diff --git "a/checkpoint-1000/trainer_state.json" "b/checkpoint-1000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1000/trainer_state.json"
@@ -0,0 +1,6099 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.7393032067276591,
+  "eval_steps": 100,
+  "global_step": 1000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.3170731707317065e-06,
+      "loss": 0.7162,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4634146341463413e-05,
+      "loss": 0.8676,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1951219512195117e-05,
+      "loss": 0.8797,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9268292682926826e-05,
+      "loss": 0.6815,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6585365853658535e-05,
+      "loss": 0.6264,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3902439024390234e-05,
+      "loss": 0.711,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.121951219512195e-05,
+      "loss": 0.8029,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.853658536585365e-05,
+      "loss": 0.5576,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.585365853658536e-05,
+      "loss": 0.7127,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.317073170731707e-05,
+      "loss": 0.5231,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.048780487804878e-05,
+      "loss": 1.1346,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.780487804878047e-05,
+      "loss": 0.6838,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.512195121951219e-05,
+      "loss": 0.8151,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0001024390243902439,
+      "loss": 0.9672,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0001097560975609756,
+      "loss": 0.7119,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0001170731707317073,
+      "loss": 0.701,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.000124390243902439,
+      "loss": 0.7036,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00013170731707317073,
+      "loss": 0.4933,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00013902439024390242,
+      "loss": 0.9684,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00014634146341463414,
+      "loss": 1.195,
+      "step": 20
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00015365853658536583,
+      "loss": 0.7636,
+      "step": 21
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00016097560975609755,
+      "loss": 0.5474,
+      "step": 22
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00016829268292682927,
+      "loss": 1.0586,
+      "step": 23
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00017560975609756094,
+      "loss": 0.4522,
+      "step": 24
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00018292682926829266,
+      "loss": 0.5121,
+      "step": 25
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00019024390243902437,
+      "loss": 0.9219,
+      "step": 26
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0001975609756097561,
+      "loss": 0.6772,
+      "step": 27
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0002048780487804878,
+      "loss": 0.7117,
+      "step": 28
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00021219512195121948,
+      "loss": 0.8291,
+      "step": 29
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0002195121951219512,
+      "loss": 0.4971,
+      "step": 30
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00022682926829268292,
+      "loss": 0.6877,
+      "step": 31
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0002341463414634146,
+      "loss": 0.4998,
+      "step": 32
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00024146341463414633,
+      "loss": 0.5154,
+      "step": 33
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.000248780487804878,
+      "loss": 0.7322,
+      "step": 34
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0002560975609756097,
+      "loss": 0.7293,
+      "step": 35
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00026341463414634146,
+      "loss": 0.7618,
+      "step": 36
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00027073170731707315,
+      "loss": 1.0949,
+      "step": 37
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00027804878048780484,
+      "loss": 0.5851,
+      "step": 38
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00028536585365853654,
+      "loss": 0.559,
+      "step": 39
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0002926829268292683,
+      "loss": 0.6606,
+      "step": 40
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0003,
+      "loss": 0.4873,
+      "step": 41
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00029999956931929273,
+      "loss": 0.3105,
+      "step": 42
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00029999827727964407,
+      "loss": 0.383,
+      "step": 43
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0002999961238884735,
+      "loss": 0.7269,
+      "step": 44
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00029999310915814663,
+      "loss": 0.5993,
+      "step": 45
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00029998923310597524,
+      "loss": 0.5674,
+      "step": 46
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0002999844957542173,
+      "loss": 0.4887,
+      "step": 47
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002999788971300765,
+      "loss": 1.0146,
+      "step": 48
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00029997243726570256,
+      "loss": 0.8546,
+      "step": 49
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002999651161981906,
+      "loss": 1.1779,
+      "step": 50
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002999569339695812,
+      "loss": 0.6413,
+      "step": 51
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00029994789062686003,
+      "loss": 0.5926,
+      "step": 52
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002999379862219577,
+      "loss": 0.828,
+      "step": 53
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002999272208117494,
+      "loss": 0.5488,
+      "step": 54
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002999155944580544,
+      "loss": 0.645,
+      "step": 55
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00029990310722763613,
+      "loss": 0.784,
+      "step": 56
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002998897591922013,
+      "loss": 0.4035,
+      "step": 57
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00029987555042839976,
+      "loss": 1.065,
+      "step": 58
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00029986048101782414,
+      "loss": 0.6875,
+      "step": 59
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002998445510470091,
+      "loss": 0.5348,
+      "step": 60
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00029982776060743106,
+      "loss": 0.7269,
+      "step": 61
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00029981010979550766,
+      "loss": 0.4813,
+      "step": 62
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.000299791598712597,
+      "loss": 0.4941,
+      "step": 63
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0002997722274649974,
+      "loss": 0.4421,
+      "step": 64
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0002997519961639464,
+      "loss": 0.3293,
+      "step": 65
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00029973090492562045,
+      "loss": 0.4955,
+      "step": 66
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0002997089538711341,
+      "loss": 0.8179,
+      "step": 67
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00029968614312653923,
+      "loss": 0.9704,
+      "step": 68
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00029966247282282453,
+      "loss": 0.6526,
+      "step": 69
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0002996379430959145,
+      "loss": 0.3936,
+      "step": 70
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.000299612554086669,
+      "loss": 0.5325,
+      "step": 71
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.000299586305940882,
+      "loss": 0.4114,
+      "step": 72
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00029955919880928124,
+      "loss": 0.4236,
+      "step": 73
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0002995312328475268,
+      "loss": 0.4274,
+      "step": 74
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0002995024082162107,
+      "loss": 0.4362,
+      "step": 75
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.000299472725080856,
+      "loss": 0.3605,
+      "step": 76
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00029944218361191514,
+      "loss": 0.5172,
+      "step": 77
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00029941078398476985,
+      "loss": 0.5467,
+      "step": 78
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00029937852637972967,
+      "loss": 0.3366,
+      "step": 79
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.000299345410982031,
+      "loss": 0.4271,
+      "step": 80
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0002993114379818359,
+      "loss": 0.9314,
+      "step": 81
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00029927660757423123,
+      "loss": 0.6369,
+      "step": 82
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0002992409199592276,
+      "loss": 0.7792,
+      "step": 83
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00029920437534175785,
+      "loss": 0.3398,
+      "step": 84
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00029916697393167615,
+      "loss": 0.6307,
+      "step": 85
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00029912871594375666,
+      "loss": 0.5283,
+      "step": 86
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0002990896015976924,
+      "loss": 0.5731,
+      "step": 87
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.000299049631118094,
+      "loss": 0.5389,
+      "step": 88
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0002990088047344883,
+      "loss": 0.6683,
+      "step": 89
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00029896712268131717,
+      "loss": 1.0819,
+      "step": 90
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00029892458519793595,
+      "loss": 0.2681,
+      "step": 91
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0002988811925286124,
+      "loss": 0.7518,
+      "step": 92
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00029883694492252474,
+      "loss": 0.5616,
+      "step": 93
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0002987918426337611,
+      "loss": 0.687,
+      "step": 94
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0002987458859213172,
+      "loss": 0.561,
+      "step": 95
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0002986990750490953,
+      "loss": 0.3038,
+      "step": 96
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00029865141028590257,
+      "loss": 0.4021,
+      "step": 97
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0002986028919054496,
+      "loss": 0.4647,
+      "step": 98
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0002985535201863488,
+      "loss": 0.3641,
+      "step": 99
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00029850329541211283,
+      "loss": 0.5877,
+      "step": 100
+    },
+    {
+      "epoch": 0.07,
+      "eval_loss": 0.7984210848808289,
+      "eval_runtime": 12.9946,
+      "eval_samples_per_second": 0.847,
+      "eval_steps_per_second": 0.847,
+      "step": 100
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00029845221787115286,
+      "loss": 0.4791,
+      "step": 101
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0002984002878567771,
+      "loss": 0.5435,
+      "step": 102
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0002983475056671889,
+      "loss": 0.346,
+      "step": 103
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00029829387160548525,
+      "loss": 0.3577,
+      "step": 104
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0002982393859796548,
+      "loss": 0.3761,
+      "step": 105
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0002982393859796548,
+      "loss": 0.7425,
+      "step": 106
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0002981840491025764,
+      "loss": 0.3844,
+      "step": 107
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00029812786129201706,
+      "loss": 0.3529,
+      "step": 108
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0002980708228706302,
+      "loss": 0.4415,
+      "step": 109
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00029801293416595374,
+      "loss": 0.4336,
+      "step": 110
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00029795419551040833,
+      "loss": 0.5404,
+      "step": 111
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00029789460724129543,
+      "loss": 0.7089,
+      "step": 112
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0002978341697007952,
+      "loss": 0.5796,
+      "step": 113
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0002977728832359648,
+      "loss": 0.6561,
+      "step": 114
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00029771074819873617,
+      "loss": 0.7428,
+      "step": 115
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00029764776494591423,
+      "loss": 0.6707,
+      "step": 116
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00029758393383917447,
+      "loss": 0.68,
+      "step": 117
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0002975192552450613,
+      "loss": 0.4643,
+      "step": 118
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0002974537295349857,
+      "loss": 0.6456,
+      "step": 119
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0002973873570852231,
+      "loss": 0.5805,
+      "step": 120
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00029732013827691125,
+      "loss": 0.3574,
+      "step": 121
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0002972520734960482,
+      "loss": 0.5281,
+      "step": 122
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00029718316313348964,
+      "loss": 0.3682,
+      "step": 123
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0002971134075849472,
+      "loss": 0.3776,
+      "step": 124
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0002970428072509857,
+      "loss": 0.8614,
+      "step": 125
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0002969713625370212,
+      "loss": 0.308,
+      "step": 126
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0002968990738533186,
+      "loss": 0.5767,
+      "step": 127
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00029682594161498896,
+      "loss": 0.4703,
+      "step": 128
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00029675196624198756,
+      "loss": 0.4235,
+      "step": 129
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00029667714815911136,
+      "loss": 0.3727,
+      "step": 130
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00029660148779599643,
+      "loss": 0.3763,
+      "step": 131
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0002965249855871155,
+      "loss": 0.5048,
+      "step": 132
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0002964476419717756,
+      "loss": 0.4571,
+      "step": 133
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0002963694573941153,
+      "loss": 0.1698,
+      "step": 134
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0002962904323031027,
+      "loss": 0.3583,
+      "step": 135
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00029621056715253206,
+      "loss": 0.3942,
+      "step": 136
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00029612986240102186,
+      "loss": 0.5731,
+      "step": 137
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0002960483185120117,
+      "loss": 0.5174,
+      "step": 138
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00029596593595376016,
+      "loss": 0.2378,
+      "step": 139
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00029588271519934145,
+      "loss": 0.6014,
+      "step": 140
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00029579865672664334,
+      "loss": 0.4233,
+      "step": 141
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0002957137610183639,
+      "loss": 0.8406,
+      "step": 142
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00029562802856200913,
+      "loss": 0.433,
+      "step": 143
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00029554145984988985,
+      "loss": 0.3892,
+      "step": 144
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00029545405537911905,
+      "loss": 0.3632,
+      "step": 145
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.000295365815651609,
+      "loss": 0.6153,
+      "step": 146
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0002952767411740683,
+      "loss": 0.3268,
+      "step": 147
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00029518683245799914,
+      "loss": 0.7161,
+      "step": 148
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00029509609001969413,
+      "loss": 0.4311,
+      "step": 149
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0002950045143802336,
+      "loss": 0.7437,
+      "step": 150
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0002949121060654821,
+      "loss": 0.3991,
+      "step": 151
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00029481886560608634,
+      "loss": 0.4817,
+      "step": 152
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00029472479353747104,
+      "loss": 0.2903,
+      "step": 153
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0002946298903998365,
+      "loss": 0.4583,
+      "step": 154
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00029453415673815555,
+      "loss": 0.4707,
+      "step": 155
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0002944375931021699,
+      "loss": 0.6428,
+      "step": 156
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0002943402000463875,
+      "loss": 0.3799,
+      "step": 157
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0002942419781300793,
+      "loss": 0.207,
+      "step": 158
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00029414292791727563,
+      "loss": 0.1921,
+      "step": 159
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00029404304997676347,
+      "loss": 0.5306,
+      "step": 160
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0002939423448820827,
+      "loss": 0.2036,
+      "step": 161
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0002938408132115233,
+      "loss": 0.5733,
+      "step": 162
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00029373845554812167,
+      "loss": 0.3922,
+      "step": 163
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0002936352724796575,
+      "loss": 0.3732,
+      "step": 164
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00029353126459865004,
+      "loss": 0.3472,
+      "step": 165
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0002934264325023553,
+      "loss": 0.2677,
+      "step": 166
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.000293320776792762,
+      "loss": 0.5054,
+      "step": 167
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00029321429807658856,
+      "loss": 0.3407,
+      "step": 168
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0002931069969652793,
+      "loss": 0.3308,
+      "step": 169
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0002929988740750011,
+      "loss": 0.4544,
+      "step": 170
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00029288993002663996,
+      "loss": 0.7072,
+      "step": 171
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00029278016544579717,
+      "loss": 0.5157,
+      "step": 172
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00029266958096278587,
+      "loss": 0.1607,
+      "step": 173
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00029255817721262744,
+      "loss": 0.3236,
+      "step": 174
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00029244595483504784,
+      "loss": 0.4895,
+      "step": 175
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.000292332914474474,
+      "loss": 0.3522,
+      "step": 176
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00029221905678002976,
+      "loss": 0.7603,
+      "step": 177
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00029210438240553276,
+      "loss": 0.502,
+      "step": 178
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00029198889200949017,
+      "loss": 0.7086,
+      "step": 179
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00029187258625509513,
+      "loss": 0.4529,
+      "step": 180
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00029175546581022295,
+      "loss": 0.5601,
+      "step": 181
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00029163753134742713,
+      "loss": 0.3547,
+      "step": 182
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0002915187835439356,
+      "loss": 0.3628,
+      "step": 183
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0002913992230816469,
+      "loss": 0.4702,
+      "step": 184
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0002912788506471262,
+      "loss": 0.6325,
+      "step": 185
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00029115766693160126,
+      "loss": 0.8866,
+      "step": 186
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00029103567263095856,
+      "loss": 0.5105,
+      "step": 187
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00029091286844573936,
+      "loss": 0.4985,
+      "step": 188
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00029078925508113564,
+      "loss": 0.3461,
+      "step": 189
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00029066483324698584,
+      "loss": 0.4485,
+      "step": 190
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0002905396036577711,
+      "loss": 0.6662,
+      "step": 191
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.000290413567032611,
+      "loss": 0.2159,
+      "step": 192
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0002902867240952595,
+      "loss": 0.2389,
+      "step": 193
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00029015907557410065,
+      "loss": 0.226,
+      "step": 194
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0002900306222021445,
+      "loss": 0.4957,
+      "step": 195
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0002899013647170229,
+      "loss": 0.3227,
+      "step": 196
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00028977130386098523,
+      "loss": 0.5794,
+      "step": 197
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002896404403808942,
+      "loss": 0.5515,
+      "step": 198
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002895087750282215,
+      "loss": 0.2853,
+      "step": 199
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002893763085590436,
+      "loss": 0.3192,
+      "step": 200
+    },
+    {
+      "epoch": 0.15,
+      "eval_loss": 0.7430515885353088,
+      "eval_runtime": 13.0035,
+      "eval_samples_per_second": 0.846,
+      "eval_steps_per_second": 0.846,
+      "step": 200
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002892430417340369,
+      "loss": 0.4321,
+      "step": 201
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002891089753184744,
+      "loss": 0.2985,
+      "step": 202
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002889741100822202,
+      "loss": 0.2427,
+      "step": 203
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002888384467997257,
+      "loss": 0.5945,
+      "step": 204
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00028870198625002516,
+      "loss": 0.6865,
+      "step": 205
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00028856472921673067,
+      "loss": 0.3938,
+      "step": 206
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00028842667648802846,
+      "loss": 0.5073,
+      "step": 207
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002882878288566738,
+      "loss": 0.4465,
+      "step": 208
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00028814818711998667,
+      "loss": 0.3318,
+      "step": 209
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0002880077520798471,
+      "loss": 0.3051,
+      "step": 210
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0002878665245426905,
+      "loss": 0.5964,
+      "step": 211
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00028772450531950325,
+      "loss": 0.6703,
+      "step": 212
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0002875816952258179,
+      "loss": 0.5141,
+      "step": 213
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0002874380950817085,
+      "loss": 0.3797,
+      "step": 214
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00028729370571178587,
+      "loss": 0.4755,
+      "step": 215
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00028714852794519286,
+      "loss": 0.2632,
+      "step": 216
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0002870025626155996,
+      "loss": 0.5007,
+      "step": 217
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00028685581056119887,
+      "loss": 0.5119,
+      "step": 218
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00028670827262470097,
+      "loss": 0.3364,
+      "step": 219
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0002865599496533292,
+      "loss": 0.3543,
+      "step": 220
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00028641084249881476,
+      "loss": 0.6461,
+      "step": 221
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.000286260952017392,
+      "loss": 0.3578,
+      "step": 222
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0002861102790697934,
+      "loss": 0.5015,
+      "step": 223
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00028595882452124474,
+      "loss": 0.2236,
+      "step": 224
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00028580658924146004,
+      "loss": 0.6078,
+      "step": 225
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00028565357410463663,
+      "loss": 0.745,
+      "step": 226
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00028549977998945,
+      "loss": 0.4428,
+      "step": 227
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.000285345207779049,
+      "loss": 0.4141,
+      "step": 228
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0002851898583610505,
+      "loss": 0.2973,
+      "step": 229
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00028503373262753456,
+      "loss": 0.3884,
+      "step": 230
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.000284876831475039,
+      "loss": 0.4595,
+      "step": 231
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0002847191558045544,
+      "loss": 0.7463,
+      "step": 232
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0002845607065215191,
+      "loss": 0.5798,
+      "step": 233
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00028440148453581375,
+      "loss": 0.4011,
+      "step": 234
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00028424149076175623,
+      "loss": 0.4742,
+      "step": 235
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00028408072611809624,
+      "loss": 0.3255,
+      "step": 236
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0002839191915280102,
+      "loss": 0.4717,
+      "step": 237
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0002837568879190958,
+      "loss": 0.4492,
+      "step": 238
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00028359381622336684,
+      "loss": 0.4957,
+      "step": 239
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00028342997737724785,
+      "loss": 0.5463,
+      "step": 240
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00028326537232156853,
+      "loss": 0.3134,
+      "step": 241
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0002831000020015585,
+      "loss": 0.4342,
+      "step": 242
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0002829338673668418,
+      "loss": 0.2409,
+      "step": 243
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0002827669693714315,
+      "loss": 0.4877,
+      "step": 244
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0002825993089737243,
+      "loss": 0.6513,
+      "step": 245
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0002824308871364948,
+      "loss": 0.5354,
+      "step": 246
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00028226170482689017,
+      "loss": 0.7068,
+      "step": 247
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00028209176301642456,
+      "loss": 0.4029,
+      "step": 248
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00028192106268097334,
+      "loss": 0.5173,
+      "step": 249
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00028174960480076776,
+      "loss": 0.3557,
+      "step": 250
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0002815773903603891,
+      "loss": 0.2404,
+      "step": 251
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0002814044203487633,
+      "loss": 0.3994,
+      "step": 252
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0002812306957591549,
+      "loss": 0.4771,
+      "step": 253
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00028105621758916165,
+      "loss": 0.6123,
+      "step": 254
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00028088098684070866,
+      "loss": 0.6972,
+      "step": 255
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0002807050045200426,
+      "loss": 0.6911,
+      "step": 256
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00028052827163772603,
+      "loss": 0.3621,
+      "step": 257
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0002803507892086315,
+      "loss": 0.3574,
+      "step": 258
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0002801725582519358,
+      "loss": 0.6216,
+      "step": 259
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0002799935797911141,
+      "loss": 0.4743,
+      "step": 260
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00027981385485393386,
+      "loss": 0.473,
+      "step": 261
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00027963338447244937,
+      "loss": 0.5412,
+      "step": 262
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00027945216968299543,
+      "loss": 0.2539,
+      "step": 263
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0002792702115261816,
+      "loss": 0.1642,
+      "step": 264
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00027908751104688604,
+      "loss": 0.4496,
+      "step": 265
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0002789040692942497,
+      "loss": 0.7403,
+      "step": 266
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0002787198873216703,
+      "loss": 0.5017,
+      "step": 267
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00027853496618679614,
+      "loss": 0.7619,
+      "step": 268
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00027834930695152,
+      "loss": 0.7091,
+      "step": 269
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0002781629106819733,
+      "loss": 0.3183,
+      "step": 270
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00027797577844851964,
+      "loss": 0.3979,
+      "step": 271
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.000277787911325749,
+      "loss": 0.6251,
+      "step": 272
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00027759931039247134,
+      "loss": 0.5845,
+      "step": 273
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00027740997673171043,
+      "loss": 0.7068,
+      "step": 274
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00027721991143069767,
+      "loss": 0.372,
+      "step": 275
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0002770291155808657,
+      "loss": 0.8528,
+      "step": 276
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0002768375902778426,
+      "loss": 0.3525,
+      "step": 277
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00027664533662144497,
+      "loss": 0.2443,
+      "step": 278
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0002764523557156721,
+      "loss": 0.6492,
+      "step": 279
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00027625864866869925,
+      "loss": 0.568,
+      "step": 280
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00027606421659287164,
+      "loss": 0.5191,
+      "step": 281
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0002758690606046979,
+      "loss": 0.4665,
+      "step": 282
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0002756731818248436,
+      "loss": 0.9246,
+      "step": 283
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0002754765813781248,
+      "loss": 0.2443,
+      "step": 284
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00027527926039350186,
+      "loss": 0.5107,
+      "step": 285
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0002750812200040727,
+      "loss": 0.2779,
+      "step": 286
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0002748824613470662,
+      "loss": 0.3306,
+      "step": 287
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.000274682985563836,
+      "loss": 0.4641,
+      "step": 288
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00027448279379985375,
+      "loss": 0.3174,
+      "step": 289
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00027428188720470247,
+      "loss": 0.3527,
+      "step": 290
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00027408026693207015,
+      "loss": 0.3878,
+      "step": 291
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0002738779341397429,
+      "loss": 0.3586,
+      "step": 292
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00027367488998959847,
+      "loss": 0.5559,
+      "step": 293
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00027347113564759945,
+      "loss": 0.739,
+      "step": 294
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00027326667228378673,
+      "loss": 0.2517,
+      "step": 295
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0002730615010722727,
+      "loss": 0.3284,
+      "step": 296
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00027285562319123447,
+      "loss": 0.7683,
+      "step": 297
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00027264903982290704,
+      "loss": 0.777,
+      "step": 298
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00027244175215357673,
+      "loss": 0.5519,
+      "step": 299
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00027223376137357425,
+      "loss": 0.6937,
+      "step": 300
+    },
+    {
+      "epoch": 0.22,
+      "eval_loss": 0.7280155420303345,
+      "eval_runtime": 13.0039,
+      "eval_samples_per_second": 0.846,
+      "eval_steps_per_second": 0.846,
+      "step": 300
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0002720250686772677,
+      "loss": 0.5711,
+      "step": 301
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0002718156752630561,
+      "loss": 0.2999,
+      "step": 302
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0002716055823333622,
+      "loss": 0.551,
+      "step": 303
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0002713947910946255,
+      "loss": 0.5503,
+      "step": 304
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00027118330275729567,
+      "loss": 0.3301,
+      "step": 305
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00027097111853582526,
+      "loss": 0.4078,
+      "step": 306
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00027075823964866307,
+      "loss": 0.2179,
+      "step": 307
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0002705446673182467,
+      "loss": 0.2818,
+      "step": 308
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.000270330402770996,
+      "loss": 0.2603,
+      "step": 309
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0002701154472373057,
+      "loss": 0.3626,
+      "step": 310
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0002698998019515385,
+      "loss": 0.3824,
+      "step": 311
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0002696834681520178,
+      "loss": 0.5088,
+      "step": 312
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0002694664470810211,
+      "loss": 0.5815,
+      "step": 313
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.000269248739984772,
+      "loss": 0.3339,
+      "step": 314
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00026903034811343393,
+      "loss": 0.4999,
+      "step": 315
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00026881127272110236,
+      "loss": 0.5024,
+      "step": 316
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00026859151506579795,
+      "loss": 0.5164,
+      "step": 317
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00026837107640945905,
+      "loss": 0.3513,
+      "step": 318
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0002681499580179347,
+      "loss": 0.2102,
+      "step": 319
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0002679281611609773,
+      "loss": 0.6105,
+      "step": 320
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00026770568711223514,
+      "loss": 0.4672,
+      "step": 321
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00026748253714924537,
+      "loss": 0.1856,
+      "step": 322
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00026725871255342635,
+      "loss": 0.2632,
+      "step": 323
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0002670342146100707,
+      "loss": 0.443,
+      "step": 324
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00026680904460833733,
+      "loss": 0.2832,
+      "step": 325
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0002665832038412447,
+      "loss": 0.4782,
+      "step": 326
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00026635669360566296,
+      "loss": 0.5187,
+      "step": 327
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0002661295152023066,
+      "loss": 0.3699,
+      "step": 328
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.000265901669935727,
+      "loss": 0.3705,
+      "step": 329
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00026567315911430497,
+      "loss": 0.4908,
+      "step": 330
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0002654439840502433,
+      "loss": 0.4913,
+      "step": 331
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.000265214146059559,
+      "loss": 0.3055,
+      "step": 332
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00026498364646207576,
+      "loss": 0.3872,
+      "step": 333
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00026475248658141686,
+      "loss": 0.3888,
+      "step": 334
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0002645206677449968,
+      "loss": 0.5704,
+      "step": 335
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0002642881912840144,
+      "loss": 0.4121,
+      "step": 336
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0002640550585334446,
+      "loss": 0.1917,
+      "step": 337
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0002638212708320312,
+      "loss": 0.7044,
+      "step": 338
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0002635868295222787,
+      "loss": 0.6185,
+      "step": 339
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00026335173595044527,
+      "loss": 0.2487,
+      "step": 340
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00026311599146653443,
+      "loss": 0.6175,
+      "step": 341
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00026287959742428744,
+      "loss": 0.4304,
+      "step": 342
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00026264255518117577,
+      "loss": 0.2045,
+      "step": 343
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00026240486609839297,
+      "loss": 0.5652,
+      "step": 344
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00026216653154084713,
+      "loss": 0.5883,
+      "step": 345
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0002619275528771528,
+      "loss": 0.2128,
+      "step": 346
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00026168793147962334,
+      "loss": 0.3772,
+      "step": 347
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0002614476687242629,
+      "loss": 0.4098,
+      "step": 348
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00026120676599075866,
+      "loss": 0.3191,
+      "step": 349
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00026096522466247265,
+      "loss": 0.7856,
+      "step": 350
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0002607230461264342,
+      "loss": 0.5807,
+      "step": 351
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0002604802317733315,
+      "loss": 0.3925,
+      "step": 352
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.000260236782997504,
+      "loss": 0.3968,
+      "step": 353
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0002599927011969343,
+      "loss": 0.6506,
+      "step": 354
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00025974798777324,
+      "loss": 0.9081,
+      "step": 355
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0002595026441316657,
+      "loss": 0.5228,
+      "step": 356
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0002592566716810752,
+      "loss": 0.3975,
+      "step": 357
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00025901007183394293,
+      "loss": 0.4149,
+      "step": 358
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0002587628460063462,
+      "loss": 0.3914,
+      "step": 359
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00025851499561795684,
+      "loss": 0.8105,
+      "step": 360
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0002582665220920334,
+      "loss": 0.6406,
+      "step": 361
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0002580174268554125,
+      "loss": 0.7106,
+      "step": 362
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.000257767711338501,
+      "loss": 0.484,
+      "step": 363
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0002575173769752677,
+      "loss": 0.375,
+      "step": 364
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0002572664252032349,
+      "loss": 0.2236,
+      "step": 365
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0002570148574634705,
+      "loss": 0.3452,
+      "step": 366
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0002567626752005794,
+      "loss": 0.2653,
+      "step": 367
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0002565098798626955,
+      "loss": 0.6595,
+      "step": 368
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00025625647290147303,
+      "loss": 0.4133,
+      "step": 369
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00025600245577207857,
+      "loss": 0.5276,
+      "step": 370
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0002557478299331825,
+      "loss": 0.3904,
+      "step": 371
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0002554925968469506,
+      "loss": 0.3788,
+      "step": 372
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00025523675797903576,
+      "loss": 0.3834,
+      "step": 373
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0002549803147985695,
+      "loss": 0.3585,
+      "step": 374
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00025472326877815354,
+      "loss": 0.2435,
+      "step": 375
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0002544656213938514,
+      "loss": 0.4661,
+      "step": 376
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0002542073741251799,
+      "loss": 0.6525,
+      "step": 377
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00025394852845510047,
+      "loss": 0.7845,
+      "step": 378
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.000253689085870011,
+      "loss": 1.0962,
+      "step": 379
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00025342904785973706,
+      "loss": 0.6062,
+      "step": 380
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00025316841591752336,
+      "loss": 0.5632,
+      "step": 381
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00025290719154002513,
+      "loss": 0.2323,
+      "step": 382
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0002526453762272998,
+      "loss": 0.6669,
+      "step": 383
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0002523829714827981,
+      "loss": 0.3308,
+      "step": 384
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00025211997881335543,
+      "loss": 0.641,
+      "step": 385
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00025185639972918346,
+      "loss": 0.5766,
+      "step": 386
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00025159223574386114,
+      "loss": 0.5476,
+      "step": 387
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00025132748837432625,
+      "loss": 0.3965,
+      "step": 388
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0002510621591408666,
+      "loss": 0.6232,
+      "step": 389
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0002507962495671113,
+      "loss": 0.4936,
+      "step": 390
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00025052976118002194,
+      "loss": 0.4642,
+      "step": 391
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00025026269550988396,
+      "loss": 0.488,
+      "step": 392
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0002499950540902978,
+      "loss": 0.8166,
+      "step": 393
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00024972683845817005,
+      "loss": 0.404,
+      "step": 394
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00024945805015370474,
+      "loss": 0.469,
+      "step": 395
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0002491886907203943,
+      "loss": 0.5544,
+      "step": 396
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0002489187617050109,
+      "loss": 0.3086,
+      "step": 397
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0002486482646575975,
+      "loss": 0.4426,
+      "step": 398
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00024837720113145886,
+      "loss": 0.5296,
+      "step": 399
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0002481055726831528,
+      "loss": 0.345,
+      "step": 400
+    },
+    {
+      "epoch": 0.3,
+      "eval_loss": 0.7337626814842224,
+      "eval_runtime": 13.0144,
+      "eval_samples_per_second": 0.845,
+      "eval_steps_per_second": 0.845,
+      "step": 400
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0002478333808724809,
+      "loss": 0.4889,
+      "step": 401
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0002475606272624801,
+      "loss": 0.2944,
+      "step": 402
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00024728731341941337,
+      "loss": 0.3656,
+      "step": 403
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0002470134409127606,
+      "loss": 0.4308,
+      "step": 404
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0002467390113152099,
+      "loss": 0.3265,
+      "step": 405
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00024646402620264834,
+      "loss": 0.166,
+      "step": 406
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00024618848715415306,
+      "loss": 0.4762,
+      "step": 407
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00024591239575198206,
+      "loss": 0.6291,
+      "step": 408
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0002456357535815652,
+      "loss": 0.5504,
+      "step": 409
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00024535856223149524,
+      "loss": 0.3328,
+      "step": 410
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0002450808232935181,
+      "loss": 0.5982,
+      "step": 411
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00024480253836252477,
+      "loss": 0.3765,
+      "step": 412
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00024452370903654114,
+      "loss": 0.5533,
+      "step": 413
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00024424433691671936,
+      "loss": 0.3661,
+      "step": 414
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00024396442360732852,
+      "loss": 0.3451,
+      "step": 415
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00024368397071574546,
+      "loss": 0.1914,
+      "step": 416
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00024340297985244556,
+      "loss": 0.2141,
+      "step": 417
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00024312145263099327,
+      "loss": 0.2406,
+      "step": 418
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00024283939066803333,
+      "loss": 0.5337,
+      "step": 419
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00024255679558328087,
+      "loss": 0.3128,
+      "step": 420
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00024227366899951265,
+      "loss": 0.2907,
+      "step": 421
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00024199001254255744,
+      "loss": 0.4188,
+      "step": 422
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0002417058278412867,
+      "loss": 0.4719,
+      "step": 423
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00024142111652760535,
+      "loss": 0.5886,
+      "step": 424
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00024113588023644237,
+      "loss": 0.5553,
+      "step": 425
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0002408501206057413,
+      "loss": 0.4051,
+      "step": 426
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00024056383927645094,
+      "loss": 0.4207,
+      "step": 427
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00024027703789251587,
+      "loss": 0.4121,
+      "step": 428
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0002399897181008671,
+      "loss": 0.4287,
+      "step": 429
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0002397018815514125,
+      "loss": 0.3167,
+      "step": 430
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00023941352989702733,
+      "loss": 0.4941,
+      "step": 431
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00023912466479354493,
+      "loss": 0.4612,
+      "step": 432
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.000238835287899747,
+      "loss": 0.6483,
+      "step": 433
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00023854540087735407,
+      "loss": 0.7222,
+      "step": 434
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00023825500539101615,
+      "loss": 0.5733,
+      "step": 435
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.000237964103108303,
+      "loss": 0.3045,
+      "step": 436
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0002376726956996947,
+      "loss": 0.6636,
+      "step": 437
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00023738078483857177,
+      "loss": 0.5089,
+      "step": 438
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00023708837220120602,
+      "loss": 0.2181,
+      "step": 439
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0002367954594667505,
+      "loss": 0.4506,
+      "step": 440
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00023650204831723008,
+      "loss": 0.6349,
+      "step": 441
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0002362081404375316,
+      "loss": 0.5496,
+      "step": 442
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00023591373751539454,
+      "loss": 0.6185,
+      "step": 443
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.000235618841241401,
+      "loss": 0.4223,
+      "step": 444
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00023532345330896617,
+      "loss": 0.3419,
+      "step": 445
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00023502757541432837,
+      "loss": 0.239,
+      "step": 446
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00023473120925653973,
+      "loss": 0.6236,
+      "step": 447
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00023443435653745605,
+      "loss": 0.6115,
+      "step": 448
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00023413701896172714,
+      "loss": 0.416,
+      "step": 449
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00023383919823678713,
+      "loss": 0.3178,
+      "step": 450
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0002335408960728445,
+      "loss": 0.3594,
+      "step": 451
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0002332421141828725,
+      "loss": 0.4471,
+      "step": 452
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00023294285428259904,
+      "loss": 0.2383,
+      "step": 453
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.000232643118090497,
+      "loss": 0.3883,
+      "step": 454
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0002323429073277743,
+      "loss": 0.5119,
+      "step": 455
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00023204222371836405,
+      "loss": 0.2078,
+      "step": 456
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0002317410689889146,
+      "loss": 0.3848,
+      "step": 457
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0002314394448687798,
+      "loss": 0.5993,
+      "step": 458
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00023113735309000886,
+      "loss": 0.539,
+      "step": 459
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00023083479538733634,
+      "loss": 0.4248,
+      "step": 460
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0002305317734981725,
+      "loss": 0.3614,
+      "step": 461
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00023022828916259306,
+      "loss": 0.5128,
+      "step": 462
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00022992434412332942,
+      "loss": 0.4425,
+      "step": 463
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0002296199401257584,
+      "loss": 0.3218,
+      "step": 464
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00022931507891789239,
+      "loss": 0.3231,
+      "step": 465
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00022900976225036923,
+      "loss": 0.7137,
+      "step": 466
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00022870399187644225,
+      "loss": 0.3365,
+      "step": 467
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00022839776955197014,
+      "loss": 0.9577,
+      "step": 468
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00022809109703540682,
+      "loss": 0.2926,
+      "step": 469
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0002277839760877915,
+      "loss": 0.2721,
+      "step": 470
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00022747640847273838,
+      "loss": 0.3,
+      "step": 471
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00022716839595642664,
+      "loss": 0.5151,
+      "step": 472
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0002268599403075903,
+      "loss": 0.4312,
+      "step": 473
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00022655104329750785,
+      "loss": 0.2957,
+      "step": 474
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00022624170669999252,
+      "loss": 0.5483,
+      "step": 475
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00022593193229138167,
+      "loss": 0.3159,
+      "step": 476
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00022562172185052677,
+      "loss": 0.2589,
+      "step": 477
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00022531107715878323,
+      "loss": 0.3596,
+      "step": 478
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.000225,
+      "loss": 0.2969,
+      "step": 479
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00022468849216050952,
+      "loss": 0.5015,
+      "step": 480
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00022437655542911729,
+      "loss": 0.6868,
+      "step": 481
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00022406419159709184,
+      "loss": 0.3607,
+      "step": 482
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00022375140245815405,
+      "loss": 0.801,
+      "step": 483
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00022343818980846738,
+      "loss": 0.3641,
+      "step": 484
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.000223124555446627,
+      "loss": 0.553,
+      "step": 485
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00022281050117364982,
+      "loss": 0.874,
+      "step": 486
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00022249602879296414,
+      "loss": 0.2427,
+      "step": 487
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0002221811401103991,
+      "loss": 0.2261,
+      "step": 488
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00022186583693417433,
+      "loss": 0.4712,
+      "step": 489
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00022155012107488988,
+      "loss": 0.6684,
+      "step": 490
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00022123399434551545,
+      "loss": 0.4416,
+      "step": 491
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00022091745856138014,
+      "loss": 0.349,
+      "step": 492
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.000220600515540162,
+      "loss": 0.2384,
+      "step": 493
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00022028316710187768,
+      "loss": 0.4482,
+      "step": 494
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00021996541506887185,
+      "loss": 0.4404,
+      "step": 495
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00021964726126580668,
+      "loss": 0.2518,
+      "step": 496
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00021932870751965167,
+      "loss": 0.2758,
+      "step": 497
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0002190097556596728,
+      "loss": 0.6988,
+      "step": 498
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0002186904075174223,
+      "loss": 0.3609,
+      "step": 499
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0002183706649267279,
+      "loss": 0.5053,
+      "step": 500
+    },
+    {
+      "epoch": 0.37,
+      "eval_loss": 0.7127830386161804,
+      "eval_runtime": 12.994,
+      "eval_samples_per_second": 0.847,
+      "eval_steps_per_second": 0.847,
+      "step": 500
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00021805052972368244,
+      "loss": 0.6729,
+      "step": 501
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0002177300037466334,
+      "loss": 0.529,
+      "step": 502
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00021740908883617217,
+      "loss": 0.5612,
+      "step": 503
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00021708778683512345,
+      "loss": 0.475,
+      "step": 504
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.000216766099588535,
+      "loss": 0.548,
+      "step": 505
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00021644402894366668,
+      "loss": 0.3165,
+      "step": 506
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00021612157674998,
+      "loss": 0.2605,
+      "step": 507
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00021579874485912743,
+      "loss": 0.4271,
+      "step": 508
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00021547553512494186,
+      "loss": 0.4293,
+      "step": 509
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00021515194940342598,
+      "loss": 0.4034,
+      "step": 510
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00021482798955274146,
+      "loss": 0.3694,
+      "step": 511
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0002145036574331984,
+      "loss": 0.1641,
+      "step": 512
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0002141789549072446,
+      "loss": 0.6786,
+      "step": 513
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00021385388383945493,
+      "loss": 0.5022,
+      "step": 514
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00021352844609652056,
+      "loss": 0.5273,
+      "step": 515
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00021320264354723824,
+      "loss": 0.6755,
+      "step": 516
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0002128764780624996,
+      "loss": 0.1676,
+      "step": 517
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00021254995151528041,
+      "loss": 0.4453,
+      "step": 518
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0002122230657806298,
+      "loss": 0.3622,
+      "step": 519
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00021189582273565942,
+      "loss": 0.8535,
+      "step": 520
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00021156822425953293,
+      "loss": 0.4528,
+      "step": 521
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00021124027223345487,
+      "loss": 0.3618,
+      "step": 522
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00021091196854066001,
+      "loss": 0.2572,
+      "step": 523
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00021058331506640255,
+      "loss": 0.5138,
+      "step": 524
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0002102543136979454,
+      "loss": 0.4255,
+      "step": 525
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00020992496632454908,
+      "loss": 0.6983,
+      "step": 526
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00020959527483746098,
+      "loss": 0.4843,
+      "step": 527
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00020926524112990465,
+      "loss": 0.708,
+      "step": 528
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00020893486709706887,
+      "loss": 0.5213,
+      "step": 529
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00020860415463609647,
+      "loss": 0.5286,
+      "step": 530
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00020827310564607384,
+      "loss": 0.6407,
+      "step": 531
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00020794172202801976,
+      "loss": 0.7612,
+      "step": 532
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0002076100056848748,
+      "loss": 0.2932,
+      "step": 533
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0002072779585214899,
+      "loss": 0.693,
+      "step": 534
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00020694558244461589,
+      "loss": 0.5049,
+      "step": 535
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00020661287936289213,
+      "loss": 0.2617,
+      "step": 536
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00020627985118683615,
+      "loss": 0.6146,
+      "step": 537
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00020594649982883198,
+      "loss": 0.2204,
+      "step": 538
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00020561282720311954,
+      "loss": 0.2412,
+      "step": 539
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0002052788352257838,
+      "loss": 0.6613,
+      "step": 540
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00020494452581474337,
+      "loss": 0.2795,
+      "step": 541
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00020460990088973975,
+      "loss": 0.2773,
+      "step": 542
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00020427496237232626,
+      "loss": 0.1297,
+      "step": 543
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00020393971218585704,
+      "loss": 0.5855,
+      "step": 544
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00020360415225547594,
+      "loss": 0.6115,
+      "step": 545
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00020326828450810544,
+      "loss": 0.6751,
+      "step": 546
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0002029321108724356,
+      "loss": 0.3769,
+      "step": 547
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00020259563327891316,
+      "loss": 0.4872,
+      "step": 548
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00020225885365973017,
+      "loss": 0.2956,
+      "step": 549
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00020192177394881317,
+      "loss": 0.1986,
+      "step": 550
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00020158439608181173,
+      "loss": 0.5691,
+      "step": 551
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00020124672199608787,
+      "loss": 0.4556,
+      "step": 552
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0002009087536307043,
+      "loss": 0.5757,
+      "step": 553
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00020057049292641386,
+      "loss": 0.5785,
+      "step": 554
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0002002319418256479,
+      "loss": 0.7793,
+      "step": 555
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00019989310227250555,
+      "loss": 0.6425,
+      "step": 556
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00019955397621274217,
+      "loss": 0.326,
+      "step": 557
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00019921456559375854,
+      "loss": 0.5331,
+      "step": 558
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001988748723645893,
+      "loss": 0.4126,
+      "step": 559
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00019853489847589212,
+      "loss": 0.2713,
+      "step": 560
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00019819464587993624,
+      "loss": 0.5311,
+      "step": 561
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00019785411653059138,
+      "loss": 0.1802,
+      "step": 562
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00019751331238331644,
+      "loss": 0.2863,
+      "step": 563
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00019717223539514846,
+      "loss": 0.2984,
+      "step": 564
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001968308875246911,
+      "loss": 0.3554,
+      "step": 565
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0001964892707321036,
+      "loss": 0.7062,
+      "step": 566
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00019614738697908946,
+      "loss": 0.5257,
+      "step": 567
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00019580523822888512,
+      "loss": 0.4245,
+      "step": 568
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00019546282644624885,
+      "loss": 0.6944,
+      "step": 569
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00019512015359744923,
+      "loss": 0.7338,
+      "step": 570
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00019477722165025418,
+      "loss": 0.3573,
+      "step": 571
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00019443403257391923,
+      "loss": 0.2873,
+      "step": 572
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00019409058833917665,
+      "loss": 0.7692,
+      "step": 573
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00019374689091822373,
+      "loss": 0.6185,
+      "step": 574
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00019340294228471192,
+      "loss": 0.2507,
+      "step": 575
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00019305874441373502,
+      "loss": 0.4036,
+      "step": 576
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00019271429928181816,
+      "loss": 0.2384,
+      "step": 577
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00019236960886690626,
+      "loss": 0.3416,
+      "step": 578
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001920246751483529,
+      "loss": 0.4421,
+      "step": 579
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00019167950010690868,
+      "loss": 0.3397,
+      "step": 580
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00019133408572470997,
+      "loss": 0.3132,
+      "step": 581
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00019098843398526765,
+      "loss": 0.4452,
+      "step": 582
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00019064254687345544,
+      "loss": 0.3581,
+      "step": 583
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00019029642637549885,
+      "loss": 0.2781,
+      "step": 584
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0001899500744789634,
+      "loss": 0.217,
+      "step": 585
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00018960349317274355,
+      "loss": 0.6642,
+      "step": 586
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00018925668444705104,
+      "loss": 0.7822,
+      "step": 587
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00018890965029340353,
+      "loss": 0.394,
+      "step": 588
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001885623927046133,
+      "loss": 0.6322,
+      "step": 589
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001882149136747755,
+      "loss": 0.3221,
+      "step": 590
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001878672151992571,
+      "loss": 0.5053,
+      "step": 591
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00018751929927468504,
+      "loss": 0.2439,
+      "step": 592
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.000187171167898935,
+      "loss": 0.2368,
+      "step": 593
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00018682282307111987,
+      "loss": 0.541,
+      "step": 594
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001864742667915783,
+      "loss": 0.2163,
+      "step": 595
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00018612550106186318,
+      "loss": 0.3043,
+      "step": 596
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00018577652788473003,
+      "loss": 0.4656,
+      "step": 597
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001854273492641258,
+      "loss": 0.5907,
+      "step": 598
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0001850779672051771,
+      "loss": 0.2994,
+      "step": 599
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00018472838371417873,
+      "loss": 0.4392,
+      "step": 600
+    },
+    {
+      "epoch": 0.44,
+      "eval_loss": 0.7241503000259399,
+      "eval_runtime": 12.9988,
+      "eval_samples_per_second": 0.846,
+      "eval_steps_per_second": 0.846,
+      "step": 600
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00018437860079858226,
+      "loss": 0.1327,
+      "step": 601
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001840286204669843,
+      "loss": 0.7039,
+      "step": 602
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00018367844472911532,
+      "loss": 0.3143,
+      "step": 603
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00018332807559582774,
+      "loss": 0.4041,
+      "step": 604
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00018297751507908452,
+      "loss": 0.4026,
+      "step": 605
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00018262676519194778,
+      "loss": 0.3177,
+      "step": 606
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001822758279485669,
+      "loss": 0.4313,
+      "step": 607
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00018192470536416728,
+      "loss": 0.5299,
+      "step": 608
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00018157339945503847,
+      "loss": 0.3415,
+      "step": 609
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001812219122385229,
+      "loss": 0.4785,
+      "step": 610
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001808702457330041,
+      "loss": 0.373,
+      "step": 611
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00018051840195789506,
+      "loss": 0.7851,
+      "step": 612
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00018016638293362683,
+      "loss": 0.4122,
+      "step": 613
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0001798141906816368,
+      "loss": 0.473,
+      "step": 614
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00017946182722435703,
+      "loss": 0.5523,
+      "step": 615
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00017910929458520274,
+      "loss": 0.429,
+      "step": 616
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00017875659478856076,
+      "loss": 0.7252,
+      "step": 617
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001784037298597777,
+      "loss": 0.8911,
+      "step": 618
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001780507018251484,
+      "loss": 0.2871,
+      "step": 619
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00017769751271190444,
+      "loss": 0.3168,
+      "step": 620
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00017734416454820225,
+      "loss": 0.6512,
+      "step": 621
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00017699065936311174,
+      "loss": 0.7975,
+      "step": 622
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00017663699918660442,
+      "loss": 0.6918,
+      "step": 623
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00017628318604954175,
+      "loss": 0.5236,
+      "step": 624
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00017592922198366364,
+      "loss": 0.2322,
+      "step": 625
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001755751090215767,
+      "loss": 0.4731,
+      "step": 626
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00017522084919674254,
+      "loss": 0.2575,
+      "step": 627
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001748664445434661,
+      "loss": 0.2666,
+      "step": 628
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00017451189709688403,
+      "loss": 0.1195,
+      "step": 629
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00017415720889295294,
+      "loss": 0.4373,
+      "step": 630
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00017380238196843767,
+      "loss": 0.3413,
+      "step": 631
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00017344741836089974,
+      "loss": 0.7706,
+      "step": 632
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001730923201086855,
+      "loss": 0.479,
+      "step": 633
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00017273708925091453,
+      "loss": 0.3097,
+      "step": 634
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00017238172782746782,
+      "loss": 0.3742,
+      "step": 635
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00017202623787897616,
+      "loss": 0.4239,
+      "step": 636
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001716706214468084,
+      "loss": 0.5373,
+      "step": 637
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00017131488057305967,
+      "loss": 0.3712,
+      "step": 638
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00017095901730053976,
+      "loss": 0.4058,
+      "step": 639
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00017060303367276121,
+      "loss": 0.1766,
+      "step": 640
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00017024693173392784,
+      "loss": 0.8582,
+      "step": 641
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00016989071352892273,
+      "loss": 0.3212,
+      "step": 642
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00016953438110329675,
+      "loss": 0.1797,
+      "step": 643
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00016917793650325644,
+      "loss": 0.6375,
+      "step": 644
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00016882138177565272,
+      "loss": 0.9199,
+      "step": 645
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001684647189679688,
+      "loss": 0.3762,
+      "step": 646
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001681079501283086,
+      "loss": 0.5697,
+      "step": 647
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00016775107730538475,
+      "loss": 0.2191,
+      "step": 648
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00016739410254850725,
+      "loss": 0.3447,
+      "step": 649
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0001670370279075712,
+      "loss": 0.7809,
+      "step": 650
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00016667985543304548,
+      "loss": 0.5296,
+      "step": 651
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00016632258717596057,
+      "loss": 0.3256,
+      "step": 652
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00016596522518789718,
+      "loss": 0.5935,
+      "step": 653
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00016560777152097413,
+      "loss": 0.4552,
+      "step": 654
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00016525022822783667,
+      "loss": 0.6108,
+      "step": 655
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00016489259736164483,
+      "loss": 0.2999,
+      "step": 656
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001645348809760615,
+      "loss": 0.3985,
+      "step": 657
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00016417708112524056,
+      "loss": 0.3948,
+      "step": 658
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00016381919986381525,
+      "loss": 0.305,
+      "step": 659
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00016346123924688638,
+      "loss": 0.5075,
+      "step": 660
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00016310320133001028,
+      "loss": 0.4921,
+      "step": 661
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00016274508816918734,
+      "loss": 0.352,
+      "step": 662
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00016238690182084986,
+      "loss": 0.1758,
+      "step": 663
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00016202864434185053,
+      "loss": 0.5357,
+      "step": 664
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00016167031778945056,
+      "loss": 0.3795,
+      "step": 665
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00016131192422130764,
+      "loss": 0.3828,
+      "step": 666
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001609534656954644,
+      "loss": 0.4688,
+      "step": 667
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00016059494427033647,
+      "loss": 0.5511,
+      "step": 668
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00016023636200470065,
+      "loss": 0.6362,
+      "step": 669
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001598777209576831,
+      "loss": 0.3887,
+      "step": 670
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00015951902318874758,
+      "loss": 0.5006,
+      "step": 671
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00015916027075768354,
+      "loss": 0.5082,
+      "step": 672
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00015880146572459433,
+      "loss": 0.3726,
+      "step": 673
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00015844261014988535,
+      "loss": 0.5567,
+      "step": 674
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0001580837060942522,
+      "loss": 0.7998,
+      "step": 675
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.000157724755618669,
+      "loss": 0.2156,
+      "step": 676
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00015736576078437625,
+      "loss": 0.3988,
+      "step": 677
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00015700672365286932,
+      "loss": 0.898,
+      "step": 678
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00015664764628588637,
+      "loss": 0.3017,
+      "step": 679
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00015628853074539675,
+      "loss": 0.4727,
+      "step": 680
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00015592937909358887,
+      "loss": 0.9664,
+      "step": 681
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00015557019339285853,
+      "loss": 0.4321,
+      "step": 682
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00015521097570579716,
+      "loss": 0.3144,
+      "step": 683
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00015485172809517978,
+      "loss": 0.352,
+      "step": 684
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00015449245262395323,
+      "loss": 0.9209,
+      "step": 685
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00015413315135522432,
+      "loss": 0.3722,
+      "step": 686
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00015377382635224806,
+      "loss": 0.8169,
+      "step": 687
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00015341447967841582,
+      "loss": 0.586,
+      "step": 688
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00015305511339724331,
+      "loss": 0.4594,
+      "step": 689
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00015269572957235877,
+      "loss": 0.4635,
+      "step": 690
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00015233633026749132,
+      "loss": 0.4394,
+      "step": 691
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00015197691754645893,
+      "loss": 0.7235,
+      "step": 692
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00015161749347315657,
+      "loss": 0.4175,
+      "step": 693
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0001512580601115444,
+      "loss": 0.2788,
+      "step": 694
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00015089861952563604,
+      "loss": 0.4044,
+      "step": 695
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00015053917377948642,
+      "loss": 0.4199,
+      "step": 696
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001501797249371802,
+      "loss": 0.2754,
+      "step": 697
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001498202750628198,
+      "loss": 0.3196,
+      "step": 698
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00014946082622051352,
+      "loss": 0.2727,
+      "step": 699
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00014910138047436394,
+      "loss": 0.4222,
+      "step": 700
+    },
+    {
+      "epoch": 0.52,
+      "eval_loss": 0.718849241733551,
+      "eval_runtime": 12.9984,
+      "eval_samples_per_second": 0.846,
+      "eval_steps_per_second": 0.846,
+      "step": 700
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00014874193988845553,
+      "loss": 0.5825,
+      "step": 701
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00014838250652684344,
+      "loss": 0.4383,
+      "step": 702
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00014802308245354104,
+      "loss": 0.3608,
+      "step": 703
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00014766366973250865,
+      "loss": 0.6402,
+      "step": 704
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001473042704276412,
+      "loss": 0.6717,
+      "step": 705
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00014694488660275666,
+      "loss": 0.4606,
+      "step": 706
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00014658552032158412,
+      "loss": 0.3563,
+      "step": 707
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001462261736477519,
+      "loss": 0.3345,
+      "step": 708
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001458668486447757,
+      "loss": 0.2947,
+      "step": 709
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001455075473760468,
+      "loss": 0.4221,
+      "step": 710
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00014514827190482022,
+      "loss": 0.7063,
+      "step": 711
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001447890242942028,
+      "loss": 0.4138,
+      "step": 712
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00014442980660714147,
+      "loss": 0.2716,
+      "step": 713
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001440706209064111,
+      "loss": 0.522,
+      "step": 714
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00014371146925460325,
+      "loss": 0.2471,
+      "step": 715
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00014335235371411363,
+      "loss": 0.2515,
+      "step": 716
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00014299327634713068,
+      "loss": 0.325,
+      "step": 717
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00014263423921562375,
+      "loss": 0.3559,
+      "step": 718
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.000142275244381331,
+      "loss": 0.311,
+      "step": 719
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001419162939057478,
+      "loss": 0.3588,
+      "step": 720
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00014155738985011462,
+      "loss": 0.4606,
+      "step": 721
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00014119853427540564,
+      "loss": 0.4991,
+      "step": 722
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00014083972924231643,
+      "loss": 0.42,
+      "step": 723
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00014048097681125242,
+      "loss": 0.2433,
+      "step": 724
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00014012227904231688,
+      "loss": 0.5278,
+      "step": 725
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00013976363799529936,
+      "loss": 0.4191,
+      "step": 726
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00013940505572966353,
+      "loss": 0.6402,
+      "step": 727
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001390465343045356,
+      "loss": 0.2295,
+      "step": 728
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00013868807577869233,
+      "loss": 0.3456,
+      "step": 729
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00013832968221054944,
+      "loss": 0.3635,
+      "step": 730
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00013797135565814944,
+      "loss": 0.4169,
+      "step": 731
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00013761309817915014,
+      "loss": 0.441,
+      "step": 732
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001372549118308127,
+      "loss": 0.4768,
+      "step": 733
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00013689679866998972,
+      "loss": 0.8404,
+      "step": 734
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00013653876075311365,
+      "loss": 0.5179,
+      "step": 735
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00013618080013618473,
+      "loss": 0.4933,
+      "step": 736
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00013582291887475947,
+      "loss": 0.4747,
+      "step": 737
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00013546511902393852,
+      "loss": 0.8171,
+      "step": 738
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00013510740263835517,
+      "loss": 0.4309,
+      "step": 739
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00013474977177216333,
+      "loss": 0.304,
+      "step": 740
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001343922284790259,
+      "loss": 0.469,
+      "step": 741
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00013403477481210285,
+      "loss": 0.5643,
+      "step": 742
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00013367741282403937,
+      "loss": 0.9558,
+      "step": 743
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001333201445669545,
+      "loss": 0.3965,
+      "step": 744
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00013296297209242872,
+      "loss": 0.409,
+      "step": 745
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00013260589745149273,
+      "loss": 0.5015,
+      "step": 746
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001322489226946152,
+      "loss": 0.4645,
+      "step": 747
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001318920498716914,
+      "loss": 0.2994,
+      "step": 748
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00013153528103203115,
+      "loss": 0.4572,
+      "step": 749
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00013117861822434723,
+      "loss": 0.556,
+      "step": 750
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00013082206349674354,
+      "loss": 0.4779,
+      "step": 751
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00013046561889670327,
+      "loss": 0.3453,
+      "step": 752
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001301092864710772,
+      "loss": 1.0891,
+      "step": 753
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00012975306826607213,
+      "loss": 0.5277,
+      "step": 754
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00012939696632723876,
+      "loss": 0.3465,
+      "step": 755
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00012904098269946024,
+      "loss": 0.3665,
+      "step": 756
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001286851194269403,
+      "loss": 0.4635,
+      "step": 757
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00012832937855319158,
+      "loss": 0.4174,
+      "step": 758
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00012797376212102387,
+      "loss": 0.847,
+      "step": 759
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00012761827217253218,
+      "loss": 0.2295,
+      "step": 760
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00012726291074908544,
+      "loss": 0.6366,
+      "step": 761
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00012690767989131447,
+      "loss": 0.2403,
+      "step": 762
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00012655258163910024,
+      "loss": 0.6297,
+      "step": 763
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001261976180315623,
+      "loss": 0.4264,
+      "step": 764
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00012584279110704704,
+      "loss": 0.7399,
+      "step": 765
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00012548810290311595,
+      "loss": 0.4965,
+      "step": 766
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00012513355545653391,
+      "loss": 0.5037,
+      "step": 767
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00012477915080325743,
+      "loss": 0.4081,
+      "step": 768
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001244248909784233,
+      "loss": 0.447,
+      "step": 769
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00012407077801633636,
+      "loss": 0.5114,
+      "step": 770
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00012371681395045826,
+      "loss": 0.4465,
+      "step": 771
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00012336300081339558,
+      "loss": 0.6708,
+      "step": 772
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00012300934063688823,
+      "loss": 0.4611,
+      "step": 773
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00012265583545179772,
+      "loss": 0.3486,
+      "step": 774
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00012230248728809556,
+      "loss": 0.4921,
+      "step": 775
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001219492981748516,
+      "loss": 0.4282,
+      "step": 776
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00012159627014022233,
+      "loss": 0.7954,
+      "step": 777
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00012124340521143926,
+      "loss": 0.6251,
+      "step": 778
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00012089070541479726,
+      "loss": 0.4673,
+      "step": 779
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.000120538172775643,
+      "loss": 0.3138,
+      "step": 780
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00012018580931836324,
+      "loss": 0.446,
+      "step": 781
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00011983361706637317,
+      "loss": 0.5042,
+      "step": 782
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00011948159804210495,
+      "loss": 0.3662,
+      "step": 783
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00011912975426699593,
+      "loss": 0.4054,
+      "step": 784
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001187780877614771,
+      "loss": 0.4903,
+      "step": 785
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00011842660054496153,
+      "loss": 0.1645,
+      "step": 786
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00011807529463583275,
+      "loss": 0.3971,
+      "step": 787
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00011772417205143306,
+      "loss": 0.4045,
+      "step": 788
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001173732348080522,
+      "loss": 0.3736,
+      "step": 789
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00011702248492091544,
+      "loss": 0.3747,
+      "step": 790
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00011667192440417226,
+      "loss": 0.4383,
+      "step": 791
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00011632155527088465,
+      "loss": 0.2069,
+      "step": 792
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00011597137953301566,
+      "loss": 0.3446,
+      "step": 793
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00011562139920141774,
+      "loss": 0.4467,
+      "step": 794
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001152716162858212,
+      "loss": 0.3552,
+      "step": 795
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00011492203279482283,
+      "loss": 0.3891,
+      "step": 796
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00011457265073587415,
+      "loss": 0.2504,
+      "step": 797
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00011422347211526993,
+      "loss": 0.4427,
+      "step": 798
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001138744989381368,
+      "loss": 0.4597,
+      "step": 799
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00011352573320842167,
+      "loss": 0.6739,
+      "step": 800
+    },
+    {
+      "epoch": 0.59,
+      "eval_loss": 0.6845341324806213,
+      "eval_runtime": 13.0104,
+      "eval_samples_per_second": 0.845,
+      "eval_steps_per_second": 0.845,
+      "step": 800
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00011317717692888012,
+      "loss": 0.3329,
+      "step": 801
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00011282883210106502,
+      "loss": 0.3141,
+      "step": 802
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00011248070072531497,
+      "loss": 0.6187,
+      "step": 803
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00011213278480074288,
+      "loss": 0.3025,
+      "step": 804
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00011178508632522448,
+      "loss": 0.3407,
+      "step": 805
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00011143760729538668,
+      "loss": 0.2956,
+      "step": 806
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00011109034970659644,
+      "loss": 0.269,
+      "step": 807
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00011074331555294897,
+      "loss": 0.2612,
+      "step": 808
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00011039650682725646,
+      "loss": 0.3399,
+      "step": 809
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00011004992552103656,
+      "loss": 0.4474,
+      "step": 810
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00010970357362450115,
+      "loss": 0.6197,
+      "step": 811
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00010935745312654454,
+      "loss": 0.2347,
+      "step": 812
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00010901156601473236,
+      "loss": 0.5961,
+      "step": 813
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00010866591427528999,
+      "loss": 0.5043,
+      "step": 814
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00010832049989309131,
+      "loss": 0.5384,
+      "step": 815
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00010797532485164707,
+      "loss": 0.5576,
+      "step": 816
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00010763039113309371,
+      "loss": 0.4151,
+      "step": 817
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00010728570071818184,
+      "loss": 0.281,
+      "step": 818
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00010694125558626498,
+      "loss": 0.389,
+      "step": 819
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001065970577152881,
+      "loss": 0.3531,
+      "step": 820
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00010625310908177625,
+      "loss": 0.2959,
+      "step": 821
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001059094116608234,
+      "loss": 0.4179,
+      "step": 822
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00010556596742608078,
+      "loss": 0.4439,
+      "step": 823
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00010522277834974585,
+      "loss": 0.2222,
+      "step": 824
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00010487984640255073,
+      "loss": 0.5375,
+      "step": 825
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00010453717355375116,
+      "loss": 0.3486,
+      "step": 826
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00010419476177111488,
+      "loss": 0.4632,
+      "step": 827
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00010385261302091059,
+      "loss": 0.526,
+      "step": 828
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00010351072926789641,
+      "loss": 0.3489,
+      "step": 829
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00010316911247530891,
+      "loss": 0.3223,
+      "step": 830
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00010282776460485157,
+      "loss": 0.9132,
+      "step": 831
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00010248668761668357,
+      "loss": 0.7139,
+      "step": 832
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00010214588346940867,
+      "loss": 0.288,
+      "step": 833
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00010180535412006374,
+      "loss": 0.2817,
+      "step": 834
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00010146510152410786,
+      "loss": 0.4616,
+      "step": 835
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00010112512763541068,
+      "loss": 0.6379,
+      "step": 836
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00010078543440624147,
+      "loss": 0.3136,
+      "step": 837
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00010044602378725777,
+      "loss": 0.2365,
+      "step": 838
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00010010689772749444,
+      "loss": 0.3048,
+      "step": 839
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 9.976805817435207e-05,
+      "loss": 0.3529,
+      "step": 840
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 9.942950707358615e-05,
+      "loss": 0.342,
+      "step": 841
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 9.909124636929564e-05,
+      "loss": 0.444,
+      "step": 842
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 9.875327800391212e-05,
+      "loss": 0.3445,
+      "step": 843
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 9.841560391818823e-05,
+      "loss": 0.3177,
+      "step": 844
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 9.807822605118682e-05,
+      "loss": 0.244,
+      "step": 845
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.774114634026979e-05,
+      "loss": 0.8378,
+      "step": 846
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.740436672108685e-05,
+      "loss": 0.4002,
+      "step": 847
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.706788912756442e-05,
+      "loss": 0.516,
+      "step": 848
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.673171549189457e-05,
+      "loss": 0.4076,
+      "step": 849
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.639584774452405e-05,
+      "loss": 0.3068,
+      "step": 850
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.606028781414292e-05,
+      "loss": 0.3099,
+      "step": 851
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.572503762767372e-05,
+      "loss": 0.5688,
+      "step": 852
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.539009911026025e-05,
+      "loss": 0.4554,
+      "step": 853
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.505547418525664e-05,
+      "loss": 0.2936,
+      "step": 854
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.472116477421619e-05,
+      "loss": 0.2868,
+      "step": 855
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.438717279688044e-05,
+      "loss": 0.7902,
+      "step": 856
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.405350017116802e-05,
+      "loss": 0.5916,
+      "step": 857
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.372014881316385e-05,
+      "loss": 0.1412,
+      "step": 858
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 9.338712063710785e-05,
+      "loss": 0.8575,
+      "step": 859
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 9.305441755538413e-05,
+      "loss": 0.593,
+      "step": 860
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 9.272204147851008e-05,
+      "loss": 0.1031,
+      "step": 861
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 9.238999431512519e-05,
+      "loss": 0.2781,
+      "step": 862
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 9.20582779719802e-05,
+      "loss": 0.3534,
+      "step": 863
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 9.172689435392617e-05,
+      "loss": 0.6448,
+      "step": 864
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 9.139584536390353e-05,
+      "loss": 0.164,
+      "step": 865
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 9.106513290293113e-05,
+      "loss": 0.3111,
+      "step": 866
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 9.073475887009532e-05,
+      "loss": 0.4406,
+      "step": 867
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 9.040472516253901e-05,
+      "loss": 0.5083,
+      "step": 868
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 9.007503367545094e-05,
+      "loss": 0.3334,
+      "step": 869
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 8.97456863020546e-05,
+      "loss": 0.6558,
+      "step": 870
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 8.941668493359741e-05,
+      "loss": 0.3639,
+      "step": 871
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 8.908803145934003e-05,
+      "loss": 0.5976,
+      "step": 872
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 8.875972776654517e-05,
+      "loss": 0.5383,
+      "step": 873
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 8.843177574046708e-05,
+      "loss": 0.3301,
+      "step": 874
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 8.810417726434054e-05,
+      "loss": 0.3455,
+      "step": 875
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 8.777693421937022e-05,
+      "loss": 0.296,
+      "step": 876
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 8.745004848471957e-05,
+      "loss": 0.2113,
+      "step": 877
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 8.712352193750043e-05,
+      "loss": 0.1236,
+      "step": 878
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 8.679735645276172e-05,
+      "loss": 0.3651,
+      "step": 879
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 8.647155390347942e-05,
+      "loss": 0.3829,
+      "step": 880
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 8.614611616054502e-05,
+      "loss": 0.55,
+      "step": 881
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 8.582104509275534e-05,
+      "loss": 0.2811,
+      "step": 882
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 8.549634256680157e-05,
+      "loss": 0.2337,
+      "step": 883
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 8.517201044725847e-05,
+      "loss": 0.2605,
+      "step": 884
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 8.4848050596574e-05,
+      "loss": 0.5998,
+      "step": 885
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 8.452446487505808e-05,
+      "loss": 0.3719,
+      "step": 886
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 8.420125514087253e-05,
+      "loss": 0.3653,
+      "step": 887
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 8.387842325002e-05,
+      "loss": 0.7566,
+      "step": 888
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 8.355597105633328e-05,
+      "loss": 0.1803,
+      "step": 889
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 8.323390041146493e-05,
+      "loss": 0.3518,
+      "step": 890
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 8.291221316487653e-05,
+      "loss": 0.5731,
+      "step": 891
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 8.259091116382784e-05,
+      "loss": 0.686,
+      "step": 892
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 8.226999625336662e-05,
+      "loss": 0.1914,
+      "step": 893
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 8.194947027631756e-05,
+      "loss": 0.4432,
+      "step": 894
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 8.162933507327212e-05,
+      "loss": 0.3462,
+      "step": 895
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 8.130959248257772e-05,
+      "loss": 0.2411,
+      "step": 896
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 8.099024434032717e-05,
+      "loss": 0.7422,
+      "step": 897
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 8.067129248034832e-05,
+      "loss": 0.5798,
+      "step": 898
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 8.035273873419333e-05,
+      "loss": 0.2258,
+      "step": 899
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 8.003458493112816e-05,
+      "loss": 0.4932,
+      "step": 900
+    },
+    {
+      "epoch": 0.67,
+      "eval_loss": 0.6635782122612,
+      "eval_runtime": 13.0807,
+      "eval_samples_per_second": 0.841,
+      "eval_steps_per_second": 0.841,
+      "step": 900
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 7.971683289812229e-05,
+      "loss": 0.4165,
+      "step": 901
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 7.939948445983799e-05,
+      "loss": 0.3897,
+      "step": 902
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 7.908254143861986e-05,
+      "loss": 0.5963,
+      "step": 903
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 7.876600565448457e-05,
+      "loss": 0.5027,
+      "step": 904
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 7.844987892511011e-05,
+      "loss": 0.287,
+      "step": 905
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 7.813416306582565e-05,
+      "loss": 0.7406,
+      "step": 906
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 7.781885988960096e-05,
+      "loss": 0.3224,
+      "step": 907
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 7.750397120703583e-05,
+      "loss": 0.3747,
+      "step": 908
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 7.718949882635011e-05,
+      "loss": 1.083,
+      "step": 909
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 7.6875444553373e-05,
+      "loss": 0.4304,
+      "step": 910
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 7.656181019153258e-05,
+      "loss": 0.6364,
+      "step": 911
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 7.624859754184588e-05,
+      "loss": 0.4104,
+      "step": 912
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 7.593580840290818e-05,
+      "loss": 0.3649,
+      "step": 913
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 7.562344457088269e-05,
+      "loss": 1.0562,
+      "step": 914
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 7.531150783949052e-05,
+      "loss": 0.3616,
+      "step": 915
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 7.500000000000002e-05,
+      "loss": 0.2649,
+      "step": 916
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 7.468892284121677e-05,
+      "loss": 0.5163,
+      "step": 917
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 7.437827814947323e-05,
+      "loss": 0.2815,
+      "step": 918
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 7.406806770861833e-05,
+      "loss": 0.4637,
+      "step": 919
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 7.375829330000747e-05,
+      "loss": 0.1545,
+      "step": 920
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 7.344895670249218e-05,
+      "loss": 0.4964,
+      "step": 921
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 7.314005969240974e-05,
+      "loss": 0.2742,
+      "step": 922
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 7.283160404357333e-05,
+      "loss": 0.7472,
+      "step": 923
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 7.252359152726156e-05,
+      "loss": 0.4042,
+      "step": 924
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 7.221602391220844e-05,
+      "loss": 0.324,
+      "step": 925
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 7.190890296459316e-05,
+      "loss": 0.616,
+      "step": 926
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 7.160223044802984e-05,
+      "loss": 0.3656,
+      "step": 927
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 7.129600812355775e-05,
+      "loss": 0.6666,
+      "step": 928
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 7.099023774963076e-05,
+      "loss": 0.1535,
+      "step": 929
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 7.068492108210756e-05,
+      "loss": 0.2176,
+      "step": 930
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 7.038005987424158e-05,
+      "loss": 0.5321,
+      "step": 931
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 7.007565587667053e-05,
+      "loss": 0.1454,
+      "step": 932
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.977171083740686e-05,
+      "loss": 0.2402,
+      "step": 933
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.946822650182751e-05,
+      "loss": 0.6484,
+      "step": 934
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.916520461266366e-05,
+      "loss": 0.442,
+      "step": 935
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.886264690999114e-05,
+      "loss": 0.1063,
+      "step": 936
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.856055513122017e-05,
+      "loss": 0.454,
+      "step": 937
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.825893101108536e-05,
+      "loss": 0.2908,
+      "step": 938
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.795777628163599e-05,
+      "loss": 0.2057,
+      "step": 939
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 6.765709267222571e-05,
+      "loss": 0.5314,
+      "step": 940
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.735688190950297e-05,
+      "loss": 0.5713,
+      "step": 941
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.705714571740096e-05,
+      "loss": 0.5988,
+      "step": 942
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.675788581712747e-05,
+      "loss": 0.2251,
+      "step": 943
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.645910392715544e-05,
+      "loss": 0.3555,
+      "step": 944
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.616080176321291e-05,
+      "loss": 0.5266,
+      "step": 945
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.586298103827286e-05,
+      "loss": 0.235,
+      "step": 946
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.556564346254394e-05,
+      "loss": 0.158,
+      "step": 947
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.526879074346027e-05,
+      "loss": 0.5409,
+      "step": 948
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.49724245856716e-05,
+      "loss": 0.3644,
+      "step": 949
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.467654669103386e-05,
+      "loss": 0.447,
+      "step": 950
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.438115875859898e-05,
+      "loss": 0.3026,
+      "step": 951
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.408626248460542e-05,
+      "loss": 0.2897,
+      "step": 952
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 6.37918595624684e-05,
+      "loss": 0.311,
+      "step": 953
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 6.349795168276994e-05,
+      "loss": 0.5142,
+      "step": 954
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 6.320454053324945e-05,
+      "loss": 0.3723,
+      "step": 955
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 6.291162779879396e-05,
+      "loss": 0.2556,
+      "step": 956
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 6.261921516142819e-05,
+      "loss": 0.6081,
+      "step": 957
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 6.232730430030535e-05,
+      "loss": 0.5688,
+      "step": 958
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 6.2035896891697e-05,
+      "loss": 0.5,
+      "step": 959
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 6.174499460898383e-05,
+      "loss": 0.5152,
+      "step": 960
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 6.145459912264596e-05,
+      "loss": 0.3877,
+      "step": 961
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 6.1164712100253e-05,
+      "loss": 0.6429,
+      "step": 962
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 6.0875335206455044e-05,
+      "loss": 0.2803,
+      "step": 963
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 6.0586470102972675e-05,
+      "loss": 0.6748,
+      "step": 964
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 6.0298118448587524e-05,
+      "loss": 0.4075,
+      "step": 965
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 6.0010281899132885e-05,
+      "loss": 0.4328,
+      "step": 966
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.972296210748414e-05,
+      "loss": 0.2727,
+      "step": 967
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.943616072354907e-05,
+      "loss": 0.5985,
+      "step": 968
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.914987939425868e-05,
+      "loss": 0.2922,
+      "step": 969
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.886411976355757e-05,
+      "loss": 0.1786,
+      "step": 970
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.857888347239457e-05,
+      "loss": 0.4119,
+      "step": 971
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.8294172158713285e-05,
+      "loss": 0.3754,
+      "step": 972
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.800998745744253e-05,
+      "loss": 0.4393,
+      "step": 973
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.7726331000487343e-05,
+      "loss": 0.1773,
+      "step": 974
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.74432044167191e-05,
+      "loss": 0.3433,
+      "step": 975
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.7160609331966646e-05,
+      "loss": 0.4045,
+      "step": 976
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.68785473690067e-05,
+      "loss": 0.5428,
+      "step": 977
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.659702014755443e-05,
+      "loss": 0.2591,
+      "step": 978
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.631602928425448e-05,
+      "loss": 0.4271,
+      "step": 979
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.603557639267146e-05,
+      "loss": 0.4947,
+      "step": 980
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.5755663083280616e-05,
+      "loss": 0.6009,
+      "step": 981
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.547629096345886e-05,
+      "loss": 0.5386,
+      "step": 982
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.51974616374752e-05,
+      "loss": 0.8077,
+      "step": 983
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.491917670648183e-05,
+      "loss": 0.5592,
+      "step": 984
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.4641437768504824e-05,
+      "loss": 0.2332,
+      "step": 985
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.436424641843475e-05,
+      "loss": 0.311,
+      "step": 986
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.408760424801792e-05,
+      "loss": 0.4153,
+      "step": 987
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.381151284584696e-05,
+      "loss": 0.3122,
+      "step": 988
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.353597379735165e-05,
+      "loss": 0.9233,
+      "step": 989
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.326098868479008e-05,
+      "loss": 0.2592,
+      "step": 990
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.29865590872394e-05,
+      "loss": 0.3064,
+      "step": 991
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.27126865805866e-05,
+      "loss": 0.2996,
+      "step": 992
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.2439372737519876e-05,
+      "loss": 0.9199,
+      "step": 993
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.2166619127519104e-05,
+      "loss": 0.2899,
+      "step": 994
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.1894427316847226e-05,
+      "loss": 0.5593,
+      "step": 995
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.1622798868541124e-05,
+      "loss": 0.3174,
+      "step": 996
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.135173534240246e-05,
+      "loss": 0.4202,
+      "step": 997
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.108123829498903e-05,
+      "loss": 0.417,
+      "step": 998
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.0811309279605675e-05,
+      "loss": 0.2984,
+      "step": 999
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.0541949846295234e-05,
+      "loss": 0.3554,
+      "step": 1000
+    },
+    {
+      "epoch": 0.74,
+      "eval_loss": 0.6549306511878967,
+      "eval_runtime": 12.9982,
+      "eval_samples_per_second": 0.846,
+      "eval_steps_per_second": 0.846,
+      "step": 1000
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1352,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "total_flos": 2.8926123468907315e+17,
+  "trial_name": null,
+  "trial_params": null
+}