "eval_loss": 1.0917216539382935,
+      "eval_runtime": 48.249,
+      "eval_samples_per_second": 12.083,
+      "eval_steps_per_second": 12.083,
+      "step": 1000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00026099725694620033,
+      "loss": 0.6934,
+      "step": 1001
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0002609221152946338,
+      "loss": 0.773,
+      "step": 1002
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0002608469121688559,
+      "loss": 0.6758,
+      "step": 1003
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0002607716476105449,
+      "loss": 0.5948,
+      "step": 1004
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0002606963216614133,
+      "loss": 0.6637,
+      "step": 1005
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00026062093436320745,
+      "loss": 0.7599,
+      "step": 1006
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00026054548575770787,
+      "loss": 0.6258,
+      "step": 1007
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00026046997588672895,
+      "loss": 0.6877,
+      "step": 1008
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00026039440479211904,
+      "loss": 0.6083,
+      "step": 1009
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0002603187725157605,
+      "loss": 0.7348,
+      "step": 1010
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00026024307909956946,
+      "loss": 0.6314,
+      "step": 1011
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.000260167324585496,
+      "loss": 0.6719,
+      "step": 1012
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0002600915090155242,
+      "loss": 0.6215,
+      "step": 1013
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0002600156324316718,
+      "loss": 0.6451,
+      "step": 1014
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0002599396948759903,
+      "loss": 0.5435,
+      "step": 1015
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0002598636963905652,
+      "loss": 0.7379,
+      "step": 1016
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00025978763701751565,
+      "loss": 0.8079,
+      "step": 1017
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00025971151679899453,
+      "loss": 0.7313,
+      "step": 1018
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0002596353357771886,
+      "loss": 0.7494,
+      "step": 1019
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00025955909399431794,
+      "loss": 0.6621,
+      "step": 1020
+    },
+    {
+      "epoch": 0.72,
+      "eval_loss": 1.0829427242279053,
+      "eval_runtime": 47.4395,
+      "eval_samples_per_second": 12.289,
+      "eval_steps_per_second": 12.289,
+      "step": 1020
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00025948279149263685,
+      "loss": 0.7054,
+      "step": 1021
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0002594064283144327,
+      "loss": 0.7649,
+      "step": 1022
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00025933000450202695,
+      "loss": 0.7151,
+      "step": 1023
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00025925352009777445,
+      "loss": 0.6973,
+      "step": 1024
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0002591769751440637,
+      "loss": 0.5321,
+      "step": 1025
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0002591003696833167,
+      "loss": 0.7191,
+      "step": 1026
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00025902370375798894,
+      "loss": 0.7028,
+      "step": 1027
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00025894697741056955,
+      "loss": 0.5966,
+      "step": 1028
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0002588701906835811,
+      "loss": 0.7734,
+      "step": 1029
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0002587933436195795,
+      "loss": 0.6521,
+      "step": 1030
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00025871643626115435,
+      "loss": 0.6522,
+      "step": 1031
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0002586394686509285,
+      "loss": 0.688,
+      "step": 1032
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00025856244083155816,
+      "loss": 0.5956,
+      "step": 1033
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00025848535284573297,
+      "loss": 0.6714,
+      "step": 1034
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0002584082047361759,
+      "loss": 0.7817,
+      "step": 1035
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0002583309965456433,
+      "loss": 0.6407,
+      "step": 1036
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0002582537283169247,
+      "loss": 0.719,
+      "step": 1037
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.000258176400092843,
+      "loss": 0.615,
+      "step": 1038
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00025809901191625436,
+      "loss": 0.5839,
+      "step": 1039
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0002580215638300481,
+      "loss": 0.5358,
+      "step": 1040
+    },
+    {
+      "epoch": 0.74,
+      "eval_loss": 1.082664132118225,
+      "eval_runtime": 47.0936,
+      "eval_samples_per_second": 12.38,
+      "eval_steps_per_second": 12.38,
+      "step": 1040
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0002579440558771468,
+      "loss": 0.6522,
+      "step": 1041
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0002578664881005061,
+      "loss": 0.6806,
+      "step": 1042
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00025778886054311496,
+      "loss": 0.6965,
+      "step": 1043
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0002577111732479954,
+      "loss": 0.6804,
+      "step": 1044
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00025763342625820246,
+      "loss": 0.6722,
+      "step": 1045
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0002575556196168244,
+      "loss": 0.7251,
+      "step": 1046
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00025747775336698256,
+      "loss": 0.6472,
+      "step": 1047
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0002573998275518311,
+      "loss": 0.7569,
+      "step": 1048
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0002573218422145574,
+      "loss": 0.7194,
+      "step": 1049
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00025724379739838174,
+      "loss": 0.7243,
+      "step": 1050
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00025716569314655747,
+      "loss": 0.6505,
+      "step": 1051
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0002570875295023707,
+      "loss": 0.6655,
+      "step": 1052
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0002570093065091406,
+      "loss": 0.5075,
+      "step": 1053
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00025693102421021907,
+      "loss": 0.6741,
+      "step": 1054
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00025685268264899117,
+      "loss": 0.7683,
+      "step": 1055
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00025677428186887446,
+      "loss": 0.7619,
+      "step": 1056
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0002566958219133196,
+      "loss": 0.7262,
+      "step": 1057
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0002566173028258099,
+      "loss": 0.731,
+      "step": 1058
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0002565387246498614,
+      "loss": 0.5833,
+      "step": 1059
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.000256460087429023,
+      "loss": 0.6532,
+      "step": 1060
+    },
+    {
+      "epoch": 0.75,
+      "eval_loss": 1.097010612487793,
+      "eval_runtime": 47.1223,
+      "eval_samples_per_second": 12.372,
+      "eval_steps_per_second": 12.372,
+      "step": 1060
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0002563813912068763,
+      "loss": 0.5901,
+      "step": 1061
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00025630263602703564,
+      "loss": 0.5804,
+      "step": 1062
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0002562238219331478,
+      "loss": 0.6274,
+      "step": 1063
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0002561449489688926,
+      "loss": 0.7524,
+      "step": 1064
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00025606601717798207,
+      "loss": 0.7542,
+      "step": 1065
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0002559870266041612,
+      "loss": 0.6828,
+      "step": 1066
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00025590797729120737,
+      "loss": 0.6393,
+      "step": 1067
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00025582886928293046,
+      "loss": 0.6877,
+      "step": 1068
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.000255749702623173,
+      "loss": 0.7499,
+      "step": 1069
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0002556704773558101,
+      "loss": 0.8339,
+      "step": 1070
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.000255591193524749,
+      "loss": 0.6404,
+      "step": 1071
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00025551185117392994,
+      "loss": 0.7891,
+      "step": 1072
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0002554324503473251,
+      "loss": 0.7533,
+      "step": 1073
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00025535299108893936,
+      "loss": 0.5187,
+      "step": 1074
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00025527347344280973,
+      "loss": 0.5826,
+      "step": 1075
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0002551938974530059,
+      "loss": 0.6974,
+      "step": 1076
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0002551142631636297,
+      "loss": 0.7383,
+      "step": 1077
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0002550345706188152,
+      "loss": 0.6973,
+      "step": 1078
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0002549548198627289,
+      "loss": 0.7623,
+      "step": 1079
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0002548750109395695,
+      "loss": 0.7133,
+      "step": 1080
+    },
+    {
+      "epoch": 0.77,
+      "eval_loss": 1.0882887840270996,
+      "eval_runtime": 47.0431,
+      "eval_samples_per_second": 12.393,
+      "eval_steps_per_second": 12.393,
+      "step": 1080
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00025479514389356806,
+      "loss": 0.8006,
+      "step": 1081
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0002547152187689876,
+      "loss": 0.6179,
+      "step": 1082
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00025463523561012355,
+      "loss": 0.7324,
+      "step": 1083
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00025455519446130335,
+      "loss": 0.7161,
+      "step": 1084
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00025447509536688674,
+      "loss": 0.6814,
+      "step": 1085
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0002543949383712655,
+      "loss": 0.7798,
+      "step": 1086
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00025431472351886337,
+      "loss": 0.6783,
+      "step": 1087
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00025423445085413635,
+      "loss": 0.732,
+      "step": 1088
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0002541541204215724,
+      "loss": 0.682,
+      "step": 1089
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0002540737322656915,
+      "loss": 0.7508,
+      "step": 1090
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00025399328643104554,
+      "loss": 0.6973,
+      "step": 1091
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0002539127829622186,
+      "loss": 0.86,
+      "step": 1092
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0002538322219038264,
+      "loss": 0.7315,
+      "step": 1093
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00025375160330051683,
+      "loss": 0.5689,
+      "step": 1094
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00025367092719696954,
+      "loss": 0.5708,
+      "step": 1095
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0002535901936378961,
+      "loss": 0.6302,
+      "step": 1096
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00025350940266803993,
+      "loss": 0.6976,
+      "step": 1097
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0002534285543321762,
+      "loss": 0.6868,
+      "step": 1098
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00025334764867511185,
+      "loss": 0.6969,
+      "step": 1099
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00025326668574168575,
+      "loss": 0.7782,
+      "step": 1100
+    },
+    {
+      "epoch": 0.78,
+      "eval_loss": 1.0843466520309448,
+      "eval_runtime": 47.0916,
+      "eval_samples_per_second": 12.38,
+      "eval_steps_per_second": 12.38,
+      "step": 1100
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0002531856655767684,
+      "loss": 0.6405,
+      "step": 1101
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00025310458822526205,
+      "loss": 0.7092,
+      "step": 1102
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0002530234537321005,
+      "loss": 0.781,
+      "step": 1103
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00025294226214224956,
+      "loss": 0.6109,
+      "step": 1104
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0002528610135007063,
+      "loss": 0.627,
+      "step": 1105
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00025277970785249966,
+      "loss": 0.5592,
+      "step": 1106
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00025269834524269,
+      "loss": 0.6787,
+      "step": 1107
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00025261692571636944,
+      "loss": 0.595,
+      "step": 1108
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0002525354493186615,
+      "loss": 0.6536,
+      "step": 1109
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00025245391609472127,
+      "loss": 0.6404,
+      "step": 1110
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00025237232608973535,
+      "loss": 0.7763,
+      "step": 1111
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00025229067934892175,
+      "loss": 0.7087,
+      "step": 1112
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00025220897591752997,
+      "loss": 0.7357,
+      "step": 1113
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0002521272158408409,
+      "loss": 0.8036,
+      "step": 1114
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0002520453991641669,
+      "loss": 0.5787,
+      "step": 1115
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00025196352593285154,
+      "loss": 0.6569,
+      "step": 1116
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0002518815961922699,
+      "loss": 0.5399,
+      "step": 1117
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0002517996099878283,
+      "loss": 0.7083,
+      "step": 1118
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00025171756736496435,
+      "loss": 0.7285,
+      "step": 1119
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.000251635468369147,
+      "loss": 0.5843,
+      "step": 1120
+    },
+    {
+      "epoch": 0.79,
+      "eval_loss": 1.0963680744171143,
+      "eval_runtime": 46.9933,
+      "eval_samples_per_second": 12.406,
+      "eval_steps_per_second": 12.406,
+      "step": 1120
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0002515533130458763,
+      "loss": 0.7,
+      "step": 1121
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00025147110144068367,
+      "loss": 0.714,
+      "step": 1122
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0002513888335991316,
+      "loss": 0.617,
+      "step": 1123
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0002513065095668138,
+      "loss": 0.6737,
+      "step": 1124
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00025122412938935517,
+      "loss": 0.7403,
+      "step": 1125
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0002511416931124116,
+      "loss": 0.6654,
+      "step": 1126
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00025105920078167024,
+      "loss": 0.7668,
+      "step": 1127
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0002509766524428492,
+      "loss": 0.7569,
+      "step": 1128
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00025089404814169754,
+      "loss": 0.608,
+      "step": 1129
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00025081138792399553,
+      "loss": 0.5664,
+      "step": 1130
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00025072867183555434,
+      "loss": 0.8207,
+      "step": 1131
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0002506458999222161,
+      "loss": 0.6275,
+      "step": 1132
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0002505630722298538,
+      "loss": 0.6397,
+      "step": 1133
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0002504801888043715,
+      "loss": 0.6254,
+      "step": 1134
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0002503972496917041,
+      "loss": 0.6148,
+      "step": 1135
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0002503142549378173,
+      "loss": 0.7381,
+      "step": 1136
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0002502312045887076,
+      "loss": 0.7652,
+      "step": 1137
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0002501480986904025,
+      "loss": 0.6307,
+      "step": 1138
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0002500649372889601,
+      "loss": 0.7062,
+      "step": 1139
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0002499817204304694,
+      "loss": 0.6582,
+      "step": 1140
+    },
+    {
+      "epoch": 0.81,
+      "eval_loss": 1.0823915004730225,
+      "eval_runtime": 48.375,
+      "eval_samples_per_second": 12.052,
+      "eval_steps_per_second": 12.052,
+      "step": 1140
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0002498984481610499,
+      "loss": 0.6163,
+      "step": 1141
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0002498151205268522,
+      "loss": 0.7662,
+      "step": 1142
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00024973173757405717,
+      "loss": 0.6953,
+      "step": 1143
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0002496482993488767,
+      "loss": 0.7779,
+      "step": 1144
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.000249564805897553,
+      "loss": 0.6921,
+      "step": 1145
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00024948125726635895,
+      "loss": 0.6451,
+      "step": 1146
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0002493976535015983,
+      "loss": 0.7701,
+      "step": 1147
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00024931399464960506,
+      "loss": 0.5965,
+      "step": 1148
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00024923028075674383,
+      "loss": 0.7844,
+      "step": 1149
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0002491465118694097,
+      "loss": 0.648,
+      "step": 1150
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00024906268803402833,
+      "loss": 0.5915,
+      "step": 1151
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0002489788092970558,
+      "loss": 0.6798,
+      "step": 1152
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0002488948757049785,
+      "loss": 0.7609,
+      "step": 1153
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00024881088730431344,
+      "loss": 0.79,
+      "step": 1154
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0002487268441416078,
+      "loss": 0.5574,
+      "step": 1155
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00024864274626343916,
+      "loss": 0.4944,
+      "step": 1156
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0002485585937164155,
+      "loss": 0.6917,
+      "step": 1157
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0002484743865471752,
+      "loss": 0.6005,
+      "step": 1158
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00024839012480238644,
+      "loss": 0.6819,
+      "step": 1159
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0002483058085287483,
+      "loss": 0.7143,
+      "step": 1160
+    },
+    {
+      "epoch": 0.82,
+      "eval_loss": 1.0836167335510254,
+      "eval_runtime": 48.2706,
+      "eval_samples_per_second": 12.078,
+      "eval_steps_per_second": 12.078,
+      "step": 1160
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0002482214377729895,
+      "loss": 0.7396,
+      "step": 1161
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0002481370125818693,
+      "loss": 0.7836,
+      "step": 1162
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00024805253300217713,
+      "loss": 0.6541,
+      "step": 1163
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0002479679990807323,
+      "loss": 0.6477,
+      "step": 1164
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0002478834108643845,
+      "loss": 0.6913,
+      "step": 1165
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00024779876840001346,
+      "loss": 0.6786,
+      "step": 1166
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00024771407173452886,
+      "loss": 0.7336,
+      "step": 1167
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0002476293209148705,
+      "loss": 0.6774,
+      "step": 1168
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00024754451598800817,
+      "loss": 0.7068,
+      "step": 1169
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00024745965700094166,
+      "loss": 0.6817,
+      "step": 1170
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0002473747440007008,
+      "loss": 0.7759,
+      "step": 1171
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00024728977703434516,
+      "loss": 0.7574,
+      "step": 1172
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00024720475614896446,
+      "loss": 0.702,
+      "step": 1173
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.000247119681391678,
+      "loss": 0.6636,
+      "step": 1174
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00024703455280963523,
+      "loss": 0.7548,
+      "step": 1175
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00024694937045001533,
+      "loss": 0.6226,
+      "step": 1176
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0002468641343600272,
+      "loss": 0.8418,
+      "step": 1177
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0002467788445869096,
+      "loss": 0.7441,
+      "step": 1178
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0002466935011779311,
+      "loss": 0.7259,
+      "step": 1179
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0002466081041803898,
+      "loss": 0.6952,
+      "step": 1180
+    },
+    {
+      "epoch": 0.84,
+      "eval_loss": 1.0832979679107666,
+      "eval_runtime": 47.8604,
+      "eval_samples_per_second": 12.181,
+      "eval_steps_per_second": 12.181,
+      "step": 1180
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0002465226536416137,
+      "loss": 0.6184,
+      "step": 1181
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0002464371496089603,
+      "loss": 0.715,
+      "step": 1182
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00024635159212981703,
+      "loss": 0.6572,
+      "step": 1183
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0002462659812516006,
+      "loss": 0.6704,
+      "step": 1184
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0002461803170217575,
+      "loss": 0.7104,
+      "step": 1185
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00024609459948776377,
+      "loss": 0.6461,
+      "step": 1186
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0002460088286971249,
+      "loss": 0.6529,
+      "step": 1187
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0002459230046973761,
+      "loss": 0.6285,
+      "step": 1188
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00024583712753608185,
+      "loss": 0.6948,
+      "step": 1189
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0002457511972608362,
+      "loss": 0.63,
+      "step": 1190
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0002456652139192626,
+      "loss": 0.6418,
+      "step": 1191
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.000245579177559014,
+      "loss": 0.557,
+      "step": 1192
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0002454930882277726,
+      "loss": 0.7011,
+      "step": 1193
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00024540694597325,
+      "loss": 0.6623,
+      "step": 1194
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00024532075084318726,
+      "loss": 0.6421,
+      "step": 1195
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00024523450288535444,
+      "loss": 0.7592,
+      "step": 1196
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0002451482021475513,
+      "loss": 0.7303,
+      "step": 1197
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0002450618486776064,
+      "loss": 0.7222,
+      "step": 1198
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00024497544252337795,
+      "loss": 0.6833,
+      "step": 1199
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00024488898373275304,
+      "loss": 0.7272,
+      "step": 1200
+    },
+    {
+      "epoch": 0.85,
+      "eval_loss": 1.0816867351531982,
+      "eval_runtime": 47.8443,
+      "eval_samples_per_second": 12.185,
+      "eval_steps_per_second": 12.185,
+      "step": 1200
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00024480247235364813,
+      "loss": 0.5681,
+      "step": 1201
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00024471590843400863,
+      "loss": 0.7089,
+      "step": 1202
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0002446292920218092,
+      "loss": 0.6905,
+      "step": 1203
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0002445426231650537,
+      "loss": 0.6904,
+      "step": 1204
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0002444559019117748,
+      "loss": 0.6911,
+      "step": 1205
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00024436912831003434,
+      "loss": 0.7497,
+      "step": 1206
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0002442823024079233,
+      "loss": 0.797,
+      "step": 1207
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0002441954242535614,
+      "loss": 0.6039,
+      "step": 1208
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0002441084938950974,
+      "loss": 0.688,
+      "step": 1209
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00024402151138070907,
+      "loss": 0.5582,
+      "step": 1210
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00024393447675860312,
+      "loss": 0.6796,
+      "step": 1211
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00024384739007701496,
+      "loss": 0.6645,
+      "step": 1212
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.000243760251384209,
+      "loss": 0.5877,
+      "step": 1213
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0002436730607284784,
+      "loss": 0.7274,
+      "step": 1214
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00024358581815814512,
+      "loss": 0.6614,
+      "step": 1215
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00024349852372155997,
+      "loss": 0.5894,
+      "step": 1216
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0002434111774671025,
+      "loss": 0.5881,
+      "step": 1217
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00024332377944318078,
+      "loss": 0.7265,
+      "step": 1218
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00024323632969823185,
+      "loss": 0.8431,
+      "step": 1219
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00024314882828072124,
+      "loss": 0.7475,
+      "step": 1220
+    },
+    {
+      "epoch": 0.86,
+      "eval_loss": 1.0790772438049316,
+      "eval_runtime": 48.0127,
+      "eval_samples_per_second": 12.143,
+      "eval_steps_per_second": 12.143,
+      "step": 1220
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00024306127523914315,
+      "loss": 0.6457,
+      "step": 1221
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0002429736706220204,
+      "loss": 0.6436,
+      "step": 1222
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00024288601447790446,
+      "loss": 0.6789,
+      "step": 1223
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0002427983068553752,
+      "loss": 0.7502,
+      "step": 1224
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00024271054780304117,
+      "loss": 0.7146,
+      "step": 1225
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00024262273736953935,
+      "loss": 0.6597,
+      "step": 1226
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0002425348756035352,
+      "loss": 0.7444,
+      "step": 1227
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0002424469625537227,
+      "loss": 0.7478,
+      "step": 1228
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00024235899826882413,
+      "loss": 0.7167,
+      "step": 1229
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0002422709827975903,
+      "loss": 0.7135,
+      "step": 1230
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0002421829161888002,
+      "loss": 0.7506,
+      "step": 1231
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00024209479849126136,
+      "loss": 0.5836,
+      "step": 1232
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00024200662975380958,
+      "loss": 0.6855,
+      "step": 1233
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0002419184100253088,
+      "loss": 0.6462,
+      "step": 1234
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00024183013935465141,
+      "loss": 0.7623,
+      "step": 1235
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00024174181779075792,
+      "loss": 0.7956,
+      "step": 1236
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0002416534453825771,
+      "loss": 0.6977,
+      "step": 1237
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00024156502217908583,
+      "loss": 0.6874,
+      "step": 1238
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00024147654822928924,
+      "loss": 0.6614,
+      "step": 1239
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00024138802358222047,
+      "loss": 0.6632,
+      "step": 1240
+    },
+    {
+      "epoch": 0.88,
+      "eval_loss": 1.0793814659118652,
+      "eval_runtime": 47.5897,
+      "eval_samples_per_second": 12.251,
+      "eval_steps_per_second": 12.251,
+      "step": 1240
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00024129944828694087,
+      "loss": 0.6338,
+      "step": 1241
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00024121082239253979,
+      "loss": 0.6874,
+      "step": 1242
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00024112214594813463,
+      "loss": 0.7036,
+      "step": 1243
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0002410334190028708,
+      "loss": 0.698,
+      "step": 1244
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00024094464160592178,
+      "loss": 0.6242,
+      "step": 1245
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00024085581380648886,
+      "loss": 0.725,
+      "step": 1246
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00024076693565380138,
+      "loss": 0.6312,
+      "step": 1247
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0002406780071971166,
+      "loss": 0.6634,
+      "step": 1248
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00024058902848571957,
+      "loss": 0.7754,
+      "step": 1249
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00024049999956892324,
+      "loss": 0.6308,
+      "step": 1250
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00024041092049606832,
+      "loss": 0.7041,
+      "step": 1251
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00024032179131652348,
+      "loss": 0.8392,
+      "step": 1252
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00024023261207968498,
+      "loss": 0.6493,
+      "step": 1253
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00024014338283497697,
+      "loss": 0.5882,
+      "step": 1254
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0002400541036318512,
+      "loss": 0.7007,
+      "step": 1255
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00023996477451978707,
+      "loss": 0.7451,
+      "step": 1256
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00023987539554829184,
+      "loss": 0.5702,
+      "step": 1257
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00023978596676690024,
+      "loss": 0.6727,
+      "step": 1258
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0002396964882251747,
+      "loss": 0.5924,
+      "step": 1259
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00023960695997270507,
+      "loss": 0.6565,
+      "step": 1260
+    },
+    {
+      "epoch": 0.89,
+      "eval_loss": 1.0851231813430786,
+      "eval_runtime": 47.0004,
+      "eval_samples_per_second": 12.404,
+      "eval_steps_per_second": 12.404,
+      "step": 1260
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00023951738205910895,
+      "loss": 0.781,
+      "step": 1261
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0002394277545340313,
+      "loss": 0.6338,
+      "step": 1262
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0002393380774471447,
+      "loss": 0.5754,
+      "step": 1263
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00023924835084814907,
+      "loss": 0.6175,
+      "step": 1264
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00023915857478677192,
+      "loss": 0.5678,
+      "step": 1265
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00023906874931276805,
+      "loss": 0.638,
+      "step": 1266
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0002389788744759197,
+      "loss": 0.511,
+      "step": 1267
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00023888895032603646,
+      "loss": 0.7268,
+      "step": 1268
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00023879897691295523,
+      "loss": 0.5996,
+      "step": 1269
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0002387089542865403,
+      "loss": 0.6624,
+      "step": 1270
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00023861888249668305,
+      "loss": 0.7584,
+      "step": 1271
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0002385287615933023,
+      "loss": 0.6154,
+      "step": 1272
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.000238438591626344,
+      "loss": 0.6076,
+      "step": 1273
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00023834837264578127,
+      "loss": 0.6373,
+      "step": 1274
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0002382581047016144,
+      "loss": 0.7189,
+      "step": 1275
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00023816778784387094,
+      "loss": 0.6716,
+      "step": 1276
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00023807742212260537,
+      "loss": 0.6188,
+      "step": 1277
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00023798700758789938,
+      "loss": 0.7345,
+      "step": 1278
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00023789654428986158,
+      "loss": 0.8079,
+      "step": 1279
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00023780603227862782,
+      "loss": 0.682,
+      "step": 1280
+    },
+    {
+      "epoch": 0.91,
+      "eval_loss": 1.0831193923950195,
+      "eval_runtime": 47.2069,
+      "eval_samples_per_second": 12.35,
+      "eval_steps_per_second": 12.35,
+      "step": 1280
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0002377154716043607,
+      "loss": 0.6521,
+      "step": 1281
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00023762486231724991,
+      "loss": 0.6951,
+      "step": 1282
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00023753420446751214,
+      "loss": 0.6459,
+      "step": 1283
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0002374434981053909,
+      "loss": 0.7327,
+      "step": 1284
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00023735274328115665,
+      "loss": 0.6521,
+      "step": 1285
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0002372619400451066,
+      "loss": 0.708,
+      "step": 1286
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00023717108844756492,
+      "loss": 0.767,
+      "step": 1287
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00023708018853888248,
+      "loss": 0.7303,
+      "step": 1288
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00023698924036943706,
+      "loss": 0.6588,
+      "step": 1289
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00023689824398963306,
+      "loss": 0.6672,
+      "step": 1290
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0002368071994499016,
+      "loss": 0.5117,
+      "step": 1291
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00023671610680070057,
+      "loss": 0.6913,
+      "step": 1292
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00023662496609251445,
+      "loss": 0.7174,
+      "step": 1293
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00023653377737585442,
+      "loss": 0.5579,
+      "step": 1294
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00023644254070125816,
+      "loss": 0.6692,
+      "step": 1295
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00023635125611929012,
+      "loss": 0.7178,
+      "step": 1296
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00023625992368054105,
+      "loss": 0.7958,
+      "step": 1297
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0002361685434356285,
+      "loss": 0.5927,
+      "step": 1298
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00023607711543519627,
+      "loss": 0.6268,
+      "step": 1299
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00023598563972991474,
+      "loss": 0.7497,
+      "step": 1300
+    },
+    {
+      "epoch": 0.92,
+      "eval_loss": 1.0784038305282593,
+      "eval_runtime": 47.9899,
+      "eval_samples_per_second": 12.148,
+      "eval_steps_per_second": 12.148,
+      "step": 1300
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00023589411637048072,
+      "loss": 0.707,
+      "step": 1301
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00023580254540761744,
+      "loss": 0.5201,
+      "step": 1302
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00023571092689207446,
+      "loss": 0.5883,
+      "step": 1303
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0002356192608746278,
+      "loss": 0.6254,
+      "step": 1304
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0002355275474060797,
+      "loss": 0.6785,
+      "step": 1305
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00023543578653725873,
+      "loss": 0.5245,
+      "step": 1306
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00023534397831901974,
+      "loss": 0.6191,
+      "step": 1307
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00023525212280224377,
+      "loss": 0.7193,
+      "step": 1308
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0002351602200378382,
+      "loss": 0.6032,
+      "step": 1309
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00023506827007673645,
+      "loss": 0.6393,
+      "step": 1310
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00023497627296989823,
+      "loss": 0.6531,
+      "step": 1311
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0002348842287683093,
+      "loss": 0.6103,
+      "step": 1312
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00023479213752298147,
+      "loss": 0.6232,
+      "step": 1313
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00023469999928495272,
+      "loss": 0.6856,
+      "step": 1314
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00023460781410528707,
+      "loss": 0.6774,
+      "step": 1315
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0002345155820350745,
+      "loss": 0.6527,
+      "step": 1316
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00023442330312543094,
+      "loss": 0.6073,
+      "step": 1317
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0002343309774274984,
+      "loss": 0.6953,
+      "step": 1318
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00023423860499244482,
+      "loss": 0.5413,
+      "step": 1319
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0002341461858714639,
+      "loss": 0.7174,
+      "step": 1320
+    },
+    {
+      "epoch": 0.94,
+      "eval_loss": 1.0791856050491333,
+      "eval_runtime": 49.4488,
+      "eval_samples_per_second": 11.79,
+      "eval_steps_per_second": 11.79,
+      "step": 1320
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00023405372011577524,
+      "loss": 0.6667,
+      "step": 1321
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00023396120777662447,
+      "loss": 0.6967,
+      "step": 1322
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00023386864890528284,
+      "loss": 0.5773,
+      "step": 1323
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00023377604355304745,
+      "loss": 0.6878,
+      "step": 1324
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0002336833917712412,
+      "loss": 0.6257,
+      "step": 1325
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0002335906936112127,
+      "loss": 0.6549,
+      "step": 1326
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0002334979491243362,
+      "loss": 0.6934,
+      "step": 1327
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00023340515836201166,
+      "loss": 0.6366,
+      "step": 1328
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00023331232137566483,
+      "loss": 0.7765,
+      "step": 1329
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00023321943821674683,
+      "loss": 0.825,
+      "step": 1330
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00023312650893673448,
+      "loss": 0.639,
+      "step": 1331
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00023303353358713027,
+      "loss": 0.6802,
+      "step": 1332
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00023294051221946208,
+      "loss": 0.7487,
+      "step": 1333
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00023284744488528332,
+      "loss": 0.6964,
+      "step": 1334
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00023275433163617295,
+      "loss": 0.7775,
+      "step": 1335
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00023266117252373518,
+      "loss": 0.6042,
+      "step": 1336
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00023256796759959993,
+      "loss": 0.6363,
+      "step": 1337
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0002324747169154223,
+      "loss": 0.6806,
+      "step": 1338
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00023238142052288283,
+      "loss": 0.7258,
+      "step": 1339
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0002322880784736873,
+      "loss": 0.7447,
+      "step": 1340
+    },
+    {
+      "epoch": 0.95,
+      "eval_loss": 1.0830106735229492,
+      "eval_runtime": 49.5568,
+      "eval_samples_per_second": 11.764,
+      "eval_steps_per_second": 11.764,
+      "step": 1340
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0002321946908195669,
+      "loss": 0.6474,
+      "step": 1341
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00023210125761227805,
+      "loss": 0.6062,
+      "step": 1342
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00023200777890360248,
+      "loss": 0.6763,
+      "step": 1343
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00023191425474534703,
+      "loss": 0.7494,
+      "step": 1344
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00023182068518934375,
+      "loss": 0.5687,
+      "step": 1345
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00023172707028744995,
+      "loss": 0.7464,
+      "step": 1346
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00023163341009154796,
+      "loss": 0.784,
+      "step": 1347
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00023153970465354524,
+      "loss": 0.6921,
+      "step": 1348
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0002314459540253744,
+      "loss": 0.7308,
+      "step": 1349
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00023135215825899295,
+      "loss": 0.7197,
+      "step": 1350
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0002312583174063835,
+      "loss": 0.6634,
+      "step": 1351
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00023116443151955374,
+      "loss": 0.7803,
+      "step": 1352
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00023107050065053613,
+      "loss": 0.6383,
+      "step": 1353
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00023097652485138824,
+      "loss": 0.6599,
+      "step": 1354
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00023088250417419235,
+      "loss": 0.6311,
+      "step": 1355
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00023078843867105575,
+      "loss": 0.6852,
+      "step": 1356
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00023069432839411055,
+      "loss": 0.7008,
+      "step": 1357
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0002306001733955137,
+      "loss": 0.6617,
+      "step": 1358
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0002305059737274469,
+      "loss": 0.7002,
+      "step": 1359
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0002304117294421165,
+      "loss": 0.6996,
+      "step": 1360
+    },
+    {
+      "epoch": 0.96,
+      "eval_loss": 1.0771316289901733,
+      "eval_runtime": 49.7649,
+      "eval_samples_per_second": 11.715,
+      "eval_steps_per_second": 11.715,
+      "step": 1360
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0002303174405917537,
+      "loss": 0.6914,
+      "step": 1361
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00023022310722861455,
+      "loss": 0.7965,
+      "step": 1362
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00023012872940497939,
+      "loss": 0.6242,
+      "step": 1363
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0002300343071731535,
+      "loss": 0.7304,
+      "step": 1364
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00022993984058546665,
+      "loss": 0.6814,
+      "step": 1365
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00022984532969427335,
+      "loss": 0.6685,
+      "step": 1366
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00022975077455195236,
+      "loss": 0.7181,
+      "step": 1367
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0002296561752109073,
+      "loss": 0.7427,
+      "step": 1368
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00022956153172356595,
+      "loss": 0.7214,
+      "step": 1369
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0002294668441423809,
+      "loss": 0.7024,
+      "step": 1370
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0002293721125198289,
+      "loss": 0.6408,
+      "step": 1371
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00022927733690841128,
+      "loss": 0.605,
+      "step": 1372
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0002291825173606537,
+      "loss": 0.5766,
+      "step": 1373
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00022908765392910604,
+      "loss": 0.6805,
+      "step": 1374
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0002289927466663428,
+      "loss": 0.8264,
+      "step": 1375
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00022889779562496238,
+      "loss": 0.5788,
+      "step": 1376
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00022880280085758786,
+      "loss": 0.8215,
+      "step": 1377
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0002287077624168662,
+      "loss": 0.7703,
+      "step": 1378
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00022861268035546882,
+      "loss": 0.5632,
+      "step": 1379
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0002285175547260911,
+      "loss": 0.763,
+      "step": 1380
+    },
+    {
+      "epoch": 0.98,
+      "eval_loss": 1.075742244720459,
+      "eval_runtime": 48.9691,
+      "eval_samples_per_second": 11.905,
+      "eval_steps_per_second": 11.905,
+      "step": 1380
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00022842238558145277,
+      "loss": 0.6453,
+      "step": 1381
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00022832717297429747,
+      "loss": 0.6527,
+      "step": 1382
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00022823191695739317,
+      "loss": 0.6642,
+      "step": 1383
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00022813661758353166,
+      "loss": 0.6017,
+      "step": 1384
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00022804127490552885,
+      "loss": 0.8221,
+      "step": 1385
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00022794588897622476,
+      "loss": 0.777,
+      "step": 1386
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00022785045984848317,
+      "loss": 0.6368,
+      "step": 1387
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00022775498757519202,
+      "loss": 0.6438,
+      "step": 1388
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00022765947220926297,
+      "loss": 0.6866,
+      "step": 1389
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00022756391380363174,
+      "loss": 0.6952,
+      "step": 1390
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00022746831241125768,
+      "loss": 0.6747,
+      "step": 1391
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00022737266808512424,
+      "loss": 0.7019,
+      "step": 1392
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0002272769808782384,
+      "loss": 0.7982,
+      "step": 1393
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00022718125084363115,
+      "loss": 0.689,
+      "step": 1394
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00022708547803435692,
+      "loss": 0.7294,
+      "step": 1395
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00022698966250349413,
+      "loss": 0.767,
+      "step": 1396
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00022689380430414475,
+      "loss": 0.7307,
+      "step": 1397
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00022679790348943435,
+      "loss": 0.6672,
+      "step": 1398
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0002267019601125123,
+      "loss": 0.7224,
+      "step": 1399
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00022660597422655132,
+      "loss": 0.7472,
+      "step": 1400
+    },
+    {
+      "epoch": 0.99,
+      "eval_loss": 1.077209711074829,
+      "eval_runtime": 48.4363,
+      "eval_samples_per_second": 12.036,
+      "eval_steps_per_second": 12.036,
+      "step": 1400
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00022650994588474787,
+      "loss": 0.7941,
+      "step": 1401
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00022641387514032185,
+      "loss": 0.7373,
+      "step": 1402
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0002263177620465167,
+      "loss": 0.7067,
+      "step": 1403
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00022622160665659925,
+      "loss": 0.641,
+      "step": 1404
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00022612540902385987,
+      "loss": 0.7532,
+      "step": 1405
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00022602916920161237,
+      "loss": 0.6334,
+      "step": 1406
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00022593288724319385,
+      "loss": 0.6611,
+      "step": 1407
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00022583656320196476,
+      "loss": 0.7375,
+      "step": 1408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.0002257401971313089,
+      "loss": 0.7868,
+      "step": 1409
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.0002256437890846334,
+      "loss": 0.6171,
+      "step": 1410
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00022554733911536859,
+      "loss": 0.5939,
+      "step": 1411
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00022545084727696813,
+      "loss": 0.6041,
+      "step": 1412
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.0002253543136229087,
+      "loss": 0.5165,
+      "step": 1413
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00022525773820669038,
+      "loss": 0.7498,
+      "step": 1414
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00022516112108183622,
+      "loss": 0.6802,
+      "step": 1415
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.0002250644623018925,
+      "loss": 0.5686,
+      "step": 1416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00022496776192042857,
+      "loss": 0.5046,
+      "step": 1417
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00022487101999103668,
+      "loss": 0.619,
+      "step": 1418
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00022477423656733241,
+      "loss": 0.5858,
+      "step": 1419
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00022467741170295399,
+      "loss": 0.5781,
+      "step": 1420
+    },
+    {
+      "epoch": 1.01,
+      "eval_loss": 1.0776679515838623,
+      "eval_runtime": 50.6361,
+      "eval_samples_per_second": 11.514,
+      "eval_steps_per_second": 11.514,
+      "step": 1420
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.0002245805454515629,
+      "loss": 0.5827,
+      "step": 1421
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.0002244836378668434,
+      "loss": 0.6471,
+      "step": 1422
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.0002243866890025027,
+      "loss": 0.6492,
+      "step": 1423
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00022428969891227085,
+      "loss": 0.5709,
+      "step": 1424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00022419266764990084,
+      "loss": 0.6936,
+      "step": 1425
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00022409559526916846,
+      "loss": 0.5598,
+      "step": 1426
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00022399848182387213,
+      "loss": 0.5272,
+      "step": 1427
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00022390132736783324,
+      "loss": 0.4035,
+      "step": 1428
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00022380413195489585,
+      "loss": 0.5484,
+      "step": 1429
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00022370689563892658,
+      "loss": 0.5948,
+      "step": 1430
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00022360961847381493,
+      "loss": 0.469,
+      "step": 1431
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00022351230051347287,
+      "loss": 0.4895,
+      "step": 1432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00022341494181183505,
+      "loss": 0.5628,
+      "step": 1433
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00022331754242285872,
+      "loss": 0.5247,
+      "step": 1434
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.0002232201024005236,
+      "loss": 0.5555,
+      "step": 1435
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00022312262179883206,
+      "loss": 0.6521,
+      "step": 1436
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00022302510067180882,
+      "loss": 0.6674,
+      "step": 1437
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00022292753907350119,
+      "loss": 0.7635,
+      "step": 1438
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00022282993705797876,
+      "loss": 0.5953,
+      "step": 1439
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00022273229467933368,
+      "loss": 0.5801,
+      "step": 1440
+    },
+    {
+      "epoch": 1.02,
+      "eval_loss": 1.0834102630615234,
+      "eval_runtime": 50.9811,
+      "eval_samples_per_second": 11.436,
+      "eval_steps_per_second": 11.436,
+      "step": 1440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00022263461199168032,
+      "loss": 0.5278,
+      "step": 1441
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00022253688904915551,
+      "loss": 0.6275,
+      "step": 1442
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00022243912590591832,
+      "loss": 0.5026,
+      "step": 1443
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00022234132261615008,
+      "loss": 0.4718,
+      "step": 1444
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00022224347923405446,
+      "loss": 0.4692,
+      "step": 1445
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00022214559581385735,
+      "loss": 0.5689,
+      "step": 1446
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00022204767240980666,
+      "loss": 0.5198,
+      "step": 1447
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00022194970907617263,
+      "loss": 0.6079,
+      "step": 1448
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00022185170586724761,
+      "loss": 0.6129,
+      "step": 1449
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00022175366283734598,
+      "loss": 0.5918,
+      "step": 1450
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00022165558004080422,
+      "loss": 0.48,
+      "step": 1451
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.0002215574575319809,
+      "loss": 0.535,
+      "step": 1452
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00022145929536525653,
+      "loss": 0.4836,
+      "step": 1453
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00022136109359503358,
+      "loss": 0.6255,
+      "step": 1454
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00022126285227573657,
+      "loss": 0.5704,
+      "step": 1455
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00022116457146181183,
+      "loss": 0.5458,
+      "step": 1456
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00022106625120772768,
+      "loss": 0.6331,
+      "step": 1457
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.0002209678915679742,
+      "loss": 0.6103,
+      "step": 1458
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.0002208694925970634,
+      "loss": 0.5651,
+      "step": 1459
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.000220771054349529,
+      "loss": 0.7059,
+      "step": 1460
+    },
+    {
+      "epoch": 1.04,
+      "eval_loss": 1.0936806201934814,
+      "eval_runtime": 51.1339,
+      "eval_samples_per_second": 11.401,
+      "eval_steps_per_second": 11.401,
+      "step": 1460
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.0002206725768799265,
+      "loss": 0.4931,
+      "step": 1461
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00022057406024283318,
+      "loss": 0.5752,
+      "step": 1462
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.000220475504492848,
+      "loss": 0.5179,
+      "step": 1463
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.0002203769096845916,
+      "loss": 0.609,
+      "step": 1464
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00022027827587270624,
+      "loss": 0.4996,
+      "step": 1465
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.0002201796031118559,
+      "loss": 0.6204,
+      "step": 1466
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.000220080891456726,
+      "loss": 0.5702,
+      "step": 1467
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.0002199821409620236,
+      "loss": 0.6268,
+      "step": 1468
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00021988335168247728,
+      "loss": 0.5835,
+      "step": 1469
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00021978452367283708,
+      "loss": 0.5916,
+      "step": 1470
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.0002196856569878746,
+      "loss": 0.4988,
+      "step": 1471
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00021958675168238273,
+      "loss": 0.5535,
+      "step": 1472
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00021948780781117584,
+      "loss": 0.5709,
+      "step": 1473
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00021938882542908974,
+      "loss": 0.5767,
+      "step": 1474
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00021928980459098144,
+      "loss": 0.6579,
+      "step": 1475
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00021919074535172936,
+      "loss": 0.4646,
+      "step": 1476
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00021909164776623318,
+      "loss": 0.5952,
+      "step": 1477
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00021899251188941388,
+      "loss": 0.5358,
+      "step": 1478
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00021889333777621353,
+      "loss": 0.5639,
+      "step": 1479
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00021879412548159557,
+      "loss": 0.6618,
+      "step": 1480
+    },
+    {
+      "epoch": 1.05,
+      "eval_loss": 1.096842646598816,
+      "eval_runtime": 51.0058,
+      "eval_samples_per_second": 11.43,
+      "eval_steps_per_second": 11.43,
+      "step": 1480
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.0002186948750605444,
+      "loss": 0.5244,
+      "step": 1481
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00021859558656806577,
+      "loss": 0.5698,
+      "step": 1482
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00021849626005918628,
+      "loss": 0.6591,
+      "step": 1483
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00021839689558895384,
+      "loss": 0.5037,
+      "step": 1484
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00021829749321243725,
+      "loss": 0.4636,
+      "step": 1485
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.0002181980529847264,
+      "loss": 0.5392,
+      "step": 1486
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00021809857496093199,
+      "loss": 0.5776,
+      "step": 1487
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00021799905919618596,
+      "loss": 0.6053,
+      "step": 1488
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00021789950574564092,
+      "loss": 0.5878,
+      "step": 1489
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00021779991466447043,
+      "loss": 0.5873,
+      "step": 1490
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.000217700286007869,
+      "loss": 0.6546,
+      "step": 1491
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00021760061983105178,
+      "loss": 0.5717,
+      "step": 1492
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00021750091618925494,
+      "loss": 0.6397,
+      "step": 1493
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00021740117513773517,
+      "loss": 0.3951,
+      "step": 1494
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00021730139673177014,
+      "loss": 0.6285,
+      "step": 1495
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00021720158102665798,
+      "loss": 0.6544,
+      "step": 1496
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00021710172807771767,
+      "loss": 0.6,
+      "step": 1497
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00021700183794028883,
+      "loss": 0.5857,
+      "step": 1498
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.0002169019106697315,
+      "loss": 0.5908,
+      "step": 1499
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00021680194632142658,
+      "loss": 0.6329,
+      "step": 1500
+    },
+    {
+      "epoch": 1.06,
+      "eval_loss": 1.094839096069336,
+      "eval_runtime": 49.755,
+      "eval_samples_per_second": 11.717,
+      "eval_steps_per_second": 11.717,
+      "step": 1500
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00021670194495077523,
+      "loss": 0.5295,
+      "step": 1501
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00021660190661319939,
+      "loss": 0.6694,
+      "step": 1502
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.00021650183136414129,
+      "loss": 0.547,
+      "step": 1503
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.0002164017192590637,
+      "loss": 0.5047,
+      "step": 1504
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.00021630157035344984,
+      "loss": 0.5559,
+      "step": 1505
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.00021620138470280332,
+      "loss": 0.5937,
+      "step": 1506
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.000216101162362648,
+      "loss": 0.6628,
+      "step": 1507
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.00021600090338852823,
+      "loss": 0.715,
+      "step": 1508
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.00021590060783600857,
+      "loss": 0.5655,
+      "step": 1509
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.00021580027576067386,
+      "loss": 0.6443,
+      "step": 1510
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.00021569990721812928,
+      "loss": 0.6226,
+      "step": 1511
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.00021559950226400002,
+      "loss": 0.4861,
+      "step": 1512
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.0002154990609539317,
+      "loss": 0.5186,
+      "step": 1513
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.00021539858334358985,
+      "loss": 0.5972,
+      "step": 1514
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.00021529806948866027,
+      "loss": 0.6928,
+      "step": 1515
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.00021519751944484882,
+      "loss": 0.6818,
+      "step": 1516
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.0002150969332678813,
+      "loss": 0.688,
+      "step": 1517
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.0002149963110135038,
+      "loss": 0.6521,
+      "step": 1518
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00021489565273748217,
+      "loss": 0.4853,
+      "step": 1519
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00021479495849560223,
+      "loss": 0.6025,
+      "step": 1520
+    },
+    {
+      "epoch": 1.08,
+      "eval_loss": 1.1004894971847534,
+      "eval_runtime": 53.7674,
+      "eval_samples_per_second": 10.843,
+      "eval_steps_per_second": 10.843,
+      "step": 1520
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00021469422834366985,
+      "loss": 0.6035,
+      "step": 1521
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00021459346233751076,
+      "loss": 0.6729,
+      "step": 1522
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00021449266053297046,
+      "loss": 0.535,
+      "step": 1523
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00021439182298591455,
+      "loss": 0.4882,
+      "step": 1524
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.0002142909497522281,
+      "loss": 0.6138,
+      "step": 1525
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.0002141900408878162,
+      "loss": 0.527,
+      "step": 1526
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00021408909644860362,
+      "loss": 0.7079,
+      "step": 1527
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00021398811649053477,
+      "loss": 0.532,
+      "step": 1528
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00021388710106957386,
+      "loss": 0.5041,
+      "step": 1529
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00021378605024170475,
+      "loss": 0.6047,
+      "step": 1530
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00021368496406293081,
+      "loss": 0.5966,
+      "step": 1531
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00021358384258927512,
+      "loss": 0.5737,
+      "step": 1532
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00021348268587678023,
+      "loss": 0.6588,
+      "step": 1533
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00021338149398150827,
+      "loss": 0.4082,
+      "step": 1534
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00021328026695954092,
+      "loss": 0.5514,
+      "step": 1535
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00021317900486697916,
+      "loss": 0.6485,
+      "step": 1536
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00021307770775994358,
+      "loss": 0.6768,
+      "step": 1537
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00021297637569457409,
+      "loss": 0.5505,
+      "step": 1538
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00021287500872703004,
+      "loss": 0.5287,
+      "step": 1539
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00021277360691349004,
+      "loss": 0.525,
+      "step": 1540
+    },
+    {
+      "epoch": 1.09,
+      "eval_loss": 1.0955358743667603,
+      "eval_runtime": 54.9258,
+      "eval_samples_per_second": 10.614,
+      "eval_steps_per_second": 10.614,
+      "step": 1540
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.0002126721703101521,
+      "loss": 0.5315,
+      "step": 1541
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00021257069897323338,
+      "loss": 0.6165,
+      "step": 1542
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.0002124691929589704,
+      "loss": 0.4919,
+      "step": 1543
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00021236765232361888,
+      "loss": 0.5447,
+      "step": 1544
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.00021226607712345364,
+      "loss": 0.494,
+      "step": 1545
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.0002121644674147689,
+      "loss": 0.4936,
+      "step": 1546
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.0002120628232538777,
+      "loss": 0.6811,
+      "step": 1547
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.00021196114469711232,
+      "loss": 0.6496,
+      "step": 1548
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.00021185943180082413,
+      "loss": 0.5582,
+      "step": 1549
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.0002117576846213835,
+      "loss": 0.7051,
+      "step": 1550
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.00021165590321517978,
+      "loss": 0.589,
+      "step": 1551
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.00021155408763862132,
+      "loss": 0.5717,
+      "step": 1552
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.0002114522379481354,
+      "loss": 0.5347,
+      "step": 1553
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.00021135035420016807,
+      "loss": 0.5374,
+      "step": 1554
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.0002112484364511846,
+      "loss": 0.5792,
+      "step": 1555
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.0002111464847576687,
+      "loss": 0.6382,
+      "step": 1556
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.00021104449917612317,
+      "loss": 0.6455,
+      "step": 1557
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.00021094247976306948,
+      "loss": 0.4594,
+      "step": 1558
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.0002108404265750479,
+      "loss": 0.6216,
+      "step": 1559
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00021073833966861732,
+      "loss": 0.6543,
+      "step": 1560
+    },
+    {
+      "epoch": 1.11,
+      "eval_loss": 1.0919363498687744,
+      "eval_runtime": 47.7489,
+      "eval_samples_per_second": 12.21,
+      "eval_steps_per_second": 12.21,
+      "step": 1560
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00021063621910035542,
+      "loss": 0.7709,
+      "step": 1561
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00021053406492685853,
+      "loss": 0.5886,
+      "step": 1562
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00021043187720474156,
+      "loss": 0.5469,
+      "step": 1563
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00021032965599063796,
+      "loss": 0.531,
+      "step": 1564
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00021022740134119994,
+      "loss": 0.679,
+      "step": 1565
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00021012511331309803,
+      "loss": 0.5541,
+      "step": 1566
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.0002100227919630213,
+      "loss": 0.5757,
+      "step": 1567
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00020992043734767738,
+      "loss": 0.522,
+      "step": 1568
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00020981804952379228,
+      "loss": 0.5151,
+      "step": 1569
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00020971562854811042,
+      "loss": 0.6289,
+      "step": 1570
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00020961317447739457,
+      "loss": 0.7154,
+      "step": 1571
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00020951068736842587,
+      "loss": 0.5404,
+      "step": 1572
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.00020940816727800377,
+      "loss": 0.6126,
+      "step": 1573
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.00020930561426294603,
+      "loss": 0.5097,
+      "step": 1574
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.0002092030283800885,
+      "loss": 0.6422,
+      "step": 1575
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.0002091004096862855,
+      "loss": 0.508,
+      "step": 1576
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.00020899775823840925,
+      "loss": 0.7503,
+      "step": 1577
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.0002088950740933504,
+      "loss": 0.6047,
+      "step": 1578
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.00020879235730801755,
+      "loss": 0.5492,
+      "step": 1579
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.0002086896079393374,
+      "loss": 0.6318,
+      "step": 1580
+    },
+    {
+      "epoch": 1.12,
+      "eval_loss": 1.0921388864517212,
+      "eval_runtime": 51.7357,
+      "eval_samples_per_second": 11.269,
+      "eval_steps_per_second": 11.269,
+      "step": 1580
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.0002085868260442548,
+      "loss": 0.5265,
+      "step": 1581
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.00020848401167973248,
+      "loss": 0.4705,
+      "step": 1582
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.00020838116490275133,
+      "loss": 0.5687,
+      "step": 1583
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.00020827828577031009,
+      "loss": 0.5737,
+      "step": 1584
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.00020817537433942546,
+      "loss": 0.58,
+      "step": 1585
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.000208072430667132,
+      "loss": 0.6476,
+      "step": 1586
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.00020796945481048234,
+      "loss": 0.5886,
+      "step": 1587
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.0002078664468265466,
+      "loss": 0.6251,
+      "step": 1588
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.00020776340677241298,
+      "loss": 0.5879,
+      "step": 1589
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.00020766033470518745,
+      "loss": 0.5474,
+      "step": 1590
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.0002075572306819935,
+      "loss": 0.5323,
+      "step": 1591
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.00020745409475997256,
+      "loss": 0.4983,
+      "step": 1592
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.0002073509269962836,
+      "loss": 0.5707,
+      "step": 1593
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.00020724772744810333,
+      "loss": 0.5318,
+      "step": 1594
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.00020714449617262602,
+      "loss": 0.6357,
+      "step": 1595
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.00020704123322706354,
+      "loss": 0.6337,
+      "step": 1596
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.00020693793866864525,
+      "loss": 0.4368,
+      "step": 1597
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.00020683461255461817,
+      "loss": 0.5902,
+      "step": 1598
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.0002067312549422467,
+      "loss": 0.4812,
+      "step": 1599
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.00020662786588881274,
+      "loss": 0.6152,
+      "step": 1600
+    },
+    {
+      "epoch": 1.13,
+      "eval_loss": 1.0973610877990723,
+      "eval_runtime": 49.0364,
+      "eval_samples_per_second": 11.889,
+      "eval_steps_per_second": 11.889,
+      "step": 1600
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00020652444545161557,
+      "loss": 0.6391,
+      "step": 1601
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.0002064209936879719,
+      "loss": 0.4768,
+      "step": 1602
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.0002063175106552158,
+      "loss": 0.5503,
+      "step": 1603
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00020621399641069862,
+      "loss": 0.5964,
+      "step": 1604
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00020611045101178916,
+      "loss": 0.7082,
+      "step": 1605
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00020600687451587324,
+      "loss": 0.466,
+      "step": 1606
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00020590326698035412,
+      "loss": 0.6567,
+      "step": 1607
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.0002057996284626522,
+      "loss": 0.652,
+      "step": 1608
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00020569595902020503,
+      "loss": 0.6165,
+      "step": 1609
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00020559225871046735,
+      "loss": 0.5823,
+      "step": 1610
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.0002054885275909109,
+      "loss": 0.5132,
+      "step": 1611
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00020538476571902467,
+      "loss": 0.5986,
+      "step": 1612
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00020528097315231441,
+      "loss": 0.5838,
+      "step": 1613
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00020517714994830328,
+      "loss": 0.5653,
+      "step": 1614
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00020507329616453103,
+      "loss": 0.5073,
+      "step": 1615
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.00020496941185855463,
+      "loss": 0.5678,
+      "step": 1616
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.0002048654970879478,
+      "loss": 0.6329,
+      "step": 1617
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.0002047615519103012,
+      "loss": 0.648,
+      "step": 1618
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.00020465757638322242,
+      "loss": 0.4906,
+      "step": 1619
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.00020455357056433577,
+      "loss": 0.6488,
+      "step": 1620
+    },
+    {
+      "epoch": 1.15,
+      "eval_loss": 1.0910096168518066,
+      "eval_runtime": 55.6274,
+      "eval_samples_per_second": 10.48,
+      "eval_steps_per_second": 10.48,
+      "step": 1620
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.00020444953451128234,
+      "loss": 0.5552,
+      "step": 1621
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.00020434546828172002,
+      "loss": 0.4332,
+      "step": 1622
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.0002042413719333235,
+      "loss": 0.5962,
+      "step": 1623
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.0002041372455237839,
+      "loss": 0.5155,
+      "step": 1624
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.00020403308911080936,
+      "loss": 0.5259,
+      "step": 1625
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.00020392890275212441,
+      "loss": 0.4806,
+      "step": 1626
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.00020382468650547022,
+      "loss": 0.5982,
+      "step": 1627
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.0002037204404286045,
+      "loss": 0.6369,
+      "step": 1628
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.00020361616457930157,
+      "loss": 0.56,
+      "step": 1629
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00020351185901535223,
+      "loss": 0.5498,
+      "step": 1630
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00020340752379456372,
+      "loss": 0.5222,
+      "step": 1631
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.0002033031589747597,
+      "loss": 0.6969,
+      "step": 1632
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00020319876461378033,
+      "loss": 0.5837,
+      "step": 1633
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00020309434076948199,
+      "loss": 0.5909,
+      "step": 1634
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00020298988749973743,
+      "loss": 0.5174,
+      "step": 1635
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00020288540486243593,
+      "loss": 0.6167,
+      "step": 1636
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00020278089291548272,
+      "loss": 0.6272,
+      "step": 1637
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.0002026763517167995,
+      "loss": 0.6179,
+      "step": 1638
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00020257178132432406,
+      "loss": 0.6105,
+      "step": 1639
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00020246718179601052,
+      "loss": 0.6168,
+      "step": 1640
+    },
+    {
+      "epoch": 1.16,
+      "eval_loss": 1.0927194356918335,
+      "eval_runtime": 47.0976,
+      "eval_samples_per_second": 12.379,
+      "eval_steps_per_second": 12.379,
+      "step": 1640
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00020236255318982894,
+      "loss": 0.5148,
+      "step": 1641
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00020225789556376568,
+      "loss": 0.641,
+      "step": 1642
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00020215320897582303,
+      "loss": 0.5992,
+      "step": 1643
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.0002020484934840195,
+      "loss": 0.5443,
+      "step": 1644
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.00020194374914638943,
+      "loss": 0.5134,
+      "step": 1645
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.0002018389760209833,
+      "loss": 0.637,
+      "step": 1646
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.00020173417416586743,
+      "loss": 0.4957,
+      "step": 1647
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.00020162934363912417,
+      "loss": 0.5196,
+      "step": 1648
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.0002015244844988517,
+      "loss": 0.6205,
+      "step": 1649
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.00020141959680316406,
+      "loss": 0.5359,
+      "step": 1650
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.00020131468061019116,
+      "loss": 0.593,
+      "step": 1651
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.00020120973597807857,
+      "loss": 0.5401,
+      "step": 1652
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.00020110476296498784,
+      "loss": 0.5923,
+      "step": 1653
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.00020099976162909602,
+      "loss": 0.4666,
+      "step": 1654
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.00020089473202859602,
+      "loss": 0.6184,
+      "step": 1655
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.00020078967422169634,
+      "loss": 0.595,
+      "step": 1656
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.00020068458826662112,
+      "loss": 0.528,
+      "step": 1657
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.0002005794742216101,
+      "loss": 0.492,
+      "step": 1658
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00020047433214491866,
+      "loss": 0.5357,
+      "step": 1659
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.0002003691620948176,
+      "loss": 0.559,
+      "step": 1660
+    },
+    {
+      "epoch": 1.18,
+      "eval_loss": 1.0928723812103271,
+      "eval_runtime": 48.0928,
+      "eval_samples_per_second": 12.122,
+      "eval_steps_per_second": 12.122,
+      "step": 1660
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.0002002639641295933,
+      "loss": 0.5241,
+      "step": 1661
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00020015873830754754,
+      "loss": 0.6363,
+      "step": 1662
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00020005348468699763,
+      "loss": 0.6653,
+      "step": 1663
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00019994820332627628,
+      "loss": 0.647,
+      "step": 1664
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00019984289428373145,
+      "loss": 0.7626,
+      "step": 1665
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00019973755761772653,
+      "loss": 0.6367,
+      "step": 1666
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00019963219338664032,
+      "loss": 0.6334,
+      "step": 1667
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00019952680164886665,
+      "loss": 0.6275,
+      "step": 1668
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00019942138246281486,
+      "loss": 0.6374,
+      "step": 1669
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.0001993159358869093,
+      "loss": 0.6698,
+      "step": 1670
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00019921046197958963,
+      "loss": 0.5068,
+      "step": 1671
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00019910496079931058,
+      "loss": 0.6298,
+      "step": 1672
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00019899943240454205,
+      "loss": 0.4854,
+      "step": 1673
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00019889387685376894,
+      "loss": 0.6441,
+      "step": 1674
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.0001987882942054913,
+      "loss": 0.572,
+      "step": 1675
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00019868268451822413,
+      "loss": 0.5392,
+      "step": 1676
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00019857704785049748,
+      "loss": 0.4915,
+      "step": 1677
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00019847138426085624,
+      "loss": 0.5839,
+      "step": 1678
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00019836569380786033,
+      "loss": 0.6578,
+      "step": 1679
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00019825997655008456,
+      "loss": 0.6186,
+      "step": 1680
+    },
+    {
+      "epoch": 1.19,
+      "eval_loss": 1.0958151817321777,
+      "eval_runtime": 48.2662,
+      "eval_samples_per_second": 12.079,
+      "eval_steps_per_second": 12.079,
+      "step": 1680
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00019815423254611846,
+      "loss": 0.5877,
+      "step": 1681
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00019804846185456657,
+      "loss": 0.5903,
+      "step": 1682
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00019794266453404807,
+      "loss": 0.4884,
+      "step": 1683
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00019783684064319696,
+      "loss": 0.5385,
+      "step": 1684
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.000197730990240662,
+      "loss": 0.6006,
+      "step": 1685
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.00019762511338510658,
+      "loss": 0.5423,
+      "step": 1686
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.00019751921013520872,
+      "loss": 0.6374,
+      "step": 1687
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.00019741328054966124,
+      "loss": 0.6309,
+      "step": 1688
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.00019730732468717133,
+      "loss": 0.6399,
+      "step": 1689
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.0001972013426064609,
+      "loss": 0.6247,
+      "step": 1690
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.00019709533436626636,
+      "loss": 0.6261,
+      "step": 1691
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.00019698930002533857,
+      "loss": 0.4662,
+      "step": 1692
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.0001968832396424429,
+      "loss": 0.586,
+      "step": 1693
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.0001967771532763591,
+      "loss": 0.6028,
+      "step": 1694
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.00019667104098588142,
+      "loss": 0.4714,
+      "step": 1695
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.00019656490282981838,
+      "loss": 0.527,
+      "step": 1696
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.0001964587388669929,
+      "loss": 0.6577,
+      "step": 1697
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.0001963525491562421,
+      "loss": 0.6602,
+      "step": 1698
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.0001962463337564175,
+      "loss": 0.5426,
+      "step": 1699
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.00019614009272638481,
+      "loss": 0.6082,
+      "step": 1700
+    },
+    {
+      "epoch": 1.21,
+      "eval_loss": 1.0884817838668823,
+      "eval_runtime": 49.24,
+      "eval_samples_per_second": 11.84,
+      "eval_steps_per_second": 11.84,
+      "step": 1700
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.0001960338261250239,
+      "loss": 0.6204,
+      "step": 1701
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.00019592753401122888,
+      "loss": 0.4837,
+      "step": 1702
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.00019582121644390793,
+      "loss": 0.7571,
+      "step": 1703
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.0001957148734819834,
+      "loss": 0.6619,
+      "step": 1704
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.00019560850518439166,
+      "loss": 0.6587,
+      "step": 1705
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.0001955021116100832,
+      "loss": 0.6691,
+      "step": 1706
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.00019539569281802236,
+      "loss": 0.5856,
+      "step": 1707
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.0001952892488671877,
+      "loss": 0.5338,
+      "step": 1708
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.00019518277981657147,
+      "loss": 0.6439,
+      "step": 1709
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.00019507628572518,
+      "loss": 0.5777,
+      "step": 1710
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.0001949697666520334,
+      "loss": 0.6245,
+      "step": 1711
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.00019486322265616578,
+      "loss": 0.4651,
+      "step": 1712
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.00019475665379662486,
+      "loss": 0.6494,
+      "step": 1713
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.00019465006013247223,
+      "loss": 0.6373,
+      "step": 1714
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.00019454344172278325,
+      "loss": 0.5584,
+      "step": 1715
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.00019443679862664694,
+      "loss": 0.5746,
+      "step": 1716
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.0001943301309031661,
+      "loss": 0.6125,
+      "step": 1717
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.00019422343861145704,
+      "loss": 0.5392,
+      "step": 1718
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.00019411672181064978,
+      "loss": 0.5458,
+      "step": 1719
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.00019400998055988796,
+      "loss": 0.6769,
+      "step": 1720
+    },
+    {
+      "epoch": 1.22,
+      "eval_loss": 1.0937696695327759,
+      "eval_runtime": 50.5527,
+      "eval_samples_per_second": 11.533,
+      "eval_steps_per_second": 11.533,
+      "step": 1720
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.00019390321491832862,
+      "loss": 0.6768,
+      "step": 1721
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.00019379642494514248,
+      "loss": 0.5996,
+      "step": 1722
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.00019368961069951365,
+      "loss": 0.559,
+      "step": 1723
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.00019358277224063968,
+      "loss": 0.5805,
+      "step": 1724
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.00019347590962773167,
+      "loss": 0.5653,
+      "step": 1725
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.0001933690229200139,
+      "loss": 0.6118,
+      "step": 1726
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.00019326211217672422,
+      "loss": 0.5735,
+      "step": 1727
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.0001931551774571136,
+      "loss": 0.5925,
+      "step": 1728
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00019304821882044646,
+      "loss": 0.6268,
+      "step": 1729
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00019294123632600035,
+      "loss": 0.614,
+      "step": 1730
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00019283423003306615,
+      "loss": 0.5044,
+      "step": 1731
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00019272720000094792,
+      "loss": 0.599,
+      "step": 1732
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00019262014628896275,
+      "loss": 0.6106,
+      "step": 1733
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.000192513068956441,
+      "loss": 0.551,
+      "step": 1734
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.0001924059680627261,
+      "loss": 0.559,
+      "step": 1735
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00019229884366717435,
+      "loss": 0.6952,
+      "step": 1736
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00019219169582915537,
+      "loss": 0.4648,
+      "step": 1737
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00019208452460805152,
+      "loss": 0.5619,
+      "step": 1738
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00019197733006325832,
+      "loss": 0.6213,
+      "step": 1739
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00019187011225418413,
+      "loss": 0.5256,
+      "step": 1740
+    },
+    {
+      "epoch": 1.23,
+      "eval_loss": 1.0908206701278687,
+      "eval_runtime": 50.3356,
+      "eval_samples_per_second": 11.582,
+      "eval_steps_per_second": 11.582,
+      "step": 1740
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00019176287124025008,
+      "loss": 0.6173,
+      "step": 1741
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.00019165560708089035,
+      "loss": 0.6461,
+      "step": 1742
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.0001915483198355519,
+      "loss": 0.5314,
+      "step": 1743
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.00019144100956369439,
+      "loss": 0.4547,
+      "step": 1744
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.0001913336763247904,
+      "loss": 0.5822,
+      "step": 1745
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.0001912263201783251,
+      "loss": 0.6169,
+      "step": 1746
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.00019111894118379634,
+      "loss": 0.5191,
+      "step": 1747
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.00019101153940071479,
+      "loss": 0.6198,
+      "step": 1748
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.0001909041148886036,
+      "loss": 0.5292,
+      "step": 1749
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.00019079666770699857,
+      "loss": 0.535,
+      "step": 1750
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.00019068919791544813,
+      "loss": 0.5327,
+      "step": 1751
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.00019058170557351313,
+      "loss": 0.5432,
+      "step": 1752
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.000190474190740767,
+      "loss": 0.5376,
+      "step": 1753
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.00019036665347679554,
+      "loss": 0.6404,
+      "step": 1754
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.00019025909384119704,
+      "loss": 0.607,
+      "step": 1755
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.0001901515118935823,
+      "loss": 0.5809,
+      "step": 1756
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00019004390769357426,
+      "loss": 0.5705,
+      "step": 1757
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00018993628130080834,
+      "loss": 0.6295,
+      "step": 1758
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.0001898286327749322,
+      "loss": 0.6257,
+      "step": 1759
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00018972096217560587,
+      "loss": 0.512,
+      "step": 1760
+    },
+    {
+      "epoch": 1.25,
+      "eval_loss": 1.0900548696517944,
+      "eval_runtime": 49.2312,
+      "eval_samples_per_second": 11.842,
+      "eval_steps_per_second": 11.842,
+      "step": 1760
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00018961326956250146,
+      "loss": 0.6692,
+      "step": 1761
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00018950555499530337,
+      "loss": 0.5409,
+      "step": 1762
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00018939781853370824,
+      "loss": 0.6741,
+      "step": 1763
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00018929006023742464,
+      "loss": 0.5716,
+      "step": 1764
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00018918228016617347,
+      "loss": 0.5257,
+      "step": 1765
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00018907447837968747,
+      "loss": 0.5506,
+      "step": 1766
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.0001889666549377116,
+      "loss": 0.5209,
+      "step": 1767
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00018885880990000275,
+      "loss": 0.6528,
+      "step": 1768
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.0001887509433263298,
+      "loss": 0.4801,
+      "step": 1769
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00018864305527647348,
+      "loss": 0.514,
+      "step": 1770
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.00018853514581022653,
+      "loss": 0.4139,
+      "step": 1771
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.00018842721498739357,
+      "loss": 0.5566,
+      "step": 1772
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.00018831926286779093,
+      "loss": 0.5583,
+      "step": 1773
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.00018821128951124688,
+      "loss": 0.6434,
+      "step": 1774
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.0001881032949776013,
+      "loss": 0.6583,
+      "step": 1775
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.00018799527932670607,
+      "loss": 0.6445,
+      "step": 1776
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.0001878872426184244,
+      "loss": 0.7203,
+      "step": 1777
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.0001877791849126316,
+      "loss": 0.5216,
+      "step": 1778
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.0001876711062692142,
+      "loss": 0.6199,
+      "step": 1779
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.00018756300674807063,
+      "loss": 0.351,
+      "step": 1780
+    },
+    {
+      "epoch": 1.26,
+      "eval_loss": 1.0978995561599731,
+      "eval_runtime": 47.4524,
+      "eval_samples_per_second": 12.286,
+      "eval_steps_per_second": 12.286,
+      "step": 1780
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.00018745488640911076,
+      "loss": 0.6729,
+      "step": 1781
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.00018734674531225605,
+      "loss": 0.6926,
+      "step": 1782
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.00018723858351743944,
+      "loss": 0.5642,
+      "step": 1783
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.0001871304010846053,
+      "loss": 0.7607,
+      "step": 1784
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.00018702219807370957,
+      "loss": 0.6865,
+      "step": 1785
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.0001869139745447194,
+      "loss": 0.6922,
+      "step": 1786
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.0001868057305576135,
+      "loss": 0.6083,
+      "step": 1787
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.00018669746617238175,
+      "loss": 0.5478,
+      "step": 1788
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.0001865891814490255,
+      "loss": 0.5005,
+      "step": 1789
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.00018648087644755726,
+      "loss": 0.7622,
+      "step": 1790
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.0001863725512280008,
+      "loss": 0.5554,
+      "step": 1791
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.00018626420585039115,
+      "loss": 0.5293,
+      "step": 1792
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.00018615584037477438,
+      "loss": 0.5803,
+      "step": 1793
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.00018604745486120787,
+      "loss": 0.6582,
+      "step": 1794
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.00018593904936975992,
+      "loss": 0.4869,
+      "step": 1795
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.00018583062396051013,
+      "loss": 0.624,
+      "step": 1796
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.00018572217869354893,
+      "loss": 0.4632,
+      "step": 1797
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.00018561371362897782,
+      "loss": 0.5311,
+      "step": 1798
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.00018550522882690933,
+      "loss": 0.5276,
+      "step": 1799
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.00018539672434746693,
+      "loss": 0.4918,
+      "step": 1800
+    },
+    {
+      "epoch": 1.28,
+      "eval_loss": 1.0977996587753296,
+      "eval_runtime": 50.4687,
+      "eval_samples_per_second": 11.552,
+      "eval_steps_per_second": 11.552,
+      "step": 1800
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.00018528820025078485,
+      "loss": 0.5862,
+      "step": 1801
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.0001851796565970083,
+      "loss": 0.6469,
+      "step": 1802
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.00018507109344629343,
+      "loss": 0.532,
+      "step": 1803
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.000184962510858807,
+      "loss": 0.5998,
+      "step": 1804
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.00018485390889472667,
+      "loss": 0.5583,
+      "step": 1805
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.00018474528761424074,
+      "loss": 0.4975,
+      "step": 1806
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.00018463664707754838,
+      "loss": 0.6029,
+      "step": 1807
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.0001845279873448592,
+      "loss": 0.5277,
+      "step": 1808
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.0001844193084763937,
+      "loss": 0.5144,
+      "step": 1809
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.0001843106105323827,
+      "loss": 0.5519,
+      "step": 1810
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.00018420189357306794,
+      "loss": 0.5969,
+      "step": 1811
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.00018409315765870136,
+      "loss": 0.6245,
+      "step": 1812
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00018398440284954568,
+      "loss": 0.6192,
+      "step": 1813
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00018387562920587388,
+      "loss": 0.6249,
+      "step": 1814
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00018376683678796948,
+      "loss": 0.6628,
+      "step": 1815
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00018365802565612646,
+      "loss": 0.436,
+      "step": 1816
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00018354919587064894,
+      "loss": 0.5617,
+      "step": 1817
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00018344034749185172,
+      "loss": 0.5271,
+      "step": 1818
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.0001833314805800596,
+      "loss": 0.5647,
+      "step": 1819
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00018322259519560786,
+      "loss": 0.54,
+      "step": 1820
+    },
+    {
+      "epoch": 1.29,
+      "eval_loss": 1.0970478057861328,
+      "eval_runtime": 54.7545,
+      "eval_samples_per_second": 10.648,
+      "eval_steps_per_second": 10.648,
+      "step": 1820
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00018311369139884187,
+      "loss": 0.5238,
+      "step": 1821
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00018300476925011728,
+      "loss": 0.5296,
+      "step": 1822
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00018289582880979993,
+      "loss": 0.6601,
+      "step": 1823
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.0001827868701382658,
+      "loss": 0.6206,
+      "step": 1824
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.0001826778932959008,
+      "loss": 0.5991,
+      "step": 1825
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00018256889834310122,
+      "loss": 0.65,
+      "step": 1826
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.00018245988534027312,
+      "loss": 0.556,
+      "step": 1827
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.00018235085434783262,
+      "loss": 0.5536,
+      "step": 1828
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.000182241805426206,
+      "loss": 0.4708,
+      "step": 1829
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.00018213273863582915,
+      "loss": 0.514,
+      "step": 1830
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.00018202365403714819,
+      "loss": 0.7325,
+      "step": 1831
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.00018191455169061884,
+      "loss": 0.6598,
+      "step": 1832
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.00018180543165670688,
+      "loss": 0.5718,
+      "step": 1833
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.00018169629399588774,
+      "loss": 0.5413,
+      "step": 1834
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.00018158713876864665,
+      "loss": 0.5362,
+      "step": 1835
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.00018147796603547865,
+      "loss": 0.4725,
+      "step": 1836
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.0001813687758568884,
+      "loss": 0.5987,
+      "step": 1837
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.0001812595682933902,
+      "loss": 0.5914,
+      "step": 1838
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.00018115034340550812,
+      "loss": 0.536,
+      "step": 1839
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.00018104110125377566,
+      "loss": 0.4865,
+      "step": 1840
+    },
+    {
+      "epoch": 1.3,
+      "eval_loss": 1.1005055904388428,
+      "eval_runtime": 48.0742,
+      "eval_samples_per_second": 12.127,
+      "eval_steps_per_second": 12.127,
+      "step": 1840
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00018093184189873606,
+      "loss": 0.6555,
+      "step": 1841
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00018082256540094196,
+      "loss": 0.7336,
+      "step": 1842
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.0001807132718209556,
+      "loss": 0.6596,
+      "step": 1843
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.0001806039612193486,
+      "loss": 0.6412,
+      "step": 1844
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.0001804946336567021,
+      "loss": 0.643,
+      "step": 1845
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00018038528919360656,
+      "loss": 0.5978,
+      "step": 1846
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00018027592789066183,
+      "loss": 0.5296,
+      "step": 1847
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00018016654980847712,
+      "loss": 0.5261,
+      "step": 1848
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.0001800571550076709,
+      "loss": 0.5704,
+      "step": 1849
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00017994774354887093,
+      "loss": 0.5677,
+      "step": 1850
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00017983831549271424,
+      "loss": 0.5956,
+      "step": 1851
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00017972887089984696,
+      "loss": 0.5644,
+      "step": 1852
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00017961940983092454,
+      "loss": 0.6385,
+      "step": 1853
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00017950993234661135,
+      "loss": 0.5025,
+      "step": 1854
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00017940043850758102,
+      "loss": 0.5746,
+      "step": 1855
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.0001792909283745162,
+      "loss": 0.5761,
+      "step": 1856
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00017918140200810857,
+      "loss": 0.5657,
+      "step": 1857
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.0001790718594690588,
+      "loss": 0.4758,
+      "step": 1858
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00017896230081807652,
+      "loss": 0.5534,
+      "step": 1859
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.0001788527261158803,
+      "loss": 0.5787,
+      "step": 1860
+    },
+    {
+      "epoch": 1.32,
+      "eval_loss": 1.0990447998046875,
+      "eval_runtime": 49.424,
+      "eval_samples_per_second": 11.796,
+      "eval_steps_per_second": 11.796,
+      "step": 1860
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00017874313542319756,
+      "loss": 0.5979,
+      "step": 1861
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.0001786335288007647,
+      "loss": 0.5585,
+      "step": 1862
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00017852390630932678,
+      "loss": 0.5751,
+      "step": 1863
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00017841426800963784,
+      "loss": 0.4355,
+      "step": 1864
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00017830461396246052,
+      "loss": 0.4937,
+      "step": 1865
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.0001781949442285663,
+      "loss": 0.6149,
+      "step": 1866
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00017808525886873526,
+      "loss": 0.5726,
+      "step": 1867
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00017797555794375628,
+      "loss": 0.5082,
+      "step": 1868
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.00017786584151442662,
+      "loss": 0.6501,
+      "step": 1869
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.00017775610964155243,
+      "loss": 0.6216,
+      "step": 1870
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.00017764636238594818,
+      "loss": 0.5988,
+      "step": 1871
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.00017753659980843705,
+      "loss": 0.5554,
+      "step": 1872
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.0001774268219698506,
+      "loss": 0.5242,
+      "step": 1873
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.00017731702893102877,
+      "loss": 0.5559,
+      "step": 1874
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.00017720722075282016,
+      "loss": 0.5783,
+      "step": 1875
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.00017709739749608157,
+      "loss": 0.626,
+      "step": 1876
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.00017698755922167822,
+      "loss": 0.5238,
+      "step": 1877
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.00017687770599048358,
+      "loss": 0.5822,
+      "step": 1878
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.0001767678378633796,
+      "loss": 0.577,
+      "step": 1879
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.00017665795490125625,
+      "loss": 0.5838,
+      "step": 1880
+    },
+    {
+      "epoch": 1.33,
+      "eval_loss": 1.0938432216644287,
+      "eval_runtime": 50.4107,
+      "eval_samples_per_second": 11.565,
+      "eval_steps_per_second": 11.565,
+      "step": 1880
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.0001765480571650119,
+      "loss": 0.6138,
+      "step": 1881
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.000176438144715553,
+      "loss": 0.5302,
+      "step": 1882
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.0001763282176137942,
+      "loss": 0.5314,
+      "step": 1883
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.0001762182759206583,
+      "loss": 0.5473,
+      "step": 1884
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00017610831969707612,
+      "loss": 0.5281,
+      "step": 1885
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00017599834900398654,
+      "loss": 0.6387,
+      "step": 1886
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00017588836390233656,
+      "loss": 0.5136,
+      "step": 1887
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00017577836445308102,
+      "loss": 0.5969,
+      "step": 1888
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00017566835071718278,
+      "loss": 0.5286,
+      "step": 1889
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00017555832275561266,
+      "loss": 0.6105,
+      "step": 1890
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00017544828062934927,
+      "loss": 0.5867,
+      "step": 1891
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00017533822439937916,
+      "loss": 0.5322,
+      "step": 1892
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00017522815412669663,
+      "loss": 0.6945,
+      "step": 1893
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00017511806987230384,
+      "loss": 0.6155,
+      "step": 1894
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00017500797169721068,
+      "loss": 0.5211,
+      "step": 1895
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00017489785966243456,
+      "loss": 0.5479,
+      "step": 1896
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00017478773382900093,
+      "loss": 0.5612,
+      "step": 1897
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00017467759425794253,
+      "loss": 0.6312,
+      "step": 1898
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00017456744101030004,
+      "loss": 0.5534,
+      "step": 1899
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.0001744572741471214,
+      "loss": 0.6168,
+      "step": 1900
+    },
+    {
+      "epoch": 1.35,
+      "eval_loss": 1.0967601537704468,
+      "eval_runtime": 50.3562,
+      "eval_samples_per_second": 11.578,
+      "eval_steps_per_second": 11.578,
+      "step": 1900
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00017434709372946233,
+      "loss": 0.6661,
+      "step": 1901
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00017423689981838598,
+      "loss": 0.5957,
+      "step": 1902
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.000174126692474963,
+      "loss": 0.6559,
+      "step": 1903
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00017401647176027143,
+      "loss": 0.5997,
+      "step": 1904
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00017390623773539678,
+      "loss": 0.5191,
+      "step": 1905
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00017379599046143197,
+      "loss": 0.6196,
+      "step": 1906
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00017368572999947707,
+      "loss": 0.5441,
+      "step": 1907
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.0001735754564106397,
+      "loss": 0.5253,
+      "step": 1908
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00017346516975603462,
+      "loss": 0.5584,
+      "step": 1909
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.0001733548700967839,
+      "loss": 0.4892,
+      "step": 1910
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00017324455749401671,
+      "loss": 0.5207,
+      "step": 1911
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00017313423200886947,
+      "loss": 0.6569,
+      "step": 1912
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00017302389370248578,
+      "loss": 0.6395,
+      "step": 1913
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00017291354263601632,
+      "loss": 0.6677,
+      "step": 1914
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00017280317887061875,
+      "loss": 0.4957,
+      "step": 1915
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00017269280246745791,
+      "loss": 0.4992,
+      "step": 1916
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.0001725824134877055,
+      "loss": 0.544,
+      "step": 1917
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00017247201199254027,
+      "loss": 0.625,
+      "step": 1918
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.000172361598043148,
+      "loss": 0.6244,
+      "step": 1919
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.0001722511717007211,
+      "loss": 0.6122,
+      "step": 1920
+    },
+    {
+      "epoch": 1.36,
+      "eval_loss": 1.099109172821045,
+      "eval_runtime": 50.329,
+      "eval_samples_per_second": 11.584,
+      "eval_steps_per_second": 11.584,
+      "step": 1920
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00017214073302645918,
+      "loss": 0.6107,
+      "step": 1921
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00017203028208156838,
+      "loss": 0.4914,
+      "step": 1922
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.0001719198189272619,
+      "loss": 0.5926,
+      "step": 1923
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00017180934362475954,
+      "loss": 0.6825,
+      "step": 1924
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.0001716988562352879,
+      "loss": 0.664,
+      "step": 1925
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00017158835682008026,
+      "loss": 0.5436,
+      "step": 1926
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00017147784544037657,
+      "loss": 0.5883,
+      "step": 1927
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00017136732215742342,
+      "loss": 0.4737,
+      "step": 1928
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00017125678703247386,
+      "loss": 0.6251,
+      "step": 1929
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00017114624012678784,
+      "loss": 0.4773,
+      "step": 1930
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00017103568150163146,
+      "loss": 0.6284,
+      "step": 1931
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.0001709251112182775,
+      "loss": 0.5929,
+      "step": 1932
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00017081452933800524,
+      "loss": 0.5722,
+      "step": 1933
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00017070393592210034,
+      "loss": 0.5837,
+      "step": 1934
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00017059333103185478,
+      "loss": 0.6329,
+      "step": 1935
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00017048271472856706,
+      "loss": 0.5034,
+      "step": 1936
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00017037208707354175,
+      "loss": 0.6004,
+      "step": 1937
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00017026144812809003,
+      "loss": 0.5349,
+      "step": 1938
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.0001701507979535291,
+      "loss": 0.5609,
+      "step": 1939
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.0001700401366111824,
+      "loss": 0.6353,
+      "step": 1940
+    },
+    {
+      "epoch": 1.38,
+      "eval_loss": 1.0890063047409058,
+      "eval_runtime": 50.2158,
+      "eval_samples_per_second": 11.61,
+      "eval_steps_per_second": 11.61,
+      "step": 1940
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00016992946416237973,
+      "loss": 0.5855,
+      "step": 1941
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00016981878066845687,
+      "loss": 0.5607,
+      "step": 1942
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.0001697080861907558,
+      "loss": 0.5345,
+      "step": 1943
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00016959738079062457,
+      "loss": 0.6367,
+      "step": 1944
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.0001694866645294173,
+      "loss": 0.5333,
+      "step": 1945
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.0001693759374684941,
+      "loss": 0.4577,
+      "step": 1946
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00016926519966922108,
+      "loss": 0.5345,
+      "step": 1947
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00016915445119297027,
+      "loss": 0.6424,
+      "step": 1948
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00016904369210111976,
+      "loss": 0.6308,
+      "step": 1949
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.0001689329224550533,
+      "loss": 0.5723,
+      "step": 1950
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00016882214231616062,
+      "loss": 0.6421,
+      "step": 1951
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00016871135174583726,
+      "loss": 0.5964,
+      "step": 1952
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00016860055080548453,
+      "loss": 0.5362,
+      "step": 1953
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.0001684897395565095,
+      "loss": 0.5417,
+      "step": 1954
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.0001683789180603249,
+      "loss": 0.5774,
+      "step": 1955
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00016826808637834925,
+      "loss": 0.6034,
+      "step": 1956
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00016815724457200655,
+      "loss": 0.632,
+      "step": 1957
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.0001680463927027266,
+      "loss": 0.627,
+      "step": 1958
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.0001679355308319446,
+      "loss": 0.6638,
+      "step": 1959
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00016782465902110137,
+      "loss": 0.6427,
+      "step": 1960
+    },
+    {
+      "epoch": 1.39,
+      "eval_loss": 1.0879871845245361,
+      "eval_runtime": 49.8874,
+      "eval_samples_per_second": 11.686,
+      "eval_steps_per_second": 11.686,
+      "step": 1960
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.0001677137773316433,
+      "loss": 0.5843,
+      "step": 1961
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00016760288582502218,
+      "loss": 0.4866,
+      "step": 1962
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00016749198456269523,
+      "loss": 0.6352,
+      "step": 1963
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.0001673810736061251,
+      "loss": 0.5431,
+      "step": 1964
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.0001672701530167799,
+      "loss": 0.586,
+      "step": 1965
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00016715922285613286,
+      "loss": 0.7694,
+      "step": 1966
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00016704828318566282,
+      "loss": 0.55,
+      "step": 1967
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00016693733406685356,
+      "loss": 0.7253,
+      "step": 1968
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00016682637556119436,
+      "loss": 0.5234,
+      "step": 1969
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00016671540773017949,
+      "loss": 0.6343,
+      "step": 1970
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00016660443063530866,
+      "loss": 0.5282,
+      "step": 1971
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00016649344433808638,
+      "loss": 0.7317,
+      "step": 1972
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00016638244890002253,
+      "loss": 0.625,
+      "step": 1973
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.000166271444382632,
+      "loss": 0.4352,
+      "step": 1974
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00016616043084743457,
+      "loss": 0.6876,
+      "step": 1975
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00016604940835595515,
+      "loss": 0.494,
+      "step": 1976
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.0001659383769697236,
+      "loss": 0.5882,
+      "step": 1977
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00016582733675027473,
+      "loss": 0.6974,
+      "step": 1978
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00016571628775914808,
+      "loss": 0.5798,
+      "step": 1979
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.0001656052300578883,
+      "loss": 0.6051,
+      "step": 1980
+    },
+    {
+      "epoch": 1.4,
+      "eval_loss": 1.0964124202728271,
+      "eval_runtime": 49.9345,
+      "eval_samples_per_second": 11.675,
+      "eval_steps_per_second": 11.675,
+      "step": 1980
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00016549416370804465,
+      "loss": 0.4925,
+      "step": 1981
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.0001653830887711714,
+      "loss": 0.5814,
+      "step": 1982
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.0001652720053088273,
+      "loss": 0.5854,
+      "step": 1983
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.0001651609133825761,
+      "loss": 0.5736,
+      "step": 1984
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.0001650498130539861,
+      "loss": 0.6087,
+      "step": 1985
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.0001649387043846302,
+      "loss": 0.6169,
+      "step": 1986
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.00016482758743608614,
+      "loss": 0.5541,
+      "step": 1987
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.000164716462269936,
+      "loss": 0.6078,
+      "step": 1988
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.0001646053289477666,
+      "loss": 0.4834,
+      "step": 1989
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.00016449418753116914,
+      "loss": 0.5419,
+      "step": 1990
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.0001643830380817394,
+      "loss": 0.6707,
+      "step": 1991
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.00016427188066107755,
+      "loss": 0.5784,
+      "step": 1992
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.00016416071533078826,
+      "loss": 0.5912,
+      "step": 1993
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.00016404954215248053,
+      "loss": 0.636,
+      "step": 1994
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.00016393836118776766,
+      "loss": 0.631,
+      "step": 1995
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00016382717249826742,
+      "loss": 0.5898,
+      "step": 1996
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00016371597614560162,
+      "loss": 0.6842,
+      "step": 1997
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00016360477219139657,
+      "loss": 0.5659,
+      "step": 1998
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00016349356069728262,
+      "loss": 0.5919,
+      "step": 1999
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.0001633823417248944,
+      "loss": 0.6461,
+      "step": 2000
+    },
+    {
+      "epoch": 1.42,
+      "eval_loss": 1.1023415327072144,
+      "eval_runtime": 51.2037,
+      "eval_samples_per_second": 11.386,
+      "eval_steps_per_second": 11.386,
+      "step": 2000
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00016327111533587062,
+      "loss": 0.713,
+      "step": 2001
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.0001631598815918541,
+      "loss": 0.6183,
+      "step": 2002
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00016304864055449183,
+      "loss": 0.4922,
+      "step": 2003
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00016293739228543467,
+      "loss": 0.5637,
+      "step": 2004
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00016282613684633765,
+      "loss": 0.5884,
+      "step": 2005
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00016271487429885976,
+      "loss": 0.6054,
+      "step": 2006
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.0001626036047046638,
+      "loss": 0.5353,
+      "step": 2007
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00016249232812541657,
+      "loss": 0.597,
+      "step": 2008
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00016238104462278877,
+      "loss": 0.6126,
+      "step": 2009
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00016226975425845486,
+      "loss": 0.534,
+      "step": 2010
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00016215845709409318,
+      "loss": 0.5377,
+      "step": 2011
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00016204715319138573,
+      "loss": 0.5654,
+      "step": 2012
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00016193584261201837,
+      "loss": 0.5163,
+      "step": 2013
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00016182452541768058,
+      "loss": 0.5636,
+      "step": 2014
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00016171320167006554,
+      "loss": 0.4149,
+      "step": 2015
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00016160187143087002,
+      "loss": 0.5924,
+      "step": 2016
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00016149053476179447,
+      "loss": 0.4811,
+      "step": 2017
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.0001613791917245428,
+      "loss": 0.5207,
+      "step": 2018
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00016126784238082253,
+      "loss": 0.5473,
+      "step": 2019
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00016115648679234466,
+      "loss": 0.6972,
+      "step": 2020
+    },
+    {
+      "epoch": 1.43,
+      "eval_loss": 1.0951440334320068,
+      "eval_runtime": 52.395,
+      "eval_samples_per_second": 11.127,
+      "eval_steps_per_second": 11.127,
+      "step": 2020
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.0001610451250208236,
+      "loss": 0.5587,
+      "step": 2021
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00016093375712797724,
+      "loss": 0.575,
+      "step": 2022
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00016082238317552682,
+      "loss": 0.6497,
+      "step": 2023
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00016071100322519703,
+      "loss": 0.6241,
+      "step": 2024
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00016059961733871572,
+      "loss": 0.5834,
+      "step": 2025
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00016048822557781424,
+      "loss": 0.5505,
+      "step": 2026
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.000160376828004227,
+      "loss": 0.5128,
+      "step": 2027
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.0001602654246796918,
+      "loss": 0.5076,
+      "step": 2028
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00016015401566594944,
+      "loss": 0.5559,
+      "step": 2029
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00016004260102474408,
+      "loss": 0.5607,
+      "step": 2030
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00015993118081782278,
+      "loss": 0.5148,
+      "step": 2031
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00015981975510693586,
+      "loss": 0.582,
+      "step": 2032
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00015970832395383665,
+      "loss": 0.7229,
+      "step": 2033
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00015959688742028146,
+      "loss": 0.5833,
+      "step": 2034
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00015948544556802956,
+      "loss": 0.5342,
+      "step": 2035
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00015937399845884325,
+      "loss": 0.4985,
+      "step": 2036
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00015926254615448768,
+      "loss": 0.606,
+      "step": 2037
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.0001591510887167309,
+      "loss": 0.5011,
+      "step": 2038
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00015903962620734383,
+      "loss": 0.5926,
+      "step": 2039
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00015892815868810012,
+      "loss": 0.6259,
+      "step": 2040
+    },
+    {
+      "epoch": 1.45,
+      "eval_loss": 1.097782850265503,
+      "eval_runtime": 50.7291,
+      "eval_samples_per_second": 11.492,
+      "eval_steps_per_second": 11.492,
+      "step": 2040
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00015881668622077632,
+      "loss": 0.4913,
+      "step": 2041
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.0001587052088671516,
+      "loss": 0.548,
+      "step": 2042
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00015859372668900786,
+      "loss": 0.5834,
+      "step": 2043
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00015848223974812973,
+      "loss": 0.5503,
+      "step": 2044
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00015837074810630454,
+      "loss": 0.6757,
+      "step": 2045
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00015825925182532202,
+      "loss": 0.4162,
+      "step": 2046
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00015814775096697467,
+      "loss": 0.582,
+      "step": 2047
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.0001580362455930574,
+      "loss": 0.5815,
+      "step": 2048
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00015792473576536767,
+      "loss": 0.5234,
+      "step": 2049
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00015781322154570545,
+      "loss": 0.4213,
+      "step": 2050
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00015770170299587305,
+      "loss": 0.5322,
+      "step": 2051
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00015759018017767528,
+      "loss": 0.5475,
+      "step": 2052
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00015747865315291927,
+      "loss": 0.5635,
+      "step": 2053
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.0001573671219834144,
+      "loss": 0.6708,
+      "step": 2054
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00015725558673097248,
+      "loss": 0.6777,
+      "step": 2055
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00015714404745740754,
+      "loss": 0.4521,
+      "step": 2056
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.0001570325042245358,
+      "loss": 0.7325,
+      "step": 2057
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00015692095709417573,
+      "loss": 0.5185,
+      "step": 2058
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00015680940612814795,
+      "loss": 0.5947,
+      "step": 2059
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00015669785138827511,
+      "loss": 0.6262,
+      "step": 2060
+    },
+    {
+      "epoch": 1.46,
+      "eval_loss": 1.0971267223358154,
+      "eval_runtime": 50.4652,
+      "eval_samples_per_second": 11.553,
+      "eval_steps_per_second": 11.553,
+      "step": 2060
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00015658629293638208,
+      "loss": 0.689,
+      "step": 2061
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.0001564747308342957,
+      "loss": 0.6049,
+      "step": 2062
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00015636316514384496,
+      "loss": 0.5391,
+      "step": 2063
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00015625159592686067,
+      "loss": 0.6408,
+      "step": 2064
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00015614002324517562,
+      "loss": 0.5653,
+      "step": 2065
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00015602844716062475,
+      "loss": 0.571,
+      "step": 2066
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.0001559168677350445,
+      "loss": 0.6625,
+      "step": 2067
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.00015580528503027354,
+      "loss": 0.5687,
+      "step": 2068
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.00015569369910815205,
+      "loss": 0.5806,
+      "step": 2069
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.0001555821100305222,
+      "loss": 0.5884,
+      "step": 2070
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.00015547051785922789,
+      "loss": 0.633,
+      "step": 2071
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.0001553589226561146,
+      "loss": 0.5046,
+      "step": 2072
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.00015524732448302957,
+      "loss": 0.5729,
+      "step": 2073
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.00015513572340182177,
+      "loss": 0.5805,
+      "step": 2074
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.00015502411947434158,
+      "loss": 0.6327,
+      "step": 2075
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.0001549125127624412,
+      "loss": 0.5049,
+      "step": 2076
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.0001548009033279742,
+      "loss": 0.6219,
+      "step": 2077
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.0001546892912327957,
+      "loss": 0.4992,
+      "step": 2078
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.00015457767653876238,
+      "loss": 0.6106,
+      "step": 2079
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.00015446605930773211,
+      "loss": 0.5518,
+      "step": 2080
+    },
+    {
+      "epoch": 1.47,
+      "eval_loss": 1.092513918876648,
+      "eval_runtime": 51.0398,
+      "eval_samples_per_second": 11.422,
+      "eval_steps_per_second": 11.422,
+      "step": 2080
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.0001543544396015646,
+      "loss": 0.5712,
+      "step": 2081
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00015424281748212043,
+      "loss": 0.7104,
+      "step": 2082
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.0001541311930112619,
+      "loss": 0.6247,
+      "step": 2083
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00015401956625085242,
+      "loss": 0.4652,
+      "step": 2084
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00015390793726275672,
+      "loss": 0.6412,
+      "step": 2085
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00015379630610884082,
+      "loss": 0.6073,
+      "step": 2086
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00015368467285097184,
+      "loss": 0.6806,
+      "step": 2087
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00015357303755101815,
+      "loss": 0.7072,
+      "step": 2088
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00015346140027084918,
+      "loss": 0.595,
+      "step": 2089
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00015334976107233552,
+      "loss": 0.6677,
+      "step": 2090
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00015323812001734877,
+      "loss": 0.508,
+      "step": 2091
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.0001531264771677616,
+      "loss": 0.5972,
+      "step": 2092
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.0001530148325854477,
+      "loss": 0.6861,
+      "step": 2093
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00015290318633228163,
+      "loss": 0.4871,
+      "step": 2094
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.0001527915384701389,
+      "loss": 0.6017,
+      "step": 2095
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.000152679889060896,
+      "loss": 0.5374,
+      "step": 2096
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00015256823816643013,
+      "loss": 0.6741,
+      "step": 2097
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00015245658584861954,
+      "loss": 0.5683,
+      "step": 2098
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00015234493216934308,
+      "loss": 0.561,
+      "step": 2099
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00015223327719048025,
+      "loss": 0.657,
+      "step": 2100
+    },
+    {
+      "epoch": 1.49,
+      "eval_loss": 1.0912736654281616,
+      "eval_runtime": 50.6193,
+      "eval_samples_per_second": 11.517,
+      "eval_steps_per_second": 11.517,
+      "step": 2100
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00015212162097391166,
+      "loss": 0.6054,
+      "step": 2101
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00015200996358151817,
+      "loss": 0.4673,
+      "step": 2102
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00015189830507518162,
+      "loss": 0.4795,
+      "step": 2103
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00015178664551678423,
+      "loss": 0.6208,
+      "step": 2104
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.000151674984968209,
+      "loss": 0.5579,
+      "step": 2105
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.0001515633234913393,
+      "loss": 0.5346,
+      "step": 2106
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00015145166114805927,
+      "loss": 0.6089,
+      "step": 2107
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00015133999800025317,
+      "loss": 0.623,
+      "step": 2108
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.00015122833410980603,
+      "loss": 0.5729,
+      "step": 2109
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.0001511166695386031,
+      "loss": 0.51,
+      "step": 2110
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.00015100500434853004,
+      "loss": 0.4772,
+      "step": 2111
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.00015089333860147296,
+      "loss": 0.571,
+      "step": 2112
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.00015078167235931816,
+      "loss": 0.5348,
+      "step": 2113
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.0001506700056839523,
+      "loss": 0.5215,
+      "step": 2114
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.0001505583386372621,
+      "loss": 0.4791,
+      "step": 2115
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.00015044667128113475,
+      "loss": 0.478,
+      "step": 2116
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.00015033500367745743,
+      "loss": 0.5658,
+      "step": 2117
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.00015022333588811756,
+      "loss": 0.6869,
+      "step": 2118
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.00015011166797500257,
+      "loss": 0.6755,
+      "step": 2119
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.00015,
+      "loss": 0.5311,
+      "step": 2120
+    },
+    {
+      "epoch": 1.5,
+      "eval_loss": 1.0915027856826782,
+      "eval_runtime": 51.0538,
+      "eval_samples_per_second": 11.419,
+      "eval_steps_per_second": 11.419,
+      "step": 2120
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.0001498883320249974,
+      "loss": 0.6028,
+      "step": 2121
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.00014977666411188247,
+      "loss": 0.57,
+      "step": 2122
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.00014966499632254254,
+      "loss": 0.6557,
+      "step": 2123
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.00014955332871886522,
+      "loss": 0.5909,
+      "step": 2124
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.00014944166136273787,
+      "loss": 0.5541,
+      "step": 2125
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.00014932999431604776,
+      "loss": 0.5245,
+      "step": 2126
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.00014921832764068182,
+      "loss": 0.4997,
+      "step": 2127
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.00014910666139852702,
+      "loss": 0.6478,
+      "step": 2128
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.0001489949956514699,
+      "loss": 0.5834,
+      "step": 2129
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.00014888333046139693,
+      "loss": 0.613,
+      "step": 2130
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.000148771665890194,
+      "loss": 0.5782,
+      "step": 2131
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.0001486600019997468,
+      "loss": 0.6888,
+      "step": 2132
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.00014854833885194078,
+      "loss": 0.4989,
+      "step": 2133
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.00014843667650866067,
+      "loss": 0.543,
+      "step": 2134
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.000148325015031791,
+      "loss": 0.4593,
+      "step": 2135
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.00014821335448321574,
+      "loss": 0.5929,
+      "step": 2136
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.0001481016949248184,
+      "loss": 0.5119,
+      "step": 2137
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.00014799003641848183,
+      "loss": 0.5225,
+      "step": 2138
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.00014787837902608834,
+      "loss": 0.5504,
+      "step": 2139
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.00014776672280951967,
+      "loss": 0.6351,
+      "step": 2140
+    },
+    {
+      "epoch": 1.52,
+      "eval_loss": 1.0936156511306763,
+      "eval_runtime": 55.2228,
+      "eval_samples_per_second": 10.557,
+      "eval_steps_per_second": 10.557,
+      "step": 2140
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.00014765506783065698,
+      "loss": 0.4782,
+      "step": 2141
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.00014754341415138043,
+      "loss": 0.5619,
+      "step": 2142
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.0001474317618335698,
+      "loss": 0.6615,
+      "step": 2143
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.00014732011093910403,
+      "loss": 0.5435,
+      "step": 2144
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.00014720846152986109,
+      "loss": 0.5615,
+      "step": 2145
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.0001470968136677184,
+      "loss": 0.5617,
+      "step": 2146
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.00014698516741455228,
+      "loss": 0.4507,
+      "step": 2147
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.0001468735228322384,
+      "loss": 0.4937,
+      "step": 2148
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.00014676187998265123,
+      "loss": 0.6191,
+      "step": 2149
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.00014665023892766445,
+      "loss": 0.7072,
+      "step": 2150
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.00014653859972915077,
+      "loss": 0.5256,
+      "step": 2151
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.00014642696244898188,
+      "loss": 0.5659,
+      "step": 2152
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.00014631532714902816,
+      "loss": 0.6362,
+      "step": 2153
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.00014620369389115918,
+      "loss": 0.6233,
+      "step": 2154
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.00014609206273724328,
+      "loss": 0.4732,
+      "step": 2155
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.00014598043374914758,
+      "loss": 0.4775,
+      "step": 2156
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.0001458688069887381,
+      "loss": 0.5827,
+      "step": 2157
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.00014575718251787955,
+      "loss": 0.5211,
+      "step": 2158
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.00014564556039843547,
+      "loss": 0.6304,
+      "step": 2159
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.00014553394069226786,
+      "loss": 0.7161,
+      "step": 2160
+    },
+    {
+      "epoch": 1.53,
+      "eval_loss": 1.0925236940383911,
+      "eval_runtime": 48.1149,
+      "eval_samples_per_second": 12.117,
+      "eval_steps_per_second": 12.117,
+      "step": 2160
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.00014542232346123764,
+      "loss": 0.5661,
+      "step": 2161
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.00014531070876720425,
+      "loss": 0.5967,
+      "step": 2162
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.00014519909667202582,
+      "loss": 0.6022,
+      "step": 2163
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.0001450874872375588,
+      "loss": 0.5653,
+      "step": 2164
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.0001449758805256584,
+      "loss": 0.6022,
+      "step": 2165
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.00014486427659817828,
+      "loss": 0.5259,
+      "step": 2166
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.00014475267551697043,
+      "loss": 0.5766,
+      "step": 2167
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.0001446410773438854,
+      "loss": 0.4481,
+      "step": 2168
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.0001445294821407721,
+      "loss": 0.5804,
+      "step": 2169
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.0001444178899694778,
+      "loss": 0.6877,
+      "step": 2170
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.00014430630089184795,
+      "loss": 0.6925,
+      "step": 2171
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.00014419471496972648,
+      "loss": 0.5714,
+      "step": 2172
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.00014408313226495545,
+      "loss": 0.5761,
+      "step": 2173
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.0001439715528393753,
+      "loss": 0.5961,
+      "step": 2174
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.00014385997675482435,
+      "loss": 0.5724,
+      "step": 2175
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.00014374840407313934,
+      "loss": 0.6372,
+      "step": 2176
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.00014363683485615506,
+      "loss": 0.5565,
+      "step": 2177
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.00014352526916570427,
+      "loss": 0.502,
+      "step": 2178
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.00014341370706361792,
+      "loss": 0.5722,
+      "step": 2179
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.00014330214861172486,
+      "loss": 0.5684,
+      "step": 2180
+    },
+    {
+      "epoch": 1.55,
+      "eval_loss": 1.0920584201812744,
+      "eval_runtime": 50.6267,
+      "eval_samples_per_second": 11.516,
+      "eval_steps_per_second": 11.516,
+      "step": 2180
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.00014319059387185207,
+      "loss": 0.5365,
+      "step": 2181
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.00014307904290582424,
+      "loss": 0.5212,
+      "step": 2182
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.00014296749577546417,
+      "loss": 0.5159,
+      "step": 2183
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.00014285595254259244,
+      "loss": 0.7308,
+      "step": 2184
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.00014274441326902752,
+      "loss": 0.4806,
+      "step": 2185
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.00014263287801658558,
+      "loss": 0.483,
+      "step": 2186
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.00014252134684708073,
+      "loss": 0.5325,
+      "step": 2187
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.0001424098198223247,
+      "loss": 0.6856,
+      "step": 2188
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.00014229829700412693,
+      "loss": 0.4694,
+      "step": 2189
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.00014218677845429453,
+      "loss": 0.4959,
+      "step": 2190
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.00014207526423463227,
+      "loss": 0.5577,
+      "step": 2191
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.00014196375440694262,
+      "loss": 0.4825,
+      "step": 2192
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.00014185224903302533,
+      "loss": 0.5109,
+      "step": 2193
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.00014174074817467795,
+      "loss": 0.5796,
+      "step": 2194
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.00014162925189369543,
+      "loss": 0.6119,
+      "step": 2195
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.00014151776025187024,
+      "loss": 0.5724,
+      "step": 2196
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.00014140627331099212,
+      "loss": 0.6852,
+      "step": 2197
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.0001412947911328484,
+      "loss": 0.5696,
+      "step": 2198
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.00014118331377922365,
+      "loss": 0.5574,
+      "step": 2199
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.00014107184131189988,
+      "loss": 0.4888,
+      "step": 2200
+    },
+    {
+      "epoch": 1.56,
+      "eval_loss": 1.0924915075302124,
+      "eval_runtime": 51.2839,
+      "eval_samples_per_second": 11.368,
+      "eval_steps_per_second": 11.368,
+      "step": 2200
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.00014096037379265615,
+      "loss": 0.4965,
+      "step": 2201
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.00014084891128326906,
+      "loss": 0.4906,
+      "step": 2202
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.00014073745384551235,
+      "loss": 0.6074,
+      "step": 2203
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.00014062600154115677,
+      "loss": 0.5852,
+      "step": 2204
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.00014051455443197042,
+      "loss": 0.6424,
+      "step": 2205
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.00014040311257971854,
+      "loss": 0.6137,
+      "step": 2206
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.00014029167604616335,
+      "loss": 0.5846,
+      "step": 2207
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.0001401802448930641,
+      "loss": 0.5233,
+      "step": 2208
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.0001400688191821772,
+      "loss": 0.6335,
+      "step": 2209
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.0001399573989752559,
+      "loss": 0.6409,
+      "step": 2210
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.00013984598433405056,
+      "loss": 0.7066,
+      "step": 2211
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.00013973457532030819,
+      "loss": 0.6514,
+      "step": 2212
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.00013962317199577294,
+      "loss": 0.5721,
+      "step": 2213
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.0001395117744221858,
+      "loss": 0.5004,
+      "step": 2214
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.00013940038266128428,
+      "loss": 0.6371,
+      "step": 2215
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.000139288996774803,
+      "loss": 0.6272,
+      "step": 2216
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.00013917761682447316,
+      "loss": 0.6745,
+      "step": 2217
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.00013906624287202279,
+      "loss": 0.6342,
+      "step": 2218
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.0001389548749791764,
+      "loss": 0.6097,
+      "step": 2219
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.00013884351320765534,
+      "loss": 0.5942,
+      "step": 2220
+    },
+    {
+      "epoch": 1.57,
+      "eval_loss": 1.0885359048843384,
+      "eval_runtime": 56.2279,
+      "eval_samples_per_second": 10.369,
+      "eval_steps_per_second": 10.369,
+      "step": 2220
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.00013873215761917745,
+      "loss": 0.583,
+      "step": 2221
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.0001386208082754572,
+      "loss": 0.7287,
+      "step": 2222
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.00013850946523820553,
+      "loss": 0.5667,
+      "step": 2223
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.00013839812856912995,
+      "loss": 0.4981,
+      "step": 2224
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.0001382867983299345,
+      "loss": 0.667,
+      "step": 2225
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.00013817547458231942,
+      "loss": 0.5836,
+      "step": 2226
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.0001380641573879816,
+      "loss": 0.5792,
+      "step": 2227
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.00013795284680861427,
+      "loss": 0.5986,
+      "step": 2228
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.00013784154290590685,
+      "loss": 0.545,
+      "step": 2229
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.00013773024574154514,
+      "loss": 0.6066,
+      "step": 2230
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.0001376189553772112,
+      "loss": 0.6131,
+      "step": 2231
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.0001375076718745834,
+      "loss": 0.6862,
+      "step": 2232
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.00013739639529533623,
+      "loss": 0.5089,
+      "step": 2233
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.00013728512570114026,
+      "loss": 0.6343,
+      "step": 2234
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.00013717386315366232,
+      "loss": 0.583,
+      "step": 2235
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.00013706260771456533,
+      "loss": 0.5676,
+      "step": 2236
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.00013695135944550817,
+      "loss": 0.7298,
+      "step": 2237
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.00013684011840814586,
+      "loss": 0.5061,
+      "step": 2238
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.00013672888466412933,
+      "loss": 0.476,
+      "step": 2239
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.0001366176582751056,
+      "loss": 0.6184,
+      "step": 2240
+    },
+    {
+      "epoch": 1.59,
+      "eval_loss": 1.0886327028274536,
+      "eval_runtime": 47.7451,
+      "eval_samples_per_second": 12.211,
+      "eval_steps_per_second": 12.211,
+      "step": 2240
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.00013650643930271736,
+      "loss": 0.5287,
+      "step": 2241
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.0001363952278086034,
+      "loss": 0.6542,
+      "step": 2242
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.00013628402385439833,
+      "loss": 0.582,
+      "step": 2243
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.0001361728275017326,
+      "loss": 0.5396,
+      "step": 2244
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.00013606163881223232,
+      "loss": 0.5974,
+      "step": 2245
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.00013595045784751944,
+      "loss": 0.6678,
+      "step": 2246
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.00013583928466921172,
+      "loss": 0.6073,
+      "step": 2247
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.00013572811933892242,
+      "loss": 0.6415,
+      "step": 2248
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.0001356169619182606,
+      "loss": 0.5761,
+      "step": 2249
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.00013550581246883083,
+      "loss": 0.6465,
+      "step": 2250
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.00013539467105223342,
+      "loss": 0.5493,
+      "step": 2251
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.00013528353773006397,
+      "loss": 0.6343,
+      "step": 2252
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.00013517241256391383,
+      "loss": 0.5872,
+      "step": 2253
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.00013506129561536973,
+      "loss": 0.5744,
+      "step": 2254
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.00013495018694601395,
+      "loss": 0.5573,
+      "step": 2255
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.00013483908661742391,
+      "loss": 0.5185,
+      "step": 2256
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.0001347279946911727,
+      "loss": 0.6325,
+      "step": 2257
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.00013461691122882864,
+      "loss": 0.4909,
+      "step": 2258
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.00013450583629195532,
+      "loss": 0.504,
+      "step": 2259
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.0001343947699421117,
+      "loss": 0.5976,
+      "step": 2260
+    },
+    {
+      "epoch": 1.6,
+      "eval_loss": 1.090318202972412,
+      "eval_runtime": 48.7558,
+      "eval_samples_per_second": 11.958,
+      "eval_steps_per_second": 11.958,
+      "step": 2260
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.0001342837122408519,
+      "loss": 0.5795,
+      "step": 2261
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.00013417266324972533,
+      "loss": 0.6043,
+      "step": 2262
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.0001340616230302764,
+      "loss": 0.5478,
+      "step": 2263
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.00013395059164404482,
+      "loss": 0.5222,
+      "step": 2264
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.0001338395691525654,
+      "loss": 0.5544,
+      "step": 2265
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.00013372855561736798,
+      "loss": 0.6309,
+      "step": 2266
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.00013361755109997744,
+      "loss": 0.6355,
+      "step": 2267
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.0001335065556619136,
+      "loss": 0.5523,
+      "step": 2268
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.00013339556936469137,
+      "loss": 0.5338,
+      "step": 2269
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.0001332845922698205,
+      "loss": 0.5503,
+      "step": 2270
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.00013317362443880567,
+      "loss": 0.6459,
+      "step": 2271
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.00013306266593314642,
+      "loss": 0.5538,
+      "step": 2272
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.00013295171681433724,
+      "loss": 0.5166,
+      "step": 2273
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.00013284077714386712,
+      "loss": 0.6625,
+      "step": 2274
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.0001327298469832201,
+      "loss": 0.5553,
+      "step": 2275
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.00013261892639387484,
+      "loss": 0.6576,
+      "step": 2276
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.00013250801543730477,
+      "loss": 0.58,
+      "step": 2277
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.00013239711417497782,
+      "loss": 0.5887,
+      "step": 2278
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.00013228622266835667,
+      "loss": 0.635,
+      "step": 2279
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.00013217534097889863,
+      "loss": 0.5321,
+      "step": 2280
+    },
+    {
+      "epoch": 1.62,
+      "eval_loss": 1.0896165370941162,
+      "eval_runtime": 49.3179,
+      "eval_samples_per_second": 11.821,
+      "eval_steps_per_second": 11.821,
+      "step": 2280
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.00013206446916805541,
+      "loss": 0.61,
+      "step": 2281
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.0001319536072972734,
+      "loss": 0.7075,
+      "step": 2282
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.0001318427554279934,
+      "loss": 0.5821,
+      "step": 2283
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.00013173191362165078,
+      "loss": 0.6198,
+      "step": 2284
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.00013162108193967508,
+      "loss": 0.7,
+      "step": 2285
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.0001315102604434905,
+      "loss": 0.6471,
+      "step": 2286
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.00013139944919451544,
+      "loss": 0.5605,
+      "step": 2287
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.00013128864825416275,
+      "loss": 0.6035,
+      "step": 2288
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.00013117785768383938,
+      "loss": 0.4734,
+      "step": 2289
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.00013106707754494668,
+      "loss": 0.553,
+      "step": 2290
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.00013095630789888027,
+      "loss": 0.6728,
+      "step": 2291
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.0001308455488070297,
+      "loss": 0.5252,
+      "step": 2292
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.00013073480033077892,
+      "loss": 0.5425,
+      "step": 2293
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.00013062406253150588,
+      "loss": 0.5428,
+      "step": 2294
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.00013051333547058273,
+      "loss": 0.6436,
+      "step": 2295
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.00013040261920937543,
+      "loss": 0.4633,
+      "step": 2296
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.0001302919138092442,
+      "loss": 0.4627,
+      "step": 2297
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.0001301812193315431,
+      "loss": 0.6299,
+      "step": 2298
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.00013007053583762027,
+      "loss": 0.6086,
+      "step": 2299
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.00012995986338881757,
+      "loss": 0.5665,
+      "step": 2300
+    },
+    {
+      "epoch": 1.63,
+      "eval_loss": 1.0861297845840454,
+      "eval_runtime": 51.6677,
+      "eval_samples_per_second": 11.284,
+      "eval_steps_per_second": 11.284,
+      "step": 2300
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.00012984920204647089,
+      "loss": 0.5121,
+      "step": 2301
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.00012973855187190997,
+      "loss": 0.6527,
+      "step": 2302
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.00012962791292645822,
+      "loss": 0.4763,
+      "step": 2303
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.00012951728527143294,
+      "loss": 0.5388,
+      "step": 2304
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.00012940666896814516,
+      "loss": 0.6602,
+      "step": 2305
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.00012929606407789963,
+      "loss": 0.6038,
+      "step": 2306
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.00012918547066199473,
+      "loss": 0.6619,
+      "step": 2307
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.00012907488878172247,
+      "loss": 0.6115,
+      "step": 2308
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.00012896431849836854,
+      "loss": 0.6404,
+      "step": 2309
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.00012885375987321219,
+      "loss": 0.6754,
+      "step": 2310
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.0001287432129675261,
+      "loss": 0.5389,
+      "step": 2311
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.00012863267784257658,
+      "loss": 0.5281,
+      "step": 2312
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.00012852215455962346,
+      "loss": 0.5566,
+      "step": 2313
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.00012841164317991974,
+      "loss": 0.6073,
+      "step": 2314
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.00012830114376471207,
+      "loss": 0.6688,
+      "step": 2315
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.00012819065637524043,
+      "loss": 0.4226,
+      "step": 2316
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.00012808018107273808,
+      "loss": 0.5567,
+      "step": 2317
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.0001279697179184316,
+      "loss": 0.6348,
+      "step": 2318
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.00012785926697354082,
+      "loss": 0.6379,
+      "step": 2319
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.00012774882829927884,
+      "loss": 0.5817,
+      "step": 2320
+    },
+    {
+      "epoch": 1.64,
+      "eval_loss": 1.0898795127868652,
+      "eval_runtime": 50.8832,
+      "eval_samples_per_second": 11.458,
+      "eval_steps_per_second": 11.458,
+      "step": 2320
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 0.00012763840195685204,
+      "loss": 0.5912,
+      "step": 2321
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 0.0001275279880074597,
+      "loss": 0.6776,
+      "step": 2322
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 0.00012741758651229446,
+      "loss": 0.4266,
+      "step": 2323
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 0.00012730719753254206,
+      "loss": 0.6574,
+      "step": 2324
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 0.00012719682112938123,
+      "loss": 0.5025,
+      "step": 2325
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 0.00012708645736398365,
+      "loss": 0.6445,
+      "step": 2326
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 0.00012697610629751417,
+      "loss": 0.5498,
+      "step": 2327
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 0.0001268657679911305,
+      "loss": 0.5472,
+      "step": 2328
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 0.00012675544250598331,
+      "loss": 0.6193,
+      "step": 2329
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 0.0001266451299032161,
+      "loss": 0.5417,
+      "step": 2330
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 0.00012653483024396533,
+      "loss": 0.7493,
+      "step": 2331
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 0.0001264245435893603,
+      "loss": 0.6265,
+      "step": 2332
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 0.00012631427000052293,
+      "loss": 0.589,
+      "step": 2333
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 0.00012620400953856803,
+      "loss": 0.616,
+      "step": 2334
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 0.00012609376226460314,
+      "loss": 0.4799,
+      "step": 2335
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 0.00012598352823972857,
+      "loss": 0.5797,
+      "step": 2336
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 0.000125873307525037,
+      "loss": 0.5894,
+      "step": 2337
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 0.000125763100181614,
+      "loss": 0.6629,
+      "step": 2338
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 0.00012565290627053767,
+      "loss": 0.533,
+      "step": 2339
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 0.0001255427258528786,
+      "loss": 0.4925,
+      "step": 2340
+    },
+    {
+      "epoch": 1.66,
+      "eval_loss": 1.089101791381836,
+      "eval_runtime": 51.3987,
+      "eval_samples_per_second": 11.343,
+      "eval_steps_per_second": 11.343,
+      "step": 2340
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 0.00012543255898969996,
+      "loss": 0.6251,
+      "step": 2341
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 0.00012532240574205741,
+      "loss": 0.6166,
+      "step": 2342
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 0.00012521226617099912,
+      "loss": 0.7698,
+      "step": 2343
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 0.00012510214033756542,
+      "loss": 0.5205,
+      "step": 2344
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 0.00012499202830278935,
+      "loss": 0.5594,
+      "step": 2345
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 0.00012488193012769608,
+      "loss": 0.6698,
+      "step": 2346
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 0.00012477184587330337,
+      "loss": 0.5754,
+      "step": 2347
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 0.00012466177560062084,
+      "loss": 0.5642,
+      "step": 2348
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 0.00012455171937065073,
+      "loss": 0.6314,
+      "step": 2349
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 0.00012444167724438737,
+      "loss": 0.6313,
+      "step": 2350
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 0.00012433164928281722,
+      "loss": 0.6279,
+      "step": 2351
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 0.00012422163554691898,
+      "loss": 0.5894,
+      "step": 2352
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 0.0001241116360976634,
+      "loss": 0.4346,
+      "step": 2353
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 0.00012400165099601346,
+      "loss": 0.5857,
+      "step": 2354
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 0.00012389168030292388,
+      "loss": 0.6526,
+      "step": 2355
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 0.00012378172407934168,
+      "loss": 0.5402,
+      "step": 2356
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 0.00012367178238620576,
+      "loss": 0.5201,
+      "step": 2357
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 0.00012356185528444698,
+      "loss": 0.6165,
+      "step": 2358
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 0.00012345194283498807,
+      "loss": 0.5454,
+      "step": 2359
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 0.00012334204509874372,
+      "loss": 0.5957,
+      "step": 2360
+    },
+    {
+      "epoch": 1.67,
+      "eval_loss": 1.0900988578796387,
+      "eval_runtime": 51.8041,
+      "eval_samples_per_second": 11.254,
+      "eval_steps_per_second": 11.254,
+      "step": 2360
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 0.00012323216213662043,
+      "loss": 0.6698,
+      "step": 2361
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 0.0001231222940095164,
+      "loss": 0.525,
+      "step": 2362
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 0.00012301244077832178,
+      "loss": 0.5949,
+      "step": 2363
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 0.0001229026025039184,
+      "loss": 0.5781,
+      "step": 2364
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 0.00012279277924717984,
+      "loss": 0.6516,
+      "step": 2365
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 0.0001226829710689712,
+      "loss": 0.6114,
+      "step": 2366
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 0.0001225731780301494,
+      "loss": 0.6341,
+      "step": 2367
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 0.00012246340019156292,
+      "loss": 0.5788,
+      "step": 2368
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 0.0001223536376140518,
+      "loss": 0.595,
+      "step": 2369
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 0.00012224389035844755,
+      "loss": 0.5634,
+      "step": 2370
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 0.00012213415848557333,
+      "loss": 0.5219,
+      "step": 2371
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 0.00012202444205624378,
+      "loss": 0.5452,
+      "step": 2372
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 0.00012191474113126472,
+      "loss": 0.6312,
+      "step": 2373
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 0.00012180505577143369,
+      "loss": 0.66,
+      "step": 2374
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 0.00012169538603753944,
+      "loss": 0.644,
+      "step": 2375
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 0.00012158573199036217,
+      "loss": 0.6053,
+      "step": 2376
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.00012147609369067321,
+      "loss": 0.5756,
+      "step": 2377
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.00012136647119923531,
+      "loss": 0.7237,
+      "step": 2378
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.00012125686457680241,
+      "loss": 0.5448,
+      "step": 2379
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.00012114727388411971,
+      "loss": 0.5024,
+      "step": 2380
+    },
+    {
+      "epoch": 1.69,
+      "eval_loss": 1.0917657613754272,
+      "eval_runtime": 55.0659,
+      "eval_samples_per_second": 10.587,
+      "eval_steps_per_second": 10.587,
+      "step": 2380
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.00012103769918192347,
+      "loss": 0.5847,
+      "step": 2381
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.00012092814053094116,
+      "loss": 0.552,
+      "step": 2382
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.00012081859799189143,
+      "loss": 0.642,
+      "step": 2383
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.00012070907162548378,
+      "loss": 0.427,
+      "step": 2384
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.00012059956149241895,
+      "loss": 0.585,
+      "step": 2385
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.00012049006765338862,
+      "loss": 0.6485,
+      "step": 2386
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.0001203805901690755,
+      "loss": 0.5941,
+      "step": 2387
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.00012027112910015301,
+      "loss": 0.533,
+      "step": 2388
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.00012016168450728575,
+      "loss": 0.5106,
+      "step": 2389
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.00012005225645112903,
+      "loss": 0.555,
+      "step": 2390
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.0001199428449923291,
+      "loss": 0.5545,
+      "step": 2391
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.00011983345019152288,
+      "loss": 0.571,
+      "step": 2392
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.00011972407210933815,
+      "loss": 0.612,
+      "step": 2393
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.00011961471080639346,
+      "loss": 0.676,
+      "step": 2394
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.0001195053663432979,
+      "loss": 0.5454,
+      "step": 2395
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.00011939603878065136,
+      "loss": 0.6757,
+      "step": 2396
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.00011928672817904436,
+      "loss": 0.6377,
+      "step": 2397
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.00011917743459905801,
+      "loss": 0.5441,
+      "step": 2398
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.00011906815810126391,
+      "loss": 0.658,
+      "step": 2399
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.0001189588987462243,
+      "loss": 0.5907,
+      "step": 2400
+    },
+    {
+      "epoch": 1.7,
+      "eval_loss": 1.0943603515625,
+      "eval_runtime": 51.1033,
+      "eval_samples_per_second": 11.408,
+      "eval_steps_per_second": 11.408,
+      "step": 2400
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.00011884965659449186,
+      "loss": 0.5607,
+      "step": 2401
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.00011874043170660981,
+      "loss": 0.561,
+      "step": 2402
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.0001186312241431116,
+      "loss": 0.5595,
+      "step": 2403
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.00011852203396452131,
+      "loss": 0.514,
+      "step": 2404
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 0.00011841286123135335,
+      "loss": 0.5128,
+      "step": 2405
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 0.00011830370600411226,
+      "loss": 0.5066,
+      "step": 2406
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 0.0001181945683432931,
+      "loss": 0.5197,
+      "step": 2407
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 0.00011808544830938112,
+      "loss": 0.4514,
+      "step": 2408
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 0.00011797634596285183,
+      "loss": 0.6285,
+      "step": 2409
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 0.00011786726136417081,
+      "loss": 0.5271,
+      "step": 2410
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 0.00011775819457379401,
+      "loss": 0.472,
+      "step": 2411
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 0.00011764914565216731,
+      "loss": 0.5634,
+      "step": 2412
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 0.00011754011465972692,
+      "loss": 0.5489,
+      "step": 2413
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 0.00011743110165689878,
+      "loss": 0.5632,
+      "step": 2414
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 0.00011732210670409916,
+      "loss": 0.4997,
+      "step": 2415
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 0.00011721312986173426,
+      "loss": 0.5725,
+      "step": 2416
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 0.00011710417119020006,
+      "loss": 0.634,
+      "step": 2417
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 0.00011699523074988269,
+      "loss": 0.6309,
+      "step": 2418
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 0.00011688630860115813,
+      "loss": 0.6044,
+      "step": 2419
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.00011677740480439215,
+      "loss": 0.6453,
+      "step": 2420
+    },
+    {
+      "epoch": 1.72,
+      "eval_loss": 1.0926622152328491,
+      "eval_runtime": 53.2806,
+      "eval_samples_per_second": 10.942,
+      "eval_steps_per_second": 10.942,
+      "step": 2420
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.00011666851941994037,
+      "loss": 0.5899,
+      "step": 2421
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.00011655965250814827,
+      "loss": 0.5627,
+      "step": 2422
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.00011645080412935101,
+      "loss": 0.4912,
+      "step": 2423
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.00011634197434387358,
+      "loss": 0.582,
+      "step": 2424
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.00011623316321203049,
+      "loss": 0.54,
+      "step": 2425
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.00011612437079412609,
+      "loss": 0.4949,
+      "step": 2426
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.00011601559715045434,
+      "loss": 0.4921,
+      "step": 2427
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.00011590684234129861,
+      "loss": 0.6616,
+      "step": 2428
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.00011579810642693205,
+      "loss": 0.645,
+      "step": 2429
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.00011568938946761725,
+      "loss": 0.5922,
+      "step": 2430
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.00011558069152360632,
+      "loss": 0.6564,
+      "step": 2431
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.00011547201265514079,
+      "loss": 0.5617,
+      "step": 2432
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.00011536335292245163,
+      "loss": 0.6692,
+      "step": 2433
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 0.0001152547123857592,
+      "loss": 0.6503,
+      "step": 2434
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 0.00011514609110527336,
+      "loss": 0.4816,
+      "step": 2435
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 0.00011503748914119301,
+      "loss": 0.7561,
+      "step": 2436
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 0.00011492890655370657,
+      "loss": 0.6644,
+      "step": 2437
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 0.0001148203434029917,
+      "loss": 0.5531,
+      "step": 2438
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 0.00011471179974921517,
+      "loss": 0.6196,
+      "step": 2439
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 0.00011460327565253308,
+      "loss": 0.5664,
+      "step": 2440
+    },
+    {
+      "epoch": 1.73,
+      "eval_loss": 1.088532567024231,
+      "eval_runtime": 55.0835,
+      "eval_samples_per_second": 10.584,
+      "eval_steps_per_second": 10.584,
+      "step": 2440
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 0.00011449477117309061,
+      "loss": 0.7278,
+      "step": 2441
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 0.00011438628637102219,
+      "loss": 0.643,
+      "step": 2442
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 0.00011427782130645109,
+      "loss": 0.525,
+      "step": 2443
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 0.00011416937603948987,
+      "loss": 0.6626,
+      "step": 2444
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 0.00011406095063024004,
+      "loss": 0.6215,
+      "step": 2445
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 0.00011395254513879219,
+      "loss": 0.6861,
+      "step": 2446
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 0.00011384415962522562,
+      "loss": 0.6001,
+      "step": 2447
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.00011373579414960886,
+      "loss": 0.5316,
+      "step": 2448
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.00011362744877199919,
+      "loss": 0.561,
+      "step": 2449
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.00011351912355244271,
+      "loss": 0.5556,
+      "step": 2450
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.00011341081855097445,
+      "loss": 0.5351,
+      "step": 2451
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.00011330253382761818,
+      "loss": 0.5942,
+      "step": 2452
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.00011319426944238652,
+      "loss": 0.4958,
+      "step": 2453
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.00011308602545528058,
+      "loss": 0.4906,
+      "step": 2454
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.00011297780192629043,
+      "loss": 0.6165,
+      "step": 2455
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.00011286959891539466,
+      "loss": 0.5917,
+      "step": 2456
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.00011276141648256057,
+      "loss": 0.5693,
+      "step": 2457
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.00011265325468774395,
+      "loss": 0.7651,
+      "step": 2458
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.00011254511359088923,
+      "loss": 0.5964,
+      "step": 2459
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.00011243699325192936,
+      "loss": 0.6266,
+      "step": 2460
+    },
+    {
+      "epoch": 1.74,
+      "eval_loss": 1.0884407758712769,
+      "eval_runtime": 51.8913,
+      "eval_samples_per_second": 11.235,
+      "eval_steps_per_second": 11.235,
+      "step": 2460
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.0001123288937307858,
+      "loss": 0.4605,
+      "step": 2461
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 0.00011222081508736842,
+      "loss": 0.5453,
+      "step": 2462
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 0.00011211275738157553,
+      "loss": 0.5034,
+      "step": 2463
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 0.00011200472067329398,
+      "loss": 0.4922,
+      "step": 2464
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 0.00011189670502239866,
+      "loss": 0.525,
+      "step": 2465
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 0.00011178871048875312,
+      "loss": 0.5778,
+      "step": 2466
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 0.00011168073713220902,
+      "loss": 0.6312,
+      "step": 2467
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 0.00011157278501260645,
+      "loss": 0.6024,
+      "step": 2468
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 0.00011146485418977344,
+      "loss": 0.4884,
+      "step": 2469
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 0.00011135694472352651,
+      "loss": 0.6132,
+      "step": 2470
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 0.00011124905667367021,
+      "loss": 0.5999,
+      "step": 2471
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 0.00011114119009999724,
+      "loss": 0.6011,
+      "step": 2472
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 0.00011103334506228839,
+      "loss": 0.5648,
+      "step": 2473
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 0.00011092552162031251,
+      "loss": 0.5677,
+      "step": 2474
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 0.00011081771983382657,
+      "loss": 0.5524,
+      "step": 2475
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.00011070993976257533,
+      "loss": 0.6795,
+      "step": 2476
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.00011060218146629174,
+      "loss": 0.6318,
+      "step": 2477
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.00011049444500469656,
+      "loss": 0.5208,
+      "step": 2478
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.00011038673043749851,
+      "loss": 0.6282,
+      "step": 2479
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.00011027903782439411,
+      "loss": 0.5673,
+      "step": 2480
+    },
+    {
+      "epoch": 1.76,
+      "eval_loss": 1.0843061208724976,
+      "eval_runtime": 50.2145,
+      "eval_samples_per_second": 11.61,
+      "eval_steps_per_second": 11.61,
+      "step": 2480
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.00011017136722506779,
+      "loss": 0.5836,
+      "step": 2481
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.00011006371869919168,
+      "loss": 0.6031,
+      "step": 2482
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.00010995609230642574,
+      "loss": 0.5097,
+      "step": 2483
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.0001098484881064177,
+      "loss": 0.5237,
+      "step": 2484
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.00010974090615880289,
+      "loss": 0.45,
+      "step": 2485
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.00010963334652320448,
+      "loss": 0.4939,
+      "step": 2486
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.00010952580925923301,
+      "loss": 0.6214,
+      "step": 2487
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.00010941829442648684,
+      "loss": 0.6141,
+      "step": 2488
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.00010931080208455182,
+      "loss": 0.5224,
+      "step": 2489
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.00010920333229300139,
+      "loss": 0.5776,
+      "step": 2490
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.00010909588511139638,
+      "loss": 0.6057,
+      "step": 2491
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.00010898846059928519,
+      "loss": 0.6543,
+      "step": 2492
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.00010888105881620362,
+      "loss": 0.5495,
+      "step": 2493
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.00010877367982167494,
+      "loss": 0.5835,
+      "step": 2494
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.00010866632367520959,
+      "loss": 0.5649,
+      "step": 2495
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.00010855899043630555,
+      "loss": 0.5829,
+      "step": 2496
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.0001084516801644481,
+      "loss": 0.638,
+      "step": 2497
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.00010834439291910963,
+      "loss": 0.5075,
+      "step": 2498
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.00010823712875974992,
+      "loss": 0.5943,
+      "step": 2499
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.00010812988774581587,
+      "loss": 0.6737,
+      "step": 2500
+    },
+    {
+      "epoch": 1.77,
+      "eval_loss": 1.0877330303192139,
+      "eval_runtime": 49.2766,
+      "eval_samples_per_second": 11.831,
+      "eval_steps_per_second": 11.831,
+      "step": 2500
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.00010802266993674165,
+      "loss": 0.7031,
+      "step": 2501
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.00010791547539194844,
+      "loss": 0.7329,
+      "step": 2502
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.00010780830417084461,
+      "loss": 0.5739,
+      "step": 2503
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.00010770115633282561,
+      "loss": 0.5045,
+      "step": 2504
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.00010759403193727395,
+      "loss": 0.6235,
+      "step": 2505
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.00010748693104355896,
+      "loss": 0.6757,
+      "step": 2506
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.00010737985371103721,
+      "loss": 0.67,
+      "step": 2507
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.0001072727999990521,
+      "loss": 0.4851,
+      "step": 2508
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.00010716576996693382,
+      "loss": 0.6236,
+      "step": 2509
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.00010705876367399965,
+      "loss": 0.6898,
+      "step": 2510
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.00010695178117955355,
+      "loss": 0.5318,
+      "step": 2511
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.0001068448225428864,
+      "loss": 0.4862,
+      "step": 2512
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.0001067378878232758,
+      "loss": 0.6183,
+      "step": 2513
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.00010663097707998608,
+      "loss": 0.5285,
+      "step": 2514
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.00010652409037226832,
+      "loss": 0.5133,
+      "step": 2515
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.00010641722775936032,
+      "loss": 0.621,
+      "step": 2516
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.00010631038930048636,
+      "loss": 0.6636,
+      "step": 2517
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 0.0001062035750548575,
+      "loss": 0.5726,
+      "step": 2518
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 0.00010609678508167138,
+      "loss": 0.5659,
+      "step": 2519
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 0.00010599001944011203,
+      "loss": 0.5098,
+      "step": 2520
+    },
+    {
+      "epoch": 1.79,
+      "eval_loss": 1.0929887294769287,
+      "eval_runtime": 51.6843,
+      "eval_samples_per_second": 11.28,
+      "eval_steps_per_second": 11.28,
+      "step": 2520
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 0.0001058832781893502,
+      "loss": 0.687,
+      "step": 2521
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 0.00010577656138854296,
+      "loss": 0.5356,
+      "step": 2522
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 0.00010566986909683392,
+      "loss": 0.5135,
+      "step": 2523
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 0.00010556320137335303,
+      "loss": 0.7238,
+      "step": 2524
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 0.00010545655827721673,
+      "loss": 0.4594,
+      "step": 2525
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 0.00010534993986752774,
+      "loss": 0.4427,
+      "step": 2526
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 0.00010524334620337514,
+      "loss": 0.6537,
+      "step": 2527
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 0.00010513677734383422,
+      "loss": 0.5698,
+      "step": 2528
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 0.00010503023334796654,
+      "loss": 0.4614,
+      "step": 2529
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 0.00010492371427482001,
+      "loss": 0.5309,
+      "step": 2530
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 0.0001048172201834285,
+      "loss": 0.6344,
+      "step": 2531
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.00010471075113281229,
+      "loss": 0.5827,
+      "step": 2532
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.00010460430718197757,
+      "loss": 0.5885,
+      "step": 2533
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.00010449788838991683,
+      "loss": 0.4883,
+      "step": 2534
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.00010439149481560832,
+      "loss": 0.5123,
+      "step": 2535
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.0001042851265180166,
+      "loss": 0.6378,
+      "step": 2536
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.00010417878355609204,
+      "loss": 0.5754,
+      "step": 2537
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.00010407246598877113,
+      "loss": 0.4651,
+      "step": 2538
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.00010396617387497609,
+      "loss": 0.5516,
+      "step": 2539
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.00010385990727361517,
+      "loss": 0.5909,
+      "step": 2540
+    },
+    {
+      "epoch": 1.8,
+      "eval_loss": 1.0919890403747559,
+      "eval_runtime": 53.0036,
+      "eval_samples_per_second": 10.999,
+      "eval_steps_per_second": 10.999,
+      "step": 2540
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.00010375366624358247,
+      "loss": 0.5565,
+      "step": 2541
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.0001036474508437579,
+      "loss": 0.5653,
+      "step": 2542
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.00010354126113300711,
+      "loss": 0.5453,
+      "step": 2543
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.00010343509717018157,
+      "loss": 0.6255,
+      "step": 2544
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.00010332895901411858,
+      "loss": 0.5834,
+      "step": 2545
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.00010322284672364088,
+      "loss": 0.5481,
+      "step": 2546
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.00010311676035755709,
+      "loss": 0.6328,
+      "step": 2547
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.00010301069997466141,
+      "loss": 0.6004,
+      "step": 2548
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.00010290466563373365,
+      "loss": 0.5441,
+      "step": 2549
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.00010279865739353909,
+      "loss": 0.5075,
+      "step": 2550
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.00010269267531282866,
+      "loss": 0.4336,
+      "step": 2551
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.00010258671945033879,
+      "loss": 0.5856,
+      "step": 2552
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.00010248078986479127,
+      "loss": 0.7359,
+      "step": 2553
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.00010237488661489343,
+      "loss": 0.4687,
+      "step": 2554
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.00010226900975933798,
+      "loss": 0.7038,
+      "step": 2555
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.00010216315935680304,
+      "loss": 0.6073,
+      "step": 2556
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.00010205733546595194,
+      "loss": 0.4907,
+      "step": 2557
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.00010195153814543341,
+      "loss": 0.5263,
+      "step": 2558
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.00010184576745388149,
+      "loss": 0.6082,
+      "step": 2559
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.00010174002344991544,
+      "loss": 0.5373,
+      "step": 2560
+    },
+    {
+      "epoch": 1.81,
+      "eval_loss": 1.0942962169647217,
+      "eval_runtime": 48.9057,
+      "eval_samples_per_second": 11.921,
+      "eval_steps_per_second": 11.921,
+      "step": 2560
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.00010163430619213966,
+      "loss": 0.4886,
+      "step": 2561
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.00010152861573914376,
+      "loss": 0.5657,
+      "step": 2562
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.00010142295214950253,
+      "loss": 0.5008,
+      "step": 2563
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.00010131731548177585,
+      "loss": 0.6415,
+      "step": 2564
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.00010121170579450868,
+      "loss": 0.5527,
+      "step": 2565
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.00010110612314623104,
+      "loss": 0.6845,
+      "step": 2566
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.000101000567595458,
+      "loss": 0.6151,
+      "step": 2567
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.00010089503920068944,
+      "loss": 0.5398,
+      "step": 2568
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.00010078953802041037,
+      "loss": 0.5267,
+      "step": 2569
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.00010068406411309068,
+      "loss": 0.5386,
+      "step": 2570
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.00010057861753718515,
+      "loss": 0.6467,
+      "step": 2571
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.00010047319835113333,
+      "loss": 0.5321,
+      "step": 2572
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.00010036780661335968,
+      "loss": 0.6161,
+      "step": 2573
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.00010026244238227347,
+      "loss": 0.6833,
+      "step": 2574
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 0.00010015710571626857,
+      "loss": 0.4621,
+      "step": 2575
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 0.00010005179667372373,
+      "loss": 0.4072,
+      "step": 2576
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 9.994651531300232e-05,
+      "loss": 0.5289,
+      "step": 2577
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 9.984126169245247e-05,
+      "loss": 0.642,
+      "step": 2578
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 9.973603587040671e-05,
+      "loss": 0.544,
+      "step": 2579
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 9.963083790518238e-05,
+      "loss": 0.4685,
+      "step": 2580
+    },
+    {
+      "epoch": 1.83,
+      "eval_loss": 1.0935213565826416,
+      "eval_runtime": 48.1332,
+      "eval_samples_per_second": 12.112,
+      "eval_steps_per_second": 12.112,
+      "step": 2580
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 9.952566785508133e-05,
+      "loss": 0.5816,
+      "step": 2581
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 9.94205257783899e-05,
+      "loss": 0.6267,
+      "step": 2582
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 9.931541173337887e-05,
+      "loss": 0.5535,
+      "step": 2583
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 9.921032577830366e-05,
+      "loss": 0.5698,
+      "step": 2584
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 9.910526797140394e-05,
+      "loss": 0.5959,
+      "step": 2585
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 9.9000238370904e-05,
+      "loss": 0.4735,
+      "step": 2586
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 9.889523703501218e-05,
+      "loss": 0.5122,
+      "step": 2587
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 9.87902640219214e-05,
+      "loss": 0.6605,
+      "step": 2588
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 9.868531938980888e-05,
+      "loss": 0.5578,
+      "step": 2589
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 9.858040319683594e-05,
+      "loss": 0.4725,
+      "step": 2590
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 9.847551550114829e-05,
+      "loss": 0.6063,
+      "step": 2591
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 9.83706563608758e-05,
+      "loss": 0.5522,
+      "step": 2592
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 9.826582583413257e-05,
+      "loss": 0.5756,
+      "step": 2593
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 9.816102397901671e-05,
+      "loss": 0.4526,
+      "step": 2594
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 9.805625085361057e-05,
+      "loss": 0.6078,
+      "step": 2595
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 9.79515065159805e-05,
+      "loss": 0.5306,
+      "step": 2596
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 9.784679102417697e-05,
+      "loss": 0.5289,
+      "step": 2597
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 9.774210443623432e-05,
+      "loss": 0.5447,
+      "step": 2598
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 9.763744681017102e-05,
+      "loss": 0.6631,
+      "step": 2599
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 9.753281820398952e-05,
+      "loss": 0.7373,
+      "step": 2600
+    },
+    {
+      "epoch": 1.84,
+      "eval_loss": 1.0865358114242554,
+      "eval_runtime": 48.2837,
+      "eval_samples_per_second": 12.074,
+      "eval_steps_per_second": 12.074,
+      "step": 2600
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 9.742821867567592e-05,
+      "loss": 0.4989,
+      "step": 2601
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 9.732364828320051e-05,
+      "loss": 0.6177,
+      "step": 2602
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 9.721910708451728e-05,
+      "loss": 0.5911,
+      "step": 2603
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 9.711459513756411e-05,
+      "loss": 0.5974,
+      "step": 2604
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 9.701011250026253e-05,
+      "loss": 0.5807,
+      "step": 2605
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 9.690565923051801e-05,
+      "loss": 0.6266,
+      "step": 2606
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 9.680123538621964e-05,
+      "loss": 0.5686,
+      "step": 2607
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 9.669684102524028e-05,
+      "loss": 0.5954,
+      "step": 2608
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 9.659247620543625e-05,
+      "loss": 0.5903,
+      "step": 2609
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 9.648814098464773e-05,
+      "loss": 0.6918,
+      "step": 2610
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 9.63838354206984e-05,
+      "loss": 0.5353,
+      "step": 2611
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 9.627955957139549e-05,
+      "loss": 0.5963,
+      "step": 2612
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 9.617531349452978e-05,
+      "loss": 0.4573,
+      "step": 2613
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 9.607109724787556e-05,
+      "loss": 0.5912,
+      "step": 2614
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 9.596691088919062e-05,
+      "loss": 0.5424,
+      "step": 2615
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 9.586275447621608e-05,
+      "loss": 0.4973,
+      "step": 2616
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 9.575862806667652e-05,
+      "loss": 0.5511,
+      "step": 2617
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 9.565453171827995e-05,
+      "loss": 0.5214,
+      "step": 2618
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 9.555046548871769e-05,
+      "loss": 0.5555,
+      "step": 2619
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 9.544642943566424e-05,
+      "loss": 0.6067,
+      "step": 2620
+    },
+    {
+      "epoch": 1.86,
+      "eval_loss": 1.0844948291778564,
+      "eval_runtime": 48.9181,
+      "eval_samples_per_second": 11.918,
+      "eval_steps_per_second": 11.918,
+      "step": 2620
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 9.534242361677756e-05,
+      "loss": 0.6386,
+      "step": 2621
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 9.523844808969878e-05,
+      "loss": 0.5754,
+      "step": 2622
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 9.51345029120522e-05,
+      "loss": 0.6138,
+      "step": 2623
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 9.503058814144537e-05,
+      "loss": 0.4591,
+      "step": 2624
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 9.49267038354689e-05,
+      "loss": 0.5523,
+      "step": 2625
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 9.482285005169674e-05,
+      "loss": 0.5851,
+      "step": 2626
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 9.471902684768556e-05,
+      "loss": 0.6735,
+      "step": 2627
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 9.461523428097534e-05,
+      "loss": 0.5289,
+      "step": 2628
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 9.451147240908906e-05,
+      "loss": 0.5775,
+      "step": 2629
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 9.440774128953265e-05,
+      "loss": 0.4975,
+      "step": 2630
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 9.430404097979495e-05,
+      "loss": 0.5719,
+      "step": 2631
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 9.420037153734778e-05,
+      "loss": 0.579,
+      "step": 2632
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 9.409673301964588e-05,
+      "loss": 0.5195,
+      "step": 2633
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 9.399312548412677e-05,
+      "loss": 0.5622,
+      "step": 2634
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 9.388954898821086e-05,
+      "loss": 0.5512,
+      "step": 2635
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 9.378600358930133e-05,
+      "loss": 0.649,
+      "step": 2636
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 9.368248934478422e-05,
+      "loss": 0.6277,
+      "step": 2637
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 9.357900631202811e-05,
+      "loss": 0.4957,
+      "step": 2638
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 9.347555454838441e-05,
+      "loss": 0.6007,
+      "step": 2639
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 9.33721341111872e-05,
+      "loss": 0.5795,
+      "step": 2640
+    },
+    {
+      "epoch": 1.87,
+      "eval_loss": 1.087457299232483,
+      "eval_runtime": 50.1053,
+      "eval_samples_per_second": 11.635,
+      "eval_steps_per_second": 11.635,
+      "step": 2640
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 9.326874505775329e-05,
+      "loss": 0.5657,
+      "step": 2641
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 9.31653874453818e-05,
+      "loss": 0.5894,
+      "step": 2642
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 9.30620613313547e-05,
+      "loss": 0.5944,
+      "step": 2643
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 9.295876677293647e-05,
+      "loss": 0.6779,
+      "step": 2644
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 9.285550382737398e-05,
+      "loss": 0.6765,
+      "step": 2645
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 9.275227255189663e-05,
+      "loss": 0.5649,
+      "step": 2646
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 9.264907300371634e-05,
+      "loss": 0.5551,
+      "step": 2647
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 9.254590524002746e-05,
+      "loss": 0.8507,
+      "step": 2648
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 9.244276931800648e-05,
+      "loss": 0.5303,
+      "step": 2649
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 9.233966529481255e-05,
+      "loss": 0.6021,
+      "step": 2650
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 9.223659322758696e-05,
+      "loss": 0.5957,
+      "step": 2651
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 9.213355317345338e-05,
+      "loss": 0.4952,
+      "step": 2652
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 9.203054518951765e-05,
+      "loss": 0.5686,
+      "step": 2653
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 9.192756933286792e-05,
+      "loss": 0.6393,
+      "step": 2654
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 9.182462566057456e-05,
+      "loss": 0.6099,
+      "step": 2655
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 9.172171422968992e-05,
+      "loss": 0.6323,
+      "step": 2656
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 9.161883509724865e-05,
+      "loss": 0.4984,
+      "step": 2657
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 9.151598832026748e-05,
+      "loss": 0.5341,
+      "step": 2658
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 9.141317395574522e-05,
+      "loss": 0.4996,
+      "step": 2659
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 9.13103920606626e-05,
+      "loss": 0.4362,
+      "step": 2660
+    },
+    {
+      "epoch": 1.89,
+      "eval_loss": 1.0860021114349365,
+      "eval_runtime": 49.1619,
+      "eval_samples_per_second": 11.859,
+      "eval_steps_per_second": 11.859,
+      "step": 2660
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 9.120764269198245e-05,
+      "loss": 0.6025,
+      "step": 2661
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 9.110492590664956e-05,
+      "loss": 0.6441,
+      "step": 2662
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 9.100224176159072e-05,
+      "loss": 0.5375,
+      "step": 2663
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 9.089959031371451e-05,
+      "loss": 0.5575,
+      "step": 2664
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 9.079697161991146e-05,
+      "loss": 0.459,
+      "step": 2665
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 9.069438573705402e-05,
+      "loss": 0.6099,
+      "step": 2666
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 9.05918327219962e-05,
+      "loss": 0.5706,
+      "step": 2667
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 9.04893126315741e-05,
+      "loss": 0.6377,
+      "step": 2668
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 9.03868255226054e-05,
+      "loss": 0.6215,
+      "step": 2669
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 9.028437145188961e-05,
+      "loss": 0.6681,
+      "step": 2670
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 9.018195047620773e-05,
+      "loss": 0.5627,
+      "step": 2671
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 9.007956265232262e-05,
+      "loss": 0.6437,
+      "step": 2672
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 8.99772080369787e-05,
+      "loss": 0.5694,
+      "step": 2673
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 8.9874886686902e-05,
+      "loss": 0.596,
+      "step": 2674
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 8.977259865880004e-05,
+      "loss": 0.6429,
+      "step": 2675
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 8.967034400936197e-05,
+      "loss": 0.5799,
+      "step": 2676
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 8.956812279525847e-05,
+      "loss": 0.5584,
+      "step": 2677
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 8.946593507314146e-05,
+      "loss": 0.5945,
+      "step": 2678
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 8.936378089964454e-05,
+      "loss": 0.4359,
+      "step": 2679
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 8.926166033138266e-05,
+      "loss": 0.6025,
+      "step": 2680
+    },
+    {
+      "epoch": 1.9,
+      "eval_loss": 1.0834202766418457,
+      "eval_runtime": 51.0086,
+      "eval_samples_per_second": 11.429,
+      "eval_steps_per_second": 11.429,
+      "step": 2680
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 8.915957342495215e-05,
+      "loss": 0.4411,
+      "step": 2681
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 8.905752023693052e-05,
+      "loss": 0.444,
+      "step": 2682
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 8.895550082387683e-05,
+      "loss": 0.5689,
+      "step": 2683
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 8.885351524233129e-05,
+      "loss": 0.4704,
+      "step": 2684
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 8.875156354881543e-05,
+      "loss": 0.5514,
+      "step": 2685
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 8.864964579983189e-05,
+      "loss": 0.6378,
+      "step": 2686
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 8.854776205186462e-05,
+      "loss": 0.553,
+      "step": 2687
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 8.84459123613787e-05,
+      "loss": 0.5774,
+      "step": 2688
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 8.834409678482021e-05,
+      "loss": 0.6136,
+      "step": 2689
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 8.824231537861648e-05,
+      "loss": 0.6466,
+      "step": 2690
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 8.814056819917585e-05,
+      "loss": 0.6702,
+      "step": 2691
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 8.803885530288769e-05,
+      "loss": 0.6422,
+      "step": 2692
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 8.793717674612232e-05,
+      "loss": 0.5391,
+      "step": 2693
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 8.78355325852311e-05,
+      "loss": 0.508,
+      "step": 2694
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 8.77339228765463e-05,
+      "loss": 0.498,
+      "step": 2695
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 8.763234767638116e-05,
+      "loss": 0.6812,
+      "step": 2696
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 8.753080704102965e-05,
+      "loss": 0.5944,
+      "step": 2697
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 8.742930102676664e-05,
+      "loss": 0.6608,
+      "step": 2698
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 8.732782968984791e-05,
+      "loss": 0.4502,
+      "step": 2699
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 8.72263930865099e-05,
+      "loss": 0.5821,
+      "step": 2700
+    },
+    {
+      "epoch": 1.91,
+      "eval_loss": 1.0828958749771118,
+      "eval_runtime": 52.3703,
+      "eval_samples_per_second": 11.132,
+      "eval_steps_per_second": 11.132,
+      "step": 2700
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 8.712499127296989e-05,
+      "loss": 0.5981,
+      "step": 2701
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 8.702362430542582e-05,
+      "loss": 0.6162,
+      "step": 2702
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 8.69222922400564e-05,
+      "loss": 0.5602,
+      "step": 2703
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 8.682099513302084e-05,
+      "loss": 0.6438,
+      "step": 2704
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 8.67197330404591e-05,
+      "loss": 0.6012,
+      "step": 2705
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 8.661850601849168e-05,
+      "loss": 0.7468,
+      "step": 2706
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 8.651731412321977e-05,
+      "loss": 0.6059,
+      "step": 2707
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 8.641615741072488e-05,
+      "loss": 0.5941,
+      "step": 2708
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 8.631503593706916e-05,
+      "loss": 0.5633,
+      "step": 2709
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 8.621394975829528e-05,
+      "loss": 0.5931,
+      "step": 2710
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 8.611289893042614e-05,
+      "loss": 0.5564,
+      "step": 2711
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 8.601188350946524e-05,
+      "loss": 0.6233,
+      "step": 2712
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 8.591090355139639e-05,
+      "loss": 0.6076,
+      "step": 2713
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 8.580995911218383e-05,
+      "loss": 0.6051,
+      "step": 2714
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 8.570905024777192e-05,
+      "loss": 0.5424,
+      "step": 2715
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 8.560817701408546e-05,
+      "loss": 0.5531,
+      "step": 2716
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 8.55073394670295e-05,
+      "loss": 0.6307,
+      "step": 2717
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 8.540653766248924e-05,
+      "loss": 0.7296,
+      "step": 2718
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 8.530577165633013e-05,
+      "loss": 0.5596,
+      "step": 2719
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 8.520504150439772e-05,
+      "loss": 0.5985,
+      "step": 2720
+    },
+    {
+      "epoch": 1.93,
+      "eval_loss": 1.0860912799835205,
+      "eval_runtime": 49.8155,
+      "eval_samples_per_second": 11.703,
+      "eval_steps_per_second": 11.703,
+      "step": 2720
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 8.510434726251778e-05,
+      "loss": 0.609,
+      "step": 2721
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 8.500368898649616e-05,
+      "loss": 0.6998,
+      "step": 2722
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 8.490306673211862e-05,
+      "loss": 0.6446,
+      "step": 2723
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 8.480248055515117e-05,
+      "loss": 0.5515,
+      "step": 2724
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 8.470193051133975e-05,
+      "loss": 0.6087,
+      "step": 2725
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 8.460141665641015e-05,
+      "loss": 0.6009,
+      "step": 2726
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 8.450093904606829e-05,
+      "loss": 0.4999,
+      "step": 2727
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 8.440049773599992e-05,
+      "loss": 0.6339,
+      "step": 2728
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 8.430009278187076e-05,
+      "loss": 0.5712,
+      "step": 2729
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 8.419972423932612e-05,
+      "loss": 0.6506,
+      "step": 2730
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 8.409939216399143e-05,
+      "loss": 0.4822,
+      "step": 2731
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 8.399909661147175e-05,
+      "loss": 0.5474,
+      "step": 2732
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 8.3898837637352e-05,
+      "loss": 0.6128,
+      "step": 2733
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 8.37986152971967e-05,
+      "loss": 0.6531,
+      "step": 2734
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 8.369842964655012e-05,
+      "loss": 0.5527,
+      "step": 2735
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 8.359828074093633e-05,
+      "loss": 0.7299,
+      "step": 2736
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 8.349816863585874e-05,
+      "loss": 0.6219,
+      "step": 2737
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 8.339809338680062e-05,
+      "loss": 0.4355,
+      "step": 2738
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 8.329805504922474e-05,
+      "loss": 0.5907,
+      "step": 2739
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 8.319805367857343e-05,
+      "loss": 0.5889,
+      "step": 2740
+    },
+    {
+      "epoch": 1.94,
+      "eval_loss": 1.0843344926834106,
+      "eval_runtime": 53.9466,
+      "eval_samples_per_second": 10.807,
+      "eval_steps_per_second": 10.807,
+      "step": 2740
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 8.309808933026845e-05,
+      "loss": 0.5841,
+      "step": 2741
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 8.299816205971114e-05,
+      "loss": 0.587,
+      "step": 2742
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 8.289827192228226e-05,
+      "loss": 0.6594,
+      "step": 2743
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 8.279841897334201e-05,
+      "loss": 0.6299,
+      "step": 2744
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 8.269860326822987e-05,
+      "loss": 0.6934,
+      "step": 2745
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 8.259882486226481e-05,
+      "loss": 0.7158,
+      "step": 2746
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 8.249908381074508e-05,
+      "loss": 0.4991,
+      "step": 2747
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 8.23993801689482e-05,
+      "loss": 0.6044,
+      "step": 2748
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 8.2299713992131e-05,
+      "loss": 0.6947,
+      "step": 2749
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 8.220008533552953e-05,
+      "loss": 0.6276,
+      "step": 2750
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 8.21004942543591e-05,
+      "loss": 0.3949,
+      "step": 2751
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 8.200094080381405e-05,
+      "loss": 0.5327,
+      "step": 2752
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 8.190142503906798e-05,
+      "loss": 0.6709,
+      "step": 2753
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 8.180194701527361e-05,
+      "loss": 0.5111,
+      "step": 2754
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 8.170250678756279e-05,
+      "loss": 0.4932,
+      "step": 2755
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 8.160310441104616e-05,
+      "loss": 0.5643,
+      "step": 2756
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 8.150373994081372e-05,
+      "loss": 0.6092,
+      "step": 2757
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 8.140441343193423e-05,
+      "loss": 0.5925,
+      "step": 2758
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 8.130512493945555e-05,
+      "loss": 0.6941,
+      "step": 2759
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 8.120587451840438e-05,
+      "loss": 0.5261,
+      "step": 2760
+    },
+    {
+      "epoch": 1.96,
+      "eval_loss": 1.081740379333496,
+      "eval_runtime": 52.7009,
+      "eval_samples_per_second": 11.062,
+      "eval_steps_per_second": 11.062,
+      "step": 2760
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 8.110666222378638e-05,
+      "loss": 0.5577,
+      "step": 2761
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 8.10074881105861e-05,
+      "loss": 0.6613,
+      "step": 2762
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 8.090835223376676e-05,
+      "loss": 0.6126,
+      "step": 2763
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 8.080925464827059e-05,
+      "loss": 0.6159,
+      "step": 2764
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 8.07101954090185e-05,
+      "loss": 0.5805,
+      "step": 2765
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 8.061117457091028e-05,
+      "loss": 0.5658,
+      "step": 2766
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 8.051219218882412e-05,
+      "loss": 0.5231,
+      "step": 2767
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 8.041324831761725e-05,
+      "loss": 0.6705,
+      "step": 2768
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 8.031434301212542e-05,
+      "loss": 0.5637,
+      "step": 2769
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 8.02154763271629e-05,
+      "loss": 0.6367,
+      "step": 2770
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 8.01166483175227e-05,
+      "loss": 0.6303,
+      "step": 2771
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 8.001785903797638e-05,
+      "loss": 0.6038,
+      "step": 2772
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 7.991910854327402e-05,
+      "loss": 0.5368,
+      "step": 2773
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 7.982039688814411e-05,
+      "loss": 0.5551,
+      "step": 2774
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 7.972172412729374e-05,
+      "loss": 0.6897,
+      "step": 2775
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 7.962309031540838e-05,
+      "loss": 0.6208,
+      "step": 2776
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 7.952449550715203e-05,
+      "loss": 0.5945,
+      "step": 2777
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 7.942593975716684e-05,
+      "loss": 0.5732,
+      "step": 2778
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 7.93274231200735e-05,
+      "loss": 0.5867,
+      "step": 2779
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 7.9228945650471e-05,
+      "loss": 0.6552,
+      "step": 2780
+    },
+    {
+      "epoch": 1.97,
+      "eval_loss": 1.0819121599197388,
+      "eval_runtime": 52.0221,
+      "eval_samples_per_second": 11.207,
+      "eval_steps_per_second": 11.207,
+      "step": 2780
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 7.913050740293657e-05,
+      "loss": 0.5705,
+      "step": 2781
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 7.903210843202574e-05,
+      "loss": 0.6197,
+      "step": 2782
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 7.893374879227227e-05,
+      "loss": 0.5437,
+      "step": 2783
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 7.883542853818816e-05,
+      "loss": 0.6202,
+      "step": 2784
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 7.873714772426342e-05,
+      "loss": 0.6434,
+      "step": 2785
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 7.86389064049664e-05,
+      "loss": 0.6076,
+      "step": 2786
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 7.854070463474346e-05,
+      "loss": 0.6028,
+      "step": 2787
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 7.84425424680191e-05,
+      "loss": 0.5576,
+      "step": 2788
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 7.834441995919575e-05,
+      "loss": 0.5032,
+      "step": 2789
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 7.824633716265399e-05,
+      "loss": 0.5847,
+      "step": 2790
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 7.81482941327524e-05,
+      "loss": 0.5651,
+      "step": 2791
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 7.805029092382735e-05,
+      "loss": 0.5856,
+      "step": 2792
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 7.795232759019334e-05,
+      "loss": 0.4629,
+      "step": 2793
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 7.785440418614266e-05,
+      "loss": 0.4663,
+      "step": 2794
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 7.775652076594554e-05,
+      "loss": 0.4623,
+      "step": 2795
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 7.765867738384993e-05,
+      "loss": 0.6689,
+      "step": 2796
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 7.75608740940817e-05,
+      "loss": 0.4442,
+      "step": 2797
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 7.746311095084449e-05,
+      "loss": 0.4191,
+      "step": 2798
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 7.736538800831964e-05,
+      "loss": 0.633,
+      "step": 2799
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 7.72677053206663e-05,
+      "loss": 0.5959,
+      "step": 2800
+    },
+    {
+      "epoch": 1.99,
+      "eval_loss": 1.0815908908843994,
+      "eval_runtime": 55.4933,
+      "eval_samples_per_second": 10.506,
+      "eval_steps_per_second": 10.506,
+      "step": 2800
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 7.717006294202116e-05,
+      "loss": 0.5983,
+      "step": 2801
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 7.70724609264988e-05,
+      "loss": 0.7114,
+      "step": 2802
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 7.697489932819114e-05,
+      "loss": 0.6138,
+      "step": 2803
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 7.68773782011679e-05,
+      "loss": 0.4792,
+      "step": 2804
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 7.677989759947633e-05,
+      "loss": 0.4508,
+      "step": 2805
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 7.668245757714129e-05,
+      "loss": 0.7298,
+      "step": 2806
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 7.658505818816495e-05,
+      "loss": 0.5663,
+      "step": 2807
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 7.648769948652713e-05,
+      "loss": 0.5153,
+      "step": 2808
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 7.639038152618505e-05,
+      "loss": 0.5217,
+      "step": 2809
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 7.629310436107342e-05,
+      "loss": 0.5486,
+      "step": 2810
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 7.619586804510416e-05,
+      "loss": 0.4922,
+      "step": 2811
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 7.609867263216671e-05,
+      "loss": 0.6291,
+      "step": 2812
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 7.600151817612788e-05,
+      "loss": 0.5628,
+      "step": 2813
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 7.590440473083157e-05,
+      "loss": 0.5611,
+      "step": 2814
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 7.580733235009913e-05,
+      "loss": 0.6037,
+      "step": 2815
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 7.571030108772914e-05,
+      "loss": 0.6763,
+      "step": 2816
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 7.561331099749735e-05,
+      "loss": 0.4548,
+      "step": 2817
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 7.551636213315663e-05,
+      "loss": 0.4728,
+      "step": 2818
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 7.54194545484371e-05,
+      "loss": 0.6367,
+      "step": 2819
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 7.532258829704599e-05,
+      "loss": 0.6805,
+      "step": 2820
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 1.0804917812347412,
+      "eval_runtime": 55.5825,
+      "eval_samples_per_second": 10.489,
+      "eval_steps_per_second": 10.489,
+      "step": 2820
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 7.522576343266759e-05,
+      "loss": 0.6268,
+      "step": 2821
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 4230,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "total_flos": 9.067433351656243e+17,
+  "trial_name": null,
+  "trial_params": null
+}